Upload saved model files
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- checkpoint-1048/config.json +130 -0
- checkpoint-1048/model.safetensors +3 -0
- checkpoint-1048/optimizer.pt +3 -0
- checkpoint-1048/rng_state.pth +3 -0
- checkpoint-1048/scaler.pt +3 -0
- checkpoint-1048/scheduler.pt +3 -0
- checkpoint-1048/special_tokens_map.json +37 -0
- checkpoint-1048/tokenizer.json +0 -0
- checkpoint-1048/tokenizer_config.json +56 -0
- checkpoint-1048/trainer_state.json +279 -0
- checkpoint-1048/training_args.bin +3 -0
- checkpoint-1048/vocab.txt +0 -0
- checkpoint-1179/config.json +130 -0
- checkpoint-1179/model.safetensors +3 -0
- checkpoint-1179/optimizer.pt +3 -0
- checkpoint-1179/rng_state.pth +3 -0
- checkpoint-1179/scaler.pt +3 -0
- checkpoint-1179/scheduler.pt +3 -0
- checkpoint-1179/special_tokens_map.json +37 -0
- checkpoint-1179/tokenizer.json +0 -0
- checkpoint-1179/tokenizer_config.json +56 -0
- checkpoint-1179/trainer_state.json +312 -0
- checkpoint-1179/training_args.bin +3 -0
- checkpoint-1179/vocab.txt +0 -0
- checkpoint-131/config.json +130 -0
- checkpoint-131/model.safetensors +3 -0
- checkpoint-131/optimizer.pt +3 -0
- checkpoint-131/rng_state.pth +3 -0
- checkpoint-131/scaler.pt +3 -0
- checkpoint-131/scheduler.pt +3 -0
- checkpoint-131/special_tokens_map.json +37 -0
- checkpoint-131/tokenizer.json +0 -0
- checkpoint-131/tokenizer_config.json +56 -0
- checkpoint-131/trainer_state.json +69 -0
- checkpoint-131/training_args.bin +3 -0
- checkpoint-131/vocab.txt +0 -0
- checkpoint-1310/config.json +130 -0
- checkpoint-1310/model.safetensors +3 -0
- checkpoint-1310/optimizer.pt +3 -0
- checkpoint-1310/rng_state.pth +3 -0
- checkpoint-1310/scaler.pt +3 -0
- checkpoint-1310/scheduler.pt +3 -0
- checkpoint-1310/special_tokens_map.json +37 -0
- checkpoint-1310/tokenizer.json +0 -0
- checkpoint-1310/tokenizer_config.json +56 -0
- checkpoint-1310/trainer_state.json +345 -0
- checkpoint-1310/training_args.bin +3 -0
- checkpoint-1310/vocab.txt +0 -0
- checkpoint-262/config.json +130 -0
- checkpoint-262/model.safetensors +3 -0
checkpoint-1048/config.json
ADDED
|
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "T1003.001",
|
| 13 |
+
"1": "T1005",
|
| 14 |
+
"2": "T1012",
|
| 15 |
+
"3": "T1016",
|
| 16 |
+
"4": "T1021.001",
|
| 17 |
+
"5": "T1027",
|
| 18 |
+
"6": "T1033",
|
| 19 |
+
"7": "T1036.005",
|
| 20 |
+
"8": "T1041",
|
| 21 |
+
"9": "T1047",
|
| 22 |
+
"10": "T1053.005",
|
| 23 |
+
"11": "T1055",
|
| 24 |
+
"12": "T1056.001",
|
| 25 |
+
"13": "T1057",
|
| 26 |
+
"14": "T1059.003",
|
| 27 |
+
"15": "T1068",
|
| 28 |
+
"16": "T1070.004",
|
| 29 |
+
"17": "T1071.001",
|
| 30 |
+
"18": "T1072",
|
| 31 |
+
"19": "T1074.001",
|
| 32 |
+
"20": "T1078",
|
| 33 |
+
"21": "T1082",
|
| 34 |
+
"22": "T1083",
|
| 35 |
+
"23": "T1090",
|
| 36 |
+
"24": "T1095",
|
| 37 |
+
"25": "T1105",
|
| 38 |
+
"26": "T1106",
|
| 39 |
+
"27": "T1110",
|
| 40 |
+
"28": "T1112",
|
| 41 |
+
"29": "T1113",
|
| 42 |
+
"30": "T1140",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1204.002",
|
| 45 |
+
"33": "T1210",
|
| 46 |
+
"34": "T1218.011",
|
| 47 |
+
"35": "T1219",
|
| 48 |
+
"36": "T1484.001",
|
| 49 |
+
"37": "T1518.001",
|
| 50 |
+
"38": "T1543.003",
|
| 51 |
+
"39": "T1547.001",
|
| 52 |
+
"40": "T1548.002",
|
| 53 |
+
"41": "T1552.001",
|
| 54 |
+
"42": "T1557.001",
|
| 55 |
+
"43": "T1562.001",
|
| 56 |
+
"44": "T1564.001",
|
| 57 |
+
"45": "T1566.001",
|
| 58 |
+
"46": "T1569.002",
|
| 59 |
+
"47": "T1570",
|
| 60 |
+
"48": "T1573.001",
|
| 61 |
+
"49": "T1574.002"
|
| 62 |
+
},
|
| 63 |
+
"initializer_range": 0.02,
|
| 64 |
+
"intermediate_size": 3072,
|
| 65 |
+
"label2id": {
|
| 66 |
+
"T1003.001": 0,
|
| 67 |
+
"T1005": 1,
|
| 68 |
+
"T1012": 2,
|
| 69 |
+
"T1016": 3,
|
| 70 |
+
"T1021.001": 4,
|
| 71 |
+
"T1027": 5,
|
| 72 |
+
"T1033": 6,
|
| 73 |
+
"T1036.005": 7,
|
| 74 |
+
"T1041": 8,
|
| 75 |
+
"T1047": 9,
|
| 76 |
+
"T1053.005": 10,
|
| 77 |
+
"T1055": 11,
|
| 78 |
+
"T1056.001": 12,
|
| 79 |
+
"T1057": 13,
|
| 80 |
+
"T1059.003": 14,
|
| 81 |
+
"T1068": 15,
|
| 82 |
+
"T1070.004": 16,
|
| 83 |
+
"T1071.001": 17,
|
| 84 |
+
"T1072": 18,
|
| 85 |
+
"T1074.001": 19,
|
| 86 |
+
"T1078": 20,
|
| 87 |
+
"T1082": 21,
|
| 88 |
+
"T1083": 22,
|
| 89 |
+
"T1090": 23,
|
| 90 |
+
"T1095": 24,
|
| 91 |
+
"T1105": 25,
|
| 92 |
+
"T1106": 26,
|
| 93 |
+
"T1110": 27,
|
| 94 |
+
"T1112": 28,
|
| 95 |
+
"T1113": 29,
|
| 96 |
+
"T1140": 30,
|
| 97 |
+
"T1190": 31,
|
| 98 |
+
"T1204.002": 32,
|
| 99 |
+
"T1210": 33,
|
| 100 |
+
"T1218.011": 34,
|
| 101 |
+
"T1219": 35,
|
| 102 |
+
"T1484.001": 36,
|
| 103 |
+
"T1518.001": 37,
|
| 104 |
+
"T1543.003": 38,
|
| 105 |
+
"T1547.001": 39,
|
| 106 |
+
"T1548.002": 40,
|
| 107 |
+
"T1552.001": 41,
|
| 108 |
+
"T1557.001": 42,
|
| 109 |
+
"T1562.001": 43,
|
| 110 |
+
"T1564.001": 44,
|
| 111 |
+
"T1566.001": 45,
|
| 112 |
+
"T1569.002": 46,
|
| 113 |
+
"T1570": 47,
|
| 114 |
+
"T1573.001": 48,
|
| 115 |
+
"T1574.002": 49
|
| 116 |
+
},
|
| 117 |
+
"layer_norm_eps": 1e-12,
|
| 118 |
+
"max_position_embeddings": 512,
|
| 119 |
+
"model_type": "bert",
|
| 120 |
+
"num_attention_heads": 12,
|
| 121 |
+
"num_hidden_layers": 12,
|
| 122 |
+
"pad_token_id": 0,
|
| 123 |
+
"position_embedding_type": "absolute",
|
| 124 |
+
"problem_type": "multi_label_classification",
|
| 125 |
+
"torch_dtype": "float32",
|
| 126 |
+
"transformers_version": "4.55.2",
|
| 127 |
+
"type_vocab_size": 2,
|
| 128 |
+
"use_cache": true,
|
| 129 |
+
"vocab_size": 30522
|
| 130 |
+
}
|
checkpoint-1048/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0cfcdc795c449fc1a254b91276ece8de434985c4a3768c3ffcb178872849d18b
|
| 3 |
+
size 438106296
|
checkpoint-1048/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c4a85a5d906639547e756dfe39e4010d891e8cf80b5eb6fc66acfab1ad5e52fc
|
| 3 |
+
size 876333626
|
checkpoint-1048/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72cdc44e5b276de44fc6c8eb2fb7847347af06e107a341615d375b97cdca96e8
|
| 3 |
+
size 14244
|
checkpoint-1048/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-1048/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8b38807a6792ac8ee07d51046be1a37c6cac67c83eb7c35b7404b57c7615dd81
|
| 3 |
+
size 1064
|
checkpoint-1048/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-1048/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1048/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-1048/trainer_state.json
ADDED
|
@@ -0,0 +1,279 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 917,
|
| 3 |
+
"best_metric": 0.4342857142857143,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-tram2_base_data/checkpoint-917",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1048,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3816793893129771,
|
| 14 |
+
"grad_norm": 36039.78125,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5416,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7633587786259542,
|
| 21 |
+
"grad_norm": 9505.58984375,
|
| 22 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 23 |
+
"loss": 0.2038,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_f1_macro": 0.0,
|
| 29 |
+
"eval_f1_micro": 0.0,
|
| 30 |
+
"eval_loss": 0.04570664092898369,
|
| 31 |
+
"eval_precision": 0.0,
|
| 32 |
+
"eval_recall": 0.0,
|
| 33 |
+
"eval_runtime": 10.7376,
|
| 34 |
+
"eval_samples_per_second": 355.759,
|
| 35 |
+
"eval_steps_per_second": 14.901,
|
| 36 |
+
"step": 131
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 1.1450381679389312,
|
| 40 |
+
"grad_norm": 9370.513671875,
|
| 41 |
+
"learning_rate": 4.981153846153847e-05,
|
| 42 |
+
"loss": 0.1094,
|
| 43 |
+
"step": 150
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 1.5267175572519083,
|
| 47 |
+
"grad_norm": 7157.39892578125,
|
| 48 |
+
"learning_rate": 4.961923076923077e-05,
|
| 49 |
+
"loss": 0.1075,
|
| 50 |
+
"step": 200
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.9083969465648853,
|
| 54 |
+
"grad_norm": 9901.6123046875,
|
| 55 |
+
"learning_rate": 4.942692307692308e-05,
|
| 56 |
+
"loss": 0.1026,
|
| 57 |
+
"step": 250
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 2.0,
|
| 61 |
+
"eval_f1_macro": 0.0,
|
| 62 |
+
"eval_f1_micro": 0.0,
|
| 63 |
+
"eval_loss": 0.04716825485229492,
|
| 64 |
+
"eval_precision": 0.0,
|
| 65 |
+
"eval_recall": 0.0,
|
| 66 |
+
"eval_runtime": 10.5515,
|
| 67 |
+
"eval_samples_per_second": 362.033,
|
| 68 |
+
"eval_steps_per_second": 15.164,
|
| 69 |
+
"step": 262
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 2.2900763358778624,
|
| 73 |
+
"grad_norm": 7338.14794921875,
|
| 74 |
+
"learning_rate": 4.923461538461539e-05,
|
| 75 |
+
"loss": 0.1003,
|
| 76 |
+
"step": 300
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.67175572519084,
|
| 80 |
+
"grad_norm": 10370.3720703125,
|
| 81 |
+
"learning_rate": 4.904230769230769e-05,
|
| 82 |
+
"loss": 0.0907,
|
| 83 |
+
"step": 350
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 3.0,
|
| 87 |
+
"eval_f1_macro": 0.04150333229523513,
|
| 88 |
+
"eval_f1_micro": 0.26657552973342447,
|
| 89 |
+
"eval_loss": 0.04108869284391403,
|
| 90 |
+
"eval_precision": 0.4431818181818182,
|
| 91 |
+
"eval_recall": 0.1906158357771261,
|
| 92 |
+
"eval_runtime": 10.6444,
|
| 93 |
+
"eval_samples_per_second": 358.876,
|
| 94 |
+
"eval_steps_per_second": 15.031,
|
| 95 |
+
"step": 393
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.053435114503817,
|
| 99 |
+
"grad_norm": 8239.685546875,
|
| 100 |
+
"learning_rate": 4.885e-05,
|
| 101 |
+
"loss": 0.0848,
|
| 102 |
+
"step": 400
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 3.435114503816794,
|
| 106 |
+
"grad_norm": 7363.1640625,
|
| 107 |
+
"learning_rate": 4.865769230769231e-05,
|
| 108 |
+
"loss": 0.074,
|
| 109 |
+
"step": 450
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 3.816793893129771,
|
| 113 |
+
"grad_norm": 10253.478515625,
|
| 114 |
+
"learning_rate": 4.8465384615384616e-05,
|
| 115 |
+
"loss": 0.0708,
|
| 116 |
+
"step": 500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 4.0,
|
| 120 |
+
"eval_f1_macro": 0.0865137040218318,
|
| 121 |
+
"eval_f1_micro": 0.31753283837806967,
|
| 122 |
+
"eval_loss": 0.03653639927506447,
|
| 123 |
+
"eval_precision": 0.38186813186813184,
|
| 124 |
+
"eval_recall": 0.27174975562072334,
|
| 125 |
+
"eval_runtime": 10.3181,
|
| 126 |
+
"eval_samples_per_second": 370.223,
|
| 127 |
+
"eval_steps_per_second": 15.507,
|
| 128 |
+
"step": 524
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 4.198473282442748,
|
| 132 |
+
"grad_norm": 7632.10791015625,
|
| 133 |
+
"learning_rate": 4.827307692307693e-05,
|
| 134 |
+
"loss": 0.0632,
|
| 135 |
+
"step": 550
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 4.580152671755725,
|
| 139 |
+
"grad_norm": 9150.0556640625,
|
| 140 |
+
"learning_rate": 4.808076923076924e-05,
|
| 141 |
+
"loss": 0.0587,
|
| 142 |
+
"step": 600
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 4.961832061068702,
|
| 146 |
+
"grad_norm": 13003.1611328125,
|
| 147 |
+
"learning_rate": 4.788846153846154e-05,
|
| 148 |
+
"loss": 0.0536,
|
| 149 |
+
"step": 650
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 5.0,
|
| 153 |
+
"eval_f1_macro": 0.17183822511375405,
|
| 154 |
+
"eval_f1_micro": 0.33932302549101545,
|
| 155 |
+
"eval_loss": 0.03477426990866661,
|
| 156 |
+
"eval_precision": 0.29635036496350364,
|
| 157 |
+
"eval_recall": 0.396871945259042,
|
| 158 |
+
"eval_runtime": 10.2992,
|
| 159 |
+
"eval_samples_per_second": 370.901,
|
| 160 |
+
"eval_steps_per_second": 15.535,
|
| 161 |
+
"step": 655
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 5.34351145038168,
|
| 165 |
+
"grad_norm": 9582.8896484375,
|
| 166 |
+
"learning_rate": 4.7696153846153846e-05,
|
| 167 |
+
"loss": 0.0466,
|
| 168 |
+
"step": 700
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 5.7251908396946565,
|
| 172 |
+
"grad_norm": 10410.0546875,
|
| 173 |
+
"learning_rate": 4.750384615384616e-05,
|
| 174 |
+
"loss": 0.0427,
|
| 175 |
+
"step": 750
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 6.0,
|
| 179 |
+
"eval_f1_macro": 0.2823004051844873,
|
| 180 |
+
"eval_f1_micro": 0.4227574750830565,
|
| 181 |
+
"eval_loss": 0.03131383657455444,
|
| 182 |
+
"eval_precision": 0.3675090252707581,
|
| 183 |
+
"eval_recall": 0.4975562072336266,
|
| 184 |
+
"eval_runtime": 10.3222,
|
| 185 |
+
"eval_samples_per_second": 370.075,
|
| 186 |
+
"eval_steps_per_second": 15.501,
|
| 187 |
+
"step": 786
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 6.106870229007634,
|
| 191 |
+
"grad_norm": 8987.109375,
|
| 192 |
+
"learning_rate": 4.731153846153846e-05,
|
| 193 |
+
"loss": 0.0402,
|
| 194 |
+
"step": 800
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 6.488549618320611,
|
| 198 |
+
"grad_norm": 4473.130859375,
|
| 199 |
+
"learning_rate": 4.711923076923077e-05,
|
| 200 |
+
"loss": 0.0351,
|
| 201 |
+
"step": 850
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 6.870229007633588,
|
| 205 |
+
"grad_norm": 10426.6162109375,
|
| 206 |
+
"learning_rate": 4.692692307692308e-05,
|
| 207 |
+
"loss": 0.0337,
|
| 208 |
+
"step": 900
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 7.0,
|
| 212 |
+
"eval_f1_macro": 0.3363462665390104,
|
| 213 |
+
"eval_f1_micro": 0.4342857142857143,
|
| 214 |
+
"eval_loss": 0.03023417480289936,
|
| 215 |
+
"eval_precision": 0.35580524344569286,
|
| 216 |
+
"eval_recall": 0.5571847507331378,
|
| 217 |
+
"eval_runtime": 10.2654,
|
| 218 |
+
"eval_samples_per_second": 372.125,
|
| 219 |
+
"eval_steps_per_second": 15.586,
|
| 220 |
+
"step": 917
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 7.251908396946565,
|
| 224 |
+
"grad_norm": 8006.98779296875,
|
| 225 |
+
"learning_rate": 4.673461538461539e-05,
|
| 226 |
+
"loss": 0.0303,
|
| 227 |
+
"step": 950
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 7.633587786259542,
|
| 231 |
+
"grad_norm": 8583.732421875,
|
| 232 |
+
"learning_rate": 4.65423076923077e-05,
|
| 233 |
+
"loss": 0.0282,
|
| 234 |
+
"step": 1000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 8.0,
|
| 238 |
+
"eval_f1_macro": 0.3390787375804454,
|
| 239 |
+
"eval_f1_micro": 0.41675248884311705,
|
| 240 |
+
"eval_loss": 0.031076261773705482,
|
| 241 |
+
"eval_precision": 0.32116402116402115,
|
| 242 |
+
"eval_recall": 0.5933528836754643,
|
| 243 |
+
"eval_runtime": 10.2543,
|
| 244 |
+
"eval_samples_per_second": 372.526,
|
| 245 |
+
"eval_steps_per_second": 15.603,
|
| 246 |
+
"step": 1048
|
| 247 |
+
}
|
| 248 |
+
],
|
| 249 |
+
"logging_steps": 50,
|
| 250 |
+
"max_steps": 13100,
|
| 251 |
+
"num_input_tokens_seen": 0,
|
| 252 |
+
"num_train_epochs": 100,
|
| 253 |
+
"save_steps": 500,
|
| 254 |
+
"stateful_callbacks": {
|
| 255 |
+
"EarlyStoppingCallback": {
|
| 256 |
+
"args": {
|
| 257 |
+
"early_stopping_patience": 3,
|
| 258 |
+
"early_stopping_threshold": 0.0
|
| 259 |
+
},
|
| 260 |
+
"attributes": {
|
| 261 |
+
"early_stopping_patience_counter": 1
|
| 262 |
+
}
|
| 263 |
+
},
|
| 264 |
+
"TrainerControl": {
|
| 265 |
+
"args": {
|
| 266 |
+
"should_epoch_stop": false,
|
| 267 |
+
"should_evaluate": false,
|
| 268 |
+
"should_log": false,
|
| 269 |
+
"should_save": true,
|
| 270 |
+
"should_training_stop": false
|
| 271 |
+
},
|
| 272 |
+
"attributes": {}
|
| 273 |
+
}
|
| 274 |
+
},
|
| 275 |
+
"total_flos": 6589034407968768.0,
|
| 276 |
+
"train_batch_size": 24,
|
| 277 |
+
"trial_name": null,
|
| 278 |
+
"trial_params": null
|
| 279 |
+
}
|
checkpoint-1048/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aee2fca7ff2c999f9ae67a083c9553706cd4b3e2f09a64c6ebac2a0bd6a82c0
|
| 3 |
+
size 5368
|
checkpoint-1048/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1179/config.json
ADDED
|
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "T1003.001",
|
| 13 |
+
"1": "T1005",
|
| 14 |
+
"2": "T1012",
|
| 15 |
+
"3": "T1016",
|
| 16 |
+
"4": "T1021.001",
|
| 17 |
+
"5": "T1027",
|
| 18 |
+
"6": "T1033",
|
| 19 |
+
"7": "T1036.005",
|
| 20 |
+
"8": "T1041",
|
| 21 |
+
"9": "T1047",
|
| 22 |
+
"10": "T1053.005",
|
| 23 |
+
"11": "T1055",
|
| 24 |
+
"12": "T1056.001",
|
| 25 |
+
"13": "T1057",
|
| 26 |
+
"14": "T1059.003",
|
| 27 |
+
"15": "T1068",
|
| 28 |
+
"16": "T1070.004",
|
| 29 |
+
"17": "T1071.001",
|
| 30 |
+
"18": "T1072",
|
| 31 |
+
"19": "T1074.001",
|
| 32 |
+
"20": "T1078",
|
| 33 |
+
"21": "T1082",
|
| 34 |
+
"22": "T1083",
|
| 35 |
+
"23": "T1090",
|
| 36 |
+
"24": "T1095",
|
| 37 |
+
"25": "T1105",
|
| 38 |
+
"26": "T1106",
|
| 39 |
+
"27": "T1110",
|
| 40 |
+
"28": "T1112",
|
| 41 |
+
"29": "T1113",
|
| 42 |
+
"30": "T1140",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1204.002",
|
| 45 |
+
"33": "T1210",
|
| 46 |
+
"34": "T1218.011",
|
| 47 |
+
"35": "T1219",
|
| 48 |
+
"36": "T1484.001",
|
| 49 |
+
"37": "T1518.001",
|
| 50 |
+
"38": "T1543.003",
|
| 51 |
+
"39": "T1547.001",
|
| 52 |
+
"40": "T1548.002",
|
| 53 |
+
"41": "T1552.001",
|
| 54 |
+
"42": "T1557.001",
|
| 55 |
+
"43": "T1562.001",
|
| 56 |
+
"44": "T1564.001",
|
| 57 |
+
"45": "T1566.001",
|
| 58 |
+
"46": "T1569.002",
|
| 59 |
+
"47": "T1570",
|
| 60 |
+
"48": "T1573.001",
|
| 61 |
+
"49": "T1574.002"
|
| 62 |
+
},
|
| 63 |
+
"initializer_range": 0.02,
|
| 64 |
+
"intermediate_size": 3072,
|
| 65 |
+
"label2id": {
|
| 66 |
+
"T1003.001": 0,
|
| 67 |
+
"T1005": 1,
|
| 68 |
+
"T1012": 2,
|
| 69 |
+
"T1016": 3,
|
| 70 |
+
"T1021.001": 4,
|
| 71 |
+
"T1027": 5,
|
| 72 |
+
"T1033": 6,
|
| 73 |
+
"T1036.005": 7,
|
| 74 |
+
"T1041": 8,
|
| 75 |
+
"T1047": 9,
|
| 76 |
+
"T1053.005": 10,
|
| 77 |
+
"T1055": 11,
|
| 78 |
+
"T1056.001": 12,
|
| 79 |
+
"T1057": 13,
|
| 80 |
+
"T1059.003": 14,
|
| 81 |
+
"T1068": 15,
|
| 82 |
+
"T1070.004": 16,
|
| 83 |
+
"T1071.001": 17,
|
| 84 |
+
"T1072": 18,
|
| 85 |
+
"T1074.001": 19,
|
| 86 |
+
"T1078": 20,
|
| 87 |
+
"T1082": 21,
|
| 88 |
+
"T1083": 22,
|
| 89 |
+
"T1090": 23,
|
| 90 |
+
"T1095": 24,
|
| 91 |
+
"T1105": 25,
|
| 92 |
+
"T1106": 26,
|
| 93 |
+
"T1110": 27,
|
| 94 |
+
"T1112": 28,
|
| 95 |
+
"T1113": 29,
|
| 96 |
+
"T1140": 30,
|
| 97 |
+
"T1190": 31,
|
| 98 |
+
"T1204.002": 32,
|
| 99 |
+
"T1210": 33,
|
| 100 |
+
"T1218.011": 34,
|
| 101 |
+
"T1219": 35,
|
| 102 |
+
"T1484.001": 36,
|
| 103 |
+
"T1518.001": 37,
|
| 104 |
+
"T1543.003": 38,
|
| 105 |
+
"T1547.001": 39,
|
| 106 |
+
"T1548.002": 40,
|
| 107 |
+
"T1552.001": 41,
|
| 108 |
+
"T1557.001": 42,
|
| 109 |
+
"T1562.001": 43,
|
| 110 |
+
"T1564.001": 44,
|
| 111 |
+
"T1566.001": 45,
|
| 112 |
+
"T1569.002": 46,
|
| 113 |
+
"T1570": 47,
|
| 114 |
+
"T1573.001": 48,
|
| 115 |
+
"T1574.002": 49
|
| 116 |
+
},
|
| 117 |
+
"layer_norm_eps": 1e-12,
|
| 118 |
+
"max_position_embeddings": 512,
|
| 119 |
+
"model_type": "bert",
|
| 120 |
+
"num_attention_heads": 12,
|
| 121 |
+
"num_hidden_layers": 12,
|
| 122 |
+
"pad_token_id": 0,
|
| 123 |
+
"position_embedding_type": "absolute",
|
| 124 |
+
"problem_type": "multi_label_classification",
|
| 125 |
+
"torch_dtype": "float32",
|
| 126 |
+
"transformers_version": "4.55.2",
|
| 127 |
+
"type_vocab_size": 2,
|
| 128 |
+
"use_cache": true,
|
| 129 |
+
"vocab_size": 30522
|
| 130 |
+
}
|
checkpoint-1179/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e5822958b26f02fa3028798bf655fdb17f1c0733f7b9d23a9f8a444a69b42a8a
|
| 3 |
+
size 438106296
|
checkpoint-1179/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef34d9bd48627f8b229cfe1fe0045bd65de0405bfa14b23bd72bcb44c7ba01d6
|
| 3 |
+
size 876333626
|
checkpoint-1179/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e9f091d267fe28797711d5a79086536ecdc91656f6d0da00338fefe355f1c9e
|
| 3 |
+
size 14244
|
checkpoint-1179/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-1179/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b16f7e1dabff2e10dc9bb8c4961288eb033720484f93d6ea2266d86beed88786
|
| 3 |
+
size 1064
|
checkpoint-1179/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-1179/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1179/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-1179/trainer_state.json
ADDED
|
@@ -0,0 +1,312 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 917,
|
| 3 |
+
"best_metric": 0.4342857142857143,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-tram2_base_data/checkpoint-917",
|
| 5 |
+
"epoch": 9.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1179,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3816793893129771,
|
| 14 |
+
"grad_norm": 36039.78125,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5416,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7633587786259542,
|
| 21 |
+
"grad_norm": 9505.58984375,
|
| 22 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 23 |
+
"loss": 0.2038,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_f1_macro": 0.0,
|
| 29 |
+
"eval_f1_micro": 0.0,
|
| 30 |
+
"eval_loss": 0.04570664092898369,
|
| 31 |
+
"eval_precision": 0.0,
|
| 32 |
+
"eval_recall": 0.0,
|
| 33 |
+
"eval_runtime": 10.7376,
|
| 34 |
+
"eval_samples_per_second": 355.759,
|
| 35 |
+
"eval_steps_per_second": 14.901,
|
| 36 |
+
"step": 131
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 1.1450381679389312,
|
| 40 |
+
"grad_norm": 9370.513671875,
|
| 41 |
+
"learning_rate": 4.981153846153847e-05,
|
| 42 |
+
"loss": 0.1094,
|
| 43 |
+
"step": 150
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 1.5267175572519083,
|
| 47 |
+
"grad_norm": 7157.39892578125,
|
| 48 |
+
"learning_rate": 4.961923076923077e-05,
|
| 49 |
+
"loss": 0.1075,
|
| 50 |
+
"step": 200
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.9083969465648853,
|
| 54 |
+
"grad_norm": 9901.6123046875,
|
| 55 |
+
"learning_rate": 4.942692307692308e-05,
|
| 56 |
+
"loss": 0.1026,
|
| 57 |
+
"step": 250
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 2.0,
|
| 61 |
+
"eval_f1_macro": 0.0,
|
| 62 |
+
"eval_f1_micro": 0.0,
|
| 63 |
+
"eval_loss": 0.04716825485229492,
|
| 64 |
+
"eval_precision": 0.0,
|
| 65 |
+
"eval_recall": 0.0,
|
| 66 |
+
"eval_runtime": 10.5515,
|
| 67 |
+
"eval_samples_per_second": 362.033,
|
| 68 |
+
"eval_steps_per_second": 15.164,
|
| 69 |
+
"step": 262
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 2.2900763358778624,
|
| 73 |
+
"grad_norm": 7338.14794921875,
|
| 74 |
+
"learning_rate": 4.923461538461539e-05,
|
| 75 |
+
"loss": 0.1003,
|
| 76 |
+
"step": 300
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.67175572519084,
|
| 80 |
+
"grad_norm": 10370.3720703125,
|
| 81 |
+
"learning_rate": 4.904230769230769e-05,
|
| 82 |
+
"loss": 0.0907,
|
| 83 |
+
"step": 350
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 3.0,
|
| 87 |
+
"eval_f1_macro": 0.04150333229523513,
|
| 88 |
+
"eval_f1_micro": 0.26657552973342447,
|
| 89 |
+
"eval_loss": 0.04108869284391403,
|
| 90 |
+
"eval_precision": 0.4431818181818182,
|
| 91 |
+
"eval_recall": 0.1906158357771261,
|
| 92 |
+
"eval_runtime": 10.6444,
|
| 93 |
+
"eval_samples_per_second": 358.876,
|
| 94 |
+
"eval_steps_per_second": 15.031,
|
| 95 |
+
"step": 393
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.053435114503817,
|
| 99 |
+
"grad_norm": 8239.685546875,
|
| 100 |
+
"learning_rate": 4.885e-05,
|
| 101 |
+
"loss": 0.0848,
|
| 102 |
+
"step": 400
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 3.435114503816794,
|
| 106 |
+
"grad_norm": 7363.1640625,
|
| 107 |
+
"learning_rate": 4.865769230769231e-05,
|
| 108 |
+
"loss": 0.074,
|
| 109 |
+
"step": 450
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 3.816793893129771,
|
| 113 |
+
"grad_norm": 10253.478515625,
|
| 114 |
+
"learning_rate": 4.8465384615384616e-05,
|
| 115 |
+
"loss": 0.0708,
|
| 116 |
+
"step": 500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 4.0,
|
| 120 |
+
"eval_f1_macro": 0.0865137040218318,
|
| 121 |
+
"eval_f1_micro": 0.31753283837806967,
|
| 122 |
+
"eval_loss": 0.03653639927506447,
|
| 123 |
+
"eval_precision": 0.38186813186813184,
|
| 124 |
+
"eval_recall": 0.27174975562072334,
|
| 125 |
+
"eval_runtime": 10.3181,
|
| 126 |
+
"eval_samples_per_second": 370.223,
|
| 127 |
+
"eval_steps_per_second": 15.507,
|
| 128 |
+
"step": 524
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 4.198473282442748,
|
| 132 |
+
"grad_norm": 7632.10791015625,
|
| 133 |
+
"learning_rate": 4.827307692307693e-05,
|
| 134 |
+
"loss": 0.0632,
|
| 135 |
+
"step": 550
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 4.580152671755725,
|
| 139 |
+
"grad_norm": 9150.0556640625,
|
| 140 |
+
"learning_rate": 4.808076923076924e-05,
|
| 141 |
+
"loss": 0.0587,
|
| 142 |
+
"step": 600
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 4.961832061068702,
|
| 146 |
+
"grad_norm": 13003.1611328125,
|
| 147 |
+
"learning_rate": 4.788846153846154e-05,
|
| 148 |
+
"loss": 0.0536,
|
| 149 |
+
"step": 650
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 5.0,
|
| 153 |
+
"eval_f1_macro": 0.17183822511375405,
|
| 154 |
+
"eval_f1_micro": 0.33932302549101545,
|
| 155 |
+
"eval_loss": 0.03477426990866661,
|
| 156 |
+
"eval_precision": 0.29635036496350364,
|
| 157 |
+
"eval_recall": 0.396871945259042,
|
| 158 |
+
"eval_runtime": 10.2992,
|
| 159 |
+
"eval_samples_per_second": 370.901,
|
| 160 |
+
"eval_steps_per_second": 15.535,
|
| 161 |
+
"step": 655
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 5.34351145038168,
|
| 165 |
+
"grad_norm": 9582.8896484375,
|
| 166 |
+
"learning_rate": 4.7696153846153846e-05,
|
| 167 |
+
"loss": 0.0466,
|
| 168 |
+
"step": 700
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 5.7251908396946565,
|
| 172 |
+
"grad_norm": 10410.0546875,
|
| 173 |
+
"learning_rate": 4.750384615384616e-05,
|
| 174 |
+
"loss": 0.0427,
|
| 175 |
+
"step": 750
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 6.0,
|
| 179 |
+
"eval_f1_macro": 0.2823004051844873,
|
| 180 |
+
"eval_f1_micro": 0.4227574750830565,
|
| 181 |
+
"eval_loss": 0.03131383657455444,
|
| 182 |
+
"eval_precision": 0.3675090252707581,
|
| 183 |
+
"eval_recall": 0.4975562072336266,
|
| 184 |
+
"eval_runtime": 10.3222,
|
| 185 |
+
"eval_samples_per_second": 370.075,
|
| 186 |
+
"eval_steps_per_second": 15.501,
|
| 187 |
+
"step": 786
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 6.106870229007634,
|
| 191 |
+
"grad_norm": 8987.109375,
|
| 192 |
+
"learning_rate": 4.731153846153846e-05,
|
| 193 |
+
"loss": 0.0402,
|
| 194 |
+
"step": 800
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 6.488549618320611,
|
| 198 |
+
"grad_norm": 4473.130859375,
|
| 199 |
+
"learning_rate": 4.711923076923077e-05,
|
| 200 |
+
"loss": 0.0351,
|
| 201 |
+
"step": 850
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 6.870229007633588,
|
| 205 |
+
"grad_norm": 10426.6162109375,
|
| 206 |
+
"learning_rate": 4.692692307692308e-05,
|
| 207 |
+
"loss": 0.0337,
|
| 208 |
+
"step": 900
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 7.0,
|
| 212 |
+
"eval_f1_macro": 0.3363462665390104,
|
| 213 |
+
"eval_f1_micro": 0.4342857142857143,
|
| 214 |
+
"eval_loss": 0.03023417480289936,
|
| 215 |
+
"eval_precision": 0.35580524344569286,
|
| 216 |
+
"eval_recall": 0.5571847507331378,
|
| 217 |
+
"eval_runtime": 10.2654,
|
| 218 |
+
"eval_samples_per_second": 372.125,
|
| 219 |
+
"eval_steps_per_second": 15.586,
|
| 220 |
+
"step": 917
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 7.251908396946565,
|
| 224 |
+
"grad_norm": 8006.98779296875,
|
| 225 |
+
"learning_rate": 4.673461538461539e-05,
|
| 226 |
+
"loss": 0.0303,
|
| 227 |
+
"step": 950
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 7.633587786259542,
|
| 231 |
+
"grad_norm": 8583.732421875,
|
| 232 |
+
"learning_rate": 4.65423076923077e-05,
|
| 233 |
+
"loss": 0.0282,
|
| 234 |
+
"step": 1000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 8.0,
|
| 238 |
+
"eval_f1_macro": 0.3390787375804454,
|
| 239 |
+
"eval_f1_micro": 0.41675248884311705,
|
| 240 |
+
"eval_loss": 0.031076261773705482,
|
| 241 |
+
"eval_precision": 0.32116402116402115,
|
| 242 |
+
"eval_recall": 0.5933528836754643,
|
| 243 |
+
"eval_runtime": 10.2543,
|
| 244 |
+
"eval_samples_per_second": 372.526,
|
| 245 |
+
"eval_steps_per_second": 15.603,
|
| 246 |
+
"step": 1048
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 8.01526717557252,
|
| 250 |
+
"grad_norm": 5588.6708984375,
|
| 251 |
+
"learning_rate": 4.635e-05,
|
| 252 |
+
"loss": 0.0281,
|
| 253 |
+
"step": 1050
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 8.396946564885496,
|
| 257 |
+
"grad_norm": 5759.66650390625,
|
| 258 |
+
"learning_rate": 4.6157692307692306e-05,
|
| 259 |
+
"loss": 0.0236,
|
| 260 |
+
"step": 1100
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 8.778625954198473,
|
| 264 |
+
"grad_norm": 7545.55810546875,
|
| 265 |
+
"learning_rate": 4.596538461538462e-05,
|
| 266 |
+
"loss": 0.0234,
|
| 267 |
+
"step": 1150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 9.0,
|
| 271 |
+
"eval_f1_macro": 0.3290051589603852,
|
| 272 |
+
"eval_f1_micro": 0.42324983943481054,
|
| 273 |
+
"eval_loss": 0.03171388432383537,
|
| 274 |
+
"eval_precision": 0.3151602104256337,
|
| 275 |
+
"eval_recall": 0.6441837732160313,
|
| 276 |
+
"eval_runtime": 10.2276,
|
| 277 |
+
"eval_samples_per_second": 373.498,
|
| 278 |
+
"eval_steps_per_second": 15.644,
|
| 279 |
+
"step": 1179
|
| 280 |
+
}
|
| 281 |
+
],
|
| 282 |
+
"logging_steps": 50,
|
| 283 |
+
"max_steps": 13100,
|
| 284 |
+
"num_input_tokens_seen": 0,
|
| 285 |
+
"num_train_epochs": 100,
|
| 286 |
+
"save_steps": 500,
|
| 287 |
+
"stateful_callbacks": {
|
| 288 |
+
"EarlyStoppingCallback": {
|
| 289 |
+
"args": {
|
| 290 |
+
"early_stopping_patience": 3,
|
| 291 |
+
"early_stopping_threshold": 0.0
|
| 292 |
+
},
|
| 293 |
+
"attributes": {
|
| 294 |
+
"early_stopping_patience_counter": 2
|
| 295 |
+
}
|
| 296 |
+
},
|
| 297 |
+
"TrainerControl": {
|
| 298 |
+
"args": {
|
| 299 |
+
"should_epoch_stop": false,
|
| 300 |
+
"should_evaluate": false,
|
| 301 |
+
"should_log": false,
|
| 302 |
+
"should_save": true,
|
| 303 |
+
"should_training_stop": false
|
| 304 |
+
},
|
| 305 |
+
"attributes": {}
|
| 306 |
+
}
|
| 307 |
+
},
|
| 308 |
+
"total_flos": 7412663708964864.0,
|
| 309 |
+
"train_batch_size": 24,
|
| 310 |
+
"trial_name": null,
|
| 311 |
+
"trial_params": null
|
| 312 |
+
}
|
checkpoint-1179/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aee2fca7ff2c999f9ae67a083c9553706cd4b3e2f09a64c6ebac2a0bd6a82c0
|
| 3 |
+
size 5368
|
checkpoint-1179/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-131/config.json
ADDED
|
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "T1003.001",
|
| 13 |
+
"1": "T1005",
|
| 14 |
+
"2": "T1012",
|
| 15 |
+
"3": "T1016",
|
| 16 |
+
"4": "T1021.001",
|
| 17 |
+
"5": "T1027",
|
| 18 |
+
"6": "T1033",
|
| 19 |
+
"7": "T1036.005",
|
| 20 |
+
"8": "T1041",
|
| 21 |
+
"9": "T1047",
|
| 22 |
+
"10": "T1053.005",
|
| 23 |
+
"11": "T1055",
|
| 24 |
+
"12": "T1056.001",
|
| 25 |
+
"13": "T1057",
|
| 26 |
+
"14": "T1059.003",
|
| 27 |
+
"15": "T1068",
|
| 28 |
+
"16": "T1070.004",
|
| 29 |
+
"17": "T1071.001",
|
| 30 |
+
"18": "T1072",
|
| 31 |
+
"19": "T1074.001",
|
| 32 |
+
"20": "T1078",
|
| 33 |
+
"21": "T1082",
|
| 34 |
+
"22": "T1083",
|
| 35 |
+
"23": "T1090",
|
| 36 |
+
"24": "T1095",
|
| 37 |
+
"25": "T1105",
|
| 38 |
+
"26": "T1106",
|
| 39 |
+
"27": "T1110",
|
| 40 |
+
"28": "T1112",
|
| 41 |
+
"29": "T1113",
|
| 42 |
+
"30": "T1140",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1204.002",
|
| 45 |
+
"33": "T1210",
|
| 46 |
+
"34": "T1218.011",
|
| 47 |
+
"35": "T1219",
|
| 48 |
+
"36": "T1484.001",
|
| 49 |
+
"37": "T1518.001",
|
| 50 |
+
"38": "T1543.003",
|
| 51 |
+
"39": "T1547.001",
|
| 52 |
+
"40": "T1548.002",
|
| 53 |
+
"41": "T1552.001",
|
| 54 |
+
"42": "T1557.001",
|
| 55 |
+
"43": "T1562.001",
|
| 56 |
+
"44": "T1564.001",
|
| 57 |
+
"45": "T1566.001",
|
| 58 |
+
"46": "T1569.002",
|
| 59 |
+
"47": "T1570",
|
| 60 |
+
"48": "T1573.001",
|
| 61 |
+
"49": "T1574.002"
|
| 62 |
+
},
|
| 63 |
+
"initializer_range": 0.02,
|
| 64 |
+
"intermediate_size": 3072,
|
| 65 |
+
"label2id": {
|
| 66 |
+
"T1003.001": 0,
|
| 67 |
+
"T1005": 1,
|
| 68 |
+
"T1012": 2,
|
| 69 |
+
"T1016": 3,
|
| 70 |
+
"T1021.001": 4,
|
| 71 |
+
"T1027": 5,
|
| 72 |
+
"T1033": 6,
|
| 73 |
+
"T1036.005": 7,
|
| 74 |
+
"T1041": 8,
|
| 75 |
+
"T1047": 9,
|
| 76 |
+
"T1053.005": 10,
|
| 77 |
+
"T1055": 11,
|
| 78 |
+
"T1056.001": 12,
|
| 79 |
+
"T1057": 13,
|
| 80 |
+
"T1059.003": 14,
|
| 81 |
+
"T1068": 15,
|
| 82 |
+
"T1070.004": 16,
|
| 83 |
+
"T1071.001": 17,
|
| 84 |
+
"T1072": 18,
|
| 85 |
+
"T1074.001": 19,
|
| 86 |
+
"T1078": 20,
|
| 87 |
+
"T1082": 21,
|
| 88 |
+
"T1083": 22,
|
| 89 |
+
"T1090": 23,
|
| 90 |
+
"T1095": 24,
|
| 91 |
+
"T1105": 25,
|
| 92 |
+
"T1106": 26,
|
| 93 |
+
"T1110": 27,
|
| 94 |
+
"T1112": 28,
|
| 95 |
+
"T1113": 29,
|
| 96 |
+
"T1140": 30,
|
| 97 |
+
"T1190": 31,
|
| 98 |
+
"T1204.002": 32,
|
| 99 |
+
"T1210": 33,
|
| 100 |
+
"T1218.011": 34,
|
| 101 |
+
"T1219": 35,
|
| 102 |
+
"T1484.001": 36,
|
| 103 |
+
"T1518.001": 37,
|
| 104 |
+
"T1543.003": 38,
|
| 105 |
+
"T1547.001": 39,
|
| 106 |
+
"T1548.002": 40,
|
| 107 |
+
"T1552.001": 41,
|
| 108 |
+
"T1557.001": 42,
|
| 109 |
+
"T1562.001": 43,
|
| 110 |
+
"T1564.001": 44,
|
| 111 |
+
"T1566.001": 45,
|
| 112 |
+
"T1569.002": 46,
|
| 113 |
+
"T1570": 47,
|
| 114 |
+
"T1573.001": 48,
|
| 115 |
+
"T1574.002": 49
|
| 116 |
+
},
|
| 117 |
+
"layer_norm_eps": 1e-12,
|
| 118 |
+
"max_position_embeddings": 512,
|
| 119 |
+
"model_type": "bert",
|
| 120 |
+
"num_attention_heads": 12,
|
| 121 |
+
"num_hidden_layers": 12,
|
| 122 |
+
"pad_token_id": 0,
|
| 123 |
+
"position_embedding_type": "absolute",
|
| 124 |
+
"problem_type": "multi_label_classification",
|
| 125 |
+
"torch_dtype": "float32",
|
| 126 |
+
"transformers_version": "4.55.2",
|
| 127 |
+
"type_vocab_size": 2,
|
| 128 |
+
"use_cache": true,
|
| 129 |
+
"vocab_size": 30522
|
| 130 |
+
}
|
checkpoint-131/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:932d9c9d68222f910360bc25b11a32355f772528e78b42c8608a153514f22ca7
|
| 3 |
+
size 438106296
|
checkpoint-131/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:934f79c25a2482f702912af7b204a859760484dfd623b22ad4aee588a9318cfc
|
| 3 |
+
size 876333626
|
checkpoint-131/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b892a0b2a0c1858081f0f8d98362d2bbb7b305a53aa9ef9213f44154195f395
|
| 3 |
+
size 14244
|
checkpoint-131/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-131/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1066e0f7a6b0ffccfd8a3169b32675cd48308393d4535ec5b4f45051c38a6f81
|
| 3 |
+
size 1064
|
checkpoint-131/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-131/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-131/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-131/trainer_state.json
ADDED
|
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 131,
|
| 3 |
+
"best_metric": 0.0,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-tram2_base_data/checkpoint-131",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 131,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3816793893129771,
|
| 14 |
+
"grad_norm": 36039.78125,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5416,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7633587786259542,
|
| 21 |
+
"grad_norm": 9505.58984375,
|
| 22 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 23 |
+
"loss": 0.2038,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_f1_macro": 0.0,
|
| 29 |
+
"eval_f1_micro": 0.0,
|
| 30 |
+
"eval_loss": 0.04570664092898369,
|
| 31 |
+
"eval_precision": 0.0,
|
| 32 |
+
"eval_recall": 0.0,
|
| 33 |
+
"eval_runtime": 10.7376,
|
| 34 |
+
"eval_samples_per_second": 355.759,
|
| 35 |
+
"eval_steps_per_second": 14.901,
|
| 36 |
+
"step": 131
|
| 37 |
+
}
|
| 38 |
+
],
|
| 39 |
+
"logging_steps": 50,
|
| 40 |
+
"max_steps": 13100,
|
| 41 |
+
"num_input_tokens_seen": 0,
|
| 42 |
+
"num_train_epochs": 100,
|
| 43 |
+
"save_steps": 500,
|
| 44 |
+
"stateful_callbacks": {
|
| 45 |
+
"EarlyStoppingCallback": {
|
| 46 |
+
"args": {
|
| 47 |
+
"early_stopping_patience": 3,
|
| 48 |
+
"early_stopping_threshold": 0.0
|
| 49 |
+
},
|
| 50 |
+
"attributes": {
|
| 51 |
+
"early_stopping_patience_counter": 0
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"TrainerControl": {
|
| 55 |
+
"args": {
|
| 56 |
+
"should_epoch_stop": false,
|
| 57 |
+
"should_evaluate": false,
|
| 58 |
+
"should_log": false,
|
| 59 |
+
"should_save": true,
|
| 60 |
+
"should_training_stop": false
|
| 61 |
+
},
|
| 62 |
+
"attributes": {}
|
| 63 |
+
}
|
| 64 |
+
},
|
| 65 |
+
"total_flos": 823629300996096.0,
|
| 66 |
+
"train_batch_size": 24,
|
| 67 |
+
"trial_name": null,
|
| 68 |
+
"trial_params": null
|
| 69 |
+
}
|
checkpoint-131/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aee2fca7ff2c999f9ae67a083c9553706cd4b3e2f09a64c6ebac2a0bd6a82c0
|
| 3 |
+
size 5368
|
checkpoint-131/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1310/config.json
ADDED
|
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "T1003.001",
|
| 13 |
+
"1": "T1005",
|
| 14 |
+
"2": "T1012",
|
| 15 |
+
"3": "T1016",
|
| 16 |
+
"4": "T1021.001",
|
| 17 |
+
"5": "T1027",
|
| 18 |
+
"6": "T1033",
|
| 19 |
+
"7": "T1036.005",
|
| 20 |
+
"8": "T1041",
|
| 21 |
+
"9": "T1047",
|
| 22 |
+
"10": "T1053.005",
|
| 23 |
+
"11": "T1055",
|
| 24 |
+
"12": "T1056.001",
|
| 25 |
+
"13": "T1057",
|
| 26 |
+
"14": "T1059.003",
|
| 27 |
+
"15": "T1068",
|
| 28 |
+
"16": "T1070.004",
|
| 29 |
+
"17": "T1071.001",
|
| 30 |
+
"18": "T1072",
|
| 31 |
+
"19": "T1074.001",
|
| 32 |
+
"20": "T1078",
|
| 33 |
+
"21": "T1082",
|
| 34 |
+
"22": "T1083",
|
| 35 |
+
"23": "T1090",
|
| 36 |
+
"24": "T1095",
|
| 37 |
+
"25": "T1105",
|
| 38 |
+
"26": "T1106",
|
| 39 |
+
"27": "T1110",
|
| 40 |
+
"28": "T1112",
|
| 41 |
+
"29": "T1113",
|
| 42 |
+
"30": "T1140",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1204.002",
|
| 45 |
+
"33": "T1210",
|
| 46 |
+
"34": "T1218.011",
|
| 47 |
+
"35": "T1219",
|
| 48 |
+
"36": "T1484.001",
|
| 49 |
+
"37": "T1518.001",
|
| 50 |
+
"38": "T1543.003",
|
| 51 |
+
"39": "T1547.001",
|
| 52 |
+
"40": "T1548.002",
|
| 53 |
+
"41": "T1552.001",
|
| 54 |
+
"42": "T1557.001",
|
| 55 |
+
"43": "T1562.001",
|
| 56 |
+
"44": "T1564.001",
|
| 57 |
+
"45": "T1566.001",
|
| 58 |
+
"46": "T1569.002",
|
| 59 |
+
"47": "T1570",
|
| 60 |
+
"48": "T1573.001",
|
| 61 |
+
"49": "T1574.002"
|
| 62 |
+
},
|
| 63 |
+
"initializer_range": 0.02,
|
| 64 |
+
"intermediate_size": 3072,
|
| 65 |
+
"label2id": {
|
| 66 |
+
"T1003.001": 0,
|
| 67 |
+
"T1005": 1,
|
| 68 |
+
"T1012": 2,
|
| 69 |
+
"T1016": 3,
|
| 70 |
+
"T1021.001": 4,
|
| 71 |
+
"T1027": 5,
|
| 72 |
+
"T1033": 6,
|
| 73 |
+
"T1036.005": 7,
|
| 74 |
+
"T1041": 8,
|
| 75 |
+
"T1047": 9,
|
| 76 |
+
"T1053.005": 10,
|
| 77 |
+
"T1055": 11,
|
| 78 |
+
"T1056.001": 12,
|
| 79 |
+
"T1057": 13,
|
| 80 |
+
"T1059.003": 14,
|
| 81 |
+
"T1068": 15,
|
| 82 |
+
"T1070.004": 16,
|
| 83 |
+
"T1071.001": 17,
|
| 84 |
+
"T1072": 18,
|
| 85 |
+
"T1074.001": 19,
|
| 86 |
+
"T1078": 20,
|
| 87 |
+
"T1082": 21,
|
| 88 |
+
"T1083": 22,
|
| 89 |
+
"T1090": 23,
|
| 90 |
+
"T1095": 24,
|
| 91 |
+
"T1105": 25,
|
| 92 |
+
"T1106": 26,
|
| 93 |
+
"T1110": 27,
|
| 94 |
+
"T1112": 28,
|
| 95 |
+
"T1113": 29,
|
| 96 |
+
"T1140": 30,
|
| 97 |
+
"T1190": 31,
|
| 98 |
+
"T1204.002": 32,
|
| 99 |
+
"T1210": 33,
|
| 100 |
+
"T1218.011": 34,
|
| 101 |
+
"T1219": 35,
|
| 102 |
+
"T1484.001": 36,
|
| 103 |
+
"T1518.001": 37,
|
| 104 |
+
"T1543.003": 38,
|
| 105 |
+
"T1547.001": 39,
|
| 106 |
+
"T1548.002": 40,
|
| 107 |
+
"T1552.001": 41,
|
| 108 |
+
"T1557.001": 42,
|
| 109 |
+
"T1562.001": 43,
|
| 110 |
+
"T1564.001": 44,
|
| 111 |
+
"T1566.001": 45,
|
| 112 |
+
"T1569.002": 46,
|
| 113 |
+
"T1570": 47,
|
| 114 |
+
"T1573.001": 48,
|
| 115 |
+
"T1574.002": 49
|
| 116 |
+
},
|
| 117 |
+
"layer_norm_eps": 1e-12,
|
| 118 |
+
"max_position_embeddings": 512,
|
| 119 |
+
"model_type": "bert",
|
| 120 |
+
"num_attention_heads": 12,
|
| 121 |
+
"num_hidden_layers": 12,
|
| 122 |
+
"pad_token_id": 0,
|
| 123 |
+
"position_embedding_type": "absolute",
|
| 124 |
+
"problem_type": "multi_label_classification",
|
| 125 |
+
"torch_dtype": "float32",
|
| 126 |
+
"transformers_version": "4.55.2",
|
| 127 |
+
"type_vocab_size": 2,
|
| 128 |
+
"use_cache": true,
|
| 129 |
+
"vocab_size": 30522
|
| 130 |
+
}
|
checkpoint-1310/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b81c8dbe0e3adead24655482edd7d58307c489f4c1bde815e2113e8d2dd770a
|
| 3 |
+
size 438106296
|
checkpoint-1310/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:171fd87005860246597f433e3603180f023b81e8a0f2505e0e67c44c4199a053
|
| 3 |
+
size 876333626
|
checkpoint-1310/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c618c5811371b0a381feb8897d7e4c7aad74dfdb7876ef33d484a4665b373580
|
| 3 |
+
size 14244
|
checkpoint-1310/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
|
| 3 |
+
size 988
|
checkpoint-1310/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6295884dcc01b54e68ae3c77a924c4f2dca4c185dfcba84534f814af9eb23a30
|
| 3 |
+
size 1064
|
checkpoint-1310/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-1310/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1310/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-1310/trainer_state.json
ADDED
|
@@ -0,0 +1,345 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 917,
|
| 3 |
+
"best_metric": 0.4342857142857143,
|
| 4 |
+
"best_model_checkpoint": "./cysecbert-ttp-tram2_base_data/checkpoint-917",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1310,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3816793893129771,
|
| 14 |
+
"grad_norm": 36039.78125,
|
| 15 |
+
"learning_rate": 2.45e-05,
|
| 16 |
+
"loss": 0.5416,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7633587786259542,
|
| 21 |
+
"grad_norm": 9505.58984375,
|
| 22 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 23 |
+
"loss": 0.2038,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_f1_macro": 0.0,
|
| 29 |
+
"eval_f1_micro": 0.0,
|
| 30 |
+
"eval_loss": 0.04570664092898369,
|
| 31 |
+
"eval_precision": 0.0,
|
| 32 |
+
"eval_recall": 0.0,
|
| 33 |
+
"eval_runtime": 10.7376,
|
| 34 |
+
"eval_samples_per_second": 355.759,
|
| 35 |
+
"eval_steps_per_second": 14.901,
|
| 36 |
+
"step": 131
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 1.1450381679389312,
|
| 40 |
+
"grad_norm": 9370.513671875,
|
| 41 |
+
"learning_rate": 4.981153846153847e-05,
|
| 42 |
+
"loss": 0.1094,
|
| 43 |
+
"step": 150
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 1.5267175572519083,
|
| 47 |
+
"grad_norm": 7157.39892578125,
|
| 48 |
+
"learning_rate": 4.961923076923077e-05,
|
| 49 |
+
"loss": 0.1075,
|
| 50 |
+
"step": 200
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.9083969465648853,
|
| 54 |
+
"grad_norm": 9901.6123046875,
|
| 55 |
+
"learning_rate": 4.942692307692308e-05,
|
| 56 |
+
"loss": 0.1026,
|
| 57 |
+
"step": 250
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 2.0,
|
| 61 |
+
"eval_f1_macro": 0.0,
|
| 62 |
+
"eval_f1_micro": 0.0,
|
| 63 |
+
"eval_loss": 0.04716825485229492,
|
| 64 |
+
"eval_precision": 0.0,
|
| 65 |
+
"eval_recall": 0.0,
|
| 66 |
+
"eval_runtime": 10.5515,
|
| 67 |
+
"eval_samples_per_second": 362.033,
|
| 68 |
+
"eval_steps_per_second": 15.164,
|
| 69 |
+
"step": 262
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 2.2900763358778624,
|
| 73 |
+
"grad_norm": 7338.14794921875,
|
| 74 |
+
"learning_rate": 4.923461538461539e-05,
|
| 75 |
+
"loss": 0.1003,
|
| 76 |
+
"step": 300
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.67175572519084,
|
| 80 |
+
"grad_norm": 10370.3720703125,
|
| 81 |
+
"learning_rate": 4.904230769230769e-05,
|
| 82 |
+
"loss": 0.0907,
|
| 83 |
+
"step": 350
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 3.0,
|
| 87 |
+
"eval_f1_macro": 0.04150333229523513,
|
| 88 |
+
"eval_f1_micro": 0.26657552973342447,
|
| 89 |
+
"eval_loss": 0.04108869284391403,
|
| 90 |
+
"eval_precision": 0.4431818181818182,
|
| 91 |
+
"eval_recall": 0.1906158357771261,
|
| 92 |
+
"eval_runtime": 10.6444,
|
| 93 |
+
"eval_samples_per_second": 358.876,
|
| 94 |
+
"eval_steps_per_second": 15.031,
|
| 95 |
+
"step": 393
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.053435114503817,
|
| 99 |
+
"grad_norm": 8239.685546875,
|
| 100 |
+
"learning_rate": 4.885e-05,
|
| 101 |
+
"loss": 0.0848,
|
| 102 |
+
"step": 400
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 3.435114503816794,
|
| 106 |
+
"grad_norm": 7363.1640625,
|
| 107 |
+
"learning_rate": 4.865769230769231e-05,
|
| 108 |
+
"loss": 0.074,
|
| 109 |
+
"step": 450
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 3.816793893129771,
|
| 113 |
+
"grad_norm": 10253.478515625,
|
| 114 |
+
"learning_rate": 4.8465384615384616e-05,
|
| 115 |
+
"loss": 0.0708,
|
| 116 |
+
"step": 500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 4.0,
|
| 120 |
+
"eval_f1_macro": 0.0865137040218318,
|
| 121 |
+
"eval_f1_micro": 0.31753283837806967,
|
| 122 |
+
"eval_loss": 0.03653639927506447,
|
| 123 |
+
"eval_precision": 0.38186813186813184,
|
| 124 |
+
"eval_recall": 0.27174975562072334,
|
| 125 |
+
"eval_runtime": 10.3181,
|
| 126 |
+
"eval_samples_per_second": 370.223,
|
| 127 |
+
"eval_steps_per_second": 15.507,
|
| 128 |
+
"step": 524
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 4.198473282442748,
|
| 132 |
+
"grad_norm": 7632.10791015625,
|
| 133 |
+
"learning_rate": 4.827307692307693e-05,
|
| 134 |
+
"loss": 0.0632,
|
| 135 |
+
"step": 550
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 4.580152671755725,
|
| 139 |
+
"grad_norm": 9150.0556640625,
|
| 140 |
+
"learning_rate": 4.808076923076924e-05,
|
| 141 |
+
"loss": 0.0587,
|
| 142 |
+
"step": 600
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 4.961832061068702,
|
| 146 |
+
"grad_norm": 13003.1611328125,
|
| 147 |
+
"learning_rate": 4.788846153846154e-05,
|
| 148 |
+
"loss": 0.0536,
|
| 149 |
+
"step": 650
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 5.0,
|
| 153 |
+
"eval_f1_macro": 0.17183822511375405,
|
| 154 |
+
"eval_f1_micro": 0.33932302549101545,
|
| 155 |
+
"eval_loss": 0.03477426990866661,
|
| 156 |
+
"eval_precision": 0.29635036496350364,
|
| 157 |
+
"eval_recall": 0.396871945259042,
|
| 158 |
+
"eval_runtime": 10.2992,
|
| 159 |
+
"eval_samples_per_second": 370.901,
|
| 160 |
+
"eval_steps_per_second": 15.535,
|
| 161 |
+
"step": 655
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 5.34351145038168,
|
| 165 |
+
"grad_norm": 9582.8896484375,
|
| 166 |
+
"learning_rate": 4.7696153846153846e-05,
|
| 167 |
+
"loss": 0.0466,
|
| 168 |
+
"step": 700
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 5.7251908396946565,
|
| 172 |
+
"grad_norm": 10410.0546875,
|
| 173 |
+
"learning_rate": 4.750384615384616e-05,
|
| 174 |
+
"loss": 0.0427,
|
| 175 |
+
"step": 750
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 6.0,
|
| 179 |
+
"eval_f1_macro": 0.2823004051844873,
|
| 180 |
+
"eval_f1_micro": 0.4227574750830565,
|
| 181 |
+
"eval_loss": 0.03131383657455444,
|
| 182 |
+
"eval_precision": 0.3675090252707581,
|
| 183 |
+
"eval_recall": 0.4975562072336266,
|
| 184 |
+
"eval_runtime": 10.3222,
|
| 185 |
+
"eval_samples_per_second": 370.075,
|
| 186 |
+
"eval_steps_per_second": 15.501,
|
| 187 |
+
"step": 786
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 6.106870229007634,
|
| 191 |
+
"grad_norm": 8987.109375,
|
| 192 |
+
"learning_rate": 4.731153846153846e-05,
|
| 193 |
+
"loss": 0.0402,
|
| 194 |
+
"step": 800
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 6.488549618320611,
|
| 198 |
+
"grad_norm": 4473.130859375,
|
| 199 |
+
"learning_rate": 4.711923076923077e-05,
|
| 200 |
+
"loss": 0.0351,
|
| 201 |
+
"step": 850
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 6.870229007633588,
|
| 205 |
+
"grad_norm": 10426.6162109375,
|
| 206 |
+
"learning_rate": 4.692692307692308e-05,
|
| 207 |
+
"loss": 0.0337,
|
| 208 |
+
"step": 900
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 7.0,
|
| 212 |
+
"eval_f1_macro": 0.3363462665390104,
|
| 213 |
+
"eval_f1_micro": 0.4342857142857143,
|
| 214 |
+
"eval_loss": 0.03023417480289936,
|
| 215 |
+
"eval_precision": 0.35580524344569286,
|
| 216 |
+
"eval_recall": 0.5571847507331378,
|
| 217 |
+
"eval_runtime": 10.2654,
|
| 218 |
+
"eval_samples_per_second": 372.125,
|
| 219 |
+
"eval_steps_per_second": 15.586,
|
| 220 |
+
"step": 917
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 7.251908396946565,
|
| 224 |
+
"grad_norm": 8006.98779296875,
|
| 225 |
+
"learning_rate": 4.673461538461539e-05,
|
| 226 |
+
"loss": 0.0303,
|
| 227 |
+
"step": 950
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 7.633587786259542,
|
| 231 |
+
"grad_norm": 8583.732421875,
|
| 232 |
+
"learning_rate": 4.65423076923077e-05,
|
| 233 |
+
"loss": 0.0282,
|
| 234 |
+
"step": 1000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 8.0,
|
| 238 |
+
"eval_f1_macro": 0.3390787375804454,
|
| 239 |
+
"eval_f1_micro": 0.41675248884311705,
|
| 240 |
+
"eval_loss": 0.031076261773705482,
|
| 241 |
+
"eval_precision": 0.32116402116402115,
|
| 242 |
+
"eval_recall": 0.5933528836754643,
|
| 243 |
+
"eval_runtime": 10.2543,
|
| 244 |
+
"eval_samples_per_second": 372.526,
|
| 245 |
+
"eval_steps_per_second": 15.603,
|
| 246 |
+
"step": 1048
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 8.01526717557252,
|
| 250 |
+
"grad_norm": 5588.6708984375,
|
| 251 |
+
"learning_rate": 4.635e-05,
|
| 252 |
+
"loss": 0.0281,
|
| 253 |
+
"step": 1050
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 8.396946564885496,
|
| 257 |
+
"grad_norm": 5759.66650390625,
|
| 258 |
+
"learning_rate": 4.6157692307692306e-05,
|
| 259 |
+
"loss": 0.0236,
|
| 260 |
+
"step": 1100
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 8.778625954198473,
|
| 264 |
+
"grad_norm": 7545.55810546875,
|
| 265 |
+
"learning_rate": 4.596538461538462e-05,
|
| 266 |
+
"loss": 0.0234,
|
| 267 |
+
"step": 1150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 9.0,
|
| 271 |
+
"eval_f1_macro": 0.3290051589603852,
|
| 272 |
+
"eval_f1_micro": 0.42324983943481054,
|
| 273 |
+
"eval_loss": 0.03171388432383537,
|
| 274 |
+
"eval_precision": 0.3151602104256337,
|
| 275 |
+
"eval_recall": 0.6441837732160313,
|
| 276 |
+
"eval_runtime": 10.2276,
|
| 277 |
+
"eval_samples_per_second": 373.498,
|
| 278 |
+
"eval_steps_per_second": 15.644,
|
| 279 |
+
"step": 1179
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 9.16030534351145,
|
| 283 |
+
"grad_norm": 5176.86767578125,
|
| 284 |
+
"learning_rate": 4.577307692307692e-05,
|
| 285 |
+
"loss": 0.0217,
|
| 286 |
+
"step": 1200
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 9.541984732824428,
|
| 290 |
+
"grad_norm": 5463.861328125,
|
| 291 |
+
"learning_rate": 4.558076923076923e-05,
|
| 292 |
+
"loss": 0.0198,
|
| 293 |
+
"step": 1250
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 9.923664122137405,
|
| 297 |
+
"grad_norm": 2985.81591796875,
|
| 298 |
+
"learning_rate": 4.538846153846154e-05,
|
| 299 |
+
"loss": 0.0183,
|
| 300 |
+
"step": 1300
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 10.0,
|
| 304 |
+
"eval_f1_macro": 0.33764633344698114,
|
| 305 |
+
"eval_f1_micro": 0.4048391856004721,
|
| 306 |
+
"eval_loss": 0.033043116331100464,
|
| 307 |
+
"eval_precision": 0.28994082840236685,
|
| 308 |
+
"eval_recall": 0.6705767350928641,
|
| 309 |
+
"eval_runtime": 10.6264,
|
| 310 |
+
"eval_samples_per_second": 359.483,
|
| 311 |
+
"eval_steps_per_second": 15.057,
|
| 312 |
+
"step": 1310
|
| 313 |
+
}
|
| 314 |
+
],
|
| 315 |
+
"logging_steps": 50,
|
| 316 |
+
"max_steps": 13100,
|
| 317 |
+
"num_input_tokens_seen": 0,
|
| 318 |
+
"num_train_epochs": 100,
|
| 319 |
+
"save_steps": 500,
|
| 320 |
+
"stateful_callbacks": {
|
| 321 |
+
"EarlyStoppingCallback": {
|
| 322 |
+
"args": {
|
| 323 |
+
"early_stopping_patience": 3,
|
| 324 |
+
"early_stopping_threshold": 0.0
|
| 325 |
+
},
|
| 326 |
+
"attributes": {
|
| 327 |
+
"early_stopping_patience_counter": 3
|
| 328 |
+
}
|
| 329 |
+
},
|
| 330 |
+
"TrainerControl": {
|
| 331 |
+
"args": {
|
| 332 |
+
"should_epoch_stop": false,
|
| 333 |
+
"should_evaluate": false,
|
| 334 |
+
"should_log": false,
|
| 335 |
+
"should_save": true,
|
| 336 |
+
"should_training_stop": true
|
| 337 |
+
},
|
| 338 |
+
"attributes": {}
|
| 339 |
+
}
|
| 340 |
+
},
|
| 341 |
+
"total_flos": 8236293009960960.0,
|
| 342 |
+
"train_batch_size": 24,
|
| 343 |
+
"trial_name": null,
|
| 344 |
+
"trial_params": null
|
| 345 |
+
}
|
checkpoint-1310/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aee2fca7ff2c999f9ae67a083c9553706cd4b3e2f09a64c6ebac2a0bd6a82c0
|
| 3 |
+
size 5368
|
checkpoint-1310/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-262/config.json
ADDED
|
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "T1003.001",
|
| 13 |
+
"1": "T1005",
|
| 14 |
+
"2": "T1012",
|
| 15 |
+
"3": "T1016",
|
| 16 |
+
"4": "T1021.001",
|
| 17 |
+
"5": "T1027",
|
| 18 |
+
"6": "T1033",
|
| 19 |
+
"7": "T1036.005",
|
| 20 |
+
"8": "T1041",
|
| 21 |
+
"9": "T1047",
|
| 22 |
+
"10": "T1053.005",
|
| 23 |
+
"11": "T1055",
|
| 24 |
+
"12": "T1056.001",
|
| 25 |
+
"13": "T1057",
|
| 26 |
+
"14": "T1059.003",
|
| 27 |
+
"15": "T1068",
|
| 28 |
+
"16": "T1070.004",
|
| 29 |
+
"17": "T1071.001",
|
| 30 |
+
"18": "T1072",
|
| 31 |
+
"19": "T1074.001",
|
| 32 |
+
"20": "T1078",
|
| 33 |
+
"21": "T1082",
|
| 34 |
+
"22": "T1083",
|
| 35 |
+
"23": "T1090",
|
| 36 |
+
"24": "T1095",
|
| 37 |
+
"25": "T1105",
|
| 38 |
+
"26": "T1106",
|
| 39 |
+
"27": "T1110",
|
| 40 |
+
"28": "T1112",
|
| 41 |
+
"29": "T1113",
|
| 42 |
+
"30": "T1140",
|
| 43 |
+
"31": "T1190",
|
| 44 |
+
"32": "T1204.002",
|
| 45 |
+
"33": "T1210",
|
| 46 |
+
"34": "T1218.011",
|
| 47 |
+
"35": "T1219",
|
| 48 |
+
"36": "T1484.001",
|
| 49 |
+
"37": "T1518.001",
|
| 50 |
+
"38": "T1543.003",
|
| 51 |
+
"39": "T1547.001",
|
| 52 |
+
"40": "T1548.002",
|
| 53 |
+
"41": "T1552.001",
|
| 54 |
+
"42": "T1557.001",
|
| 55 |
+
"43": "T1562.001",
|
| 56 |
+
"44": "T1564.001",
|
| 57 |
+
"45": "T1566.001",
|
| 58 |
+
"46": "T1569.002",
|
| 59 |
+
"47": "T1570",
|
| 60 |
+
"48": "T1573.001",
|
| 61 |
+
"49": "T1574.002"
|
| 62 |
+
},
|
| 63 |
+
"initializer_range": 0.02,
|
| 64 |
+
"intermediate_size": 3072,
|
| 65 |
+
"label2id": {
|
| 66 |
+
"T1003.001": 0,
|
| 67 |
+
"T1005": 1,
|
| 68 |
+
"T1012": 2,
|
| 69 |
+
"T1016": 3,
|
| 70 |
+
"T1021.001": 4,
|
| 71 |
+
"T1027": 5,
|
| 72 |
+
"T1033": 6,
|
| 73 |
+
"T1036.005": 7,
|
| 74 |
+
"T1041": 8,
|
| 75 |
+
"T1047": 9,
|
| 76 |
+
"T1053.005": 10,
|
| 77 |
+
"T1055": 11,
|
| 78 |
+
"T1056.001": 12,
|
| 79 |
+
"T1057": 13,
|
| 80 |
+
"T1059.003": 14,
|
| 81 |
+
"T1068": 15,
|
| 82 |
+
"T1070.004": 16,
|
| 83 |
+
"T1071.001": 17,
|
| 84 |
+
"T1072": 18,
|
| 85 |
+
"T1074.001": 19,
|
| 86 |
+
"T1078": 20,
|
| 87 |
+
"T1082": 21,
|
| 88 |
+
"T1083": 22,
|
| 89 |
+
"T1090": 23,
|
| 90 |
+
"T1095": 24,
|
| 91 |
+
"T1105": 25,
|
| 92 |
+
"T1106": 26,
|
| 93 |
+
"T1110": 27,
|
| 94 |
+
"T1112": 28,
|
| 95 |
+
"T1113": 29,
|
| 96 |
+
"T1140": 30,
|
| 97 |
+
"T1190": 31,
|
| 98 |
+
"T1204.002": 32,
|
| 99 |
+
"T1210": 33,
|
| 100 |
+
"T1218.011": 34,
|
| 101 |
+
"T1219": 35,
|
| 102 |
+
"T1484.001": 36,
|
| 103 |
+
"T1518.001": 37,
|
| 104 |
+
"T1543.003": 38,
|
| 105 |
+
"T1547.001": 39,
|
| 106 |
+
"T1548.002": 40,
|
| 107 |
+
"T1552.001": 41,
|
| 108 |
+
"T1557.001": 42,
|
| 109 |
+
"T1562.001": 43,
|
| 110 |
+
"T1564.001": 44,
|
| 111 |
+
"T1566.001": 45,
|
| 112 |
+
"T1569.002": 46,
|
| 113 |
+
"T1570": 47,
|
| 114 |
+
"T1573.001": 48,
|
| 115 |
+
"T1574.002": 49
|
| 116 |
+
},
|
| 117 |
+
"layer_norm_eps": 1e-12,
|
| 118 |
+
"max_position_embeddings": 512,
|
| 119 |
+
"model_type": "bert",
|
| 120 |
+
"num_attention_heads": 12,
|
| 121 |
+
"num_hidden_layers": 12,
|
| 122 |
+
"pad_token_id": 0,
|
| 123 |
+
"position_embedding_type": "absolute",
|
| 124 |
+
"problem_type": "multi_label_classification",
|
| 125 |
+
"torch_dtype": "float32",
|
| 126 |
+
"transformers_version": "4.55.2",
|
| 127 |
+
"type_vocab_size": 2,
|
| 128 |
+
"use_cache": true,
|
| 129 |
+
"vocab_size": 30522
|
| 130 |
+
}
|
checkpoint-262/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2466a3e5355848fc410d9a9d3af85a5e20a977e5c5d89e0d5e2173a6a40a47a4
|
| 3 |
+
size 438106296
|