Upload folder using huggingface_hub
Browse files- README.md +36 -0
- checkpoint-8670/config.json +39 -0
- checkpoint-8670/model.safetensors +3 -0
- checkpoint-8670/optimizer.pt +3 -0
- checkpoint-8670/rng_state.pth +3 -0
- checkpoint-8670/scheduler.pt +3 -0
- checkpoint-8670/trainer_state.json +2482 -0
- checkpoint-8670/training_args.bin +3 -0
- config.json +39 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- runs/Aug08_04-59-57_r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4/events.out.tfevents.1723093198.r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4.101.0 +2 -2
- runs/Aug08_04-59-57_r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4/events.out.tfevents.1723096032.r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4.101.1 +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +57 -0
- training_args.bin +3 -0
- training_params.json +30 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- text-classification
|
| 6 |
+
base_model: FacebookAI/roberta-base
|
| 7 |
+
widget:
|
| 8 |
+
- text: "I love AutoTrain"
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# Model Trained Using AutoTrain
|
| 12 |
+
|
| 13 |
+
- Problem type: Text Classification
|
| 14 |
+
|
| 15 |
+
## Validation Metrics
|
| 16 |
+
loss: 0.1770678013563156
|
| 17 |
+
|
| 18 |
+
f1_macro: 0.7558411985220287
|
| 19 |
+
|
| 20 |
+
f1_micro: 0.9476355247981546
|
| 21 |
+
|
| 22 |
+
f1_weighted: 0.9440988511640933
|
| 23 |
+
|
| 24 |
+
precision_macro: 0.8003866545209276
|
| 25 |
+
|
| 26 |
+
precision_micro: 0.9476355247981546
|
| 27 |
+
|
| 28 |
+
precision_weighted: 0.9423080358163488
|
| 29 |
+
|
| 30 |
+
recall_macro: 0.7247604343765456
|
| 31 |
+
|
| 32 |
+
recall_micro: 0.9476355247981546
|
| 33 |
+
|
| 34 |
+
recall_weighted: 0.9476355247981546
|
| 35 |
+
|
| 36 |
+
accuracy: 0.9476355247981546
|
checkpoint-8670/config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "FacebookAI/roberta-base",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"RobertaForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_probs_dropout_prob": 0.1,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"classifier_dropout": null,
|
| 10 |
+
"eos_token_id": 2,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": "1",
|
| 16 |
+
"1": "3",
|
| 17 |
+
"2": "5"
|
| 18 |
+
},
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 3072,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"1": 0,
|
| 23 |
+
"3": 1,
|
| 24 |
+
"5": 2
|
| 25 |
+
},
|
| 26 |
+
"layer_norm_eps": 1e-05,
|
| 27 |
+
"max_position_embeddings": 514,
|
| 28 |
+
"model_type": "roberta",
|
| 29 |
+
"num_attention_heads": 12,
|
| 30 |
+
"num_hidden_layers": 12,
|
| 31 |
+
"pad_token_id": 1,
|
| 32 |
+
"position_embedding_type": "absolute",
|
| 33 |
+
"problem_type": "single_label_classification",
|
| 34 |
+
"torch_dtype": "float32",
|
| 35 |
+
"transformers_version": "4.43.1",
|
| 36 |
+
"type_vocab_size": 1,
|
| 37 |
+
"use_cache": true,
|
| 38 |
+
"vocab_size": 50265
|
| 39 |
+
}
|
checkpoint-8670/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37427f4f57e436f313489f2927538add4c967871063912df29ba2fbf93a7a8aa
|
| 3 |
+
size 498615900
|
checkpoint-8670/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e33df1880a4c0c81695a23b8c8f008e80b86035574d7df5c0de6cee17b1c4fa
|
| 3 |
+
size 997351674
|
checkpoint-8670/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a9f214cc30351af425941469d729b93ec669dbdec82aaae50bebd6171965e8c
|
| 3 |
+
size 14244
|
checkpoint-8670/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb9232613ac41617352a07721dfe6f53dbb3bd9e7d08357b4fab7ed9a8b887cc
|
| 3 |
+
size 1064
|
checkpoint-8670/trainer_state.json
ADDED
|
@@ -0,0 +1,2482 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.1770678013563156,
|
| 3 |
+
"best_model_checkpoint": "autotrain-qhv86-nrgf8/checkpoint-8670",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 8670,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0028835063437139563,
|
| 13 |
+
"grad_norm": 4.133541107177734,
|
| 14 |
+
"learning_rate": 1.922337562475971e-07,
|
| 15 |
+
"loss": 1.0488,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0057670126874279125,
|
| 20 |
+
"grad_norm": 5.84380578994751,
|
| 21 |
+
"learning_rate": 3.844675124951942e-07,
|
| 22 |
+
"loss": 1.037,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00865051903114187,
|
| 27 |
+
"grad_norm": 5.987768650054932,
|
| 28 |
+
"learning_rate": 5.767012687427913e-07,
|
| 29 |
+
"loss": 1.0177,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.011534025374855825,
|
| 34 |
+
"grad_norm": 6.162261962890625,
|
| 35 |
+
"learning_rate": 7.689350249903884e-07,
|
| 36 |
+
"loss": 1.0047,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.01441753171856978,
|
| 41 |
+
"grad_norm": 5.814749240875244,
|
| 42 |
+
"learning_rate": 9.534794309880816e-07,
|
| 43 |
+
"loss": 0.952,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01730103806228374,
|
| 48 |
+
"grad_norm": 7.353080749511719,
|
| 49 |
+
"learning_rate": 1.1457131872356786e-06,
|
| 50 |
+
"loss": 0.9035,
|
| 51 |
+
"step": 150
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.020184544405997693,
|
| 55 |
+
"grad_norm": 5.703307628631592,
|
| 56 |
+
"learning_rate": 1.3379469434832756e-06,
|
| 57 |
+
"loss": 0.8376,
|
| 58 |
+
"step": 175
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02306805074971165,
|
| 62 |
+
"grad_norm": 5.434878826141357,
|
| 63 |
+
"learning_rate": 1.530180699730873e-06,
|
| 64 |
+
"loss": 0.7262,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.025951557093425604,
|
| 69 |
+
"grad_norm": 5.220206260681152,
|
| 70 |
+
"learning_rate": 1.72241445597847e-06,
|
| 71 |
+
"loss": 0.6554,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.02883506343713956,
|
| 76 |
+
"grad_norm": 12.712615013122559,
|
| 77 |
+
"learning_rate": 1.9069588619761632e-06,
|
| 78 |
+
"loss": 0.5453,
|
| 79 |
+
"step": 250
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.031718569780853516,
|
| 83 |
+
"grad_norm": 5.674898624420166,
|
| 84 |
+
"learning_rate": 2.09919261822376e-06,
|
| 85 |
+
"loss": 0.4761,
|
| 86 |
+
"step": 275
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03460207612456748,
|
| 90 |
+
"grad_norm": 8.469904899597168,
|
| 91 |
+
"learning_rate": 2.2914263744713573e-06,
|
| 92 |
+
"loss": 0.3886,
|
| 93 |
+
"step": 300
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03748558246828143,
|
| 97 |
+
"grad_norm": 5.936707973480225,
|
| 98 |
+
"learning_rate": 2.4836601307189544e-06,
|
| 99 |
+
"loss": 0.5193,
|
| 100 |
+
"step": 325
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.040369088811995385,
|
| 104 |
+
"grad_norm": 2.5938639640808105,
|
| 105 |
+
"learning_rate": 2.6682045367166475e-06,
|
| 106 |
+
"loss": 0.3869,
|
| 107 |
+
"step": 350
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.04325259515570934,
|
| 111 |
+
"grad_norm": 19.329137802124023,
|
| 112 |
+
"learning_rate": 2.860438292964245e-06,
|
| 113 |
+
"loss": 0.4093,
|
| 114 |
+
"step": 375
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0461361014994233,
|
| 118 |
+
"grad_norm": 14.663346290588379,
|
| 119 |
+
"learning_rate": 3.052672049211842e-06,
|
| 120 |
+
"loss": 0.3598,
|
| 121 |
+
"step": 400
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.049019607843137254,
|
| 125 |
+
"grad_norm": 11.453632354736328,
|
| 126 |
+
"learning_rate": 3.244905805459439e-06,
|
| 127 |
+
"loss": 0.411,
|
| 128 |
+
"step": 425
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.05190311418685121,
|
| 132 |
+
"grad_norm": 12.628795623779297,
|
| 133 |
+
"learning_rate": 3.4371395617070357e-06,
|
| 134 |
+
"loss": 0.3114,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.05478662053056517,
|
| 139 |
+
"grad_norm": 1.187443733215332,
|
| 140 |
+
"learning_rate": 3.6293733179546333e-06,
|
| 141 |
+
"loss": 0.2535,
|
| 142 |
+
"step": 475
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.05767012687427912,
|
| 146 |
+
"grad_norm": 2.074822425842285,
|
| 147 |
+
"learning_rate": 3.821607074202231e-06,
|
| 148 |
+
"loss": 0.3137,
|
| 149 |
+
"step": 500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.06055363321799308,
|
| 153 |
+
"grad_norm": 0.4427667558193207,
|
| 154 |
+
"learning_rate": 4.0138408304498275e-06,
|
| 155 |
+
"loss": 0.4233,
|
| 156 |
+
"step": 525
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.06343713956170703,
|
| 160 |
+
"grad_norm": 1.1009535789489746,
|
| 161 |
+
"learning_rate": 4.206074586697424e-06,
|
| 162 |
+
"loss": 0.2299,
|
| 163 |
+
"step": 550
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.06632064590542099,
|
| 167 |
+
"grad_norm": 0.233256533741951,
|
| 168 |
+
"learning_rate": 4.398308342945021e-06,
|
| 169 |
+
"loss": 0.2383,
|
| 170 |
+
"step": 575
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.06920415224913495,
|
| 174 |
+
"grad_norm": 11.742915153503418,
|
| 175 |
+
"learning_rate": 4.5751633986928105e-06,
|
| 176 |
+
"loss": 0.3222,
|
| 177 |
+
"step": 600
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.07208765859284891,
|
| 181 |
+
"grad_norm": 5.9581170082092285,
|
| 182 |
+
"learning_rate": 4.767397154940408e-06,
|
| 183 |
+
"loss": 0.4124,
|
| 184 |
+
"step": 625
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.07497116493656286,
|
| 188 |
+
"grad_norm": 0.9550909399986267,
|
| 189 |
+
"learning_rate": 4.959630911188005e-06,
|
| 190 |
+
"loss": 0.1909,
|
| 191 |
+
"step": 650
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.07785467128027682,
|
| 195 |
+
"grad_norm": 0.3025433123111725,
|
| 196 |
+
"learning_rate": 5.1441753171856975e-06,
|
| 197 |
+
"loss": 0.1892,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.08073817762399077,
|
| 202 |
+
"grad_norm": 68.54451751708984,
|
| 203 |
+
"learning_rate": 5.336409073433295e-06,
|
| 204 |
+
"loss": 0.3271,
|
| 205 |
+
"step": 700
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.08362168396770472,
|
| 209 |
+
"grad_norm": 1.1613526344299316,
|
| 210 |
+
"learning_rate": 5.528642829680893e-06,
|
| 211 |
+
"loss": 0.3139,
|
| 212 |
+
"step": 725
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.08650519031141868,
|
| 216 |
+
"grad_norm": 184.1776123046875,
|
| 217 |
+
"learning_rate": 5.72087658592849e-06,
|
| 218 |
+
"loss": 0.387,
|
| 219 |
+
"step": 750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.08938869665513265,
|
| 223 |
+
"grad_norm": 27.971179962158203,
|
| 224 |
+
"learning_rate": 5.913110342176087e-06,
|
| 225 |
+
"loss": 0.2067,
|
| 226 |
+
"step": 775
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.0922722029988466,
|
| 230 |
+
"grad_norm": 0.9812828302383423,
|
| 231 |
+
"learning_rate": 6.105344098423684e-06,
|
| 232 |
+
"loss": 0.4529,
|
| 233 |
+
"step": 800
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.09515570934256055,
|
| 237 |
+
"grad_norm": 45.9796028137207,
|
| 238 |
+
"learning_rate": 6.29757785467128e-06,
|
| 239 |
+
"loss": 0.2389,
|
| 240 |
+
"step": 825
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.09803921568627451,
|
| 244 |
+
"grad_norm": 1.3149181604385376,
|
| 245 |
+
"learning_rate": 6.489811610918878e-06,
|
| 246 |
+
"loss": 0.2239,
|
| 247 |
+
"step": 850
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.10092272202998846,
|
| 251 |
+
"grad_norm": 41.83066940307617,
|
| 252 |
+
"learning_rate": 6.6820453671664755e-06,
|
| 253 |
+
"loss": 0.3085,
|
| 254 |
+
"step": 875
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.10380622837370242,
|
| 258 |
+
"grad_norm": 17.775442123413086,
|
| 259 |
+
"learning_rate": 6.874279123414071e-06,
|
| 260 |
+
"loss": 0.1772,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.10668973471741638,
|
| 265 |
+
"grad_norm": 6.698100566864014,
|
| 266 |
+
"learning_rate": 7.066512879661669e-06,
|
| 267 |
+
"loss": 0.2728,
|
| 268 |
+
"step": 925
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.10957324106113034,
|
| 272 |
+
"grad_norm": 8.300532341003418,
|
| 273 |
+
"learning_rate": 7.2587466359092665e-06,
|
| 274 |
+
"loss": 0.3344,
|
| 275 |
+
"step": 950
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.11245674740484429,
|
| 279 |
+
"grad_norm": 22.10728645324707,
|
| 280 |
+
"learning_rate": 7.450980392156863e-06,
|
| 281 |
+
"loss": 0.2625,
|
| 282 |
+
"step": 975
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.11534025374855825,
|
| 286 |
+
"grad_norm": 0.9620298147201538,
|
| 287 |
+
"learning_rate": 7.643214148404462e-06,
|
| 288 |
+
"loss": 0.2913,
|
| 289 |
+
"step": 1000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.1182237600922722,
|
| 293 |
+
"grad_norm": 0.5038890242576599,
|
| 294 |
+
"learning_rate": 7.835447904652057e-06,
|
| 295 |
+
"loss": 0.2463,
|
| 296 |
+
"step": 1025
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.12110726643598616,
|
| 300 |
+
"grad_norm": 12.068671226501465,
|
| 301 |
+
"learning_rate": 8.027681660899655e-06,
|
| 302 |
+
"loss": 0.3073,
|
| 303 |
+
"step": 1050
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.12399077277970011,
|
| 307 |
+
"grad_norm": 203.9915313720703,
|
| 308 |
+
"learning_rate": 8.219915417147252e-06,
|
| 309 |
+
"loss": 0.2792,
|
| 310 |
+
"step": 1075
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.12687427912341406,
|
| 314 |
+
"grad_norm": 0.9805781841278076,
|
| 315 |
+
"learning_rate": 8.412149173394849e-06,
|
| 316 |
+
"loss": 0.277,
|
| 317 |
+
"step": 1100
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.12975778546712802,
|
| 321 |
+
"grad_norm": 63.801979064941406,
|
| 322 |
+
"learning_rate": 8.604382929642447e-06,
|
| 323 |
+
"loss": 0.3467,
|
| 324 |
+
"step": 1125
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.13264129181084197,
|
| 328 |
+
"grad_norm": 11.357747077941895,
|
| 329 |
+
"learning_rate": 8.796616685890042e-06,
|
| 330 |
+
"loss": 0.2091,
|
| 331 |
+
"step": 1150
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.13552479815455595,
|
| 335 |
+
"grad_norm": 4.0652384757995605,
|
| 336 |
+
"learning_rate": 8.98885044213764e-06,
|
| 337 |
+
"loss": 0.2469,
|
| 338 |
+
"step": 1175
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1384083044982699,
|
| 342 |
+
"grad_norm": 43.28067398071289,
|
| 343 |
+
"learning_rate": 9.181084198385237e-06,
|
| 344 |
+
"loss": 0.2779,
|
| 345 |
+
"step": 1200
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.14129181084198386,
|
| 349 |
+
"grad_norm": 17.829866409301758,
|
| 350 |
+
"learning_rate": 9.373317954632834e-06,
|
| 351 |
+
"loss": 0.263,
|
| 352 |
+
"step": 1225
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.14417531718569782,
|
| 356 |
+
"grad_norm": 0.10307206958532333,
|
| 357 |
+
"learning_rate": 9.565551710880432e-06,
|
| 358 |
+
"loss": 0.3491,
|
| 359 |
+
"step": 1250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.14705882352941177,
|
| 363 |
+
"grad_norm": 11.474695205688477,
|
| 364 |
+
"learning_rate": 9.757785467128029e-06,
|
| 365 |
+
"loss": 0.3839,
|
| 366 |
+
"step": 1275
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.14994232987312572,
|
| 370 |
+
"grad_norm": 20.019237518310547,
|
| 371 |
+
"learning_rate": 9.950019223375626e-06,
|
| 372 |
+
"loss": 0.3165,
|
| 373 |
+
"step": 1300
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.15282583621683968,
|
| 377 |
+
"grad_norm": 1.2238942384719849,
|
| 378 |
+
"learning_rate": 1.0142252979623222e-05,
|
| 379 |
+
"loss": 0.2102,
|
| 380 |
+
"step": 1325
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.15570934256055363,
|
| 384 |
+
"grad_norm": 0.36761274933815,
|
| 385 |
+
"learning_rate": 1.0334486735870819e-05,
|
| 386 |
+
"loss": 0.3574,
|
| 387 |
+
"step": 1350
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.15859284890426759,
|
| 391 |
+
"grad_norm": 20.99927520751953,
|
| 392 |
+
"learning_rate": 1.0526720492118416e-05,
|
| 393 |
+
"loss": 0.2431,
|
| 394 |
+
"step": 1375
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.16147635524798154,
|
| 398 |
+
"grad_norm": 0.13160750269889832,
|
| 399 |
+
"learning_rate": 1.0718954248366014e-05,
|
| 400 |
+
"loss": 0.1396,
|
| 401 |
+
"step": 1400
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.1643598615916955,
|
| 405 |
+
"grad_norm": 12.785110473632812,
|
| 406 |
+
"learning_rate": 1.0911188004613611e-05,
|
| 407 |
+
"loss": 0.238,
|
| 408 |
+
"step": 1425
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.16724336793540945,
|
| 412 |
+
"grad_norm": 0.129859060049057,
|
| 413 |
+
"learning_rate": 1.1103421760861208e-05,
|
| 414 |
+
"loss": 0.2905,
|
| 415 |
+
"step": 1450
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.1701268742791234,
|
| 419 |
+
"grad_norm": 1.9923458099365234,
|
| 420 |
+
"learning_rate": 1.1295655517108806e-05,
|
| 421 |
+
"loss": 0.2477,
|
| 422 |
+
"step": 1475
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.17301038062283736,
|
| 426 |
+
"grad_norm": 6.433657646179199,
|
| 427 |
+
"learning_rate": 1.1487889273356403e-05,
|
| 428 |
+
"loss": 0.2733,
|
| 429 |
+
"step": 1500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.17589388696655134,
|
| 433 |
+
"grad_norm": 9.380339622497559,
|
| 434 |
+
"learning_rate": 1.1680123029604e-05,
|
| 435 |
+
"loss": 0.1776,
|
| 436 |
+
"step": 1525
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.1787773933102653,
|
| 440 |
+
"grad_norm": 20.437368392944336,
|
| 441 |
+
"learning_rate": 1.1872356785851598e-05,
|
| 442 |
+
"loss": 0.1776,
|
| 443 |
+
"step": 1550
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.18166089965397925,
|
| 447 |
+
"grad_norm": 4.278858184814453,
|
| 448 |
+
"learning_rate": 1.2064590542099193e-05,
|
| 449 |
+
"loss": 0.2345,
|
| 450 |
+
"step": 1575
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.1845444059976932,
|
| 454 |
+
"grad_norm": 0.23265202343463898,
|
| 455 |
+
"learning_rate": 1.225682429834679e-05,
|
| 456 |
+
"loss": 0.1966,
|
| 457 |
+
"step": 1600
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.18742791234140715,
|
| 461 |
+
"grad_norm": 0.028090018779039383,
|
| 462 |
+
"learning_rate": 1.2449058054594388e-05,
|
| 463 |
+
"loss": 0.1882,
|
| 464 |
+
"step": 1625
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.1903114186851211,
|
| 468 |
+
"grad_norm": 4.165157318115234,
|
| 469 |
+
"learning_rate": 1.2641291810841985e-05,
|
| 470 |
+
"loss": 0.2044,
|
| 471 |
+
"step": 1650
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.19319492502883506,
|
| 475 |
+
"grad_norm": 7.842674255371094,
|
| 476 |
+
"learning_rate": 1.2833525567089581e-05,
|
| 477 |
+
"loss": 0.1453,
|
| 478 |
+
"step": 1675
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.19607843137254902,
|
| 482 |
+
"grad_norm": 0.10485371202230453,
|
| 483 |
+
"learning_rate": 1.302575932333718e-05,
|
| 484 |
+
"loss": 0.3435,
|
| 485 |
+
"step": 1700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.19896193771626297,
|
| 489 |
+
"grad_norm": 3.000262498855591,
|
| 490 |
+
"learning_rate": 1.3217993079584777e-05,
|
| 491 |
+
"loss": 0.2614,
|
| 492 |
+
"step": 1725
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.20184544405997693,
|
| 496 |
+
"grad_norm": 25.10281753540039,
|
| 497 |
+
"learning_rate": 1.3410226835832373e-05,
|
| 498 |
+
"loss": 0.1128,
|
| 499 |
+
"step": 1750
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.20472895040369088,
|
| 503 |
+
"grad_norm": 4.00576114654541,
|
| 504 |
+
"learning_rate": 1.3602460592079972e-05,
|
| 505 |
+
"loss": 0.3431,
|
| 506 |
+
"step": 1775
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.20761245674740483,
|
| 510 |
+
"grad_norm": 0.17649517953395844,
|
| 511 |
+
"learning_rate": 1.3794694348327567e-05,
|
| 512 |
+
"loss": 0.1869,
|
| 513 |
+
"step": 1800
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.2104959630911188,
|
| 517 |
+
"grad_norm": 6.081534385681152,
|
| 518 |
+
"learning_rate": 1.3986928104575163e-05,
|
| 519 |
+
"loss": 0.2695,
|
| 520 |
+
"step": 1825
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.21337946943483277,
|
| 524 |
+
"grad_norm": 47.501277923583984,
|
| 525 |
+
"learning_rate": 1.417916186082276e-05,
|
| 526 |
+
"loss": 0.401,
|
| 527 |
+
"step": 1850
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.21626297577854672,
|
| 531 |
+
"grad_norm": 23.75909996032715,
|
| 532 |
+
"learning_rate": 1.4371395617070359e-05,
|
| 533 |
+
"loss": 0.1943,
|
| 534 |
+
"step": 1875
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.21914648212226068,
|
| 538 |
+
"grad_norm": 5.0438642501831055,
|
| 539 |
+
"learning_rate": 1.4563629373317955e-05,
|
| 540 |
+
"loss": 0.2749,
|
| 541 |
+
"step": 1900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.22202998846597463,
|
| 545 |
+
"grad_norm": 0.12808991968631744,
|
| 546 |
+
"learning_rate": 1.4755863129565554e-05,
|
| 547 |
+
"loss": 0.1551,
|
| 548 |
+
"step": 1925
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.22491349480968859,
|
| 552 |
+
"grad_norm": 0.8075173497200012,
|
| 553 |
+
"learning_rate": 1.494809688581315e-05,
|
| 554 |
+
"loss": 0.2959,
|
| 555 |
+
"step": 1950
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.22779700115340254,
|
| 559 |
+
"grad_norm": 0.17163802683353424,
|
| 560 |
+
"learning_rate": 1.5140330642060747e-05,
|
| 561 |
+
"loss": 0.1598,
|
| 562 |
+
"step": 1975
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.2306805074971165,
|
| 566 |
+
"grad_norm": 0.15211321413516998,
|
| 567 |
+
"learning_rate": 1.5332564398308344e-05,
|
| 568 |
+
"loss": 0.3563,
|
| 569 |
+
"step": 2000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.23356401384083045,
|
| 573 |
+
"grad_norm": 0.9268957376480103,
|
| 574 |
+
"learning_rate": 1.5524798154555942e-05,
|
| 575 |
+
"loss": 0.2784,
|
| 576 |
+
"step": 2025
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.2364475201845444,
|
| 580 |
+
"grad_norm": 5.549417495727539,
|
| 581 |
+
"learning_rate": 1.5717031910803537e-05,
|
| 582 |
+
"loss": 0.2874,
|
| 583 |
+
"step": 2050
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.23933102652825836,
|
| 587 |
+
"grad_norm": 0.09005352854728699,
|
| 588 |
+
"learning_rate": 1.5909265667051136e-05,
|
| 589 |
+
"loss": 0.1934,
|
| 590 |
+
"step": 2075
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.2422145328719723,
|
| 594 |
+
"grad_norm": 31.793176651000977,
|
| 595 |
+
"learning_rate": 1.610149942329873e-05,
|
| 596 |
+
"loss": 0.3498,
|
| 597 |
+
"step": 2100
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.24509803921568626,
|
| 601 |
+
"grad_norm": 15.986053466796875,
|
| 602 |
+
"learning_rate": 1.629373317954633e-05,
|
| 603 |
+
"loss": 0.1998,
|
| 604 |
+
"step": 2125
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.24798154555940022,
|
| 608 |
+
"grad_norm": 0.5927138924598694,
|
| 609 |
+
"learning_rate": 1.6485966935793928e-05,
|
| 610 |
+
"loss": 0.3051,
|
| 611 |
+
"step": 2150
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.2508650519031142,
|
| 615 |
+
"grad_norm": 35.84630584716797,
|
| 616 |
+
"learning_rate": 1.6678200692041523e-05,
|
| 617 |
+
"loss": 0.2784,
|
| 618 |
+
"step": 2175
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.2537485582468281,
|
| 622 |
+
"grad_norm": 25.285512924194336,
|
| 623 |
+
"learning_rate": 1.687043444828912e-05,
|
| 624 |
+
"loss": 0.3806,
|
| 625 |
+
"step": 2200
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2566320645905421,
|
| 629 |
+
"grad_norm": 0.07709932327270508,
|
| 630 |
+
"learning_rate": 1.706266820453672e-05,
|
| 631 |
+
"loss": 0.3747,
|
| 632 |
+
"step": 2225
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.25951557093425603,
|
| 636 |
+
"grad_norm": 7.287086486816406,
|
| 637 |
+
"learning_rate": 1.7254901960784314e-05,
|
| 638 |
+
"loss": 0.2194,
|
| 639 |
+
"step": 2250
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.26239907727797,
|
| 643 |
+
"grad_norm": 14.211732864379883,
|
| 644 |
+
"learning_rate": 1.7447135717031913e-05,
|
| 645 |
+
"loss": 0.1749,
|
| 646 |
+
"step": 2275
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.26528258362168394,
|
| 650 |
+
"grad_norm": 0.8980517387390137,
|
| 651 |
+
"learning_rate": 1.7639369473279508e-05,
|
| 652 |
+
"loss": 0.1473,
|
| 653 |
+
"step": 2300
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.2681660899653979,
|
| 657 |
+
"grad_norm": 9.516524314880371,
|
| 658 |
+
"learning_rate": 1.7831603229527106e-05,
|
| 659 |
+
"loss": 0.2874,
|
| 660 |
+
"step": 2325
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.2710495963091119,
|
| 664 |
+
"grad_norm": 0.02932104468345642,
|
| 665 |
+
"learning_rate": 1.8023836985774705e-05,
|
| 666 |
+
"loss": 0.1767,
|
| 667 |
+
"step": 2350
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.27393310265282583,
|
| 671 |
+
"grad_norm": 9.359821319580078,
|
| 672 |
+
"learning_rate": 1.82160707420223e-05,
|
| 673 |
+
"loss": 0.3452,
|
| 674 |
+
"step": 2375
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.2768166089965398,
|
| 678 |
+
"grad_norm": 0.9531604051589966,
|
| 679 |
+
"learning_rate": 1.8408304498269898e-05,
|
| 680 |
+
"loss": 0.1393,
|
| 681 |
+
"step": 2400
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.27970011534025374,
|
| 685 |
+
"grad_norm": 22.104854583740234,
|
| 686 |
+
"learning_rate": 1.8600538254517497e-05,
|
| 687 |
+
"loss": 0.1646,
|
| 688 |
+
"step": 2425
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.2825836216839677,
|
| 692 |
+
"grad_norm": 0.1896578073501587,
|
| 693 |
+
"learning_rate": 1.879277201076509e-05,
|
| 694 |
+
"loss": 0.2958,
|
| 695 |
+
"step": 2450
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.28546712802768165,
|
| 699 |
+
"grad_norm": 0.6035724878311157,
|
| 700 |
+
"learning_rate": 1.898500576701269e-05,
|
| 701 |
+
"loss": 0.2007,
|
| 702 |
+
"step": 2475
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.28835063437139563,
|
| 706 |
+
"grad_norm": 4.53132963180542,
|
| 707 |
+
"learning_rate": 1.917723952326029e-05,
|
| 708 |
+
"loss": 0.2649,
|
| 709 |
+
"step": 2500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.29123414071510956,
|
| 713 |
+
"grad_norm": 0.08580818772315979,
|
| 714 |
+
"learning_rate": 1.9369473279507883e-05,
|
| 715 |
+
"loss": 0.2153,
|
| 716 |
+
"step": 2525
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.29411764705882354,
|
| 720 |
+
"grad_norm": 25.713481903076172,
|
| 721 |
+
"learning_rate": 1.956170703575548e-05,
|
| 722 |
+
"loss": 0.2382,
|
| 723 |
+
"step": 2550
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.29700115340253747,
|
| 727 |
+
"grad_norm": 1.4123493432998657,
|
| 728 |
+
"learning_rate": 1.9753940792003077e-05,
|
| 729 |
+
"loss": 0.2435,
|
| 730 |
+
"step": 2575
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.29988465974625145,
|
| 734 |
+
"grad_norm": 1.7834199666976929,
|
| 735 |
+
"learning_rate": 1.9946174548250675e-05,
|
| 736 |
+
"loss": 0.2232,
|
| 737 |
+
"step": 2600
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.3027681660899654,
|
| 741 |
+
"grad_norm": 0.0898483544588089,
|
| 742 |
+
"learning_rate": 1.9984621299500194e-05,
|
| 743 |
+
"loss": 0.2478,
|
| 744 |
+
"step": 2625
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.30565167243367936,
|
| 748 |
+
"grad_norm": 27.21086311340332,
|
| 749 |
+
"learning_rate": 1.996326199325046e-05,
|
| 750 |
+
"loss": 0.3023,
|
| 751 |
+
"step": 2650
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.30853517877739334,
|
| 755 |
+
"grad_norm": 130.15350341796875,
|
| 756 |
+
"learning_rate": 1.9941902687000727e-05,
|
| 757 |
+
"loss": 0.2167,
|
| 758 |
+
"step": 2675
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.31141868512110726,
|
| 762 |
+
"grad_norm": 4.993441581726074,
|
| 763 |
+
"learning_rate": 1.9920543380750994e-05,
|
| 764 |
+
"loss": 0.1558,
|
| 765 |
+
"step": 2700
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.31430219146482125,
|
| 769 |
+
"grad_norm": 6.58321475982666,
|
| 770 |
+
"learning_rate": 1.989918407450126e-05,
|
| 771 |
+
"loss": 0.2632,
|
| 772 |
+
"step": 2725
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.31718569780853517,
|
| 776 |
+
"grad_norm": 0.23823672533035278,
|
| 777 |
+
"learning_rate": 1.9877824768251528e-05,
|
| 778 |
+
"loss": 0.2257,
|
| 779 |
+
"step": 2750
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.32006920415224915,
|
| 783 |
+
"grad_norm": 4.666306495666504,
|
| 784 |
+
"learning_rate": 1.9856465462001795e-05,
|
| 785 |
+
"loss": 0.3007,
|
| 786 |
+
"step": 2775
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.3229527104959631,
|
| 790 |
+
"grad_norm": 11.329859733581543,
|
| 791 |
+
"learning_rate": 1.983510615575206e-05,
|
| 792 |
+
"loss": 0.283,
|
| 793 |
+
"step": 2800
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.32583621683967706,
|
| 797 |
+
"grad_norm": 0.03293532133102417,
|
| 798 |
+
"learning_rate": 1.981460122175232e-05,
|
| 799 |
+
"loss": 0.3227,
|
| 800 |
+
"step": 2825
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.328719723183391,
|
| 804 |
+
"grad_norm": 36.910587310791016,
|
| 805 |
+
"learning_rate": 1.9793241915502585e-05,
|
| 806 |
+
"loss": 0.3972,
|
| 807 |
+
"step": 2850
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.33160322952710497,
|
| 811 |
+
"grad_norm": 4.214820384979248,
|
| 812 |
+
"learning_rate": 1.9771882609252852e-05,
|
| 813 |
+
"loss": 0.3538,
|
| 814 |
+
"step": 2875
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.3344867358708189,
|
| 818 |
+
"grad_norm": 0.247139111161232,
|
| 819 |
+
"learning_rate": 1.975052330300312e-05,
|
| 820 |
+
"loss": 0.1739,
|
| 821 |
+
"step": 2900
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.3373702422145329,
|
| 825 |
+
"grad_norm": 0.11523030698299408,
|
| 826 |
+
"learning_rate": 1.9729163996753386e-05,
|
| 827 |
+
"loss": 0.2811,
|
| 828 |
+
"step": 2925
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.3402537485582468,
|
| 832 |
+
"grad_norm": 0.057359687983989716,
|
| 833 |
+
"learning_rate": 1.9707804690503653e-05,
|
| 834 |
+
"loss": 0.2256,
|
| 835 |
+
"step": 2950
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.3431372549019608,
|
| 839 |
+
"grad_norm": 15.97446346282959,
|
| 840 |
+
"learning_rate": 1.968644538425392e-05,
|
| 841 |
+
"loss": 0.3174,
|
| 842 |
+
"step": 2975
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.3460207612456747,
|
| 846 |
+
"grad_norm": 13.150789260864258,
|
| 847 |
+
"learning_rate": 1.9665086078004187e-05,
|
| 848 |
+
"loss": 0.203,
|
| 849 |
+
"step": 3000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.3489042675893887,
|
| 853 |
+
"grad_norm": 13.900849342346191,
|
| 854 |
+
"learning_rate": 1.9643726771754453e-05,
|
| 855 |
+
"loss": 0.2407,
|
| 856 |
+
"step": 3025
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.3517877739331027,
|
| 860 |
+
"grad_norm": 10.278796195983887,
|
| 861 |
+
"learning_rate": 1.962236746550472e-05,
|
| 862 |
+
"loss": 0.365,
|
| 863 |
+
"step": 3050
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.3546712802768166,
|
| 867 |
+
"grad_norm": 13.592562675476074,
|
| 868 |
+
"learning_rate": 1.9601008159254987e-05,
|
| 869 |
+
"loss": 0.2901,
|
| 870 |
+
"step": 3075
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.3575547866205306,
|
| 874 |
+
"grad_norm": 11.419650077819824,
|
| 875 |
+
"learning_rate": 1.9579648853005254e-05,
|
| 876 |
+
"loss": 0.1789,
|
| 877 |
+
"step": 3100
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.3604382929642445,
|
| 881 |
+
"grad_norm": 0.6274221539497375,
|
| 882 |
+
"learning_rate": 1.955828954675552e-05,
|
| 883 |
+
"loss": 0.3775,
|
| 884 |
+
"step": 3125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.3633217993079585,
|
| 888 |
+
"grad_norm": 5.571254253387451,
|
| 889 |
+
"learning_rate": 1.9536930240505788e-05,
|
| 890 |
+
"loss": 0.2945,
|
| 891 |
+
"step": 3150
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.3662053056516724,
|
| 895 |
+
"grad_norm": 73.99065399169922,
|
| 896 |
+
"learning_rate": 1.9515570934256058e-05,
|
| 897 |
+
"loss": 0.2594,
|
| 898 |
+
"step": 3175
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.3690888119953864,
|
| 902 |
+
"grad_norm": 10.681817054748535,
|
| 903 |
+
"learning_rate": 1.9494211628006325e-05,
|
| 904 |
+
"loss": 0.2704,
|
| 905 |
+
"step": 3200
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.3719723183391003,
|
| 909 |
+
"grad_norm": 9.517922401428223,
|
| 910 |
+
"learning_rate": 1.947285232175659e-05,
|
| 911 |
+
"loss": 0.2103,
|
| 912 |
+
"step": 3225
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.3748558246828143,
|
| 916 |
+
"grad_norm": 9.301681518554688,
|
| 917 |
+
"learning_rate": 1.945149301550686e-05,
|
| 918 |
+
"loss": 0.3217,
|
| 919 |
+
"step": 3250
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.37773933102652824,
|
| 923 |
+
"grad_norm": 0.2770113945007324,
|
| 924 |
+
"learning_rate": 1.9430133709257125e-05,
|
| 925 |
+
"loss": 0.2023,
|
| 926 |
+
"step": 3275
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.3806228373702422,
|
| 930 |
+
"grad_norm": 0.03743845969438553,
|
| 931 |
+
"learning_rate": 1.9408774403007392e-05,
|
| 932 |
+
"loss": 0.1819,
|
| 933 |
+
"step": 3300
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.38350634371395614,
|
| 937 |
+
"grad_norm": 0.24937543272972107,
|
| 938 |
+
"learning_rate": 1.938741509675766e-05,
|
| 939 |
+
"loss": 0.2014,
|
| 940 |
+
"step": 3325
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.3863898500576701,
|
| 944 |
+
"grad_norm": 19.140220642089844,
|
| 945 |
+
"learning_rate": 1.9366055790507926e-05,
|
| 946 |
+
"loss": 0.1368,
|
| 947 |
+
"step": 3350
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.3892733564013841,
|
| 951 |
+
"grad_norm": 11.055630683898926,
|
| 952 |
+
"learning_rate": 1.9344696484258193e-05,
|
| 953 |
+
"loss": 0.2868,
|
| 954 |
+
"step": 3375
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.39215686274509803,
|
| 958 |
+
"grad_norm": 2.5183160305023193,
|
| 959 |
+
"learning_rate": 1.932333717800846e-05,
|
| 960 |
+
"loss": 0.1538,
|
| 961 |
+
"step": 3400
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.395040369088812,
|
| 965 |
+
"grad_norm": 26.567752838134766,
|
| 966 |
+
"learning_rate": 1.9301977871758726e-05,
|
| 967 |
+
"loss": 0.3423,
|
| 968 |
+
"step": 3425
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.39792387543252594,
|
| 972 |
+
"grad_norm": 0.10602366179227829,
|
| 973 |
+
"learning_rate": 1.9280618565508993e-05,
|
| 974 |
+
"loss": 0.2114,
|
| 975 |
+
"step": 3450
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.4008073817762399,
|
| 979 |
+
"grad_norm": 12.951467514038086,
|
| 980 |
+
"learning_rate": 1.925925925925926e-05,
|
| 981 |
+
"loss": 0.2475,
|
| 982 |
+
"step": 3475
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.40369088811995385,
|
| 986 |
+
"grad_norm": 13.720992088317871,
|
| 987 |
+
"learning_rate": 1.9237899953009527e-05,
|
| 988 |
+
"loss": 0.3702,
|
| 989 |
+
"step": 3500
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.40657439446366783,
|
| 993 |
+
"grad_norm": 8.615378379821777,
|
| 994 |
+
"learning_rate": 1.9216540646759797e-05,
|
| 995 |
+
"loss": 0.2491,
|
| 996 |
+
"step": 3525
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.40945790080738176,
|
| 1000 |
+
"grad_norm": 1.0278849601745605,
|
| 1001 |
+
"learning_rate": 1.9195181340510064e-05,
|
| 1002 |
+
"loss": 0.2841,
|
| 1003 |
+
"step": 3550
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.41234140715109574,
|
| 1007 |
+
"grad_norm": 2.592742681503296,
|
| 1008 |
+
"learning_rate": 1.917382203426033e-05,
|
| 1009 |
+
"loss": 0.2825,
|
| 1010 |
+
"step": 3575
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.41522491349480967,
|
| 1014 |
+
"grad_norm": 17.225446701049805,
|
| 1015 |
+
"learning_rate": 1.9152462728010598e-05,
|
| 1016 |
+
"loss": 0.2233,
|
| 1017 |
+
"step": 3600
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.41810841983852365,
|
| 1021 |
+
"grad_norm": 30.14980125427246,
|
| 1022 |
+
"learning_rate": 1.9131103421760864e-05,
|
| 1023 |
+
"loss": 0.2972,
|
| 1024 |
+
"step": 3625
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.4209919261822376,
|
| 1028 |
+
"grad_norm": 3.5756497383117676,
|
| 1029 |
+
"learning_rate": 1.910974411551113e-05,
|
| 1030 |
+
"loss": 0.2904,
|
| 1031 |
+
"step": 3650
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.42387543252595156,
|
| 1035 |
+
"grad_norm": 13.067952156066895,
|
| 1036 |
+
"learning_rate": 1.9088384809261398e-05,
|
| 1037 |
+
"loss": 0.3018,
|
| 1038 |
+
"step": 3675
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.42675893886966554,
|
| 1042 |
+
"grad_norm": 20.728599548339844,
|
| 1043 |
+
"learning_rate": 1.9067025503011665e-05,
|
| 1044 |
+
"loss": 0.2865,
|
| 1045 |
+
"step": 3700
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.42964244521337946,
|
| 1049 |
+
"grad_norm": 197.9759063720703,
|
| 1050 |
+
"learning_rate": 1.9045666196761932e-05,
|
| 1051 |
+
"loss": 0.2326,
|
| 1052 |
+
"step": 3725
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.43252595155709345,
|
| 1056 |
+
"grad_norm": 0.11359433829784393,
|
| 1057 |
+
"learning_rate": 1.90243068905122e-05,
|
| 1058 |
+
"loss": 0.2059,
|
| 1059 |
+
"step": 3750
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.4354094579008074,
|
| 1063 |
+
"grad_norm": 36.202911376953125,
|
| 1064 |
+
"learning_rate": 1.9002947584262465e-05,
|
| 1065 |
+
"loss": 0.2523,
|
| 1066 |
+
"step": 3775
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.43829296424452135,
|
| 1070 |
+
"grad_norm": 0.4303267300128937,
|
| 1071 |
+
"learning_rate": 1.8981588278012732e-05,
|
| 1072 |
+
"loss": 0.1389,
|
| 1073 |
+
"step": 3800
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.4411764705882353,
|
| 1077 |
+
"grad_norm": 36.952667236328125,
|
| 1078 |
+
"learning_rate": 1.8960228971763e-05,
|
| 1079 |
+
"loss": 0.3338,
|
| 1080 |
+
"step": 3825
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.44405997693194926,
|
| 1084 |
+
"grad_norm": 0.025168880820274353,
|
| 1085 |
+
"learning_rate": 1.8938869665513266e-05,
|
| 1086 |
+
"loss": 0.198,
|
| 1087 |
+
"step": 3850
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.4469434832756632,
|
| 1091 |
+
"grad_norm": 0.18423987925052643,
|
| 1092 |
+
"learning_rate": 1.8917510359263533e-05,
|
| 1093 |
+
"loss": 0.2642,
|
| 1094 |
+
"step": 3875
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.44982698961937717,
|
| 1098 |
+
"grad_norm": 0.37645941972732544,
|
| 1099 |
+
"learning_rate": 1.88961510530138e-05,
|
| 1100 |
+
"loss": 0.2285,
|
| 1101 |
+
"step": 3900
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.4527104959630911,
|
| 1105 |
+
"grad_norm": 0.11676705628633499,
|
| 1106 |
+
"learning_rate": 1.8874791746764067e-05,
|
| 1107 |
+
"loss": 0.1215,
|
| 1108 |
+
"step": 3925
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.4555940023068051,
|
| 1112 |
+
"grad_norm": 8.879549980163574,
|
| 1113 |
+
"learning_rate": 1.8853432440514333e-05,
|
| 1114 |
+
"loss": 0.4279,
|
| 1115 |
+
"step": 3950
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.458477508650519,
|
| 1119 |
+
"grad_norm": 0.5608970522880554,
|
| 1120 |
+
"learning_rate": 1.88320731342646e-05,
|
| 1121 |
+
"loss": 0.1822,
|
| 1122 |
+
"step": 3975
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.461361014994233,
|
| 1126 |
+
"grad_norm": 10.533734321594238,
|
| 1127 |
+
"learning_rate": 1.8810713828014867e-05,
|
| 1128 |
+
"loss": 0.2495,
|
| 1129 |
+
"step": 4000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.46424452133794697,
|
| 1133 |
+
"grad_norm": 0.26254749298095703,
|
| 1134 |
+
"learning_rate": 1.8789354521765134e-05,
|
| 1135 |
+
"loss": 0.2471,
|
| 1136 |
+
"step": 4025
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.4671280276816609,
|
| 1140 |
+
"grad_norm": 37.459617614746094,
|
| 1141 |
+
"learning_rate": 1.87679952155154e-05,
|
| 1142 |
+
"loss": 0.2435,
|
| 1143 |
+
"step": 4050
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.4700115340253749,
|
| 1147 |
+
"grad_norm": 16.732284545898438,
|
| 1148 |
+
"learning_rate": 1.8746635909265668e-05,
|
| 1149 |
+
"loss": 0.2939,
|
| 1150 |
+
"step": 4075
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.4728950403690888,
|
| 1154 |
+
"grad_norm": 2.694005012512207,
|
| 1155 |
+
"learning_rate": 1.8725276603015934e-05,
|
| 1156 |
+
"loss": 0.284,
|
| 1157 |
+
"step": 4100
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.4757785467128028,
|
| 1161 |
+
"grad_norm": 9.272911071777344,
|
| 1162 |
+
"learning_rate": 1.8703917296766205e-05,
|
| 1163 |
+
"loss": 0.1847,
|
| 1164 |
+
"step": 4125
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.4786620530565167,
|
| 1168 |
+
"grad_norm": 63.758750915527344,
|
| 1169 |
+
"learning_rate": 1.868255799051647e-05,
|
| 1170 |
+
"loss": 0.1465,
|
| 1171 |
+
"step": 4150
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4815455594002307,
|
| 1175 |
+
"grad_norm": 1.7441495656967163,
|
| 1176 |
+
"learning_rate": 1.866119868426674e-05,
|
| 1177 |
+
"loss": 0.3875,
|
| 1178 |
+
"step": 4175
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.4844290657439446,
|
| 1182 |
+
"grad_norm": 0.08567396551370621,
|
| 1183 |
+
"learning_rate": 1.8639839378017005e-05,
|
| 1184 |
+
"loss": 0.1664,
|
| 1185 |
+
"step": 4200
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.4873125720876586,
|
| 1189 |
+
"grad_norm": 0.11183150857686996,
|
| 1190 |
+
"learning_rate": 1.8618480071767272e-05,
|
| 1191 |
+
"loss": 0.3229,
|
| 1192 |
+
"step": 4225
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.49019607843137253,
|
| 1196 |
+
"grad_norm": 62.44700622558594,
|
| 1197 |
+
"learning_rate": 1.859712076551754e-05,
|
| 1198 |
+
"loss": 0.197,
|
| 1199 |
+
"step": 4250
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.4930795847750865,
|
| 1203 |
+
"grad_norm": 36.57319259643555,
|
| 1204 |
+
"learning_rate": 1.8575761459267806e-05,
|
| 1205 |
+
"loss": 0.1885,
|
| 1206 |
+
"step": 4275
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.49596309111880044,
|
| 1210 |
+
"grad_norm": 105.54853057861328,
|
| 1211 |
+
"learning_rate": 1.8554402153018073e-05,
|
| 1212 |
+
"loss": 0.1823,
|
| 1213 |
+
"step": 4300
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.4988465974625144,
|
| 1217 |
+
"grad_norm": 0.05175294354557991,
|
| 1218 |
+
"learning_rate": 1.853304284676834e-05,
|
| 1219 |
+
"loss": 0.2568,
|
| 1220 |
+
"step": 4325
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.5017301038062284,
|
| 1224 |
+
"grad_norm": 7.137915134429932,
|
| 1225 |
+
"learning_rate": 1.8511683540518606e-05,
|
| 1226 |
+
"loss": 0.1355,
|
| 1227 |
+
"step": 4350
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.5046136101499423,
|
| 1231 |
+
"grad_norm": 0.1737750619649887,
|
| 1232 |
+
"learning_rate": 1.8490324234268873e-05,
|
| 1233 |
+
"loss": 0.2962,
|
| 1234 |
+
"step": 4375
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.5074971164936563,
|
| 1238 |
+
"grad_norm": 1.8034484386444092,
|
| 1239 |
+
"learning_rate": 1.846896492801914e-05,
|
| 1240 |
+
"loss": 0.3601,
|
| 1241 |
+
"step": 4400
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.5103806228373703,
|
| 1245 |
+
"grad_norm": 1.1106842756271362,
|
| 1246 |
+
"learning_rate": 1.8447605621769407e-05,
|
| 1247 |
+
"loss": 0.1803,
|
| 1248 |
+
"step": 4425
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.5132641291810842,
|
| 1252 |
+
"grad_norm": 4.245251178741455,
|
| 1253 |
+
"learning_rate": 1.8426246315519674e-05,
|
| 1254 |
+
"loss": 0.2903,
|
| 1255 |
+
"step": 4450
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.5161476355247981,
|
| 1259 |
+
"grad_norm": 19.86756706237793,
|
| 1260 |
+
"learning_rate": 1.840488700926994e-05,
|
| 1261 |
+
"loss": 0.3144,
|
| 1262 |
+
"step": 4475
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.5190311418685121,
|
| 1266 |
+
"grad_norm": 10.5123291015625,
|
| 1267 |
+
"learning_rate": 1.8383527703020207e-05,
|
| 1268 |
+
"loss": 0.3279,
|
| 1269 |
+
"step": 4500
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.5219146482122261,
|
| 1273 |
+
"grad_norm": 24.970521926879883,
|
| 1274 |
+
"learning_rate": 1.8362168396770474e-05,
|
| 1275 |
+
"loss": 0.1741,
|
| 1276 |
+
"step": 4525
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.52479815455594,
|
| 1280 |
+
"grad_norm": 0.4247228503227234,
|
| 1281 |
+
"learning_rate": 1.834080909052074e-05,
|
| 1282 |
+
"loss": 0.1565,
|
| 1283 |
+
"step": 4550
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.527681660899654,
|
| 1287 |
+
"grad_norm": 0.09796195477247238,
|
| 1288 |
+
"learning_rate": 1.8319449784271008e-05,
|
| 1289 |
+
"loss": 0.2446,
|
| 1290 |
+
"step": 4575
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.5305651672433679,
|
| 1294 |
+
"grad_norm": 0.6912326216697693,
|
| 1295 |
+
"learning_rate": 1.8298090478021275e-05,
|
| 1296 |
+
"loss": 0.2786,
|
| 1297 |
+
"step": 4600
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.5334486735870819,
|
| 1301 |
+
"grad_norm": 0.15354254841804504,
|
| 1302 |
+
"learning_rate": 1.827673117177154e-05,
|
| 1303 |
+
"loss": 0.2418,
|
| 1304 |
+
"step": 4625
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.5363321799307958,
|
| 1308 |
+
"grad_norm": 43.35624313354492,
|
| 1309 |
+
"learning_rate": 1.825537186552181e-05,
|
| 1310 |
+
"loss": 0.3739,
|
| 1311 |
+
"step": 4650
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.5392156862745098,
|
| 1315 |
+
"grad_norm": 3.5612680912017822,
|
| 1316 |
+
"learning_rate": 1.8234012559272075e-05,
|
| 1317 |
+
"loss": 0.2555,
|
| 1318 |
+
"step": 4675
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.5420991926182238,
|
| 1322 |
+
"grad_norm": 41.634315490722656,
|
| 1323 |
+
"learning_rate": 1.8212653253022342e-05,
|
| 1324 |
+
"loss": 0.2539,
|
| 1325 |
+
"step": 4700
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.5449826989619377,
|
| 1329 |
+
"grad_norm": 77.00975036621094,
|
| 1330 |
+
"learning_rate": 1.819129394677261e-05,
|
| 1331 |
+
"loss": 0.2101,
|
| 1332 |
+
"step": 4725
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.5478662053056517,
|
| 1336 |
+
"grad_norm": 94.3470458984375,
|
| 1337 |
+
"learning_rate": 1.8169934640522876e-05,
|
| 1338 |
+
"loss": 0.236,
|
| 1339 |
+
"step": 4750
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.5507497116493656,
|
| 1343 |
+
"grad_norm": 3.7230870723724365,
|
| 1344 |
+
"learning_rate": 1.8148575334273143e-05,
|
| 1345 |
+
"loss": 0.2221,
|
| 1346 |
+
"step": 4775
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.5536332179930796,
|
| 1350 |
+
"grad_norm": 28.07904624938965,
|
| 1351 |
+
"learning_rate": 1.812721602802341e-05,
|
| 1352 |
+
"loss": 0.2298,
|
| 1353 |
+
"step": 4800
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.5565167243367936,
|
| 1357 |
+
"grad_norm": 38.550270080566406,
|
| 1358 |
+
"learning_rate": 1.810585672177368e-05,
|
| 1359 |
+
"loss": 0.2298,
|
| 1360 |
+
"step": 4825
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.5594002306805075,
|
| 1364 |
+
"grad_norm": 0.02074141800403595,
|
| 1365 |
+
"learning_rate": 1.8084497415523946e-05,
|
| 1366 |
+
"loss": 0.2115,
|
| 1367 |
+
"step": 4850
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.5622837370242214,
|
| 1371 |
+
"grad_norm": 3.2749974727630615,
|
| 1372 |
+
"learning_rate": 1.8063138109274213e-05,
|
| 1373 |
+
"loss": 0.2015,
|
| 1374 |
+
"step": 4875
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.5651672433679354,
|
| 1378 |
+
"grad_norm": 10.315749168395996,
|
| 1379 |
+
"learning_rate": 1.804177880302448e-05,
|
| 1380 |
+
"loss": 0.2693,
|
| 1381 |
+
"step": 4900
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.5680507497116494,
|
| 1385 |
+
"grad_norm": 10.362585067749023,
|
| 1386 |
+
"learning_rate": 1.8020419496774747e-05,
|
| 1387 |
+
"loss": 0.1759,
|
| 1388 |
+
"step": 4925
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.5709342560553633,
|
| 1392 |
+
"grad_norm": 117.65594482421875,
|
| 1393 |
+
"learning_rate": 1.7999060190525014e-05,
|
| 1394 |
+
"loss": 0.2608,
|
| 1395 |
+
"step": 4950
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.5738177623990772,
|
| 1399 |
+
"grad_norm": 0.16872438788414001,
|
| 1400 |
+
"learning_rate": 1.797770088427528e-05,
|
| 1401 |
+
"loss": 0.2088,
|
| 1402 |
+
"step": 4975
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.5767012687427913,
|
| 1406 |
+
"grad_norm": 19.122615814208984,
|
| 1407 |
+
"learning_rate": 1.7956341578025548e-05,
|
| 1408 |
+
"loss": 0.327,
|
| 1409 |
+
"step": 5000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.5795847750865052,
|
| 1413 |
+
"grad_norm": 0.4338647425174713,
|
| 1414 |
+
"learning_rate": 1.7934982271775814e-05,
|
| 1415 |
+
"loss": 0.2677,
|
| 1416 |
+
"step": 5025
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.5824682814302191,
|
| 1420 |
+
"grad_norm": 0.321367084980011,
|
| 1421 |
+
"learning_rate": 1.791362296552608e-05,
|
| 1422 |
+
"loss": 0.1996,
|
| 1423 |
+
"step": 5050
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.5853517877739332,
|
| 1427 |
+
"grad_norm": 0.06689249724149704,
|
| 1428 |
+
"learning_rate": 1.7892263659276348e-05,
|
| 1429 |
+
"loss": 0.1648,
|
| 1430 |
+
"step": 5075
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.5882352941176471,
|
| 1434 |
+
"grad_norm": 0.08134772628545761,
|
| 1435 |
+
"learning_rate": 1.7870904353026615e-05,
|
| 1436 |
+
"loss": 0.1596,
|
| 1437 |
+
"step": 5100
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.591118800461361,
|
| 1441 |
+
"grad_norm": 0.6613795757293701,
|
| 1442 |
+
"learning_rate": 1.7849545046776882e-05,
|
| 1443 |
+
"loss": 0.3882,
|
| 1444 |
+
"step": 5125
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.5940023068050749,
|
| 1448 |
+
"grad_norm": 9.452569961547852,
|
| 1449 |
+
"learning_rate": 1.782818574052715e-05,
|
| 1450 |
+
"loss": 0.2335,
|
| 1451 |
+
"step": 5150
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.596885813148789,
|
| 1455 |
+
"grad_norm": 3.534229040145874,
|
| 1456 |
+
"learning_rate": 1.7806826434277415e-05,
|
| 1457 |
+
"loss": 0.2462,
|
| 1458 |
+
"step": 5175
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.5997693194925029,
|
| 1462 |
+
"grad_norm": 0.705708920955658,
|
| 1463 |
+
"learning_rate": 1.7785467128027682e-05,
|
| 1464 |
+
"loss": 0.3247,
|
| 1465 |
+
"step": 5200
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.6026528258362168,
|
| 1469 |
+
"grad_norm": 16.955322265625,
|
| 1470 |
+
"learning_rate": 1.776410782177795e-05,
|
| 1471 |
+
"loss": 0.3191,
|
| 1472 |
+
"step": 5225
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.6055363321799307,
|
| 1476 |
+
"grad_norm": 44.89308166503906,
|
| 1477 |
+
"learning_rate": 1.7742748515528216e-05,
|
| 1478 |
+
"loss": 0.2556,
|
| 1479 |
+
"step": 5250
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.6084198385236448,
|
| 1483 |
+
"grad_norm": 0.16740162670612335,
|
| 1484 |
+
"learning_rate": 1.7721389209278483e-05,
|
| 1485 |
+
"loss": 0.2197,
|
| 1486 |
+
"step": 5275
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.6113033448673587,
|
| 1490 |
+
"grad_norm": 3.059319019317627,
|
| 1491 |
+
"learning_rate": 1.770002990302875e-05,
|
| 1492 |
+
"loss": 0.2797,
|
| 1493 |
+
"step": 5300
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.6141868512110726,
|
| 1497 |
+
"grad_norm": 0.18199850618839264,
|
| 1498 |
+
"learning_rate": 1.7678670596779017e-05,
|
| 1499 |
+
"loss": 0.2133,
|
| 1500 |
+
"step": 5325
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.6170703575547867,
|
| 1504 |
+
"grad_norm": 0.4918987452983856,
|
| 1505 |
+
"learning_rate": 1.7657311290529283e-05,
|
| 1506 |
+
"loss": 0.2481,
|
| 1507 |
+
"step": 5350
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.6199538638985006,
|
| 1511 |
+
"grad_norm": 98.67976379394531,
|
| 1512 |
+
"learning_rate": 1.763595198427955e-05,
|
| 1513 |
+
"loss": 0.2215,
|
| 1514 |
+
"step": 5375
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.6228373702422145,
|
| 1518 |
+
"grad_norm": 0.5225765109062195,
|
| 1519 |
+
"learning_rate": 1.7614592678029817e-05,
|
| 1520 |
+
"loss": 0.1051,
|
| 1521 |
+
"step": 5400
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.6257208765859285,
|
| 1525 |
+
"grad_norm": 0.03558097034692764,
|
| 1526 |
+
"learning_rate": 1.7593233371780084e-05,
|
| 1527 |
+
"loss": 0.2843,
|
| 1528 |
+
"step": 5425
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.6286043829296425,
|
| 1532 |
+
"grad_norm": 123.46427917480469,
|
| 1533 |
+
"learning_rate": 1.757187406553035e-05,
|
| 1534 |
+
"loss": 0.1533,
|
| 1535 |
+
"step": 5450
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.6314878892733564,
|
| 1539 |
+
"grad_norm": 0.40241536498069763,
|
| 1540 |
+
"learning_rate": 1.7550514759280618e-05,
|
| 1541 |
+
"loss": 0.1457,
|
| 1542 |
+
"step": 5475
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.6343713956170703,
|
| 1546 |
+
"grad_norm": 2.7131290435791016,
|
| 1547 |
+
"learning_rate": 1.7529155453030884e-05,
|
| 1548 |
+
"loss": 0.2867,
|
| 1549 |
+
"step": 5500
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.6372549019607843,
|
| 1553 |
+
"grad_norm": 0.1088755801320076,
|
| 1554 |
+
"learning_rate": 1.7507796146781155e-05,
|
| 1555 |
+
"loss": 0.1878,
|
| 1556 |
+
"step": 5525
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.6401384083044983,
|
| 1560 |
+
"grad_norm": 13.675771713256836,
|
| 1561 |
+
"learning_rate": 1.748643684053142e-05,
|
| 1562 |
+
"loss": 0.2924,
|
| 1563 |
+
"step": 5550
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.6430219146482122,
|
| 1567 |
+
"grad_norm": 7.713492393493652,
|
| 1568 |
+
"learning_rate": 1.746507753428169e-05,
|
| 1569 |
+
"loss": 0.1444,
|
| 1570 |
+
"step": 5575
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.6459054209919262,
|
| 1574 |
+
"grad_norm": 74.01490020751953,
|
| 1575 |
+
"learning_rate": 1.7443718228031955e-05,
|
| 1576 |
+
"loss": 0.3885,
|
| 1577 |
+
"step": 5600
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.6487889273356401,
|
| 1581 |
+
"grad_norm": 0.263332337141037,
|
| 1582 |
+
"learning_rate": 1.7422358921782222e-05,
|
| 1583 |
+
"loss": 0.2199,
|
| 1584 |
+
"step": 5625
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.6516724336793541,
|
| 1588 |
+
"grad_norm": 10.242852210998535,
|
| 1589 |
+
"learning_rate": 1.740099961553249e-05,
|
| 1590 |
+
"loss": 0.1626,
|
| 1591 |
+
"step": 5650
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.654555940023068,
|
| 1595 |
+
"grad_norm": 0.14904502034187317,
|
| 1596 |
+
"learning_rate": 1.7379640309282756e-05,
|
| 1597 |
+
"loss": 0.1444,
|
| 1598 |
+
"step": 5675
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.657439446366782,
|
| 1602 |
+
"grad_norm": 0.26692208647727966,
|
| 1603 |
+
"learning_rate": 1.7358281003033023e-05,
|
| 1604 |
+
"loss": 0.3067,
|
| 1605 |
+
"step": 5700
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.660322952710496,
|
| 1609 |
+
"grad_norm": 5.127139091491699,
|
| 1610 |
+
"learning_rate": 1.733692169678329e-05,
|
| 1611 |
+
"loss": 0.0814,
|
| 1612 |
+
"step": 5725
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.6632064590542099,
|
| 1616 |
+
"grad_norm": 3.4886350631713867,
|
| 1617 |
+
"learning_rate": 1.7315562390533556e-05,
|
| 1618 |
+
"loss": 0.3612,
|
| 1619 |
+
"step": 5750
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.6660899653979239,
|
| 1623 |
+
"grad_norm": 0.24010740220546722,
|
| 1624 |
+
"learning_rate": 1.7294203084283826e-05,
|
| 1625 |
+
"loss": 0.1789,
|
| 1626 |
+
"step": 5775
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.6689734717416378,
|
| 1630 |
+
"grad_norm": 0.20936521887779236,
|
| 1631 |
+
"learning_rate": 1.7272843778034093e-05,
|
| 1632 |
+
"loss": 0.2636,
|
| 1633 |
+
"step": 5800
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.6718569780853518,
|
| 1637 |
+
"grad_norm": 14.584404945373535,
|
| 1638 |
+
"learning_rate": 1.725148447178436e-05,
|
| 1639 |
+
"loss": 0.2459,
|
| 1640 |
+
"step": 5825
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.6747404844290658,
|
| 1644 |
+
"grad_norm": 0.12056291848421097,
|
| 1645 |
+
"learning_rate": 1.7230125165534627e-05,
|
| 1646 |
+
"loss": 0.2672,
|
| 1647 |
+
"step": 5850
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.6776239907727797,
|
| 1651 |
+
"grad_norm": 0.03764568269252777,
|
| 1652 |
+
"learning_rate": 1.7208765859284894e-05,
|
| 1653 |
+
"loss": 0.3007,
|
| 1654 |
+
"step": 5875
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.6805074971164936,
|
| 1658 |
+
"grad_norm": 0.06103919446468353,
|
| 1659 |
+
"learning_rate": 1.718740655303516e-05,
|
| 1660 |
+
"loss": 0.1368,
|
| 1661 |
+
"step": 5900
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.6833910034602076,
|
| 1665 |
+
"grad_norm": 145.5305633544922,
|
| 1666 |
+
"learning_rate": 1.7166047246785428e-05,
|
| 1667 |
+
"loss": 0.2379,
|
| 1668 |
+
"step": 5925
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.6862745098039216,
|
| 1672 |
+
"grad_norm": 2.428950786590576,
|
| 1673 |
+
"learning_rate": 1.7144687940535694e-05,
|
| 1674 |
+
"loss": 0.1346,
|
| 1675 |
+
"step": 5950
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.6891580161476355,
|
| 1679 |
+
"grad_norm": 0.829033613204956,
|
| 1680 |
+
"learning_rate": 1.712332863428596e-05,
|
| 1681 |
+
"loss": 0.2341,
|
| 1682 |
+
"step": 5975
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.6920415224913494,
|
| 1686 |
+
"grad_norm": 21.751039505004883,
|
| 1687 |
+
"learning_rate": 1.7101969328036228e-05,
|
| 1688 |
+
"loss": 0.1783,
|
| 1689 |
+
"step": 6000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.6949250288350635,
|
| 1693 |
+
"grad_norm": 0.04146712273359299,
|
| 1694 |
+
"learning_rate": 1.7080610021786495e-05,
|
| 1695 |
+
"loss": 0.2396,
|
| 1696 |
+
"step": 6025
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.6978085351787774,
|
| 1700 |
+
"grad_norm": 0.16122393310070038,
|
| 1701 |
+
"learning_rate": 1.7059250715536762e-05,
|
| 1702 |
+
"loss": 0.1342,
|
| 1703 |
+
"step": 6050
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.7006920415224913,
|
| 1707 |
+
"grad_norm": 7.1467695236206055,
|
| 1708 |
+
"learning_rate": 1.703789140928703e-05,
|
| 1709 |
+
"loss": 0.2858,
|
| 1710 |
+
"step": 6075
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.7035755478662054,
|
| 1714 |
+
"grad_norm": 6.679579734802246,
|
| 1715 |
+
"learning_rate": 1.7016532103037295e-05,
|
| 1716 |
+
"loss": 0.1869,
|
| 1717 |
+
"step": 6100
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.7064590542099193,
|
| 1721 |
+
"grad_norm": 1.2571030855178833,
|
| 1722 |
+
"learning_rate": 1.6995172796787562e-05,
|
| 1723 |
+
"loss": 0.2108,
|
| 1724 |
+
"step": 6125
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.7093425605536332,
|
| 1728 |
+
"grad_norm": 208.1612548828125,
|
| 1729 |
+
"learning_rate": 1.697381349053783e-05,
|
| 1730 |
+
"loss": 0.1907,
|
| 1731 |
+
"step": 6150
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.7122260668973471,
|
| 1735 |
+
"grad_norm": 0.8248432874679565,
|
| 1736 |
+
"learning_rate": 1.6952454184288096e-05,
|
| 1737 |
+
"loss": 0.2966,
|
| 1738 |
+
"step": 6175
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.7151095732410612,
|
| 1742 |
+
"grad_norm": 0.13479585945606232,
|
| 1743 |
+
"learning_rate": 1.6931094878038363e-05,
|
| 1744 |
+
"loss": 0.2029,
|
| 1745 |
+
"step": 6200
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.7179930795847751,
|
| 1749 |
+
"grad_norm": 0.06604406982660294,
|
| 1750 |
+
"learning_rate": 1.690973557178863e-05,
|
| 1751 |
+
"loss": 0.1891,
|
| 1752 |
+
"step": 6225
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.720876585928489,
|
| 1756 |
+
"grad_norm": 0.03318684920668602,
|
| 1757 |
+
"learning_rate": 1.6888376265538896e-05,
|
| 1758 |
+
"loss": 0.1293,
|
| 1759 |
+
"step": 6250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.723760092272203,
|
| 1763 |
+
"grad_norm": 0.02965410053730011,
|
| 1764 |
+
"learning_rate": 1.6867016959289163e-05,
|
| 1765 |
+
"loss": 0.2063,
|
| 1766 |
+
"step": 6275
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.726643598615917,
|
| 1770 |
+
"grad_norm": 145.8665771484375,
|
| 1771 |
+
"learning_rate": 1.684565765303943e-05,
|
| 1772 |
+
"loss": 0.181,
|
| 1773 |
+
"step": 6300
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.7295271049596309,
|
| 1777 |
+
"grad_norm": 0.11629751324653625,
|
| 1778 |
+
"learning_rate": 1.6824298346789697e-05,
|
| 1779 |
+
"loss": 0.3259,
|
| 1780 |
+
"step": 6325
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.7324106113033448,
|
| 1784 |
+
"grad_norm": 0.23574000597000122,
|
| 1785 |
+
"learning_rate": 1.6802939040539964e-05,
|
| 1786 |
+
"loss": 0.1905,
|
| 1787 |
+
"step": 6350
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.7352941176470589,
|
| 1791 |
+
"grad_norm": 0.10753282159566879,
|
| 1792 |
+
"learning_rate": 1.678157973429023e-05,
|
| 1793 |
+
"loss": 0.205,
|
| 1794 |
+
"step": 6375
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.7381776239907728,
|
| 1798 |
+
"grad_norm": 3.6417784690856934,
|
| 1799 |
+
"learning_rate": 1.6760220428040498e-05,
|
| 1800 |
+
"loss": 0.1898,
|
| 1801 |
+
"step": 6400
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.7410611303344867,
|
| 1805 |
+
"grad_norm": 12.807069778442383,
|
| 1806 |
+
"learning_rate": 1.6738861121790764e-05,
|
| 1807 |
+
"loss": 0.1732,
|
| 1808 |
+
"step": 6425
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.7439446366782007,
|
| 1812 |
+
"grad_norm": 0.03930257260799408,
|
| 1813 |
+
"learning_rate": 1.671750181554103e-05,
|
| 1814 |
+
"loss": 0.1412,
|
| 1815 |
+
"step": 6450
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.7468281430219147,
|
| 1819 |
+
"grad_norm": 0.15581054985523224,
|
| 1820 |
+
"learning_rate": 1.66961425092913e-05,
|
| 1821 |
+
"loss": 0.1932,
|
| 1822 |
+
"step": 6475
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.7497116493656286,
|
| 1826 |
+
"grad_norm": 53.44426727294922,
|
| 1827 |
+
"learning_rate": 1.6674783203041568e-05,
|
| 1828 |
+
"loss": 0.256,
|
| 1829 |
+
"step": 6500
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.7525951557093425,
|
| 1833 |
+
"grad_norm": 7.01704216003418,
|
| 1834 |
+
"learning_rate": 1.6653423896791835e-05,
|
| 1835 |
+
"loss": 0.3403,
|
| 1836 |
+
"step": 6525
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.7554786620530565,
|
| 1840 |
+
"grad_norm": 0.0836925357580185,
|
| 1841 |
+
"learning_rate": 1.6632064590542102e-05,
|
| 1842 |
+
"loss": 0.13,
|
| 1843 |
+
"step": 6550
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.7583621683967705,
|
| 1847 |
+
"grad_norm": 21.222312927246094,
|
| 1848 |
+
"learning_rate": 1.661070528429237e-05,
|
| 1849 |
+
"loss": 0.2272,
|
| 1850 |
+
"step": 6575
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.7612456747404844,
|
| 1854 |
+
"grad_norm": 0.7010261416435242,
|
| 1855 |
+
"learning_rate": 1.6590200350292626e-05,
|
| 1856 |
+
"loss": 0.3342,
|
| 1857 |
+
"step": 6600
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.7641291810841984,
|
| 1861 |
+
"grad_norm": 0.15556718409061432,
|
| 1862 |
+
"learning_rate": 1.6568841044042893e-05,
|
| 1863 |
+
"loss": 0.307,
|
| 1864 |
+
"step": 6625
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.7670126874279123,
|
| 1868 |
+
"grad_norm": 0.05323057994246483,
|
| 1869 |
+
"learning_rate": 1.654748173779316e-05,
|
| 1870 |
+
"loss": 0.3262,
|
| 1871 |
+
"step": 6650
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.7698961937716263,
|
| 1875 |
+
"grad_norm": 0.16455906629562378,
|
| 1876 |
+
"learning_rate": 1.6526122431543426e-05,
|
| 1877 |
+
"loss": 0.1733,
|
| 1878 |
+
"step": 6675
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.7727797001153403,
|
| 1882 |
+
"grad_norm": 15.758971214294434,
|
| 1883 |
+
"learning_rate": 1.6504763125293693e-05,
|
| 1884 |
+
"loss": 0.3104,
|
| 1885 |
+
"step": 6700
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.7756632064590542,
|
| 1889 |
+
"grad_norm": 71.95575714111328,
|
| 1890 |
+
"learning_rate": 1.648340381904396e-05,
|
| 1891 |
+
"loss": 0.3078,
|
| 1892 |
+
"step": 6725
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.7785467128027682,
|
| 1896 |
+
"grad_norm": 0.15171046555042267,
|
| 1897 |
+
"learning_rate": 1.6462044512794227e-05,
|
| 1898 |
+
"loss": 0.2523,
|
| 1899 |
+
"step": 6750
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.7814302191464821,
|
| 1903 |
+
"grad_norm": 0.1595706045627594,
|
| 1904 |
+
"learning_rate": 1.6440685206544494e-05,
|
| 1905 |
+
"loss": 0.3617,
|
| 1906 |
+
"step": 6775
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.7843137254901961,
|
| 1910 |
+
"grad_norm": 29.170223236083984,
|
| 1911 |
+
"learning_rate": 1.641932590029476e-05,
|
| 1912 |
+
"loss": 0.2655,
|
| 1913 |
+
"step": 6800
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.78719723183391,
|
| 1917 |
+
"grad_norm": 18.99031639099121,
|
| 1918 |
+
"learning_rate": 1.6397966594045028e-05,
|
| 1919 |
+
"loss": 0.1425,
|
| 1920 |
+
"step": 6825
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.790080738177624,
|
| 1924 |
+
"grad_norm": 16.746435165405273,
|
| 1925 |
+
"learning_rate": 1.6376607287795294e-05,
|
| 1926 |
+
"loss": 0.2725,
|
| 1927 |
+
"step": 6850
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.792964244521338,
|
| 1931 |
+
"grad_norm": 0.4288238286972046,
|
| 1932 |
+
"learning_rate": 1.635524798154556e-05,
|
| 1933 |
+
"loss": 0.3489,
|
| 1934 |
+
"step": 6875
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.7958477508650519,
|
| 1938 |
+
"grad_norm": 8.68108081817627,
|
| 1939 |
+
"learning_rate": 1.6333888675295828e-05,
|
| 1940 |
+
"loss": 0.1509,
|
| 1941 |
+
"step": 6900
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.7987312572087658,
|
| 1945 |
+
"grad_norm": 0.11659915000200272,
|
| 1946 |
+
"learning_rate": 1.6312529369046095e-05,
|
| 1947 |
+
"loss": 0.2226,
|
| 1948 |
+
"step": 6925
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.8016147635524798,
|
| 1952 |
+
"grad_norm": 65.89250946044922,
|
| 1953 |
+
"learning_rate": 1.6291170062796362e-05,
|
| 1954 |
+
"loss": 0.1936,
|
| 1955 |
+
"step": 6950
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.8044982698961938,
|
| 1959 |
+
"grad_norm": 19.225807189941406,
|
| 1960 |
+
"learning_rate": 1.626981075654663e-05,
|
| 1961 |
+
"loss": 0.1961,
|
| 1962 |
+
"step": 6975
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.8073817762399077,
|
| 1966 |
+
"grad_norm": 0.07038139551877975,
|
| 1967 |
+
"learning_rate": 1.6248451450296895e-05,
|
| 1968 |
+
"loss": 0.3018,
|
| 1969 |
+
"step": 7000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.8102652825836217,
|
| 1973 |
+
"grad_norm": 11.083084106445312,
|
| 1974 |
+
"learning_rate": 1.6227092144047162e-05,
|
| 1975 |
+
"loss": 0.2297,
|
| 1976 |
+
"step": 7025
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.8131487889273357,
|
| 1980 |
+
"grad_norm": 14.312478065490723,
|
| 1981 |
+
"learning_rate": 1.620573283779743e-05,
|
| 1982 |
+
"loss": 0.2694,
|
| 1983 |
+
"step": 7050
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.8160322952710496,
|
| 1987 |
+
"grad_norm": 38.1439323425293,
|
| 1988 |
+
"learning_rate": 1.6184373531547696e-05,
|
| 1989 |
+
"loss": 0.1831,
|
| 1990 |
+
"step": 7075
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.8189158016147635,
|
| 1994 |
+
"grad_norm": 3.1983442306518555,
|
| 1995 |
+
"learning_rate": 1.6163014225297963e-05,
|
| 1996 |
+
"loss": 0.3203,
|
| 1997 |
+
"step": 7100
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.8217993079584776,
|
| 2001 |
+
"grad_norm": 0.1014777421951294,
|
| 2002 |
+
"learning_rate": 1.614165491904823e-05,
|
| 2003 |
+
"loss": 0.1245,
|
| 2004 |
+
"step": 7125
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.8246828143021915,
|
| 2008 |
+
"grad_norm": 6.031269550323486,
|
| 2009 |
+
"learning_rate": 1.6120295612798496e-05,
|
| 2010 |
+
"loss": 0.2052,
|
| 2011 |
+
"step": 7150
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.8275663206459054,
|
| 2015 |
+
"grad_norm": 29.11400032043457,
|
| 2016 |
+
"learning_rate": 1.6098936306548763e-05,
|
| 2017 |
+
"loss": 0.2821,
|
| 2018 |
+
"step": 7175
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.8304498269896193,
|
| 2022 |
+
"grad_norm": 23.76592254638672,
|
| 2023 |
+
"learning_rate": 1.607757700029903e-05,
|
| 2024 |
+
"loss": 0.1827,
|
| 2025 |
+
"step": 7200
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.8333333333333334,
|
| 2029 |
+
"grad_norm": 0.33321645855903625,
|
| 2030 |
+
"learning_rate": 1.6056217694049297e-05,
|
| 2031 |
+
"loss": 0.2097,
|
| 2032 |
+
"step": 7225
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.8362168396770473,
|
| 2036 |
+
"grad_norm": 0.26732248067855835,
|
| 2037 |
+
"learning_rate": 1.6034858387799567e-05,
|
| 2038 |
+
"loss": 0.2476,
|
| 2039 |
+
"step": 7250
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.8391003460207612,
|
| 2043 |
+
"grad_norm": 0.2259816974401474,
|
| 2044 |
+
"learning_rate": 1.6013499081549834e-05,
|
| 2045 |
+
"loss": 0.13,
|
| 2046 |
+
"step": 7275
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.8419838523644751,
|
| 2050 |
+
"grad_norm": 0.049072183668613434,
|
| 2051 |
+
"learning_rate": 1.59921397753001e-05,
|
| 2052 |
+
"loss": 0.1653,
|
| 2053 |
+
"step": 7300
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.8448673587081892,
|
| 2057 |
+
"grad_norm": 29.999534606933594,
|
| 2058 |
+
"learning_rate": 1.5970780469050368e-05,
|
| 2059 |
+
"loss": 0.3749,
|
| 2060 |
+
"step": 7325
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.8477508650519031,
|
| 2064 |
+
"grad_norm": 0.13831402361392975,
|
| 2065 |
+
"learning_rate": 1.5949421162800635e-05,
|
| 2066 |
+
"loss": 0.2006,
|
| 2067 |
+
"step": 7350
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.850634371395617,
|
| 2071 |
+
"grad_norm": 68.95947265625,
|
| 2072 |
+
"learning_rate": 1.59280618565509e-05,
|
| 2073 |
+
"loss": 0.2685,
|
| 2074 |
+
"step": 7375
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.8535178777393311,
|
| 2078 |
+
"grad_norm": 24.01173973083496,
|
| 2079 |
+
"learning_rate": 1.590670255030117e-05,
|
| 2080 |
+
"loss": 0.175,
|
| 2081 |
+
"step": 7400
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.856401384083045,
|
| 2085 |
+
"grad_norm": 0.2649136185646057,
|
| 2086 |
+
"learning_rate": 1.5885343244051435e-05,
|
| 2087 |
+
"loss": 0.2323,
|
| 2088 |
+
"step": 7425
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.8592848904267589,
|
| 2092 |
+
"grad_norm": 0.20093220472335815,
|
| 2093 |
+
"learning_rate": 1.5863983937801702e-05,
|
| 2094 |
+
"loss": 0.1903,
|
| 2095 |
+
"step": 7450
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.8621683967704729,
|
| 2099 |
+
"grad_norm": 0.5193009972572327,
|
| 2100 |
+
"learning_rate": 1.584262463155197e-05,
|
| 2101 |
+
"loss": 0.0966,
|
| 2102 |
+
"step": 7475
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.8650519031141869,
|
| 2106 |
+
"grad_norm": 22.47789764404297,
|
| 2107 |
+
"learning_rate": 1.5821265325302236e-05,
|
| 2108 |
+
"loss": 0.2615,
|
| 2109 |
+
"step": 7500
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.8679354094579008,
|
| 2113 |
+
"grad_norm": 0.21877521276474,
|
| 2114 |
+
"learning_rate": 1.5799906019052503e-05,
|
| 2115 |
+
"loss": 0.1871,
|
| 2116 |
+
"step": 7525
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.8708189158016147,
|
| 2120 |
+
"grad_norm": 5.362409591674805,
|
| 2121 |
+
"learning_rate": 1.577854671280277e-05,
|
| 2122 |
+
"loss": 0.1467,
|
| 2123 |
+
"step": 7550
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.8737024221453287,
|
| 2127 |
+
"grad_norm": 11.162101745605469,
|
| 2128 |
+
"learning_rate": 1.5757187406553036e-05,
|
| 2129 |
+
"loss": 0.3552,
|
| 2130 |
+
"step": 7575
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.8765859284890427,
|
| 2134 |
+
"grad_norm": 0.07614091038703918,
|
| 2135 |
+
"learning_rate": 1.5735828100303303e-05,
|
| 2136 |
+
"loss": 0.1898,
|
| 2137 |
+
"step": 7600
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.8794694348327566,
|
| 2141 |
+
"grad_norm": 1.4044957160949707,
|
| 2142 |
+
"learning_rate": 1.571446879405357e-05,
|
| 2143 |
+
"loss": 0.2969,
|
| 2144 |
+
"step": 7625
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.8823529411764706,
|
| 2148 |
+
"grad_norm": 17.89237403869629,
|
| 2149 |
+
"learning_rate": 1.5693109487803837e-05,
|
| 2150 |
+
"loss": 0.2066,
|
| 2151 |
+
"step": 7650
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.8852364475201846,
|
| 2155 |
+
"grad_norm": 29.870235443115234,
|
| 2156 |
+
"learning_rate": 1.5671750181554104e-05,
|
| 2157 |
+
"loss": 0.2441,
|
| 2158 |
+
"step": 7675
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.8881199538638985,
|
| 2162 |
+
"grad_norm": 51.64949035644531,
|
| 2163 |
+
"learning_rate": 1.565039087530437e-05,
|
| 2164 |
+
"loss": 0.25,
|
| 2165 |
+
"step": 7700
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.8910034602076125,
|
| 2169 |
+
"grad_norm": 0.9088483452796936,
|
| 2170 |
+
"learning_rate": 1.5629031569054637e-05,
|
| 2171 |
+
"loss": 0.2572,
|
| 2172 |
+
"step": 7725
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.8938869665513264,
|
| 2176 |
+
"grad_norm": 0.20605720579624176,
|
| 2177 |
+
"learning_rate": 1.5607672262804904e-05,
|
| 2178 |
+
"loss": 0.1968,
|
| 2179 |
+
"step": 7750
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.8967704728950404,
|
| 2183 |
+
"grad_norm": 73.5993881225586,
|
| 2184 |
+
"learning_rate": 1.558631295655517e-05,
|
| 2185 |
+
"loss": 0.1527,
|
| 2186 |
+
"step": 7775
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.8996539792387543,
|
| 2190 |
+
"grad_norm": 238.90350341796875,
|
| 2191 |
+
"learning_rate": 1.5564953650305438e-05,
|
| 2192 |
+
"loss": 0.3636,
|
| 2193 |
+
"step": 7800
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.9025374855824683,
|
| 2197 |
+
"grad_norm": 0.11910497397184372,
|
| 2198 |
+
"learning_rate": 1.5543594344055705e-05,
|
| 2199 |
+
"loss": 0.2838,
|
| 2200 |
+
"step": 7825
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.9054209919261822,
|
| 2204 |
+
"grad_norm": 0.027470508590340614,
|
| 2205 |
+
"learning_rate": 1.552223503780597e-05,
|
| 2206 |
+
"loss": 0.1275,
|
| 2207 |
+
"step": 7850
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.9083044982698962,
|
| 2211 |
+
"grad_norm": 0.052310917526483536,
|
| 2212 |
+
"learning_rate": 1.550087573155624e-05,
|
| 2213 |
+
"loss": 0.1901,
|
| 2214 |
+
"step": 7875
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.9111880046136102,
|
| 2218 |
+
"grad_norm": 1.4135714769363403,
|
| 2219 |
+
"learning_rate": 1.5479516425306505e-05,
|
| 2220 |
+
"loss": 0.2446,
|
| 2221 |
+
"step": 7900
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.9140715109573241,
|
| 2225 |
+
"grad_norm": 0.08867527544498444,
|
| 2226 |
+
"learning_rate": 1.5458157119056772e-05,
|
| 2227 |
+
"loss": 0.2053,
|
| 2228 |
+
"step": 7925
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.916955017301038,
|
| 2232 |
+
"grad_norm": 24.488529205322266,
|
| 2233 |
+
"learning_rate": 1.5436797812807042e-05,
|
| 2234 |
+
"loss": 0.1357,
|
| 2235 |
+
"step": 7950
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.919838523644752,
|
| 2239 |
+
"grad_norm": 58.86060333251953,
|
| 2240 |
+
"learning_rate": 1.541543850655731e-05,
|
| 2241 |
+
"loss": 0.3632,
|
| 2242 |
+
"step": 7975
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.922722029988466,
|
| 2246 |
+
"grad_norm": 0.019439009949564934,
|
| 2247 |
+
"learning_rate": 1.5394079200307576e-05,
|
| 2248 |
+
"loss": 0.1422,
|
| 2249 |
+
"step": 8000
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.9256055363321799,
|
| 2253 |
+
"grad_norm": 1.5843712091445923,
|
| 2254 |
+
"learning_rate": 1.5372719894057843e-05,
|
| 2255 |
+
"loss": 0.2779,
|
| 2256 |
+
"step": 8025
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.9284890426758939,
|
| 2260 |
+
"grad_norm": 0.5175610184669495,
|
| 2261 |
+
"learning_rate": 1.535136058780811e-05,
|
| 2262 |
+
"loss": 0.3052,
|
| 2263 |
+
"step": 8050
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.9313725490196079,
|
| 2267 |
+
"grad_norm": 0.0682525783777237,
|
| 2268 |
+
"learning_rate": 1.5330001281558376e-05,
|
| 2269 |
+
"loss": 0.1329,
|
| 2270 |
+
"step": 8075
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.9342560553633218,
|
| 2274 |
+
"grad_norm": 0.02290676161646843,
|
| 2275 |
+
"learning_rate": 1.5308641975308643e-05,
|
| 2276 |
+
"loss": 0.2221,
|
| 2277 |
+
"step": 8100
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.9371395617070357,
|
| 2281 |
+
"grad_norm": 0.12559133768081665,
|
| 2282 |
+
"learning_rate": 1.528728266905891e-05,
|
| 2283 |
+
"loss": 0.11,
|
| 2284 |
+
"step": 8125
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.9400230680507498,
|
| 2288 |
+
"grad_norm": 3.0466935634613037,
|
| 2289 |
+
"learning_rate": 1.5265923362809177e-05,
|
| 2290 |
+
"loss": 0.2059,
|
| 2291 |
+
"step": 8150
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.9429065743944637,
|
| 2295 |
+
"grad_norm": 9.555996894836426,
|
| 2296 |
+
"learning_rate": 1.5244564056559444e-05,
|
| 2297 |
+
"loss": 0.2314,
|
| 2298 |
+
"step": 8175
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.9457900807381776,
|
| 2302 |
+
"grad_norm": 0.06787367910146713,
|
| 2303 |
+
"learning_rate": 1.522320475030971e-05,
|
| 2304 |
+
"loss": 0.2672,
|
| 2305 |
+
"step": 8200
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.9486735870818915,
|
| 2309 |
+
"grad_norm": 12.74453353881836,
|
| 2310 |
+
"learning_rate": 1.5201845444059978e-05,
|
| 2311 |
+
"loss": 0.3794,
|
| 2312 |
+
"step": 8225
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.9515570934256056,
|
| 2316 |
+
"grad_norm": 0.1670641452074051,
|
| 2317 |
+
"learning_rate": 1.5180486137810244e-05,
|
| 2318 |
+
"loss": 0.1821,
|
| 2319 |
+
"step": 8250
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.9544405997693195,
|
| 2323 |
+
"grad_norm": 0.08948147296905518,
|
| 2324 |
+
"learning_rate": 1.5159126831560511e-05,
|
| 2325 |
+
"loss": 0.1434,
|
| 2326 |
+
"step": 8275
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.9573241061130334,
|
| 2330 |
+
"grad_norm": 0.451598584651947,
|
| 2331 |
+
"learning_rate": 1.5137767525310778e-05,
|
| 2332 |
+
"loss": 0.2515,
|
| 2333 |
+
"step": 8300
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.9602076124567474,
|
| 2337 |
+
"grad_norm": 4.159804821014404,
|
| 2338 |
+
"learning_rate": 1.5116408219061045e-05,
|
| 2339 |
+
"loss": 0.3622,
|
| 2340 |
+
"step": 8325
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.9630911188004614,
|
| 2344 |
+
"grad_norm": 3.510255813598633,
|
| 2345 |
+
"learning_rate": 1.5095048912811312e-05,
|
| 2346 |
+
"loss": 0.2408,
|
| 2347 |
+
"step": 8350
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.9659746251441753,
|
| 2351 |
+
"grad_norm": 0.15341836214065552,
|
| 2352 |
+
"learning_rate": 1.5073689606561582e-05,
|
| 2353 |
+
"loss": 0.2448,
|
| 2354 |
+
"step": 8375
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.9688581314878892,
|
| 2358 |
+
"grad_norm": 20.573637008666992,
|
| 2359 |
+
"learning_rate": 1.5052330300311849e-05,
|
| 2360 |
+
"loss": 0.2034,
|
| 2361 |
+
"step": 8400
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.9717416378316033,
|
| 2365 |
+
"grad_norm": 2.2208144664764404,
|
| 2366 |
+
"learning_rate": 1.5030970994062116e-05,
|
| 2367 |
+
"loss": 0.1718,
|
| 2368 |
+
"step": 8425
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.9746251441753172,
|
| 2372 |
+
"grad_norm": 9.478900909423828,
|
| 2373 |
+
"learning_rate": 1.5009611687812383e-05,
|
| 2374 |
+
"loss": 0.2298,
|
| 2375 |
+
"step": 8450
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.9775086505190311,
|
| 2379 |
+
"grad_norm": 0.03624691069126129,
|
| 2380 |
+
"learning_rate": 1.498825238156265e-05,
|
| 2381 |
+
"loss": 0.2407,
|
| 2382 |
+
"step": 8475
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.9803921568627451,
|
| 2386 |
+
"grad_norm": 0.11539937555789948,
|
| 2387 |
+
"learning_rate": 1.4966893075312916e-05,
|
| 2388 |
+
"loss": 0.2593,
|
| 2389 |
+
"step": 8500
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.9832756632064591,
|
| 2393 |
+
"grad_norm": 3.431764602661133,
|
| 2394 |
+
"learning_rate": 1.4945533769063183e-05,
|
| 2395 |
+
"loss": 0.2113,
|
| 2396 |
+
"step": 8525
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.986159169550173,
|
| 2400 |
+
"grad_norm": 5.679684162139893,
|
| 2401 |
+
"learning_rate": 1.492417446281345e-05,
|
| 2402 |
+
"loss": 0.245,
|
| 2403 |
+
"step": 8550
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.989042675893887,
|
| 2407 |
+
"grad_norm": 40.45989227294922,
|
| 2408 |
+
"learning_rate": 1.4902815156563717e-05,
|
| 2409 |
+
"loss": 0.1563,
|
| 2410 |
+
"step": 8575
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.9919261822376009,
|
| 2414 |
+
"grad_norm": 8.676360130310059,
|
| 2415 |
+
"learning_rate": 1.4881455850313984e-05,
|
| 2416 |
+
"loss": 0.2528,
|
| 2417 |
+
"step": 8600
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.9948096885813149,
|
| 2421 |
+
"grad_norm": 0.08342110365629196,
|
| 2422 |
+
"learning_rate": 1.486009654406425e-05,
|
| 2423 |
+
"loss": 0.2102,
|
| 2424 |
+
"step": 8625
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.9976931949250288,
|
| 2428 |
+
"grad_norm": 0.11708399653434753,
|
| 2429 |
+
"learning_rate": 1.4838737237814517e-05,
|
| 2430 |
+
"loss": 0.278,
|
| 2431 |
+
"step": 8650
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 1.0,
|
| 2435 |
+
"eval_accuracy": 0.9476355247981546,
|
| 2436 |
+
"eval_f1_macro": 0.7558411985220287,
|
| 2437 |
+
"eval_f1_micro": 0.9476355247981546,
|
| 2438 |
+
"eval_f1_weighted": 0.9440988511640933,
|
| 2439 |
+
"eval_loss": 0.1770678013563156,
|
| 2440 |
+
"eval_precision_macro": 0.8003866545209276,
|
| 2441 |
+
"eval_precision_micro": 0.9476355247981546,
|
| 2442 |
+
"eval_precision_weighted": 0.9423080358163488,
|
| 2443 |
+
"eval_recall_macro": 0.7247604343765456,
|
| 2444 |
+
"eval_recall_micro": 0.9476355247981546,
|
| 2445 |
+
"eval_recall_weighted": 0.9476355247981546,
|
| 2446 |
+
"eval_runtime": 21.7574,
|
| 2447 |
+
"eval_samples_per_second": 398.485,
|
| 2448 |
+
"eval_steps_per_second": 24.911,
|
| 2449 |
+
"step": 8670
|
| 2450 |
+
}
|
| 2451 |
+
],
|
| 2452 |
+
"logging_steps": 25,
|
| 2453 |
+
"max_steps": 26010,
|
| 2454 |
+
"num_input_tokens_seen": 0,
|
| 2455 |
+
"num_train_epochs": 3,
|
| 2456 |
+
"save_steps": 500,
|
| 2457 |
+
"stateful_callbacks": {
|
| 2458 |
+
"EarlyStoppingCallback": {
|
| 2459 |
+
"args": {
|
| 2460 |
+
"early_stopping_patience": 5,
|
| 2461 |
+
"early_stopping_threshold": 0.01
|
| 2462 |
+
},
|
| 2463 |
+
"attributes": {
|
| 2464 |
+
"early_stopping_patience_counter": 0
|
| 2465 |
+
}
|
| 2466 |
+
},
|
| 2467 |
+
"TrainerControl": {
|
| 2468 |
+
"args": {
|
| 2469 |
+
"should_epoch_stop": false,
|
| 2470 |
+
"should_evaluate": false,
|
| 2471 |
+
"should_log": false,
|
| 2472 |
+
"should_save": true,
|
| 2473 |
+
"should_training_stop": false
|
| 2474 |
+
},
|
| 2475 |
+
"attributes": {}
|
| 2476 |
+
}
|
| 2477 |
+
},
|
| 2478 |
+
"total_flos": 4562189328340224.0,
|
| 2479 |
+
"train_batch_size": 8,
|
| 2480 |
+
"trial_name": null,
|
| 2481 |
+
"trial_params": null
|
| 2482 |
+
}
|
checkpoint-8670/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b465d8bad10fe8145104c31999e081366c5e664fbba52f270fcb5f4e46cd56a3
|
| 3 |
+
size 5240
|
config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "FacebookAI/roberta-base",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"RobertaForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_probs_dropout_prob": 0.1,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"classifier_dropout": null,
|
| 10 |
+
"eos_token_id": 2,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": "1",
|
| 16 |
+
"1": "3",
|
| 17 |
+
"2": "5"
|
| 18 |
+
},
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 3072,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"1": 0,
|
| 23 |
+
"3": 1,
|
| 24 |
+
"5": 2
|
| 25 |
+
},
|
| 26 |
+
"layer_norm_eps": 1e-05,
|
| 27 |
+
"max_position_embeddings": 514,
|
| 28 |
+
"model_type": "roberta",
|
| 29 |
+
"num_attention_heads": 12,
|
| 30 |
+
"num_hidden_layers": 12,
|
| 31 |
+
"pad_token_id": 1,
|
| 32 |
+
"position_embedding_type": "absolute",
|
| 33 |
+
"problem_type": "single_label_classification",
|
| 34 |
+
"torch_dtype": "float32",
|
| 35 |
+
"transformers_version": "4.43.1",
|
| 36 |
+
"type_vocab_size": 1,
|
| 37 |
+
"use_cache": true,
|
| 38 |
+
"vocab_size": 50265
|
| 39 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37427f4f57e436f313489f2927538add4c967871063912df29ba2fbf93a7a8aa
|
| 3 |
+
size 498615900
|
runs/Aug08_04-59-57_r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4/events.out.tfevents.1723093198.r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4.101.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd5a4cf2c2ebcb2b2e714d7abb6091273472e75075d107d5c82d63a008dde407
|
| 3 |
+
size 228943
|
runs/Aug08_04-59-57_r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4/events.out.tfevents.1723096032.r-riken01-flan-t5-base-h8osfsgf-e6f8c-z8ly4.101.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:512609ff2ff60075e23ea7a87ab247d57c680f027b18bcfa030fe65d790e0847
|
| 3 |
+
size 936
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<pad>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"50264": {
|
| 37 |
+
"content": "<mask>",
|
| 38 |
+
"lstrip": true,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"bos_token": "<s>",
|
| 46 |
+
"clean_up_tokenization_spaces": true,
|
| 47 |
+
"cls_token": "<s>",
|
| 48 |
+
"eos_token": "</s>",
|
| 49 |
+
"errors": "replace",
|
| 50 |
+
"mask_token": "<mask>",
|
| 51 |
+
"model_max_length": 512,
|
| 52 |
+
"pad_token": "<pad>",
|
| 53 |
+
"sep_token": "</s>",
|
| 54 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 55 |
+
"trim_offsets": true,
|
| 56 |
+
"unk_token": "<unk>"
|
| 57 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b465d8bad10fe8145104c31999e081366c5e664fbba52f270fcb5f4e46cd56a3
|
| 3 |
+
size 5240
|
training_params.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"data_path": "autotrain-qhv86-nrgf8/autotrain-data",
|
| 3 |
+
"model": "FacebookAI/roberta-base",
|
| 4 |
+
"lr": 2e-05,
|
| 5 |
+
"epochs": 3,
|
| 6 |
+
"max_seq_length": 128,
|
| 7 |
+
"batch_size": 8,
|
| 8 |
+
"warmup_ratio": 0.1,
|
| 9 |
+
"gradient_accumulation": 1,
|
| 10 |
+
"optimizer": "adamw_torch",
|
| 11 |
+
"scheduler": "linear",
|
| 12 |
+
"weight_decay": 0.0,
|
| 13 |
+
"max_grad_norm": 1.0,
|
| 14 |
+
"seed": 42,
|
| 15 |
+
"train_split": "train",
|
| 16 |
+
"valid_split": "validation",
|
| 17 |
+
"text_column": "autotrain_text",
|
| 18 |
+
"target_column": "autotrain_label",
|
| 19 |
+
"logging_steps": -1,
|
| 20 |
+
"project_name": "autotrain-qhv86-nrgf8",
|
| 21 |
+
"auto_find_batch_size": false,
|
| 22 |
+
"mixed_precision": "fp16",
|
| 23 |
+
"save_total_limit": 1,
|
| 24 |
+
"push_to_hub": true,
|
| 25 |
+
"eval_strategy": "epoch",
|
| 26 |
+
"username": "riken01",
|
| 27 |
+
"log": "tensorboard",
|
| 28 |
+
"early_stopping_patience": 5,
|
| 29 |
+
"early_stopping_threshold": 0.01
|
| 30 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|