Upload saved model files
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- checkpoint-104/config.json +152 -0
- checkpoint-104/model.safetensors +3 -0
- checkpoint-104/optimizer.pt +3 -0
- checkpoint-104/rng_state.pth +3 -0
- checkpoint-104/scaler.pt +3 -0
- checkpoint-104/scheduler.pt +3 -0
- checkpoint-104/special_tokens_map.json +37 -0
- checkpoint-104/tokenizer.json +0 -0
- checkpoint-104/tokenizer_config.json +56 -0
- checkpoint-104/trainer_state.json +81 -0
- checkpoint-104/training_args.bin +3 -0
- checkpoint-104/vocab.txt +0 -0
- checkpoint-1040/config.json +152 -0
- checkpoint-1040/model.safetensors +3 -0
- checkpoint-1040/optimizer.pt +3 -0
- checkpoint-1040/rng_state.pth +3 -0
- checkpoint-1040/scaler.pt +3 -0
- checkpoint-1040/scheduler.pt +3 -0
- checkpoint-1040/special_tokens_map.json +37 -0
- checkpoint-1040/tokenizer.json +0 -0
- checkpoint-1040/tokenizer_config.json +56 -0
- checkpoint-1040/trainer_state.json +423 -0
- checkpoint-1040/training_args.bin +3 -0
- checkpoint-1040/vocab.txt +0 -0
- checkpoint-1092/config.json +152 -0
- checkpoint-1092/model.safetensors +3 -0
- checkpoint-1092/optimizer.pt +3 -0
- checkpoint-1092/rng_state.pth +3 -0
- checkpoint-1092/scaler.pt +3 -0
- checkpoint-1092/scheduler.pt +3 -0
- checkpoint-1092/special_tokens_map.json +37 -0
- checkpoint-1092/tokenizer.json +0 -0
- checkpoint-1092/tokenizer_config.json +56 -0
- checkpoint-1092/trainer_state.json +442 -0
- checkpoint-1092/training_args.bin +3 -0
- checkpoint-1092/vocab.txt +0 -0
- checkpoint-1144/config.json +152 -0
- checkpoint-1144/model.safetensors +3 -0
- checkpoint-1144/optimizer.pt +3 -0
- checkpoint-1144/rng_state.pth +3 -0
- checkpoint-1144/scaler.pt +3 -0
- checkpoint-1144/scheduler.pt +3 -0
- checkpoint-1144/special_tokens_map.json +37 -0
- checkpoint-1144/tokenizer.json +0 -0
- checkpoint-1144/tokenizer_config.json +56 -0
- checkpoint-1144/trainer_state.json +461 -0
- checkpoint-1144/training_args.bin +3 -0
- checkpoint-1144/vocab.txt +0 -0
- checkpoint-1196/config.json +152 -0
- checkpoint-1196/model.safetensors +3 -0
checkpoint-104/config.json
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "T1005",
|
| 13 |
+
"1": "T1021",
|
| 14 |
+
"2": "T1027",
|
| 15 |
+
"3": "T1033",
|
| 16 |
+
"4": "T1036",
|
| 17 |
+
"5": "T1041",
|
| 18 |
+
"6": "T1046",
|
| 19 |
+
"7": "T1048",
|
| 20 |
+
"8": "T1049",
|
| 21 |
+
"9": "T1053",
|
| 22 |
+
"10": "T1055",
|
| 23 |
+
"11": "T1056",
|
| 24 |
+
"12": "T1057",
|
| 25 |
+
"13": "T1059",
|
| 26 |
+
"14": "T1070",
|
| 27 |
+
"15": "T1071",
|
| 28 |
+
"16": "T1074",
|
| 29 |
+
"17": "T1078",
|
| 30 |
+
"18": "T1082",
|
| 31 |
+
"19": "T1083",
|
| 32 |
+
"20": "T1098",
|
| 33 |
+
"21": "T1102",
|
| 34 |
+
"22": "T1105",
|
| 35 |
+
"23": "T1110",
|
| 36 |
+
"24": "T1113",
|
| 37 |
+
"25": "T1114",
|
| 38 |
+
"26": "T1115",
|
| 39 |
+
"27": "T1132",
|
| 40 |
+
"28": "T1137",
|
| 41 |
+
"29": "T1140",
|
| 42 |
+
"30": "T1189",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1195",
|
| 45 |
+
"33": "T1203",
|
| 46 |
+
"34": "T1204",
|
| 47 |
+
"35": "T1218",
|
| 48 |
+
"36": "T1486",
|
| 49 |
+
"37": "T1491",
|
| 50 |
+
"38": "T1496",
|
| 51 |
+
"39": "T1497",
|
| 52 |
+
"40": "T1499",
|
| 53 |
+
"41": "T1528",
|
| 54 |
+
"42": "T1539",
|
| 55 |
+
"43": "T1547",
|
| 56 |
+
"44": "T1555",
|
| 57 |
+
"45": "T1557",
|
| 58 |
+
"46": "T1562",
|
| 59 |
+
"47": "T1564",
|
| 60 |
+
"48": "T1566",
|
| 61 |
+
"49": "T1567",
|
| 62 |
+
"50": "T1573",
|
| 63 |
+
"51": "T1574",
|
| 64 |
+
"52": "T1583",
|
| 65 |
+
"53": "T1586",
|
| 66 |
+
"54": "T1589",
|
| 67 |
+
"55": "T1606",
|
| 68 |
+
"56": "T1608",
|
| 69 |
+
"57": "T1614",
|
| 70 |
+
"58": "T1620",
|
| 71 |
+
"59": "T1623.001",
|
| 72 |
+
"60": "T1631.001"
|
| 73 |
+
},
|
| 74 |
+
"initializer_range": 0.02,
|
| 75 |
+
"intermediate_size": 3072,
|
| 76 |
+
"label2id": {
|
| 77 |
+
"T1005": 0,
|
| 78 |
+
"T1021": 1,
|
| 79 |
+
"T1027": 2,
|
| 80 |
+
"T1033": 3,
|
| 81 |
+
"T1036": 4,
|
| 82 |
+
"T1041": 5,
|
| 83 |
+
"T1046": 6,
|
| 84 |
+
"T1048": 7,
|
| 85 |
+
"T1049": 8,
|
| 86 |
+
"T1053": 9,
|
| 87 |
+
"T1055": 10,
|
| 88 |
+
"T1056": 11,
|
| 89 |
+
"T1057": 12,
|
| 90 |
+
"T1059": 13,
|
| 91 |
+
"T1070": 14,
|
| 92 |
+
"T1071": 15,
|
| 93 |
+
"T1074": 16,
|
| 94 |
+
"T1078": 17,
|
| 95 |
+
"T1082": 18,
|
| 96 |
+
"T1083": 19,
|
| 97 |
+
"T1098": 20,
|
| 98 |
+
"T1102": 21,
|
| 99 |
+
"T1105": 22,
|
| 100 |
+
"T1110": 23,
|
| 101 |
+
"T1113": 24,
|
| 102 |
+
"T1114": 25,
|
| 103 |
+
"T1115": 26,
|
| 104 |
+
"T1132": 27,
|
| 105 |
+
"T1137": 28,
|
| 106 |
+
"T1140": 29,
|
| 107 |
+
"T1189": 30,
|
| 108 |
+
"T1190": 31,
|
| 109 |
+
"T1195": 32,
|
| 110 |
+
"T1203": 33,
|
| 111 |
+
"T1204": 34,
|
| 112 |
+
"T1218": 35,
|
| 113 |
+
"T1486": 36,
|
| 114 |
+
"T1491": 37,
|
| 115 |
+
"T1496": 38,
|
| 116 |
+
"T1497": 39,
|
| 117 |
+
"T1499": 40,
|
| 118 |
+
"T1528": 41,
|
| 119 |
+
"T1539": 42,
|
| 120 |
+
"T1547": 43,
|
| 121 |
+
"T1555": 44,
|
| 122 |
+
"T1557": 45,
|
| 123 |
+
"T1562": 46,
|
| 124 |
+
"T1564": 47,
|
| 125 |
+
"T1566": 48,
|
| 126 |
+
"T1567": 49,
|
| 127 |
+
"T1573": 50,
|
| 128 |
+
"T1574": 51,
|
| 129 |
+
"T1583": 52,
|
| 130 |
+
"T1586": 53,
|
| 131 |
+
"T1589": 54,
|
| 132 |
+
"T1606": 55,
|
| 133 |
+
"T1608": 56,
|
| 134 |
+
"T1614": 57,
|
| 135 |
+
"T1620": 58,
|
| 136 |
+
"T1623.001": 59,
|
| 137 |
+
"T1631.001": 60
|
| 138 |
+
},
|
| 139 |
+
"layer_norm_eps": 1e-12,
|
| 140 |
+
"max_position_embeddings": 512,
|
| 141 |
+
"model_type": "bert",
|
| 142 |
+
"num_attention_heads": 12,
|
| 143 |
+
"num_hidden_layers": 12,
|
| 144 |
+
"pad_token_id": 0,
|
| 145 |
+
"position_embedding_type": "absolute",
|
| 146 |
+
"problem_type": "multi_label_classification",
|
| 147 |
+
"torch_dtype": "float32",
|
| 148 |
+
"transformers_version": "4.55.2",
|
| 149 |
+
"type_vocab_size": 2,
|
| 150 |
+
"use_cache": true,
|
| 151 |
+
"vocab_size": 30522
|
| 152 |
+
}
|
checkpoint-104/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7df2fe92b1e0876a7c8d36194c053ae9bcbfd001ece2aa27e0155ad80d63eb4f
|
| 3 |
+
size 438140132
|
checkpoint-104/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b6fc31b3eb14051342e97bf41918a5fdf23de8ad17d018cc2b365fee6382a97
|
| 3 |
+
size 876401210
|
checkpoint-104/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:736858f0ba9663f7ef1e0de90df5597ae3a880c27538fc5eb96e2cb1485eb4e5
|
| 3 |
+
size 14244
|
checkpoint-104/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-104/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:279117f7d4cc0f14fd1502f1e503dda0c2168ebe4feb368500cd5598d6b2c574
|
| 3 |
+
size 1064
|
checkpoint-104/special_tokens_map.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": {
|
| 3 |
+
"content": "[CLS]",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"mask_token": {
|
| 10 |
+
"content": "[MASK]",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"sep_token": {
|
| 24 |
+
"content": "[SEP]",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"unk_token": {
|
| 31 |
+
"content": "[UNK]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
}
|
| 37 |
+
}
|
checkpoint-104/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-104/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": false,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"extra_special_tokens": {},
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"pad_token": "[PAD]",
|
| 51 |
+
"sep_token": "[SEP]",
|
| 52 |
+
"strip_accents": null,
|
| 53 |
+
"tokenize_chinese_chars": true,
|
| 54 |
+
"tokenizer_class": "BertTokenizer",
|
| 55 |
+
"unk_token": "[UNK]"
|
| 56 |
+
}
|
checkpoint-104/trainer_state.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 52,
|
| 3 |
+
"best_metric": 0.0,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-annoctr_step2/checkpoint-52",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 104,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.9615384615384616,
|
| 14 |
+
"grad_norm": 35970.37109375,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5612,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_f1_macro": 0.0,
|
| 22 |
+
"eval_f1_micro": 0.0,
|
| 23 |
+
"eval_loss": 0.33502161502838135,
|
| 24 |
+
"eval_precision": 0.0,
|
| 25 |
+
"eval_recall": 0.0,
|
| 26 |
+
"eval_runtime": 1.6495,
|
| 27 |
+
"eval_samples_per_second": 375.276,
|
| 28 |
+
"eval_steps_per_second": 15.763,
|
| 29 |
+
"step": 52
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.9230769230769231,
|
| 33 |
+
"grad_norm": 10958.1689453125,
|
| 34 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 35 |
+
"loss": 0.214,
|
| 36 |
+
"step": 100
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 2.0,
|
| 40 |
+
"eval_f1_macro": 0.0,
|
| 41 |
+
"eval_f1_micro": 0.0,
|
| 42 |
+
"eval_loss": 0.09798076748847961,
|
| 43 |
+
"eval_precision": 0.0,
|
| 44 |
+
"eval_recall": 0.0,
|
| 45 |
+
"eval_runtime": 1.6513,
|
| 46 |
+
"eval_samples_per_second": 374.861,
|
| 47 |
+
"eval_steps_per_second": 15.745,
|
| 48 |
+
"step": 104
|
| 49 |
+
}
|
| 50 |
+
],
|
| 51 |
+
"logging_steps": 50,
|
| 52 |
+
"max_steps": 5200,
|
| 53 |
+
"num_input_tokens_seen": 0,
|
| 54 |
+
"num_train_epochs": 100,
|
| 55 |
+
"save_steps": 500,
|
| 56 |
+
"stateful_callbacks": {
|
| 57 |
+
"EarlyStoppingCallback": {
|
| 58 |
+
"args": {
|
| 59 |
+
"early_stopping_patience": 5,
|
| 60 |
+
"early_stopping_threshold": 0.0
|
| 61 |
+
},
|
| 62 |
+
"attributes": {
|
| 63 |
+
"early_stopping_patience_counter": 1
|
| 64 |
+
}
|
| 65 |
+
},
|
| 66 |
+
"TrainerControl": {
|
| 67 |
+
"args": {
|
| 68 |
+
"should_epoch_stop": false,
|
| 69 |
+
"should_evaluate": false,
|
| 70 |
+
"should_log": false,
|
| 71 |
+
"should_save": true,
|
| 72 |
+
"should_training_stop": false
|
| 73 |
+
},
|
| 74 |
+
"attributes": {}
|
| 75 |
+
}
|
| 76 |
+
},
|
| 77 |
+
"total_flos": 653914080718848.0,
|
| 78 |
+
"train_batch_size": 24,
|
| 79 |
+
"trial_name": null,
|
| 80 |
+
"trial_params": null
|
| 81 |
+
}
|
checkpoint-104/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f05e8606862008bfc17115034db9429cc42bab3677cf65b2b782cae0ed9dfed
|
| 3 |
+
size 5368
|
checkpoint-104/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1040/config.json
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "T1005",
|
| 13 |
+
"1": "T1021",
|
| 14 |
+
"2": "T1027",
|
| 15 |
+
"3": "T1033",
|
| 16 |
+
"4": "T1036",
|
| 17 |
+
"5": "T1041",
|
| 18 |
+
"6": "T1046",
|
| 19 |
+
"7": "T1048",
|
| 20 |
+
"8": "T1049",
|
| 21 |
+
"9": "T1053",
|
| 22 |
+
"10": "T1055",
|
| 23 |
+
"11": "T1056",
|
| 24 |
+
"12": "T1057",
|
| 25 |
+
"13": "T1059",
|
| 26 |
+
"14": "T1070",
|
| 27 |
+
"15": "T1071",
|
| 28 |
+
"16": "T1074",
|
| 29 |
+
"17": "T1078",
|
| 30 |
+
"18": "T1082",
|
| 31 |
+
"19": "T1083",
|
| 32 |
+
"20": "T1098",
|
| 33 |
+
"21": "T1102",
|
| 34 |
+
"22": "T1105",
|
| 35 |
+
"23": "T1110",
|
| 36 |
+
"24": "T1113",
|
| 37 |
+
"25": "T1114",
|
| 38 |
+
"26": "T1115",
|
| 39 |
+
"27": "T1132",
|
| 40 |
+
"28": "T1137",
|
| 41 |
+
"29": "T1140",
|
| 42 |
+
"30": "T1189",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1195",
|
| 45 |
+
"33": "T1203",
|
| 46 |
+
"34": "T1204",
|
| 47 |
+
"35": "T1218",
|
| 48 |
+
"36": "T1486",
|
| 49 |
+
"37": "T1491",
|
| 50 |
+
"38": "T1496",
|
| 51 |
+
"39": "T1497",
|
| 52 |
+
"40": "T1499",
|
| 53 |
+
"41": "T1528",
|
| 54 |
+
"42": "T1539",
|
| 55 |
+
"43": "T1547",
|
| 56 |
+
"44": "T1555",
|
| 57 |
+
"45": "T1557",
|
| 58 |
+
"46": "T1562",
|
| 59 |
+
"47": "T1564",
|
| 60 |
+
"48": "T1566",
|
| 61 |
+
"49": "T1567",
|
| 62 |
+
"50": "T1573",
|
| 63 |
+
"51": "T1574",
|
| 64 |
+
"52": "T1583",
|
| 65 |
+
"53": "T1586",
|
| 66 |
+
"54": "T1589",
|
| 67 |
+
"55": "T1606",
|
| 68 |
+
"56": "T1608",
|
| 69 |
+
"57": "T1614",
|
| 70 |
+
"58": "T1620",
|
| 71 |
+
"59": "T1623.001",
|
| 72 |
+
"60": "T1631.001"
|
| 73 |
+
},
|
| 74 |
+
"initializer_range": 0.02,
|
| 75 |
+
"intermediate_size": 3072,
|
| 76 |
+
"label2id": {
|
| 77 |
+
"T1005": 0,
|
| 78 |
+
"T1021": 1,
|
| 79 |
+
"T1027": 2,
|
| 80 |
+
"T1033": 3,
|
| 81 |
+
"T1036": 4,
|
| 82 |
+
"T1041": 5,
|
| 83 |
+
"T1046": 6,
|
| 84 |
+
"T1048": 7,
|
| 85 |
+
"T1049": 8,
|
| 86 |
+
"T1053": 9,
|
| 87 |
+
"T1055": 10,
|
| 88 |
+
"T1056": 11,
|
| 89 |
+
"T1057": 12,
|
| 90 |
+
"T1059": 13,
|
| 91 |
+
"T1070": 14,
|
| 92 |
+
"T1071": 15,
|
| 93 |
+
"T1074": 16,
|
| 94 |
+
"T1078": 17,
|
| 95 |
+
"T1082": 18,
|
| 96 |
+
"T1083": 19,
|
| 97 |
+
"T1098": 20,
|
| 98 |
+
"T1102": 21,
|
| 99 |
+
"T1105": 22,
|
| 100 |
+
"T1110": 23,
|
| 101 |
+
"T1113": 24,
|
| 102 |
+
"T1114": 25,
|
| 103 |
+
"T1115": 26,
|
| 104 |
+
"T1132": 27,
|
| 105 |
+
"T1137": 28,
|
| 106 |
+
"T1140": 29,
|
| 107 |
+
"T1189": 30,
|
| 108 |
+
"T1190": 31,
|
| 109 |
+
"T1195": 32,
|
| 110 |
+
"T1203": 33,
|
| 111 |
+
"T1204": 34,
|
| 112 |
+
"T1218": 35,
|
| 113 |
+
"T1486": 36,
|
| 114 |
+
"T1491": 37,
|
| 115 |
+
"T1496": 38,
|
| 116 |
+
"T1497": 39,
|
| 117 |
+
"T1499": 40,
|
| 118 |
+
"T1528": 41,
|
| 119 |
+
"T1539": 42,
|
| 120 |
+
"T1547": 43,
|
| 121 |
+
"T1555": 44,
|
| 122 |
+
"T1557": 45,
|
| 123 |
+
"T1562": 46,
|
| 124 |
+
"T1564": 47,
|
| 125 |
+
"T1566": 48,
|
| 126 |
+
"T1567": 49,
|
| 127 |
+
"T1573": 50,
|
| 128 |
+
"T1574": 51,
|
| 129 |
+
"T1583": 52,
|
| 130 |
+
"T1586": 53,
|
| 131 |
+
"T1589": 54,
|
| 132 |
+
"T1606": 55,
|
| 133 |
+
"T1608": 56,
|
| 134 |
+
"T1614": 57,
|
| 135 |
+
"T1620": 58,
|
| 136 |
+
"T1623.001": 59,
|
| 137 |
+
"T1631.001": 60
|
| 138 |
+
},
|
| 139 |
+
"layer_norm_eps": 1e-12,
|
| 140 |
+
"max_position_embeddings": 512,
|
| 141 |
+
"model_type": "bert",
|
| 142 |
+
"num_attention_heads": 12,
|
| 143 |
+
"num_hidden_layers": 12,
|
| 144 |
+
"pad_token_id": 0,
|
| 145 |
+
"position_embedding_type": "absolute",
|
| 146 |
+
"problem_type": "multi_label_classification",
|
| 147 |
+
"torch_dtype": "float32",
|
| 148 |
+
"transformers_version": "4.55.2",
|
| 149 |
+
"type_vocab_size": 2,
|
| 150 |
+
"use_cache": true,
|
| 151 |
+
"vocab_size": 30522
|
| 152 |
+
}
|
checkpoint-1040/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b896e079d9638f588e023e01f960378b8d230266d7f5f4deea75180d814f7944
|
| 3 |
+
size 438140132
|
checkpoint-1040/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:933df3f27a3b805d3c573fb31e6305a70b24efec69367e894738e2f1ce088a77
|
| 3 |
+
size 876401210
|
checkpoint-1040/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ace5adf5b6afced75a55131c69efa38017a45543db99fff32a13305899ced141
|
| 3 |
+
size 14244
|
checkpoint-1040/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-1040/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9dd306eccedd3d701a952a9eb783d3bc211c2cad0601f18a968e26cc19d0363b
|
| 3 |
+
size 1064
|
checkpoint-1040/special_tokens_map.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": {
|
| 3 |
+
"content": "[CLS]",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"mask_token": {
|
| 10 |
+
"content": "[MASK]",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"sep_token": {
|
| 24 |
+
"content": "[SEP]",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"unk_token": {
|
| 31 |
+
"content": "[UNK]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
}
|
| 37 |
+
}
|
checkpoint-1040/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1040/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": false,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"extra_special_tokens": {},
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"pad_token": "[PAD]",
|
| 51 |
+
"sep_token": "[SEP]",
|
| 52 |
+
"strip_accents": null,
|
| 53 |
+
"tokenize_chinese_chars": true,
|
| 54 |
+
"tokenizer_class": "BertTokenizer",
|
| 55 |
+
"unk_token": "[UNK]"
|
| 56 |
+
}
|
checkpoint-1040/trainer_state.json
ADDED
|
@@ -0,0 +1,423 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 988,
|
| 3 |
+
"best_metric": 0.6585365853658537,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-annoctr_step2/checkpoint-988",
|
| 5 |
+
"epoch": 20.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1040,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.9615384615384616,
|
| 14 |
+
"grad_norm": 35970.37109375,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5612,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_f1_macro": 0.0,
|
| 22 |
+
"eval_f1_micro": 0.0,
|
| 23 |
+
"eval_loss": 0.33502161502838135,
|
| 24 |
+
"eval_precision": 0.0,
|
| 25 |
+
"eval_recall": 0.0,
|
| 26 |
+
"eval_runtime": 1.6495,
|
| 27 |
+
"eval_samples_per_second": 375.276,
|
| 28 |
+
"eval_steps_per_second": 15.763,
|
| 29 |
+
"step": 52
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.9230769230769231,
|
| 33 |
+
"grad_norm": 10958.1689453125,
|
| 34 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 35 |
+
"loss": 0.214,
|
| 36 |
+
"step": 100
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 2.0,
|
| 40 |
+
"eval_f1_macro": 0.0,
|
| 41 |
+
"eval_f1_micro": 0.0,
|
| 42 |
+
"eval_loss": 0.09798076748847961,
|
| 43 |
+
"eval_precision": 0.0,
|
| 44 |
+
"eval_recall": 0.0,
|
| 45 |
+
"eval_runtime": 1.6513,
|
| 46 |
+
"eval_samples_per_second": 374.861,
|
| 47 |
+
"eval_steps_per_second": 15.745,
|
| 48 |
+
"step": 104
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 2.8846153846153846,
|
| 52 |
+
"grad_norm": 5634.60546875,
|
| 53 |
+
"learning_rate": 4.951960784313726e-05,
|
| 54 |
+
"loss": 0.0873,
|
| 55 |
+
"step": 150
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 3.0,
|
| 59 |
+
"eval_f1_macro": 0.0,
|
| 60 |
+
"eval_f1_micro": 0.0,
|
| 61 |
+
"eval_loss": 0.07998213171958923,
|
| 62 |
+
"eval_precision": 0.0,
|
| 63 |
+
"eval_recall": 0.0,
|
| 64 |
+
"eval_runtime": 1.6587,
|
| 65 |
+
"eval_samples_per_second": 373.195,
|
| 66 |
+
"eval_steps_per_second": 15.675,
|
| 67 |
+
"step": 156
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 3.8461538461538463,
|
| 71 |
+
"grad_norm": 6965.6220703125,
|
| 72 |
+
"learning_rate": 4.9029411764705883e-05,
|
| 73 |
+
"loss": 0.0791,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.0,
|
| 78 |
+
"eval_f1_macro": 0.0,
|
| 79 |
+
"eval_f1_micro": 0.0,
|
| 80 |
+
"eval_loss": 0.07983831316232681,
|
| 81 |
+
"eval_precision": 0.0,
|
| 82 |
+
"eval_recall": 0.0,
|
| 83 |
+
"eval_runtime": 1.6476,
|
| 84 |
+
"eval_samples_per_second": 375.688,
|
| 85 |
+
"eval_steps_per_second": 15.78,
|
| 86 |
+
"step": 208
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 4.8076923076923075,
|
| 90 |
+
"grad_norm": 4831.40478515625,
|
| 91 |
+
"learning_rate": 4.8539215686274515e-05,
|
| 92 |
+
"loss": 0.0777,
|
| 93 |
+
"step": 250
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 5.0,
|
| 97 |
+
"eval_f1_macro": 0.0,
|
| 98 |
+
"eval_f1_micro": 0.0,
|
| 99 |
+
"eval_loss": 0.07502060383558273,
|
| 100 |
+
"eval_precision": 0.0,
|
| 101 |
+
"eval_recall": 0.0,
|
| 102 |
+
"eval_runtime": 1.6796,
|
| 103 |
+
"eval_samples_per_second": 368.535,
|
| 104 |
+
"eval_steps_per_second": 15.48,
|
| 105 |
+
"step": 260
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 5.769230769230769,
|
| 109 |
+
"grad_norm": 7363.94580078125,
|
| 110 |
+
"learning_rate": 4.804901960784314e-05,
|
| 111 |
+
"loss": 0.0716,
|
| 112 |
+
"step": 300
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 6.0,
|
| 116 |
+
"eval_f1_macro": 0.038923099933562195,
|
| 117 |
+
"eval_f1_micro": 0.47416413373860183,
|
| 118 |
+
"eval_loss": 0.0661635547876358,
|
| 119 |
+
"eval_precision": 0.9069767441860465,
|
| 120 |
+
"eval_recall": 0.32098765432098764,
|
| 121 |
+
"eval_runtime": 1.6554,
|
| 122 |
+
"eval_samples_per_second": 373.937,
|
| 123 |
+
"eval_steps_per_second": 15.707,
|
| 124 |
+
"step": 312
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 6.730769230769231,
|
| 128 |
+
"grad_norm": 6673.61376953125,
|
| 129 |
+
"learning_rate": 4.7558823529411766e-05,
|
| 130 |
+
"loss": 0.0612,
|
| 131 |
+
"step": 350
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 7.0,
|
| 135 |
+
"eval_f1_macro": 0.04364161045234521,
|
| 136 |
+
"eval_f1_micro": 0.502835538752363,
|
| 137 |
+
"eval_loss": 0.061204444617033005,
|
| 138 |
+
"eval_precision": 0.8085106382978723,
|
| 139 |
+
"eval_recall": 0.36488340192043894,
|
| 140 |
+
"eval_runtime": 1.6748,
|
| 141 |
+
"eval_samples_per_second": 369.593,
|
| 142 |
+
"eval_steps_per_second": 15.524,
|
| 143 |
+
"step": 364
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 7.6923076923076925,
|
| 147 |
+
"grad_norm": 6426.9541015625,
|
| 148 |
+
"learning_rate": 4.70686274509804e-05,
|
| 149 |
+
"loss": 0.0526,
|
| 150 |
+
"step": 400
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 8.0,
|
| 154 |
+
"eval_f1_macro": 0.07288718124374782,
|
| 155 |
+
"eval_f1_micro": 0.5546522131887985,
|
| 156 |
+
"eval_loss": 0.05695081874728203,
|
| 157 |
+
"eval_precision": 0.8121693121693122,
|
| 158 |
+
"eval_recall": 0.42112482853223593,
|
| 159 |
+
"eval_runtime": 1.665,
|
| 160 |
+
"eval_samples_per_second": 371.782,
|
| 161 |
+
"eval_steps_per_second": 15.616,
|
| 162 |
+
"step": 416
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 8.653846153846153,
|
| 166 |
+
"grad_norm": 9758.2529296875,
|
| 167 |
+
"learning_rate": 4.6578431372549016e-05,
|
| 168 |
+
"loss": 0.0433,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 9.0,
|
| 173 |
+
"eval_f1_macro": 0.07471540750966602,
|
| 174 |
+
"eval_f1_micro": 0.545950864422202,
|
| 175 |
+
"eval_loss": 0.05365221947431564,
|
| 176 |
+
"eval_precision": 0.8108108108108109,
|
| 177 |
+
"eval_recall": 0.411522633744856,
|
| 178 |
+
"eval_runtime": 1.6749,
|
| 179 |
+
"eval_samples_per_second": 369.564,
|
| 180 |
+
"eval_steps_per_second": 15.523,
|
| 181 |
+
"step": 468
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 9.615384615384615,
|
| 185 |
+
"grad_norm": 3793.51953125,
|
| 186 |
+
"learning_rate": 4.608823529411765e-05,
|
| 187 |
+
"loss": 0.0397,
|
| 188 |
+
"step": 500
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 10.0,
|
| 192 |
+
"eval_f1_macro": 0.12734138202225653,
|
| 193 |
+
"eval_f1_micro": 0.5939086294416244,
|
| 194 |
+
"eval_loss": 0.052679501473903656,
|
| 195 |
+
"eval_precision": 0.7748344370860927,
|
| 196 |
+
"eval_recall": 0.48148148148148145,
|
| 197 |
+
"eval_runtime": 1.6561,
|
| 198 |
+
"eval_samples_per_second": 373.762,
|
| 199 |
+
"eval_steps_per_second": 15.699,
|
| 200 |
+
"step": 520
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 10.576923076923077,
|
| 204 |
+
"grad_norm": 4446.76611328125,
|
| 205 |
+
"learning_rate": 4.559803921568628e-05,
|
| 206 |
+
"loss": 0.0329,
|
| 207 |
+
"step": 550
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 11.0,
|
| 211 |
+
"eval_f1_macro": 0.1493812021209619,
|
| 212 |
+
"eval_f1_micro": 0.609735269000854,
|
| 213 |
+
"eval_loss": 0.05011816695332527,
|
| 214 |
+
"eval_precision": 0.8076923076923077,
|
| 215 |
+
"eval_recall": 0.4897119341563786,
|
| 216 |
+
"eval_runtime": 1.6806,
|
| 217 |
+
"eval_samples_per_second": 368.324,
|
| 218 |
+
"eval_steps_per_second": 15.471,
|
| 219 |
+
"step": 572
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 11.538461538461538,
|
| 223 |
+
"grad_norm": 5606.7880859375,
|
| 224 |
+
"learning_rate": 4.51078431372549e-05,
|
| 225 |
+
"loss": 0.0286,
|
| 226 |
+
"step": 600
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 12.0,
|
| 230 |
+
"eval_f1_macro": 0.19655294907558352,
|
| 231 |
+
"eval_f1_micro": 0.6096959737058341,
|
| 232 |
+
"eval_loss": 0.05061562359333038,
|
| 233 |
+
"eval_precision": 0.7602459016393442,
|
| 234 |
+
"eval_recall": 0.5089163237311386,
|
| 235 |
+
"eval_runtime": 1.6713,
|
| 236 |
+
"eval_samples_per_second": 370.373,
|
| 237 |
+
"eval_steps_per_second": 15.557,
|
| 238 |
+
"step": 624
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 12.5,
|
| 242 |
+
"grad_norm": 5415.85546875,
|
| 243 |
+
"learning_rate": 4.461764705882353e-05,
|
| 244 |
+
"loss": 0.0254,
|
| 245 |
+
"step": 650
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 13.0,
|
| 249 |
+
"eval_f1_macro": 0.20270281237441773,
|
| 250 |
+
"eval_f1_micro": 0.6230831315577078,
|
| 251 |
+
"eval_loss": 0.04895725101232529,
|
| 252 |
+
"eval_precision": 0.7568627450980392,
|
| 253 |
+
"eval_recall": 0.5294924554183813,
|
| 254 |
+
"eval_runtime": 2.2359,
|
| 255 |
+
"eval_samples_per_second": 276.851,
|
| 256 |
+
"eval_steps_per_second": 11.629,
|
| 257 |
+
"step": 676
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 13.461538461538462,
|
| 261 |
+
"grad_norm": 6214.2744140625,
|
| 262 |
+
"learning_rate": 4.412745098039216e-05,
|
| 263 |
+
"loss": 0.023,
|
| 264 |
+
"step": 700
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 14.0,
|
| 268 |
+
"eval_f1_macro": 0.21593787964288247,
|
| 269 |
+
"eval_f1_micro": 0.6309904153354633,
|
| 270 |
+
"eval_loss": 0.047714490443468094,
|
| 271 |
+
"eval_precision": 0.7552581261950286,
|
| 272 |
+
"eval_recall": 0.541838134430727,
|
| 273 |
+
"eval_runtime": 1.7054,
|
| 274 |
+
"eval_samples_per_second": 362.956,
|
| 275 |
+
"eval_steps_per_second": 15.245,
|
| 276 |
+
"step": 728
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 14.423076923076923,
|
| 280 |
+
"grad_norm": 5322.5205078125,
|
| 281 |
+
"learning_rate": 4.363725490196079e-05,
|
| 282 |
+
"loss": 0.0202,
|
| 283 |
+
"step": 750
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 15.0,
|
| 287 |
+
"eval_f1_macro": 0.2127885503653234,
|
| 288 |
+
"eval_f1_micro": 0.6297739672642245,
|
| 289 |
+
"eval_loss": 0.04875025525689125,
|
| 290 |
+
"eval_precision": 0.7292418772563177,
|
| 291 |
+
"eval_recall": 0.5541838134430727,
|
| 292 |
+
"eval_runtime": 1.6728,
|
| 293 |
+
"eval_samples_per_second": 370.029,
|
| 294 |
+
"eval_steps_per_second": 15.542,
|
| 295 |
+
"step": 780
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 15.384615384615385,
|
| 299 |
+
"grad_norm": 3209.152587890625,
|
| 300 |
+
"learning_rate": 4.3147058823529413e-05,
|
| 301 |
+
"loss": 0.0186,
|
| 302 |
+
"step": 800
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 16.0,
|
| 306 |
+
"eval_f1_macro": 0.2234013473846607,
|
| 307 |
+
"eval_f1_micro": 0.6476484194294526,
|
| 308 |
+
"eval_loss": 0.04788675159215927,
|
| 309 |
+
"eval_precision": 0.7394366197183099,
|
| 310 |
+
"eval_recall": 0.5761316872427984,
|
| 311 |
+
"eval_runtime": 1.6703,
|
| 312 |
+
"eval_samples_per_second": 370.591,
|
| 313 |
+
"eval_steps_per_second": 15.566,
|
| 314 |
+
"step": 832
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 16.346153846153847,
|
| 318 |
+
"grad_norm": 6189.79296875,
|
| 319 |
+
"learning_rate": 4.265686274509804e-05,
|
| 320 |
+
"loss": 0.0166,
|
| 321 |
+
"step": 850
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 17.0,
|
| 325 |
+
"eval_f1_macro": 0.2461245877407925,
|
| 326 |
+
"eval_f1_micro": 0.6392067124332571,
|
| 327 |
+
"eval_loss": 0.047994960099458694,
|
| 328 |
+
"eval_precision": 0.7199312714776632,
|
| 329 |
+
"eval_recall": 0.5747599451303155,
|
| 330 |
+
"eval_runtime": 1.6784,
|
| 331 |
+
"eval_samples_per_second": 368.8,
|
| 332 |
+
"eval_steps_per_second": 15.491,
|
| 333 |
+
"step": 884
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 17.307692307692307,
|
| 337 |
+
"grad_norm": 2584.114013671875,
|
| 338 |
+
"learning_rate": 4.216666666666667e-05,
|
| 339 |
+
"loss": 0.0148,
|
| 340 |
+
"step": 900
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 18.0,
|
| 344 |
+
"eval_f1_macro": 0.2672202679954545,
|
| 345 |
+
"eval_f1_micro": 0.6533742331288344,
|
| 346 |
+
"eval_loss": 0.04750063270330429,
|
| 347 |
+
"eval_precision": 0.7408695652173913,
|
| 348 |
+
"eval_recall": 0.5843621399176955,
|
| 349 |
+
"eval_runtime": 1.6744,
|
| 350 |
+
"eval_samples_per_second": 369.683,
|
| 351 |
+
"eval_steps_per_second": 15.528,
|
| 352 |
+
"step": 936
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 18.26923076923077,
|
| 356 |
+
"grad_norm": 2617.191650390625,
|
| 357 |
+
"learning_rate": 4.1676470588235296e-05,
|
| 358 |
+
"loss": 0.0128,
|
| 359 |
+
"step": 950
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 19.0,
|
| 363 |
+
"eval_f1_macro": 0.25478648592377506,
|
| 364 |
+
"eval_f1_micro": 0.6585365853658537,
|
| 365 |
+
"eval_loss": 0.04794125631451607,
|
| 366 |
+
"eval_precision": 0.7409948542024014,
|
| 367 |
+
"eval_recall": 0.5925925925925926,
|
| 368 |
+
"eval_runtime": 1.6751,
|
| 369 |
+
"eval_samples_per_second": 369.525,
|
| 370 |
+
"eval_steps_per_second": 15.521,
|
| 371 |
+
"step": 988
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 19.23076923076923,
|
| 375 |
+
"grad_norm": 2930.471923828125,
|
| 376 |
+
"learning_rate": 4.118627450980392e-05,
|
| 377 |
+
"loss": 0.0116,
|
| 378 |
+
"step": 1000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 20.0,
|
| 382 |
+
"eval_f1_macro": 0.2563341548388003,
|
| 383 |
+
"eval_f1_micro": 0.6470143613000756,
|
| 384 |
+
"eval_loss": 0.04850601404905319,
|
| 385 |
+
"eval_precision": 0.7205387205387206,
|
| 386 |
+
"eval_recall": 0.5871056241426612,
|
| 387 |
+
"eval_runtime": 1.6688,
|
| 388 |
+
"eval_samples_per_second": 370.936,
|
| 389 |
+
"eval_steps_per_second": 15.58,
|
| 390 |
+
"step": 1040
|
| 391 |
+
}
|
| 392 |
+
],
|
| 393 |
+
"logging_steps": 50,
|
| 394 |
+
"max_steps": 5200,
|
| 395 |
+
"num_input_tokens_seen": 0,
|
| 396 |
+
"num_train_epochs": 100,
|
| 397 |
+
"save_steps": 500,
|
| 398 |
+
"stateful_callbacks": {
|
| 399 |
+
"EarlyStoppingCallback": {
|
| 400 |
+
"args": {
|
| 401 |
+
"early_stopping_patience": 5,
|
| 402 |
+
"early_stopping_threshold": 0.0
|
| 403 |
+
},
|
| 404 |
+
"attributes": {
|
| 405 |
+
"early_stopping_patience_counter": 1
|
| 406 |
+
}
|
| 407 |
+
},
|
| 408 |
+
"TrainerControl": {
|
| 409 |
+
"args": {
|
| 410 |
+
"should_epoch_stop": false,
|
| 411 |
+
"should_evaluate": false,
|
| 412 |
+
"should_log": false,
|
| 413 |
+
"should_save": true,
|
| 414 |
+
"should_training_stop": false
|
| 415 |
+
},
|
| 416 |
+
"attributes": {}
|
| 417 |
+
}
|
| 418 |
+
},
|
| 419 |
+
"total_flos": 6539140807188480.0,
|
| 420 |
+
"train_batch_size": 24,
|
| 421 |
+
"trial_name": null,
|
| 422 |
+
"trial_params": null
|
| 423 |
+
}
|
checkpoint-1040/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f05e8606862008bfc17115034db9429cc42bab3677cf65b2b782cae0ed9dfed
|
| 3 |
+
size 5368
|
checkpoint-1040/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1092/config.json
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "T1005",
|
| 13 |
+
"1": "T1021",
|
| 14 |
+
"2": "T1027",
|
| 15 |
+
"3": "T1033",
|
| 16 |
+
"4": "T1036",
|
| 17 |
+
"5": "T1041",
|
| 18 |
+
"6": "T1046",
|
| 19 |
+
"7": "T1048",
|
| 20 |
+
"8": "T1049",
|
| 21 |
+
"9": "T1053",
|
| 22 |
+
"10": "T1055",
|
| 23 |
+
"11": "T1056",
|
| 24 |
+
"12": "T1057",
|
| 25 |
+
"13": "T1059",
|
| 26 |
+
"14": "T1070",
|
| 27 |
+
"15": "T1071",
|
| 28 |
+
"16": "T1074",
|
| 29 |
+
"17": "T1078",
|
| 30 |
+
"18": "T1082",
|
| 31 |
+
"19": "T1083",
|
| 32 |
+
"20": "T1098",
|
| 33 |
+
"21": "T1102",
|
| 34 |
+
"22": "T1105",
|
| 35 |
+
"23": "T1110",
|
| 36 |
+
"24": "T1113",
|
| 37 |
+
"25": "T1114",
|
| 38 |
+
"26": "T1115",
|
| 39 |
+
"27": "T1132",
|
| 40 |
+
"28": "T1137",
|
| 41 |
+
"29": "T1140",
|
| 42 |
+
"30": "T1189",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1195",
|
| 45 |
+
"33": "T1203",
|
| 46 |
+
"34": "T1204",
|
| 47 |
+
"35": "T1218",
|
| 48 |
+
"36": "T1486",
|
| 49 |
+
"37": "T1491",
|
| 50 |
+
"38": "T1496",
|
| 51 |
+
"39": "T1497",
|
| 52 |
+
"40": "T1499",
|
| 53 |
+
"41": "T1528",
|
| 54 |
+
"42": "T1539",
|
| 55 |
+
"43": "T1547",
|
| 56 |
+
"44": "T1555",
|
| 57 |
+
"45": "T1557",
|
| 58 |
+
"46": "T1562",
|
| 59 |
+
"47": "T1564",
|
| 60 |
+
"48": "T1566",
|
| 61 |
+
"49": "T1567",
|
| 62 |
+
"50": "T1573",
|
| 63 |
+
"51": "T1574",
|
| 64 |
+
"52": "T1583",
|
| 65 |
+
"53": "T1586",
|
| 66 |
+
"54": "T1589",
|
| 67 |
+
"55": "T1606",
|
| 68 |
+
"56": "T1608",
|
| 69 |
+
"57": "T1614",
|
| 70 |
+
"58": "T1620",
|
| 71 |
+
"59": "T1623.001",
|
| 72 |
+
"60": "T1631.001"
|
| 73 |
+
},
|
| 74 |
+
"initializer_range": 0.02,
|
| 75 |
+
"intermediate_size": 3072,
|
| 76 |
+
"label2id": {
|
| 77 |
+
"T1005": 0,
|
| 78 |
+
"T1021": 1,
|
| 79 |
+
"T1027": 2,
|
| 80 |
+
"T1033": 3,
|
| 81 |
+
"T1036": 4,
|
| 82 |
+
"T1041": 5,
|
| 83 |
+
"T1046": 6,
|
| 84 |
+
"T1048": 7,
|
| 85 |
+
"T1049": 8,
|
| 86 |
+
"T1053": 9,
|
| 87 |
+
"T1055": 10,
|
| 88 |
+
"T1056": 11,
|
| 89 |
+
"T1057": 12,
|
| 90 |
+
"T1059": 13,
|
| 91 |
+
"T1070": 14,
|
| 92 |
+
"T1071": 15,
|
| 93 |
+
"T1074": 16,
|
| 94 |
+
"T1078": 17,
|
| 95 |
+
"T1082": 18,
|
| 96 |
+
"T1083": 19,
|
| 97 |
+
"T1098": 20,
|
| 98 |
+
"T1102": 21,
|
| 99 |
+
"T1105": 22,
|
| 100 |
+
"T1110": 23,
|
| 101 |
+
"T1113": 24,
|
| 102 |
+
"T1114": 25,
|
| 103 |
+
"T1115": 26,
|
| 104 |
+
"T1132": 27,
|
| 105 |
+
"T1137": 28,
|
| 106 |
+
"T1140": 29,
|
| 107 |
+
"T1189": 30,
|
| 108 |
+
"T1190": 31,
|
| 109 |
+
"T1195": 32,
|
| 110 |
+
"T1203": 33,
|
| 111 |
+
"T1204": 34,
|
| 112 |
+
"T1218": 35,
|
| 113 |
+
"T1486": 36,
|
| 114 |
+
"T1491": 37,
|
| 115 |
+
"T1496": 38,
|
| 116 |
+
"T1497": 39,
|
| 117 |
+
"T1499": 40,
|
| 118 |
+
"T1528": 41,
|
| 119 |
+
"T1539": 42,
|
| 120 |
+
"T1547": 43,
|
| 121 |
+
"T1555": 44,
|
| 122 |
+
"T1557": 45,
|
| 123 |
+
"T1562": 46,
|
| 124 |
+
"T1564": 47,
|
| 125 |
+
"T1566": 48,
|
| 126 |
+
"T1567": 49,
|
| 127 |
+
"T1573": 50,
|
| 128 |
+
"T1574": 51,
|
| 129 |
+
"T1583": 52,
|
| 130 |
+
"T1586": 53,
|
| 131 |
+
"T1589": 54,
|
| 132 |
+
"T1606": 55,
|
| 133 |
+
"T1608": 56,
|
| 134 |
+
"T1614": 57,
|
| 135 |
+
"T1620": 58,
|
| 136 |
+
"T1623.001": 59,
|
| 137 |
+
"T1631.001": 60
|
| 138 |
+
},
|
| 139 |
+
"layer_norm_eps": 1e-12,
|
| 140 |
+
"max_position_embeddings": 512,
|
| 141 |
+
"model_type": "bert",
|
| 142 |
+
"num_attention_heads": 12,
|
| 143 |
+
"num_hidden_layers": 12,
|
| 144 |
+
"pad_token_id": 0,
|
| 145 |
+
"position_embedding_type": "absolute",
|
| 146 |
+
"problem_type": "multi_label_classification",
|
| 147 |
+
"torch_dtype": "float32",
|
| 148 |
+
"transformers_version": "4.55.2",
|
| 149 |
+
"type_vocab_size": 2,
|
| 150 |
+
"use_cache": true,
|
| 151 |
+
"vocab_size": 30522
|
| 152 |
+
}
|
checkpoint-1092/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c9fb7b7785f54b70538249b6e13313b0908ac89988469530be141016d608dfcd
|
| 3 |
+
size 438140132
|
checkpoint-1092/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc910d9686defdb41ffd883291715f2b22578d048fb693839b20d4792359cac4
|
| 3 |
+
size 876401210
|
checkpoint-1092/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5988312761947fb69c8c90536e0665fe3c0ab245c725afc50794702d835f93b2
|
| 3 |
+
size 14244
|
checkpoint-1092/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-1092/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdd7f8b20fb7f9669142e0bbf4b736114d8316d103a75f1eb23facb1d1d7569f
|
| 3 |
+
size 1064
|
checkpoint-1092/special_tokens_map.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": {
|
| 3 |
+
"content": "[CLS]",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"mask_token": {
|
| 10 |
+
"content": "[MASK]",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"sep_token": {
|
| 24 |
+
"content": "[SEP]",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"unk_token": {
|
| 31 |
+
"content": "[UNK]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
}
|
| 37 |
+
}
|
checkpoint-1092/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1092/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": false,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"extra_special_tokens": {},
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"pad_token": "[PAD]",
|
| 51 |
+
"sep_token": "[SEP]",
|
| 52 |
+
"strip_accents": null,
|
| 53 |
+
"tokenize_chinese_chars": true,
|
| 54 |
+
"tokenizer_class": "BertTokenizer",
|
| 55 |
+
"unk_token": "[UNK]"
|
| 56 |
+
}
|
checkpoint-1092/trainer_state.json
ADDED
|
@@ -0,0 +1,442 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 988,
|
| 3 |
+
"best_metric": 0.6585365853658537,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-annoctr_step2/checkpoint-988",
|
| 5 |
+
"epoch": 21.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1092,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.9615384615384616,
|
| 14 |
+
"grad_norm": 35970.37109375,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5612,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_f1_macro": 0.0,
|
| 22 |
+
"eval_f1_micro": 0.0,
|
| 23 |
+
"eval_loss": 0.33502161502838135,
|
| 24 |
+
"eval_precision": 0.0,
|
| 25 |
+
"eval_recall": 0.0,
|
| 26 |
+
"eval_runtime": 1.6495,
|
| 27 |
+
"eval_samples_per_second": 375.276,
|
| 28 |
+
"eval_steps_per_second": 15.763,
|
| 29 |
+
"step": 52
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.9230769230769231,
|
| 33 |
+
"grad_norm": 10958.1689453125,
|
| 34 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 35 |
+
"loss": 0.214,
|
| 36 |
+
"step": 100
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 2.0,
|
| 40 |
+
"eval_f1_macro": 0.0,
|
| 41 |
+
"eval_f1_micro": 0.0,
|
| 42 |
+
"eval_loss": 0.09798076748847961,
|
| 43 |
+
"eval_precision": 0.0,
|
| 44 |
+
"eval_recall": 0.0,
|
| 45 |
+
"eval_runtime": 1.6513,
|
| 46 |
+
"eval_samples_per_second": 374.861,
|
| 47 |
+
"eval_steps_per_second": 15.745,
|
| 48 |
+
"step": 104
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 2.8846153846153846,
|
| 52 |
+
"grad_norm": 5634.60546875,
|
| 53 |
+
"learning_rate": 4.951960784313726e-05,
|
| 54 |
+
"loss": 0.0873,
|
| 55 |
+
"step": 150
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 3.0,
|
| 59 |
+
"eval_f1_macro": 0.0,
|
| 60 |
+
"eval_f1_micro": 0.0,
|
| 61 |
+
"eval_loss": 0.07998213171958923,
|
| 62 |
+
"eval_precision": 0.0,
|
| 63 |
+
"eval_recall": 0.0,
|
| 64 |
+
"eval_runtime": 1.6587,
|
| 65 |
+
"eval_samples_per_second": 373.195,
|
| 66 |
+
"eval_steps_per_second": 15.675,
|
| 67 |
+
"step": 156
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 3.8461538461538463,
|
| 71 |
+
"grad_norm": 6965.6220703125,
|
| 72 |
+
"learning_rate": 4.9029411764705883e-05,
|
| 73 |
+
"loss": 0.0791,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.0,
|
| 78 |
+
"eval_f1_macro": 0.0,
|
| 79 |
+
"eval_f1_micro": 0.0,
|
| 80 |
+
"eval_loss": 0.07983831316232681,
|
| 81 |
+
"eval_precision": 0.0,
|
| 82 |
+
"eval_recall": 0.0,
|
| 83 |
+
"eval_runtime": 1.6476,
|
| 84 |
+
"eval_samples_per_second": 375.688,
|
| 85 |
+
"eval_steps_per_second": 15.78,
|
| 86 |
+
"step": 208
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 4.8076923076923075,
|
| 90 |
+
"grad_norm": 4831.40478515625,
|
| 91 |
+
"learning_rate": 4.8539215686274515e-05,
|
| 92 |
+
"loss": 0.0777,
|
| 93 |
+
"step": 250
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 5.0,
|
| 97 |
+
"eval_f1_macro": 0.0,
|
| 98 |
+
"eval_f1_micro": 0.0,
|
| 99 |
+
"eval_loss": 0.07502060383558273,
|
| 100 |
+
"eval_precision": 0.0,
|
| 101 |
+
"eval_recall": 0.0,
|
| 102 |
+
"eval_runtime": 1.6796,
|
| 103 |
+
"eval_samples_per_second": 368.535,
|
| 104 |
+
"eval_steps_per_second": 15.48,
|
| 105 |
+
"step": 260
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 5.769230769230769,
|
| 109 |
+
"grad_norm": 7363.94580078125,
|
| 110 |
+
"learning_rate": 4.804901960784314e-05,
|
| 111 |
+
"loss": 0.0716,
|
| 112 |
+
"step": 300
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 6.0,
|
| 116 |
+
"eval_f1_macro": 0.038923099933562195,
|
| 117 |
+
"eval_f1_micro": 0.47416413373860183,
|
| 118 |
+
"eval_loss": 0.0661635547876358,
|
| 119 |
+
"eval_precision": 0.9069767441860465,
|
| 120 |
+
"eval_recall": 0.32098765432098764,
|
| 121 |
+
"eval_runtime": 1.6554,
|
| 122 |
+
"eval_samples_per_second": 373.937,
|
| 123 |
+
"eval_steps_per_second": 15.707,
|
| 124 |
+
"step": 312
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 6.730769230769231,
|
| 128 |
+
"grad_norm": 6673.61376953125,
|
| 129 |
+
"learning_rate": 4.7558823529411766e-05,
|
| 130 |
+
"loss": 0.0612,
|
| 131 |
+
"step": 350
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 7.0,
|
| 135 |
+
"eval_f1_macro": 0.04364161045234521,
|
| 136 |
+
"eval_f1_micro": 0.502835538752363,
|
| 137 |
+
"eval_loss": 0.061204444617033005,
|
| 138 |
+
"eval_precision": 0.8085106382978723,
|
| 139 |
+
"eval_recall": 0.36488340192043894,
|
| 140 |
+
"eval_runtime": 1.6748,
|
| 141 |
+
"eval_samples_per_second": 369.593,
|
| 142 |
+
"eval_steps_per_second": 15.524,
|
| 143 |
+
"step": 364
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 7.6923076923076925,
|
| 147 |
+
"grad_norm": 6426.9541015625,
|
| 148 |
+
"learning_rate": 4.70686274509804e-05,
|
| 149 |
+
"loss": 0.0526,
|
| 150 |
+
"step": 400
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 8.0,
|
| 154 |
+
"eval_f1_macro": 0.07288718124374782,
|
| 155 |
+
"eval_f1_micro": 0.5546522131887985,
|
| 156 |
+
"eval_loss": 0.05695081874728203,
|
| 157 |
+
"eval_precision": 0.8121693121693122,
|
| 158 |
+
"eval_recall": 0.42112482853223593,
|
| 159 |
+
"eval_runtime": 1.665,
|
| 160 |
+
"eval_samples_per_second": 371.782,
|
| 161 |
+
"eval_steps_per_second": 15.616,
|
| 162 |
+
"step": 416
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 8.653846153846153,
|
| 166 |
+
"grad_norm": 9758.2529296875,
|
| 167 |
+
"learning_rate": 4.6578431372549016e-05,
|
| 168 |
+
"loss": 0.0433,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 9.0,
|
| 173 |
+
"eval_f1_macro": 0.07471540750966602,
|
| 174 |
+
"eval_f1_micro": 0.545950864422202,
|
| 175 |
+
"eval_loss": 0.05365221947431564,
|
| 176 |
+
"eval_precision": 0.8108108108108109,
|
| 177 |
+
"eval_recall": 0.411522633744856,
|
| 178 |
+
"eval_runtime": 1.6749,
|
| 179 |
+
"eval_samples_per_second": 369.564,
|
| 180 |
+
"eval_steps_per_second": 15.523,
|
| 181 |
+
"step": 468
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 9.615384615384615,
|
| 185 |
+
"grad_norm": 3793.51953125,
|
| 186 |
+
"learning_rate": 4.608823529411765e-05,
|
| 187 |
+
"loss": 0.0397,
|
| 188 |
+
"step": 500
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 10.0,
|
| 192 |
+
"eval_f1_macro": 0.12734138202225653,
|
| 193 |
+
"eval_f1_micro": 0.5939086294416244,
|
| 194 |
+
"eval_loss": 0.052679501473903656,
|
| 195 |
+
"eval_precision": 0.7748344370860927,
|
| 196 |
+
"eval_recall": 0.48148148148148145,
|
| 197 |
+
"eval_runtime": 1.6561,
|
| 198 |
+
"eval_samples_per_second": 373.762,
|
| 199 |
+
"eval_steps_per_second": 15.699,
|
| 200 |
+
"step": 520
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 10.576923076923077,
|
| 204 |
+
"grad_norm": 4446.76611328125,
|
| 205 |
+
"learning_rate": 4.559803921568628e-05,
|
| 206 |
+
"loss": 0.0329,
|
| 207 |
+
"step": 550
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 11.0,
|
| 211 |
+
"eval_f1_macro": 0.1493812021209619,
|
| 212 |
+
"eval_f1_micro": 0.609735269000854,
|
| 213 |
+
"eval_loss": 0.05011816695332527,
|
| 214 |
+
"eval_precision": 0.8076923076923077,
|
| 215 |
+
"eval_recall": 0.4897119341563786,
|
| 216 |
+
"eval_runtime": 1.6806,
|
| 217 |
+
"eval_samples_per_second": 368.324,
|
| 218 |
+
"eval_steps_per_second": 15.471,
|
| 219 |
+
"step": 572
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 11.538461538461538,
|
| 223 |
+
"grad_norm": 5606.7880859375,
|
| 224 |
+
"learning_rate": 4.51078431372549e-05,
|
| 225 |
+
"loss": 0.0286,
|
| 226 |
+
"step": 600
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 12.0,
|
| 230 |
+
"eval_f1_macro": 0.19655294907558352,
|
| 231 |
+
"eval_f1_micro": 0.6096959737058341,
|
| 232 |
+
"eval_loss": 0.05061562359333038,
|
| 233 |
+
"eval_precision": 0.7602459016393442,
|
| 234 |
+
"eval_recall": 0.5089163237311386,
|
| 235 |
+
"eval_runtime": 1.6713,
|
| 236 |
+
"eval_samples_per_second": 370.373,
|
| 237 |
+
"eval_steps_per_second": 15.557,
|
| 238 |
+
"step": 624
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 12.5,
|
| 242 |
+
"grad_norm": 5415.85546875,
|
| 243 |
+
"learning_rate": 4.461764705882353e-05,
|
| 244 |
+
"loss": 0.0254,
|
| 245 |
+
"step": 650
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 13.0,
|
| 249 |
+
"eval_f1_macro": 0.20270281237441773,
|
| 250 |
+
"eval_f1_micro": 0.6230831315577078,
|
| 251 |
+
"eval_loss": 0.04895725101232529,
|
| 252 |
+
"eval_precision": 0.7568627450980392,
|
| 253 |
+
"eval_recall": 0.5294924554183813,
|
| 254 |
+
"eval_runtime": 2.2359,
|
| 255 |
+
"eval_samples_per_second": 276.851,
|
| 256 |
+
"eval_steps_per_second": 11.629,
|
| 257 |
+
"step": 676
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 13.461538461538462,
|
| 261 |
+
"grad_norm": 6214.2744140625,
|
| 262 |
+
"learning_rate": 4.412745098039216e-05,
|
| 263 |
+
"loss": 0.023,
|
| 264 |
+
"step": 700
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 14.0,
|
| 268 |
+
"eval_f1_macro": 0.21593787964288247,
|
| 269 |
+
"eval_f1_micro": 0.6309904153354633,
|
| 270 |
+
"eval_loss": 0.047714490443468094,
|
| 271 |
+
"eval_precision": 0.7552581261950286,
|
| 272 |
+
"eval_recall": 0.541838134430727,
|
| 273 |
+
"eval_runtime": 1.7054,
|
| 274 |
+
"eval_samples_per_second": 362.956,
|
| 275 |
+
"eval_steps_per_second": 15.245,
|
| 276 |
+
"step": 728
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 14.423076923076923,
|
| 280 |
+
"grad_norm": 5322.5205078125,
|
| 281 |
+
"learning_rate": 4.363725490196079e-05,
|
| 282 |
+
"loss": 0.0202,
|
| 283 |
+
"step": 750
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 15.0,
|
| 287 |
+
"eval_f1_macro": 0.2127885503653234,
|
| 288 |
+
"eval_f1_micro": 0.6297739672642245,
|
| 289 |
+
"eval_loss": 0.04875025525689125,
|
| 290 |
+
"eval_precision": 0.7292418772563177,
|
| 291 |
+
"eval_recall": 0.5541838134430727,
|
| 292 |
+
"eval_runtime": 1.6728,
|
| 293 |
+
"eval_samples_per_second": 370.029,
|
| 294 |
+
"eval_steps_per_second": 15.542,
|
| 295 |
+
"step": 780
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 15.384615384615385,
|
| 299 |
+
"grad_norm": 3209.152587890625,
|
| 300 |
+
"learning_rate": 4.3147058823529413e-05,
|
| 301 |
+
"loss": 0.0186,
|
| 302 |
+
"step": 800
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 16.0,
|
| 306 |
+
"eval_f1_macro": 0.2234013473846607,
|
| 307 |
+
"eval_f1_micro": 0.6476484194294526,
|
| 308 |
+
"eval_loss": 0.04788675159215927,
|
| 309 |
+
"eval_precision": 0.7394366197183099,
|
| 310 |
+
"eval_recall": 0.5761316872427984,
|
| 311 |
+
"eval_runtime": 1.6703,
|
| 312 |
+
"eval_samples_per_second": 370.591,
|
| 313 |
+
"eval_steps_per_second": 15.566,
|
| 314 |
+
"step": 832
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 16.346153846153847,
|
| 318 |
+
"grad_norm": 6189.79296875,
|
| 319 |
+
"learning_rate": 4.265686274509804e-05,
|
| 320 |
+
"loss": 0.0166,
|
| 321 |
+
"step": 850
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 17.0,
|
| 325 |
+
"eval_f1_macro": 0.2461245877407925,
|
| 326 |
+
"eval_f1_micro": 0.6392067124332571,
|
| 327 |
+
"eval_loss": 0.047994960099458694,
|
| 328 |
+
"eval_precision": 0.7199312714776632,
|
| 329 |
+
"eval_recall": 0.5747599451303155,
|
| 330 |
+
"eval_runtime": 1.6784,
|
| 331 |
+
"eval_samples_per_second": 368.8,
|
| 332 |
+
"eval_steps_per_second": 15.491,
|
| 333 |
+
"step": 884
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 17.307692307692307,
|
| 337 |
+
"grad_norm": 2584.114013671875,
|
| 338 |
+
"learning_rate": 4.216666666666667e-05,
|
| 339 |
+
"loss": 0.0148,
|
| 340 |
+
"step": 900
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 18.0,
|
| 344 |
+
"eval_f1_macro": 0.2672202679954545,
|
| 345 |
+
"eval_f1_micro": 0.6533742331288344,
|
| 346 |
+
"eval_loss": 0.04750063270330429,
|
| 347 |
+
"eval_precision": 0.7408695652173913,
|
| 348 |
+
"eval_recall": 0.5843621399176955,
|
| 349 |
+
"eval_runtime": 1.6744,
|
| 350 |
+
"eval_samples_per_second": 369.683,
|
| 351 |
+
"eval_steps_per_second": 15.528,
|
| 352 |
+
"step": 936
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 18.26923076923077,
|
| 356 |
+
"grad_norm": 2617.191650390625,
|
| 357 |
+
"learning_rate": 4.1676470588235296e-05,
|
| 358 |
+
"loss": 0.0128,
|
| 359 |
+
"step": 950
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 19.0,
|
| 363 |
+
"eval_f1_macro": 0.25478648592377506,
|
| 364 |
+
"eval_f1_micro": 0.6585365853658537,
|
| 365 |
+
"eval_loss": 0.04794125631451607,
|
| 366 |
+
"eval_precision": 0.7409948542024014,
|
| 367 |
+
"eval_recall": 0.5925925925925926,
|
| 368 |
+
"eval_runtime": 1.6751,
|
| 369 |
+
"eval_samples_per_second": 369.525,
|
| 370 |
+
"eval_steps_per_second": 15.521,
|
| 371 |
+
"step": 988
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 19.23076923076923,
|
| 375 |
+
"grad_norm": 2930.471923828125,
|
| 376 |
+
"learning_rate": 4.118627450980392e-05,
|
| 377 |
+
"loss": 0.0116,
|
| 378 |
+
"step": 1000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 20.0,
|
| 382 |
+
"eval_f1_macro": 0.2563341548388003,
|
| 383 |
+
"eval_f1_micro": 0.6470143613000756,
|
| 384 |
+
"eval_loss": 0.04850601404905319,
|
| 385 |
+
"eval_precision": 0.7205387205387206,
|
| 386 |
+
"eval_recall": 0.5871056241426612,
|
| 387 |
+
"eval_runtime": 1.6688,
|
| 388 |
+
"eval_samples_per_second": 370.936,
|
| 389 |
+
"eval_steps_per_second": 15.58,
|
| 390 |
+
"step": 1040
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 20.192307692307693,
|
| 394 |
+
"grad_norm": 2841.39111328125,
|
| 395 |
+
"learning_rate": 4.069607843137255e-05,
|
| 396 |
+
"loss": 0.0106,
|
| 397 |
+
"step": 1050
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 21.0,
|
| 401 |
+
"eval_f1_macro": 0.2594183553248198,
|
| 402 |
+
"eval_f1_micro": 0.6483600305110603,
|
| 403 |
+
"eval_loss": 0.04854930564761162,
|
| 404 |
+
"eval_precision": 0.7302405498281787,
|
| 405 |
+
"eval_recall": 0.5829903978052127,
|
| 406 |
+
"eval_runtime": 1.6818,
|
| 407 |
+
"eval_samples_per_second": 368.068,
|
| 408 |
+
"eval_steps_per_second": 15.46,
|
| 409 |
+
"step": 1092
|
| 410 |
+
}
|
| 411 |
+
],
|
| 412 |
+
"logging_steps": 50,
|
| 413 |
+
"max_steps": 5200,
|
| 414 |
+
"num_input_tokens_seen": 0,
|
| 415 |
+
"num_train_epochs": 100,
|
| 416 |
+
"save_steps": 500,
|
| 417 |
+
"stateful_callbacks": {
|
| 418 |
+
"EarlyStoppingCallback": {
|
| 419 |
+
"args": {
|
| 420 |
+
"early_stopping_patience": 5,
|
| 421 |
+
"early_stopping_threshold": 0.0
|
| 422 |
+
},
|
| 423 |
+
"attributes": {
|
| 424 |
+
"early_stopping_patience_counter": 2
|
| 425 |
+
}
|
| 426 |
+
},
|
| 427 |
+
"TrainerControl": {
|
| 428 |
+
"args": {
|
| 429 |
+
"should_epoch_stop": false,
|
| 430 |
+
"should_evaluate": false,
|
| 431 |
+
"should_log": false,
|
| 432 |
+
"should_save": true,
|
| 433 |
+
"should_training_stop": false
|
| 434 |
+
},
|
| 435 |
+
"attributes": {}
|
| 436 |
+
}
|
| 437 |
+
},
|
| 438 |
+
"total_flos": 6866097847547904.0,
|
| 439 |
+
"train_batch_size": 24,
|
| 440 |
+
"trial_name": null,
|
| 441 |
+
"trial_params": null
|
| 442 |
+
}
|
checkpoint-1092/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f05e8606862008bfc17115034db9429cc42bab3677cf65b2b782cae0ed9dfed
|
| 3 |
+
size 5368
|
checkpoint-1092/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1144/config.json
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "T1005",
|
| 13 |
+
"1": "T1021",
|
| 14 |
+
"2": "T1027",
|
| 15 |
+
"3": "T1033",
|
| 16 |
+
"4": "T1036",
|
| 17 |
+
"5": "T1041",
|
| 18 |
+
"6": "T1046",
|
| 19 |
+
"7": "T1048",
|
| 20 |
+
"8": "T1049",
|
| 21 |
+
"9": "T1053",
|
| 22 |
+
"10": "T1055",
|
| 23 |
+
"11": "T1056",
|
| 24 |
+
"12": "T1057",
|
| 25 |
+
"13": "T1059",
|
| 26 |
+
"14": "T1070",
|
| 27 |
+
"15": "T1071",
|
| 28 |
+
"16": "T1074",
|
| 29 |
+
"17": "T1078",
|
| 30 |
+
"18": "T1082",
|
| 31 |
+
"19": "T1083",
|
| 32 |
+
"20": "T1098",
|
| 33 |
+
"21": "T1102",
|
| 34 |
+
"22": "T1105",
|
| 35 |
+
"23": "T1110",
|
| 36 |
+
"24": "T1113",
|
| 37 |
+
"25": "T1114",
|
| 38 |
+
"26": "T1115",
|
| 39 |
+
"27": "T1132",
|
| 40 |
+
"28": "T1137",
|
| 41 |
+
"29": "T1140",
|
| 42 |
+
"30": "T1189",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1195",
|
| 45 |
+
"33": "T1203",
|
| 46 |
+
"34": "T1204",
|
| 47 |
+
"35": "T1218",
|
| 48 |
+
"36": "T1486",
|
| 49 |
+
"37": "T1491",
|
| 50 |
+
"38": "T1496",
|
| 51 |
+
"39": "T1497",
|
| 52 |
+
"40": "T1499",
|
| 53 |
+
"41": "T1528",
|
| 54 |
+
"42": "T1539",
|
| 55 |
+
"43": "T1547",
|
| 56 |
+
"44": "T1555",
|
| 57 |
+
"45": "T1557",
|
| 58 |
+
"46": "T1562",
|
| 59 |
+
"47": "T1564",
|
| 60 |
+
"48": "T1566",
|
| 61 |
+
"49": "T1567",
|
| 62 |
+
"50": "T1573",
|
| 63 |
+
"51": "T1574",
|
| 64 |
+
"52": "T1583",
|
| 65 |
+
"53": "T1586",
|
| 66 |
+
"54": "T1589",
|
| 67 |
+
"55": "T1606",
|
| 68 |
+
"56": "T1608",
|
| 69 |
+
"57": "T1614",
|
| 70 |
+
"58": "T1620",
|
| 71 |
+
"59": "T1623.001",
|
| 72 |
+
"60": "T1631.001"
|
| 73 |
+
},
|
| 74 |
+
"initializer_range": 0.02,
|
| 75 |
+
"intermediate_size": 3072,
|
| 76 |
+
"label2id": {
|
| 77 |
+
"T1005": 0,
|
| 78 |
+
"T1021": 1,
|
| 79 |
+
"T1027": 2,
|
| 80 |
+
"T1033": 3,
|
| 81 |
+
"T1036": 4,
|
| 82 |
+
"T1041": 5,
|
| 83 |
+
"T1046": 6,
|
| 84 |
+
"T1048": 7,
|
| 85 |
+
"T1049": 8,
|
| 86 |
+
"T1053": 9,
|
| 87 |
+
"T1055": 10,
|
| 88 |
+
"T1056": 11,
|
| 89 |
+
"T1057": 12,
|
| 90 |
+
"T1059": 13,
|
| 91 |
+
"T1070": 14,
|
| 92 |
+
"T1071": 15,
|
| 93 |
+
"T1074": 16,
|
| 94 |
+
"T1078": 17,
|
| 95 |
+
"T1082": 18,
|
| 96 |
+
"T1083": 19,
|
| 97 |
+
"T1098": 20,
|
| 98 |
+
"T1102": 21,
|
| 99 |
+
"T1105": 22,
|
| 100 |
+
"T1110": 23,
|
| 101 |
+
"T1113": 24,
|
| 102 |
+
"T1114": 25,
|
| 103 |
+
"T1115": 26,
|
| 104 |
+
"T1132": 27,
|
| 105 |
+
"T1137": 28,
|
| 106 |
+
"T1140": 29,
|
| 107 |
+
"T1189": 30,
|
| 108 |
+
"T1190": 31,
|
| 109 |
+
"T1195": 32,
|
| 110 |
+
"T1203": 33,
|
| 111 |
+
"T1204": 34,
|
| 112 |
+
"T1218": 35,
|
| 113 |
+
"T1486": 36,
|
| 114 |
+
"T1491": 37,
|
| 115 |
+
"T1496": 38,
|
| 116 |
+
"T1497": 39,
|
| 117 |
+
"T1499": 40,
|
| 118 |
+
"T1528": 41,
|
| 119 |
+
"T1539": 42,
|
| 120 |
+
"T1547": 43,
|
| 121 |
+
"T1555": 44,
|
| 122 |
+
"T1557": 45,
|
| 123 |
+
"T1562": 46,
|
| 124 |
+
"T1564": 47,
|
| 125 |
+
"T1566": 48,
|
| 126 |
+
"T1567": 49,
|
| 127 |
+
"T1573": 50,
|
| 128 |
+
"T1574": 51,
|
| 129 |
+
"T1583": 52,
|
| 130 |
+
"T1586": 53,
|
| 131 |
+
"T1589": 54,
|
| 132 |
+
"T1606": 55,
|
| 133 |
+
"T1608": 56,
|
| 134 |
+
"T1614": 57,
|
| 135 |
+
"T1620": 58,
|
| 136 |
+
"T1623.001": 59,
|
| 137 |
+
"T1631.001": 60
|
| 138 |
+
},
|
| 139 |
+
"layer_norm_eps": 1e-12,
|
| 140 |
+
"max_position_embeddings": 512,
|
| 141 |
+
"model_type": "bert",
|
| 142 |
+
"num_attention_heads": 12,
|
| 143 |
+
"num_hidden_layers": 12,
|
| 144 |
+
"pad_token_id": 0,
|
| 145 |
+
"position_embedding_type": "absolute",
|
| 146 |
+
"problem_type": "multi_label_classification",
|
| 147 |
+
"torch_dtype": "float32",
|
| 148 |
+
"transformers_version": "4.55.2",
|
| 149 |
+
"type_vocab_size": 2,
|
| 150 |
+
"use_cache": true,
|
| 151 |
+
"vocab_size": 30522
|
| 152 |
+
}
|
checkpoint-1144/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53292d8292e5eb8a125e883da4859b67e125c2bee626e465817107b0b47253db
|
| 3 |
+
size 438140132
|
checkpoint-1144/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b2648001073013c1199c297f7ad1e2eae82e586f6a7e90e534e4623fff849df
|
| 3 |
+
size 876401210
|
checkpoint-1144/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d08838643a7c90e6466d579ae7c80616e23efb3384b0f77a49923144d1295434
|
| 3 |
+
size 14244
|
checkpoint-1144/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-1144/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fbc4ce20d003fb79e97683490de717af355afe30a4bbaf6a848254efd86ee97
|
| 3 |
+
size 1064
|
checkpoint-1144/special_tokens_map.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": {
|
| 3 |
+
"content": "[CLS]",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"mask_token": {
|
| 10 |
+
"content": "[MASK]",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"sep_token": {
|
| 24 |
+
"content": "[SEP]",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"unk_token": {
|
| 31 |
+
"content": "[UNK]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
}
|
| 37 |
+
}
|
checkpoint-1144/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1144/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": false,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"extra_special_tokens": {},
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"pad_token": "[PAD]",
|
| 51 |
+
"sep_token": "[SEP]",
|
| 52 |
+
"strip_accents": null,
|
| 53 |
+
"tokenize_chinese_chars": true,
|
| 54 |
+
"tokenizer_class": "BertTokenizer",
|
| 55 |
+
"unk_token": "[UNK]"
|
| 56 |
+
}
|
checkpoint-1144/trainer_state.json
ADDED
|
@@ -0,0 +1,461 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 988,
|
| 3 |
+
"best_metric": 0.6585365853658537,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-annoctr_step2/checkpoint-988",
|
| 5 |
+
"epoch": 22.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1144,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.9615384615384616,
|
| 14 |
+
"grad_norm": 35970.37109375,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5612,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_f1_macro": 0.0,
|
| 22 |
+
"eval_f1_micro": 0.0,
|
| 23 |
+
"eval_loss": 0.33502161502838135,
|
| 24 |
+
"eval_precision": 0.0,
|
| 25 |
+
"eval_recall": 0.0,
|
| 26 |
+
"eval_runtime": 1.6495,
|
| 27 |
+
"eval_samples_per_second": 375.276,
|
| 28 |
+
"eval_steps_per_second": 15.763,
|
| 29 |
+
"step": 52
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.9230769230769231,
|
| 33 |
+
"grad_norm": 10958.1689453125,
|
| 34 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 35 |
+
"loss": 0.214,
|
| 36 |
+
"step": 100
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 2.0,
|
| 40 |
+
"eval_f1_macro": 0.0,
|
| 41 |
+
"eval_f1_micro": 0.0,
|
| 42 |
+
"eval_loss": 0.09798076748847961,
|
| 43 |
+
"eval_precision": 0.0,
|
| 44 |
+
"eval_recall": 0.0,
|
| 45 |
+
"eval_runtime": 1.6513,
|
| 46 |
+
"eval_samples_per_second": 374.861,
|
| 47 |
+
"eval_steps_per_second": 15.745,
|
| 48 |
+
"step": 104
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 2.8846153846153846,
|
| 52 |
+
"grad_norm": 5634.60546875,
|
| 53 |
+
"learning_rate": 4.951960784313726e-05,
|
| 54 |
+
"loss": 0.0873,
|
| 55 |
+
"step": 150
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 3.0,
|
| 59 |
+
"eval_f1_macro": 0.0,
|
| 60 |
+
"eval_f1_micro": 0.0,
|
| 61 |
+
"eval_loss": 0.07998213171958923,
|
| 62 |
+
"eval_precision": 0.0,
|
| 63 |
+
"eval_recall": 0.0,
|
| 64 |
+
"eval_runtime": 1.6587,
|
| 65 |
+
"eval_samples_per_second": 373.195,
|
| 66 |
+
"eval_steps_per_second": 15.675,
|
| 67 |
+
"step": 156
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 3.8461538461538463,
|
| 71 |
+
"grad_norm": 6965.6220703125,
|
| 72 |
+
"learning_rate": 4.9029411764705883e-05,
|
| 73 |
+
"loss": 0.0791,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.0,
|
| 78 |
+
"eval_f1_macro": 0.0,
|
| 79 |
+
"eval_f1_micro": 0.0,
|
| 80 |
+
"eval_loss": 0.07983831316232681,
|
| 81 |
+
"eval_precision": 0.0,
|
| 82 |
+
"eval_recall": 0.0,
|
| 83 |
+
"eval_runtime": 1.6476,
|
| 84 |
+
"eval_samples_per_second": 375.688,
|
| 85 |
+
"eval_steps_per_second": 15.78,
|
| 86 |
+
"step": 208
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 4.8076923076923075,
|
| 90 |
+
"grad_norm": 4831.40478515625,
|
| 91 |
+
"learning_rate": 4.8539215686274515e-05,
|
| 92 |
+
"loss": 0.0777,
|
| 93 |
+
"step": 250
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 5.0,
|
| 97 |
+
"eval_f1_macro": 0.0,
|
| 98 |
+
"eval_f1_micro": 0.0,
|
| 99 |
+
"eval_loss": 0.07502060383558273,
|
| 100 |
+
"eval_precision": 0.0,
|
| 101 |
+
"eval_recall": 0.0,
|
| 102 |
+
"eval_runtime": 1.6796,
|
| 103 |
+
"eval_samples_per_second": 368.535,
|
| 104 |
+
"eval_steps_per_second": 15.48,
|
| 105 |
+
"step": 260
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 5.769230769230769,
|
| 109 |
+
"grad_norm": 7363.94580078125,
|
| 110 |
+
"learning_rate": 4.804901960784314e-05,
|
| 111 |
+
"loss": 0.0716,
|
| 112 |
+
"step": 300
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 6.0,
|
| 116 |
+
"eval_f1_macro": 0.038923099933562195,
|
| 117 |
+
"eval_f1_micro": 0.47416413373860183,
|
| 118 |
+
"eval_loss": 0.0661635547876358,
|
| 119 |
+
"eval_precision": 0.9069767441860465,
|
| 120 |
+
"eval_recall": 0.32098765432098764,
|
| 121 |
+
"eval_runtime": 1.6554,
|
| 122 |
+
"eval_samples_per_second": 373.937,
|
| 123 |
+
"eval_steps_per_second": 15.707,
|
| 124 |
+
"step": 312
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 6.730769230769231,
|
| 128 |
+
"grad_norm": 6673.61376953125,
|
| 129 |
+
"learning_rate": 4.7558823529411766e-05,
|
| 130 |
+
"loss": 0.0612,
|
| 131 |
+
"step": 350
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 7.0,
|
| 135 |
+
"eval_f1_macro": 0.04364161045234521,
|
| 136 |
+
"eval_f1_micro": 0.502835538752363,
|
| 137 |
+
"eval_loss": 0.061204444617033005,
|
| 138 |
+
"eval_precision": 0.8085106382978723,
|
| 139 |
+
"eval_recall": 0.36488340192043894,
|
| 140 |
+
"eval_runtime": 1.6748,
|
| 141 |
+
"eval_samples_per_second": 369.593,
|
| 142 |
+
"eval_steps_per_second": 15.524,
|
| 143 |
+
"step": 364
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 7.6923076923076925,
|
| 147 |
+
"grad_norm": 6426.9541015625,
|
| 148 |
+
"learning_rate": 4.70686274509804e-05,
|
| 149 |
+
"loss": 0.0526,
|
| 150 |
+
"step": 400
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 8.0,
|
| 154 |
+
"eval_f1_macro": 0.07288718124374782,
|
| 155 |
+
"eval_f1_micro": 0.5546522131887985,
|
| 156 |
+
"eval_loss": 0.05695081874728203,
|
| 157 |
+
"eval_precision": 0.8121693121693122,
|
| 158 |
+
"eval_recall": 0.42112482853223593,
|
| 159 |
+
"eval_runtime": 1.665,
|
| 160 |
+
"eval_samples_per_second": 371.782,
|
| 161 |
+
"eval_steps_per_second": 15.616,
|
| 162 |
+
"step": 416
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 8.653846153846153,
|
| 166 |
+
"grad_norm": 9758.2529296875,
|
| 167 |
+
"learning_rate": 4.6578431372549016e-05,
|
| 168 |
+
"loss": 0.0433,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 9.0,
|
| 173 |
+
"eval_f1_macro": 0.07471540750966602,
|
| 174 |
+
"eval_f1_micro": 0.545950864422202,
|
| 175 |
+
"eval_loss": 0.05365221947431564,
|
| 176 |
+
"eval_precision": 0.8108108108108109,
|
| 177 |
+
"eval_recall": 0.411522633744856,
|
| 178 |
+
"eval_runtime": 1.6749,
|
| 179 |
+
"eval_samples_per_second": 369.564,
|
| 180 |
+
"eval_steps_per_second": 15.523,
|
| 181 |
+
"step": 468
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 9.615384615384615,
|
| 185 |
+
"grad_norm": 3793.51953125,
|
| 186 |
+
"learning_rate": 4.608823529411765e-05,
|
| 187 |
+
"loss": 0.0397,
|
| 188 |
+
"step": 500
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 10.0,
|
| 192 |
+
"eval_f1_macro": 0.12734138202225653,
|
| 193 |
+
"eval_f1_micro": 0.5939086294416244,
|
| 194 |
+
"eval_loss": 0.052679501473903656,
|
| 195 |
+
"eval_precision": 0.7748344370860927,
|
| 196 |
+
"eval_recall": 0.48148148148148145,
|
| 197 |
+
"eval_runtime": 1.6561,
|
| 198 |
+
"eval_samples_per_second": 373.762,
|
| 199 |
+
"eval_steps_per_second": 15.699,
|
| 200 |
+
"step": 520
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 10.576923076923077,
|
| 204 |
+
"grad_norm": 4446.76611328125,
|
| 205 |
+
"learning_rate": 4.559803921568628e-05,
|
| 206 |
+
"loss": 0.0329,
|
| 207 |
+
"step": 550
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 11.0,
|
| 211 |
+
"eval_f1_macro": 0.1493812021209619,
|
| 212 |
+
"eval_f1_micro": 0.609735269000854,
|
| 213 |
+
"eval_loss": 0.05011816695332527,
|
| 214 |
+
"eval_precision": 0.8076923076923077,
|
| 215 |
+
"eval_recall": 0.4897119341563786,
|
| 216 |
+
"eval_runtime": 1.6806,
|
| 217 |
+
"eval_samples_per_second": 368.324,
|
| 218 |
+
"eval_steps_per_second": 15.471,
|
| 219 |
+
"step": 572
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 11.538461538461538,
|
| 223 |
+
"grad_norm": 5606.7880859375,
|
| 224 |
+
"learning_rate": 4.51078431372549e-05,
|
| 225 |
+
"loss": 0.0286,
|
| 226 |
+
"step": 600
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 12.0,
|
| 230 |
+
"eval_f1_macro": 0.19655294907558352,
|
| 231 |
+
"eval_f1_micro": 0.6096959737058341,
|
| 232 |
+
"eval_loss": 0.05061562359333038,
|
| 233 |
+
"eval_precision": 0.7602459016393442,
|
| 234 |
+
"eval_recall": 0.5089163237311386,
|
| 235 |
+
"eval_runtime": 1.6713,
|
| 236 |
+
"eval_samples_per_second": 370.373,
|
| 237 |
+
"eval_steps_per_second": 15.557,
|
| 238 |
+
"step": 624
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 12.5,
|
| 242 |
+
"grad_norm": 5415.85546875,
|
| 243 |
+
"learning_rate": 4.461764705882353e-05,
|
| 244 |
+
"loss": 0.0254,
|
| 245 |
+
"step": 650
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 13.0,
|
| 249 |
+
"eval_f1_macro": 0.20270281237441773,
|
| 250 |
+
"eval_f1_micro": 0.6230831315577078,
|
| 251 |
+
"eval_loss": 0.04895725101232529,
|
| 252 |
+
"eval_precision": 0.7568627450980392,
|
| 253 |
+
"eval_recall": 0.5294924554183813,
|
| 254 |
+
"eval_runtime": 2.2359,
|
| 255 |
+
"eval_samples_per_second": 276.851,
|
| 256 |
+
"eval_steps_per_second": 11.629,
|
| 257 |
+
"step": 676
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 13.461538461538462,
|
| 261 |
+
"grad_norm": 6214.2744140625,
|
| 262 |
+
"learning_rate": 4.412745098039216e-05,
|
| 263 |
+
"loss": 0.023,
|
| 264 |
+
"step": 700
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 14.0,
|
| 268 |
+
"eval_f1_macro": 0.21593787964288247,
|
| 269 |
+
"eval_f1_micro": 0.6309904153354633,
|
| 270 |
+
"eval_loss": 0.047714490443468094,
|
| 271 |
+
"eval_precision": 0.7552581261950286,
|
| 272 |
+
"eval_recall": 0.541838134430727,
|
| 273 |
+
"eval_runtime": 1.7054,
|
| 274 |
+
"eval_samples_per_second": 362.956,
|
| 275 |
+
"eval_steps_per_second": 15.245,
|
| 276 |
+
"step": 728
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 14.423076923076923,
|
| 280 |
+
"grad_norm": 5322.5205078125,
|
| 281 |
+
"learning_rate": 4.363725490196079e-05,
|
| 282 |
+
"loss": 0.0202,
|
| 283 |
+
"step": 750
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 15.0,
|
| 287 |
+
"eval_f1_macro": 0.2127885503653234,
|
| 288 |
+
"eval_f1_micro": 0.6297739672642245,
|
| 289 |
+
"eval_loss": 0.04875025525689125,
|
| 290 |
+
"eval_precision": 0.7292418772563177,
|
| 291 |
+
"eval_recall": 0.5541838134430727,
|
| 292 |
+
"eval_runtime": 1.6728,
|
| 293 |
+
"eval_samples_per_second": 370.029,
|
| 294 |
+
"eval_steps_per_second": 15.542,
|
| 295 |
+
"step": 780
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 15.384615384615385,
|
| 299 |
+
"grad_norm": 3209.152587890625,
|
| 300 |
+
"learning_rate": 4.3147058823529413e-05,
|
| 301 |
+
"loss": 0.0186,
|
| 302 |
+
"step": 800
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 16.0,
|
| 306 |
+
"eval_f1_macro": 0.2234013473846607,
|
| 307 |
+
"eval_f1_micro": 0.6476484194294526,
|
| 308 |
+
"eval_loss": 0.04788675159215927,
|
| 309 |
+
"eval_precision": 0.7394366197183099,
|
| 310 |
+
"eval_recall": 0.5761316872427984,
|
| 311 |
+
"eval_runtime": 1.6703,
|
| 312 |
+
"eval_samples_per_second": 370.591,
|
| 313 |
+
"eval_steps_per_second": 15.566,
|
| 314 |
+
"step": 832
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 16.346153846153847,
|
| 318 |
+
"grad_norm": 6189.79296875,
|
| 319 |
+
"learning_rate": 4.265686274509804e-05,
|
| 320 |
+
"loss": 0.0166,
|
| 321 |
+
"step": 850
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 17.0,
|
| 325 |
+
"eval_f1_macro": 0.2461245877407925,
|
| 326 |
+
"eval_f1_micro": 0.6392067124332571,
|
| 327 |
+
"eval_loss": 0.047994960099458694,
|
| 328 |
+
"eval_precision": 0.7199312714776632,
|
| 329 |
+
"eval_recall": 0.5747599451303155,
|
| 330 |
+
"eval_runtime": 1.6784,
|
| 331 |
+
"eval_samples_per_second": 368.8,
|
| 332 |
+
"eval_steps_per_second": 15.491,
|
| 333 |
+
"step": 884
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 17.307692307692307,
|
| 337 |
+
"grad_norm": 2584.114013671875,
|
| 338 |
+
"learning_rate": 4.216666666666667e-05,
|
| 339 |
+
"loss": 0.0148,
|
| 340 |
+
"step": 900
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 18.0,
|
| 344 |
+
"eval_f1_macro": 0.2672202679954545,
|
| 345 |
+
"eval_f1_micro": 0.6533742331288344,
|
| 346 |
+
"eval_loss": 0.04750063270330429,
|
| 347 |
+
"eval_precision": 0.7408695652173913,
|
| 348 |
+
"eval_recall": 0.5843621399176955,
|
| 349 |
+
"eval_runtime": 1.6744,
|
| 350 |
+
"eval_samples_per_second": 369.683,
|
| 351 |
+
"eval_steps_per_second": 15.528,
|
| 352 |
+
"step": 936
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 18.26923076923077,
|
| 356 |
+
"grad_norm": 2617.191650390625,
|
| 357 |
+
"learning_rate": 4.1676470588235296e-05,
|
| 358 |
+
"loss": 0.0128,
|
| 359 |
+
"step": 950
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 19.0,
|
| 363 |
+
"eval_f1_macro": 0.25478648592377506,
|
| 364 |
+
"eval_f1_micro": 0.6585365853658537,
|
| 365 |
+
"eval_loss": 0.04794125631451607,
|
| 366 |
+
"eval_precision": 0.7409948542024014,
|
| 367 |
+
"eval_recall": 0.5925925925925926,
|
| 368 |
+
"eval_runtime": 1.6751,
|
| 369 |
+
"eval_samples_per_second": 369.525,
|
| 370 |
+
"eval_steps_per_second": 15.521,
|
| 371 |
+
"step": 988
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 19.23076923076923,
|
| 375 |
+
"grad_norm": 2930.471923828125,
|
| 376 |
+
"learning_rate": 4.118627450980392e-05,
|
| 377 |
+
"loss": 0.0116,
|
| 378 |
+
"step": 1000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 20.0,
|
| 382 |
+
"eval_f1_macro": 0.2563341548388003,
|
| 383 |
+
"eval_f1_micro": 0.6470143613000756,
|
| 384 |
+
"eval_loss": 0.04850601404905319,
|
| 385 |
+
"eval_precision": 0.7205387205387206,
|
| 386 |
+
"eval_recall": 0.5871056241426612,
|
| 387 |
+
"eval_runtime": 1.6688,
|
| 388 |
+
"eval_samples_per_second": 370.936,
|
| 389 |
+
"eval_steps_per_second": 15.58,
|
| 390 |
+
"step": 1040
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 20.192307692307693,
|
| 394 |
+
"grad_norm": 2841.39111328125,
|
| 395 |
+
"learning_rate": 4.069607843137255e-05,
|
| 396 |
+
"loss": 0.0106,
|
| 397 |
+
"step": 1050
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 21.0,
|
| 401 |
+
"eval_f1_macro": 0.2594183553248198,
|
| 402 |
+
"eval_f1_micro": 0.6483600305110603,
|
| 403 |
+
"eval_loss": 0.04854930564761162,
|
| 404 |
+
"eval_precision": 0.7302405498281787,
|
| 405 |
+
"eval_recall": 0.5829903978052127,
|
| 406 |
+
"eval_runtime": 1.6818,
|
| 407 |
+
"eval_samples_per_second": 368.068,
|
| 408 |
+
"eval_steps_per_second": 15.46,
|
| 409 |
+
"step": 1092
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 21.153846153846153,
|
| 413 |
+
"grad_norm": 2682.0166015625,
|
| 414 |
+
"learning_rate": 4.020588235294118e-05,
|
| 415 |
+
"loss": 0.0096,
|
| 416 |
+
"step": 1100
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 22.0,
|
| 420 |
+
"eval_f1_macro": 0.2448587100305824,
|
| 421 |
+
"eval_f1_micro": 0.6454478164322723,
|
| 422 |
+
"eval_loss": 0.048451464623212814,
|
| 423 |
+
"eval_precision": 0.7009646302250804,
|
| 424 |
+
"eval_recall": 0.598079561042524,
|
| 425 |
+
"eval_runtime": 1.6836,
|
| 426 |
+
"eval_samples_per_second": 367.655,
|
| 427 |
+
"eval_steps_per_second": 15.443,
|
| 428 |
+
"step": 1144
|
| 429 |
+
}
|
| 430 |
+
],
|
| 431 |
+
"logging_steps": 50,
|
| 432 |
+
"max_steps": 5200,
|
| 433 |
+
"num_input_tokens_seen": 0,
|
| 434 |
+
"num_train_epochs": 100,
|
| 435 |
+
"save_steps": 500,
|
| 436 |
+
"stateful_callbacks": {
|
| 437 |
+
"EarlyStoppingCallback": {
|
| 438 |
+
"args": {
|
| 439 |
+
"early_stopping_patience": 5,
|
| 440 |
+
"early_stopping_threshold": 0.0
|
| 441 |
+
},
|
| 442 |
+
"attributes": {
|
| 443 |
+
"early_stopping_patience_counter": 3
|
| 444 |
+
}
|
| 445 |
+
},
|
| 446 |
+
"TrainerControl": {
|
| 447 |
+
"args": {
|
| 448 |
+
"should_epoch_stop": false,
|
| 449 |
+
"should_evaluate": false,
|
| 450 |
+
"should_log": false,
|
| 451 |
+
"should_save": true,
|
| 452 |
+
"should_training_stop": false
|
| 453 |
+
},
|
| 454 |
+
"attributes": {}
|
| 455 |
+
}
|
| 456 |
+
},
|
| 457 |
+
"total_flos": 7193054887907328.0,
|
| 458 |
+
"train_batch_size": 24,
|
| 459 |
+
"trial_name": null,
|
| 460 |
+
"trial_params": null
|
| 461 |
+
}
|
checkpoint-1144/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f05e8606862008bfc17115034db9429cc42bab3677cf65b2b782cae0ed9dfed
|
| 3 |
+
size 5368
|
checkpoint-1144/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1196/config.json
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "T1005",
|
| 13 |
+
"1": "T1021",
|
| 14 |
+
"2": "T1027",
|
| 15 |
+
"3": "T1033",
|
| 16 |
+
"4": "T1036",
|
| 17 |
+
"5": "T1041",
|
| 18 |
+
"6": "T1046",
|
| 19 |
+
"7": "T1048",
|
| 20 |
+
"8": "T1049",
|
| 21 |
+
"9": "T1053",
|
| 22 |
+
"10": "T1055",
|
| 23 |
+
"11": "T1056",
|
| 24 |
+
"12": "T1057",
|
| 25 |
+
"13": "T1059",
|
| 26 |
+
"14": "T1070",
|
| 27 |
+
"15": "T1071",
|
| 28 |
+
"16": "T1074",
|
| 29 |
+
"17": "T1078",
|
| 30 |
+
"18": "T1082",
|
| 31 |
+
"19": "T1083",
|
| 32 |
+
"20": "T1098",
|
| 33 |
+
"21": "T1102",
|
| 34 |
+
"22": "T1105",
|
| 35 |
+
"23": "T1110",
|
| 36 |
+
"24": "T1113",
|
| 37 |
+
"25": "T1114",
|
| 38 |
+
"26": "T1115",
|
| 39 |
+
"27": "T1132",
|
| 40 |
+
"28": "T1137",
|
| 41 |
+
"29": "T1140",
|
| 42 |
+
"30": "T1189",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1195",
|
| 45 |
+
"33": "T1203",
|
| 46 |
+
"34": "T1204",
|
| 47 |
+
"35": "T1218",
|
| 48 |
+
"36": "T1486",
|
| 49 |
+
"37": "T1491",
|
| 50 |
+
"38": "T1496",
|
| 51 |
+
"39": "T1497",
|
| 52 |
+
"40": "T1499",
|
| 53 |
+
"41": "T1528",
|
| 54 |
+
"42": "T1539",
|
| 55 |
+
"43": "T1547",
|
| 56 |
+
"44": "T1555",
|
| 57 |
+
"45": "T1557",
|
| 58 |
+
"46": "T1562",
|
| 59 |
+
"47": "T1564",
|
| 60 |
+
"48": "T1566",
|
| 61 |
+
"49": "T1567",
|
| 62 |
+
"50": "T1573",
|
| 63 |
+
"51": "T1574",
|
| 64 |
+
"52": "T1583",
|
| 65 |
+
"53": "T1586",
|
| 66 |
+
"54": "T1589",
|
| 67 |
+
"55": "T1606",
|
| 68 |
+
"56": "T1608",
|
| 69 |
+
"57": "T1614",
|
| 70 |
+
"58": "T1620",
|
| 71 |
+
"59": "T1623.001",
|
| 72 |
+
"60": "T1631.001"
|
| 73 |
+
},
|
| 74 |
+
"initializer_range": 0.02,
|
| 75 |
+
"intermediate_size": 3072,
|
| 76 |
+
"label2id": {
|
| 77 |
+
"T1005": 0,
|
| 78 |
+
"T1021": 1,
|
| 79 |
+
"T1027": 2,
|
| 80 |
+
"T1033": 3,
|
| 81 |
+
"T1036": 4,
|
| 82 |
+
"T1041": 5,
|
| 83 |
+
"T1046": 6,
|
| 84 |
+
"T1048": 7,
|
| 85 |
+
"T1049": 8,
|
| 86 |
+
"T1053": 9,
|
| 87 |
+
"T1055": 10,
|
| 88 |
+
"T1056": 11,
|
| 89 |
+
"T1057": 12,
|
| 90 |
+
"T1059": 13,
|
| 91 |
+
"T1070": 14,
|
| 92 |
+
"T1071": 15,
|
| 93 |
+
"T1074": 16,
|
| 94 |
+
"T1078": 17,
|
| 95 |
+
"T1082": 18,
|
| 96 |
+
"T1083": 19,
|
| 97 |
+
"T1098": 20,
|
| 98 |
+
"T1102": 21,
|
| 99 |
+
"T1105": 22,
|
| 100 |
+
"T1110": 23,
|
| 101 |
+
"T1113": 24,
|
| 102 |
+
"T1114": 25,
|
| 103 |
+
"T1115": 26,
|
| 104 |
+
"T1132": 27,
|
| 105 |
+
"T1137": 28,
|
| 106 |
+
"T1140": 29,
|
| 107 |
+
"T1189": 30,
|
| 108 |
+
"T1190": 31,
|
| 109 |
+
"T1195": 32,
|
| 110 |
+
"T1203": 33,
|
| 111 |
+
"T1204": 34,
|
| 112 |
+
"T1218": 35,
|
| 113 |
+
"T1486": 36,
|
| 114 |
+
"T1491": 37,
|
| 115 |
+
"T1496": 38,
|
| 116 |
+
"T1497": 39,
|
| 117 |
+
"T1499": 40,
|
| 118 |
+
"T1528": 41,
|
| 119 |
+
"T1539": 42,
|
| 120 |
+
"T1547": 43,
|
| 121 |
+
"T1555": 44,
|
| 122 |
+
"T1557": 45,
|
| 123 |
+
"T1562": 46,
|
| 124 |
+
"T1564": 47,
|
| 125 |
+
"T1566": 48,
|
| 126 |
+
"T1567": 49,
|
| 127 |
+
"T1573": 50,
|
| 128 |
+
"T1574": 51,
|
| 129 |
+
"T1583": 52,
|
| 130 |
+
"T1586": 53,
|
| 131 |
+
"T1589": 54,
|
| 132 |
+
"T1606": 55,
|
| 133 |
+
"T1608": 56,
|
| 134 |
+
"T1614": 57,
|
| 135 |
+
"T1620": 58,
|
| 136 |
+
"T1623.001": 59,
|
| 137 |
+
"T1631.001": 60
|
| 138 |
+
},
|
| 139 |
+
"layer_norm_eps": 1e-12,
|
| 140 |
+
"max_position_embeddings": 512,
|
| 141 |
+
"model_type": "bert",
|
| 142 |
+
"num_attention_heads": 12,
|
| 143 |
+
"num_hidden_layers": 12,
|
| 144 |
+
"pad_token_id": 0,
|
| 145 |
+
"position_embedding_type": "absolute",
|
| 146 |
+
"problem_type": "multi_label_classification",
|
| 147 |
+
"torch_dtype": "float32",
|
| 148 |
+
"transformers_version": "4.55.2",
|
| 149 |
+
"type_vocab_size": 2,
|
| 150 |
+
"use_cache": true,
|
| 151 |
+
"vocab_size": 30522
|
| 152 |
+
}
|
checkpoint-1196/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:775d3d1a0bd8afff119f35e3fd6c8d5b7964891b3a9abbf7765ec715353f2f14
|
| 3 |
+
size 438140132
|