Commit ·
8229cbb
1
Parent(s): a5dba1d
Training in progress, epoch 1
Browse files- .gitignore +1 -0
- config.json +118 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +9 -0
- tokenizer.json +0 -0
- tokenizer_config.json +16 -0
- training_args.bin +3 -0
.gitignore
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
checkpoint-*/
|
config.json
ADDED
|
@@ -0,0 +1,118 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Fsoft-AIC/videberta-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DebertaV2ForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_head_size": 64,
|
| 7 |
+
"attention_probs_dropout_prob": 0.1,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "O",
|
| 13 |
+
"1": "I-UNITCALIBRATOR",
|
| 14 |
+
"2": "I-TREATMENT",
|
| 15 |
+
"3": "I-TRANSPORTATION",
|
| 16 |
+
"4": "I-SURGERY",
|
| 17 |
+
"5": "I-PREVENTIVEMED",
|
| 18 |
+
"6": "I-PERSONALCARE",
|
| 19 |
+
"7": "I-ORGANIZATION",
|
| 20 |
+
"8": "I-ORGAN",
|
| 21 |
+
"9": "I-OCCUPATION",
|
| 22 |
+
"10": "I-NAME",
|
| 23 |
+
"11": "I-MEDDEVICETECHNIQUE",
|
| 24 |
+
"12": "I-LOCATION",
|
| 25 |
+
"13": "I-GENDER",
|
| 26 |
+
"14": "I-FOODDRINK",
|
| 27 |
+
"15": "I-DRUGCHEMICAL",
|
| 28 |
+
"16": "I-DISEASESYMTOM",
|
| 29 |
+
"17": "I-DIAGNOSTICS",
|
| 30 |
+
"18": "I-DATETIME",
|
| 31 |
+
"19": "I-AGE",
|
| 32 |
+
"20": "B-UNITCALIBRATOR",
|
| 33 |
+
"21": "B-TREATMENT",
|
| 34 |
+
"22": "B-TRANSPORTATION",
|
| 35 |
+
"23": "B-SURGERY",
|
| 36 |
+
"24": "B-PREVENTIVEMED",
|
| 37 |
+
"25": "B-PERSONALCARE",
|
| 38 |
+
"26": "B-ORGANIZATION",
|
| 39 |
+
"27": "B-ORGAN",
|
| 40 |
+
"28": "B-OCCUPATION",
|
| 41 |
+
"29": "B-NAME",
|
| 42 |
+
"30": "B-MEDDEVICETECHNIQUE",
|
| 43 |
+
"31": "B-LOCATION",
|
| 44 |
+
"32": "B-GENDER",
|
| 45 |
+
"33": "B-FOODDRINK",
|
| 46 |
+
"34": "B-DRUGCHEMICAL",
|
| 47 |
+
"35": "B-DISEASESYMTOM",
|
| 48 |
+
"36": "B-DIAGNOSTICS",
|
| 49 |
+
"37": "B-DATETIME",
|
| 50 |
+
"38": "B-AGE"
|
| 51 |
+
},
|
| 52 |
+
"initializer_range": 0.02,
|
| 53 |
+
"intermediate_size": 3072,
|
| 54 |
+
"label2id": {
|
| 55 |
+
"B-AGE": 38,
|
| 56 |
+
"B-DATETIME": 37,
|
| 57 |
+
"B-DIAGNOSTICS": 36,
|
| 58 |
+
"B-DISEASESYMTOM": 35,
|
| 59 |
+
"B-DRUGCHEMICAL": 34,
|
| 60 |
+
"B-FOODDRINK": 33,
|
| 61 |
+
"B-GENDER": 32,
|
| 62 |
+
"B-LOCATION": 31,
|
| 63 |
+
"B-MEDDEVICETECHNIQUE": 30,
|
| 64 |
+
"B-NAME": 29,
|
| 65 |
+
"B-OCCUPATION": 28,
|
| 66 |
+
"B-ORGAN": 27,
|
| 67 |
+
"B-ORGANIZATION": 26,
|
| 68 |
+
"B-PERSONALCARE": 25,
|
| 69 |
+
"B-PREVENTIVEMED": 24,
|
| 70 |
+
"B-SURGERY": 23,
|
| 71 |
+
"B-TRANSPORTATION": 22,
|
| 72 |
+
"B-TREATMENT": 21,
|
| 73 |
+
"B-UNITCALIBRATOR": 20,
|
| 74 |
+
"I-AGE": 19,
|
| 75 |
+
"I-DATETIME": 18,
|
| 76 |
+
"I-DIAGNOSTICS": 17,
|
| 77 |
+
"I-DISEASESYMTOM": 16,
|
| 78 |
+
"I-DRUGCHEMICAL": 15,
|
| 79 |
+
"I-FOODDRINK": 14,
|
| 80 |
+
"I-GENDER": 13,
|
| 81 |
+
"I-LOCATION": 12,
|
| 82 |
+
"I-MEDDEVICETECHNIQUE": 11,
|
| 83 |
+
"I-NAME": 10,
|
| 84 |
+
"I-OCCUPATION": 9,
|
| 85 |
+
"I-ORGAN": 8,
|
| 86 |
+
"I-ORGANIZATION": 7,
|
| 87 |
+
"I-PERSONALCARE": 6,
|
| 88 |
+
"I-PREVENTIVEMED": 5,
|
| 89 |
+
"I-SURGERY": 4,
|
| 90 |
+
"I-TRANSPORTATION": 3,
|
| 91 |
+
"I-TREATMENT": 2,
|
| 92 |
+
"I-UNITCALIBRATOR": 1,
|
| 93 |
+
"O": 0
|
| 94 |
+
},
|
| 95 |
+
"layer_norm_eps": 1e-07,
|
| 96 |
+
"max_position_embeddings": 512,
|
| 97 |
+
"max_relative_positions": -1,
|
| 98 |
+
"model_type": "deberta-v2",
|
| 99 |
+
"norm_rel_ebd": "layer_norm",
|
| 100 |
+
"num_attention_heads": 12,
|
| 101 |
+
"num_hidden_layers": 12,
|
| 102 |
+
"pad_token_id": 0,
|
| 103 |
+
"pooler_dropout": 0,
|
| 104 |
+
"pooler_hidden_act": "gelu",
|
| 105 |
+
"pooler_hidden_size": 768,
|
| 106 |
+
"pos_att_type": [
|
| 107 |
+
"p2c",
|
| 108 |
+
"c2p"
|
| 109 |
+
],
|
| 110 |
+
"position_biased_input": false,
|
| 111 |
+
"position_buckets": 256,
|
| 112 |
+
"relative_attention": true,
|
| 113 |
+
"share_att_key": true,
|
| 114 |
+
"torch_dtype": "float32",
|
| 115 |
+
"transformers_version": "4.32.0.dev0",
|
| 116 |
+
"type_vocab_size": 0,
|
| 117 |
+
"vocab_size": 128000
|
| 118 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11a1938f8948ec1abcea96a785ed966231e645b651044909773cb343b8a66474
|
| 3 |
+
size 735208362
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "[CLS]",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"eos_token": "[SEP]",
|
| 5 |
+
"mask_token": "[MASK]",
|
| 6 |
+
"pad_token": "[PAD]",
|
| 7 |
+
"sep_token": "[SEP]",
|
| 8 |
+
"unk_token": "[UNK]"
|
| 9 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "[CLS]",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"eos_token": "[SEP]",
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"sp_model_kwargs": {},
|
| 12 |
+
"split_by_punct": false,
|
| 13 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 14 |
+
"unk_token": "[UNK]",
|
| 15 |
+
"vocab_type": "spm"
|
| 16 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d7561fa94dd84913549ba9e239e3cfdd465578de8f442d7451c89577ef42474
|
| 3 |
+
size 4408
|