Upload folder using huggingface_hub
Browse files- README.md +36 -0
- checkpoint-6201/config.json +55 -0
- checkpoint-6201/model.safetensors +3 -0
- checkpoint-6201/optimizer.pt +3 -0
- checkpoint-6201/rng_state.pth +3 -0
- checkpoint-6201/scheduler.pt +3 -0
- checkpoint-6201/trainer_state.json +1832 -0
- checkpoint-6201/training_args.bin +3 -0
- config.json +55 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- runs/Nov01_08-49-04_r-kafikani-longformer-va66dsc0-0e056-fjgin/events.out.tfevents.1730450948.r-kafikani-longformer-va66dsc0-0e056-fjgin.465.0 +2 -2
- runs/Nov01_08-49-04_r-kafikani-longformer-va66dsc0-0e056-fjgin/events.out.tfevents.1730870444.r-kafikani-longformer-va66dsc0-0e056-fjgin.465.1 +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: allenai/longformer-base-4096
|
| 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: 1.307055115699768
|
| 17 |
+
|
| 18 |
+
f1_macro: 0.5244016249451032
|
| 19 |
+
|
| 20 |
+
f1_micro: 0.7504835589941973
|
| 21 |
+
|
| 22 |
+
f1_weighted: 0.714761195760481
|
| 23 |
+
|
| 24 |
+
precision_macro: 0.5012229210342417
|
| 25 |
+
|
| 26 |
+
precision_micro: 0.7504835589941973
|
| 27 |
+
|
| 28 |
+
precision_weighted: 0.6860840439724423
|
| 29 |
+
|
| 30 |
+
recall_macro: 0.5532259049014222
|
| 31 |
+
|
| 32 |
+
recall_micro: 0.7504835589941973
|
| 33 |
+
|
| 34 |
+
recall_weighted: 0.7504835589941973
|
| 35 |
+
|
| 36 |
+
accuracy: 0.7504835589941973
|
checkpoint-6201/config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "allenai/longformer-base-4096",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"LongformerForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_mode": "longformer",
|
| 8 |
+
"attention_probs_dropout_prob": 0.1,
|
| 9 |
+
"attention_window": [
|
| 10 |
+
512,
|
| 11 |
+
512,
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512
|
| 22 |
+
],
|
| 23 |
+
"bos_token_id": 0,
|
| 24 |
+
"eos_token_id": 2,
|
| 25 |
+
"gradient_checkpointing": false,
|
| 26 |
+
"hidden_act": "gelu",
|
| 27 |
+
"hidden_dropout_prob": 0.1,
|
| 28 |
+
"hidden_size": 768,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "negative",
|
| 31 |
+
"1": "neutral",
|
| 32 |
+
"2": "positive"
|
| 33 |
+
},
|
| 34 |
+
"ignore_attention_mask": false,
|
| 35 |
+
"initializer_range": 0.02,
|
| 36 |
+
"intermediate_size": 3072,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"negative": 0,
|
| 39 |
+
"neutral": 1,
|
| 40 |
+
"positive": 2
|
| 41 |
+
},
|
| 42 |
+
"layer_norm_eps": 1e-05,
|
| 43 |
+
"max_position_embeddings": 4098,
|
| 44 |
+
"model_type": "longformer",
|
| 45 |
+
"num_attention_heads": 12,
|
| 46 |
+
"num_hidden_layers": 12,
|
| 47 |
+
"onnx_export": false,
|
| 48 |
+
"pad_token_id": 1,
|
| 49 |
+
"problem_type": "single_label_classification",
|
| 50 |
+
"sep_token_id": 2,
|
| 51 |
+
"torch_dtype": "float32",
|
| 52 |
+
"transformers_version": "4.45.0",
|
| 53 |
+
"type_vocab_size": 1,
|
| 54 |
+
"vocab_size": 50265
|
| 55 |
+
}
|
checkpoint-6201/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:322dae133d3ce467a820df6105d8c7973c40d913fb2ce1117801d5c7b57c139f
|
| 3 |
+
size 594681260
|
checkpoint-6201/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b69385ae3f0d8b25a998fc24255da1a205051f30b2e71450956ea059fb08b4e
|
| 3 |
+
size 1189514810
|
checkpoint-6201/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:66988915a0a37f08074f2bbf10041158cae43ca697efd6a7f943e09e0e6bd7e4
|
| 3 |
+
size 13990
|
checkpoint-6201/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8311d9af8bc0939f4f57ff934d5efeb067fdfe2d024c64e6f9d34a9e095e315b
|
| 3 |
+
size 1064
|
checkpoint-6201/trainer_state.json
ADDED
|
@@ -0,0 +1,1832 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.307055115699768,
|
| 3 |
+
"best_model_checkpoint": "autotrain-iinjh-0wh75/checkpoint-6201",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 6201,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.012094823415578132,
|
| 13 |
+
"grad_norm": 10.670669555664062,
|
| 14 |
+
"learning_rate": 2.012882447665056e-06,
|
| 15 |
+
"loss": 1.0613,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.024189646831156264,
|
| 20 |
+
"grad_norm": 17.839488983154297,
|
| 21 |
+
"learning_rate": 4.025764895330112e-06,
|
| 22 |
+
"loss": 1.0539,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.036284470246734396,
|
| 27 |
+
"grad_norm": 9.838277816772461,
|
| 28 |
+
"learning_rate": 6.038647342995169e-06,
|
| 29 |
+
"loss": 1.0026,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04837929366231253,
|
| 34 |
+
"grad_norm": 18.464771270751953,
|
| 35 |
+
"learning_rate": 8.051529790660225e-06,
|
| 36 |
+
"loss": 0.9234,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06047411707789066,
|
| 41 |
+
"grad_norm": 14.502671241760254,
|
| 42 |
+
"learning_rate": 1.0064412238325282e-05,
|
| 43 |
+
"loss": 1.2115,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07256894049346879,
|
| 48 |
+
"grad_norm": 13.285282135009766,
|
| 49 |
+
"learning_rate": 1.2077294685990338e-05,
|
| 50 |
+
"loss": 1.0436,
|
| 51 |
+
"step": 150
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08466376390904692,
|
| 55 |
+
"grad_norm": 10.353692054748535,
|
| 56 |
+
"learning_rate": 1.4090177133655394e-05,
|
| 57 |
+
"loss": 0.9726,
|
| 58 |
+
"step": 175
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.09675858732462506,
|
| 62 |
+
"grad_norm": 8.703958511352539,
|
| 63 |
+
"learning_rate": 1.610305958132045e-05,
|
| 64 |
+
"loss": 0.8265,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.10885341074020319,
|
| 69 |
+
"grad_norm": 13.100640296936035,
|
| 70 |
+
"learning_rate": 1.8115942028985507e-05,
|
| 71 |
+
"loss": 1.2616,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.12094823415578132,
|
| 76 |
+
"grad_norm": 20.051851272583008,
|
| 77 |
+
"learning_rate": 2.0128824476650564e-05,
|
| 78 |
+
"loss": 1.0757,
|
| 79 |
+
"step": 250
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.13304305757135945,
|
| 83 |
+
"grad_norm": 10.080187797546387,
|
| 84 |
+
"learning_rate": 2.214170692431562e-05,
|
| 85 |
+
"loss": 1.1058,
|
| 86 |
+
"step": 275
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.14513788098693758,
|
| 90 |
+
"grad_norm": 1.0888252258300781,
|
| 91 |
+
"learning_rate": 2.4154589371980676e-05,
|
| 92 |
+
"loss": 1.0442,
|
| 93 |
+
"step": 300
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.15723270440251572,
|
| 97 |
+
"grad_norm": 216.78359985351562,
|
| 98 |
+
"learning_rate": 2.6167471819645733e-05,
|
| 99 |
+
"loss": 0.4581,
|
| 100 |
+
"step": 325
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.16932752781809385,
|
| 104 |
+
"grad_norm": 3.2515599727630615,
|
| 105 |
+
"learning_rate": 2.8180354267310787e-05,
|
| 106 |
+
"loss": 1.7329,
|
| 107 |
+
"step": 350
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.18142235123367198,
|
| 111 |
+
"grad_norm": 4.606472969055176,
|
| 112 |
+
"learning_rate": 3.0193236714975848e-05,
|
| 113 |
+
"loss": 1.8193,
|
| 114 |
+
"step": 375
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1935171746492501,
|
| 118 |
+
"grad_norm": 4.541548252105713,
|
| 119 |
+
"learning_rate": 3.22061191626409e-05,
|
| 120 |
+
"loss": 1.9815,
|
| 121 |
+
"step": 400
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.20561199806482824,
|
| 125 |
+
"grad_norm": 0.2636414170265198,
|
| 126 |
+
"learning_rate": 3.421900161030596e-05,
|
| 127 |
+
"loss": 1.1492,
|
| 128 |
+
"step": 425
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.21770682148040638,
|
| 132 |
+
"grad_norm": 5.543747425079346,
|
| 133 |
+
"learning_rate": 3.6231884057971014e-05,
|
| 134 |
+
"loss": 1.4995,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2298016448959845,
|
| 139 |
+
"grad_norm": 29.730493545532227,
|
| 140 |
+
"learning_rate": 3.824476650563607e-05,
|
| 141 |
+
"loss": 1.736,
|
| 142 |
+
"step": 475
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.24189646831156264,
|
| 146 |
+
"grad_norm": 0.5111683011054993,
|
| 147 |
+
"learning_rate": 4.025764895330113e-05,
|
| 148 |
+
"loss": 1.7303,
|
| 149 |
+
"step": 500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2539912917271408,
|
| 153 |
+
"grad_norm": 1.9221380949020386,
|
| 154 |
+
"learning_rate": 4.2270531400966186e-05,
|
| 155 |
+
"loss": 1.3904,
|
| 156 |
+
"step": 525
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2660861151427189,
|
| 160 |
+
"grad_norm": 0.305082768201828,
|
| 161 |
+
"learning_rate": 4.428341384863124e-05,
|
| 162 |
+
"loss": 1.6028,
|
| 163 |
+
"step": 550
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.27818093855829706,
|
| 167 |
+
"grad_norm": 49.53160858154297,
|
| 168 |
+
"learning_rate": 4.62962962962963e-05,
|
| 169 |
+
"loss": 1.5488,
|
| 170 |
+
"step": 575
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.29027576197387517,
|
| 174 |
+
"grad_norm": 0.5875459313392639,
|
| 175 |
+
"learning_rate": 4.830917874396135e-05,
|
| 176 |
+
"loss": 2.5069,
|
| 177 |
+
"step": 600
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.30237058538945333,
|
| 181 |
+
"grad_norm": 0.1444765031337738,
|
| 182 |
+
"learning_rate": 4.996415770609319e-05,
|
| 183 |
+
"loss": 1.1429,
|
| 184 |
+
"step": 625
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.31446540880503143,
|
| 188 |
+
"grad_norm": 6.752070903778076,
|
| 189 |
+
"learning_rate": 4.974014336917563e-05,
|
| 190 |
+
"loss": 1.1868,
|
| 191 |
+
"step": 650
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3265602322206096,
|
| 195 |
+
"grad_norm": 2.889754056930542,
|
| 196 |
+
"learning_rate": 4.951612903225807e-05,
|
| 197 |
+
"loss": 1.2563,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.3386550556361877,
|
| 202 |
+
"grad_norm": 0.23045390844345093,
|
| 203 |
+
"learning_rate": 4.92921146953405e-05,
|
| 204 |
+
"loss": 1.5123,
|
| 205 |
+
"step": 700
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.35074987905176586,
|
| 209 |
+
"grad_norm": 0.20135998725891113,
|
| 210 |
+
"learning_rate": 4.906810035842294e-05,
|
| 211 |
+
"loss": 1.7428,
|
| 212 |
+
"step": 725
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.36284470246734396,
|
| 216 |
+
"grad_norm": 0.06913596391677856,
|
| 217 |
+
"learning_rate": 4.884408602150538e-05,
|
| 218 |
+
"loss": 1.2342,
|
| 219 |
+
"step": 750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3749395258829221,
|
| 223 |
+
"grad_norm": 2.32104754447937,
|
| 224 |
+
"learning_rate": 4.8620071684587816e-05,
|
| 225 |
+
"loss": 1.4426,
|
| 226 |
+
"step": 775
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3870343492985002,
|
| 230 |
+
"grad_norm": 1.5299639701843262,
|
| 231 |
+
"learning_rate": 4.8396057347670255e-05,
|
| 232 |
+
"loss": 1.3624,
|
| 233 |
+
"step": 800
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3991291727140784,
|
| 237 |
+
"grad_norm": 0.35129979252815247,
|
| 238 |
+
"learning_rate": 4.8172043010752693e-05,
|
| 239 |
+
"loss": 1.9102,
|
| 240 |
+
"step": 825
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4112239961296565,
|
| 244 |
+
"grad_norm": 11.400070190429688,
|
| 245 |
+
"learning_rate": 4.7948028673835125e-05,
|
| 246 |
+
"loss": 1.5198,
|
| 247 |
+
"step": 850
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.42331881954523465,
|
| 251 |
+
"grad_norm": 13.278685569763184,
|
| 252 |
+
"learning_rate": 4.7724014336917564e-05,
|
| 253 |
+
"loss": 1.146,
|
| 254 |
+
"step": 875
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.43541364296081275,
|
| 258 |
+
"grad_norm": 14.47873592376709,
|
| 259 |
+
"learning_rate": 4.75e-05,
|
| 260 |
+
"loss": 0.8003,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4475084663763909,
|
| 265 |
+
"grad_norm": 13.323905944824219,
|
| 266 |
+
"learning_rate": 4.727598566308244e-05,
|
| 267 |
+
"loss": 1.0161,
|
| 268 |
+
"step": 925
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.459603289791969,
|
| 272 |
+
"grad_norm": 29.149803161621094,
|
| 273 |
+
"learning_rate": 4.705197132616488e-05,
|
| 274 |
+
"loss": 2.1808,
|
| 275 |
+
"step": 950
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4716981132075472,
|
| 279 |
+
"grad_norm": 14.055343627929688,
|
| 280 |
+
"learning_rate": 4.682795698924731e-05,
|
| 281 |
+
"loss": 1.9727,
|
| 282 |
+
"step": 975
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4837929366231253,
|
| 286 |
+
"grad_norm": 13.733134269714355,
|
| 287 |
+
"learning_rate": 4.660394265232975e-05,
|
| 288 |
+
"loss": 1.3969,
|
| 289 |
+
"step": 1000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.49588776003870344,
|
| 293 |
+
"grad_norm": 12.515105247497559,
|
| 294 |
+
"learning_rate": 4.637992831541219e-05,
|
| 295 |
+
"loss": 1.1516,
|
| 296 |
+
"step": 1025
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.5079825834542816,
|
| 300 |
+
"grad_norm": 20.534067153930664,
|
| 301 |
+
"learning_rate": 4.615591397849463e-05,
|
| 302 |
+
"loss": 2.2554,
|
| 303 |
+
"step": 1050
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5200774068698597,
|
| 307 |
+
"grad_norm": 16.800443649291992,
|
| 308 |
+
"learning_rate": 4.5931899641577066e-05,
|
| 309 |
+
"loss": 1.6032,
|
| 310 |
+
"step": 1075
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5321722302854378,
|
| 314 |
+
"grad_norm": 2.256089448928833,
|
| 315 |
+
"learning_rate": 4.57078853046595e-05,
|
| 316 |
+
"loss": 1.7488,
|
| 317 |
+
"step": 1100
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5442670537010159,
|
| 321 |
+
"grad_norm": 13.562662124633789,
|
| 322 |
+
"learning_rate": 4.548387096774194e-05,
|
| 323 |
+
"loss": 1.2443,
|
| 324 |
+
"step": 1125
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5563618771165941,
|
| 328 |
+
"grad_norm": 12.777020454406738,
|
| 329 |
+
"learning_rate": 4.5259856630824375e-05,
|
| 330 |
+
"loss": 1.3996,
|
| 331 |
+
"step": 1150
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5684567005321722,
|
| 335 |
+
"grad_norm": 11.443547248840332,
|
| 336 |
+
"learning_rate": 4.503584229390681e-05,
|
| 337 |
+
"loss": 1.2757,
|
| 338 |
+
"step": 1175
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5805515239477503,
|
| 342 |
+
"grad_norm": 3.251011848449707,
|
| 343 |
+
"learning_rate": 4.481182795698925e-05,
|
| 344 |
+
"loss": 1.0835,
|
| 345 |
+
"step": 1200
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5926463473633284,
|
| 349 |
+
"grad_norm": 12.59429931640625,
|
| 350 |
+
"learning_rate": 4.458781362007169e-05,
|
| 351 |
+
"loss": 1.4228,
|
| 352 |
+
"step": 1225
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6047411707789067,
|
| 356 |
+
"grad_norm": 1.7253248691558838,
|
| 357 |
+
"learning_rate": 4.436379928315412e-05,
|
| 358 |
+
"loss": 1.7177,
|
| 359 |
+
"step": 1250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6168359941944848,
|
| 363 |
+
"grad_norm": 15.0354642868042,
|
| 364 |
+
"learning_rate": 4.413978494623656e-05,
|
| 365 |
+
"loss": 1.7989,
|
| 366 |
+
"step": 1275
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6289308176100629,
|
| 370 |
+
"grad_norm": 13.84261417388916,
|
| 371 |
+
"learning_rate": 4.3915770609318994e-05,
|
| 372 |
+
"loss": 1.4853,
|
| 373 |
+
"step": 1300
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6410256410256411,
|
| 377 |
+
"grad_norm": 13.903498649597168,
|
| 378 |
+
"learning_rate": 4.369175627240143e-05,
|
| 379 |
+
"loss": 1.1268,
|
| 380 |
+
"step": 1325
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6531204644412192,
|
| 384 |
+
"grad_norm": 1.7544844150543213,
|
| 385 |
+
"learning_rate": 4.346774193548388e-05,
|
| 386 |
+
"loss": 1.3478,
|
| 387 |
+
"step": 1350
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.6652152878567973,
|
| 391 |
+
"grad_norm": 16.601131439208984,
|
| 392 |
+
"learning_rate": 4.324372759856631e-05,
|
| 393 |
+
"loss": 1.5037,
|
| 394 |
+
"step": 1375
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6773101112723754,
|
| 398 |
+
"grad_norm": 0.15384919941425323,
|
| 399 |
+
"learning_rate": 4.301971326164875e-05,
|
| 400 |
+
"loss": 1.8697,
|
| 401 |
+
"step": 1400
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6894049346879536,
|
| 405 |
+
"grad_norm": 16.029470443725586,
|
| 406 |
+
"learning_rate": 4.279569892473119e-05,
|
| 407 |
+
"loss": 1.7013,
|
| 408 |
+
"step": 1425
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.7014997581035317,
|
| 412 |
+
"grad_norm": 1.9329349994659424,
|
| 413 |
+
"learning_rate": 4.257168458781362e-05,
|
| 414 |
+
"loss": 1.852,
|
| 415 |
+
"step": 1450
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7135945815191098,
|
| 419 |
+
"grad_norm": 34.523712158203125,
|
| 420 |
+
"learning_rate": 4.234767025089606e-05,
|
| 421 |
+
"loss": 1.1739,
|
| 422 |
+
"step": 1475
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7256894049346879,
|
| 426 |
+
"grad_norm": 18.080400466918945,
|
| 427 |
+
"learning_rate": 4.2123655913978496e-05,
|
| 428 |
+
"loss": 1.1161,
|
| 429 |
+
"step": 1500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7377842283502661,
|
| 433 |
+
"grad_norm": 13.704357147216797,
|
| 434 |
+
"learning_rate": 4.1899641577060935e-05,
|
| 435 |
+
"loss": 1.1861,
|
| 436 |
+
"step": 1525
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7498790517658442,
|
| 440 |
+
"grad_norm": 10.594738006591797,
|
| 441 |
+
"learning_rate": 4.167562724014337e-05,
|
| 442 |
+
"loss": 1.2265,
|
| 443 |
+
"step": 1550
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.7619738751814223,
|
| 447 |
+
"grad_norm": 1.4613639116287231,
|
| 448 |
+
"learning_rate": 4.1451612903225805e-05,
|
| 449 |
+
"loss": 1.173,
|
| 450 |
+
"step": 1575
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7740686985970004,
|
| 454 |
+
"grad_norm": 0.7370879054069519,
|
| 455 |
+
"learning_rate": 4.1227598566308244e-05,
|
| 456 |
+
"loss": 1.9062,
|
| 457 |
+
"step": 1600
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7861635220125787,
|
| 461 |
+
"grad_norm": 1.4284896850585938,
|
| 462 |
+
"learning_rate": 4.100358422939068e-05,
|
| 463 |
+
"loss": 1.9307,
|
| 464 |
+
"step": 1625
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7982583454281568,
|
| 468 |
+
"grad_norm": 3.1330041885375977,
|
| 469 |
+
"learning_rate": 4.077956989247312e-05,
|
| 470 |
+
"loss": 1.4191,
|
| 471 |
+
"step": 1650
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.8103531688437349,
|
| 475 |
+
"grad_norm": 25.01806640625,
|
| 476 |
+
"learning_rate": 4.055555555555556e-05,
|
| 477 |
+
"loss": 1.8695,
|
| 478 |
+
"step": 1675
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.822447992259313,
|
| 482 |
+
"grad_norm": 24.93403434753418,
|
| 483 |
+
"learning_rate": 4.0331541218638e-05,
|
| 484 |
+
"loss": 1.0304,
|
| 485 |
+
"step": 1700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8345428156748912,
|
| 489 |
+
"grad_norm": 0.39635589718818665,
|
| 490 |
+
"learning_rate": 4.010752688172043e-05,
|
| 491 |
+
"loss": 1.5612,
|
| 492 |
+
"step": 1725
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8466376390904693,
|
| 496 |
+
"grad_norm": 0.25706031918525696,
|
| 497 |
+
"learning_rate": 3.988351254480287e-05,
|
| 498 |
+
"loss": 1.6167,
|
| 499 |
+
"step": 1750
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.8587324625060474,
|
| 503 |
+
"grad_norm": 204.7973175048828,
|
| 504 |
+
"learning_rate": 3.965949820788531e-05,
|
| 505 |
+
"loss": 1.033,
|
| 506 |
+
"step": 1775
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.8708272859216255,
|
| 510 |
+
"grad_norm": 0.1547696739435196,
|
| 511 |
+
"learning_rate": 3.9435483870967746e-05,
|
| 512 |
+
"loss": 0.9405,
|
| 513 |
+
"step": 1800
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.8829221093372037,
|
| 517 |
+
"grad_norm": 25.602577209472656,
|
| 518 |
+
"learning_rate": 3.9211469534050185e-05,
|
| 519 |
+
"loss": 1.2914,
|
| 520 |
+
"step": 1825
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.8950169327527818,
|
| 524 |
+
"grad_norm": 25.09197425842285,
|
| 525 |
+
"learning_rate": 3.8987455197132616e-05,
|
| 526 |
+
"loss": 0.8189,
|
| 527 |
+
"step": 1850
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.9071117561683599,
|
| 531 |
+
"grad_norm": 0.08802329748868942,
|
| 532 |
+
"learning_rate": 3.8763440860215055e-05,
|
| 533 |
+
"loss": 1.7206,
|
| 534 |
+
"step": 1875
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.919206579583938,
|
| 538 |
+
"grad_norm": 23.29674530029297,
|
| 539 |
+
"learning_rate": 3.8539426523297494e-05,
|
| 540 |
+
"loss": 2.5356,
|
| 541 |
+
"step": 1900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9313014029995162,
|
| 545 |
+
"grad_norm": 24.580432891845703,
|
| 546 |
+
"learning_rate": 3.8315412186379926e-05,
|
| 547 |
+
"loss": 1.6752,
|
| 548 |
+
"step": 1925
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9433962264150944,
|
| 552 |
+
"grad_norm": 2.269275665283203,
|
| 553 |
+
"learning_rate": 3.809139784946237e-05,
|
| 554 |
+
"loss": 1.9433,
|
| 555 |
+
"step": 1950
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.9554910498306725,
|
| 559 |
+
"grad_norm": 3.1146695613861084,
|
| 560 |
+
"learning_rate": 3.786738351254481e-05,
|
| 561 |
+
"loss": 1.562,
|
| 562 |
+
"step": 1975
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.9675858732462506,
|
| 566 |
+
"grad_norm": 0.2862231135368347,
|
| 567 |
+
"learning_rate": 3.764336917562724e-05,
|
| 568 |
+
"loss": 0.6273,
|
| 569 |
+
"step": 2000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.9796806966618288,
|
| 573 |
+
"grad_norm": 0.39030689001083374,
|
| 574 |
+
"learning_rate": 3.741935483870968e-05,
|
| 575 |
+
"loss": 1.7117,
|
| 576 |
+
"step": 2025
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.9917755200774069,
|
| 580 |
+
"grad_norm": 25.142972946166992,
|
| 581 |
+
"learning_rate": 3.719534050179211e-05,
|
| 582 |
+
"loss": 1.8362,
|
| 583 |
+
"step": 2050
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.0,
|
| 587 |
+
"eval_accuracy": 0.6054158607350096,
|
| 588 |
+
"eval_f1_macro": 0.41404572439055193,
|
| 589 |
+
"eval_f1_micro": 0.6054158607350096,
|
| 590 |
+
"eval_f1_weighted": 0.5589758737227555,
|
| 591 |
+
"eval_loss": 1.5838946104049683,
|
| 592 |
+
"eval_precision_macro": 0.46238977618202576,
|
| 593 |
+
"eval_precision_micro": 0.6054158607350096,
|
| 594 |
+
"eval_precision_weighted": 0.6418784559670553,
|
| 595 |
+
"eval_recall_macro": 0.46003784410669596,
|
| 596 |
+
"eval_recall_micro": 0.6054158607350096,
|
| 597 |
+
"eval_recall_weighted": 0.6054158607350096,
|
| 598 |
+
"eval_runtime": 7116.055,
|
| 599 |
+
"eval_samples_per_second": 0.073,
|
| 600 |
+
"eval_steps_per_second": 0.036,
|
| 601 |
+
"step": 2067
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 1.003870343492985,
|
| 605 |
+
"grad_norm": 23.918968200683594,
|
| 606 |
+
"learning_rate": 3.697132616487455e-05,
|
| 607 |
+
"loss": 1.7229,
|
| 608 |
+
"step": 2075
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.0159651669085632,
|
| 612 |
+
"grad_norm": 24.551406860351562,
|
| 613 |
+
"learning_rate": 3.6747311827956996e-05,
|
| 614 |
+
"loss": 0.7897,
|
| 615 |
+
"step": 2100
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.0280599903241412,
|
| 619 |
+
"grad_norm": 0.6501819491386414,
|
| 620 |
+
"learning_rate": 3.652329749103943e-05,
|
| 621 |
+
"loss": 1.4349,
|
| 622 |
+
"step": 2125
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.0401548137397194,
|
| 626 |
+
"grad_norm": 0.4270811676979065,
|
| 627 |
+
"learning_rate": 3.6299283154121866e-05,
|
| 628 |
+
"loss": 1.4771,
|
| 629 |
+
"step": 2150
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.0522496371552976,
|
| 633 |
+
"grad_norm": 25.46353530883789,
|
| 634 |
+
"learning_rate": 3.6075268817204305e-05,
|
| 635 |
+
"loss": 1.6769,
|
| 636 |
+
"step": 2175
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.0643444605708756,
|
| 640 |
+
"grad_norm": 24.238040924072266,
|
| 641 |
+
"learning_rate": 3.585125448028674e-05,
|
| 642 |
+
"loss": 2.2981,
|
| 643 |
+
"step": 2200
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 1.0764392839864538,
|
| 647 |
+
"grad_norm": 24.645816802978516,
|
| 648 |
+
"learning_rate": 3.5627240143369176e-05,
|
| 649 |
+
"loss": 1.9655,
|
| 650 |
+
"step": 2225
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.0885341074020318,
|
| 654 |
+
"grad_norm": 0.16327662765979767,
|
| 655 |
+
"learning_rate": 3.5403225806451614e-05,
|
| 656 |
+
"loss": 1.2296,
|
| 657 |
+
"step": 2250
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 1.10062893081761,
|
| 661 |
+
"grad_norm": 3.4786908626556396,
|
| 662 |
+
"learning_rate": 3.517921146953405e-05,
|
| 663 |
+
"loss": 1.3128,
|
| 664 |
+
"step": 2275
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.1127237542331883,
|
| 668 |
+
"grad_norm": 0.18501241505146027,
|
| 669 |
+
"learning_rate": 3.495519713261649e-05,
|
| 670 |
+
"loss": 1.1883,
|
| 671 |
+
"step": 2300
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.1248185776487662,
|
| 675 |
+
"grad_norm": 0.35296130180358887,
|
| 676 |
+
"learning_rate": 3.473118279569892e-05,
|
| 677 |
+
"loss": 1.5325,
|
| 678 |
+
"step": 2325
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.1369134010643445,
|
| 682 |
+
"grad_norm": 0.24219243228435516,
|
| 683 |
+
"learning_rate": 3.450716845878136e-05,
|
| 684 |
+
"loss": 0.8073,
|
| 685 |
+
"step": 2350
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 1.1490082244799227,
|
| 689 |
+
"grad_norm": 0.16509315371513367,
|
| 690 |
+
"learning_rate": 3.42831541218638e-05,
|
| 691 |
+
"loss": 0.4453,
|
| 692 |
+
"step": 2375
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.1611030478955007,
|
| 696 |
+
"grad_norm": 25.64842987060547,
|
| 697 |
+
"learning_rate": 3.405913978494624e-05,
|
| 698 |
+
"loss": 1.8413,
|
| 699 |
+
"step": 2400
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 1.1731978713110789,
|
| 703 |
+
"grad_norm": 0.2987683117389679,
|
| 704 |
+
"learning_rate": 3.383512544802868e-05,
|
| 705 |
+
"loss": 1.3856,
|
| 706 |
+
"step": 2425
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.185292694726657,
|
| 710 |
+
"grad_norm": 0.1390484720468521,
|
| 711 |
+
"learning_rate": 3.3611111111111116e-05,
|
| 712 |
+
"loss": 1.0936,
|
| 713 |
+
"step": 2450
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.197387518142235,
|
| 717 |
+
"grad_norm": 152.2466278076172,
|
| 718 |
+
"learning_rate": 3.338709677419355e-05,
|
| 719 |
+
"loss": 1.29,
|
| 720 |
+
"step": 2475
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.2094823415578133,
|
| 724 |
+
"grad_norm": 0.5964694619178772,
|
| 725 |
+
"learning_rate": 3.316308243727599e-05,
|
| 726 |
+
"loss": 2.0098,
|
| 727 |
+
"step": 2500
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 1.2215771649733913,
|
| 731 |
+
"grad_norm": 0.7620899677276611,
|
| 732 |
+
"learning_rate": 3.2939068100358426e-05,
|
| 733 |
+
"loss": 0.9353,
|
| 734 |
+
"step": 2525
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.2336719883889695,
|
| 738 |
+
"grad_norm": 0.4900791347026825,
|
| 739 |
+
"learning_rate": 3.2715053763440864e-05,
|
| 740 |
+
"loss": 1.1481,
|
| 741 |
+
"step": 2550
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 1.2457668118045477,
|
| 745 |
+
"grad_norm": 0.8056408762931824,
|
| 746 |
+
"learning_rate": 3.24910394265233e-05,
|
| 747 |
+
"loss": 1.6924,
|
| 748 |
+
"step": 2575
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 1.2578616352201257,
|
| 752 |
+
"grad_norm": 0.34215426445007324,
|
| 753 |
+
"learning_rate": 3.2267025089605735e-05,
|
| 754 |
+
"loss": 1.5408,
|
| 755 |
+
"step": 2600
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.269956458635704,
|
| 759 |
+
"grad_norm": 0.32340100407600403,
|
| 760 |
+
"learning_rate": 3.204301075268817e-05,
|
| 761 |
+
"loss": 1.9354,
|
| 762 |
+
"step": 2625
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.282051282051282,
|
| 766 |
+
"grad_norm": 24.62626075744629,
|
| 767 |
+
"learning_rate": 3.181899641577061e-05,
|
| 768 |
+
"loss": 1.9551,
|
| 769 |
+
"step": 2650
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 1.2941461054668602,
|
| 773 |
+
"grad_norm": 1.058280348777771,
|
| 774 |
+
"learning_rate": 3.1594982078853044e-05,
|
| 775 |
+
"loss": 1.3507,
|
| 776 |
+
"step": 2675
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.3062409288824384,
|
| 780 |
+
"grad_norm": 1.5900629758834839,
|
| 781 |
+
"learning_rate": 3.137096774193549e-05,
|
| 782 |
+
"loss": 2.1607,
|
| 783 |
+
"step": 2700
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 1.3183357522980166,
|
| 787 |
+
"grad_norm": 24.93858528137207,
|
| 788 |
+
"learning_rate": 3.114695340501792e-05,
|
| 789 |
+
"loss": 1.328,
|
| 790 |
+
"step": 2725
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.3304305757135946,
|
| 794 |
+
"grad_norm": 24.877849578857422,
|
| 795 |
+
"learning_rate": 3.092293906810036e-05,
|
| 796 |
+
"loss": 1.6651,
|
| 797 |
+
"step": 2750
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 1.3425253991291728,
|
| 801 |
+
"grad_norm": 0.48590317368507385,
|
| 802 |
+
"learning_rate": 3.06989247311828e-05,
|
| 803 |
+
"loss": 0.6887,
|
| 804 |
+
"step": 2775
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 1.3546202225447508,
|
| 808 |
+
"grad_norm": 0.09108546376228333,
|
| 809 |
+
"learning_rate": 3.0474910394265234e-05,
|
| 810 |
+
"loss": 0.9513,
|
| 811 |
+
"step": 2800
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 1.366715045960329,
|
| 815 |
+
"grad_norm": 0.23763756453990936,
|
| 816 |
+
"learning_rate": 3.0250896057347672e-05,
|
| 817 |
+
"loss": 1.5187,
|
| 818 |
+
"step": 2825
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.3788098693759072,
|
| 822 |
+
"grad_norm": 0.2408967763185501,
|
| 823 |
+
"learning_rate": 3.002688172043011e-05,
|
| 824 |
+
"loss": 1.3683,
|
| 825 |
+
"step": 2850
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 1.3909046927914852,
|
| 829 |
+
"grad_norm": 2.7443923950195312,
|
| 830 |
+
"learning_rate": 2.9802867383512546e-05,
|
| 831 |
+
"loss": 1.7777,
|
| 832 |
+
"step": 2875
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 1.4029995162070634,
|
| 836 |
+
"grad_norm": 26.244659423828125,
|
| 837 |
+
"learning_rate": 2.9578853046594985e-05,
|
| 838 |
+
"loss": 2.1378,
|
| 839 |
+
"step": 2900
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 1.4150943396226414,
|
| 843 |
+
"grad_norm": 158.08462524414062,
|
| 844 |
+
"learning_rate": 2.9354838709677417e-05,
|
| 845 |
+
"loss": 1.28,
|
| 846 |
+
"step": 2925
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.4271891630382196,
|
| 850 |
+
"grad_norm": 27.04939842224121,
|
| 851 |
+
"learning_rate": 2.913082437275986e-05,
|
| 852 |
+
"loss": 0.8604,
|
| 853 |
+
"step": 2950
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 1.4392839864537978,
|
| 857 |
+
"grad_norm": 0.1699460744857788,
|
| 858 |
+
"learning_rate": 2.8906810035842297e-05,
|
| 859 |
+
"loss": 2.7105,
|
| 860 |
+
"step": 2975
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.4513788098693758,
|
| 864 |
+
"grad_norm": 23.923328399658203,
|
| 865 |
+
"learning_rate": 2.868279569892473e-05,
|
| 866 |
+
"loss": 1.3738,
|
| 867 |
+
"step": 3000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1.463473633284954,
|
| 871 |
+
"grad_norm": 23.95648765563965,
|
| 872 |
+
"learning_rate": 2.845878136200717e-05,
|
| 873 |
+
"loss": 1.4941,
|
| 874 |
+
"step": 3025
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1.475568456700532,
|
| 878 |
+
"grad_norm": 0.05188923701643944,
|
| 879 |
+
"learning_rate": 2.823476702508961e-05,
|
| 880 |
+
"loss": 0.9538,
|
| 881 |
+
"step": 3050
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.4876632801161103,
|
| 885 |
+
"grad_norm": 2.9171254634857178,
|
| 886 |
+
"learning_rate": 2.801075268817204e-05,
|
| 887 |
+
"loss": 1.7475,
|
| 888 |
+
"step": 3075
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1.4997581035316885,
|
| 892 |
+
"grad_norm": 6.571992874145508,
|
| 893 |
+
"learning_rate": 2.7786738351254484e-05,
|
| 894 |
+
"loss": 1.1107,
|
| 895 |
+
"step": 3100
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 1.5118529269472667,
|
| 899 |
+
"grad_norm": 26.21474838256836,
|
| 900 |
+
"learning_rate": 2.7562724014336922e-05,
|
| 901 |
+
"loss": 1.1843,
|
| 902 |
+
"step": 3125
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.5239477503628447,
|
| 906 |
+
"grad_norm": 0.16156063973903656,
|
| 907 |
+
"learning_rate": 2.7338709677419354e-05,
|
| 908 |
+
"loss": 1.5585,
|
| 909 |
+
"step": 3150
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1.5360425737784227,
|
| 913 |
+
"grad_norm": 0.13359041512012482,
|
| 914 |
+
"learning_rate": 2.7114695340501796e-05,
|
| 915 |
+
"loss": 1.1763,
|
| 916 |
+
"step": 3175
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1.548137397194001,
|
| 920 |
+
"grad_norm": 0.2726369798183441,
|
| 921 |
+
"learning_rate": 2.6890681003584228e-05,
|
| 922 |
+
"loss": 1.6273,
|
| 923 |
+
"step": 3200
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.5602322206095791,
|
| 927 |
+
"grad_norm": 0.41835281252861023,
|
| 928 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 929 |
+
"loss": 1.3093,
|
| 930 |
+
"step": 3225
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 1.5723270440251573,
|
| 934 |
+
"grad_norm": 0.2902648448944092,
|
| 935 |
+
"learning_rate": 2.6442652329749105e-05,
|
| 936 |
+
"loss": 0.8197,
|
| 937 |
+
"step": 3250
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 1.5844218674407353,
|
| 941 |
+
"grad_norm": 0.3506704866886139,
|
| 942 |
+
"learning_rate": 2.621863799283154e-05,
|
| 943 |
+
"loss": 0.8084,
|
| 944 |
+
"step": 3275
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.5965166908563135,
|
| 948 |
+
"grad_norm": 0.07854276150465012,
|
| 949 |
+
"learning_rate": 2.599462365591398e-05,
|
| 950 |
+
"loss": 0.6503,
|
| 951 |
+
"step": 3300
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 1.6086115142718915,
|
| 955 |
+
"grad_norm": 0.24219046533107758,
|
| 956 |
+
"learning_rate": 2.5770609318996418e-05,
|
| 957 |
+
"loss": 0.9614,
|
| 958 |
+
"step": 3325
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.6207063376874697,
|
| 962 |
+
"grad_norm": 26.107500076293945,
|
| 963 |
+
"learning_rate": 2.5546594982078853e-05,
|
| 964 |
+
"loss": 1.6199,
|
| 965 |
+
"step": 3350
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.632801161103048,
|
| 969 |
+
"grad_norm": 26.309776306152344,
|
| 970 |
+
"learning_rate": 2.532258064516129e-05,
|
| 971 |
+
"loss": 1.1816,
|
| 972 |
+
"step": 3375
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 1.6448959845186262,
|
| 976 |
+
"grad_norm": 0.2523309886455536,
|
| 977 |
+
"learning_rate": 2.5098566308243727e-05,
|
| 978 |
+
"loss": 1.1721,
|
| 979 |
+
"step": 3400
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 1.6569908079342042,
|
| 983 |
+
"grad_norm": 26.82816505432129,
|
| 984 |
+
"learning_rate": 2.4874551971326165e-05,
|
| 985 |
+
"loss": 0.696,
|
| 986 |
+
"step": 3425
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.6690856313497822,
|
| 990 |
+
"grad_norm": 0.15189094841480255,
|
| 991 |
+
"learning_rate": 2.46505376344086e-05,
|
| 992 |
+
"loss": 1.1234,
|
| 993 |
+
"step": 3450
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.6811804547653604,
|
| 997 |
+
"grad_norm": 0.19785544276237488,
|
| 998 |
+
"learning_rate": 2.4426523297491043e-05,
|
| 999 |
+
"loss": 1.6574,
|
| 1000 |
+
"step": 3475
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.6932752781809386,
|
| 1004 |
+
"grad_norm": 0.66933673620224,
|
| 1005 |
+
"learning_rate": 2.4202508960573478e-05,
|
| 1006 |
+
"loss": 2.0303,
|
| 1007 |
+
"step": 3500
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 1.7053701015965168,
|
| 1011 |
+
"grad_norm": 23.572277069091797,
|
| 1012 |
+
"learning_rate": 2.3978494623655913e-05,
|
| 1013 |
+
"loss": 2.2247,
|
| 1014 |
+
"step": 3525
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 1.7174649250120948,
|
| 1018 |
+
"grad_norm": 0.5169154405593872,
|
| 1019 |
+
"learning_rate": 2.3754480286738355e-05,
|
| 1020 |
+
"loss": 0.7378,
|
| 1021 |
+
"step": 3550
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 1.7295597484276728,
|
| 1025 |
+
"grad_norm": 0.28006285429000854,
|
| 1026 |
+
"learning_rate": 2.353046594982079e-05,
|
| 1027 |
+
"loss": 1.823,
|
| 1028 |
+
"step": 3575
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.741654571843251,
|
| 1032 |
+
"grad_norm": 0.47209933400154114,
|
| 1033 |
+
"learning_rate": 2.3306451612903226e-05,
|
| 1034 |
+
"loss": 2.0323,
|
| 1035 |
+
"step": 3600
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 1.7537493952588292,
|
| 1039 |
+
"grad_norm": 0.23533566296100616,
|
| 1040 |
+
"learning_rate": 2.3082437275985664e-05,
|
| 1041 |
+
"loss": 0.5909,
|
| 1042 |
+
"step": 3625
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 1.7658442186744074,
|
| 1046 |
+
"grad_norm": 0.17464753985404968,
|
| 1047 |
+
"learning_rate": 2.2858422939068103e-05,
|
| 1048 |
+
"loss": 1.0246,
|
| 1049 |
+
"step": 3650
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.7779390420899854,
|
| 1053 |
+
"grad_norm": 0.18633712828159332,
|
| 1054 |
+
"learning_rate": 2.2634408602150538e-05,
|
| 1055 |
+
"loss": 0.7297,
|
| 1056 |
+
"step": 3675
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 1.7900338655055636,
|
| 1060 |
+
"grad_norm": 0.19488303363323212,
|
| 1061 |
+
"learning_rate": 2.2410394265232977e-05,
|
| 1062 |
+
"loss": 1.0616,
|
| 1063 |
+
"step": 3700
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 1.8021286889211416,
|
| 1067 |
+
"grad_norm": 0.21784909069538116,
|
| 1068 |
+
"learning_rate": 2.2186379928315412e-05,
|
| 1069 |
+
"loss": 1.5585,
|
| 1070 |
+
"step": 3725
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.8142235123367199,
|
| 1074 |
+
"grad_norm": 0.2788207530975342,
|
| 1075 |
+
"learning_rate": 2.196236559139785e-05,
|
| 1076 |
+
"loss": 1.201,
|
| 1077 |
+
"step": 3750
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.826318335752298,
|
| 1081 |
+
"grad_norm": 24.099077224731445,
|
| 1082 |
+
"learning_rate": 2.173835125448029e-05,
|
| 1083 |
+
"loss": 2.076,
|
| 1084 |
+
"step": 3775
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 1.8384131591678763,
|
| 1088 |
+
"grad_norm": 0.18294040858745575,
|
| 1089 |
+
"learning_rate": 2.1514336917562725e-05,
|
| 1090 |
+
"loss": 1.384,
|
| 1091 |
+
"step": 3800
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.8505079825834543,
|
| 1095 |
+
"grad_norm": 24.9930477142334,
|
| 1096 |
+
"learning_rate": 2.129032258064516e-05,
|
| 1097 |
+
"loss": 1.0263,
|
| 1098 |
+
"step": 3825
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.8626028059990323,
|
| 1102 |
+
"grad_norm": 0.19692493975162506,
|
| 1103 |
+
"learning_rate": 2.1066308243727602e-05,
|
| 1104 |
+
"loss": 1.4213,
|
| 1105 |
+
"step": 3850
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.8746976294146105,
|
| 1109 |
+
"grad_norm": 0.32238996028900146,
|
| 1110 |
+
"learning_rate": 2.0842293906810037e-05,
|
| 1111 |
+
"loss": 0.9261,
|
| 1112 |
+
"step": 3875
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.8867924528301887,
|
| 1116 |
+
"grad_norm": 0.16958071291446686,
|
| 1117 |
+
"learning_rate": 2.0618279569892472e-05,
|
| 1118 |
+
"loss": 0.4504,
|
| 1119 |
+
"step": 3900
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.898887276245767,
|
| 1123 |
+
"grad_norm": 0.187217578291893,
|
| 1124 |
+
"learning_rate": 2.039426523297491e-05,
|
| 1125 |
+
"loss": 1.5949,
|
| 1126 |
+
"step": 3925
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.910982099661345,
|
| 1130 |
+
"grad_norm": 0.644129753112793,
|
| 1131 |
+
"learning_rate": 2.017025089605735e-05,
|
| 1132 |
+
"loss": 2.5123,
|
| 1133 |
+
"step": 3950
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.9230769230769231,
|
| 1137 |
+
"grad_norm": 0.2620396018028259,
|
| 1138 |
+
"learning_rate": 1.9946236559139785e-05,
|
| 1139 |
+
"loss": 1.0002,
|
| 1140 |
+
"step": 3975
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 1.9351717464925011,
|
| 1144 |
+
"grad_norm": 0.2608689069747925,
|
| 1145 |
+
"learning_rate": 1.9722222222222224e-05,
|
| 1146 |
+
"loss": 1.3737,
|
| 1147 |
+
"step": 4000
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 1.9472665699080793,
|
| 1151 |
+
"grad_norm": 24.944629669189453,
|
| 1152 |
+
"learning_rate": 1.9498207885304662e-05,
|
| 1153 |
+
"loss": 1.1471,
|
| 1154 |
+
"step": 4025
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.9593613933236576,
|
| 1158 |
+
"grad_norm": 23.796247482299805,
|
| 1159 |
+
"learning_rate": 1.9274193548387097e-05,
|
| 1160 |
+
"loss": 2.678,
|
| 1161 |
+
"step": 4050
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.9714562167392358,
|
| 1165 |
+
"grad_norm": 0.3903225362300873,
|
| 1166 |
+
"learning_rate": 1.9050179211469536e-05,
|
| 1167 |
+
"loss": 1.7349,
|
| 1168 |
+
"step": 4075
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 1.9835510401548138,
|
| 1172 |
+
"grad_norm": 23.73769760131836,
|
| 1173 |
+
"learning_rate": 1.882616487455197e-05,
|
| 1174 |
+
"loss": 1.1653,
|
| 1175 |
+
"step": 4100
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.9956458635703918,
|
| 1179 |
+
"grad_norm": 22.70951271057129,
|
| 1180 |
+
"learning_rate": 1.860215053763441e-05,
|
| 1181 |
+
"loss": 1.9917,
|
| 1182 |
+
"step": 4125
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.0,
|
| 1186 |
+
"eval_accuracy": 0.6731141199226306,
|
| 1187 |
+
"eval_f1_macro": 0.4615303615303616,
|
| 1188 |
+
"eval_f1_micro": 0.6731141199226306,
|
| 1189 |
+
"eval_f1_weighted": 0.6321285489564019,
|
| 1190 |
+
"eval_loss": 1.3909244537353516,
|
| 1191 |
+
"eval_precision_macro": 0.4655498128675815,
|
| 1192 |
+
"eval_precision_micro": 0.6731141199226306,
|
| 1193 |
+
"eval_precision_weighted": 0.6287401060434273,
|
| 1194 |
+
"eval_recall_macro": 0.4847830067753159,
|
| 1195 |
+
"eval_recall_micro": 0.6731141199226306,
|
| 1196 |
+
"eval_recall_weighted": 0.6731141199226306,
|
| 1197 |
+
"eval_runtime": 6650.8901,
|
| 1198 |
+
"eval_samples_per_second": 0.078,
|
| 1199 |
+
"eval_steps_per_second": 0.039,
|
| 1200 |
+
"step": 4134
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.00774068698597,
|
| 1204 |
+
"grad_norm": 0.9518312811851501,
|
| 1205 |
+
"learning_rate": 1.837813620071685e-05,
|
| 1206 |
+
"loss": 0.8235,
|
| 1207 |
+
"step": 4150
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.019835510401548,
|
| 1211 |
+
"grad_norm": 0.2822599411010742,
|
| 1212 |
+
"learning_rate": 1.8154121863799284e-05,
|
| 1213 |
+
"loss": 0.7381,
|
| 1214 |
+
"step": 4175
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.0319303338171264,
|
| 1218 |
+
"grad_norm": 0.2420988231897354,
|
| 1219 |
+
"learning_rate": 1.793010752688172e-05,
|
| 1220 |
+
"loss": 1.0998,
|
| 1221 |
+
"step": 4200
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.0440251572327046,
|
| 1225 |
+
"grad_norm": 0.23866595327854156,
|
| 1226 |
+
"learning_rate": 1.770609318996416e-05,
|
| 1227 |
+
"loss": 1.5128,
|
| 1228 |
+
"step": 4225
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.0561199806482824,
|
| 1232 |
+
"grad_norm": 0.20324940979480743,
|
| 1233 |
+
"learning_rate": 1.7482078853046596e-05,
|
| 1234 |
+
"loss": 0.6954,
|
| 1235 |
+
"step": 4250
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.0682148040638606,
|
| 1239 |
+
"grad_norm": 0.23257912695407867,
|
| 1240 |
+
"learning_rate": 1.725806451612903e-05,
|
| 1241 |
+
"loss": 2.1005,
|
| 1242 |
+
"step": 4275
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.080309627479439,
|
| 1246 |
+
"grad_norm": 0.28216153383255005,
|
| 1247 |
+
"learning_rate": 1.703405017921147e-05,
|
| 1248 |
+
"loss": 0.6184,
|
| 1249 |
+
"step": 4300
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.092404450895017,
|
| 1253 |
+
"grad_norm": 0.2500057518482208,
|
| 1254 |
+
"learning_rate": 1.681003584229391e-05,
|
| 1255 |
+
"loss": 1.82,
|
| 1256 |
+
"step": 4325
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.1044992743105952,
|
| 1260 |
+
"grad_norm": 0.442717045545578,
|
| 1261 |
+
"learning_rate": 1.6586021505376344e-05,
|
| 1262 |
+
"loss": 2.2307,
|
| 1263 |
+
"step": 4350
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.116594097726173,
|
| 1267 |
+
"grad_norm": 44.982303619384766,
|
| 1268 |
+
"learning_rate": 1.6362007168458783e-05,
|
| 1269 |
+
"loss": 1.16,
|
| 1270 |
+
"step": 4375
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.1286889211417512,
|
| 1274 |
+
"grad_norm": 34.185569763183594,
|
| 1275 |
+
"learning_rate": 1.6137992831541218e-05,
|
| 1276 |
+
"loss": 1.8138,
|
| 1277 |
+
"step": 4400
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.1407837445573294,
|
| 1281 |
+
"grad_norm": 0.5167245864868164,
|
| 1282 |
+
"learning_rate": 1.5913978494623657e-05,
|
| 1283 |
+
"loss": 1.5199,
|
| 1284 |
+
"step": 4425
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.1528785679729077,
|
| 1288 |
+
"grad_norm": 25.86376190185547,
|
| 1289 |
+
"learning_rate": 1.5689964157706095e-05,
|
| 1290 |
+
"loss": 1.1301,
|
| 1291 |
+
"step": 4450
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.164973391388486,
|
| 1295 |
+
"grad_norm": 0.7654205560684204,
|
| 1296 |
+
"learning_rate": 1.546594982078853e-05,
|
| 1297 |
+
"loss": 0.9802,
|
| 1298 |
+
"step": 4475
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.1770682148040637,
|
| 1302 |
+
"grad_norm": 0.22726771235466003,
|
| 1303 |
+
"learning_rate": 1.5241935483870967e-05,
|
| 1304 |
+
"loss": 1.3874,
|
| 1305 |
+
"step": 4500
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.189163038219642,
|
| 1309 |
+
"grad_norm": 0.3605280816555023,
|
| 1310 |
+
"learning_rate": 1.5017921146953406e-05,
|
| 1311 |
+
"loss": 1.9075,
|
| 1312 |
+
"step": 4525
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.20125786163522,
|
| 1316 |
+
"grad_norm": 0.28213346004486084,
|
| 1317 |
+
"learning_rate": 1.4793906810035843e-05,
|
| 1318 |
+
"loss": 1.0298,
|
| 1319 |
+
"step": 4550
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.2133526850507983,
|
| 1323 |
+
"grad_norm": 0.18000219762325287,
|
| 1324 |
+
"learning_rate": 1.456989247311828e-05,
|
| 1325 |
+
"loss": 1.5587,
|
| 1326 |
+
"step": 4575
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.2254475084663765,
|
| 1330 |
+
"grad_norm": 25.92272186279297,
|
| 1331 |
+
"learning_rate": 1.4345878136200718e-05,
|
| 1332 |
+
"loss": 1.1183,
|
| 1333 |
+
"step": 4600
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.2375423318819547,
|
| 1337 |
+
"grad_norm": 0.14410781860351562,
|
| 1338 |
+
"learning_rate": 1.4121863799283155e-05,
|
| 1339 |
+
"loss": 1.6175,
|
| 1340 |
+
"step": 4625
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.2496371552975325,
|
| 1344 |
+
"grad_norm": 0.16350312530994415,
|
| 1345 |
+
"learning_rate": 1.3897849462365592e-05,
|
| 1346 |
+
"loss": 1.2121,
|
| 1347 |
+
"step": 4650
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.2617319787131107,
|
| 1351 |
+
"grad_norm": 48.49122619628906,
|
| 1352 |
+
"learning_rate": 1.367383512544803e-05,
|
| 1353 |
+
"loss": 1.9733,
|
| 1354 |
+
"step": 4675
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.273826802128689,
|
| 1358 |
+
"grad_norm": 0.27995195984840393,
|
| 1359 |
+
"learning_rate": 1.3449820788530468e-05,
|
| 1360 |
+
"loss": 0.8415,
|
| 1361 |
+
"step": 4700
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.285921625544267,
|
| 1365 |
+
"grad_norm": 0.35172316431999207,
|
| 1366 |
+
"learning_rate": 1.3225806451612905e-05,
|
| 1367 |
+
"loss": 1.1809,
|
| 1368 |
+
"step": 4725
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.2980164489598454,
|
| 1372 |
+
"grad_norm": 0.33857014775276184,
|
| 1373 |
+
"learning_rate": 1.300179211469534e-05,
|
| 1374 |
+
"loss": 0.992,
|
| 1375 |
+
"step": 4750
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.310111272375423,
|
| 1379 |
+
"grad_norm": 0.9367401003837585,
|
| 1380 |
+
"learning_rate": 1.2777777777777777e-05,
|
| 1381 |
+
"loss": 1.4335,
|
| 1382 |
+
"step": 4775
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.3222060957910013,
|
| 1386 |
+
"grad_norm": 0.2820034921169281,
|
| 1387 |
+
"learning_rate": 1.2553763440860217e-05,
|
| 1388 |
+
"loss": 0.6884,
|
| 1389 |
+
"step": 4800
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.3343009192065796,
|
| 1393 |
+
"grad_norm": 0.37094831466674805,
|
| 1394 |
+
"learning_rate": 1.2329749103942653e-05,
|
| 1395 |
+
"loss": 1.672,
|
| 1396 |
+
"step": 4825
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.3463957426221578,
|
| 1400 |
+
"grad_norm": 0.32186359167099,
|
| 1401 |
+
"learning_rate": 1.210573476702509e-05,
|
| 1402 |
+
"loss": 1.2877,
|
| 1403 |
+
"step": 4850
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.358490566037736,
|
| 1407 |
+
"grad_norm": 54.59464645385742,
|
| 1408 |
+
"learning_rate": 1.1881720430107528e-05,
|
| 1409 |
+
"loss": 1.83,
|
| 1410 |
+
"step": 4875
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.370585389453314,
|
| 1414 |
+
"grad_norm": 0.36730891466140747,
|
| 1415 |
+
"learning_rate": 1.1657706093189963e-05,
|
| 1416 |
+
"loss": 1.6348,
|
| 1417 |
+
"step": 4900
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.382680212868892,
|
| 1421 |
+
"grad_norm": 0.23865163326263428,
|
| 1422 |
+
"learning_rate": 1.1433691756272402e-05,
|
| 1423 |
+
"loss": 1.2163,
|
| 1424 |
+
"step": 4925
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.39477503628447,
|
| 1428 |
+
"grad_norm": 27.761423110961914,
|
| 1429 |
+
"learning_rate": 1.1209677419354839e-05,
|
| 1430 |
+
"loss": 0.821,
|
| 1431 |
+
"step": 4950
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.4068698597000484,
|
| 1435 |
+
"grad_norm": 0.162750706076622,
|
| 1436 |
+
"learning_rate": 1.0985663082437276e-05,
|
| 1437 |
+
"loss": 0.7309,
|
| 1438 |
+
"step": 4975
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.4189646831156266,
|
| 1442 |
+
"grad_norm": 25.199321746826172,
|
| 1443 |
+
"learning_rate": 1.0761648745519713e-05,
|
| 1444 |
+
"loss": 0.7073,
|
| 1445 |
+
"step": 5000
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.431059506531205,
|
| 1449 |
+
"grad_norm": 0.1526053547859192,
|
| 1450 |
+
"learning_rate": 1.0537634408602151e-05,
|
| 1451 |
+
"loss": 1.2978,
|
| 1452 |
+
"step": 5025
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.4431543299467826,
|
| 1456 |
+
"grad_norm": 0.1368321031332016,
|
| 1457 |
+
"learning_rate": 1.0313620071684588e-05,
|
| 1458 |
+
"loss": 1.6805,
|
| 1459 |
+
"step": 5050
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.455249153362361,
|
| 1463 |
+
"grad_norm": 0.23694339394569397,
|
| 1464 |
+
"learning_rate": 1.0089605734767025e-05,
|
| 1465 |
+
"loss": 1.2977,
|
| 1466 |
+
"step": 5075
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.467343976777939,
|
| 1470 |
+
"grad_norm": 0.7356523871421814,
|
| 1471 |
+
"learning_rate": 9.865591397849464e-06,
|
| 1472 |
+
"loss": 1.8933,
|
| 1473 |
+
"step": 5100
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.4794388001935173,
|
| 1477 |
+
"grad_norm": 0.3766566812992096,
|
| 1478 |
+
"learning_rate": 9.6415770609319e-06,
|
| 1479 |
+
"loss": 1.2868,
|
| 1480 |
+
"step": 5125
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.4915336236090955,
|
| 1484 |
+
"grad_norm": 0.14354300498962402,
|
| 1485 |
+
"learning_rate": 9.417562724014338e-06,
|
| 1486 |
+
"loss": 1.5716,
|
| 1487 |
+
"step": 5150
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.5036284470246732,
|
| 1491 |
+
"grad_norm": 0.2801443934440613,
|
| 1492 |
+
"learning_rate": 9.193548387096775e-06,
|
| 1493 |
+
"loss": 1.1805,
|
| 1494 |
+
"step": 5175
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.5157232704402515,
|
| 1498 |
+
"grad_norm": 26.191604614257812,
|
| 1499 |
+
"learning_rate": 8.969534050179212e-06,
|
| 1500 |
+
"loss": 1.5799,
|
| 1501 |
+
"step": 5200
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.5278180938558297,
|
| 1505 |
+
"grad_norm": 0.29439038038253784,
|
| 1506 |
+
"learning_rate": 8.745519713261649e-06,
|
| 1507 |
+
"loss": 0.7503,
|
| 1508 |
+
"step": 5225
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.539912917271408,
|
| 1512 |
+
"grad_norm": 0.25986921787261963,
|
| 1513 |
+
"learning_rate": 8.521505376344087e-06,
|
| 1514 |
+
"loss": 1.5983,
|
| 1515 |
+
"step": 5250
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.552007740686986,
|
| 1519 |
+
"grad_norm": 0.37119609117507935,
|
| 1520 |
+
"learning_rate": 8.297491039426524e-06,
|
| 1521 |
+
"loss": 1.3133,
|
| 1522 |
+
"step": 5275
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.564102564102564,
|
| 1526 |
+
"grad_norm": 0.29802656173706055,
|
| 1527 |
+
"learning_rate": 8.073476702508961e-06,
|
| 1528 |
+
"loss": 1.2844,
|
| 1529 |
+
"step": 5300
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.576197387518142,
|
| 1533 |
+
"grad_norm": 0.22544977068901062,
|
| 1534 |
+
"learning_rate": 7.849462365591398e-06,
|
| 1535 |
+
"loss": 1.5217,
|
| 1536 |
+
"step": 5325
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.5882922109337203,
|
| 1540 |
+
"grad_norm": 0.2867962718009949,
|
| 1541 |
+
"learning_rate": 7.625448028673836e-06,
|
| 1542 |
+
"loss": 0.637,
|
| 1543 |
+
"step": 5350
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.6003870343492985,
|
| 1547 |
+
"grad_norm": 0.25557562708854675,
|
| 1548 |
+
"learning_rate": 7.401433691756272e-06,
|
| 1549 |
+
"loss": 1.1052,
|
| 1550 |
+
"step": 5375
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.6124818577648767,
|
| 1554 |
+
"grad_norm": 25.903554916381836,
|
| 1555 |
+
"learning_rate": 7.177419354838711e-06,
|
| 1556 |
+
"loss": 1.2246,
|
| 1557 |
+
"step": 5400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.6245766811804545,
|
| 1561 |
+
"grad_norm": 24.891738891601562,
|
| 1562 |
+
"learning_rate": 6.953405017921147e-06,
|
| 1563 |
+
"loss": 1.3789,
|
| 1564 |
+
"step": 5425
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.636671504596033,
|
| 1568 |
+
"grad_norm": 0.23630130290985107,
|
| 1569 |
+
"learning_rate": 6.7293906810035845e-06,
|
| 1570 |
+
"loss": 1.1069,
|
| 1571 |
+
"step": 5450
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.648766328011611,
|
| 1575 |
+
"grad_norm": 0.27478015422821045,
|
| 1576 |
+
"learning_rate": 6.5053763440860214e-06,
|
| 1577 |
+
"loss": 0.5998,
|
| 1578 |
+
"step": 5475
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.660861151427189,
|
| 1582 |
+
"grad_norm": 25.27565574645996,
|
| 1583 |
+
"learning_rate": 6.281362007168459e-06,
|
| 1584 |
+
"loss": 1.2044,
|
| 1585 |
+
"step": 5500
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.6729559748427674,
|
| 1589 |
+
"grad_norm": 25.68202781677246,
|
| 1590 |
+
"learning_rate": 6.057347670250896e-06,
|
| 1591 |
+
"loss": 1.0829,
|
| 1592 |
+
"step": 5525
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.6850507982583456,
|
| 1596 |
+
"grad_norm": 0.17175991833209991,
|
| 1597 |
+
"learning_rate": 5.833333333333334e-06,
|
| 1598 |
+
"loss": 1.8226,
|
| 1599 |
+
"step": 5550
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.697145621673924,
|
| 1603 |
+
"grad_norm": 0.19636030495166779,
|
| 1604 |
+
"learning_rate": 5.609318996415771e-06,
|
| 1605 |
+
"loss": 0.8312,
|
| 1606 |
+
"step": 5575
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.7092404450895016,
|
| 1610 |
+
"grad_norm": 0.16146661341190338,
|
| 1611 |
+
"learning_rate": 5.385304659498208e-06,
|
| 1612 |
+
"loss": 0.8461,
|
| 1613 |
+
"step": 5600
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.72133526850508,
|
| 1617 |
+
"grad_norm": 0.17099538445472717,
|
| 1618 |
+
"learning_rate": 5.161290322580646e-06,
|
| 1619 |
+
"loss": 0.667,
|
| 1620 |
+
"step": 5625
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.733430091920658,
|
| 1624 |
+
"grad_norm": 0.48382991552352905,
|
| 1625 |
+
"learning_rate": 4.9372759856630825e-06,
|
| 1626 |
+
"loss": 0.6893,
|
| 1627 |
+
"step": 5650
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.745524915336236,
|
| 1631 |
+
"grad_norm": 25.40802574157715,
|
| 1632 |
+
"learning_rate": 4.7132616487455195e-06,
|
| 1633 |
+
"loss": 1.7727,
|
| 1634 |
+
"step": 5675
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.7576197387518144,
|
| 1638 |
+
"grad_norm": 45.407470703125,
|
| 1639 |
+
"learning_rate": 4.489247311827957e-06,
|
| 1640 |
+
"loss": 1.7406,
|
| 1641 |
+
"step": 5700
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.769714562167392,
|
| 1645 |
+
"grad_norm": 0.16706956923007965,
|
| 1646 |
+
"learning_rate": 4.265232974910394e-06,
|
| 1647 |
+
"loss": 1.0508,
|
| 1648 |
+
"step": 5725
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.7818093855829704,
|
| 1652 |
+
"grad_norm": 0.14837704598903656,
|
| 1653 |
+
"learning_rate": 4.041218637992832e-06,
|
| 1654 |
+
"loss": 0.9122,
|
| 1655 |
+
"step": 5750
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.7939042089985486,
|
| 1659 |
+
"grad_norm": 0.19028626382350922,
|
| 1660 |
+
"learning_rate": 3.817204301075269e-06,
|
| 1661 |
+
"loss": 0.9756,
|
| 1662 |
+
"step": 5775
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.805999032414127,
|
| 1666 |
+
"grad_norm": 29.7191162109375,
|
| 1667 |
+
"learning_rate": 3.593189964157706e-06,
|
| 1668 |
+
"loss": 1.4132,
|
| 1669 |
+
"step": 5800
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.818093855829705,
|
| 1673 |
+
"grad_norm": 25.76336097717285,
|
| 1674 |
+
"learning_rate": 3.3691756272401432e-06,
|
| 1675 |
+
"loss": 1.9683,
|
| 1676 |
+
"step": 5825
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.830188679245283,
|
| 1680 |
+
"grad_norm": 0.4251428544521332,
|
| 1681 |
+
"learning_rate": 3.1451612903225806e-06,
|
| 1682 |
+
"loss": 0.6718,
|
| 1683 |
+
"step": 5850
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.842283502660861,
|
| 1687 |
+
"grad_norm": 0.26770126819610596,
|
| 1688 |
+
"learning_rate": 2.921146953405018e-06,
|
| 1689 |
+
"loss": 0.6973,
|
| 1690 |
+
"step": 5875
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.8543783260764393,
|
| 1694 |
+
"grad_norm": 0.7866289615631104,
|
| 1695 |
+
"learning_rate": 2.6971326164874553e-06,
|
| 1696 |
+
"loss": 0.7933,
|
| 1697 |
+
"step": 5900
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.8664731494920175,
|
| 1701 |
+
"grad_norm": 46.420658111572266,
|
| 1702 |
+
"learning_rate": 2.4731182795698927e-06,
|
| 1703 |
+
"loss": 2.0253,
|
| 1704 |
+
"step": 5925
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.8785679729075957,
|
| 1708 |
+
"grad_norm": 0.2426643818616867,
|
| 1709 |
+
"learning_rate": 2.2491039426523296e-06,
|
| 1710 |
+
"loss": 1.0285,
|
| 1711 |
+
"step": 5950
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.8906627963231735,
|
| 1715 |
+
"grad_norm": 34.91679382324219,
|
| 1716 |
+
"learning_rate": 2.025089605734767e-06,
|
| 1717 |
+
"loss": 1.3188,
|
| 1718 |
+
"step": 5975
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.9027576197387517,
|
| 1722 |
+
"grad_norm": 0.22075743973255157,
|
| 1723 |
+
"learning_rate": 1.8010752688172043e-06,
|
| 1724 |
+
"loss": 1.6779,
|
| 1725 |
+
"step": 6000
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.91485244315433,
|
| 1729 |
+
"grad_norm": 25.831626892089844,
|
| 1730 |
+
"learning_rate": 1.577060931899642e-06,
|
| 1731 |
+
"loss": 1.2721,
|
| 1732 |
+
"step": 6025
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.926947266569908,
|
| 1736 |
+
"grad_norm": 0.1505293846130371,
|
| 1737 |
+
"learning_rate": 1.3530465949820788e-06,
|
| 1738 |
+
"loss": 2.161,
|
| 1739 |
+
"step": 6050
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.9390420899854863,
|
| 1743 |
+
"grad_norm": 0.18153157830238342,
|
| 1744 |
+
"learning_rate": 1.1290322580645162e-06,
|
| 1745 |
+
"loss": 1.1947,
|
| 1746 |
+
"step": 6075
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.951136913401064,
|
| 1750 |
+
"grad_norm": 25.609390258789062,
|
| 1751 |
+
"learning_rate": 9.050179211469536e-07,
|
| 1752 |
+
"loss": 0.6296,
|
| 1753 |
+
"step": 6100
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.9632317368166423,
|
| 1757 |
+
"grad_norm": 0.3544562757015228,
|
| 1758 |
+
"learning_rate": 6.810035842293907e-07,
|
| 1759 |
+
"loss": 1.2337,
|
| 1760 |
+
"step": 6125
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.9753265602322205,
|
| 1764 |
+
"grad_norm": 25.893449783325195,
|
| 1765 |
+
"learning_rate": 4.5698924731182797e-07,
|
| 1766 |
+
"loss": 1.7994,
|
| 1767 |
+
"step": 6150
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.9874213836477987,
|
| 1771 |
+
"grad_norm": 1.0282210111618042,
|
| 1772 |
+
"learning_rate": 2.3297491039426527e-07,
|
| 1773 |
+
"loss": 1.2547,
|
| 1774 |
+
"step": 6175
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.999516207063377,
|
| 1778 |
+
"grad_norm": 0.2006077915430069,
|
| 1779 |
+
"learning_rate": 8.960573476702509e-09,
|
| 1780 |
+
"loss": 0.87,
|
| 1781 |
+
"step": 6200
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 3.0,
|
| 1785 |
+
"eval_accuracy": 0.7504835589941973,
|
| 1786 |
+
"eval_f1_macro": 0.5244016249451032,
|
| 1787 |
+
"eval_f1_micro": 0.7504835589941973,
|
| 1788 |
+
"eval_f1_weighted": 0.714761195760481,
|
| 1789 |
+
"eval_loss": 1.307055115699768,
|
| 1790 |
+
"eval_precision_macro": 0.5012229210342417,
|
| 1791 |
+
"eval_precision_micro": 0.7504835589941973,
|
| 1792 |
+
"eval_precision_weighted": 0.6860840439724423,
|
| 1793 |
+
"eval_recall_macro": 0.5532259049014222,
|
| 1794 |
+
"eval_recall_micro": 0.7504835589941973,
|
| 1795 |
+
"eval_recall_weighted": 0.7504835589941973,
|
| 1796 |
+
"eval_runtime": 6673.8059,
|
| 1797 |
+
"eval_samples_per_second": 0.077,
|
| 1798 |
+
"eval_steps_per_second": 0.039,
|
| 1799 |
+
"step": 6201
|
| 1800 |
+
}
|
| 1801 |
+
],
|
| 1802 |
+
"logging_steps": 25,
|
| 1803 |
+
"max_steps": 6201,
|
| 1804 |
+
"num_input_tokens_seen": 0,
|
| 1805 |
+
"num_train_epochs": 3,
|
| 1806 |
+
"save_steps": 500,
|
| 1807 |
+
"stateful_callbacks": {
|
| 1808 |
+
"EarlyStoppingCallback": {
|
| 1809 |
+
"args": {
|
| 1810 |
+
"early_stopping_patience": 5,
|
| 1811 |
+
"early_stopping_threshold": 0.01
|
| 1812 |
+
},
|
| 1813 |
+
"attributes": {
|
| 1814 |
+
"early_stopping_patience_counter": 0
|
| 1815 |
+
}
|
| 1816 |
+
},
|
| 1817 |
+
"TrainerControl": {
|
| 1818 |
+
"args": {
|
| 1819 |
+
"should_epoch_stop": false,
|
| 1820 |
+
"should_evaluate": false,
|
| 1821 |
+
"should_log": false,
|
| 1822 |
+
"should_save": true,
|
| 1823 |
+
"should_training_stop": true
|
| 1824 |
+
},
|
| 1825 |
+
"attributes": {}
|
| 1826 |
+
}
|
| 1827 |
+
},
|
| 1828 |
+
"total_flos": 1.6292664567668736e+16,
|
| 1829 |
+
"train_batch_size": 1,
|
| 1830 |
+
"trial_name": null,
|
| 1831 |
+
"trial_params": null
|
| 1832 |
+
}
|
checkpoint-6201/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cab6e4694c69fe840eda2a160ba8d0e45613204efc679c3084b6aa54b4cd418c
|
| 3 |
+
size 5240
|
config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "allenai/longformer-base-4096",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"LongformerForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_mode": "longformer",
|
| 8 |
+
"attention_probs_dropout_prob": 0.1,
|
| 9 |
+
"attention_window": [
|
| 10 |
+
512,
|
| 11 |
+
512,
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512
|
| 22 |
+
],
|
| 23 |
+
"bos_token_id": 0,
|
| 24 |
+
"eos_token_id": 2,
|
| 25 |
+
"gradient_checkpointing": false,
|
| 26 |
+
"hidden_act": "gelu",
|
| 27 |
+
"hidden_dropout_prob": 0.1,
|
| 28 |
+
"hidden_size": 768,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "negative",
|
| 31 |
+
"1": "neutral",
|
| 32 |
+
"2": "positive"
|
| 33 |
+
},
|
| 34 |
+
"ignore_attention_mask": false,
|
| 35 |
+
"initializer_range": 0.02,
|
| 36 |
+
"intermediate_size": 3072,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"negative": 0,
|
| 39 |
+
"neutral": 1,
|
| 40 |
+
"positive": 2
|
| 41 |
+
},
|
| 42 |
+
"layer_norm_eps": 1e-05,
|
| 43 |
+
"max_position_embeddings": 4098,
|
| 44 |
+
"model_type": "longformer",
|
| 45 |
+
"num_attention_heads": 12,
|
| 46 |
+
"num_hidden_layers": 12,
|
| 47 |
+
"onnx_export": false,
|
| 48 |
+
"pad_token_id": 1,
|
| 49 |
+
"problem_type": "single_label_classification",
|
| 50 |
+
"sep_token_id": 2,
|
| 51 |
+
"torch_dtype": "float32",
|
| 52 |
+
"transformers_version": "4.45.0",
|
| 53 |
+
"type_vocab_size": 1,
|
| 54 |
+
"vocab_size": 50265
|
| 55 |
+
}
|
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 sha256:322dae133d3ce467a820df6105d8c7973c40d913fb2ce1117801d5c7b57c139f
|
| 3 |
+
size 594681260
|
runs/Nov01_08-49-04_r-kafikani-longformer-va66dsc0-0e056-fjgin/events.out.tfevents.1730450948.r-kafikani-longformer-va66dsc0-0e056-fjgin.465.0
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:cfdf7e559ab1ee1a97c3a4c242ada4f77098d01eb43f82c4493442ef596eef4a
|
| 3 |
+
size 60472
|
runs/Nov01_08-49-04_r-kafikani-longformer-va66dsc0-0e056-fjgin/events.out.tfevents.1730870444.r-kafikani-longformer-va66dsc0-0e056-fjgin.465.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:232870b490734f0c789391c0d5c8c5bdeca5087a951488bf94a57efb56d76b58
|
| 3 |
+
size 921
|
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": 1000000000000000019884624838656,
|
| 52 |
+
"pad_token": "<pad>",
|
| 53 |
+
"sep_token": "</s>",
|
| 54 |
+
"tokenizer_class": "LongformerTokenizer",
|
| 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:cab6e4694c69fe840eda2a160ba8d0e45613204efc679c3084b6aa54b4cd418c
|
| 3 |
+
size 5240
|
training_params.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"data_path": "autotrain-iinjh-0wh75/autotrain-data",
|
| 3 |
+
"model": "allenai/longformer-base-4096",
|
| 4 |
+
"lr": 5e-05,
|
| 5 |
+
"epochs": 3,
|
| 6 |
+
"max_seq_length": 4096,
|
| 7 |
+
"batch_size": 1,
|
| 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-iinjh-0wh75",
|
| 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": "kafikani",
|
| 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
|
|
|