jinmang2 commited on
Commit
5ead116
1 Parent(s): 2888a99

add ner model configs

Browse files
ner/en/roberta.base/config.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "_name_or_path": "jinmang2/tmp",
3
  "architectures": [
4
- "RobertaForCharNER"
5
  ],
6
  "attention_probs_dropout_prob": 0.1,
7
  "bos_token_id": 0,
@@ -11,80 +11,80 @@
11
  "hidden_dropout_prob": 0.1,
12
  "hidden_size": 768,
13
  "id2label": {
14
- "0": "LABEL_0",
15
- "1": "LABEL_1",
16
- "2": "LABEL_2",
17
- "3": "LABEL_3",
18
- "4": "LABEL_4",
19
- "5": "LABEL_5",
20
- "6": "LABEL_6",
21
- "7": "LABEL_7",
22
- "8": "LABEL_8",
23
- "9": "LABEL_9",
24
- "10": "LABEL_10",
25
- "11": "LABEL_11",
26
- "12": "LABEL_12",
27
- "13": "LABEL_13",
28
- "14": "LABEL_14",
29
- "15": "LABEL_15",
30
- "16": "LABEL_16",
31
- "17": "LABEL_17",
32
- "18": "LABEL_18",
33
- "19": "LABEL_19",
34
- "20": "LABEL_20",
35
- "21": "LABEL_21",
36
- "22": "LABEL_22",
37
- "23": "LABEL_23",
38
- "24": "LABEL_24",
39
- "25": "LABEL_25",
40
- "26": "LABEL_26",
41
- "27": "LABEL_27",
42
- "28": "LABEL_28",
43
- "29": "LABEL_29",
44
- "30": "LABEL_30",
45
- "31": "LABEL_31",
46
- "32": "LABEL_32",
47
- "33": "LABEL_33",
48
- "34": "LABEL_34"
49
  },
50
  "initializer_range": 0.02,
51
  "intermediate_size": 3072,
52
  "label2id": {
53
- "LABEL_0": 0,
54
- "LABEL_1": 1,
55
- "LABEL_10": 10,
56
- "LABEL_11": 11,
57
- "LABEL_12": 12,
58
- "LABEL_13": 13,
59
- "LABEL_14": 14,
60
- "LABEL_15": 15,
61
- "LABEL_16": 16,
62
- "LABEL_17": 17,
63
- "LABEL_18": 18,
64
- "LABEL_19": 19,
65
- "LABEL_2": 2,
66
- "LABEL_20": 20,
67
- "LABEL_21": 21,
68
- "LABEL_22": 22,
69
- "LABEL_23": 23,
70
- "LABEL_24": 24,
71
- "LABEL_25": 25,
72
- "LABEL_26": 26,
73
- "LABEL_27": 27,
74
- "LABEL_28": 28,
75
- "LABEL_29": 29,
76
- "LABEL_3": 3,
77
- "LABEL_30": 30,
78
- "LABEL_31": 31,
79
- "LABEL_32": 32,
80
- "LABEL_33": 33,
81
- "LABEL_34": 34,
82
- "LABEL_4": 4,
83
- "LABEL_5": 5,
84
- "LABEL_6": 6,
85
- "LABEL_7": 7,
86
- "LABEL_8": 8,
87
- "LABEL_9": 9
88
  },
89
  "layer_norm_eps": 1e-05,
90
  "max_position_embeddings": 514,
@@ -94,7 +94,7 @@
94
  "pad_token_id": 1,
95
  "position_embedding_type": "absolute",
96
  "torch_dtype": "float32",
97
- "transformers_version": "4.15.0",
98
  "type_vocab_size": 1,
99
  "use_cache": true,
100
  "vocab_size": 50265
 
1
  {
2
+ "_name_or_path": "jinmang2/dooly-hub/ner/en/roberta.base",
3
  "architectures": [
4
+ "RobertaForSequenceTagging"
5
  ],
6
  "attention_probs_dropout_prob": 0.1,
7
  "bos_token_id": 0,
 
11
  "hidden_dropout_prob": 0.1,
12
  "hidden_size": 768,
13
  "id2label": {
14
+ "0": "O",
15
+ "1": "I-ORG",
16
+ "2": "I-PERSON",
17
+ "3": "I-DATE",
18
+ "4": "I-GPE",
19
+ "5": "B-ORG",
20
+ "6": "B-GPE",
21
+ "7": "B-PERSON",
22
+ "8": "B-DATE",
23
+ "9": "I-MONEY",
24
+ "10": "B-CARDINAL",
25
+ "11": "I-PERCENT",
26
+ "12": "B-NORP",
27
+ "13": "I-CARDINAL",
28
+ "14": "B-MONEY",
29
+ "15": "B-PERCENT",
30
+ "16": "I-TIME",
31
+ "17": "I-LOC",
32
+ "18": "I-FAC",
33
+ "19": "I-QUANTITY",
34
+ "20": "I-NORP",
35
+ "21": "I-EVENT",
36
+ "22": "B-ORDINAL",
37
+ "23": "I-PRODUCT",
38
+ "24": "B-LOC",
39
+ "25": "B-TIME",
40
+ "26": "I-LAW",
41
+ "27": "B-QUANTITY",
42
+ "28": "B-FAC",
43
+ "29": "B-PRODUCT",
44
+ "30": "B-EVENT",
45
+ "31": "I-ORDINAL",
46
+ "32": "B-LAW",
47
+ "33": "B-LANGUAGE",
48
+ "34": "I-LANGUAGE"
49
  },
50
  "initializer_range": 0.02,
51
  "intermediate_size": 3072,
52
  "label2id": {
53
+ "O": 0,
54
+ "I-ORG": 1,
55
+ "I-PERSON": 2,
56
+ "I-DATE": 3,
57
+ "I-GPE": 4,
58
+ "B-ORG": 5,
59
+ "B-GPE": 6,
60
+ "B-PERSON": 7,
61
+ "B-DATE": 8,
62
+ "I-MONEY": 9,
63
+ "B-CARDINAL": 10,
64
+ "I-PERCENT": 11,
65
+ "B-NORP": 12,
66
+ "I-CARDINAL": 13,
67
+ "B-MONEY": 14,
68
+ "B-PERCENT": 15,
69
+ "I-TIME": 16,
70
+ "I-LOC": 17,
71
+ "I-FAC": 18,
72
+ "I-QUANTITY": 19,
73
+ "I-NORP": 20,
74
+ "I-EVENT": 21,
75
+ "B-ORDINAL": 22,
76
+ "I-PRODUCT": 23,
77
+ "B-LOC": 24,
78
+ "B-TIME": 25,
79
+ "I-LAW": 26,
80
+ "B-QUANTITY": 27,
81
+ "B-FAC": 28,
82
+ "B-PRODUCT": 29,
83
+ "B-EVENT": 30,
84
+ "I-ORDINAL": 31,
85
+ "B-LAW": 32,
86
+ "B-LANGUAGE": 33,
87
+ "I-LANGUAGE": 34
88
  },
89
  "layer_norm_eps": 1e-05,
90
  "max_position_embeddings": 514,
 
94
  "pad_token_id": 1,
95
  "position_embedding_type": "absolute",
96
  "torch_dtype": "float32",
97
+ "transformers_version": "4.16.2",
98
  "type_vocab_size": 1,
99
  "use_cache": true,
100
  "vocab_size": 50265
ner/ja/jaberta.base/config.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "jinmang2/dooly-hub/ner/ja/jaberta.base",
3
+ "architectures": [
4
+ "RobertaForSequenceTagging"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "O",
15
+ "1": "B-LOCATION",
16
+ "2": "I-LOCATION",
17
+ "3": "I-ARTIFACT",
18
+ "4": "I-ORGANIZATION",
19
+ "5": "I-DATE",
20
+ "6": "B-DATE",
21
+ "7": "I-PERSON",
22
+ "8": "B-ORGANIZATION",
23
+ "9": "B-ARTIFACT",
24
+ "10": "B-PERSON",
25
+ "11": "I-OPTIONAL",
26
+ "12": "B-OPTIONAL",
27
+ "13": "I-MONEY",
28
+ "14": "I-PERCENT",
29
+ "15": "B-MONEY",
30
+ "16": "I-TIME",
31
+ "17": "B-PERCENT",
32
+ "18": "B-TIME"
33
+ },
34
+ "initializer_range": 0.02,
35
+ "intermediate_size": 3072,
36
+ "label2id": {
37
+ "O": 0,
38
+ "B-LOCATION": 1,
39
+ "I-LOCATION": 2,
40
+ "I-ARTIFACT": 3,
41
+ "I-ORGANIZATION": 4,
42
+ "I-DATE": 5,
43
+ "B-DATE": 6,
44
+ "I-PERSON": 7,
45
+ "B-ORGANIZATION": 8,
46
+ "B-ARTIFACT": 9,
47
+ "B-PERSON": 10,
48
+ "I-OPTIONAL": 11,
49
+ "B-OPTIONAL": 12,
50
+ "I-MONEY": 13,
51
+ "I-PERCENT": 14,
52
+ "B-MONEY": 15,
53
+ "I-TIME": 16,
54
+ "B-PERCENT": 17,
55
+ "B-TIME": 18
56
+ },
57
+ "layer_norm_eps": 1e-05,
58
+ "max_position_embeddings": 514,
59
+ "model_type": "roberta",
60
+ "num_attention_heads": 12,
61
+ "num_hidden_layers": 12,
62
+ "pad_token_id": 1,
63
+ "position_embedding_type": "absolute",
64
+ "torch_dtype": "float32",
65
+ "transformers_version": "4.16.2",
66
+ "type_vocab_size": 1,
67
+ "use_cache": true,
68
+ "vocab_size": 32000
69
+ }
ner/ko/charbert.base/config.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "_name_or_path": "jinmang2/dooly-hub/ner/ko",
3
  "architectures": [
4
- "RobertaForCharNER"
5
  ],
6
  "attention_probs_dropout_prob": 0.1,
7
  "bos_token_id": 0,
 
1
  {
2
+ "_name_or_path": "jinmang2/dooly-hub/ner/ko/charbert.base",
3
  "architectures": [
4
+ "RobertaForSequenceTagging"
5
  ],
6
  "attention_probs_dropout_prob": 0.1,
7
  "bos_token_id": 0,
ner/zh/zhberta.base/config.json ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "jinmang2/dooly-hub/ner/zh/zhberta.base",
3
+ "architectures": [
4
+ "RobertaForSequenceTagging"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "O",
15
+ "1": "I-ORG",
16
+ "2": "I-PERSON",
17
+ "3": "I-GPE",
18
+ "4": "I-DATE",
19
+ "5": "B-GPE",
20
+ "6": "B-PERSON",
21
+ "7": "B-ORG",
22
+ "8": "B-DATE",
23
+ "9": "B-CARDINAL",
24
+ "10": "I-MONEY",
25
+ "11": "I-CARDINAL",
26
+ "12": "I-EVENT",
27
+ "13": "I-TIME",
28
+ "14": "I-FAC",
29
+ "15": "I-LOC",
30
+ "16": "I-PERCENT",
31
+ "17": "I-NORP",
32
+ "18": "I-QUANTITY",
33
+ "19": "B-LOC",
34
+ "20": "B-NORP",
35
+ "21": "B-TIME",
36
+ "22": "I-LAW",
37
+ "23": "B-FAC",
38
+ "24": "B-MONEY",
39
+ "25": "I-ORDINAL",
40
+ "26": "B-ORDINAL",
41
+ "27": "B-EVENT",
42
+ "28": "I-PRODUCT",
43
+ "29": "B-QUANTITY",
44
+ "30": "B-PERCENT",
45
+ "31": "I-LANGUAGE",
46
+ "32": "B-PRODUCT",
47
+ "33": "B-LANGUAGE",
48
+ "34": "B-LAW"
49
+ },
50
+ "initializer_range": 0.02,
51
+ "intermediate_size": 3072,
52
+ "label2id": {
53
+ "O": 0,
54
+ "I-ORG": 1,
55
+ "I-PERSON": 2,
56
+ "I-GPE": 3,
57
+ "I-DATE": 4,
58
+ "B-GPE": 5,
59
+ "B-PERSON": 6,
60
+ "B-ORG": 7,
61
+ "B-DATE": 8,
62
+ "B-CARDINAL": 9,
63
+ "I-MONEY": 10,
64
+ "I-CARDINAL": 11,
65
+ "I-EVENT": 12,
66
+ "I-TIME": 13,
67
+ "I-FAC": 14,
68
+ "I-LOC": 15,
69
+ "I-PERCENT": 16,
70
+ "I-NORP": 17,
71
+ "I-QUANTITY": 18,
72
+ "B-LOC": 19,
73
+ "B-NORP": 20,
74
+ "B-TIME": 21,
75
+ "I-LAW": 22,
76
+ "B-FAC": 23,
77
+ "B-MONEY": 24,
78
+ "I-ORDINAL": 25,
79
+ "B-ORDINAL": 26,
80
+ "B-EVENT": 27,
81
+ "I-PRODUCT": 28,
82
+ "B-QUANTITY": 29,
83
+ "B-PERCENT": 30,
84
+ "I-LANGUAGE": 31,
85
+ "B-PRODUCT": 32,
86
+ "B-LANGUAGE": 33,
87
+ "B-LAW": 34
88
+ },
89
+ "layer_norm_eps": 1e-05,
90
+ "max_position_embeddings": 514,
91
+ "model_type": "roberta",
92
+ "num_attention_heads": 12,
93
+ "num_hidden_layers": 12,
94
+ "pad_token_id": 1,
95
+ "position_embedding_type": "absolute",
96
+ "torch_dtype": "float32",
97
+ "transformers_version": "4.16.2",
98
+ "type_vocab_size": 1,
99
+ "use_cache": true,
100
+ "vocab_size": 21128
101
+ }