WilliamHH commited on
Commit
3619203
·
verified ·
1 Parent(s): 968266f

Training in progress, step 500

Browse files
config.json ADDED
@@ -0,0 +1,352 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "ModernBertForSequenceClassification"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 50281,
8
+ "classifier_activation": "gelu",
9
+ "classifier_bias": false,
10
+ "classifier_dropout": 0.0,
11
+ "classifier_pooling": "mean",
12
+ "cls_token_id": 50281,
13
+ "decoder_bias": true,
14
+ "deterministic_flash_attn": false,
15
+ "dtype": "float32",
16
+ "embedding_dropout": 0.0,
17
+ "eos_token_id": 50282,
18
+ "global_attn_every_n_layers": 3,
19
+ "global_rope_theta": 160000.0,
20
+ "gradient_checkpointing": false,
21
+ "hidden_activation": "gelu",
22
+ "hidden_size": 1024,
23
+ "id2label": {
24
+ "0": "LABEL_0",
25
+ "1": "LABEL_1",
26
+ "2": "LABEL_2",
27
+ "3": "LABEL_3",
28
+ "4": "LABEL_4",
29
+ "5": "LABEL_5",
30
+ "6": "LABEL_6",
31
+ "7": "LABEL_7",
32
+ "8": "LABEL_8",
33
+ "9": "LABEL_9",
34
+ "10": "LABEL_10",
35
+ "11": "LABEL_11",
36
+ "12": "LABEL_12",
37
+ "13": "LABEL_13",
38
+ "14": "LABEL_14",
39
+ "15": "LABEL_15",
40
+ "16": "LABEL_16",
41
+ "17": "LABEL_17",
42
+ "18": "LABEL_18",
43
+ "19": "LABEL_19",
44
+ "20": "LABEL_20",
45
+ "21": "LABEL_21",
46
+ "22": "LABEL_22",
47
+ "23": "LABEL_23",
48
+ "24": "LABEL_24",
49
+ "25": "LABEL_25",
50
+ "26": "LABEL_26",
51
+ "27": "LABEL_27",
52
+ "28": "LABEL_28",
53
+ "29": "LABEL_29",
54
+ "30": "LABEL_30",
55
+ "31": "LABEL_31",
56
+ "32": "LABEL_32",
57
+ "33": "LABEL_33",
58
+ "34": "LABEL_34",
59
+ "35": "LABEL_35",
60
+ "36": "LABEL_36",
61
+ "37": "LABEL_37",
62
+ "38": "LABEL_38",
63
+ "39": "LABEL_39",
64
+ "40": "LABEL_40",
65
+ "41": "LABEL_41",
66
+ "42": "LABEL_42",
67
+ "43": "LABEL_43",
68
+ "44": "LABEL_44",
69
+ "45": "LABEL_45",
70
+ "46": "LABEL_46",
71
+ "47": "LABEL_47",
72
+ "48": "LABEL_48",
73
+ "49": "LABEL_49",
74
+ "50": "LABEL_50",
75
+ "51": "LABEL_51",
76
+ "52": "LABEL_52",
77
+ "53": "LABEL_53",
78
+ "54": "LABEL_54",
79
+ "55": "LABEL_55",
80
+ "56": "LABEL_56",
81
+ "57": "LABEL_57",
82
+ "58": "LABEL_58",
83
+ "59": "LABEL_59",
84
+ "60": "LABEL_60",
85
+ "61": "LABEL_61",
86
+ "62": "LABEL_62",
87
+ "63": "LABEL_63",
88
+ "64": "LABEL_64",
89
+ "65": "LABEL_65",
90
+ "66": "LABEL_66",
91
+ "67": "LABEL_67",
92
+ "68": "LABEL_68",
93
+ "69": "LABEL_69",
94
+ "70": "LABEL_70",
95
+ "71": "LABEL_71",
96
+ "72": "LABEL_72",
97
+ "73": "LABEL_73",
98
+ "74": "LABEL_74",
99
+ "75": "LABEL_75",
100
+ "76": "LABEL_76",
101
+ "77": "LABEL_77",
102
+ "78": "LABEL_78",
103
+ "79": "LABEL_79",
104
+ "80": "LABEL_80",
105
+ "81": "LABEL_81",
106
+ "82": "LABEL_82",
107
+ "83": "LABEL_83",
108
+ "84": "LABEL_84",
109
+ "85": "LABEL_85",
110
+ "86": "LABEL_86",
111
+ "87": "LABEL_87",
112
+ "88": "LABEL_88",
113
+ "89": "LABEL_89",
114
+ "90": "LABEL_90",
115
+ "91": "LABEL_91",
116
+ "92": "LABEL_92",
117
+ "93": "LABEL_93",
118
+ "94": "LABEL_94",
119
+ "95": "LABEL_95",
120
+ "96": "LABEL_96",
121
+ "97": "LABEL_97",
122
+ "98": "LABEL_98",
123
+ "99": "LABEL_99",
124
+ "100": "LABEL_100",
125
+ "101": "LABEL_101",
126
+ "102": "LABEL_102",
127
+ "103": "LABEL_103",
128
+ "104": "LABEL_104",
129
+ "105": "LABEL_105",
130
+ "106": "LABEL_106",
131
+ "107": "LABEL_107",
132
+ "108": "LABEL_108",
133
+ "109": "LABEL_109",
134
+ "110": "LABEL_110",
135
+ "111": "LABEL_111",
136
+ "112": "LABEL_112",
137
+ "113": "LABEL_113",
138
+ "114": "LABEL_114",
139
+ "115": "LABEL_115",
140
+ "116": "LABEL_116",
141
+ "117": "LABEL_117",
142
+ "118": "LABEL_118",
143
+ "119": "LABEL_119",
144
+ "120": "LABEL_120",
145
+ "121": "LABEL_121",
146
+ "122": "LABEL_122",
147
+ "123": "LABEL_123",
148
+ "124": "LABEL_124",
149
+ "125": "LABEL_125",
150
+ "126": "LABEL_126",
151
+ "127": "LABEL_127",
152
+ "128": "LABEL_128",
153
+ "129": "LABEL_129",
154
+ "130": "LABEL_130",
155
+ "131": "LABEL_131",
156
+ "132": "LABEL_132",
157
+ "133": "LABEL_133",
158
+ "134": "LABEL_134",
159
+ "135": "LABEL_135",
160
+ "136": "LABEL_136",
161
+ "137": "LABEL_137",
162
+ "138": "LABEL_138",
163
+ "139": "LABEL_139",
164
+ "140": "LABEL_140",
165
+ "141": "LABEL_141",
166
+ "142": "LABEL_142",
167
+ "143": "LABEL_143",
168
+ "144": "LABEL_144",
169
+ "145": "LABEL_145",
170
+ "146": "LABEL_146",
171
+ "147": "LABEL_147",
172
+ "148": "LABEL_148",
173
+ "149": "LABEL_149",
174
+ "150": "LABEL_150"
175
+ },
176
+ "initializer_cutoff_factor": 2.0,
177
+ "initializer_range": 0.02,
178
+ "intermediate_size": 2624,
179
+ "label2id": {
180
+ "LABEL_0": 0,
181
+ "LABEL_1": 1,
182
+ "LABEL_10": 10,
183
+ "LABEL_100": 100,
184
+ "LABEL_101": 101,
185
+ "LABEL_102": 102,
186
+ "LABEL_103": 103,
187
+ "LABEL_104": 104,
188
+ "LABEL_105": 105,
189
+ "LABEL_106": 106,
190
+ "LABEL_107": 107,
191
+ "LABEL_108": 108,
192
+ "LABEL_109": 109,
193
+ "LABEL_11": 11,
194
+ "LABEL_110": 110,
195
+ "LABEL_111": 111,
196
+ "LABEL_112": 112,
197
+ "LABEL_113": 113,
198
+ "LABEL_114": 114,
199
+ "LABEL_115": 115,
200
+ "LABEL_116": 116,
201
+ "LABEL_117": 117,
202
+ "LABEL_118": 118,
203
+ "LABEL_119": 119,
204
+ "LABEL_12": 12,
205
+ "LABEL_120": 120,
206
+ "LABEL_121": 121,
207
+ "LABEL_122": 122,
208
+ "LABEL_123": 123,
209
+ "LABEL_124": 124,
210
+ "LABEL_125": 125,
211
+ "LABEL_126": 126,
212
+ "LABEL_127": 127,
213
+ "LABEL_128": 128,
214
+ "LABEL_129": 129,
215
+ "LABEL_13": 13,
216
+ "LABEL_130": 130,
217
+ "LABEL_131": 131,
218
+ "LABEL_132": 132,
219
+ "LABEL_133": 133,
220
+ "LABEL_134": 134,
221
+ "LABEL_135": 135,
222
+ "LABEL_136": 136,
223
+ "LABEL_137": 137,
224
+ "LABEL_138": 138,
225
+ "LABEL_139": 139,
226
+ "LABEL_14": 14,
227
+ "LABEL_140": 140,
228
+ "LABEL_141": 141,
229
+ "LABEL_142": 142,
230
+ "LABEL_143": 143,
231
+ "LABEL_144": 144,
232
+ "LABEL_145": 145,
233
+ "LABEL_146": 146,
234
+ "LABEL_147": 147,
235
+ "LABEL_148": 148,
236
+ "LABEL_149": 149,
237
+ "LABEL_15": 15,
238
+ "LABEL_150": 150,
239
+ "LABEL_16": 16,
240
+ "LABEL_17": 17,
241
+ "LABEL_18": 18,
242
+ "LABEL_19": 19,
243
+ "LABEL_2": 2,
244
+ "LABEL_20": 20,
245
+ "LABEL_21": 21,
246
+ "LABEL_22": 22,
247
+ "LABEL_23": 23,
248
+ "LABEL_24": 24,
249
+ "LABEL_25": 25,
250
+ "LABEL_26": 26,
251
+ "LABEL_27": 27,
252
+ "LABEL_28": 28,
253
+ "LABEL_29": 29,
254
+ "LABEL_3": 3,
255
+ "LABEL_30": 30,
256
+ "LABEL_31": 31,
257
+ "LABEL_32": 32,
258
+ "LABEL_33": 33,
259
+ "LABEL_34": 34,
260
+ "LABEL_35": 35,
261
+ "LABEL_36": 36,
262
+ "LABEL_37": 37,
263
+ "LABEL_38": 38,
264
+ "LABEL_39": 39,
265
+ "LABEL_4": 4,
266
+ "LABEL_40": 40,
267
+ "LABEL_41": 41,
268
+ "LABEL_42": 42,
269
+ "LABEL_43": 43,
270
+ "LABEL_44": 44,
271
+ "LABEL_45": 45,
272
+ "LABEL_46": 46,
273
+ "LABEL_47": 47,
274
+ "LABEL_48": 48,
275
+ "LABEL_49": 49,
276
+ "LABEL_5": 5,
277
+ "LABEL_50": 50,
278
+ "LABEL_51": 51,
279
+ "LABEL_52": 52,
280
+ "LABEL_53": 53,
281
+ "LABEL_54": 54,
282
+ "LABEL_55": 55,
283
+ "LABEL_56": 56,
284
+ "LABEL_57": 57,
285
+ "LABEL_58": 58,
286
+ "LABEL_59": 59,
287
+ "LABEL_6": 6,
288
+ "LABEL_60": 60,
289
+ "LABEL_61": 61,
290
+ "LABEL_62": 62,
291
+ "LABEL_63": 63,
292
+ "LABEL_64": 64,
293
+ "LABEL_65": 65,
294
+ "LABEL_66": 66,
295
+ "LABEL_67": 67,
296
+ "LABEL_68": 68,
297
+ "LABEL_69": 69,
298
+ "LABEL_7": 7,
299
+ "LABEL_70": 70,
300
+ "LABEL_71": 71,
301
+ "LABEL_72": 72,
302
+ "LABEL_73": 73,
303
+ "LABEL_74": 74,
304
+ "LABEL_75": 75,
305
+ "LABEL_76": 76,
306
+ "LABEL_77": 77,
307
+ "LABEL_78": 78,
308
+ "LABEL_79": 79,
309
+ "LABEL_8": 8,
310
+ "LABEL_80": 80,
311
+ "LABEL_81": 81,
312
+ "LABEL_82": 82,
313
+ "LABEL_83": 83,
314
+ "LABEL_84": 84,
315
+ "LABEL_85": 85,
316
+ "LABEL_86": 86,
317
+ "LABEL_87": 87,
318
+ "LABEL_88": 88,
319
+ "LABEL_89": 89,
320
+ "LABEL_9": 9,
321
+ "LABEL_90": 90,
322
+ "LABEL_91": 91,
323
+ "LABEL_92": 92,
324
+ "LABEL_93": 93,
325
+ "LABEL_94": 94,
326
+ "LABEL_95": 95,
327
+ "LABEL_96": 96,
328
+ "LABEL_97": 97,
329
+ "LABEL_98": 98,
330
+ "LABEL_99": 99
331
+ },
332
+ "layer_norm_eps": 1e-05,
333
+ "local_attention": 128,
334
+ "local_rope_theta": 10000.0,
335
+ "max_position_embeddings": 8192,
336
+ "mlp_bias": false,
337
+ "mlp_dropout": 0.0,
338
+ "model_type": "modernbert",
339
+ "norm_bias": false,
340
+ "norm_eps": 1e-05,
341
+ "num_attention_heads": 16,
342
+ "num_hidden_layers": 28,
343
+ "pad_token_id": 50283,
344
+ "position_embedding_type": "absolute",
345
+ "problem_type": "single_label_classification",
346
+ "repad_logits_with_grad": false,
347
+ "sep_token_id": 50282,
348
+ "sparse_pred_ignore_index": -100,
349
+ "sparse_prediction": false,
350
+ "transformers_version": "4.57.0",
351
+ "vocab_size": 50368
352
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad29bb54027c872b816d306dfdaf50f92b081201f4acd1284055f89f0e98936c
3
+ size 1583962548
runs/Oct13_13-53-24_0aee9d3f3417/events.out.tfevents.1760363607.0aee9d3f3417.805.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:948f812f3633706acbcb35041b280e6cdc2ef05d235653bb6cfbf233c27f4d2b
3
+ size 13351
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccfc02b601d7318637b13b6785918a648133d377d3ef6f7f5f795f4ed2608c9b
3
+ size 5841