lfoppiano commited on
Commit
4dea189
·
verified ·
1 Parent(s): 970783a

Upload folder using huggingface_hub

Browse files
software-BERT/config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "model_name": "software-BERT",
3
  "architecture": "BERT",
4
  "embeddings_name": null,
5
  "char_vocab_size": 549,
@@ -35,4 +35,4 @@
35
  "I-<version>": 8,
36
  "O": 9
37
  }
38
- }
 
1
  {
2
+ "model_name": "grobid-software_all-BERT",
3
  "architecture": "BERT",
4
  "embeddings_name": null,
5
  "char_vocab_size": 549,
 
35
  "I-<version>": 8,
36
  "O": 9
37
  }
38
+ }
software-BERT/model_weights.hdf5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:14c716b7ab2d945de3a838e009a5ebb4303292ec454d5c078e9dfb441a987ae8
3
  size 440060072
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f143d8a874cc3b03b4c50ae55e66cb06bb54ccc65221bbede645dc962edd106b
3
  size 440060072
software-BERT/transformer-config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/srv/storage/datalake@storage2.nancy.grid5000.fr/lfoppiano/embeddings/scibert_scivocab_cased",
3
  "attention_probs_dropout_prob": 0.1,
4
  "classifier_dropout": null,
5
  "hidden_act": "gelu",
 
1
  {
2
+ "_name_or_path": "/srv/storage/datalake2@storage1.nancy.grid5000.fr/lfoppiano/embeddings/scibert_scivocab_cased",
3
  "attention_probs_dropout_prob": 0.1,
4
  "classifier_dropout": null,
5
  "hidden_act": "gelu",
software-BERT_CRF/config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "model_name": "software-BERT_CRF",
3
  "architecture": "BERT_CRF",
4
  "embeddings_name": null,
5
  "char_vocab_size": 549,
@@ -35,4 +35,4 @@
35
  "I-<version>": 8,
36
  "O": 9
37
  }
38
- }
 
1
  {
2
+ "model_name": "grobid-software_all-BERT_CRF",
3
  "architecture": "BERT_CRF",
4
  "embeddings_name": null,
5
  "char_vocab_size": 549,
 
35
  "I-<version>": 8,
36
  "O": 9
37
  }
38
+ }
software-BERT_CRF/model_weights.hdf5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1a3c199e07db398c4c659e360a6451facd87aa1797bbc48f27ceabc3bf9fc618
3
  size 440058504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bfe3fd1cf3ce313a7afc2d43d93ba90abf02ea1acd010bf1bb53d4c16151cb6f
3
  size 440058504
software-BERT_CRF/transformer-config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/srv/storage/datalake@storage2.nancy.grid5000.fr/lfoppiano/embeddings/scibert_scivocab_cased",
3
  "attention_probs_dropout_prob": 0.1,
4
  "classifier_dropout": null,
5
  "hidden_act": "gelu",
 
1
  {
2
+ "_name_or_path": "/srv/storage/datalake2@storage1.nancy.grid5000.fr/lfoppiano/embeddings/scibert_scivocab_cased",
3
  "attention_probs_dropout_prob": 0.1,
4
  "classifier_dropout": null,
5
  "hidden_act": "gelu",
software-type-BERT/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "grobid-software-type-all-BERT",
3
+ "architecture": "BERT",
4
+ "embeddings_name": null,
5
+ "char_vocab_size": 296,
6
+ "case_vocab_size": 8,
7
+ "char_embedding_size": 25,
8
+ "num_char_lstm_units": 25,
9
+ "max_char_length": 30,
10
+ "features_vocabulary_size": 12,
11
+ "features_indices": null,
12
+ "features_embedding_size": 4,
13
+ "features_lstm_units": 4,
14
+ "max_sequence_length": 512,
15
+ "word_embedding_size": 0,
16
+ "num_word_lstm_units": 100,
17
+ "case_embedding_size": 5,
18
+ "dropout": 0.5,
19
+ "recurrent_dropout": 0.5,
20
+ "use_crf": false,
21
+ "use_chain_crf": false,
22
+ "fold_number": 1,
23
+ "batch_size": 8,
24
+ "transformer_name": "michiyasunaga/LinkBERT-basecased",
25
+ "use_ELMo": false,
26
+ "labels": {
27
+ "<PAD>": 0,
28
+ "B-<component>": 1,
29
+ "B-<environment>": 2,
30
+ "B-<implicit>": 3,
31
+ "B-<language>": 4,
32
+ "I-<component>": 5,
33
+ "I-<environment>": 6,
34
+ "I-<implicit>": 7,
35
+ "I-<language>": 8,
36
+ "O": 9
37
+ }
38
+ }
software-type-BERT/model_weights.hdf5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a191a4fbd1b1eda23a1a700de58caed1ba4a45bc1bfa8486110daa1eef0fe464
3
+ size 433547432
software-type-BERT/preprocessor.json ADDED
@@ -0,0 +1,343 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "padding": true,
3
+ "return_lengths": false,
4
+ "return_word_embeddings": false,
5
+ "return_casing": false,
6
+ "return_features": false,
7
+ "return_chars": false,
8
+ "return_bert_embeddings": true,
9
+ "vocab_char": {
10
+ "<PAD>": 0,
11
+ "<UNK>": 1,
12
+ "!": 2,
13
+ "\"": 3,
14
+ "#": 4,
15
+ "$": 5,
16
+ "%": 6,
17
+ "&": 7,
18
+ "'": 8,
19
+ "(": 9,
20
+ ")": 10,
21
+ "*": 11,
22
+ "+": 12,
23
+ ",": 13,
24
+ "-": 14,
25
+ ".": 15,
26
+ "/": 16,
27
+ "0": 17,
28
+ "1": 18,
29
+ "2": 19,
30
+ "3": 20,
31
+ "4": 21,
32
+ "5": 22,
33
+ "6": 23,
34
+ "7": 24,
35
+ "8": 25,
36
+ "9": 26,
37
+ ":": 27,
38
+ ";": 28,
39
+ "<": 29,
40
+ "=": 30,
41
+ ">": 31,
42
+ "?": 32,
43
+ "@": 33,
44
+ "A": 34,
45
+ "B": 35,
46
+ "C": 36,
47
+ "D": 37,
48
+ "E": 38,
49
+ "F": 39,
50
+ "G": 40,
51
+ "H": 41,
52
+ "I": 42,
53
+ "J": 43,
54
+ "K": 44,
55
+ "L": 45,
56
+ "M": 46,
57
+ "N": 47,
58
+ "O": 48,
59
+ "P": 49,
60
+ "Q": 50,
61
+ "R": 51,
62
+ "S": 52,
63
+ "T": 53,
64
+ "U": 54,
65
+ "V": 55,
66
+ "W": 56,
67
+ "X": 57,
68
+ "Y": 58,
69
+ "Z": 59,
70
+ "[": 60,
71
+ "]": 61,
72
+ "^": 62,
73
+ "_": 63,
74
+ "`": 64,
75
+ "a": 65,
76
+ "b": 66,
77
+ "c": 67,
78
+ "d": 68,
79
+ "e": 69,
80
+ "f": 70,
81
+ "g": 71,
82
+ "h": 72,
83
+ "i": 73,
84
+ "j": 74,
85
+ "k": 75,
86
+ "l": 76,
87
+ "m": 77,
88
+ "n": 78,
89
+ "o": 79,
90
+ "p": 80,
91
+ "q": 81,
92
+ "r": 82,
93
+ "s": 83,
94
+ "t": 84,
95
+ "u": 85,
96
+ "v": 86,
97
+ "w": 87,
98
+ "x": 88,
99
+ "y": 89,
100
+ "z": 90,
101
+ "{": 91,
102
+ "|": 92,
103
+ "}": 93,
104
+ "~": 94,
105
+ "\u00a1": 95,
106
+ "\u00a2": 96,
107
+ "\u00a3": 97,
108
+ "\u00a7": 98,
109
+ "\u00a8": 99,
110
+ "\u00a9": 100,
111
+ "\u00ae": 101,
112
+ "\u00af": 102,
113
+ "\u00b0": 103,
114
+ "\u00b1": 104,
115
+ "\u00b2": 105,
116
+ "\u00b4": 106,
117
+ "\u00b5": 107,
118
+ "\u00b7": 108,
119
+ "\u00b8": 109,
120
+ "\u00b9": 110,
121
+ "\u00ba": 111,
122
+ "\u00bb": 112,
123
+ "\u00bc": 113,
124
+ "\u00bd": 114,
125
+ "\u00c0": 115,
126
+ "\u00c1": 116,
127
+ "\u00c2": 117,
128
+ "\u00c3": 118,
129
+ "\u00c5": 119,
130
+ "\u00c9": 120,
131
+ "\u00d2": 121,
132
+ "\u00d4": 122,
133
+ "\u00d5": 123,
134
+ "\u00d7": 124,
135
+ "\u00dc": 125,
136
+ "\u00de": 126,
137
+ "\u00df": 127,
138
+ "\u00e1": 128,
139
+ "\u00e2": 129,
140
+ "\u00e3": 130,
141
+ "\u00e4": 131,
142
+ "\u00e5": 132,
143
+ "\u00e7": 133,
144
+ "\u00e8": 134,
145
+ "\u00e9": 135,
146
+ "\u00ea": 136,
147
+ "\u00ed": 137,
148
+ "\u00ef": 138,
149
+ "\u00f0": 139,
150
+ "\u00f3": 140,
151
+ "\u00f4": 141,
152
+ "\u00f6": 142,
153
+ "\u00f8": 143,
154
+ "\u00fa": 144,
155
+ "\u00fc": 145,
156
+ "\u00fe": 146,
157
+ "\u0107": 147,
158
+ "\u010c": 148,
159
+ "\u010d": 149,
160
+ "\u0113": 150,
161
+ "\u0117": 151,
162
+ "\u0131": 152,
163
+ "\u0144": 153,
164
+ "\u0160": 154,
165
+ "\u0161": 155,
166
+ "\u0177": 156,
167
+ "\u017e": 157,
168
+ "\u0192": 158,
169
+ "\u01eb": 159,
170
+ "\u0251": 160,
171
+ "\u025b": 161,
172
+ "\u02c2": 162,
173
+ "\u02da": 163,
174
+ "\u02dc": 164,
175
+ "\u030a": 165,
176
+ "\u031d": 166,
177
+ "\u0357": 167,
178
+ "\u0358": 168,
179
+ "\u0374": 169,
180
+ "\u0394": 170,
181
+ "\u0397": 171,
182
+ "\u03a6": 172,
183
+ "\u03a9": 173,
184
+ "\u03b1": 174,
185
+ "\u03b2": 175,
186
+ "\u03b3": 176,
187
+ "\u03b4": 177,
188
+ "\u03b5": 178,
189
+ "\u03b6": 179,
190
+ "\u03b7": 180,
191
+ "\u03b8": 181,
192
+ "\u03ba": 182,
193
+ "\u03bb": 183,
194
+ "\u03bc": 184,
195
+ "\u03bd": 185,
196
+ "\u03be": 186,
197
+ "\u03c0": 187,
198
+ "\u03c1": 188,
199
+ "\u03c3": 189,
200
+ "\u03c4": 190,
201
+ "\u03c5": 191,
202
+ "\u03c6": 192,
203
+ "\u03c7": 193,
204
+ "\u03c8": 194,
205
+ "\u03c9": 195,
206
+ "\u03d2": 196,
207
+ "\u03d5": 197,
208
+ "\u03e9": 198,
209
+ "\u03ea": 199,
210
+ "\u03eb": 200,
211
+ "\u03ed": 201,
212
+ "\u03ee": 202,
213
+ "\u03f3": 203,
214
+ "\u03fd": 204,
215
+ "\u03fe": 205,
216
+ "\u0408": 206,
217
+ "\u0412": 207,
218
+ "\u0545": 208,
219
+ "\u0546": 209,
220
+ "\u0609": 210,
221
+ "\u060a": 211,
222
+ "\u060e": 212,
223
+ "\u1e91": 213,
224
+ "\u2020": 214,
225
+ "\u2021": 215,
226
+ "\u2022": 216,
227
+ "\u2026": 217,
228
+ "\u2032": 218,
229
+ "\u2033": 219,
230
+ "\u2044": 220,
231
+ "\u2074": 221,
232
+ "\u2075": 222,
233
+ "\u2081": 223,
234
+ "\u20ac": 224,
235
+ "\u2103": 225,
236
+ "\u2122": 226,
237
+ "\u2150": 227,
238
+ "\u2192": 228,
239
+ "\u2202": 229,
240
+ "\u2206": 230,
241
+ "\u2208": 231,
242
+ "\u2212": 232,
243
+ "\u2215": 233,
244
+ "\u221a": 234,
245
+ "\u221e": 235,
246
+ "\u223c": 236,
247
+ "\u2248": 237,
248
+ "\u2260": 238,
249
+ "\u2261": 239,
250
+ "\u2264": 240,
251
+ "\u2265": 241,
252
+ "\u22c5": 242,
253
+ "\u232c": 243,
254
+ "\u233a": 244,
255
+ "\u233d": 245,
256
+ "\u2423": 246,
257
+ "\u2424": 247,
258
+ "\u2425": 248,
259
+ "\u24c7": 249,
260
+ "\u2500": 250,
261
+ "\u2502": 251,
262
+ "\u25b3": 252,
263
+ "\u25b8": 253,
264
+ "\u2a7e": 254,
265
+ "\ue024": 255,
266
+ "\ue02c": 256,
267
+ "\ue02e": 257,
268
+ "\ue032": 258,
269
+ "\ue039": 259,
270
+ "\ue061": 260,
271
+ "\ue062": 261,
272
+ "\ue067": 262,
273
+ "\ue06d": 263,
274
+ "\ue074": 264,
275
+ "\ue07a": 265,
276
+ "\ue093": 266,
277
+ "\ue09d": 267,
278
+ "\ue103": 268,
279
+ "\ue104": 269,
280
+ "\uf062": 270,
281
+ "\uf06c": 271,
282
+ "\uf06d": 272,
283
+ "\uf077": 273,
284
+ "\uf0a3": 274,
285
+ "\uf0b0": 275,
286
+ "\uf0b1": 276,
287
+ "\uf0b4": 277,
288
+ "\uf0b7": 278,
289
+ "\uf0b9": 279,
290
+ "\uf643": 280,
291
+ "\uf644": 281,
292
+ "\uf645": 282,
293
+ "\uf64b": 283,
294
+ "\uf64c": 284,
295
+ "\uf76d": 285,
296
+ "\uff1a": 286,
297
+ "\uff1c": 287,
298
+ "\uff1e": 288,
299
+ "\ufffd": 289,
300
+ "\ud835\udc42": 290,
301
+ "\ud835\udc56": 291,
302
+ "\ud835\udc57": 292,
303
+ "\ud835\udc61": 293,
304
+ "\ud835\udefc": 294,
305
+ "\ud835\udeff": 295
306
+ },
307
+ "vocab_tag": {
308
+ "<PAD>": 0,
309
+ "B-<component>": 1,
310
+ "B-<environment>": 2,
311
+ "B-<implicit>": 3,
312
+ "B-<language>": 4,
313
+ "I-<component>": 5,
314
+ "I-<environment>": 6,
315
+ "I-<implicit>": 7,
316
+ "I-<language>": 8,
317
+ "O": 9
318
+ },
319
+ "vocab_case": [
320
+ "<PAD>",
321
+ "numeric",
322
+ "allLower",
323
+ "allUpper",
324
+ "initialUpper",
325
+ "other",
326
+ "mainly_numeric",
327
+ "contains_digit"
328
+ ],
329
+ "max_char_length": 30,
330
+ "feature_preprocessor": null,
331
+ "indice_tag": {
332
+ "0": "<PAD>",
333
+ "1": "B-<component>",
334
+ "2": "B-<environment>",
335
+ "3": "B-<implicit>",
336
+ "4": "B-<language>",
337
+ "5": "I-<component>",
338
+ "6": "I-<environment>",
339
+ "7": "I-<implicit>",
340
+ "8": "I-<language>",
341
+ "9": "O"
342
+ }
343
+ }
software-type-BERT/transformer-config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/srv/storage/[email protected]/lfoppiano/embeddings/LinkBERT-base",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "pad_token_id": 0,
20
+ "position_embedding_type": "absolute",
21
+ "transformers_version": "4.33.2",
22
+ "type_vocab_size": 2,
23
+ "use_cache": true,
24
+ "vocab_size": 28996
25
+ }
software-type-BERT/transformer-tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
software-type-BERT/transformer-tokenizer/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
software-type-BERT/transformer-tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "clean_up_tokenization_spaces": true,
4
+ "cls_token": "[CLS]",
5
+ "do_lower_case": false,
6
+ "mask_token": "[MASK]",
7
+ "max_length": 512,
8
+ "model_max_length": 512,
9
+ "pad_token": "[PAD]",
10
+ "sep_token": "[SEP]",
11
+ "strip_accents": null,
12
+ "tokenize_chinese_chars": true,
13
+ "tokenizer_class": "BertTokenizer",
14
+ "unk_token": "[UNK]"
15
+ }
software-type-BERT/transformer-tokenizer/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
software-type-BERT_CRF/config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "model_name": "software-type-BERT_CRF",
3
  "architecture": "BERT_CRF",
4
  "embeddings_name": null,
5
  "char_vocab_size": 296,
@@ -35,4 +35,4 @@
35
  "I-<language>": 8,
36
  "O": 9
37
  }
38
- }
 
1
  {
2
+ "model_name": "grobid-software-type-all-BERT_CRF",
3
  "architecture": "BERT_CRF",
4
  "embeddings_name": null,
5
  "char_vocab_size": 296,
 
35
  "I-<language>": 8,
36
  "O": 9
37
  }
38
+ }
software-type-BERT_CRF/model_weights.hdf5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:739b00c7e174fab6831c263c11da7e38ae5ad17b752af74a1f0eff62b4aebb01
3
  size 433545864
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51622a068f47b42d066afd922b550aaa7818d4ce3d8716d62e7d042ee44e8c4b
3
  size 433545864
software-type-BERT_CRF/transformer-config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/srv/storage/datalake@storage2.nancy.grid5000.fr/lfoppiano/embeddings/LinkBERT-base",
3
  "architectures": [
4
  "BertModel"
5
  ],
 
1
  {
2
+ "_name_or_path": "/srv/storage/datalake2@storage1.nancy.grid5000.fr/lfoppiano/embeddings/LinkBERT-base",
3
  "architectures": [
4
  "BertModel"
5
  ],