Upload expA_realonly — Macro F1=0.7497
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- .gitattributes +1 -0
- README.md +44 -0
- checkpoint-1150/config.json +93 -0
- checkpoint-1150/model.safetensors +3 -0
- checkpoint-1150/optimizer.pt +3 -0
- checkpoint-1150/rng_state.pth +3 -0
- checkpoint-1150/scheduler.pt +3 -0
- checkpoint-1150/tokenizer.json +0 -0
- checkpoint-1150/tokenizer_config.json +16 -0
- checkpoint-1150/trainer_state.json +648 -0
- checkpoint-1150/training_args.bin +3 -0
- checkpoint-1200/config.json +93 -0
- checkpoint-1200/model.safetensors +3 -0
- checkpoint-1200/optimizer.pt +3 -0
- checkpoint-1200/rng_state.pth +3 -0
- checkpoint-1200/scheduler.pt +3 -0
- checkpoint-1200/tokenizer.json +0 -0
- checkpoint-1200/tokenizer_config.json +16 -0
- checkpoint-1200/trainer_state.json +674 -0
- checkpoint-1200/training_args.bin +3 -0
- checkpoint-1250/config.json +93 -0
- checkpoint-1250/model.safetensors +3 -0
- checkpoint-1250/optimizer.pt +3 -0
- checkpoint-1250/rng_state.pth +3 -0
- checkpoint-1250/scheduler.pt +3 -0
- checkpoint-1250/tokenizer.json +0 -0
- checkpoint-1250/tokenizer_config.json +16 -0
- checkpoint-1250/trainer_state.json +700 -0
- checkpoint-1250/training_args.bin +3 -0
- checkpoint-1290/config.json +93 -0
- checkpoint-1290/model.safetensors +3 -0
- checkpoint-1290/optimizer.pt +3 -0
- checkpoint-1290/rng_state.pth +3 -0
- checkpoint-1290/scheduler.pt +3 -0
- checkpoint-1290/tokenizer.json +0 -0
- checkpoint-1290/tokenizer_config.json +16 -0
- checkpoint-1290/trainer_state.json +707 -0
- checkpoint-1290/training_args.bin +3 -0
- checkpoint-900/config.json +93 -0
- checkpoint-900/model.safetensors +3 -0
- checkpoint-900/optimizer.pt +3 -0
- checkpoint-900/rng_state.pth +3 -0
- checkpoint-900/scheduler.pt +3 -0
- checkpoint-900/tokenizer.json +0 -0
- checkpoint-900/tokenizer_config.json +16 -0
- checkpoint-900/trainer_state.json +518 -0
- checkpoint-900/training_args.bin +3 -0
- config.json +93 -0
- confusion_matrix.png +3 -0
- model.safetensors +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
confusion_matrix.png filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: en
|
| 3 |
+
tags:
|
| 4 |
+
- disfluency-detection
|
| 5 |
+
- token-classification
|
| 6 |
+
- modernbert
|
| 7 |
+
- speech-pathology
|
| 8 |
+
datasets:
|
| 9 |
+
- arielcerdap/disfluency-fluencybank
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# ModernBERT Disfluency Detection — Exp A (Real Only)
|
| 13 |
+
|
| 14 |
+
Fine-tuned from [answerdotai/ModernBERT-base](https://huggingface.co/answerdotai/ModernBERT-base) on **real-only** FluencyBank Timestamped data.
|
| 15 |
+
|
| 16 |
+
## Dataset
|
| 17 |
+
- Config: `real_only` de `arielcerdap/disfluency-fluencybank`
|
| 18 |
+
- Train: 2737 segmentos (100% reales)
|
| 19 |
+
- Val/Test: idénticos a Exp B para comparación directa
|
| 20 |
+
|
| 21 |
+
## Labels
|
| 22 |
+
O · FP (filled pause) · RP (repetition) · RV (revision) · PW (partial word)
|
| 23 |
+
|
| 24 |
+
## Test Results
|
| 25 |
+
| Label | P | R | F1 | Support |
|
| 26 |
+
|---|---|---|---|---|
|
| 27 |
+
| Label | P | R | F1 | Support |
|
| 28 |
+
|---|---|---|---|---|
|
| 29 |
+
| O | 0.9783 | 0.9860 | 0.9821 | 3704 |
|
| 30 |
+
| FP | 0.9888 | 1.0000 | 0.9944 | 176 |
|
| 31 |
+
| RP | 0.7784 | 0.8276 | 0.8022 | 174 |
|
| 32 |
+
| RV | 0.3425 | 0.2907 | 0.3145 | 86 |
|
| 33 |
+
| PW | 0.9510 | 0.8326 | 0.8879 | 233 |
|
| 34 |
+
|
| 35 |
+
**Macro F1 (4 disfluencias):** 0.7497
|
| 36 |
+
**Binary F1:** 0.8902
|
| 37 |
+
|
| 38 |
+
## Hyperparameters
|
| 39 |
+
- learning_rate: 5e-05
|
| 40 |
+
- epochs: 15
|
| 41 |
+
- warmup_steps: 191
|
| 42 |
+
- weight_decay: 0.1
|
| 43 |
+
- focal_loss_gamma: 3.0 (adaptive)
|
| 44 |
+
- class_weights: O=1.0, FP=3.0, RP=6.0, RV=20.0, PW=5.0
|
checkpoint-1150/config.json
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.4,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"gradient_checkpointing": false,
|
| 20 |
+
"hidden_activation": "gelu",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "O",
|
| 24 |
+
"1": "FP",
|
| 25 |
+
"2": "RP",
|
| 26 |
+
"3": "RV",
|
| 27 |
+
"4": "PW"
|
| 28 |
+
},
|
| 29 |
+
"initializer_cutoff_factor": 2.0,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 1152,
|
| 32 |
+
"label2id": {
|
| 33 |
+
"FP": 1,
|
| 34 |
+
"O": 0,
|
| 35 |
+
"PW": 4,
|
| 36 |
+
"RP": 2,
|
| 37 |
+
"RV": 3
|
| 38 |
+
},
|
| 39 |
+
"layer_norm_eps": 1e-05,
|
| 40 |
+
"layer_types": [
|
| 41 |
+
"full_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"full_attention"
|
| 63 |
+
],
|
| 64 |
+
"local_attention": 128,
|
| 65 |
+
"max_position_embeddings": 8192,
|
| 66 |
+
"mlp_bias": false,
|
| 67 |
+
"mlp_dropout": 0.0,
|
| 68 |
+
"model_type": "modernbert",
|
| 69 |
+
"norm_bias": false,
|
| 70 |
+
"norm_eps": 1e-05,
|
| 71 |
+
"num_attention_heads": 12,
|
| 72 |
+
"num_hidden_layers": 22,
|
| 73 |
+
"pad_token_id": 50283,
|
| 74 |
+
"position_embedding_type": "absolute",
|
| 75 |
+
"repad_logits_with_grad": false,
|
| 76 |
+
"rope_parameters": {
|
| 77 |
+
"full_attention": {
|
| 78 |
+
"rope_theta": 160000.0,
|
| 79 |
+
"rope_type": "default"
|
| 80 |
+
},
|
| 81 |
+
"sliding_attention": {
|
| 82 |
+
"rope_theta": 10000.0,
|
| 83 |
+
"rope_type": "default"
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"sep_token_id": 50282,
|
| 87 |
+
"sparse_pred_ignore_index": -100,
|
| 88 |
+
"sparse_prediction": false,
|
| 89 |
+
"tie_word_embeddings": true,
|
| 90 |
+
"transformers_version": "5.0.0",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"vocab_size": 50368
|
| 93 |
+
}
|
checkpoint-1150/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9fe599a9062d1719b36d7d693acb165b9ceafb108745f3aaa96e7c3fbd6d80df
|
| 3 |
+
size 598449012
|
checkpoint-1150/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58e2e22c08d1f829c9d785c3007330b102ec531b9b4e3820da84947c055d7a6e
|
| 3 |
+
size 535150859
|
checkpoint-1150/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b460537a835f7b49654e8de244b9d79b4b0a060dfef532490dc117a2d7554dc2
|
| 3 |
+
size 14709
|
checkpoint-1150/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ca20107812a9e91be5d7494a624d5425b7eb5a4c316b6395d44eeb4545bb878
|
| 3 |
+
size 1465
|
checkpoint-1150/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1150/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"mask_token": "[MASK]",
|
| 7 |
+
"model_input_names": [
|
| 8 |
+
"input_ids",
|
| 9 |
+
"attention_mask"
|
| 10 |
+
],
|
| 11 |
+
"model_max_length": 8192,
|
| 12 |
+
"pad_token": "[PAD]",
|
| 13 |
+
"sep_token": "[SEP]",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend",
|
| 15 |
+
"unk_token": "[UNK]"
|
| 16 |
+
}
|
checkpoint-1150/trainer_state.json
ADDED
|
@@ -0,0 +1,648 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 900,
|
| 3 |
+
"best_metric": 0.8228043143297381,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/expA_realonly/checkpoint-900",
|
| 5 |
+
"epoch": 13.372093023255815,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 1150,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011627906976744186,
|
| 14 |
+
"grad_norm": 4.9740142822265625,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.317932665348053,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.29069767441860467,
|
| 21 |
+
"grad_norm": 1.5566426515579224,
|
| 22 |
+
"learning_rate": 6.282722513089005e-06,
|
| 23 |
+
"loss": 0.3199571371078491,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5813953488372093,
|
| 28 |
+
"grad_norm": 1.6379677057266235,
|
| 29 |
+
"learning_rate": 1.2827225130890053e-05,
|
| 30 |
+
"loss": 0.24125898361206055,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5813953488372093,
|
| 35 |
+
"eval_accuracy": 0.8287126584998925,
|
| 36 |
+
"eval_f1": 0.08611410118406888,
|
| 37 |
+
"eval_loss": 0.10772959887981415,
|
| 38 |
+
"eval_precision": 0.15810276679841898,
|
| 39 |
+
"eval_recall": 0.05917159763313609,
|
| 40 |
+
"eval_runtime": 4.2273,
|
| 41 |
+
"eval_samples_per_second": 80.667,
|
| 42 |
+
"eval_steps_per_second": 1.419,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.872093023255814,
|
| 47 |
+
"grad_norm": 8.703161239624023,
|
| 48 |
+
"learning_rate": 1.93717277486911e-05,
|
| 49 |
+
"loss": 0.20946041107177735,
|
| 50 |
+
"step": 75
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.1627906976744187,
|
| 54 |
+
"grad_norm": 0.9212121963500977,
|
| 55 |
+
"learning_rate": 2.591623036649215e-05,
|
| 56 |
+
"loss": 0.19275867462158203,
|
| 57 |
+
"step": 100
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.1627906976744187,
|
| 61 |
+
"eval_accuracy": 0.7635933806146572,
|
| 62 |
+
"eval_f1": 0.3981945837512538,
|
| 63 |
+
"eval_loss": 0.07719173282384872,
|
| 64 |
+
"eval_precision": 0.30121396054628224,
|
| 65 |
+
"eval_recall": 0.5872781065088757,
|
| 66 |
+
"eval_runtime": 1.6016,
|
| 67 |
+
"eval_samples_per_second": 212.909,
|
| 68 |
+
"eval_steps_per_second": 3.746,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.4534883720930232,
|
| 73 |
+
"grad_norm": 1.46477210521698,
|
| 74 |
+
"learning_rate": 3.246073298429319e-05,
|
| 75 |
+
"loss": 0.14603865623474122,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.744186046511628,
|
| 80 |
+
"grad_norm": 3.729416608810425,
|
| 81 |
+
"learning_rate": 3.900523560209424e-05,
|
| 82 |
+
"loss": 0.13855588912963868,
|
| 83 |
+
"step": 150
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.744186046511628,
|
| 87 |
+
"eval_accuracy": 0.81130453470879,
|
| 88 |
+
"eval_f1": 0.46785521339816316,
|
| 89 |
+
"eval_loss": 0.06223325803875923,
|
| 90 |
+
"eval_precision": 0.36851063829787234,
|
| 91 |
+
"eval_recall": 0.6405325443786982,
|
| 92 |
+
"eval_runtime": 1.5906,
|
| 93 |
+
"eval_samples_per_second": 214.385,
|
| 94 |
+
"eval_steps_per_second": 3.772,
|
| 95 |
+
"step": 150
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0348837209302326,
|
| 99 |
+
"grad_norm": 1.2140589952468872,
|
| 100 |
+
"learning_rate": 4.554973821989529e-05,
|
| 101 |
+
"loss": 0.09624839782714843,
|
| 102 |
+
"step": 175
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.3255813953488373,
|
| 106 |
+
"grad_norm": 0.875888466835022,
|
| 107 |
+
"learning_rate": 4.999346304613061e-05,
|
| 108 |
+
"loss": 0.0821054744720459,
|
| 109 |
+
"step": 200
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3255813953488373,
|
| 113 |
+
"eval_accuracy": 0.7872340425531915,
|
| 114 |
+
"eval_f1": 0.4847619047619048,
|
| 115 |
+
"eval_loss": 0.04723217338323593,
|
| 116 |
+
"eval_precision": 0.3574438202247191,
|
| 117 |
+
"eval_recall": 0.7529585798816568,
|
| 118 |
+
"eval_runtime": 1.6255,
|
| 119 |
+
"eval_samples_per_second": 209.777,
|
| 120 |
+
"eval_steps_per_second": 3.691,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.616279069767442,
|
| 125 |
+
"grad_norm": 1.0979472398757935,
|
| 126 |
+
"learning_rate": 4.98888472605987e-05,
|
| 127 |
+
"loss": 0.08414460182189941,
|
| 128 |
+
"step": 225
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.9069767441860463,
|
| 132 |
+
"grad_norm": 0.49558770656585693,
|
| 133 |
+
"learning_rate": 4.965717280180432e-05,
|
| 134 |
+
"loss": 0.08369583129882813,
|
| 135 |
+
"step": 250
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.9069767441860463,
|
| 139 |
+
"eval_accuracy": 0.6211046636578551,
|
| 140 |
+
"eval_f1": 0.3483915126625599,
|
| 141 |
+
"eval_loss": 0.045315902680158615,
|
| 142 |
+
"eval_precision": 0.22662511130899377,
|
| 143 |
+
"eval_recall": 0.7529585798816568,
|
| 144 |
+
"eval_runtime": 1.6274,
|
| 145 |
+
"eval_samples_per_second": 209.539,
|
| 146 |
+
"eval_steps_per_second": 3.687,
|
| 147 |
+
"step": 250
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.197674418604651,
|
| 151 |
+
"grad_norm": 0.8114362955093384,
|
| 152 |
+
"learning_rate": 4.9299622378173246e-05,
|
| 153 |
+
"loss": 0.05607881069183349,
|
| 154 |
+
"step": 275
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.488372093023256,
|
| 158 |
+
"grad_norm": 0.6278374195098877,
|
| 159 |
+
"learning_rate": 4.881802130053548e-05,
|
| 160 |
+
"loss": 0.043469696044921874,
|
| 161 |
+
"step": 300
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 3.488372093023256,
|
| 165 |
+
"eval_accuracy": 0.8082957231893402,
|
| 166 |
+
"eval_f1": 0.5285784074255008,
|
| 167 |
+
"eval_loss": 0.04308881610631943,
|
| 168 |
+
"eval_precision": 0.39460247994164843,
|
| 169 |
+
"eval_recall": 0.8002958579881657,
|
| 170 |
+
"eval_runtime": 1.6193,
|
| 171 |
+
"eval_samples_per_second": 210.585,
|
| 172 |
+
"eval_steps_per_second": 3.705,
|
| 173 |
+
"step": 300
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.7790697674418605,
|
| 177 |
+
"grad_norm": 1.0219486951828003,
|
| 178 |
+
"learning_rate": 4.821482816383218e-05,
|
| 179 |
+
"loss": 0.047760300636291504,
|
| 180 |
+
"step": 325
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.069767441860465,
|
| 184 |
+
"grad_norm": 0.509274959564209,
|
| 185 |
+
"learning_rate": 4.749312229587908e-05,
|
| 186 |
+
"loss": 0.03177810907363891,
|
| 187 |
+
"step": 350
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 4.069767441860465,
|
| 191 |
+
"eval_accuracy": 0.9196217494089834,
|
| 192 |
+
"eval_f1": 0.7152317880794702,
|
| 193 |
+
"eval_loss": 0.04378104954957962,
|
| 194 |
+
"eval_precision": 0.6474820143884892,
|
| 195 |
+
"eval_recall": 0.7988165680473372,
|
| 196 |
+
"eval_runtime": 1.6177,
|
| 197 |
+
"eval_samples_per_second": 210.792,
|
| 198 |
+
"eval_steps_per_second": 3.709,
|
| 199 |
+
"step": 350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.3604651162790695,
|
| 203 |
+
"grad_norm": 0.47098585963249207,
|
| 204 |
+
"learning_rate": 4.6656588037260834e-05,
|
| 205 |
+
"loss": 0.020146708488464355,
|
| 206 |
+
"step": 375
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.651162790697675,
|
| 210 |
+
"grad_norm": 0.7534321546554565,
|
| 211 |
+
"learning_rate": 4.570949593260837e-05,
|
| 212 |
+
"loss": 0.018921481370925905,
|
| 213 |
+
"step": 400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.651162790697675,
|
| 217 |
+
"eval_accuracy": 0.9097356544165055,
|
| 218 |
+
"eval_f1": 0.6987341772151898,
|
| 219 |
+
"eval_loss": 0.05498046055436134,
|
| 220 |
+
"eval_precision": 0.6106194690265486,
|
| 221 |
+
"eval_recall": 0.8165680473372781,
|
| 222 |
+
"eval_runtime": 1.6157,
|
| 223 |
+
"eval_samples_per_second": 211.05,
|
| 224 |
+
"eval_steps_per_second": 3.713,
|
| 225 |
+
"step": 400
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.941860465116279,
|
| 229 |
+
"grad_norm": 0.7499828934669495,
|
| 230 |
+
"learning_rate": 4.465668092927841e-05,
|
| 231 |
+
"loss": 0.01998784065246582,
|
| 232 |
+
"step": 425
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 5.232558139534884,
|
| 236 |
+
"grad_norm": 0.24809418618679047,
|
| 237 |
+
"learning_rate": 4.350351769473198e-05,
|
| 238 |
+
"loss": 0.01218403697013855,
|
| 239 |
+
"step": 450
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 5.232558139534884,
|
| 243 |
+
"eval_accuracy": 0.9084461637653127,
|
| 244 |
+
"eval_f1": 0.6831746031746032,
|
| 245 |
+
"eval_loss": 0.05535344406962395,
|
| 246 |
+
"eval_precision": 0.5984427141268076,
|
| 247 |
+
"eval_recall": 0.7958579881656804,
|
| 248 |
+
"eval_runtime": 1.6123,
|
| 249 |
+
"eval_samples_per_second": 211.498,
|
| 250 |
+
"eval_steps_per_second": 3.721,
|
| 251 |
+
"step": 450
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 5.523255813953488,
|
| 255 |
+
"grad_norm": 0.3144132196903229,
|
| 256 |
+
"learning_rate": 4.2255893178617745e-05,
|
| 257 |
+
"loss": 0.012306832075119019,
|
| 258 |
+
"step": 475
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 5.813953488372093,
|
| 262 |
+
"grad_norm": 0.1607646942138672,
|
| 263 |
+
"learning_rate": 4.092017655963198e-05,
|
| 264 |
+
"loss": 0.010539753437042236,
|
| 265 |
+
"step": 500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 5.813953488372093,
|
| 269 |
+
"eval_accuracy": 0.9527186761229315,
|
| 270 |
+
"eval_f1": 0.8059040590405904,
|
| 271 |
+
"eval_loss": 0.07191809266805649,
|
| 272 |
+
"eval_precision": 0.8041237113402062,
|
| 273 |
+
"eval_recall": 0.8076923076923077,
|
| 274 |
+
"eval_runtime": 1.6569,
|
| 275 |
+
"eval_samples_per_second": 205.806,
|
| 276 |
+
"eval_steps_per_second": 3.621,
|
| 277 |
+
"step": 500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.104651162790698,
|
| 281 |
+
"grad_norm": 0.19801200926303864,
|
| 282 |
+
"learning_rate": 3.9503186730577816e-05,
|
| 283 |
+
"loss": 0.008691226243972778,
|
| 284 |
+
"step": 525
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.395348837209302,
|
| 288 |
+
"grad_norm": 0.4177681505680084,
|
| 289 |
+
"learning_rate": 3.801215748761385e-05,
|
| 290 |
+
"loss": 0.004750225245952606,
|
| 291 |
+
"step": 550
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 6.395348837209302,
|
| 295 |
+
"eval_accuracy": 0.9458413926499033,
|
| 296 |
+
"eval_f1": 0.7863866763215062,
|
| 297 |
+
"eval_loss": 0.07378304749727249,
|
| 298 |
+
"eval_precision": 0.7702127659574468,
|
| 299 |
+
"eval_recall": 0.8032544378698225,
|
| 300 |
+
"eval_runtime": 1.6214,
|
| 301 |
+
"eval_samples_per_second": 210.313,
|
| 302 |
+
"eval_steps_per_second": 3.701,
|
| 303 |
+
"step": 550
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.686046511627907,
|
| 307 |
+
"grad_norm": 0.30753573775291443,
|
| 308 |
+
"learning_rate": 3.645470060140278e-05,
|
| 309 |
+
"loss": 0.005743271708488464,
|
| 310 |
+
"step": 575
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.976744186046512,
|
| 314 |
+
"grad_norm": 0.13237181305885315,
|
| 315 |
+
"learning_rate": 3.4838766958683304e-05,
|
| 316 |
+
"loss": 0.006215065121650696,
|
| 317 |
+
"step": 600
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.976744186046512,
|
| 321 |
+
"eval_accuracy": 0.9447668171072426,
|
| 322 |
+
"eval_f1": 0.7789934354485777,
|
| 323 |
+
"eval_loss": 0.062308527529239655,
|
| 324 |
+
"eval_precision": 0.7683453237410072,
|
| 325 |
+
"eval_recall": 0.7899408284023669,
|
| 326 |
+
"eval_runtime": 1.6197,
|
| 327 |
+
"eval_samples_per_second": 210.528,
|
| 328 |
+
"eval_steps_per_second": 3.704,
|
| 329 |
+
"step": 600
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 7.267441860465116,
|
| 333 |
+
"grad_norm": 0.023809686303138733,
|
| 334 |
+
"learning_rate": 3.317260597263932e-05,
|
| 335 |
+
"loss": 0.002318141460418701,
|
| 336 |
+
"step": 625
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.558139534883721,
|
| 340 |
+
"grad_norm": 0.03782212734222412,
|
| 341 |
+
"learning_rate": 3.146472346927845e-05,
|
| 342 |
+
"loss": 0.003790220320224762,
|
| 343 |
+
"step": 650
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 7.558139534883721,
|
| 347 |
+
"eval_accuracy": 0.9529335912314636,
|
| 348 |
+
"eval_f1": 0.8062827225130891,
|
| 349 |
+
"eval_loss": 0.07683796435594559,
|
| 350 |
+
"eval_precision": 0.8154311649016641,
|
| 351 |
+
"eval_recall": 0.7973372781065089,
|
| 352 |
+
"eval_runtime": 1.6235,
|
| 353 |
+
"eval_samples_per_second": 210.043,
|
| 354 |
+
"eval_steps_per_second": 3.696,
|
| 355 |
+
"step": 650
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 7.848837209302325,
|
| 359 |
+
"grad_norm": 0.037971507757902145,
|
| 360 |
+
"learning_rate": 2.9723838264811545e-05,
|
| 361 |
+
"loss": 0.0018679146468639374,
|
| 362 |
+
"step": 675
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 8.13953488372093,
|
| 366 |
+
"grad_norm": 0.6083823442459106,
|
| 367 |
+
"learning_rate": 2.7958837655707814e-05,
|
| 368 |
+
"loss": 0.0019237904250621796,
|
| 369 |
+
"step": 700
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 8.13953488372093,
|
| 373 |
+
"eval_accuracy": 0.9565871480765098,
|
| 374 |
+
"eval_f1": 0.8152091254752851,
|
| 375 |
+
"eval_loss": 0.08732757717370987,
|
| 376 |
+
"eval_precision": 0.838810641627543,
|
| 377 |
+
"eval_recall": 0.7928994082840237,
|
| 378 |
+
"eval_runtime": 1.6124,
|
| 379 |
+
"eval_samples_per_second": 211.484,
|
| 380 |
+
"eval_steps_per_second": 3.721,
|
| 381 |
+
"step": 700
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.430232558139535,
|
| 385 |
+
"grad_norm": 0.022007286548614502,
|
| 386 |
+
"learning_rate": 2.6178732048650694e-05,
|
| 387 |
+
"loss": 0.0010710173845291137,
|
| 388 |
+
"step": 725
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 8.720930232558139,
|
| 392 |
+
"grad_norm": 0.09439104795455933,
|
| 393 |
+
"learning_rate": 2.4392608962010652e-05,
|
| 394 |
+
"loss": 0.0006907544285058976,
|
| 395 |
+
"step": 750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.720930232558139,
|
| 399 |
+
"eval_accuracy": 0.9449817322157748,
|
| 400 |
+
"eval_f1": 0.7829848594087959,
|
| 401 |
+
"eval_loss": 0.08308640867471695,
|
| 402 |
+
"eval_precision": 0.7637130801687764,
|
| 403 |
+
"eval_recall": 0.8032544378698225,
|
| 404 |
+
"eval_runtime": 1.6359,
|
| 405 |
+
"eval_samples_per_second": 208.445,
|
| 406 |
+
"eval_steps_per_second": 3.668,
|
| 407 |
+
"step": 750
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 9.011627906976743,
|
| 411 |
+
"grad_norm": 0.018290311098098755,
|
| 412 |
+
"learning_rate": 2.2609586633659256e-05,
|
| 413 |
+
"loss": 0.0005397907271981239,
|
| 414 |
+
"step": 775
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 9.30232558139535,
|
| 418 |
+
"grad_norm": 0.002839206252247095,
|
| 419 |
+
"learning_rate": 2.0838767471959014e-05,
|
| 420 |
+
"loss": 0.0002632186934351921,
|
| 421 |
+
"step": 800
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 9.30232558139535,
|
| 425 |
+
"eval_accuracy": 0.9544379969911885,
|
| 426 |
+
"eval_f1": 0.8081880212282032,
|
| 427 |
+
"eval_loss": 0.09754534065723419,
|
| 428 |
+
"eval_precision": 0.8289269051321928,
|
| 429 |
+
"eval_recall": 0.7884615384615384,
|
| 430 |
+
"eval_runtime": 1.6188,
|
| 431 |
+
"eval_samples_per_second": 210.653,
|
| 432 |
+
"eval_steps_per_second": 3.706,
|
| 433 |
+
"step": 800
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 9.593023255813954,
|
| 437 |
+
"grad_norm": 0.04134448245167732,
|
| 438 |
+
"learning_rate": 1.9089191587563417e-05,
|
| 439 |
+
"loss": 0.0005076020956039428,
|
| 440 |
+
"step": 825
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 9.883720930232558,
|
| 444 |
+
"grad_norm": 0.08968573063611984,
|
| 445 |
+
"learning_rate": 1.7369790643249573e-05,
|
| 446 |
+
"loss": 0.0005874237418174743,
|
| 447 |
+
"step": 850
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 9.883720930232558,
|
| 451 |
+
"eval_accuracy": 0.9548678272082527,
|
| 452 |
+
"eval_f1": 0.8093797276853252,
|
| 453 |
+
"eval_loss": 0.09803693741559982,
|
| 454 |
+
"eval_precision": 0.8281733746130031,
|
| 455 |
+
"eval_recall": 0.7914201183431953,
|
| 456 |
+
"eval_runtime": 1.6209,
|
| 457 |
+
"eval_samples_per_second": 210.379,
|
| 458 |
+
"eval_steps_per_second": 3.702,
|
| 459 |
+
"step": 850
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 10.174418604651162,
|
| 463 |
+
"grad_norm": 0.02889181859791279,
|
| 464 |
+
"learning_rate": 1.5689342257382207e-05,
|
| 465 |
+
"loss": 0.0004838050529360771,
|
| 466 |
+
"step": 875
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.465116279069768,
|
| 470 |
+
"grad_norm": 0.020970497280359268,
|
| 471 |
+
"learning_rate": 1.4056425193781048e-05,
|
| 472 |
+
"loss": 0.0001906752586364746,
|
| 473 |
+
"step": 900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 10.465116279069768,
|
| 477 |
+
"eval_accuracy": 0.9587362991618311,
|
| 478 |
+
"eval_f1": 0.8228043143297381,
|
| 479 |
+
"eval_loss": 0.10382715612649918,
|
| 480 |
+
"eval_precision": 0.8585209003215434,
|
| 481 |
+
"eval_recall": 0.7899408284023669,
|
| 482 |
+
"eval_runtime": 1.6341,
|
| 483 |
+
"eval_samples_per_second": 208.675,
|
| 484 |
+
"eval_steps_per_second": 3.672,
|
| 485 |
+
"step": 900
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 10.755813953488373,
|
| 489 |
+
"grad_norm": 0.028520096093416214,
|
| 490 |
+
"learning_rate": 1.2479375566749694e-05,
|
| 491 |
+
"loss": 0.00012492083944380284,
|
| 492 |
+
"step": 925
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 11.046511627906977,
|
| 496 |
+
"grad_norm": 0.03838531672954559,
|
| 497 |
+
"learning_rate": 1.0966244284840926e-05,
|
| 498 |
+
"loss": 0.00028255857527256014,
|
| 499 |
+
"step": 950
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 11.046511627906977,
|
| 503 |
+
"eval_accuracy": 0.9580915538362347,
|
| 504 |
+
"eval_f1": 0.819369715603382,
|
| 505 |
+
"eval_loss": 0.10540532320737839,
|
| 506 |
+
"eval_precision": 0.8528,
|
| 507 |
+
"eval_recall": 0.7884615384615384,
|
| 508 |
+
"eval_runtime": 1.6495,
|
| 509 |
+
"eval_samples_per_second": 206.73,
|
| 510 |
+
"eval_steps_per_second": 3.637,
|
| 511 |
+
"step": 950
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 11.337209302325581,
|
| 515 |
+
"grad_norm": 0.05398673936724663,
|
| 516 |
+
"learning_rate": 9.524755950610204e-06,
|
| 517 |
+
"loss": 0.00014415644109249114,
|
| 518 |
+
"step": 975
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 11.627906976744185,
|
| 522 |
+
"grad_norm": 0.026561638340353966,
|
| 523 |
+
"learning_rate": 8.162269426175681e-06,
|
| 524 |
+
"loss": 9.150288999080658e-05,
|
| 525 |
+
"step": 1000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 11.627906976744185,
|
| 529 |
+
"eval_accuracy": 0.9574468085106383,
|
| 530 |
+
"eval_f1": 0.8180428134556575,
|
| 531 |
+
"eval_loss": 0.10544126480817795,
|
| 532 |
+
"eval_precision": 0.8465189873417721,
|
| 533 |
+
"eval_recall": 0.7914201183431953,
|
| 534 |
+
"eval_runtime": 1.6274,
|
| 535 |
+
"eval_samples_per_second": 209.539,
|
| 536 |
+
"eval_steps_per_second": 3.687,
|
| 537 |
+
"step": 1000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 11.918604651162791,
|
| 541 |
+
"grad_norm": 0.010734460316598415,
|
| 542 |
+
"learning_rate": 6.885740265899526e-06,
|
| 543 |
+
"loss": 4.033456556499005e-05,
|
| 544 |
+
"step": 1025
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 12.209302325581396,
|
| 548 |
+
"grad_norm": 0.02024712599813938,
|
| 549 |
+
"learning_rate": 5.701685207973243e-06,
|
| 550 |
+
"loss": 6.52561942115426e-05,
|
| 551 |
+
"step": 1050
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 12.209302325581396,
|
| 555 |
+
"eval_accuracy": 0.9576617236191705,
|
| 556 |
+
"eval_f1": 0.8190184049079755,
|
| 557 |
+
"eval_loss": 0.10666228085756302,
|
| 558 |
+
"eval_precision": 0.8503184713375797,
|
| 559 |
+
"eval_recall": 0.7899408284023669,
|
| 560 |
+
"eval_runtime": 1.6269,
|
| 561 |
+
"eval_samples_per_second": 209.598,
|
| 562 |
+
"eval_steps_per_second": 3.688,
|
| 563 |
+
"step": 1050
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 12.5,
|
| 567 |
+
"grad_norm": 0.034469809383153915,
|
| 568 |
+
"learning_rate": 4.616148906179082e-06,
|
| 569 |
+
"loss": 7.551711052656174e-05,
|
| 570 |
+
"step": 1075
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 12.790697674418604,
|
| 574 |
+
"grad_norm": 0.004050907213240862,
|
| 575 |
+
"learning_rate": 3.6346730716634025e-06,
|
| 576 |
+
"loss": 7.49318953603506e-05,
|
| 577 |
+
"step": 1100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 12.790697674418604,
|
| 581 |
+
"eval_accuracy": 0.9565871480765098,
|
| 582 |
+
"eval_f1": 0.8146453089244853,
|
| 583 |
+
"eval_loss": 0.10619472712278366,
|
| 584 |
+
"eval_precision": 0.8409448818897638,
|
| 585 |
+
"eval_recall": 0.7899408284023669,
|
| 586 |
+
"eval_runtime": 1.6196,
|
| 587 |
+
"eval_samples_per_second": 210.546,
|
| 588 |
+
"eval_steps_per_second": 3.705,
|
| 589 |
+
"step": 1100
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 13.081395348837209,
|
| 593 |
+
"grad_norm": 0.11917345970869064,
|
| 594 |
+
"learning_rate": 2.7622681822545764e-06,
|
| 595 |
+
"loss": 8.815748617053032e-05,
|
| 596 |
+
"step": 1125
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 13.372093023255815,
|
| 600 |
+
"grad_norm": 0.02682509273290634,
|
| 601 |
+
"learning_rate": 2.0033879037506004e-06,
|
| 602 |
+
"loss": 6.015756167471409e-05,
|
| 603 |
+
"step": 1150
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 13.372093023255815,
|
| 607 |
+
"eval_accuracy": 0.9583064689447668,
|
| 608 |
+
"eval_f1": 0.8205521472392638,
|
| 609 |
+
"eval_loss": 0.10740578919649124,
|
| 610 |
+
"eval_precision": 0.8519108280254777,
|
| 611 |
+
"eval_recall": 0.7914201183431953,
|
| 612 |
+
"eval_runtime": 1.6217,
|
| 613 |
+
"eval_samples_per_second": 210.273,
|
| 614 |
+
"eval_steps_per_second": 3.7,
|
| 615 |
+
"step": 1150
|
| 616 |
+
}
|
| 617 |
+
],
|
| 618 |
+
"logging_steps": 25,
|
| 619 |
+
"max_steps": 1290,
|
| 620 |
+
"num_input_tokens_seen": 0,
|
| 621 |
+
"num_train_epochs": 15,
|
| 622 |
+
"save_steps": 50,
|
| 623 |
+
"stateful_callbacks": {
|
| 624 |
+
"EarlyStoppingCallback": {
|
| 625 |
+
"args": {
|
| 626 |
+
"early_stopping_patience": 10,
|
| 627 |
+
"early_stopping_threshold": 0.0
|
| 628 |
+
},
|
| 629 |
+
"attributes": {
|
| 630 |
+
"early_stopping_patience_counter": 5
|
| 631 |
+
}
|
| 632 |
+
},
|
| 633 |
+
"TrainerControl": {
|
| 634 |
+
"args": {
|
| 635 |
+
"should_epoch_stop": false,
|
| 636 |
+
"should_evaluate": false,
|
| 637 |
+
"should_log": false,
|
| 638 |
+
"should_save": true,
|
| 639 |
+
"should_training_stop": false
|
| 640 |
+
},
|
| 641 |
+
"attributes": {}
|
| 642 |
+
}
|
| 643 |
+
},
|
| 644 |
+
"total_flos": 6236850188428800.0,
|
| 645 |
+
"train_batch_size": 16,
|
| 646 |
+
"trial_name": null,
|
| 647 |
+
"trial_params": null
|
| 648 |
+
}
|
checkpoint-1150/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5071855a54d36018bc5c30ca34d9fef42b69fb3deda057af4ffc70193237d5c4
|
| 3 |
+
size 5201
|
checkpoint-1200/config.json
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.4,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"gradient_checkpointing": false,
|
| 20 |
+
"hidden_activation": "gelu",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "O",
|
| 24 |
+
"1": "FP",
|
| 25 |
+
"2": "RP",
|
| 26 |
+
"3": "RV",
|
| 27 |
+
"4": "PW"
|
| 28 |
+
},
|
| 29 |
+
"initializer_cutoff_factor": 2.0,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 1152,
|
| 32 |
+
"label2id": {
|
| 33 |
+
"FP": 1,
|
| 34 |
+
"O": 0,
|
| 35 |
+
"PW": 4,
|
| 36 |
+
"RP": 2,
|
| 37 |
+
"RV": 3
|
| 38 |
+
},
|
| 39 |
+
"layer_norm_eps": 1e-05,
|
| 40 |
+
"layer_types": [
|
| 41 |
+
"full_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"full_attention"
|
| 63 |
+
],
|
| 64 |
+
"local_attention": 128,
|
| 65 |
+
"max_position_embeddings": 8192,
|
| 66 |
+
"mlp_bias": false,
|
| 67 |
+
"mlp_dropout": 0.0,
|
| 68 |
+
"model_type": "modernbert",
|
| 69 |
+
"norm_bias": false,
|
| 70 |
+
"norm_eps": 1e-05,
|
| 71 |
+
"num_attention_heads": 12,
|
| 72 |
+
"num_hidden_layers": 22,
|
| 73 |
+
"pad_token_id": 50283,
|
| 74 |
+
"position_embedding_type": "absolute",
|
| 75 |
+
"repad_logits_with_grad": false,
|
| 76 |
+
"rope_parameters": {
|
| 77 |
+
"full_attention": {
|
| 78 |
+
"rope_theta": 160000.0,
|
| 79 |
+
"rope_type": "default"
|
| 80 |
+
},
|
| 81 |
+
"sliding_attention": {
|
| 82 |
+
"rope_theta": 10000.0,
|
| 83 |
+
"rope_type": "default"
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"sep_token_id": 50282,
|
| 87 |
+
"sparse_pred_ignore_index": -100,
|
| 88 |
+
"sparse_prediction": false,
|
| 89 |
+
"tie_word_embeddings": true,
|
| 90 |
+
"transformers_version": "5.0.0",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"vocab_size": 50368
|
| 93 |
+
}
|
checkpoint-1200/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:192d234fce88e3c0a53584d9886186f0bc98018d8baab24a7589d863a55e18bf
|
| 3 |
+
size 598449012
|
checkpoint-1200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1632f2d45550d119f075f652bd63b324e7023adc6af0ab5f7309173a376c872
|
| 3 |
+
size 535150859
|
checkpoint-1200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3346e012b94a0e3b8710b5afadeb0801fec62951ffc4a696ed752d55bf580c93
|
| 3 |
+
size 14709
|
checkpoint-1200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:875cbc43a94946c9471d4378d3a7262ff116a77879c41bf7197e6426eff5ed7d
|
| 3 |
+
size 1465
|
checkpoint-1200/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1200/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"mask_token": "[MASK]",
|
| 7 |
+
"model_input_names": [
|
| 8 |
+
"input_ids",
|
| 9 |
+
"attention_mask"
|
| 10 |
+
],
|
| 11 |
+
"model_max_length": 8192,
|
| 12 |
+
"pad_token": "[PAD]",
|
| 13 |
+
"sep_token": "[SEP]",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend",
|
| 15 |
+
"unk_token": "[UNK]"
|
| 16 |
+
}
|
checkpoint-1200/trainer_state.json
ADDED
|
@@ -0,0 +1,674 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 900,
|
| 3 |
+
"best_metric": 0.8228043143297381,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/expA_realonly/checkpoint-900",
|
| 5 |
+
"epoch": 13.953488372093023,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 1200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011627906976744186,
|
| 14 |
+
"grad_norm": 4.9740142822265625,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.317932665348053,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.29069767441860467,
|
| 21 |
+
"grad_norm": 1.5566426515579224,
|
| 22 |
+
"learning_rate": 6.282722513089005e-06,
|
| 23 |
+
"loss": 0.3199571371078491,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5813953488372093,
|
| 28 |
+
"grad_norm": 1.6379677057266235,
|
| 29 |
+
"learning_rate": 1.2827225130890053e-05,
|
| 30 |
+
"loss": 0.24125898361206055,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5813953488372093,
|
| 35 |
+
"eval_accuracy": 0.8287126584998925,
|
| 36 |
+
"eval_f1": 0.08611410118406888,
|
| 37 |
+
"eval_loss": 0.10772959887981415,
|
| 38 |
+
"eval_precision": 0.15810276679841898,
|
| 39 |
+
"eval_recall": 0.05917159763313609,
|
| 40 |
+
"eval_runtime": 4.2273,
|
| 41 |
+
"eval_samples_per_second": 80.667,
|
| 42 |
+
"eval_steps_per_second": 1.419,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.872093023255814,
|
| 47 |
+
"grad_norm": 8.703161239624023,
|
| 48 |
+
"learning_rate": 1.93717277486911e-05,
|
| 49 |
+
"loss": 0.20946041107177735,
|
| 50 |
+
"step": 75
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.1627906976744187,
|
| 54 |
+
"grad_norm": 0.9212121963500977,
|
| 55 |
+
"learning_rate": 2.591623036649215e-05,
|
| 56 |
+
"loss": 0.19275867462158203,
|
| 57 |
+
"step": 100
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.1627906976744187,
|
| 61 |
+
"eval_accuracy": 0.7635933806146572,
|
| 62 |
+
"eval_f1": 0.3981945837512538,
|
| 63 |
+
"eval_loss": 0.07719173282384872,
|
| 64 |
+
"eval_precision": 0.30121396054628224,
|
| 65 |
+
"eval_recall": 0.5872781065088757,
|
| 66 |
+
"eval_runtime": 1.6016,
|
| 67 |
+
"eval_samples_per_second": 212.909,
|
| 68 |
+
"eval_steps_per_second": 3.746,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.4534883720930232,
|
| 73 |
+
"grad_norm": 1.46477210521698,
|
| 74 |
+
"learning_rate": 3.246073298429319e-05,
|
| 75 |
+
"loss": 0.14603865623474122,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.744186046511628,
|
| 80 |
+
"grad_norm": 3.729416608810425,
|
| 81 |
+
"learning_rate": 3.900523560209424e-05,
|
| 82 |
+
"loss": 0.13855588912963868,
|
| 83 |
+
"step": 150
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.744186046511628,
|
| 87 |
+
"eval_accuracy": 0.81130453470879,
|
| 88 |
+
"eval_f1": 0.46785521339816316,
|
| 89 |
+
"eval_loss": 0.06223325803875923,
|
| 90 |
+
"eval_precision": 0.36851063829787234,
|
| 91 |
+
"eval_recall": 0.6405325443786982,
|
| 92 |
+
"eval_runtime": 1.5906,
|
| 93 |
+
"eval_samples_per_second": 214.385,
|
| 94 |
+
"eval_steps_per_second": 3.772,
|
| 95 |
+
"step": 150
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0348837209302326,
|
| 99 |
+
"grad_norm": 1.2140589952468872,
|
| 100 |
+
"learning_rate": 4.554973821989529e-05,
|
| 101 |
+
"loss": 0.09624839782714843,
|
| 102 |
+
"step": 175
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.3255813953488373,
|
| 106 |
+
"grad_norm": 0.875888466835022,
|
| 107 |
+
"learning_rate": 4.999346304613061e-05,
|
| 108 |
+
"loss": 0.0821054744720459,
|
| 109 |
+
"step": 200
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3255813953488373,
|
| 113 |
+
"eval_accuracy": 0.7872340425531915,
|
| 114 |
+
"eval_f1": 0.4847619047619048,
|
| 115 |
+
"eval_loss": 0.04723217338323593,
|
| 116 |
+
"eval_precision": 0.3574438202247191,
|
| 117 |
+
"eval_recall": 0.7529585798816568,
|
| 118 |
+
"eval_runtime": 1.6255,
|
| 119 |
+
"eval_samples_per_second": 209.777,
|
| 120 |
+
"eval_steps_per_second": 3.691,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.616279069767442,
|
| 125 |
+
"grad_norm": 1.0979472398757935,
|
| 126 |
+
"learning_rate": 4.98888472605987e-05,
|
| 127 |
+
"loss": 0.08414460182189941,
|
| 128 |
+
"step": 225
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.9069767441860463,
|
| 132 |
+
"grad_norm": 0.49558770656585693,
|
| 133 |
+
"learning_rate": 4.965717280180432e-05,
|
| 134 |
+
"loss": 0.08369583129882813,
|
| 135 |
+
"step": 250
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.9069767441860463,
|
| 139 |
+
"eval_accuracy": 0.6211046636578551,
|
| 140 |
+
"eval_f1": 0.3483915126625599,
|
| 141 |
+
"eval_loss": 0.045315902680158615,
|
| 142 |
+
"eval_precision": 0.22662511130899377,
|
| 143 |
+
"eval_recall": 0.7529585798816568,
|
| 144 |
+
"eval_runtime": 1.6274,
|
| 145 |
+
"eval_samples_per_second": 209.539,
|
| 146 |
+
"eval_steps_per_second": 3.687,
|
| 147 |
+
"step": 250
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.197674418604651,
|
| 151 |
+
"grad_norm": 0.8114362955093384,
|
| 152 |
+
"learning_rate": 4.9299622378173246e-05,
|
| 153 |
+
"loss": 0.05607881069183349,
|
| 154 |
+
"step": 275
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.488372093023256,
|
| 158 |
+
"grad_norm": 0.6278374195098877,
|
| 159 |
+
"learning_rate": 4.881802130053548e-05,
|
| 160 |
+
"loss": 0.043469696044921874,
|
| 161 |
+
"step": 300
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 3.488372093023256,
|
| 165 |
+
"eval_accuracy": 0.8082957231893402,
|
| 166 |
+
"eval_f1": 0.5285784074255008,
|
| 167 |
+
"eval_loss": 0.04308881610631943,
|
| 168 |
+
"eval_precision": 0.39460247994164843,
|
| 169 |
+
"eval_recall": 0.8002958579881657,
|
| 170 |
+
"eval_runtime": 1.6193,
|
| 171 |
+
"eval_samples_per_second": 210.585,
|
| 172 |
+
"eval_steps_per_second": 3.705,
|
| 173 |
+
"step": 300
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.7790697674418605,
|
| 177 |
+
"grad_norm": 1.0219486951828003,
|
| 178 |
+
"learning_rate": 4.821482816383218e-05,
|
| 179 |
+
"loss": 0.047760300636291504,
|
| 180 |
+
"step": 325
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.069767441860465,
|
| 184 |
+
"grad_norm": 0.509274959564209,
|
| 185 |
+
"learning_rate": 4.749312229587908e-05,
|
| 186 |
+
"loss": 0.03177810907363891,
|
| 187 |
+
"step": 350
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 4.069767441860465,
|
| 191 |
+
"eval_accuracy": 0.9196217494089834,
|
| 192 |
+
"eval_f1": 0.7152317880794702,
|
| 193 |
+
"eval_loss": 0.04378104954957962,
|
| 194 |
+
"eval_precision": 0.6474820143884892,
|
| 195 |
+
"eval_recall": 0.7988165680473372,
|
| 196 |
+
"eval_runtime": 1.6177,
|
| 197 |
+
"eval_samples_per_second": 210.792,
|
| 198 |
+
"eval_steps_per_second": 3.709,
|
| 199 |
+
"step": 350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.3604651162790695,
|
| 203 |
+
"grad_norm": 0.47098585963249207,
|
| 204 |
+
"learning_rate": 4.6656588037260834e-05,
|
| 205 |
+
"loss": 0.020146708488464355,
|
| 206 |
+
"step": 375
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.651162790697675,
|
| 210 |
+
"grad_norm": 0.7534321546554565,
|
| 211 |
+
"learning_rate": 4.570949593260837e-05,
|
| 212 |
+
"loss": 0.018921481370925905,
|
| 213 |
+
"step": 400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.651162790697675,
|
| 217 |
+
"eval_accuracy": 0.9097356544165055,
|
| 218 |
+
"eval_f1": 0.6987341772151898,
|
| 219 |
+
"eval_loss": 0.05498046055436134,
|
| 220 |
+
"eval_precision": 0.6106194690265486,
|
| 221 |
+
"eval_recall": 0.8165680473372781,
|
| 222 |
+
"eval_runtime": 1.6157,
|
| 223 |
+
"eval_samples_per_second": 211.05,
|
| 224 |
+
"eval_steps_per_second": 3.713,
|
| 225 |
+
"step": 400
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.941860465116279,
|
| 229 |
+
"grad_norm": 0.7499828934669495,
|
| 230 |
+
"learning_rate": 4.465668092927841e-05,
|
| 231 |
+
"loss": 0.01998784065246582,
|
| 232 |
+
"step": 425
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 5.232558139534884,
|
| 236 |
+
"grad_norm": 0.24809418618679047,
|
| 237 |
+
"learning_rate": 4.350351769473198e-05,
|
| 238 |
+
"loss": 0.01218403697013855,
|
| 239 |
+
"step": 450
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 5.232558139534884,
|
| 243 |
+
"eval_accuracy": 0.9084461637653127,
|
| 244 |
+
"eval_f1": 0.6831746031746032,
|
| 245 |
+
"eval_loss": 0.05535344406962395,
|
| 246 |
+
"eval_precision": 0.5984427141268076,
|
| 247 |
+
"eval_recall": 0.7958579881656804,
|
| 248 |
+
"eval_runtime": 1.6123,
|
| 249 |
+
"eval_samples_per_second": 211.498,
|
| 250 |
+
"eval_steps_per_second": 3.721,
|
| 251 |
+
"step": 450
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 5.523255813953488,
|
| 255 |
+
"grad_norm": 0.3144132196903229,
|
| 256 |
+
"learning_rate": 4.2255893178617745e-05,
|
| 257 |
+
"loss": 0.012306832075119019,
|
| 258 |
+
"step": 475
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 5.813953488372093,
|
| 262 |
+
"grad_norm": 0.1607646942138672,
|
| 263 |
+
"learning_rate": 4.092017655963198e-05,
|
| 264 |
+
"loss": 0.010539753437042236,
|
| 265 |
+
"step": 500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 5.813953488372093,
|
| 269 |
+
"eval_accuracy": 0.9527186761229315,
|
| 270 |
+
"eval_f1": 0.8059040590405904,
|
| 271 |
+
"eval_loss": 0.07191809266805649,
|
| 272 |
+
"eval_precision": 0.8041237113402062,
|
| 273 |
+
"eval_recall": 0.8076923076923077,
|
| 274 |
+
"eval_runtime": 1.6569,
|
| 275 |
+
"eval_samples_per_second": 205.806,
|
| 276 |
+
"eval_steps_per_second": 3.621,
|
| 277 |
+
"step": 500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.104651162790698,
|
| 281 |
+
"grad_norm": 0.19801200926303864,
|
| 282 |
+
"learning_rate": 3.9503186730577816e-05,
|
| 283 |
+
"loss": 0.008691226243972778,
|
| 284 |
+
"step": 525
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.395348837209302,
|
| 288 |
+
"grad_norm": 0.4177681505680084,
|
| 289 |
+
"learning_rate": 3.801215748761385e-05,
|
| 290 |
+
"loss": 0.004750225245952606,
|
| 291 |
+
"step": 550
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 6.395348837209302,
|
| 295 |
+
"eval_accuracy": 0.9458413926499033,
|
| 296 |
+
"eval_f1": 0.7863866763215062,
|
| 297 |
+
"eval_loss": 0.07378304749727249,
|
| 298 |
+
"eval_precision": 0.7702127659574468,
|
| 299 |
+
"eval_recall": 0.8032544378698225,
|
| 300 |
+
"eval_runtime": 1.6214,
|
| 301 |
+
"eval_samples_per_second": 210.313,
|
| 302 |
+
"eval_steps_per_second": 3.701,
|
| 303 |
+
"step": 550
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.686046511627907,
|
| 307 |
+
"grad_norm": 0.30753573775291443,
|
| 308 |
+
"learning_rate": 3.645470060140278e-05,
|
| 309 |
+
"loss": 0.005743271708488464,
|
| 310 |
+
"step": 575
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.976744186046512,
|
| 314 |
+
"grad_norm": 0.13237181305885315,
|
| 315 |
+
"learning_rate": 3.4838766958683304e-05,
|
| 316 |
+
"loss": 0.006215065121650696,
|
| 317 |
+
"step": 600
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.976744186046512,
|
| 321 |
+
"eval_accuracy": 0.9447668171072426,
|
| 322 |
+
"eval_f1": 0.7789934354485777,
|
| 323 |
+
"eval_loss": 0.062308527529239655,
|
| 324 |
+
"eval_precision": 0.7683453237410072,
|
| 325 |
+
"eval_recall": 0.7899408284023669,
|
| 326 |
+
"eval_runtime": 1.6197,
|
| 327 |
+
"eval_samples_per_second": 210.528,
|
| 328 |
+
"eval_steps_per_second": 3.704,
|
| 329 |
+
"step": 600
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 7.267441860465116,
|
| 333 |
+
"grad_norm": 0.023809686303138733,
|
| 334 |
+
"learning_rate": 3.317260597263932e-05,
|
| 335 |
+
"loss": 0.002318141460418701,
|
| 336 |
+
"step": 625
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.558139534883721,
|
| 340 |
+
"grad_norm": 0.03782212734222412,
|
| 341 |
+
"learning_rate": 3.146472346927845e-05,
|
| 342 |
+
"loss": 0.003790220320224762,
|
| 343 |
+
"step": 650
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 7.558139534883721,
|
| 347 |
+
"eval_accuracy": 0.9529335912314636,
|
| 348 |
+
"eval_f1": 0.8062827225130891,
|
| 349 |
+
"eval_loss": 0.07683796435594559,
|
| 350 |
+
"eval_precision": 0.8154311649016641,
|
| 351 |
+
"eval_recall": 0.7973372781065089,
|
| 352 |
+
"eval_runtime": 1.6235,
|
| 353 |
+
"eval_samples_per_second": 210.043,
|
| 354 |
+
"eval_steps_per_second": 3.696,
|
| 355 |
+
"step": 650
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 7.848837209302325,
|
| 359 |
+
"grad_norm": 0.037971507757902145,
|
| 360 |
+
"learning_rate": 2.9723838264811545e-05,
|
| 361 |
+
"loss": 0.0018679146468639374,
|
| 362 |
+
"step": 675
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 8.13953488372093,
|
| 366 |
+
"grad_norm": 0.6083823442459106,
|
| 367 |
+
"learning_rate": 2.7958837655707814e-05,
|
| 368 |
+
"loss": 0.0019237904250621796,
|
| 369 |
+
"step": 700
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 8.13953488372093,
|
| 373 |
+
"eval_accuracy": 0.9565871480765098,
|
| 374 |
+
"eval_f1": 0.8152091254752851,
|
| 375 |
+
"eval_loss": 0.08732757717370987,
|
| 376 |
+
"eval_precision": 0.838810641627543,
|
| 377 |
+
"eval_recall": 0.7928994082840237,
|
| 378 |
+
"eval_runtime": 1.6124,
|
| 379 |
+
"eval_samples_per_second": 211.484,
|
| 380 |
+
"eval_steps_per_second": 3.721,
|
| 381 |
+
"step": 700
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.430232558139535,
|
| 385 |
+
"grad_norm": 0.022007286548614502,
|
| 386 |
+
"learning_rate": 2.6178732048650694e-05,
|
| 387 |
+
"loss": 0.0010710173845291137,
|
| 388 |
+
"step": 725
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 8.720930232558139,
|
| 392 |
+
"grad_norm": 0.09439104795455933,
|
| 393 |
+
"learning_rate": 2.4392608962010652e-05,
|
| 394 |
+
"loss": 0.0006907544285058976,
|
| 395 |
+
"step": 750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.720930232558139,
|
| 399 |
+
"eval_accuracy": 0.9449817322157748,
|
| 400 |
+
"eval_f1": 0.7829848594087959,
|
| 401 |
+
"eval_loss": 0.08308640867471695,
|
| 402 |
+
"eval_precision": 0.7637130801687764,
|
| 403 |
+
"eval_recall": 0.8032544378698225,
|
| 404 |
+
"eval_runtime": 1.6359,
|
| 405 |
+
"eval_samples_per_second": 208.445,
|
| 406 |
+
"eval_steps_per_second": 3.668,
|
| 407 |
+
"step": 750
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 9.011627906976743,
|
| 411 |
+
"grad_norm": 0.018290311098098755,
|
| 412 |
+
"learning_rate": 2.2609586633659256e-05,
|
| 413 |
+
"loss": 0.0005397907271981239,
|
| 414 |
+
"step": 775
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 9.30232558139535,
|
| 418 |
+
"grad_norm": 0.002839206252247095,
|
| 419 |
+
"learning_rate": 2.0838767471959014e-05,
|
| 420 |
+
"loss": 0.0002632186934351921,
|
| 421 |
+
"step": 800
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 9.30232558139535,
|
| 425 |
+
"eval_accuracy": 0.9544379969911885,
|
| 426 |
+
"eval_f1": 0.8081880212282032,
|
| 427 |
+
"eval_loss": 0.09754534065723419,
|
| 428 |
+
"eval_precision": 0.8289269051321928,
|
| 429 |
+
"eval_recall": 0.7884615384615384,
|
| 430 |
+
"eval_runtime": 1.6188,
|
| 431 |
+
"eval_samples_per_second": 210.653,
|
| 432 |
+
"eval_steps_per_second": 3.706,
|
| 433 |
+
"step": 800
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 9.593023255813954,
|
| 437 |
+
"grad_norm": 0.04134448245167732,
|
| 438 |
+
"learning_rate": 1.9089191587563417e-05,
|
| 439 |
+
"loss": 0.0005076020956039428,
|
| 440 |
+
"step": 825
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 9.883720930232558,
|
| 444 |
+
"grad_norm": 0.08968573063611984,
|
| 445 |
+
"learning_rate": 1.7369790643249573e-05,
|
| 446 |
+
"loss": 0.0005874237418174743,
|
| 447 |
+
"step": 850
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 9.883720930232558,
|
| 451 |
+
"eval_accuracy": 0.9548678272082527,
|
| 452 |
+
"eval_f1": 0.8093797276853252,
|
| 453 |
+
"eval_loss": 0.09803693741559982,
|
| 454 |
+
"eval_precision": 0.8281733746130031,
|
| 455 |
+
"eval_recall": 0.7914201183431953,
|
| 456 |
+
"eval_runtime": 1.6209,
|
| 457 |
+
"eval_samples_per_second": 210.379,
|
| 458 |
+
"eval_steps_per_second": 3.702,
|
| 459 |
+
"step": 850
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 10.174418604651162,
|
| 463 |
+
"grad_norm": 0.02889181859791279,
|
| 464 |
+
"learning_rate": 1.5689342257382207e-05,
|
| 465 |
+
"loss": 0.0004838050529360771,
|
| 466 |
+
"step": 875
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.465116279069768,
|
| 470 |
+
"grad_norm": 0.020970497280359268,
|
| 471 |
+
"learning_rate": 1.4056425193781048e-05,
|
| 472 |
+
"loss": 0.0001906752586364746,
|
| 473 |
+
"step": 900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 10.465116279069768,
|
| 477 |
+
"eval_accuracy": 0.9587362991618311,
|
| 478 |
+
"eval_f1": 0.8228043143297381,
|
| 479 |
+
"eval_loss": 0.10382715612649918,
|
| 480 |
+
"eval_precision": 0.8585209003215434,
|
| 481 |
+
"eval_recall": 0.7899408284023669,
|
| 482 |
+
"eval_runtime": 1.6341,
|
| 483 |
+
"eval_samples_per_second": 208.675,
|
| 484 |
+
"eval_steps_per_second": 3.672,
|
| 485 |
+
"step": 900
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 10.755813953488373,
|
| 489 |
+
"grad_norm": 0.028520096093416214,
|
| 490 |
+
"learning_rate": 1.2479375566749694e-05,
|
| 491 |
+
"loss": 0.00012492083944380284,
|
| 492 |
+
"step": 925
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 11.046511627906977,
|
| 496 |
+
"grad_norm": 0.03838531672954559,
|
| 497 |
+
"learning_rate": 1.0966244284840926e-05,
|
| 498 |
+
"loss": 0.00028255857527256014,
|
| 499 |
+
"step": 950
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 11.046511627906977,
|
| 503 |
+
"eval_accuracy": 0.9580915538362347,
|
| 504 |
+
"eval_f1": 0.819369715603382,
|
| 505 |
+
"eval_loss": 0.10540532320737839,
|
| 506 |
+
"eval_precision": 0.8528,
|
| 507 |
+
"eval_recall": 0.7884615384615384,
|
| 508 |
+
"eval_runtime": 1.6495,
|
| 509 |
+
"eval_samples_per_second": 206.73,
|
| 510 |
+
"eval_steps_per_second": 3.637,
|
| 511 |
+
"step": 950
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 11.337209302325581,
|
| 515 |
+
"grad_norm": 0.05398673936724663,
|
| 516 |
+
"learning_rate": 9.524755950610204e-06,
|
| 517 |
+
"loss": 0.00014415644109249114,
|
| 518 |
+
"step": 975
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 11.627906976744185,
|
| 522 |
+
"grad_norm": 0.026561638340353966,
|
| 523 |
+
"learning_rate": 8.162269426175681e-06,
|
| 524 |
+
"loss": 9.150288999080658e-05,
|
| 525 |
+
"step": 1000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 11.627906976744185,
|
| 529 |
+
"eval_accuracy": 0.9574468085106383,
|
| 530 |
+
"eval_f1": 0.8180428134556575,
|
| 531 |
+
"eval_loss": 0.10544126480817795,
|
| 532 |
+
"eval_precision": 0.8465189873417721,
|
| 533 |
+
"eval_recall": 0.7914201183431953,
|
| 534 |
+
"eval_runtime": 1.6274,
|
| 535 |
+
"eval_samples_per_second": 209.539,
|
| 536 |
+
"eval_steps_per_second": 3.687,
|
| 537 |
+
"step": 1000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 11.918604651162791,
|
| 541 |
+
"grad_norm": 0.010734460316598415,
|
| 542 |
+
"learning_rate": 6.885740265899526e-06,
|
| 543 |
+
"loss": 4.033456556499005e-05,
|
| 544 |
+
"step": 1025
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 12.209302325581396,
|
| 548 |
+
"grad_norm": 0.02024712599813938,
|
| 549 |
+
"learning_rate": 5.701685207973243e-06,
|
| 550 |
+
"loss": 6.52561942115426e-05,
|
| 551 |
+
"step": 1050
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 12.209302325581396,
|
| 555 |
+
"eval_accuracy": 0.9576617236191705,
|
| 556 |
+
"eval_f1": 0.8190184049079755,
|
| 557 |
+
"eval_loss": 0.10666228085756302,
|
| 558 |
+
"eval_precision": 0.8503184713375797,
|
| 559 |
+
"eval_recall": 0.7899408284023669,
|
| 560 |
+
"eval_runtime": 1.6269,
|
| 561 |
+
"eval_samples_per_second": 209.598,
|
| 562 |
+
"eval_steps_per_second": 3.688,
|
| 563 |
+
"step": 1050
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 12.5,
|
| 567 |
+
"grad_norm": 0.034469809383153915,
|
| 568 |
+
"learning_rate": 4.616148906179082e-06,
|
| 569 |
+
"loss": 7.551711052656174e-05,
|
| 570 |
+
"step": 1075
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 12.790697674418604,
|
| 574 |
+
"grad_norm": 0.004050907213240862,
|
| 575 |
+
"learning_rate": 3.6346730716634025e-06,
|
| 576 |
+
"loss": 7.49318953603506e-05,
|
| 577 |
+
"step": 1100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 12.790697674418604,
|
| 581 |
+
"eval_accuracy": 0.9565871480765098,
|
| 582 |
+
"eval_f1": 0.8146453089244853,
|
| 583 |
+
"eval_loss": 0.10619472712278366,
|
| 584 |
+
"eval_precision": 0.8409448818897638,
|
| 585 |
+
"eval_recall": 0.7899408284023669,
|
| 586 |
+
"eval_runtime": 1.6196,
|
| 587 |
+
"eval_samples_per_second": 210.546,
|
| 588 |
+
"eval_steps_per_second": 3.705,
|
| 589 |
+
"step": 1100
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 13.081395348837209,
|
| 593 |
+
"grad_norm": 0.11917345970869064,
|
| 594 |
+
"learning_rate": 2.7622681822545764e-06,
|
| 595 |
+
"loss": 8.815748617053032e-05,
|
| 596 |
+
"step": 1125
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 13.372093023255815,
|
| 600 |
+
"grad_norm": 0.02682509273290634,
|
| 601 |
+
"learning_rate": 2.0033879037506004e-06,
|
| 602 |
+
"loss": 6.015756167471409e-05,
|
| 603 |
+
"step": 1150
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 13.372093023255815,
|
| 607 |
+
"eval_accuracy": 0.9583064689447668,
|
| 608 |
+
"eval_f1": 0.8205521472392638,
|
| 609 |
+
"eval_loss": 0.10740578919649124,
|
| 610 |
+
"eval_precision": 0.8519108280254777,
|
| 611 |
+
"eval_recall": 0.7914201183431953,
|
| 612 |
+
"eval_runtime": 1.6217,
|
| 613 |
+
"eval_samples_per_second": 210.273,
|
| 614 |
+
"eval_steps_per_second": 3.7,
|
| 615 |
+
"step": 1150
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 13.662790697674419,
|
| 619 |
+
"grad_norm": 0.004100531339645386,
|
| 620 |
+
"learning_rate": 1.361906353756917e-06,
|
| 621 |
+
"loss": 4.690625239163637e-05,
|
| 622 |
+
"step": 1175
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 13.953488372093023,
|
| 626 |
+
"grad_norm": 0.008385015651583672,
|
| 627 |
+
"learning_rate": 8.410983241436132e-07,
|
| 628 |
+
"loss": 6.560925394296646e-05,
|
| 629 |
+
"step": 1200
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 13.953488372093023,
|
| 633 |
+
"eval_accuracy": 0.9578766387277026,
|
| 634 |
+
"eval_f1": 0.8181120491174213,
|
| 635 |
+
"eval_loss": 0.10830119252204895,
|
| 636 |
+
"eval_precision": 0.8500797448165869,
|
| 637 |
+
"eval_recall": 0.7884615384615384,
|
| 638 |
+
"eval_runtime": 1.6315,
|
| 639 |
+
"eval_samples_per_second": 209.012,
|
| 640 |
+
"eval_steps_per_second": 3.678,
|
| 641 |
+
"step": 1200
|
| 642 |
+
}
|
| 643 |
+
],
|
| 644 |
+
"logging_steps": 25,
|
| 645 |
+
"max_steps": 1290,
|
| 646 |
+
"num_input_tokens_seen": 0,
|
| 647 |
+
"num_train_epochs": 15,
|
| 648 |
+
"save_steps": 50,
|
| 649 |
+
"stateful_callbacks": {
|
| 650 |
+
"EarlyStoppingCallback": {
|
| 651 |
+
"args": {
|
| 652 |
+
"early_stopping_patience": 10,
|
| 653 |
+
"early_stopping_threshold": 0.0
|
| 654 |
+
},
|
| 655 |
+
"attributes": {
|
| 656 |
+
"early_stopping_patience_counter": 6
|
| 657 |
+
}
|
| 658 |
+
},
|
| 659 |
+
"TrainerControl": {
|
| 660 |
+
"args": {
|
| 661 |
+
"should_epoch_stop": false,
|
| 662 |
+
"should_evaluate": false,
|
| 663 |
+
"should_log": false,
|
| 664 |
+
"should_save": true,
|
| 665 |
+
"should_training_stop": false
|
| 666 |
+
},
|
| 667 |
+
"attributes": {}
|
| 668 |
+
}
|
| 669 |
+
},
|
| 670 |
+
"total_flos": 6509462134924800.0,
|
| 671 |
+
"train_batch_size": 16,
|
| 672 |
+
"trial_name": null,
|
| 673 |
+
"trial_params": null
|
| 674 |
+
}
|
checkpoint-1200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5071855a54d36018bc5c30ca34d9fef42b69fb3deda057af4ffc70193237d5c4
|
| 3 |
+
size 5201
|
checkpoint-1250/config.json
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.4,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"gradient_checkpointing": false,
|
| 20 |
+
"hidden_activation": "gelu",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "O",
|
| 24 |
+
"1": "FP",
|
| 25 |
+
"2": "RP",
|
| 26 |
+
"3": "RV",
|
| 27 |
+
"4": "PW"
|
| 28 |
+
},
|
| 29 |
+
"initializer_cutoff_factor": 2.0,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 1152,
|
| 32 |
+
"label2id": {
|
| 33 |
+
"FP": 1,
|
| 34 |
+
"O": 0,
|
| 35 |
+
"PW": 4,
|
| 36 |
+
"RP": 2,
|
| 37 |
+
"RV": 3
|
| 38 |
+
},
|
| 39 |
+
"layer_norm_eps": 1e-05,
|
| 40 |
+
"layer_types": [
|
| 41 |
+
"full_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"full_attention"
|
| 63 |
+
],
|
| 64 |
+
"local_attention": 128,
|
| 65 |
+
"max_position_embeddings": 8192,
|
| 66 |
+
"mlp_bias": false,
|
| 67 |
+
"mlp_dropout": 0.0,
|
| 68 |
+
"model_type": "modernbert",
|
| 69 |
+
"norm_bias": false,
|
| 70 |
+
"norm_eps": 1e-05,
|
| 71 |
+
"num_attention_heads": 12,
|
| 72 |
+
"num_hidden_layers": 22,
|
| 73 |
+
"pad_token_id": 50283,
|
| 74 |
+
"position_embedding_type": "absolute",
|
| 75 |
+
"repad_logits_with_grad": false,
|
| 76 |
+
"rope_parameters": {
|
| 77 |
+
"full_attention": {
|
| 78 |
+
"rope_theta": 160000.0,
|
| 79 |
+
"rope_type": "default"
|
| 80 |
+
},
|
| 81 |
+
"sliding_attention": {
|
| 82 |
+
"rope_theta": 10000.0,
|
| 83 |
+
"rope_type": "default"
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"sep_token_id": 50282,
|
| 87 |
+
"sparse_pred_ignore_index": -100,
|
| 88 |
+
"sparse_prediction": false,
|
| 89 |
+
"tie_word_embeddings": true,
|
| 90 |
+
"transformers_version": "5.0.0",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"vocab_size": 50368
|
| 93 |
+
}
|
checkpoint-1250/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09581240c6dc5c8b58b4e5b94a41c7a53e1528bc5fc0ab3742bd8869048b558e
|
| 3 |
+
size 598449012
|
checkpoint-1250/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c65641ba7ee7c24fbb634d7c2c686e52088e8ad3271ab615835e9b5ec7f3aa2
|
| 3 |
+
size 535150859
|
checkpoint-1250/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7bdaec2d9eed960550d0e19b70cb2eea8832b93c6127669a2a582c2090899806
|
| 3 |
+
size 14709
|
checkpoint-1250/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f7ba0c5f31bfcb1b4183cfd6b9b630ac6985851ed8d739f00668a5d279cbf6a
|
| 3 |
+
size 1465
|
checkpoint-1250/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1250/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"mask_token": "[MASK]",
|
| 7 |
+
"model_input_names": [
|
| 8 |
+
"input_ids",
|
| 9 |
+
"attention_mask"
|
| 10 |
+
],
|
| 11 |
+
"model_max_length": 8192,
|
| 12 |
+
"pad_token": "[PAD]",
|
| 13 |
+
"sep_token": "[SEP]",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend",
|
| 15 |
+
"unk_token": "[UNK]"
|
| 16 |
+
}
|
checkpoint-1250/trainer_state.json
ADDED
|
@@ -0,0 +1,700 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 900,
|
| 3 |
+
"best_metric": 0.8228043143297381,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/expA_realonly/checkpoint-900",
|
| 5 |
+
"epoch": 14.534883720930232,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 1250,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011627906976744186,
|
| 14 |
+
"grad_norm": 4.9740142822265625,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.317932665348053,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.29069767441860467,
|
| 21 |
+
"grad_norm": 1.5566426515579224,
|
| 22 |
+
"learning_rate": 6.282722513089005e-06,
|
| 23 |
+
"loss": 0.3199571371078491,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5813953488372093,
|
| 28 |
+
"grad_norm": 1.6379677057266235,
|
| 29 |
+
"learning_rate": 1.2827225130890053e-05,
|
| 30 |
+
"loss": 0.24125898361206055,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5813953488372093,
|
| 35 |
+
"eval_accuracy": 0.8287126584998925,
|
| 36 |
+
"eval_f1": 0.08611410118406888,
|
| 37 |
+
"eval_loss": 0.10772959887981415,
|
| 38 |
+
"eval_precision": 0.15810276679841898,
|
| 39 |
+
"eval_recall": 0.05917159763313609,
|
| 40 |
+
"eval_runtime": 4.2273,
|
| 41 |
+
"eval_samples_per_second": 80.667,
|
| 42 |
+
"eval_steps_per_second": 1.419,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.872093023255814,
|
| 47 |
+
"grad_norm": 8.703161239624023,
|
| 48 |
+
"learning_rate": 1.93717277486911e-05,
|
| 49 |
+
"loss": 0.20946041107177735,
|
| 50 |
+
"step": 75
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.1627906976744187,
|
| 54 |
+
"grad_norm": 0.9212121963500977,
|
| 55 |
+
"learning_rate": 2.591623036649215e-05,
|
| 56 |
+
"loss": 0.19275867462158203,
|
| 57 |
+
"step": 100
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.1627906976744187,
|
| 61 |
+
"eval_accuracy": 0.7635933806146572,
|
| 62 |
+
"eval_f1": 0.3981945837512538,
|
| 63 |
+
"eval_loss": 0.07719173282384872,
|
| 64 |
+
"eval_precision": 0.30121396054628224,
|
| 65 |
+
"eval_recall": 0.5872781065088757,
|
| 66 |
+
"eval_runtime": 1.6016,
|
| 67 |
+
"eval_samples_per_second": 212.909,
|
| 68 |
+
"eval_steps_per_second": 3.746,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.4534883720930232,
|
| 73 |
+
"grad_norm": 1.46477210521698,
|
| 74 |
+
"learning_rate": 3.246073298429319e-05,
|
| 75 |
+
"loss": 0.14603865623474122,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.744186046511628,
|
| 80 |
+
"grad_norm": 3.729416608810425,
|
| 81 |
+
"learning_rate": 3.900523560209424e-05,
|
| 82 |
+
"loss": 0.13855588912963868,
|
| 83 |
+
"step": 150
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.744186046511628,
|
| 87 |
+
"eval_accuracy": 0.81130453470879,
|
| 88 |
+
"eval_f1": 0.46785521339816316,
|
| 89 |
+
"eval_loss": 0.06223325803875923,
|
| 90 |
+
"eval_precision": 0.36851063829787234,
|
| 91 |
+
"eval_recall": 0.6405325443786982,
|
| 92 |
+
"eval_runtime": 1.5906,
|
| 93 |
+
"eval_samples_per_second": 214.385,
|
| 94 |
+
"eval_steps_per_second": 3.772,
|
| 95 |
+
"step": 150
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0348837209302326,
|
| 99 |
+
"grad_norm": 1.2140589952468872,
|
| 100 |
+
"learning_rate": 4.554973821989529e-05,
|
| 101 |
+
"loss": 0.09624839782714843,
|
| 102 |
+
"step": 175
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.3255813953488373,
|
| 106 |
+
"grad_norm": 0.875888466835022,
|
| 107 |
+
"learning_rate": 4.999346304613061e-05,
|
| 108 |
+
"loss": 0.0821054744720459,
|
| 109 |
+
"step": 200
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3255813953488373,
|
| 113 |
+
"eval_accuracy": 0.7872340425531915,
|
| 114 |
+
"eval_f1": 0.4847619047619048,
|
| 115 |
+
"eval_loss": 0.04723217338323593,
|
| 116 |
+
"eval_precision": 0.3574438202247191,
|
| 117 |
+
"eval_recall": 0.7529585798816568,
|
| 118 |
+
"eval_runtime": 1.6255,
|
| 119 |
+
"eval_samples_per_second": 209.777,
|
| 120 |
+
"eval_steps_per_second": 3.691,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.616279069767442,
|
| 125 |
+
"grad_norm": 1.0979472398757935,
|
| 126 |
+
"learning_rate": 4.98888472605987e-05,
|
| 127 |
+
"loss": 0.08414460182189941,
|
| 128 |
+
"step": 225
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.9069767441860463,
|
| 132 |
+
"grad_norm": 0.49558770656585693,
|
| 133 |
+
"learning_rate": 4.965717280180432e-05,
|
| 134 |
+
"loss": 0.08369583129882813,
|
| 135 |
+
"step": 250
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.9069767441860463,
|
| 139 |
+
"eval_accuracy": 0.6211046636578551,
|
| 140 |
+
"eval_f1": 0.3483915126625599,
|
| 141 |
+
"eval_loss": 0.045315902680158615,
|
| 142 |
+
"eval_precision": 0.22662511130899377,
|
| 143 |
+
"eval_recall": 0.7529585798816568,
|
| 144 |
+
"eval_runtime": 1.6274,
|
| 145 |
+
"eval_samples_per_second": 209.539,
|
| 146 |
+
"eval_steps_per_second": 3.687,
|
| 147 |
+
"step": 250
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.197674418604651,
|
| 151 |
+
"grad_norm": 0.8114362955093384,
|
| 152 |
+
"learning_rate": 4.9299622378173246e-05,
|
| 153 |
+
"loss": 0.05607881069183349,
|
| 154 |
+
"step": 275
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.488372093023256,
|
| 158 |
+
"grad_norm": 0.6278374195098877,
|
| 159 |
+
"learning_rate": 4.881802130053548e-05,
|
| 160 |
+
"loss": 0.043469696044921874,
|
| 161 |
+
"step": 300
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 3.488372093023256,
|
| 165 |
+
"eval_accuracy": 0.8082957231893402,
|
| 166 |
+
"eval_f1": 0.5285784074255008,
|
| 167 |
+
"eval_loss": 0.04308881610631943,
|
| 168 |
+
"eval_precision": 0.39460247994164843,
|
| 169 |
+
"eval_recall": 0.8002958579881657,
|
| 170 |
+
"eval_runtime": 1.6193,
|
| 171 |
+
"eval_samples_per_second": 210.585,
|
| 172 |
+
"eval_steps_per_second": 3.705,
|
| 173 |
+
"step": 300
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.7790697674418605,
|
| 177 |
+
"grad_norm": 1.0219486951828003,
|
| 178 |
+
"learning_rate": 4.821482816383218e-05,
|
| 179 |
+
"loss": 0.047760300636291504,
|
| 180 |
+
"step": 325
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.069767441860465,
|
| 184 |
+
"grad_norm": 0.509274959564209,
|
| 185 |
+
"learning_rate": 4.749312229587908e-05,
|
| 186 |
+
"loss": 0.03177810907363891,
|
| 187 |
+
"step": 350
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 4.069767441860465,
|
| 191 |
+
"eval_accuracy": 0.9196217494089834,
|
| 192 |
+
"eval_f1": 0.7152317880794702,
|
| 193 |
+
"eval_loss": 0.04378104954957962,
|
| 194 |
+
"eval_precision": 0.6474820143884892,
|
| 195 |
+
"eval_recall": 0.7988165680473372,
|
| 196 |
+
"eval_runtime": 1.6177,
|
| 197 |
+
"eval_samples_per_second": 210.792,
|
| 198 |
+
"eval_steps_per_second": 3.709,
|
| 199 |
+
"step": 350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.3604651162790695,
|
| 203 |
+
"grad_norm": 0.47098585963249207,
|
| 204 |
+
"learning_rate": 4.6656588037260834e-05,
|
| 205 |
+
"loss": 0.020146708488464355,
|
| 206 |
+
"step": 375
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.651162790697675,
|
| 210 |
+
"grad_norm": 0.7534321546554565,
|
| 211 |
+
"learning_rate": 4.570949593260837e-05,
|
| 212 |
+
"loss": 0.018921481370925905,
|
| 213 |
+
"step": 400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.651162790697675,
|
| 217 |
+
"eval_accuracy": 0.9097356544165055,
|
| 218 |
+
"eval_f1": 0.6987341772151898,
|
| 219 |
+
"eval_loss": 0.05498046055436134,
|
| 220 |
+
"eval_precision": 0.6106194690265486,
|
| 221 |
+
"eval_recall": 0.8165680473372781,
|
| 222 |
+
"eval_runtime": 1.6157,
|
| 223 |
+
"eval_samples_per_second": 211.05,
|
| 224 |
+
"eval_steps_per_second": 3.713,
|
| 225 |
+
"step": 400
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.941860465116279,
|
| 229 |
+
"grad_norm": 0.7499828934669495,
|
| 230 |
+
"learning_rate": 4.465668092927841e-05,
|
| 231 |
+
"loss": 0.01998784065246582,
|
| 232 |
+
"step": 425
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 5.232558139534884,
|
| 236 |
+
"grad_norm": 0.24809418618679047,
|
| 237 |
+
"learning_rate": 4.350351769473198e-05,
|
| 238 |
+
"loss": 0.01218403697013855,
|
| 239 |
+
"step": 450
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 5.232558139534884,
|
| 243 |
+
"eval_accuracy": 0.9084461637653127,
|
| 244 |
+
"eval_f1": 0.6831746031746032,
|
| 245 |
+
"eval_loss": 0.05535344406962395,
|
| 246 |
+
"eval_precision": 0.5984427141268076,
|
| 247 |
+
"eval_recall": 0.7958579881656804,
|
| 248 |
+
"eval_runtime": 1.6123,
|
| 249 |
+
"eval_samples_per_second": 211.498,
|
| 250 |
+
"eval_steps_per_second": 3.721,
|
| 251 |
+
"step": 450
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 5.523255813953488,
|
| 255 |
+
"grad_norm": 0.3144132196903229,
|
| 256 |
+
"learning_rate": 4.2255893178617745e-05,
|
| 257 |
+
"loss": 0.012306832075119019,
|
| 258 |
+
"step": 475
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 5.813953488372093,
|
| 262 |
+
"grad_norm": 0.1607646942138672,
|
| 263 |
+
"learning_rate": 4.092017655963198e-05,
|
| 264 |
+
"loss": 0.010539753437042236,
|
| 265 |
+
"step": 500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 5.813953488372093,
|
| 269 |
+
"eval_accuracy": 0.9527186761229315,
|
| 270 |
+
"eval_f1": 0.8059040590405904,
|
| 271 |
+
"eval_loss": 0.07191809266805649,
|
| 272 |
+
"eval_precision": 0.8041237113402062,
|
| 273 |
+
"eval_recall": 0.8076923076923077,
|
| 274 |
+
"eval_runtime": 1.6569,
|
| 275 |
+
"eval_samples_per_second": 205.806,
|
| 276 |
+
"eval_steps_per_second": 3.621,
|
| 277 |
+
"step": 500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.104651162790698,
|
| 281 |
+
"grad_norm": 0.19801200926303864,
|
| 282 |
+
"learning_rate": 3.9503186730577816e-05,
|
| 283 |
+
"loss": 0.008691226243972778,
|
| 284 |
+
"step": 525
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.395348837209302,
|
| 288 |
+
"grad_norm": 0.4177681505680084,
|
| 289 |
+
"learning_rate": 3.801215748761385e-05,
|
| 290 |
+
"loss": 0.004750225245952606,
|
| 291 |
+
"step": 550
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 6.395348837209302,
|
| 295 |
+
"eval_accuracy": 0.9458413926499033,
|
| 296 |
+
"eval_f1": 0.7863866763215062,
|
| 297 |
+
"eval_loss": 0.07378304749727249,
|
| 298 |
+
"eval_precision": 0.7702127659574468,
|
| 299 |
+
"eval_recall": 0.8032544378698225,
|
| 300 |
+
"eval_runtime": 1.6214,
|
| 301 |
+
"eval_samples_per_second": 210.313,
|
| 302 |
+
"eval_steps_per_second": 3.701,
|
| 303 |
+
"step": 550
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.686046511627907,
|
| 307 |
+
"grad_norm": 0.30753573775291443,
|
| 308 |
+
"learning_rate": 3.645470060140278e-05,
|
| 309 |
+
"loss": 0.005743271708488464,
|
| 310 |
+
"step": 575
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.976744186046512,
|
| 314 |
+
"grad_norm": 0.13237181305885315,
|
| 315 |
+
"learning_rate": 3.4838766958683304e-05,
|
| 316 |
+
"loss": 0.006215065121650696,
|
| 317 |
+
"step": 600
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.976744186046512,
|
| 321 |
+
"eval_accuracy": 0.9447668171072426,
|
| 322 |
+
"eval_f1": 0.7789934354485777,
|
| 323 |
+
"eval_loss": 0.062308527529239655,
|
| 324 |
+
"eval_precision": 0.7683453237410072,
|
| 325 |
+
"eval_recall": 0.7899408284023669,
|
| 326 |
+
"eval_runtime": 1.6197,
|
| 327 |
+
"eval_samples_per_second": 210.528,
|
| 328 |
+
"eval_steps_per_second": 3.704,
|
| 329 |
+
"step": 600
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 7.267441860465116,
|
| 333 |
+
"grad_norm": 0.023809686303138733,
|
| 334 |
+
"learning_rate": 3.317260597263932e-05,
|
| 335 |
+
"loss": 0.002318141460418701,
|
| 336 |
+
"step": 625
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.558139534883721,
|
| 340 |
+
"grad_norm": 0.03782212734222412,
|
| 341 |
+
"learning_rate": 3.146472346927845e-05,
|
| 342 |
+
"loss": 0.003790220320224762,
|
| 343 |
+
"step": 650
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 7.558139534883721,
|
| 347 |
+
"eval_accuracy": 0.9529335912314636,
|
| 348 |
+
"eval_f1": 0.8062827225130891,
|
| 349 |
+
"eval_loss": 0.07683796435594559,
|
| 350 |
+
"eval_precision": 0.8154311649016641,
|
| 351 |
+
"eval_recall": 0.7973372781065089,
|
| 352 |
+
"eval_runtime": 1.6235,
|
| 353 |
+
"eval_samples_per_second": 210.043,
|
| 354 |
+
"eval_steps_per_second": 3.696,
|
| 355 |
+
"step": 650
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 7.848837209302325,
|
| 359 |
+
"grad_norm": 0.037971507757902145,
|
| 360 |
+
"learning_rate": 2.9723838264811545e-05,
|
| 361 |
+
"loss": 0.0018679146468639374,
|
| 362 |
+
"step": 675
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 8.13953488372093,
|
| 366 |
+
"grad_norm": 0.6083823442459106,
|
| 367 |
+
"learning_rate": 2.7958837655707814e-05,
|
| 368 |
+
"loss": 0.0019237904250621796,
|
| 369 |
+
"step": 700
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 8.13953488372093,
|
| 373 |
+
"eval_accuracy": 0.9565871480765098,
|
| 374 |
+
"eval_f1": 0.8152091254752851,
|
| 375 |
+
"eval_loss": 0.08732757717370987,
|
| 376 |
+
"eval_precision": 0.838810641627543,
|
| 377 |
+
"eval_recall": 0.7928994082840237,
|
| 378 |
+
"eval_runtime": 1.6124,
|
| 379 |
+
"eval_samples_per_second": 211.484,
|
| 380 |
+
"eval_steps_per_second": 3.721,
|
| 381 |
+
"step": 700
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.430232558139535,
|
| 385 |
+
"grad_norm": 0.022007286548614502,
|
| 386 |
+
"learning_rate": 2.6178732048650694e-05,
|
| 387 |
+
"loss": 0.0010710173845291137,
|
| 388 |
+
"step": 725
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 8.720930232558139,
|
| 392 |
+
"grad_norm": 0.09439104795455933,
|
| 393 |
+
"learning_rate": 2.4392608962010652e-05,
|
| 394 |
+
"loss": 0.0006907544285058976,
|
| 395 |
+
"step": 750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.720930232558139,
|
| 399 |
+
"eval_accuracy": 0.9449817322157748,
|
| 400 |
+
"eval_f1": 0.7829848594087959,
|
| 401 |
+
"eval_loss": 0.08308640867471695,
|
| 402 |
+
"eval_precision": 0.7637130801687764,
|
| 403 |
+
"eval_recall": 0.8032544378698225,
|
| 404 |
+
"eval_runtime": 1.6359,
|
| 405 |
+
"eval_samples_per_second": 208.445,
|
| 406 |
+
"eval_steps_per_second": 3.668,
|
| 407 |
+
"step": 750
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 9.011627906976743,
|
| 411 |
+
"grad_norm": 0.018290311098098755,
|
| 412 |
+
"learning_rate": 2.2609586633659256e-05,
|
| 413 |
+
"loss": 0.0005397907271981239,
|
| 414 |
+
"step": 775
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 9.30232558139535,
|
| 418 |
+
"grad_norm": 0.002839206252247095,
|
| 419 |
+
"learning_rate": 2.0838767471959014e-05,
|
| 420 |
+
"loss": 0.0002632186934351921,
|
| 421 |
+
"step": 800
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 9.30232558139535,
|
| 425 |
+
"eval_accuracy": 0.9544379969911885,
|
| 426 |
+
"eval_f1": 0.8081880212282032,
|
| 427 |
+
"eval_loss": 0.09754534065723419,
|
| 428 |
+
"eval_precision": 0.8289269051321928,
|
| 429 |
+
"eval_recall": 0.7884615384615384,
|
| 430 |
+
"eval_runtime": 1.6188,
|
| 431 |
+
"eval_samples_per_second": 210.653,
|
| 432 |
+
"eval_steps_per_second": 3.706,
|
| 433 |
+
"step": 800
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 9.593023255813954,
|
| 437 |
+
"grad_norm": 0.04134448245167732,
|
| 438 |
+
"learning_rate": 1.9089191587563417e-05,
|
| 439 |
+
"loss": 0.0005076020956039428,
|
| 440 |
+
"step": 825
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 9.883720930232558,
|
| 444 |
+
"grad_norm": 0.08968573063611984,
|
| 445 |
+
"learning_rate": 1.7369790643249573e-05,
|
| 446 |
+
"loss": 0.0005874237418174743,
|
| 447 |
+
"step": 850
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 9.883720930232558,
|
| 451 |
+
"eval_accuracy": 0.9548678272082527,
|
| 452 |
+
"eval_f1": 0.8093797276853252,
|
| 453 |
+
"eval_loss": 0.09803693741559982,
|
| 454 |
+
"eval_precision": 0.8281733746130031,
|
| 455 |
+
"eval_recall": 0.7914201183431953,
|
| 456 |
+
"eval_runtime": 1.6209,
|
| 457 |
+
"eval_samples_per_second": 210.379,
|
| 458 |
+
"eval_steps_per_second": 3.702,
|
| 459 |
+
"step": 850
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 10.174418604651162,
|
| 463 |
+
"grad_norm": 0.02889181859791279,
|
| 464 |
+
"learning_rate": 1.5689342257382207e-05,
|
| 465 |
+
"loss": 0.0004838050529360771,
|
| 466 |
+
"step": 875
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.465116279069768,
|
| 470 |
+
"grad_norm": 0.020970497280359268,
|
| 471 |
+
"learning_rate": 1.4056425193781048e-05,
|
| 472 |
+
"loss": 0.0001906752586364746,
|
| 473 |
+
"step": 900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 10.465116279069768,
|
| 477 |
+
"eval_accuracy": 0.9587362991618311,
|
| 478 |
+
"eval_f1": 0.8228043143297381,
|
| 479 |
+
"eval_loss": 0.10382715612649918,
|
| 480 |
+
"eval_precision": 0.8585209003215434,
|
| 481 |
+
"eval_recall": 0.7899408284023669,
|
| 482 |
+
"eval_runtime": 1.6341,
|
| 483 |
+
"eval_samples_per_second": 208.675,
|
| 484 |
+
"eval_steps_per_second": 3.672,
|
| 485 |
+
"step": 900
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 10.755813953488373,
|
| 489 |
+
"grad_norm": 0.028520096093416214,
|
| 490 |
+
"learning_rate": 1.2479375566749694e-05,
|
| 491 |
+
"loss": 0.00012492083944380284,
|
| 492 |
+
"step": 925
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 11.046511627906977,
|
| 496 |
+
"grad_norm": 0.03838531672954559,
|
| 497 |
+
"learning_rate": 1.0966244284840926e-05,
|
| 498 |
+
"loss": 0.00028255857527256014,
|
| 499 |
+
"step": 950
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 11.046511627906977,
|
| 503 |
+
"eval_accuracy": 0.9580915538362347,
|
| 504 |
+
"eval_f1": 0.819369715603382,
|
| 505 |
+
"eval_loss": 0.10540532320737839,
|
| 506 |
+
"eval_precision": 0.8528,
|
| 507 |
+
"eval_recall": 0.7884615384615384,
|
| 508 |
+
"eval_runtime": 1.6495,
|
| 509 |
+
"eval_samples_per_second": 206.73,
|
| 510 |
+
"eval_steps_per_second": 3.637,
|
| 511 |
+
"step": 950
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 11.337209302325581,
|
| 515 |
+
"grad_norm": 0.05398673936724663,
|
| 516 |
+
"learning_rate": 9.524755950610204e-06,
|
| 517 |
+
"loss": 0.00014415644109249114,
|
| 518 |
+
"step": 975
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 11.627906976744185,
|
| 522 |
+
"grad_norm": 0.026561638340353966,
|
| 523 |
+
"learning_rate": 8.162269426175681e-06,
|
| 524 |
+
"loss": 9.150288999080658e-05,
|
| 525 |
+
"step": 1000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 11.627906976744185,
|
| 529 |
+
"eval_accuracy": 0.9574468085106383,
|
| 530 |
+
"eval_f1": 0.8180428134556575,
|
| 531 |
+
"eval_loss": 0.10544126480817795,
|
| 532 |
+
"eval_precision": 0.8465189873417721,
|
| 533 |
+
"eval_recall": 0.7914201183431953,
|
| 534 |
+
"eval_runtime": 1.6274,
|
| 535 |
+
"eval_samples_per_second": 209.539,
|
| 536 |
+
"eval_steps_per_second": 3.687,
|
| 537 |
+
"step": 1000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 11.918604651162791,
|
| 541 |
+
"grad_norm": 0.010734460316598415,
|
| 542 |
+
"learning_rate": 6.885740265899526e-06,
|
| 543 |
+
"loss": 4.033456556499005e-05,
|
| 544 |
+
"step": 1025
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 12.209302325581396,
|
| 548 |
+
"grad_norm": 0.02024712599813938,
|
| 549 |
+
"learning_rate": 5.701685207973243e-06,
|
| 550 |
+
"loss": 6.52561942115426e-05,
|
| 551 |
+
"step": 1050
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 12.209302325581396,
|
| 555 |
+
"eval_accuracy": 0.9576617236191705,
|
| 556 |
+
"eval_f1": 0.8190184049079755,
|
| 557 |
+
"eval_loss": 0.10666228085756302,
|
| 558 |
+
"eval_precision": 0.8503184713375797,
|
| 559 |
+
"eval_recall": 0.7899408284023669,
|
| 560 |
+
"eval_runtime": 1.6269,
|
| 561 |
+
"eval_samples_per_second": 209.598,
|
| 562 |
+
"eval_steps_per_second": 3.688,
|
| 563 |
+
"step": 1050
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 12.5,
|
| 567 |
+
"grad_norm": 0.034469809383153915,
|
| 568 |
+
"learning_rate": 4.616148906179082e-06,
|
| 569 |
+
"loss": 7.551711052656174e-05,
|
| 570 |
+
"step": 1075
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 12.790697674418604,
|
| 574 |
+
"grad_norm": 0.004050907213240862,
|
| 575 |
+
"learning_rate": 3.6346730716634025e-06,
|
| 576 |
+
"loss": 7.49318953603506e-05,
|
| 577 |
+
"step": 1100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 12.790697674418604,
|
| 581 |
+
"eval_accuracy": 0.9565871480765098,
|
| 582 |
+
"eval_f1": 0.8146453089244853,
|
| 583 |
+
"eval_loss": 0.10619472712278366,
|
| 584 |
+
"eval_precision": 0.8409448818897638,
|
| 585 |
+
"eval_recall": 0.7899408284023669,
|
| 586 |
+
"eval_runtime": 1.6196,
|
| 587 |
+
"eval_samples_per_second": 210.546,
|
| 588 |
+
"eval_steps_per_second": 3.705,
|
| 589 |
+
"step": 1100
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 13.081395348837209,
|
| 593 |
+
"grad_norm": 0.11917345970869064,
|
| 594 |
+
"learning_rate": 2.7622681822545764e-06,
|
| 595 |
+
"loss": 8.815748617053032e-05,
|
| 596 |
+
"step": 1125
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 13.372093023255815,
|
| 600 |
+
"grad_norm": 0.02682509273290634,
|
| 601 |
+
"learning_rate": 2.0033879037506004e-06,
|
| 602 |
+
"loss": 6.015756167471409e-05,
|
| 603 |
+
"step": 1150
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 13.372093023255815,
|
| 607 |
+
"eval_accuracy": 0.9583064689447668,
|
| 608 |
+
"eval_f1": 0.8205521472392638,
|
| 609 |
+
"eval_loss": 0.10740578919649124,
|
| 610 |
+
"eval_precision": 0.8519108280254777,
|
| 611 |
+
"eval_recall": 0.7914201183431953,
|
| 612 |
+
"eval_runtime": 1.6217,
|
| 613 |
+
"eval_samples_per_second": 210.273,
|
| 614 |
+
"eval_steps_per_second": 3.7,
|
| 615 |
+
"step": 1150
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 13.662790697674419,
|
| 619 |
+
"grad_norm": 0.004100531339645386,
|
| 620 |
+
"learning_rate": 1.361906353756917e-06,
|
| 621 |
+
"loss": 4.690625239163637e-05,
|
| 622 |
+
"step": 1175
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 13.953488372093023,
|
| 626 |
+
"grad_norm": 0.008385015651583672,
|
| 627 |
+
"learning_rate": 8.410983241436132e-07,
|
| 628 |
+
"loss": 6.560925394296646e-05,
|
| 629 |
+
"step": 1200
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 13.953488372093023,
|
| 633 |
+
"eval_accuracy": 0.9578766387277026,
|
| 634 |
+
"eval_f1": 0.8181120491174213,
|
| 635 |
+
"eval_loss": 0.10830119252204895,
|
| 636 |
+
"eval_precision": 0.8500797448165869,
|
| 637 |
+
"eval_recall": 0.7884615384615384,
|
| 638 |
+
"eval_runtime": 1.6315,
|
| 639 |
+
"eval_samples_per_second": 209.012,
|
| 640 |
+
"eval_steps_per_second": 3.678,
|
| 641 |
+
"step": 1200
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 14.244186046511627,
|
| 645 |
+
"grad_norm": 0.003520221682265401,
|
| 646 |
+
"learning_rate": 4.436225630870927e-07,
|
| 647 |
+
"loss": 0.00012058223597705364,
|
| 648 |
+
"step": 1225
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 14.534883720930232,
|
| 652 |
+
"grad_norm": 0.00905795767903328,
|
| 653 |
+
"learning_rate": 1.7150820204230865e-07,
|
| 654 |
+
"loss": 8.818465285003186e-05,
|
| 655 |
+
"step": 1250
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 14.534883720930232,
|
| 659 |
+
"eval_accuracy": 0.9578766387277026,
|
| 660 |
+
"eval_f1": 0.8187403993855606,
|
| 661 |
+
"eval_loss": 0.10815286636352539,
|
| 662 |
+
"eval_precision": 0.8514376996805112,
|
| 663 |
+
"eval_recall": 0.7884615384615384,
|
| 664 |
+
"eval_runtime": 1.6307,
|
| 665 |
+
"eval_samples_per_second": 209.118,
|
| 666 |
+
"eval_steps_per_second": 3.68,
|
| 667 |
+
"step": 1250
|
| 668 |
+
}
|
| 669 |
+
],
|
| 670 |
+
"logging_steps": 25,
|
| 671 |
+
"max_steps": 1290,
|
| 672 |
+
"num_input_tokens_seen": 0,
|
| 673 |
+
"num_train_epochs": 15,
|
| 674 |
+
"save_steps": 50,
|
| 675 |
+
"stateful_callbacks": {
|
| 676 |
+
"EarlyStoppingCallback": {
|
| 677 |
+
"args": {
|
| 678 |
+
"early_stopping_patience": 10,
|
| 679 |
+
"early_stopping_threshold": 0.0
|
| 680 |
+
},
|
| 681 |
+
"attributes": {
|
| 682 |
+
"early_stopping_patience_counter": 7
|
| 683 |
+
}
|
| 684 |
+
},
|
| 685 |
+
"TrainerControl": {
|
| 686 |
+
"args": {
|
| 687 |
+
"should_epoch_stop": false,
|
| 688 |
+
"should_evaluate": false,
|
| 689 |
+
"should_log": false,
|
| 690 |
+
"should_save": true,
|
| 691 |
+
"should_training_stop": false
|
| 692 |
+
},
|
| 693 |
+
"attributes": {}
|
| 694 |
+
}
|
| 695 |
+
},
|
| 696 |
+
"total_flos": 6779518344422400.0,
|
| 697 |
+
"train_batch_size": 16,
|
| 698 |
+
"trial_name": null,
|
| 699 |
+
"trial_params": null
|
| 700 |
+
}
|
checkpoint-1250/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5071855a54d36018bc5c30ca34d9fef42b69fb3deda057af4ffc70193237d5c4
|
| 3 |
+
size 5201
|
checkpoint-1290/config.json
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.4,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"gradient_checkpointing": false,
|
| 20 |
+
"hidden_activation": "gelu",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "O",
|
| 24 |
+
"1": "FP",
|
| 25 |
+
"2": "RP",
|
| 26 |
+
"3": "RV",
|
| 27 |
+
"4": "PW"
|
| 28 |
+
},
|
| 29 |
+
"initializer_cutoff_factor": 2.0,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 1152,
|
| 32 |
+
"label2id": {
|
| 33 |
+
"FP": 1,
|
| 34 |
+
"O": 0,
|
| 35 |
+
"PW": 4,
|
| 36 |
+
"RP": 2,
|
| 37 |
+
"RV": 3
|
| 38 |
+
},
|
| 39 |
+
"layer_norm_eps": 1e-05,
|
| 40 |
+
"layer_types": [
|
| 41 |
+
"full_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"full_attention"
|
| 63 |
+
],
|
| 64 |
+
"local_attention": 128,
|
| 65 |
+
"max_position_embeddings": 8192,
|
| 66 |
+
"mlp_bias": false,
|
| 67 |
+
"mlp_dropout": 0.0,
|
| 68 |
+
"model_type": "modernbert",
|
| 69 |
+
"norm_bias": false,
|
| 70 |
+
"norm_eps": 1e-05,
|
| 71 |
+
"num_attention_heads": 12,
|
| 72 |
+
"num_hidden_layers": 22,
|
| 73 |
+
"pad_token_id": 50283,
|
| 74 |
+
"position_embedding_type": "absolute",
|
| 75 |
+
"repad_logits_with_grad": false,
|
| 76 |
+
"rope_parameters": {
|
| 77 |
+
"full_attention": {
|
| 78 |
+
"rope_theta": 160000.0,
|
| 79 |
+
"rope_type": "default"
|
| 80 |
+
},
|
| 81 |
+
"sliding_attention": {
|
| 82 |
+
"rope_theta": 10000.0,
|
| 83 |
+
"rope_type": "default"
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"sep_token_id": 50282,
|
| 87 |
+
"sparse_pred_ignore_index": -100,
|
| 88 |
+
"sparse_prediction": false,
|
| 89 |
+
"tie_word_embeddings": true,
|
| 90 |
+
"transformers_version": "5.0.0",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"vocab_size": 50368
|
| 93 |
+
}
|
checkpoint-1290/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcf0010171b4f8ae74bc0a9404acd43c43af4d80dddb7e09b0c0b6869ef892b4
|
| 3 |
+
size 598449012
|
checkpoint-1290/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2bc640cb921976850ac3b758c0e4e59bbdcd79d21ee8a30628d399276e70f72b
|
| 3 |
+
size 535150859
|
checkpoint-1290/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6cf1a66cc75ea89e6d13a532e3309a73f39df97be9642d5bb258af4a75b423c6
|
| 3 |
+
size 14709
|
checkpoint-1290/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a746d7e4a1d22729f55d871883d7524e4414d4d2354d4dc07d961fc5518e6046
|
| 3 |
+
size 1465
|
checkpoint-1290/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1290/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"mask_token": "[MASK]",
|
| 7 |
+
"model_input_names": [
|
| 8 |
+
"input_ids",
|
| 9 |
+
"attention_mask"
|
| 10 |
+
],
|
| 11 |
+
"model_max_length": 8192,
|
| 12 |
+
"pad_token": "[PAD]",
|
| 13 |
+
"sep_token": "[SEP]",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend",
|
| 15 |
+
"unk_token": "[UNK]"
|
| 16 |
+
}
|
checkpoint-1290/trainer_state.json
ADDED
|
@@ -0,0 +1,707 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 900,
|
| 3 |
+
"best_metric": 0.8228043143297381,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/expA_realonly/checkpoint-900",
|
| 5 |
+
"epoch": 15.0,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 1290,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011627906976744186,
|
| 14 |
+
"grad_norm": 4.9740142822265625,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.317932665348053,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.29069767441860467,
|
| 21 |
+
"grad_norm": 1.5566426515579224,
|
| 22 |
+
"learning_rate": 6.282722513089005e-06,
|
| 23 |
+
"loss": 0.3199571371078491,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5813953488372093,
|
| 28 |
+
"grad_norm": 1.6379677057266235,
|
| 29 |
+
"learning_rate": 1.2827225130890053e-05,
|
| 30 |
+
"loss": 0.24125898361206055,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5813953488372093,
|
| 35 |
+
"eval_accuracy": 0.8287126584998925,
|
| 36 |
+
"eval_f1": 0.08611410118406888,
|
| 37 |
+
"eval_loss": 0.10772959887981415,
|
| 38 |
+
"eval_precision": 0.15810276679841898,
|
| 39 |
+
"eval_recall": 0.05917159763313609,
|
| 40 |
+
"eval_runtime": 4.2273,
|
| 41 |
+
"eval_samples_per_second": 80.667,
|
| 42 |
+
"eval_steps_per_second": 1.419,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.872093023255814,
|
| 47 |
+
"grad_norm": 8.703161239624023,
|
| 48 |
+
"learning_rate": 1.93717277486911e-05,
|
| 49 |
+
"loss": 0.20946041107177735,
|
| 50 |
+
"step": 75
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.1627906976744187,
|
| 54 |
+
"grad_norm": 0.9212121963500977,
|
| 55 |
+
"learning_rate": 2.591623036649215e-05,
|
| 56 |
+
"loss": 0.19275867462158203,
|
| 57 |
+
"step": 100
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.1627906976744187,
|
| 61 |
+
"eval_accuracy": 0.7635933806146572,
|
| 62 |
+
"eval_f1": 0.3981945837512538,
|
| 63 |
+
"eval_loss": 0.07719173282384872,
|
| 64 |
+
"eval_precision": 0.30121396054628224,
|
| 65 |
+
"eval_recall": 0.5872781065088757,
|
| 66 |
+
"eval_runtime": 1.6016,
|
| 67 |
+
"eval_samples_per_second": 212.909,
|
| 68 |
+
"eval_steps_per_second": 3.746,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.4534883720930232,
|
| 73 |
+
"grad_norm": 1.46477210521698,
|
| 74 |
+
"learning_rate": 3.246073298429319e-05,
|
| 75 |
+
"loss": 0.14603865623474122,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.744186046511628,
|
| 80 |
+
"grad_norm": 3.729416608810425,
|
| 81 |
+
"learning_rate": 3.900523560209424e-05,
|
| 82 |
+
"loss": 0.13855588912963868,
|
| 83 |
+
"step": 150
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.744186046511628,
|
| 87 |
+
"eval_accuracy": 0.81130453470879,
|
| 88 |
+
"eval_f1": 0.46785521339816316,
|
| 89 |
+
"eval_loss": 0.06223325803875923,
|
| 90 |
+
"eval_precision": 0.36851063829787234,
|
| 91 |
+
"eval_recall": 0.6405325443786982,
|
| 92 |
+
"eval_runtime": 1.5906,
|
| 93 |
+
"eval_samples_per_second": 214.385,
|
| 94 |
+
"eval_steps_per_second": 3.772,
|
| 95 |
+
"step": 150
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0348837209302326,
|
| 99 |
+
"grad_norm": 1.2140589952468872,
|
| 100 |
+
"learning_rate": 4.554973821989529e-05,
|
| 101 |
+
"loss": 0.09624839782714843,
|
| 102 |
+
"step": 175
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.3255813953488373,
|
| 106 |
+
"grad_norm": 0.875888466835022,
|
| 107 |
+
"learning_rate": 4.999346304613061e-05,
|
| 108 |
+
"loss": 0.0821054744720459,
|
| 109 |
+
"step": 200
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3255813953488373,
|
| 113 |
+
"eval_accuracy": 0.7872340425531915,
|
| 114 |
+
"eval_f1": 0.4847619047619048,
|
| 115 |
+
"eval_loss": 0.04723217338323593,
|
| 116 |
+
"eval_precision": 0.3574438202247191,
|
| 117 |
+
"eval_recall": 0.7529585798816568,
|
| 118 |
+
"eval_runtime": 1.6255,
|
| 119 |
+
"eval_samples_per_second": 209.777,
|
| 120 |
+
"eval_steps_per_second": 3.691,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.616279069767442,
|
| 125 |
+
"grad_norm": 1.0979472398757935,
|
| 126 |
+
"learning_rate": 4.98888472605987e-05,
|
| 127 |
+
"loss": 0.08414460182189941,
|
| 128 |
+
"step": 225
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.9069767441860463,
|
| 132 |
+
"grad_norm": 0.49558770656585693,
|
| 133 |
+
"learning_rate": 4.965717280180432e-05,
|
| 134 |
+
"loss": 0.08369583129882813,
|
| 135 |
+
"step": 250
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.9069767441860463,
|
| 139 |
+
"eval_accuracy": 0.6211046636578551,
|
| 140 |
+
"eval_f1": 0.3483915126625599,
|
| 141 |
+
"eval_loss": 0.045315902680158615,
|
| 142 |
+
"eval_precision": 0.22662511130899377,
|
| 143 |
+
"eval_recall": 0.7529585798816568,
|
| 144 |
+
"eval_runtime": 1.6274,
|
| 145 |
+
"eval_samples_per_second": 209.539,
|
| 146 |
+
"eval_steps_per_second": 3.687,
|
| 147 |
+
"step": 250
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.197674418604651,
|
| 151 |
+
"grad_norm": 0.8114362955093384,
|
| 152 |
+
"learning_rate": 4.9299622378173246e-05,
|
| 153 |
+
"loss": 0.05607881069183349,
|
| 154 |
+
"step": 275
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.488372093023256,
|
| 158 |
+
"grad_norm": 0.6278374195098877,
|
| 159 |
+
"learning_rate": 4.881802130053548e-05,
|
| 160 |
+
"loss": 0.043469696044921874,
|
| 161 |
+
"step": 300
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 3.488372093023256,
|
| 165 |
+
"eval_accuracy": 0.8082957231893402,
|
| 166 |
+
"eval_f1": 0.5285784074255008,
|
| 167 |
+
"eval_loss": 0.04308881610631943,
|
| 168 |
+
"eval_precision": 0.39460247994164843,
|
| 169 |
+
"eval_recall": 0.8002958579881657,
|
| 170 |
+
"eval_runtime": 1.6193,
|
| 171 |
+
"eval_samples_per_second": 210.585,
|
| 172 |
+
"eval_steps_per_second": 3.705,
|
| 173 |
+
"step": 300
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.7790697674418605,
|
| 177 |
+
"grad_norm": 1.0219486951828003,
|
| 178 |
+
"learning_rate": 4.821482816383218e-05,
|
| 179 |
+
"loss": 0.047760300636291504,
|
| 180 |
+
"step": 325
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.069767441860465,
|
| 184 |
+
"grad_norm": 0.509274959564209,
|
| 185 |
+
"learning_rate": 4.749312229587908e-05,
|
| 186 |
+
"loss": 0.03177810907363891,
|
| 187 |
+
"step": 350
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 4.069767441860465,
|
| 191 |
+
"eval_accuracy": 0.9196217494089834,
|
| 192 |
+
"eval_f1": 0.7152317880794702,
|
| 193 |
+
"eval_loss": 0.04378104954957962,
|
| 194 |
+
"eval_precision": 0.6474820143884892,
|
| 195 |
+
"eval_recall": 0.7988165680473372,
|
| 196 |
+
"eval_runtime": 1.6177,
|
| 197 |
+
"eval_samples_per_second": 210.792,
|
| 198 |
+
"eval_steps_per_second": 3.709,
|
| 199 |
+
"step": 350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.3604651162790695,
|
| 203 |
+
"grad_norm": 0.47098585963249207,
|
| 204 |
+
"learning_rate": 4.6656588037260834e-05,
|
| 205 |
+
"loss": 0.020146708488464355,
|
| 206 |
+
"step": 375
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.651162790697675,
|
| 210 |
+
"grad_norm": 0.7534321546554565,
|
| 211 |
+
"learning_rate": 4.570949593260837e-05,
|
| 212 |
+
"loss": 0.018921481370925905,
|
| 213 |
+
"step": 400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.651162790697675,
|
| 217 |
+
"eval_accuracy": 0.9097356544165055,
|
| 218 |
+
"eval_f1": 0.6987341772151898,
|
| 219 |
+
"eval_loss": 0.05498046055436134,
|
| 220 |
+
"eval_precision": 0.6106194690265486,
|
| 221 |
+
"eval_recall": 0.8165680473372781,
|
| 222 |
+
"eval_runtime": 1.6157,
|
| 223 |
+
"eval_samples_per_second": 211.05,
|
| 224 |
+
"eval_steps_per_second": 3.713,
|
| 225 |
+
"step": 400
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.941860465116279,
|
| 229 |
+
"grad_norm": 0.7499828934669495,
|
| 230 |
+
"learning_rate": 4.465668092927841e-05,
|
| 231 |
+
"loss": 0.01998784065246582,
|
| 232 |
+
"step": 425
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 5.232558139534884,
|
| 236 |
+
"grad_norm": 0.24809418618679047,
|
| 237 |
+
"learning_rate": 4.350351769473198e-05,
|
| 238 |
+
"loss": 0.01218403697013855,
|
| 239 |
+
"step": 450
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 5.232558139534884,
|
| 243 |
+
"eval_accuracy": 0.9084461637653127,
|
| 244 |
+
"eval_f1": 0.6831746031746032,
|
| 245 |
+
"eval_loss": 0.05535344406962395,
|
| 246 |
+
"eval_precision": 0.5984427141268076,
|
| 247 |
+
"eval_recall": 0.7958579881656804,
|
| 248 |
+
"eval_runtime": 1.6123,
|
| 249 |
+
"eval_samples_per_second": 211.498,
|
| 250 |
+
"eval_steps_per_second": 3.721,
|
| 251 |
+
"step": 450
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 5.523255813953488,
|
| 255 |
+
"grad_norm": 0.3144132196903229,
|
| 256 |
+
"learning_rate": 4.2255893178617745e-05,
|
| 257 |
+
"loss": 0.012306832075119019,
|
| 258 |
+
"step": 475
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 5.813953488372093,
|
| 262 |
+
"grad_norm": 0.1607646942138672,
|
| 263 |
+
"learning_rate": 4.092017655963198e-05,
|
| 264 |
+
"loss": 0.010539753437042236,
|
| 265 |
+
"step": 500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 5.813953488372093,
|
| 269 |
+
"eval_accuracy": 0.9527186761229315,
|
| 270 |
+
"eval_f1": 0.8059040590405904,
|
| 271 |
+
"eval_loss": 0.07191809266805649,
|
| 272 |
+
"eval_precision": 0.8041237113402062,
|
| 273 |
+
"eval_recall": 0.8076923076923077,
|
| 274 |
+
"eval_runtime": 1.6569,
|
| 275 |
+
"eval_samples_per_second": 205.806,
|
| 276 |
+
"eval_steps_per_second": 3.621,
|
| 277 |
+
"step": 500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.104651162790698,
|
| 281 |
+
"grad_norm": 0.19801200926303864,
|
| 282 |
+
"learning_rate": 3.9503186730577816e-05,
|
| 283 |
+
"loss": 0.008691226243972778,
|
| 284 |
+
"step": 525
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.395348837209302,
|
| 288 |
+
"grad_norm": 0.4177681505680084,
|
| 289 |
+
"learning_rate": 3.801215748761385e-05,
|
| 290 |
+
"loss": 0.004750225245952606,
|
| 291 |
+
"step": 550
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 6.395348837209302,
|
| 295 |
+
"eval_accuracy": 0.9458413926499033,
|
| 296 |
+
"eval_f1": 0.7863866763215062,
|
| 297 |
+
"eval_loss": 0.07378304749727249,
|
| 298 |
+
"eval_precision": 0.7702127659574468,
|
| 299 |
+
"eval_recall": 0.8032544378698225,
|
| 300 |
+
"eval_runtime": 1.6214,
|
| 301 |
+
"eval_samples_per_second": 210.313,
|
| 302 |
+
"eval_steps_per_second": 3.701,
|
| 303 |
+
"step": 550
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.686046511627907,
|
| 307 |
+
"grad_norm": 0.30753573775291443,
|
| 308 |
+
"learning_rate": 3.645470060140278e-05,
|
| 309 |
+
"loss": 0.005743271708488464,
|
| 310 |
+
"step": 575
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.976744186046512,
|
| 314 |
+
"grad_norm": 0.13237181305885315,
|
| 315 |
+
"learning_rate": 3.4838766958683304e-05,
|
| 316 |
+
"loss": 0.006215065121650696,
|
| 317 |
+
"step": 600
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.976744186046512,
|
| 321 |
+
"eval_accuracy": 0.9447668171072426,
|
| 322 |
+
"eval_f1": 0.7789934354485777,
|
| 323 |
+
"eval_loss": 0.062308527529239655,
|
| 324 |
+
"eval_precision": 0.7683453237410072,
|
| 325 |
+
"eval_recall": 0.7899408284023669,
|
| 326 |
+
"eval_runtime": 1.6197,
|
| 327 |
+
"eval_samples_per_second": 210.528,
|
| 328 |
+
"eval_steps_per_second": 3.704,
|
| 329 |
+
"step": 600
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 7.267441860465116,
|
| 333 |
+
"grad_norm": 0.023809686303138733,
|
| 334 |
+
"learning_rate": 3.317260597263932e-05,
|
| 335 |
+
"loss": 0.002318141460418701,
|
| 336 |
+
"step": 625
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.558139534883721,
|
| 340 |
+
"grad_norm": 0.03782212734222412,
|
| 341 |
+
"learning_rate": 3.146472346927845e-05,
|
| 342 |
+
"loss": 0.003790220320224762,
|
| 343 |
+
"step": 650
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 7.558139534883721,
|
| 347 |
+
"eval_accuracy": 0.9529335912314636,
|
| 348 |
+
"eval_f1": 0.8062827225130891,
|
| 349 |
+
"eval_loss": 0.07683796435594559,
|
| 350 |
+
"eval_precision": 0.8154311649016641,
|
| 351 |
+
"eval_recall": 0.7973372781065089,
|
| 352 |
+
"eval_runtime": 1.6235,
|
| 353 |
+
"eval_samples_per_second": 210.043,
|
| 354 |
+
"eval_steps_per_second": 3.696,
|
| 355 |
+
"step": 650
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 7.848837209302325,
|
| 359 |
+
"grad_norm": 0.037971507757902145,
|
| 360 |
+
"learning_rate": 2.9723838264811545e-05,
|
| 361 |
+
"loss": 0.0018679146468639374,
|
| 362 |
+
"step": 675
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 8.13953488372093,
|
| 366 |
+
"grad_norm": 0.6083823442459106,
|
| 367 |
+
"learning_rate": 2.7958837655707814e-05,
|
| 368 |
+
"loss": 0.0019237904250621796,
|
| 369 |
+
"step": 700
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 8.13953488372093,
|
| 373 |
+
"eval_accuracy": 0.9565871480765098,
|
| 374 |
+
"eval_f1": 0.8152091254752851,
|
| 375 |
+
"eval_loss": 0.08732757717370987,
|
| 376 |
+
"eval_precision": 0.838810641627543,
|
| 377 |
+
"eval_recall": 0.7928994082840237,
|
| 378 |
+
"eval_runtime": 1.6124,
|
| 379 |
+
"eval_samples_per_second": 211.484,
|
| 380 |
+
"eval_steps_per_second": 3.721,
|
| 381 |
+
"step": 700
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.430232558139535,
|
| 385 |
+
"grad_norm": 0.022007286548614502,
|
| 386 |
+
"learning_rate": 2.6178732048650694e-05,
|
| 387 |
+
"loss": 0.0010710173845291137,
|
| 388 |
+
"step": 725
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 8.720930232558139,
|
| 392 |
+
"grad_norm": 0.09439104795455933,
|
| 393 |
+
"learning_rate": 2.4392608962010652e-05,
|
| 394 |
+
"loss": 0.0006907544285058976,
|
| 395 |
+
"step": 750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.720930232558139,
|
| 399 |
+
"eval_accuracy": 0.9449817322157748,
|
| 400 |
+
"eval_f1": 0.7829848594087959,
|
| 401 |
+
"eval_loss": 0.08308640867471695,
|
| 402 |
+
"eval_precision": 0.7637130801687764,
|
| 403 |
+
"eval_recall": 0.8032544378698225,
|
| 404 |
+
"eval_runtime": 1.6359,
|
| 405 |
+
"eval_samples_per_second": 208.445,
|
| 406 |
+
"eval_steps_per_second": 3.668,
|
| 407 |
+
"step": 750
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 9.011627906976743,
|
| 411 |
+
"grad_norm": 0.018290311098098755,
|
| 412 |
+
"learning_rate": 2.2609586633659256e-05,
|
| 413 |
+
"loss": 0.0005397907271981239,
|
| 414 |
+
"step": 775
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 9.30232558139535,
|
| 418 |
+
"grad_norm": 0.002839206252247095,
|
| 419 |
+
"learning_rate": 2.0838767471959014e-05,
|
| 420 |
+
"loss": 0.0002632186934351921,
|
| 421 |
+
"step": 800
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 9.30232558139535,
|
| 425 |
+
"eval_accuracy": 0.9544379969911885,
|
| 426 |
+
"eval_f1": 0.8081880212282032,
|
| 427 |
+
"eval_loss": 0.09754534065723419,
|
| 428 |
+
"eval_precision": 0.8289269051321928,
|
| 429 |
+
"eval_recall": 0.7884615384615384,
|
| 430 |
+
"eval_runtime": 1.6188,
|
| 431 |
+
"eval_samples_per_second": 210.653,
|
| 432 |
+
"eval_steps_per_second": 3.706,
|
| 433 |
+
"step": 800
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 9.593023255813954,
|
| 437 |
+
"grad_norm": 0.04134448245167732,
|
| 438 |
+
"learning_rate": 1.9089191587563417e-05,
|
| 439 |
+
"loss": 0.0005076020956039428,
|
| 440 |
+
"step": 825
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 9.883720930232558,
|
| 444 |
+
"grad_norm": 0.08968573063611984,
|
| 445 |
+
"learning_rate": 1.7369790643249573e-05,
|
| 446 |
+
"loss": 0.0005874237418174743,
|
| 447 |
+
"step": 850
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 9.883720930232558,
|
| 451 |
+
"eval_accuracy": 0.9548678272082527,
|
| 452 |
+
"eval_f1": 0.8093797276853252,
|
| 453 |
+
"eval_loss": 0.09803693741559982,
|
| 454 |
+
"eval_precision": 0.8281733746130031,
|
| 455 |
+
"eval_recall": 0.7914201183431953,
|
| 456 |
+
"eval_runtime": 1.6209,
|
| 457 |
+
"eval_samples_per_second": 210.379,
|
| 458 |
+
"eval_steps_per_second": 3.702,
|
| 459 |
+
"step": 850
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 10.174418604651162,
|
| 463 |
+
"grad_norm": 0.02889181859791279,
|
| 464 |
+
"learning_rate": 1.5689342257382207e-05,
|
| 465 |
+
"loss": 0.0004838050529360771,
|
| 466 |
+
"step": 875
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.465116279069768,
|
| 470 |
+
"grad_norm": 0.020970497280359268,
|
| 471 |
+
"learning_rate": 1.4056425193781048e-05,
|
| 472 |
+
"loss": 0.0001906752586364746,
|
| 473 |
+
"step": 900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 10.465116279069768,
|
| 477 |
+
"eval_accuracy": 0.9587362991618311,
|
| 478 |
+
"eval_f1": 0.8228043143297381,
|
| 479 |
+
"eval_loss": 0.10382715612649918,
|
| 480 |
+
"eval_precision": 0.8585209003215434,
|
| 481 |
+
"eval_recall": 0.7899408284023669,
|
| 482 |
+
"eval_runtime": 1.6341,
|
| 483 |
+
"eval_samples_per_second": 208.675,
|
| 484 |
+
"eval_steps_per_second": 3.672,
|
| 485 |
+
"step": 900
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 10.755813953488373,
|
| 489 |
+
"grad_norm": 0.028520096093416214,
|
| 490 |
+
"learning_rate": 1.2479375566749694e-05,
|
| 491 |
+
"loss": 0.00012492083944380284,
|
| 492 |
+
"step": 925
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 11.046511627906977,
|
| 496 |
+
"grad_norm": 0.03838531672954559,
|
| 497 |
+
"learning_rate": 1.0966244284840926e-05,
|
| 498 |
+
"loss": 0.00028255857527256014,
|
| 499 |
+
"step": 950
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 11.046511627906977,
|
| 503 |
+
"eval_accuracy": 0.9580915538362347,
|
| 504 |
+
"eval_f1": 0.819369715603382,
|
| 505 |
+
"eval_loss": 0.10540532320737839,
|
| 506 |
+
"eval_precision": 0.8528,
|
| 507 |
+
"eval_recall": 0.7884615384615384,
|
| 508 |
+
"eval_runtime": 1.6495,
|
| 509 |
+
"eval_samples_per_second": 206.73,
|
| 510 |
+
"eval_steps_per_second": 3.637,
|
| 511 |
+
"step": 950
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 11.337209302325581,
|
| 515 |
+
"grad_norm": 0.05398673936724663,
|
| 516 |
+
"learning_rate": 9.524755950610204e-06,
|
| 517 |
+
"loss": 0.00014415644109249114,
|
| 518 |
+
"step": 975
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 11.627906976744185,
|
| 522 |
+
"grad_norm": 0.026561638340353966,
|
| 523 |
+
"learning_rate": 8.162269426175681e-06,
|
| 524 |
+
"loss": 9.150288999080658e-05,
|
| 525 |
+
"step": 1000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 11.627906976744185,
|
| 529 |
+
"eval_accuracy": 0.9574468085106383,
|
| 530 |
+
"eval_f1": 0.8180428134556575,
|
| 531 |
+
"eval_loss": 0.10544126480817795,
|
| 532 |
+
"eval_precision": 0.8465189873417721,
|
| 533 |
+
"eval_recall": 0.7914201183431953,
|
| 534 |
+
"eval_runtime": 1.6274,
|
| 535 |
+
"eval_samples_per_second": 209.539,
|
| 536 |
+
"eval_steps_per_second": 3.687,
|
| 537 |
+
"step": 1000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 11.918604651162791,
|
| 541 |
+
"grad_norm": 0.010734460316598415,
|
| 542 |
+
"learning_rate": 6.885740265899526e-06,
|
| 543 |
+
"loss": 4.033456556499005e-05,
|
| 544 |
+
"step": 1025
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 12.209302325581396,
|
| 548 |
+
"grad_norm": 0.02024712599813938,
|
| 549 |
+
"learning_rate": 5.701685207973243e-06,
|
| 550 |
+
"loss": 6.52561942115426e-05,
|
| 551 |
+
"step": 1050
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 12.209302325581396,
|
| 555 |
+
"eval_accuracy": 0.9576617236191705,
|
| 556 |
+
"eval_f1": 0.8190184049079755,
|
| 557 |
+
"eval_loss": 0.10666228085756302,
|
| 558 |
+
"eval_precision": 0.8503184713375797,
|
| 559 |
+
"eval_recall": 0.7899408284023669,
|
| 560 |
+
"eval_runtime": 1.6269,
|
| 561 |
+
"eval_samples_per_second": 209.598,
|
| 562 |
+
"eval_steps_per_second": 3.688,
|
| 563 |
+
"step": 1050
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 12.5,
|
| 567 |
+
"grad_norm": 0.034469809383153915,
|
| 568 |
+
"learning_rate": 4.616148906179082e-06,
|
| 569 |
+
"loss": 7.551711052656174e-05,
|
| 570 |
+
"step": 1075
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 12.790697674418604,
|
| 574 |
+
"grad_norm": 0.004050907213240862,
|
| 575 |
+
"learning_rate": 3.6346730716634025e-06,
|
| 576 |
+
"loss": 7.49318953603506e-05,
|
| 577 |
+
"step": 1100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 12.790697674418604,
|
| 581 |
+
"eval_accuracy": 0.9565871480765098,
|
| 582 |
+
"eval_f1": 0.8146453089244853,
|
| 583 |
+
"eval_loss": 0.10619472712278366,
|
| 584 |
+
"eval_precision": 0.8409448818897638,
|
| 585 |
+
"eval_recall": 0.7899408284023669,
|
| 586 |
+
"eval_runtime": 1.6196,
|
| 587 |
+
"eval_samples_per_second": 210.546,
|
| 588 |
+
"eval_steps_per_second": 3.705,
|
| 589 |
+
"step": 1100
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 13.081395348837209,
|
| 593 |
+
"grad_norm": 0.11917345970869064,
|
| 594 |
+
"learning_rate": 2.7622681822545764e-06,
|
| 595 |
+
"loss": 8.815748617053032e-05,
|
| 596 |
+
"step": 1125
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 13.372093023255815,
|
| 600 |
+
"grad_norm": 0.02682509273290634,
|
| 601 |
+
"learning_rate": 2.0033879037506004e-06,
|
| 602 |
+
"loss": 6.015756167471409e-05,
|
| 603 |
+
"step": 1150
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 13.372093023255815,
|
| 607 |
+
"eval_accuracy": 0.9583064689447668,
|
| 608 |
+
"eval_f1": 0.8205521472392638,
|
| 609 |
+
"eval_loss": 0.10740578919649124,
|
| 610 |
+
"eval_precision": 0.8519108280254777,
|
| 611 |
+
"eval_recall": 0.7914201183431953,
|
| 612 |
+
"eval_runtime": 1.6217,
|
| 613 |
+
"eval_samples_per_second": 210.273,
|
| 614 |
+
"eval_steps_per_second": 3.7,
|
| 615 |
+
"step": 1150
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 13.662790697674419,
|
| 619 |
+
"grad_norm": 0.004100531339645386,
|
| 620 |
+
"learning_rate": 1.361906353756917e-06,
|
| 621 |
+
"loss": 4.690625239163637e-05,
|
| 622 |
+
"step": 1175
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 13.953488372093023,
|
| 626 |
+
"grad_norm": 0.008385015651583672,
|
| 627 |
+
"learning_rate": 8.410983241436132e-07,
|
| 628 |
+
"loss": 6.560925394296646e-05,
|
| 629 |
+
"step": 1200
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 13.953488372093023,
|
| 633 |
+
"eval_accuracy": 0.9578766387277026,
|
| 634 |
+
"eval_f1": 0.8181120491174213,
|
| 635 |
+
"eval_loss": 0.10830119252204895,
|
| 636 |
+
"eval_precision": 0.8500797448165869,
|
| 637 |
+
"eval_recall": 0.7884615384615384,
|
| 638 |
+
"eval_runtime": 1.6315,
|
| 639 |
+
"eval_samples_per_second": 209.012,
|
| 640 |
+
"eval_steps_per_second": 3.678,
|
| 641 |
+
"step": 1200
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 14.244186046511627,
|
| 645 |
+
"grad_norm": 0.003520221682265401,
|
| 646 |
+
"learning_rate": 4.436225630870927e-07,
|
| 647 |
+
"loss": 0.00012058223597705364,
|
| 648 |
+
"step": 1225
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 14.534883720930232,
|
| 652 |
+
"grad_norm": 0.00905795767903328,
|
| 653 |
+
"learning_rate": 1.7150820204230865e-07,
|
| 654 |
+
"loss": 8.818465285003186e-05,
|
| 655 |
+
"step": 1250
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 14.534883720930232,
|
| 659 |
+
"eval_accuracy": 0.9578766387277026,
|
| 660 |
+
"eval_f1": 0.8187403993855606,
|
| 661 |
+
"eval_loss": 0.10815286636352539,
|
| 662 |
+
"eval_precision": 0.8514376996805112,
|
| 663 |
+
"eval_recall": 0.7884615384615384,
|
| 664 |
+
"eval_runtime": 1.6307,
|
| 665 |
+
"eval_samples_per_second": 209.118,
|
| 666 |
+
"eval_steps_per_second": 3.68,
|
| 667 |
+
"step": 1250
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 14.825581395348838,
|
| 671 |
+
"grad_norm": 0.006475365720689297,
|
| 672 |
+
"learning_rate": 2.6144396936325645e-08,
|
| 673 |
+
"loss": 4.936248995363712e-05,
|
| 674 |
+
"step": 1275
|
| 675 |
+
}
|
| 676 |
+
],
|
| 677 |
+
"logging_steps": 25,
|
| 678 |
+
"max_steps": 1290,
|
| 679 |
+
"num_input_tokens_seen": 0,
|
| 680 |
+
"num_train_epochs": 15,
|
| 681 |
+
"save_steps": 50,
|
| 682 |
+
"stateful_callbacks": {
|
| 683 |
+
"EarlyStoppingCallback": {
|
| 684 |
+
"args": {
|
| 685 |
+
"early_stopping_patience": 10,
|
| 686 |
+
"early_stopping_threshold": 0.0
|
| 687 |
+
},
|
| 688 |
+
"attributes": {
|
| 689 |
+
"early_stopping_patience_counter": 7
|
| 690 |
+
}
|
| 691 |
+
},
|
| 692 |
+
"TrainerControl": {
|
| 693 |
+
"args": {
|
| 694 |
+
"should_epoch_stop": false,
|
| 695 |
+
"should_evaluate": false,
|
| 696 |
+
"should_log": false,
|
| 697 |
+
"should_save": true,
|
| 698 |
+
"should_training_stop": true
|
| 699 |
+
},
|
| 700 |
+
"attributes": {}
|
| 701 |
+
}
|
| 702 |
+
},
|
| 703 |
+
"total_flos": 6995052164620800.0,
|
| 704 |
+
"train_batch_size": 16,
|
| 705 |
+
"trial_name": null,
|
| 706 |
+
"trial_params": null
|
| 707 |
+
}
|
checkpoint-1290/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5071855a54d36018bc5c30ca34d9fef42b69fb3deda057af4ffc70193237d5c4
|
| 3 |
+
size 5201
|
checkpoint-900/config.json
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.4,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"gradient_checkpointing": false,
|
| 20 |
+
"hidden_activation": "gelu",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "O",
|
| 24 |
+
"1": "FP",
|
| 25 |
+
"2": "RP",
|
| 26 |
+
"3": "RV",
|
| 27 |
+
"4": "PW"
|
| 28 |
+
},
|
| 29 |
+
"initializer_cutoff_factor": 2.0,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 1152,
|
| 32 |
+
"label2id": {
|
| 33 |
+
"FP": 1,
|
| 34 |
+
"O": 0,
|
| 35 |
+
"PW": 4,
|
| 36 |
+
"RP": 2,
|
| 37 |
+
"RV": 3
|
| 38 |
+
},
|
| 39 |
+
"layer_norm_eps": 1e-05,
|
| 40 |
+
"layer_types": [
|
| 41 |
+
"full_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"full_attention"
|
| 63 |
+
],
|
| 64 |
+
"local_attention": 128,
|
| 65 |
+
"max_position_embeddings": 8192,
|
| 66 |
+
"mlp_bias": false,
|
| 67 |
+
"mlp_dropout": 0.0,
|
| 68 |
+
"model_type": "modernbert",
|
| 69 |
+
"norm_bias": false,
|
| 70 |
+
"norm_eps": 1e-05,
|
| 71 |
+
"num_attention_heads": 12,
|
| 72 |
+
"num_hidden_layers": 22,
|
| 73 |
+
"pad_token_id": 50283,
|
| 74 |
+
"position_embedding_type": "absolute",
|
| 75 |
+
"repad_logits_with_grad": false,
|
| 76 |
+
"rope_parameters": {
|
| 77 |
+
"full_attention": {
|
| 78 |
+
"rope_theta": 160000.0,
|
| 79 |
+
"rope_type": "default"
|
| 80 |
+
},
|
| 81 |
+
"sliding_attention": {
|
| 82 |
+
"rope_theta": 10000.0,
|
| 83 |
+
"rope_type": "default"
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"sep_token_id": 50282,
|
| 87 |
+
"sparse_pred_ignore_index": -100,
|
| 88 |
+
"sparse_prediction": false,
|
| 89 |
+
"tie_word_embeddings": true,
|
| 90 |
+
"transformers_version": "5.0.0",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"vocab_size": 50368
|
| 93 |
+
}
|
checkpoint-900/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:21c65b89ca04d84973d10cf676f8e009915c6acb1626e669150301db7180e8ea
|
| 3 |
+
size 598449012
|
checkpoint-900/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e4a4f864256c83543f011ceb16a229dd53801f1880b99a72d8c3fe2bcc91b0d
|
| 3 |
+
size 535150859
|
checkpoint-900/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ecd50dbc2959bbba3f714fc9997c16fc15e2f3441b4818a8e36d461bd20caea8
|
| 3 |
+
size 14709
|
checkpoint-900/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69966330a69f7b6a2710b6064c2261635cce76869812d3e21fb60dd080336f56
|
| 3 |
+
size 1465
|
checkpoint-900/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-900/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"mask_token": "[MASK]",
|
| 7 |
+
"model_input_names": [
|
| 8 |
+
"input_ids",
|
| 9 |
+
"attention_mask"
|
| 10 |
+
],
|
| 11 |
+
"model_max_length": 8192,
|
| 12 |
+
"pad_token": "[PAD]",
|
| 13 |
+
"sep_token": "[SEP]",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend",
|
| 15 |
+
"unk_token": "[UNK]"
|
| 16 |
+
}
|
checkpoint-900/trainer_state.json
ADDED
|
@@ -0,0 +1,518 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 900,
|
| 3 |
+
"best_metric": 0.8228043143297381,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/expA_realonly/checkpoint-900",
|
| 5 |
+
"epoch": 10.465116279069768,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 900,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011627906976744186,
|
| 14 |
+
"grad_norm": 4.9740142822265625,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.317932665348053,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.29069767441860467,
|
| 21 |
+
"grad_norm": 1.5566426515579224,
|
| 22 |
+
"learning_rate": 6.282722513089005e-06,
|
| 23 |
+
"loss": 0.3199571371078491,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5813953488372093,
|
| 28 |
+
"grad_norm": 1.6379677057266235,
|
| 29 |
+
"learning_rate": 1.2827225130890053e-05,
|
| 30 |
+
"loss": 0.24125898361206055,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5813953488372093,
|
| 35 |
+
"eval_accuracy": 0.8287126584998925,
|
| 36 |
+
"eval_f1": 0.08611410118406888,
|
| 37 |
+
"eval_loss": 0.10772959887981415,
|
| 38 |
+
"eval_precision": 0.15810276679841898,
|
| 39 |
+
"eval_recall": 0.05917159763313609,
|
| 40 |
+
"eval_runtime": 4.2273,
|
| 41 |
+
"eval_samples_per_second": 80.667,
|
| 42 |
+
"eval_steps_per_second": 1.419,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.872093023255814,
|
| 47 |
+
"grad_norm": 8.703161239624023,
|
| 48 |
+
"learning_rate": 1.93717277486911e-05,
|
| 49 |
+
"loss": 0.20946041107177735,
|
| 50 |
+
"step": 75
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.1627906976744187,
|
| 54 |
+
"grad_norm": 0.9212121963500977,
|
| 55 |
+
"learning_rate": 2.591623036649215e-05,
|
| 56 |
+
"loss": 0.19275867462158203,
|
| 57 |
+
"step": 100
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.1627906976744187,
|
| 61 |
+
"eval_accuracy": 0.7635933806146572,
|
| 62 |
+
"eval_f1": 0.3981945837512538,
|
| 63 |
+
"eval_loss": 0.07719173282384872,
|
| 64 |
+
"eval_precision": 0.30121396054628224,
|
| 65 |
+
"eval_recall": 0.5872781065088757,
|
| 66 |
+
"eval_runtime": 1.6016,
|
| 67 |
+
"eval_samples_per_second": 212.909,
|
| 68 |
+
"eval_steps_per_second": 3.746,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.4534883720930232,
|
| 73 |
+
"grad_norm": 1.46477210521698,
|
| 74 |
+
"learning_rate": 3.246073298429319e-05,
|
| 75 |
+
"loss": 0.14603865623474122,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.744186046511628,
|
| 80 |
+
"grad_norm": 3.729416608810425,
|
| 81 |
+
"learning_rate": 3.900523560209424e-05,
|
| 82 |
+
"loss": 0.13855588912963868,
|
| 83 |
+
"step": 150
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.744186046511628,
|
| 87 |
+
"eval_accuracy": 0.81130453470879,
|
| 88 |
+
"eval_f1": 0.46785521339816316,
|
| 89 |
+
"eval_loss": 0.06223325803875923,
|
| 90 |
+
"eval_precision": 0.36851063829787234,
|
| 91 |
+
"eval_recall": 0.6405325443786982,
|
| 92 |
+
"eval_runtime": 1.5906,
|
| 93 |
+
"eval_samples_per_second": 214.385,
|
| 94 |
+
"eval_steps_per_second": 3.772,
|
| 95 |
+
"step": 150
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0348837209302326,
|
| 99 |
+
"grad_norm": 1.2140589952468872,
|
| 100 |
+
"learning_rate": 4.554973821989529e-05,
|
| 101 |
+
"loss": 0.09624839782714843,
|
| 102 |
+
"step": 175
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.3255813953488373,
|
| 106 |
+
"grad_norm": 0.875888466835022,
|
| 107 |
+
"learning_rate": 4.999346304613061e-05,
|
| 108 |
+
"loss": 0.0821054744720459,
|
| 109 |
+
"step": 200
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3255813953488373,
|
| 113 |
+
"eval_accuracy": 0.7872340425531915,
|
| 114 |
+
"eval_f1": 0.4847619047619048,
|
| 115 |
+
"eval_loss": 0.04723217338323593,
|
| 116 |
+
"eval_precision": 0.3574438202247191,
|
| 117 |
+
"eval_recall": 0.7529585798816568,
|
| 118 |
+
"eval_runtime": 1.6255,
|
| 119 |
+
"eval_samples_per_second": 209.777,
|
| 120 |
+
"eval_steps_per_second": 3.691,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.616279069767442,
|
| 125 |
+
"grad_norm": 1.0979472398757935,
|
| 126 |
+
"learning_rate": 4.98888472605987e-05,
|
| 127 |
+
"loss": 0.08414460182189941,
|
| 128 |
+
"step": 225
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.9069767441860463,
|
| 132 |
+
"grad_norm": 0.49558770656585693,
|
| 133 |
+
"learning_rate": 4.965717280180432e-05,
|
| 134 |
+
"loss": 0.08369583129882813,
|
| 135 |
+
"step": 250
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.9069767441860463,
|
| 139 |
+
"eval_accuracy": 0.6211046636578551,
|
| 140 |
+
"eval_f1": 0.3483915126625599,
|
| 141 |
+
"eval_loss": 0.045315902680158615,
|
| 142 |
+
"eval_precision": 0.22662511130899377,
|
| 143 |
+
"eval_recall": 0.7529585798816568,
|
| 144 |
+
"eval_runtime": 1.6274,
|
| 145 |
+
"eval_samples_per_second": 209.539,
|
| 146 |
+
"eval_steps_per_second": 3.687,
|
| 147 |
+
"step": 250
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.197674418604651,
|
| 151 |
+
"grad_norm": 0.8114362955093384,
|
| 152 |
+
"learning_rate": 4.9299622378173246e-05,
|
| 153 |
+
"loss": 0.05607881069183349,
|
| 154 |
+
"step": 275
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.488372093023256,
|
| 158 |
+
"grad_norm": 0.6278374195098877,
|
| 159 |
+
"learning_rate": 4.881802130053548e-05,
|
| 160 |
+
"loss": 0.043469696044921874,
|
| 161 |
+
"step": 300
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 3.488372093023256,
|
| 165 |
+
"eval_accuracy": 0.8082957231893402,
|
| 166 |
+
"eval_f1": 0.5285784074255008,
|
| 167 |
+
"eval_loss": 0.04308881610631943,
|
| 168 |
+
"eval_precision": 0.39460247994164843,
|
| 169 |
+
"eval_recall": 0.8002958579881657,
|
| 170 |
+
"eval_runtime": 1.6193,
|
| 171 |
+
"eval_samples_per_second": 210.585,
|
| 172 |
+
"eval_steps_per_second": 3.705,
|
| 173 |
+
"step": 300
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.7790697674418605,
|
| 177 |
+
"grad_norm": 1.0219486951828003,
|
| 178 |
+
"learning_rate": 4.821482816383218e-05,
|
| 179 |
+
"loss": 0.047760300636291504,
|
| 180 |
+
"step": 325
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.069767441860465,
|
| 184 |
+
"grad_norm": 0.509274959564209,
|
| 185 |
+
"learning_rate": 4.749312229587908e-05,
|
| 186 |
+
"loss": 0.03177810907363891,
|
| 187 |
+
"step": 350
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 4.069767441860465,
|
| 191 |
+
"eval_accuracy": 0.9196217494089834,
|
| 192 |
+
"eval_f1": 0.7152317880794702,
|
| 193 |
+
"eval_loss": 0.04378104954957962,
|
| 194 |
+
"eval_precision": 0.6474820143884892,
|
| 195 |
+
"eval_recall": 0.7988165680473372,
|
| 196 |
+
"eval_runtime": 1.6177,
|
| 197 |
+
"eval_samples_per_second": 210.792,
|
| 198 |
+
"eval_steps_per_second": 3.709,
|
| 199 |
+
"step": 350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.3604651162790695,
|
| 203 |
+
"grad_norm": 0.47098585963249207,
|
| 204 |
+
"learning_rate": 4.6656588037260834e-05,
|
| 205 |
+
"loss": 0.020146708488464355,
|
| 206 |
+
"step": 375
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.651162790697675,
|
| 210 |
+
"grad_norm": 0.7534321546554565,
|
| 211 |
+
"learning_rate": 4.570949593260837e-05,
|
| 212 |
+
"loss": 0.018921481370925905,
|
| 213 |
+
"step": 400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.651162790697675,
|
| 217 |
+
"eval_accuracy": 0.9097356544165055,
|
| 218 |
+
"eval_f1": 0.6987341772151898,
|
| 219 |
+
"eval_loss": 0.05498046055436134,
|
| 220 |
+
"eval_precision": 0.6106194690265486,
|
| 221 |
+
"eval_recall": 0.8165680473372781,
|
| 222 |
+
"eval_runtime": 1.6157,
|
| 223 |
+
"eval_samples_per_second": 211.05,
|
| 224 |
+
"eval_steps_per_second": 3.713,
|
| 225 |
+
"step": 400
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.941860465116279,
|
| 229 |
+
"grad_norm": 0.7499828934669495,
|
| 230 |
+
"learning_rate": 4.465668092927841e-05,
|
| 231 |
+
"loss": 0.01998784065246582,
|
| 232 |
+
"step": 425
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 5.232558139534884,
|
| 236 |
+
"grad_norm": 0.24809418618679047,
|
| 237 |
+
"learning_rate": 4.350351769473198e-05,
|
| 238 |
+
"loss": 0.01218403697013855,
|
| 239 |
+
"step": 450
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 5.232558139534884,
|
| 243 |
+
"eval_accuracy": 0.9084461637653127,
|
| 244 |
+
"eval_f1": 0.6831746031746032,
|
| 245 |
+
"eval_loss": 0.05535344406962395,
|
| 246 |
+
"eval_precision": 0.5984427141268076,
|
| 247 |
+
"eval_recall": 0.7958579881656804,
|
| 248 |
+
"eval_runtime": 1.6123,
|
| 249 |
+
"eval_samples_per_second": 211.498,
|
| 250 |
+
"eval_steps_per_second": 3.721,
|
| 251 |
+
"step": 450
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 5.523255813953488,
|
| 255 |
+
"grad_norm": 0.3144132196903229,
|
| 256 |
+
"learning_rate": 4.2255893178617745e-05,
|
| 257 |
+
"loss": 0.012306832075119019,
|
| 258 |
+
"step": 475
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 5.813953488372093,
|
| 262 |
+
"grad_norm": 0.1607646942138672,
|
| 263 |
+
"learning_rate": 4.092017655963198e-05,
|
| 264 |
+
"loss": 0.010539753437042236,
|
| 265 |
+
"step": 500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 5.813953488372093,
|
| 269 |
+
"eval_accuracy": 0.9527186761229315,
|
| 270 |
+
"eval_f1": 0.8059040590405904,
|
| 271 |
+
"eval_loss": 0.07191809266805649,
|
| 272 |
+
"eval_precision": 0.8041237113402062,
|
| 273 |
+
"eval_recall": 0.8076923076923077,
|
| 274 |
+
"eval_runtime": 1.6569,
|
| 275 |
+
"eval_samples_per_second": 205.806,
|
| 276 |
+
"eval_steps_per_second": 3.621,
|
| 277 |
+
"step": 500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.104651162790698,
|
| 281 |
+
"grad_norm": 0.19801200926303864,
|
| 282 |
+
"learning_rate": 3.9503186730577816e-05,
|
| 283 |
+
"loss": 0.008691226243972778,
|
| 284 |
+
"step": 525
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.395348837209302,
|
| 288 |
+
"grad_norm": 0.4177681505680084,
|
| 289 |
+
"learning_rate": 3.801215748761385e-05,
|
| 290 |
+
"loss": 0.004750225245952606,
|
| 291 |
+
"step": 550
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 6.395348837209302,
|
| 295 |
+
"eval_accuracy": 0.9458413926499033,
|
| 296 |
+
"eval_f1": 0.7863866763215062,
|
| 297 |
+
"eval_loss": 0.07378304749727249,
|
| 298 |
+
"eval_precision": 0.7702127659574468,
|
| 299 |
+
"eval_recall": 0.8032544378698225,
|
| 300 |
+
"eval_runtime": 1.6214,
|
| 301 |
+
"eval_samples_per_second": 210.313,
|
| 302 |
+
"eval_steps_per_second": 3.701,
|
| 303 |
+
"step": 550
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.686046511627907,
|
| 307 |
+
"grad_norm": 0.30753573775291443,
|
| 308 |
+
"learning_rate": 3.645470060140278e-05,
|
| 309 |
+
"loss": 0.005743271708488464,
|
| 310 |
+
"step": 575
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.976744186046512,
|
| 314 |
+
"grad_norm": 0.13237181305885315,
|
| 315 |
+
"learning_rate": 3.4838766958683304e-05,
|
| 316 |
+
"loss": 0.006215065121650696,
|
| 317 |
+
"step": 600
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.976744186046512,
|
| 321 |
+
"eval_accuracy": 0.9447668171072426,
|
| 322 |
+
"eval_f1": 0.7789934354485777,
|
| 323 |
+
"eval_loss": 0.062308527529239655,
|
| 324 |
+
"eval_precision": 0.7683453237410072,
|
| 325 |
+
"eval_recall": 0.7899408284023669,
|
| 326 |
+
"eval_runtime": 1.6197,
|
| 327 |
+
"eval_samples_per_second": 210.528,
|
| 328 |
+
"eval_steps_per_second": 3.704,
|
| 329 |
+
"step": 600
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 7.267441860465116,
|
| 333 |
+
"grad_norm": 0.023809686303138733,
|
| 334 |
+
"learning_rate": 3.317260597263932e-05,
|
| 335 |
+
"loss": 0.002318141460418701,
|
| 336 |
+
"step": 625
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.558139534883721,
|
| 340 |
+
"grad_norm": 0.03782212734222412,
|
| 341 |
+
"learning_rate": 3.146472346927845e-05,
|
| 342 |
+
"loss": 0.003790220320224762,
|
| 343 |
+
"step": 650
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 7.558139534883721,
|
| 347 |
+
"eval_accuracy": 0.9529335912314636,
|
| 348 |
+
"eval_f1": 0.8062827225130891,
|
| 349 |
+
"eval_loss": 0.07683796435594559,
|
| 350 |
+
"eval_precision": 0.8154311649016641,
|
| 351 |
+
"eval_recall": 0.7973372781065089,
|
| 352 |
+
"eval_runtime": 1.6235,
|
| 353 |
+
"eval_samples_per_second": 210.043,
|
| 354 |
+
"eval_steps_per_second": 3.696,
|
| 355 |
+
"step": 650
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 7.848837209302325,
|
| 359 |
+
"grad_norm": 0.037971507757902145,
|
| 360 |
+
"learning_rate": 2.9723838264811545e-05,
|
| 361 |
+
"loss": 0.0018679146468639374,
|
| 362 |
+
"step": 675
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 8.13953488372093,
|
| 366 |
+
"grad_norm": 0.6083823442459106,
|
| 367 |
+
"learning_rate": 2.7958837655707814e-05,
|
| 368 |
+
"loss": 0.0019237904250621796,
|
| 369 |
+
"step": 700
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 8.13953488372093,
|
| 373 |
+
"eval_accuracy": 0.9565871480765098,
|
| 374 |
+
"eval_f1": 0.8152091254752851,
|
| 375 |
+
"eval_loss": 0.08732757717370987,
|
| 376 |
+
"eval_precision": 0.838810641627543,
|
| 377 |
+
"eval_recall": 0.7928994082840237,
|
| 378 |
+
"eval_runtime": 1.6124,
|
| 379 |
+
"eval_samples_per_second": 211.484,
|
| 380 |
+
"eval_steps_per_second": 3.721,
|
| 381 |
+
"step": 700
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.430232558139535,
|
| 385 |
+
"grad_norm": 0.022007286548614502,
|
| 386 |
+
"learning_rate": 2.6178732048650694e-05,
|
| 387 |
+
"loss": 0.0010710173845291137,
|
| 388 |
+
"step": 725
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 8.720930232558139,
|
| 392 |
+
"grad_norm": 0.09439104795455933,
|
| 393 |
+
"learning_rate": 2.4392608962010652e-05,
|
| 394 |
+
"loss": 0.0006907544285058976,
|
| 395 |
+
"step": 750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 8.720930232558139,
|
| 399 |
+
"eval_accuracy": 0.9449817322157748,
|
| 400 |
+
"eval_f1": 0.7829848594087959,
|
| 401 |
+
"eval_loss": 0.08308640867471695,
|
| 402 |
+
"eval_precision": 0.7637130801687764,
|
| 403 |
+
"eval_recall": 0.8032544378698225,
|
| 404 |
+
"eval_runtime": 1.6359,
|
| 405 |
+
"eval_samples_per_second": 208.445,
|
| 406 |
+
"eval_steps_per_second": 3.668,
|
| 407 |
+
"step": 750
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 9.011627906976743,
|
| 411 |
+
"grad_norm": 0.018290311098098755,
|
| 412 |
+
"learning_rate": 2.2609586633659256e-05,
|
| 413 |
+
"loss": 0.0005397907271981239,
|
| 414 |
+
"step": 775
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 9.30232558139535,
|
| 418 |
+
"grad_norm": 0.002839206252247095,
|
| 419 |
+
"learning_rate": 2.0838767471959014e-05,
|
| 420 |
+
"loss": 0.0002632186934351921,
|
| 421 |
+
"step": 800
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 9.30232558139535,
|
| 425 |
+
"eval_accuracy": 0.9544379969911885,
|
| 426 |
+
"eval_f1": 0.8081880212282032,
|
| 427 |
+
"eval_loss": 0.09754534065723419,
|
| 428 |
+
"eval_precision": 0.8289269051321928,
|
| 429 |
+
"eval_recall": 0.7884615384615384,
|
| 430 |
+
"eval_runtime": 1.6188,
|
| 431 |
+
"eval_samples_per_second": 210.653,
|
| 432 |
+
"eval_steps_per_second": 3.706,
|
| 433 |
+
"step": 800
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 9.593023255813954,
|
| 437 |
+
"grad_norm": 0.04134448245167732,
|
| 438 |
+
"learning_rate": 1.9089191587563417e-05,
|
| 439 |
+
"loss": 0.0005076020956039428,
|
| 440 |
+
"step": 825
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 9.883720930232558,
|
| 444 |
+
"grad_norm": 0.08968573063611984,
|
| 445 |
+
"learning_rate": 1.7369790643249573e-05,
|
| 446 |
+
"loss": 0.0005874237418174743,
|
| 447 |
+
"step": 850
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 9.883720930232558,
|
| 451 |
+
"eval_accuracy": 0.9548678272082527,
|
| 452 |
+
"eval_f1": 0.8093797276853252,
|
| 453 |
+
"eval_loss": 0.09803693741559982,
|
| 454 |
+
"eval_precision": 0.8281733746130031,
|
| 455 |
+
"eval_recall": 0.7914201183431953,
|
| 456 |
+
"eval_runtime": 1.6209,
|
| 457 |
+
"eval_samples_per_second": 210.379,
|
| 458 |
+
"eval_steps_per_second": 3.702,
|
| 459 |
+
"step": 850
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 10.174418604651162,
|
| 463 |
+
"grad_norm": 0.02889181859791279,
|
| 464 |
+
"learning_rate": 1.5689342257382207e-05,
|
| 465 |
+
"loss": 0.0004838050529360771,
|
| 466 |
+
"step": 875
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.465116279069768,
|
| 470 |
+
"grad_norm": 0.020970497280359268,
|
| 471 |
+
"learning_rate": 1.4056425193781048e-05,
|
| 472 |
+
"loss": 0.0001906752586364746,
|
| 473 |
+
"step": 900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 10.465116279069768,
|
| 477 |
+
"eval_accuracy": 0.9587362991618311,
|
| 478 |
+
"eval_f1": 0.8228043143297381,
|
| 479 |
+
"eval_loss": 0.10382715612649918,
|
| 480 |
+
"eval_precision": 0.8585209003215434,
|
| 481 |
+
"eval_recall": 0.7899408284023669,
|
| 482 |
+
"eval_runtime": 1.6341,
|
| 483 |
+
"eval_samples_per_second": 208.675,
|
| 484 |
+
"eval_steps_per_second": 3.672,
|
| 485 |
+
"step": 900
|
| 486 |
+
}
|
| 487 |
+
],
|
| 488 |
+
"logging_steps": 25,
|
| 489 |
+
"max_steps": 1290,
|
| 490 |
+
"num_input_tokens_seen": 0,
|
| 491 |
+
"num_train_epochs": 15,
|
| 492 |
+
"save_steps": 50,
|
| 493 |
+
"stateful_callbacks": {
|
| 494 |
+
"EarlyStoppingCallback": {
|
| 495 |
+
"args": {
|
| 496 |
+
"early_stopping_patience": 10,
|
| 497 |
+
"early_stopping_threshold": 0.0
|
| 498 |
+
},
|
| 499 |
+
"attributes": {
|
| 500 |
+
"early_stopping_patience_counter": 0
|
| 501 |
+
}
|
| 502 |
+
},
|
| 503 |
+
"TrainerControl": {
|
| 504 |
+
"args": {
|
| 505 |
+
"should_epoch_stop": false,
|
| 506 |
+
"should_evaluate": false,
|
| 507 |
+
"should_log": false,
|
| 508 |
+
"should_save": true,
|
| 509 |
+
"should_training_stop": false
|
| 510 |
+
},
|
| 511 |
+
"attributes": {}
|
| 512 |
+
}
|
| 513 |
+
},
|
| 514 |
+
"total_flos": 4881457666944000.0,
|
| 515 |
+
"train_batch_size": 16,
|
| 516 |
+
"trial_name": null,
|
| 517 |
+
"trial_params": null
|
| 518 |
+
}
|
checkpoint-900/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5071855a54d36018bc5c30ca34d9fef42b69fb3deda057af4ffc70193237d5c4
|
| 3 |
+
size 5201
|
config.json
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.4,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"gradient_checkpointing": false,
|
| 20 |
+
"hidden_activation": "gelu",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "O",
|
| 24 |
+
"1": "FP",
|
| 25 |
+
"2": "RP",
|
| 26 |
+
"3": "RV",
|
| 27 |
+
"4": "PW"
|
| 28 |
+
},
|
| 29 |
+
"initializer_cutoff_factor": 2.0,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 1152,
|
| 32 |
+
"label2id": {
|
| 33 |
+
"FP": 1,
|
| 34 |
+
"O": 0,
|
| 35 |
+
"PW": 4,
|
| 36 |
+
"RP": 2,
|
| 37 |
+
"RV": 3
|
| 38 |
+
},
|
| 39 |
+
"layer_norm_eps": 1e-05,
|
| 40 |
+
"layer_types": [
|
| 41 |
+
"full_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"sliding_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"full_attention"
|
| 63 |
+
],
|
| 64 |
+
"local_attention": 128,
|
| 65 |
+
"max_position_embeddings": 8192,
|
| 66 |
+
"mlp_bias": false,
|
| 67 |
+
"mlp_dropout": 0.0,
|
| 68 |
+
"model_type": "modernbert",
|
| 69 |
+
"norm_bias": false,
|
| 70 |
+
"norm_eps": 1e-05,
|
| 71 |
+
"num_attention_heads": 12,
|
| 72 |
+
"num_hidden_layers": 22,
|
| 73 |
+
"pad_token_id": 50283,
|
| 74 |
+
"position_embedding_type": "absolute",
|
| 75 |
+
"repad_logits_with_grad": false,
|
| 76 |
+
"rope_parameters": {
|
| 77 |
+
"full_attention": {
|
| 78 |
+
"rope_theta": 160000.0,
|
| 79 |
+
"rope_type": "default"
|
| 80 |
+
},
|
| 81 |
+
"sliding_attention": {
|
| 82 |
+
"rope_theta": 10000.0,
|
| 83 |
+
"rope_type": "default"
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"sep_token_id": 50282,
|
| 87 |
+
"sparse_pred_ignore_index": -100,
|
| 88 |
+
"sparse_prediction": false,
|
| 89 |
+
"tie_word_embeddings": true,
|
| 90 |
+
"transformers_version": "5.0.0",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"vocab_size": 50368
|
| 93 |
+
}
|
confusion_matrix.png
ADDED
|
Git LFS Details
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:21c65b89ca04d84973d10cf676f8e009915c6acb1626e669150301db7180e8ea
|
| 3 |
+
size 598449012
|