Upload 6 files
Browse files- config.json +172 -0
- model.safetensors +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +60 -0
- trainer_state.json +141 -0
config.json
ADDED
|
@@ -0,0 +1,172 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/muril-large-cased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"embedding_size": 1024,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "B-AerospaceManufacturer",
|
| 14 |
+
"1": "B-AnatomicalStructure",
|
| 15 |
+
"2": "B-ArtWork",
|
| 16 |
+
"3": "B-Artist",
|
| 17 |
+
"4": "B-Athlete",
|
| 18 |
+
"5": "B-CarManufacturer",
|
| 19 |
+
"6": "B-Cleric",
|
| 20 |
+
"7": "B-Clothing",
|
| 21 |
+
"8": "B-Disease",
|
| 22 |
+
"9": "B-Drink",
|
| 23 |
+
"10": "B-Facility",
|
| 24 |
+
"11": "B-Food",
|
| 25 |
+
"12": "B-HumanSettlement",
|
| 26 |
+
"13": "B-MedicalProcedure",
|
| 27 |
+
"14": "B-Medication/Vaccine",
|
| 28 |
+
"15": "B-MusicalGRP",
|
| 29 |
+
"16": "B-MusicalWork",
|
| 30 |
+
"17": "B-ORG",
|
| 31 |
+
"18": "B-OtherLOC",
|
| 32 |
+
"19": "B-OtherPER",
|
| 33 |
+
"20": "B-OtherPROD",
|
| 34 |
+
"21": "B-Politician",
|
| 35 |
+
"22": "B-PrivateCORP",
|
| 36 |
+
"23": "B-PrivateCorp",
|
| 37 |
+
"24": "B-PublicCORP",
|
| 38 |
+
"25": "B-PublicCorp",
|
| 39 |
+
"26": "B-Scientist",
|
| 40 |
+
"27": "B-Software",
|
| 41 |
+
"28": "B-SportsGRP",
|
| 42 |
+
"29": "B-SportsManager",
|
| 43 |
+
"30": "B-Station",
|
| 44 |
+
"31": "B-Symptom",
|
| 45 |
+
"32": "B-Vehicle",
|
| 46 |
+
"33": "B-VisualWork",
|
| 47 |
+
"34": "B-WrittenWork",
|
| 48 |
+
"35": "I-AerospaceManufacturer",
|
| 49 |
+
"36": "I-AnatomicalStructure",
|
| 50 |
+
"37": "I-ArtWork",
|
| 51 |
+
"38": "I-Artist",
|
| 52 |
+
"39": "I-Athlete",
|
| 53 |
+
"40": "I-CarManufacturer",
|
| 54 |
+
"41": "I-Cleric",
|
| 55 |
+
"42": "I-Clothing",
|
| 56 |
+
"43": "I-Disease",
|
| 57 |
+
"44": "I-Drink",
|
| 58 |
+
"45": "I-Facility",
|
| 59 |
+
"46": "I-Food",
|
| 60 |
+
"47": "I-HumanSettlement",
|
| 61 |
+
"48": "I-MedicalProcedure",
|
| 62 |
+
"49": "I-Medication/Vaccine",
|
| 63 |
+
"50": "I-MusicalGRP",
|
| 64 |
+
"51": "I-MusicalWork",
|
| 65 |
+
"52": "I-ORG",
|
| 66 |
+
"53": "I-OtherLOC",
|
| 67 |
+
"54": "I-OtherPER",
|
| 68 |
+
"55": "I-OtherPROD",
|
| 69 |
+
"56": "I-Politician",
|
| 70 |
+
"57": "I-PrivateCORP",
|
| 71 |
+
"58": "I-PrivateCorp",
|
| 72 |
+
"59": "I-PublicCORP",
|
| 73 |
+
"60": "I-PublicCorp",
|
| 74 |
+
"61": "I-Scientist",
|
| 75 |
+
"62": "I-Software",
|
| 76 |
+
"63": "I-SportsGRP",
|
| 77 |
+
"64": "I-SportsManager",
|
| 78 |
+
"65": "I-Station",
|
| 79 |
+
"66": "I-Symptom",
|
| 80 |
+
"67": "I-Vehicle",
|
| 81 |
+
"68": "I-VisualWork",
|
| 82 |
+
"69": "I-WrittenWork",
|
| 83 |
+
"70": "O"
|
| 84 |
+
},
|
| 85 |
+
"initializer_range": 0.02,
|
| 86 |
+
"intermediate_size": 4096,
|
| 87 |
+
"label2id": {
|
| 88 |
+
"B-AerospaceManufacturer": 0,
|
| 89 |
+
"B-AnatomicalStructure": 1,
|
| 90 |
+
"B-ArtWork": 2,
|
| 91 |
+
"B-Artist": 3,
|
| 92 |
+
"B-Athlete": 4,
|
| 93 |
+
"B-CarManufacturer": 5,
|
| 94 |
+
"B-Cleric": 6,
|
| 95 |
+
"B-Clothing": 7,
|
| 96 |
+
"B-Disease": 8,
|
| 97 |
+
"B-Drink": 9,
|
| 98 |
+
"B-Facility": 10,
|
| 99 |
+
"B-Food": 11,
|
| 100 |
+
"B-HumanSettlement": 12,
|
| 101 |
+
"B-MedicalProcedure": 13,
|
| 102 |
+
"B-Medication/Vaccine": 14,
|
| 103 |
+
"B-MusicalGRP": 15,
|
| 104 |
+
"B-MusicalWork": 16,
|
| 105 |
+
"B-ORG": 17,
|
| 106 |
+
"B-OtherLOC": 18,
|
| 107 |
+
"B-OtherPER": 19,
|
| 108 |
+
"B-OtherPROD": 20,
|
| 109 |
+
"B-Politician": 21,
|
| 110 |
+
"B-PrivateCORP": 22,
|
| 111 |
+
"B-PrivateCorp": 23,
|
| 112 |
+
"B-PublicCORP": 24,
|
| 113 |
+
"B-PublicCorp": 25,
|
| 114 |
+
"B-Scientist": 26,
|
| 115 |
+
"B-Software": 27,
|
| 116 |
+
"B-SportsGRP": 28,
|
| 117 |
+
"B-SportsManager": 29,
|
| 118 |
+
"B-Station": 30,
|
| 119 |
+
"B-Symptom": 31,
|
| 120 |
+
"B-Vehicle": 32,
|
| 121 |
+
"B-VisualWork": 33,
|
| 122 |
+
"B-WrittenWork": 34,
|
| 123 |
+
"I-AerospaceManufacturer": 35,
|
| 124 |
+
"I-AnatomicalStructure": 36,
|
| 125 |
+
"I-ArtWork": 37,
|
| 126 |
+
"I-Artist": 38,
|
| 127 |
+
"I-Athlete": 39,
|
| 128 |
+
"I-CarManufacturer": 40,
|
| 129 |
+
"I-Cleric": 41,
|
| 130 |
+
"I-Clothing": 42,
|
| 131 |
+
"I-Disease": 43,
|
| 132 |
+
"I-Drink": 44,
|
| 133 |
+
"I-Facility": 45,
|
| 134 |
+
"I-Food": 46,
|
| 135 |
+
"I-HumanSettlement": 47,
|
| 136 |
+
"I-MedicalProcedure": 48,
|
| 137 |
+
"I-Medication/Vaccine": 49,
|
| 138 |
+
"I-MusicalGRP": 50,
|
| 139 |
+
"I-MusicalWork": 51,
|
| 140 |
+
"I-ORG": 52,
|
| 141 |
+
"I-OtherLOC": 53,
|
| 142 |
+
"I-OtherPER": 54,
|
| 143 |
+
"I-OtherPROD": 55,
|
| 144 |
+
"I-Politician": 56,
|
| 145 |
+
"I-PrivateCORP": 57,
|
| 146 |
+
"I-PrivateCorp": 58,
|
| 147 |
+
"I-PublicCORP": 59,
|
| 148 |
+
"I-PublicCorp": 60,
|
| 149 |
+
"I-Scientist": 61,
|
| 150 |
+
"I-Software": 62,
|
| 151 |
+
"I-SportsGRP": 63,
|
| 152 |
+
"I-SportsManager": 64,
|
| 153 |
+
"I-Station": 65,
|
| 154 |
+
"I-Symptom": 66,
|
| 155 |
+
"I-Vehicle": 67,
|
| 156 |
+
"I-VisualWork": 68,
|
| 157 |
+
"I-WrittenWork": 69,
|
| 158 |
+
"O": 70
|
| 159 |
+
},
|
| 160 |
+
"layer_norm_eps": 1e-12,
|
| 161 |
+
"max_position_embeddings": 512,
|
| 162 |
+
"model_type": "bert",
|
| 163 |
+
"num_attention_heads": 16,
|
| 164 |
+
"num_hidden_layers": 24,
|
| 165 |
+
"pad_token_id": 0,
|
| 166 |
+
"position_embedding_type": "absolute",
|
| 167 |
+
"torch_dtype": "float32",
|
| 168 |
+
"transformers_version": "4.48.3",
|
| 169 |
+
"type_vocab_size": 2,
|
| 170 |
+
"use_cache": true,
|
| 171 |
+
"vocab_size": 197285
|
| 172 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c17260fe7deb02bc655fd9839ad5dc9e59bd2bff43f78dfe191c61b80c8bd13a
|
| 3 |
+
size 2019768756
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "[PAD]",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"100": {
|
| 13 |
+
"content": "[UNK]",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"103": {
|
| 21 |
+
"content": "[MASK]",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"104": {
|
| 29 |
+
"content": "[CLS]",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"105": {
|
| 37 |
+
"content": "[SEP]",
|
| 38 |
+
"lstrip": false,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "[CLS]",
|
| 47 |
+
"do_basic_tokenize": true,
|
| 48 |
+
"do_lower_case": false,
|
| 49 |
+
"extra_special_tokens": {},
|
| 50 |
+
"lowercase": false,
|
| 51 |
+
"mask_token": "[MASK]",
|
| 52 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 53 |
+
"never_split": null,
|
| 54 |
+
"pad_token": "[PAD]",
|
| 55 |
+
"sep_token": "[SEP]",
|
| 56 |
+
"strip_accents": false,
|
| 57 |
+
"tokenize_chinese_chars": true,
|
| 58 |
+
"tokenizer_class": "BertTokenizer",
|
| 59 |
+
"unk_token": "[UNK]"
|
| 60 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,141 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5666028954056137,
|
| 3 |
+
"best_model_checkpoint": "./models/brx_TrA50_MR/checkpoint-2214",
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2214,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.997289972899729,
|
| 13 |
+
"grad_norm": 3.5397324562072754,
|
| 14 |
+
"learning_rate": 4.168925022583559e-05,
|
| 15 |
+
"loss": 0.8438,
|
| 16 |
+
"step": 368
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_f1": 0.44267854576807014,
|
| 21 |
+
"eval_loss": 0.5134469270706177,
|
| 22 |
+
"eval_precision": 0.4302273615999106,
|
| 23 |
+
"eval_recall": 0.4558719071859832,
|
| 24 |
+
"eval_runtime": 71.5939,
|
| 25 |
+
"eval_samples_per_second": 330.321,
|
| 26 |
+
"eval_steps_per_second": 5.168,
|
| 27 |
+
"step": 369
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 1.994579945799458,
|
| 31 |
+
"grad_norm": 2.107616901397705,
|
| 32 |
+
"learning_rate": 3.3378500451671184e-05,
|
| 33 |
+
"loss": 0.4624,
|
| 34 |
+
"step": 736
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.0,
|
| 38 |
+
"eval_f1": 0.5124409543580949,
|
| 39 |
+
"eval_loss": 0.4316766858100891,
|
| 40 |
+
"eval_precision": 0.5210793611943951,
|
| 41 |
+
"eval_recall": 0.504084290280573,
|
| 42 |
+
"eval_runtime": 71.7509,
|
| 43 |
+
"eval_samples_per_second": 329.599,
|
| 44 |
+
"eval_steps_per_second": 5.157,
|
| 45 |
+
"step": 738
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.991869918699187,
|
| 49 |
+
"grad_norm": 3.3447747230529785,
|
| 50 |
+
"learning_rate": 2.5067750677506774e-05,
|
| 51 |
+
"loss": 0.3523,
|
| 52 |
+
"step": 1104
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 3.0,
|
| 56 |
+
"eval_f1": 0.5369874838152783,
|
| 57 |
+
"eval_loss": 0.4135218858718872,
|
| 58 |
+
"eval_precision": 0.5224511505514808,
|
| 59 |
+
"eval_recall": 0.5523558659879247,
|
| 60 |
+
"eval_runtime": 71.9772,
|
| 61 |
+
"eval_samples_per_second": 328.562,
|
| 62 |
+
"eval_steps_per_second": 5.141,
|
| 63 |
+
"step": 1107
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 3.989159891598916,
|
| 67 |
+
"grad_norm": 2.4756720066070557,
|
| 68 |
+
"learning_rate": 1.675700090334237e-05,
|
| 69 |
+
"loss": 0.275,
|
| 70 |
+
"step": 1472
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 4.0,
|
| 74 |
+
"eval_f1": 0.550384120140627,
|
| 75 |
+
"eval_loss": 0.4026535451412201,
|
| 76 |
+
"eval_precision": 0.5383657411338485,
|
| 77 |
+
"eval_recall": 0.5629513436723097,
|
| 78 |
+
"eval_runtime": 71.9543,
|
| 79 |
+
"eval_samples_per_second": 328.667,
|
| 80 |
+
"eval_steps_per_second": 5.142,
|
| 81 |
+
"step": 1476
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 4.9864498644986455,
|
| 85 |
+
"grad_norm": 1.6352941989898682,
|
| 86 |
+
"learning_rate": 8.446251129177959e-06,
|
| 87 |
+
"loss": 0.219,
|
| 88 |
+
"step": 1840
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 5.0,
|
| 92 |
+
"eval_f1": 0.5614785049658729,
|
| 93 |
+
"eval_loss": 0.4078289270401001,
|
| 94 |
+
"eval_precision": 0.5413975983072737,
|
| 95 |
+
"eval_recall": 0.583106428317746,
|
| 96 |
+
"eval_runtime": 71.8827,
|
| 97 |
+
"eval_samples_per_second": 328.994,
|
| 98 |
+
"eval_steps_per_second": 5.147,
|
| 99 |
+
"step": 1845
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 5.983739837398374,
|
| 103 |
+
"grad_norm": 1.7785389423370361,
|
| 104 |
+
"learning_rate": 1.3550135501355015e-07,
|
| 105 |
+
"loss": 0.1772,
|
| 106 |
+
"step": 2208
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 6.0,
|
| 110 |
+
"eval_f1": 0.5666028954056137,
|
| 111 |
+
"eval_loss": 0.41467106342315674,
|
| 112 |
+
"eval_precision": 0.5473327081149666,
|
| 113 |
+
"eval_recall": 0.5872795075174618,
|
| 114 |
+
"eval_runtime": 71.6677,
|
| 115 |
+
"eval_samples_per_second": 329.981,
|
| 116 |
+
"eval_steps_per_second": 5.163,
|
| 117 |
+
"step": 2214
|
| 118 |
+
}
|
| 119 |
+
],
|
| 120 |
+
"logging_steps": 368,
|
| 121 |
+
"max_steps": 2214,
|
| 122 |
+
"num_input_tokens_seen": 0,
|
| 123 |
+
"num_train_epochs": 6,
|
| 124 |
+
"save_steps": 500,
|
| 125 |
+
"stateful_callbacks": {
|
| 126 |
+
"TrainerControl": {
|
| 127 |
+
"args": {
|
| 128 |
+
"should_epoch_stop": false,
|
| 129 |
+
"should_evaluate": false,
|
| 130 |
+
"should_log": false,
|
| 131 |
+
"should_save": true,
|
| 132 |
+
"should_training_stop": true
|
| 133 |
+
},
|
| 134 |
+
"attributes": {}
|
| 135 |
+
}
|
| 136 |
+
},
|
| 137 |
+
"total_flos": 2.9681397967022096e+16,
|
| 138 |
+
"train_batch_size": 64,
|
| 139 |
+
"trial_name": null,
|
| 140 |
+
"trial_params": null
|
| 141 |
+
}
|