Upload fine-tuned DistilBERT NER model (PER/LOC/ORG/DATE BIO, test F1 micro 0.85)
Browse files- config.json +50 -0
- eval_results.json +101 -0
- labels.json +24 -0
- model.safetensors +3 -0
- runs/checkpoint-1650/config.json +50 -0
- runs/checkpoint-1650/model.safetensors +3 -0
- runs/checkpoint-1650/tokenizer.json +0 -0
- runs/checkpoint-1650/tokenizer_config.json +15 -0
- runs/checkpoint-1650/trainer_state.json +320 -0
- runs/checkpoint-1650/training_args.bin +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +15 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_dim": 3072,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "O",
|
| 15 |
+
"1": "B-PER",
|
| 16 |
+
"2": "I-PER",
|
| 17 |
+
"3": "B-LOC",
|
| 18 |
+
"4": "I-LOC",
|
| 19 |
+
"5": "B-ORG",
|
| 20 |
+
"6": "I-ORG",
|
| 21 |
+
"7": "B-DATE",
|
| 22 |
+
"8": "I-DATE"
|
| 23 |
+
},
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-DATE": 7,
|
| 27 |
+
"B-LOC": 3,
|
| 28 |
+
"B-ORG": 5,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-DATE": 8,
|
| 31 |
+
"I-LOC": 4,
|
| 32 |
+
"I-ORG": 6,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"max_position_embeddings": 512,
|
| 37 |
+
"model_type": "distilbert",
|
| 38 |
+
"n_heads": 12,
|
| 39 |
+
"n_layers": 6,
|
| 40 |
+
"output_past": true,
|
| 41 |
+
"pad_token_id": 0,
|
| 42 |
+
"qa_dropout": 0.1,
|
| 43 |
+
"seq_classif_dropout": 0.2,
|
| 44 |
+
"sinusoidal_pos_embds": false,
|
| 45 |
+
"tie_weights_": true,
|
| 46 |
+
"tie_word_embeddings": true,
|
| 47 |
+
"transformers_version": "5.8.0",
|
| 48 |
+
"use_cache": false,
|
| 49 |
+
"vocab_size": 119547
|
| 50 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,101 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_name": "distilbert-base-multilingual-cased",
|
| 3 |
+
"task": "ner",
|
| 4 |
+
"num_labels": 9,
|
| 5 |
+
"labels": {
|
| 6 |
+
"O": 0,
|
| 7 |
+
"B-PER": 1,
|
| 8 |
+
"I-PER": 2,
|
| 9 |
+
"B-LOC": 3,
|
| 10 |
+
"I-LOC": 4,
|
| 11 |
+
"B-ORG": 5,
|
| 12 |
+
"I-ORG": 6,
|
| 13 |
+
"B-DATE": 7,
|
| 14 |
+
"I-DATE": 8
|
| 15 |
+
},
|
| 16 |
+
"test_metrics": {
|
| 17 |
+
"test_loss": 0.2599596679210663,
|
| 18 |
+
"test_model_preparation_time": 0.001,
|
| 19 |
+
"test_f1": 0.8454746136865342,
|
| 20 |
+
"test_precision": 0.8299024918743229,
|
| 21 |
+
"test_recall": 0.8616422947131609,
|
| 22 |
+
"test_runtime": 5.9639,
|
| 23 |
+
"test_samples_per_second": 110.666,
|
| 24 |
+
"test_steps_per_second": 7.042,
|
| 25 |
+
"epoch": 0.0
|
| 26 |
+
},
|
| 27 |
+
"classification_report": {
|
| 28 |
+
"DATE": {
|
| 29 |
+
"precision": 0.9848484848484849,
|
| 30 |
+
"recall": 1.0,
|
| 31 |
+
"f1-score": 0.9923664122137404,
|
| 32 |
+
"support": 65
|
| 33 |
+
},
|
| 34 |
+
"LOC": {
|
| 35 |
+
"precision": 0.8625,
|
| 36 |
+
"recall": 0.8961038961038961,
|
| 37 |
+
"f1-score": 0.8789808917197451,
|
| 38 |
+
"support": 308
|
| 39 |
+
},
|
| 40 |
+
"ORG": {
|
| 41 |
+
"precision": 0.7113821138211383,
|
| 42 |
+
"recall": 0.7383966244725738,
|
| 43 |
+
"f1-score": 0.7246376811594203,
|
| 44 |
+
"support": 237
|
| 45 |
+
},
|
| 46 |
+
"PER": {
|
| 47 |
+
"precision": 0.8591065292096219,
|
| 48 |
+
"recall": 0.8960573476702509,
|
| 49 |
+
"f1-score": 0.8771929824561404,
|
| 50 |
+
"support": 279
|
| 51 |
+
},
|
| 52 |
+
"micro avg": {
|
| 53 |
+
"precision": 0.8299024918743229,
|
| 54 |
+
"recall": 0.8616422947131609,
|
| 55 |
+
"f1-score": 0.8454746136865342,
|
| 56 |
+
"support": 889
|
| 57 |
+
},
|
| 58 |
+
"macro avg": {
|
| 59 |
+
"precision": 0.8544592819698112,
|
| 60 |
+
"recall": 0.8826394670616802,
|
| 61 |
+
"f1-score": 0.8682944918872616,
|
| 62 |
+
"support": 889
|
| 63 |
+
},
|
| 64 |
+
"weighted avg": {
|
| 65 |
+
"precision": 0.830093851676317,
|
| 66 |
+
"recall": 0.8616422947131609,
|
| 67 |
+
"f1-score": 0.8455634465507541,
|
| 68 |
+
"support": 889
|
| 69 |
+
}
|
| 70 |
+
},
|
| 71 |
+
"per_language": {
|
| 72 |
+
"AR": {
|
| 73 |
+
"n": 220,
|
| 74 |
+
"f1": 0.8510638297872339,
|
| 75 |
+
"precision": 0.8421052631578947,
|
| 76 |
+
"recall": 0.8602150537634409
|
| 77 |
+
},
|
| 78 |
+
"EN": {
|
| 79 |
+
"n": 220,
|
| 80 |
+
"f1": 0.8226059654631083,
|
| 81 |
+
"precision": 0.8061538461538461,
|
| 82 |
+
"recall": 0.8397435897435898
|
| 83 |
+
},
|
| 84 |
+
"FR": {
|
| 85 |
+
"n": 220,
|
| 86 |
+
"f1": 0.8641571194762685,
|
| 87 |
+
"precision": 0.8434504792332268,
|
| 88 |
+
"recall": 0.8859060402684564
|
| 89 |
+
}
|
| 90 |
+
},
|
| 91 |
+
"training": {
|
| 92 |
+
"epochs": 5,
|
| 93 |
+
"per_device_batch": 8,
|
| 94 |
+
"grad_accum": 2,
|
| 95 |
+
"effective_batch": 16,
|
| 96 |
+
"learning_rate": 2e-05,
|
| 97 |
+
"warmup_steps": 100,
|
| 98 |
+
"fp16": true,
|
| 99 |
+
"note": "Recovered via finalize_ner.py after train_ner.py crashed on json.dumps (numpy int64 in seqeval report 'support'). Model itself was fully trained and saved; this script only regenerates eval_results.json."
|
| 100 |
+
}
|
| 101 |
+
}
|
labels.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_to_id": {
|
| 3 |
+
"O": 0,
|
| 4 |
+
"B-PER": 1,
|
| 5 |
+
"I-PER": 2,
|
| 6 |
+
"B-LOC": 3,
|
| 7 |
+
"I-LOC": 4,
|
| 8 |
+
"B-ORG": 5,
|
| 9 |
+
"I-ORG": 6,
|
| 10 |
+
"B-DATE": 7,
|
| 11 |
+
"I-DATE": 8
|
| 12 |
+
},
|
| 13 |
+
"id_to_label": {
|
| 14 |
+
"0": "O",
|
| 15 |
+
"1": "B-PER",
|
| 16 |
+
"2": "I-PER",
|
| 17 |
+
"3": "B-LOC",
|
| 18 |
+
"4": "I-LOC",
|
| 19 |
+
"5": "B-ORG",
|
| 20 |
+
"6": "I-ORG",
|
| 21 |
+
"7": "B-DATE",
|
| 22 |
+
"8": "I-DATE"
|
| 23 |
+
}
|
| 24 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7095fe51877c44015a3ee0ce70bf87d50681c119facd49ae888e88a60b3ddf51
|
| 3 |
+
size 538976356
|
runs/checkpoint-1650/config.json
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_dim": 3072,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "O",
|
| 15 |
+
"1": "B-PER",
|
| 16 |
+
"2": "I-PER",
|
| 17 |
+
"3": "B-LOC",
|
| 18 |
+
"4": "I-LOC",
|
| 19 |
+
"5": "B-ORG",
|
| 20 |
+
"6": "I-ORG",
|
| 21 |
+
"7": "B-DATE",
|
| 22 |
+
"8": "I-DATE"
|
| 23 |
+
},
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"label2id": {
|
| 26 |
+
"B-DATE": 7,
|
| 27 |
+
"B-LOC": 3,
|
| 28 |
+
"B-ORG": 5,
|
| 29 |
+
"B-PER": 1,
|
| 30 |
+
"I-DATE": 8,
|
| 31 |
+
"I-LOC": 4,
|
| 32 |
+
"I-ORG": 6,
|
| 33 |
+
"I-PER": 2,
|
| 34 |
+
"O": 0
|
| 35 |
+
},
|
| 36 |
+
"max_position_embeddings": 512,
|
| 37 |
+
"model_type": "distilbert",
|
| 38 |
+
"n_heads": 12,
|
| 39 |
+
"n_layers": 6,
|
| 40 |
+
"output_past": true,
|
| 41 |
+
"pad_token_id": 0,
|
| 42 |
+
"qa_dropout": 0.1,
|
| 43 |
+
"seq_classif_dropout": 0.2,
|
| 44 |
+
"sinusoidal_pos_embds": false,
|
| 45 |
+
"tie_weights_": true,
|
| 46 |
+
"tie_word_embeddings": true,
|
| 47 |
+
"transformers_version": "5.8.0",
|
| 48 |
+
"use_cache": false,
|
| 49 |
+
"vocab_size": 119547
|
| 50 |
+
}
|
runs/checkpoint-1650/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7095fe51877c44015a3ee0ce70bf87d50681c119facd49ae888e88a60b3ddf51
|
| 3 |
+
size 538976356
|
runs/checkpoint-1650/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
runs/checkpoint-1650/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": false,
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"local_files_only": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
runs/checkpoint-1650/trainer_state.json
ADDED
|
@@ -0,0 +1,320 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1650,
|
| 3 |
+
"best_metric": 0.825503355704698,
|
| 4 |
+
"best_model_checkpoint": "/home/momen/Desktop/aispire/multilingual-intent-ner-chatbot/models/ner_model/runs/checkpoint-1650",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1650,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.15151515151515152,
|
| 14 |
+
"grad_norm": 6.311799049377441,
|
| 15 |
+
"learning_rate": 9.600000000000001e-06,
|
| 16 |
+
"loss": 3.912173156738281,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.30303030303030304,
|
| 21 |
+
"grad_norm": 9.783806800842285,
|
| 22 |
+
"learning_rate": 1.9600000000000002e-05,
|
| 23 |
+
"loss": 2.426840362548828,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.45454545454545453,
|
| 28 |
+
"grad_norm": 7.24477481842041,
|
| 29 |
+
"learning_rate": 1.9380645161290325e-05,
|
| 30 |
+
"loss": 1.1294728851318359,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.6060606060606061,
|
| 35 |
+
"grad_norm": 12.536176681518555,
|
| 36 |
+
"learning_rate": 1.8748387096774195e-05,
|
| 37 |
+
"loss": 0.7891785430908204,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7575757575757576,
|
| 42 |
+
"grad_norm": 10.589644432067871,
|
| 43 |
+
"learning_rate": 1.8103225806451614e-05,
|
| 44 |
+
"loss": 0.7708683013916016,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.9090909090909091,
|
| 49 |
+
"grad_norm": 13.25113296508789,
|
| 50 |
+
"learning_rate": 1.7470967741935483e-05,
|
| 51 |
+
"loss": 0.6683406066894532,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_f1": 0.7807840971838763,
|
| 57 |
+
"eval_loss": 0.29953959584236145,
|
| 58 |
+
"eval_precision": 0.7634989200863931,
|
| 59 |
+
"eval_recall": 0.7988700564971751,
|
| 60 |
+
"eval_runtime": 6.7464,
|
| 61 |
+
"eval_samples_per_second": 97.831,
|
| 62 |
+
"eval_steps_per_second": 6.226,
|
| 63 |
+
"step": 330
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0606060606060606,
|
| 67 |
+
"grad_norm": 11.886589050292969,
|
| 68 |
+
"learning_rate": 1.6825806451612906e-05,
|
| 69 |
+
"loss": 0.6518353271484375,
|
| 70 |
+
"step": 350
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.2121212121212122,
|
| 74 |
+
"grad_norm": 12.366310119628906,
|
| 75 |
+
"learning_rate": 1.6180645161290325e-05,
|
| 76 |
+
"loss": 0.4751667785644531,
|
| 77 |
+
"step": 400
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.3636363636363638,
|
| 81 |
+
"grad_norm": 17.39374351501465,
|
| 82 |
+
"learning_rate": 1.553548387096774e-05,
|
| 83 |
+
"loss": 0.47883995056152345,
|
| 84 |
+
"step": 450
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.5151515151515151,
|
| 88 |
+
"grad_norm": 11.148221969604492,
|
| 89 |
+
"learning_rate": 1.4890322580645162e-05,
|
| 90 |
+
"loss": 0.44007171630859376,
|
| 91 |
+
"step": 500
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.6666666666666665,
|
| 95 |
+
"grad_norm": 2.922976493835449,
|
| 96 |
+
"learning_rate": 1.4245161290322583e-05,
|
| 97 |
+
"loss": 0.5034889221191406,
|
| 98 |
+
"step": 550
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.8181818181818183,
|
| 102 |
+
"grad_norm": 14.21239185333252,
|
| 103 |
+
"learning_rate": 1.3600000000000002e-05,
|
| 104 |
+
"loss": 0.48983699798583985,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.9696969696969697,
|
| 109 |
+
"grad_norm": 5.757568836212158,
|
| 110 |
+
"learning_rate": 1.295483870967742e-05,
|
| 111 |
+
"loss": 0.47100028991699217,
|
| 112 |
+
"step": 650
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_f1": 0.8089635854341737,
|
| 117 |
+
"eval_loss": 0.27586379647254944,
|
| 118 |
+
"eval_precision": 0.8022222222222222,
|
| 119 |
+
"eval_recall": 0.8158192090395481,
|
| 120 |
+
"eval_runtime": 6.7528,
|
| 121 |
+
"eval_samples_per_second": 97.737,
|
| 122 |
+
"eval_steps_per_second": 6.22,
|
| 123 |
+
"step": 660
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.121212121212121,
|
| 127 |
+
"grad_norm": 4.0693488121032715,
|
| 128 |
+
"learning_rate": 1.230967741935484e-05,
|
| 129 |
+
"loss": 0.32343727111816406,
|
| 130 |
+
"step": 700
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.2727272727272725,
|
| 134 |
+
"grad_norm": 12.245550155639648,
|
| 135 |
+
"learning_rate": 1.166451612903226e-05,
|
| 136 |
+
"loss": 0.3425853729248047,
|
| 137 |
+
"step": 750
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.4242424242424243,
|
| 141 |
+
"grad_norm": 5.691013813018799,
|
| 142 |
+
"learning_rate": 1.1019354838709679e-05,
|
| 143 |
+
"loss": 0.3368482971191406,
|
| 144 |
+
"step": 800
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.5757575757575757,
|
| 148 |
+
"grad_norm": 6.1310577392578125,
|
| 149 |
+
"learning_rate": 1.0374193548387096e-05,
|
| 150 |
+
"loss": 0.26243484497070313,
|
| 151 |
+
"step": 850
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.7272727272727275,
|
| 155 |
+
"grad_norm": 9.599594116210938,
|
| 156 |
+
"learning_rate": 9.729032258064517e-06,
|
| 157 |
+
"loss": 0.2967669677734375,
|
| 158 |
+
"step": 900
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.878787878787879,
|
| 162 |
+
"grad_norm": 22.023439407348633,
|
| 163 |
+
"learning_rate": 9.083870967741936e-06,
|
| 164 |
+
"loss": 0.3708695983886719,
|
| 165 |
+
"step": 950
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 3.0,
|
| 169 |
+
"eval_f1": 0.8176733780760626,
|
| 170 |
+
"eval_loss": 0.2901476323604584,
|
| 171 |
+
"eval_precision": 0.8095238095238095,
|
| 172 |
+
"eval_recall": 0.8259887005649718,
|
| 173 |
+
"eval_runtime": 6.7458,
|
| 174 |
+
"eval_samples_per_second": 97.839,
|
| 175 |
+
"eval_steps_per_second": 6.226,
|
| 176 |
+
"step": 990
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.0303030303030303,
|
| 180 |
+
"grad_norm": 1.546368956565857,
|
| 181 |
+
"learning_rate": 8.438709677419355e-06,
|
| 182 |
+
"loss": 0.2276275062561035,
|
| 183 |
+
"step": 1000
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.1818181818181817,
|
| 187 |
+
"grad_norm": 6.266290187835693,
|
| 188 |
+
"learning_rate": 7.793548387096775e-06,
|
| 189 |
+
"loss": 0.22509269714355468,
|
| 190 |
+
"step": 1050
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.3333333333333335,
|
| 194 |
+
"grad_norm": 13.151640892028809,
|
| 195 |
+
"learning_rate": 7.148387096774194e-06,
|
| 196 |
+
"loss": 0.24928258895874023,
|
| 197 |
+
"step": 1100
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.484848484848485,
|
| 201 |
+
"grad_norm": 2.640580415725708,
|
| 202 |
+
"learning_rate": 6.503225806451613e-06,
|
| 203 |
+
"loss": 0.2151968002319336,
|
| 204 |
+
"step": 1150
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.6363636363636362,
|
| 208 |
+
"grad_norm": 9.910027503967285,
|
| 209 |
+
"learning_rate": 5.858064516129033e-06,
|
| 210 |
+
"loss": 0.19832775115966797,
|
| 211 |
+
"step": 1200
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.787878787878788,
|
| 215 |
+
"grad_norm": 5.850687026977539,
|
| 216 |
+
"learning_rate": 5.2129032258064514e-06,
|
| 217 |
+
"loss": 0.22372671127319335,
|
| 218 |
+
"step": 1250
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.9393939393939394,
|
| 222 |
+
"grad_norm": 9.980695724487305,
|
| 223 |
+
"learning_rate": 4.5677419354838715e-06,
|
| 224 |
+
"loss": 0.206069393157959,
|
| 225 |
+
"step": 1300
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.0,
|
| 229 |
+
"eval_f1": 0.8181309457190823,
|
| 230 |
+
"eval_loss": 0.31706130504608154,
|
| 231 |
+
"eval_precision": 0.8104212860310421,
|
| 232 |
+
"eval_recall": 0.8259887005649718,
|
| 233 |
+
"eval_runtime": 6.7515,
|
| 234 |
+
"eval_samples_per_second": 97.756,
|
| 235 |
+
"eval_steps_per_second": 6.221,
|
| 236 |
+
"step": 1320
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 4.090909090909091,
|
| 240 |
+
"grad_norm": 1.9647992849349976,
|
| 241 |
+
"learning_rate": 3.922580645161291e-06,
|
| 242 |
+
"loss": 0.17119260787963866,
|
| 243 |
+
"step": 1350
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.242424242424242,
|
| 247 |
+
"grad_norm": 2.3536336421966553,
|
| 248 |
+
"learning_rate": 3.2774193548387103e-06,
|
| 249 |
+
"loss": 0.15796515464782715,
|
| 250 |
+
"step": 1400
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.393939393939394,
|
| 254 |
+
"grad_norm": 3.601684331893921,
|
| 255 |
+
"learning_rate": 2.6322580645161295e-06,
|
| 256 |
+
"loss": 0.1238516902923584,
|
| 257 |
+
"step": 1450
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.545454545454545,
|
| 261 |
+
"grad_norm": 7.281225681304932,
|
| 262 |
+
"learning_rate": 1.9870967741935487e-06,
|
| 263 |
+
"loss": 0.1787614631652832,
|
| 264 |
+
"step": 1500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.696969696969697,
|
| 268 |
+
"grad_norm": 2.253016948699951,
|
| 269 |
+
"learning_rate": 1.3419354838709679e-06,
|
| 270 |
+
"loss": 0.1587313461303711,
|
| 271 |
+
"step": 1550
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.848484848484849,
|
| 275 |
+
"grad_norm": 4.587268829345703,
|
| 276 |
+
"learning_rate": 6.96774193548387e-07,
|
| 277 |
+
"loss": 0.17344398498535157,
|
| 278 |
+
"step": 1600
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 5.0,
|
| 282 |
+
"grad_norm": 21.149198532104492,
|
| 283 |
+
"learning_rate": 5.1612903225806456e-08,
|
| 284 |
+
"loss": 0.1394967555999756,
|
| 285 |
+
"step": 1650
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 5.0,
|
| 289 |
+
"eval_f1": 0.825503355704698,
|
| 290 |
+
"eval_loss": 0.3302437961101532,
|
| 291 |
+
"eval_precision": 0.8172757475083057,
|
| 292 |
+
"eval_recall": 0.8338983050847457,
|
| 293 |
+
"eval_runtime": 6.6687,
|
| 294 |
+
"eval_samples_per_second": 98.969,
|
| 295 |
+
"eval_steps_per_second": 6.298,
|
| 296 |
+
"step": 1650
|
| 297 |
+
}
|
| 298 |
+
],
|
| 299 |
+
"logging_steps": 50,
|
| 300 |
+
"max_steps": 1650,
|
| 301 |
+
"num_input_tokens_seen": 0,
|
| 302 |
+
"num_train_epochs": 5,
|
| 303 |
+
"save_steps": 500,
|
| 304 |
+
"stateful_callbacks": {
|
| 305 |
+
"TrainerControl": {
|
| 306 |
+
"args": {
|
| 307 |
+
"should_epoch_stop": false,
|
| 308 |
+
"should_evaluate": false,
|
| 309 |
+
"should_log": false,
|
| 310 |
+
"should_save": true,
|
| 311 |
+
"should_training_stop": true
|
| 312 |
+
},
|
| 313 |
+
"attributes": {}
|
| 314 |
+
}
|
| 315 |
+
},
|
| 316 |
+
"total_flos": 163028747086128.0,
|
| 317 |
+
"train_batch_size": 8,
|
| 318 |
+
"trial_name": null,
|
| 319 |
+
"trial_params": null
|
| 320 |
+
}
|
runs/checkpoint-1650/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:803cfa755e3dcf2d3aae0449370784b6b25f3836f966c134439d33134a98dd93
|
| 3 |
+
size 5329
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": false,
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"local_files_only": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:803cfa755e3dcf2d3aae0449370784b6b25f3836f966c134439d33134a98dd93
|
| 3 |
+
size 5329
|