Theoreticallyhugo commited on
Commit
c388e25
1 Parent(s): 49e8f58

Training in progress, epoch 10, checkpoint

Browse files
checkpoint-810/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "allenai/longformer-base-4096",
3
+ "architectures": [
4
+ "LongformerForTokenClassification"
5
+ ],
6
+ "attention_mode": "longformer",
7
+ "attention_probs_dropout_prob": 0.1,
8
+ "attention_window": [
9
+ 700,
10
+ 700,
11
+ 700,
12
+ 700,
13
+ 700,
14
+ 700,
15
+ 700,
16
+ 700,
17
+ 700,
18
+ 700,
19
+ 700,
20
+ 700
21
+ ],
22
+ "bos_token_id": 0,
23
+ "eos_token_id": 2,
24
+ "gradient_checkpointing": false,
25
+ "hidden_act": "gelu",
26
+ "hidden_dropout_prob": 0.1,
27
+ "hidden_size": 768,
28
+ "id2label": {
29
+ "0": "O",
30
+ "1": "B",
31
+ "2": "I"
32
+ },
33
+ "ignore_attention_mask": false,
34
+ "initializer_range": 0.02,
35
+ "intermediate_size": 3072,
36
+ "label2id": {
37
+ "B": 1,
38
+ "I": 2,
39
+ "O": 0
40
+ },
41
+ "layer_norm_eps": 1e-05,
42
+ "max_position_embeddings": 4098,
43
+ "model_type": "longformer",
44
+ "num_attention_heads": 12,
45
+ "num_hidden_layers": 12,
46
+ "onnx_export": false,
47
+ "pad_token_id": 1,
48
+ "sep_token_id": 2,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.38.2",
51
+ "type_vocab_size": 1,
52
+ "vocab_size": 50265
53
+ }
checkpoint-810/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-810/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bfab6dcca559f8e8bcec4cb735066969ef5d718827e72bdca4eba874c6aa313
3
+ size 592318676
checkpoint-810/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf8b12394dd1e2159879297504a4f5507318d3c3d813197526dc064ad65a7f23
3
+ size 1014645498
checkpoint-810/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05f741fe22f4ca5655fda1f9ef1ecba0015571f3b8afe49bf335811d1e94a138
3
+ size 14244
checkpoint-810/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e415b77a665cacd845500f2c1b0e83fd6d1720a5953f0905e282772d30c115c
3
+ size 1064
checkpoint-810/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
checkpoint-810/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-810/tokenizer_config.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<s>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<pad>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "50264": {
37
+ "content": "<mask>",
38
+ "lstrip": true,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ }
44
+ },
45
+ "bos_token": "<s>",
46
+ "clean_up_tokenization_spaces": true,
47
+ "cls_token": "<s>",
48
+ "eos_token": "</s>",
49
+ "errors": "replace",
50
+ "mask_token": "<mask>",
51
+ "model_max_length": 4096,
52
+ "pad_token": "<pad>",
53
+ "sep_token": "</s>",
54
+ "tokenizer_class": "LongformerTokenizer",
55
+ "trim_offsets": true,
56
+ "unk_token": "<unk>"
57
+ }
checkpoint-810/trainer_state.json ADDED
@@ -0,0 +1,418 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 10.0,
5
+ "eval_steps": 500,
6
+ "global_step": 810,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 1.0,
13
+ "eval_B": {
14
+ "f1-score": 0.8087666523420713,
15
+ "precision": 0.8189730200174065,
16
+ "recall": 0.7988115449915111,
17
+ "support": 1178.0
18
+ },
19
+ "eval_I": {
20
+ "f1-score": 0.9374597034171502,
21
+ "precision": 0.9521418826739427,
22
+ "recall": 0.9232234509762421,
23
+ "support": 18899.0
24
+ },
25
+ "eval_O": {
26
+ "f1-score": 0.8881362400419788,
27
+ "precision": 0.863303347862376,
28
+ "recall": 0.9144400785854617,
29
+ "support": 10180.0
30
+ },
31
+ "eval_accuracy": 0.9154245298608586,
32
+ "eval_loss": 0.2228638380765915,
33
+ "eval_macro avg": {
34
+ "f1-score": 0.8781208652670669,
35
+ "precision": 0.8781394168512416,
36
+ "recall": 0.8788250248510717,
37
+ "support": 30257.0
38
+ },
39
+ "eval_runtime": 4.7167,
40
+ "eval_samples_per_second": 17.173,
41
+ "eval_steps_per_second": 2.332,
42
+ "eval_weighted avg": {
43
+ "f1-score": 0.9158543469268937,
44
+ "precision": 0.9170673807540184,
45
+ "recall": 0.9154245298608586,
46
+ "support": 30257.0
47
+ },
48
+ "step": 81
49
+ },
50
+ {
51
+ "epoch": 2.0,
52
+ "eval_B": {
53
+ "f1-score": 0.8401515151515152,
54
+ "precision": 0.7585499316005472,
55
+ "recall": 0.9414261460101867,
56
+ "support": 1178.0
57
+ },
58
+ "eval_I": {
59
+ "f1-score": 0.9568235603470944,
60
+ "precision": 0.9510219016256338,
61
+ "recall": 0.9626964389650247,
62
+ "support": 18899.0
63
+ },
64
+ "eval_O": {
65
+ "f1-score": 0.9185648054827655,
66
+ "precision": 0.9430877483443708,
67
+ "recall": 0.8952848722986247,
68
+ "support": 10180.0
69
+ },
70
+ "eval_accuracy": 0.9391876260038999,
71
+ "eval_loss": 0.16692574322223663,
72
+ "eval_macro avg": {
73
+ "f1-score": 0.905179960327125,
74
+ "precision": 0.8842198605235172,
75
+ "recall": 0.9331358190912787,
76
+ "support": 30257.0
77
+ },
78
+ "eval_runtime": 4.7116,
79
+ "eval_samples_per_second": 17.192,
80
+ "eval_steps_per_second": 2.335,
81
+ "eval_weighted avg": {
82
+ "f1-score": 0.9394089523635119,
83
+ "precision": 0.9408589092241132,
84
+ "recall": 0.9391876260038999,
85
+ "support": 30257.0
86
+ },
87
+ "step": 162
88
+ },
89
+ {
90
+ "epoch": 3.0,
91
+ "eval_B": {
92
+ "f1-score": 0.8612244897959184,
93
+ "precision": 0.8294025157232704,
94
+ "recall": 0.8955857385398981,
95
+ "support": 1178.0
96
+ },
97
+ "eval_I": {
98
+ "f1-score": 0.951106371634231,
99
+ "precision": 0.9584654236741712,
100
+ "recall": 0.9438594634636753,
101
+ "support": 18899.0
102
+ },
103
+ "eval_O": {
104
+ "f1-score": 0.9106743212999903,
105
+ "precision": 0.9021592442645074,
106
+ "recall": 0.9193516699410609,
107
+ "support": 10180.0
108
+ },
109
+ "eval_accuracy": 0.9337343424662061,
110
+ "eval_loss": 0.1742253601551056,
111
+ "eval_macro avg": {
112
+ "f1-score": 0.9076683942433799,
113
+ "precision": 0.8966757278873163,
114
+ "recall": 0.9195989573148781,
115
+ "support": 30257.0
116
+ },
117
+ "eval_runtime": 4.7094,
118
+ "eval_samples_per_second": 17.2,
119
+ "eval_steps_per_second": 2.336,
120
+ "eval_weighted avg": {
121
+ "f1-score": 0.9340035812317422,
122
+ "precision": 0.9344963252190521,
123
+ "recall": 0.9337343424662061,
124
+ "support": 30257.0
125
+ },
126
+ "step": 243
127
+ },
128
+ {
129
+ "epoch": 4.0,
130
+ "eval_B": {
131
+ "f1-score": 0.8717948717948719,
132
+ "precision": 0.8254931714719271,
133
+ "recall": 0.9235993208828522,
134
+ "support": 1178.0
135
+ },
136
+ "eval_I": {
137
+ "f1-score": 0.9561271586371415,
138
+ "precision": 0.9374586871408959,
139
+ "recall": 0.9755542621302714,
140
+ "support": 18899.0
141
+ },
142
+ "eval_O": {
143
+ "f1-score": 0.9117828500925355,
144
+ "precision": 0.956427955133736,
145
+ "recall": 0.8711198428290766,
146
+ "support": 10180.0
147
+ },
148
+ "eval_accuracy": 0.9383944211256899,
149
+ "eval_loss": 0.18792153894901276,
150
+ "eval_macro avg": {
151
+ "f1-score": 0.9132349601748496,
152
+ "precision": 0.9064599379155197,
153
+ "recall": 0.9234244752807333,
154
+ "support": 30257.0
155
+ },
156
+ "eval_runtime": 4.7342,
157
+ "eval_samples_per_second": 17.11,
158
+ "eval_steps_per_second": 2.324,
159
+ "eval_weighted avg": {
160
+ "f1-score": 0.9379241479327001,
161
+ "precision": 0.9394817486046585,
162
+ "recall": 0.9383944211256899,
163
+ "support": 30257.0
164
+ },
165
+ "step": 324
166
+ },
167
+ {
168
+ "epoch": 5.0,
169
+ "eval_B": {
170
+ "f1-score": 0.8742949234488315,
171
+ "precision": 0.8320552147239264,
172
+ "recall": 0.9210526315789473,
173
+ "support": 1178.0
174
+ },
175
+ "eval_I": {
176
+ "f1-score": 0.9583957528450158,
177
+ "precision": 0.9521148825065274,
178
+ "recall": 0.9647600402137679,
179
+ "support": 18899.0
180
+ },
181
+ "eval_O": {
182
+ "f1-score": 0.9198818996146724,
183
+ "precision": 0.9375701315923697,
184
+ "recall": 0.9028487229862475,
185
+ "support": 10180.0
186
+ },
187
+ "eval_accuracy": 0.9422282447037049,
188
+ "eval_loss": 0.19990424811840057,
189
+ "eval_macro avg": {
190
+ "f1-score": 0.9175241919695066,
191
+ "precision": 0.9072467429409411,
192
+ "recall": 0.9295537982596542,
193
+ "support": 30257.0
194
+ },
195
+ "eval_runtime": 4.7101,
196
+ "eval_samples_per_second": 17.197,
197
+ "eval_steps_per_second": 2.335,
198
+ "eval_weighted avg": {
199
+ "f1-score": 0.9421634164298524,
200
+ "precision": 0.9425469857238316,
201
+ "recall": 0.9422282447037049,
202
+ "support": 30257.0
203
+ },
204
+ "step": 405
205
+ },
206
+ {
207
+ "epoch": 6.0,
208
+ "eval_B": {
209
+ "f1-score": 0.8808247422680412,
210
+ "precision": 0.8564554931836408,
211
+ "recall": 0.9066213921901528,
212
+ "support": 1178.0
213
+ },
214
+ "eval_I": {
215
+ "f1-score": 0.957406827435014,
216
+ "precision": 0.9446361435855178,
217
+ "recall": 0.9705275411397428,
218
+ "support": 18899.0
219
+ },
220
+ "eval_O": {
221
+ "f1-score": 0.9163000050574014,
222
+ "precision": 0.9443344105076619,
223
+ "recall": 0.8898821218074656,
224
+ "support": 10180.0
225
+ },
226
+ "eval_accuracy": 0.9409062365733549,
227
+ "eval_loss": 0.20615626871585846,
228
+ "eval_macro avg": {
229
+ "f1-score": 0.9181771915868189,
230
+ "precision": 0.9151420157589403,
231
+ "recall": 0.922343685045787,
232
+ "support": 30257.0
233
+ },
234
+ "eval_runtime": 4.6949,
235
+ "eval_samples_per_second": 17.253,
236
+ "eval_steps_per_second": 2.343,
237
+ "eval_weighted avg": {
238
+ "f1-score": 0.9405948120953971,
239
+ "precision": 0.941101475610967,
240
+ "recall": 0.9409062365733549,
241
+ "support": 30257.0
242
+ },
243
+ "step": 486
244
+ },
245
+ {
246
+ "epoch": 6.17,
247
+ "grad_norm": 2.4104158878326416,
248
+ "learning_rate": 1.7530864197530865e-05,
249
+ "loss": 0.1662,
250
+ "step": 500
251
+ },
252
+ {
253
+ "epoch": 7.0,
254
+ "eval_B": {
255
+ "f1-score": 0.8836820083682009,
256
+ "precision": 0.8712871287128713,
257
+ "recall": 0.8964346349745331,
258
+ "support": 1178.0
259
+ },
260
+ "eval_I": {
261
+ "f1-score": 0.9614174269469787,
262
+ "precision": 0.9609854091774159,
263
+ "recall": 0.9618498333245146,
264
+ "support": 18899.0
265
+ },
266
+ "eval_O": {
267
+ "f1-score": 0.9271751440248166,
268
+ "precision": 0.9295093296475466,
269
+ "recall": 0.924852652259332,
270
+ "support": 10180.0
271
+ },
272
+ "eval_accuracy": 0.9468552731599299,
273
+ "eval_loss": 0.22925782203674316,
274
+ "eval_macro avg": {
275
+ "f1-score": 0.9240915264466655,
276
+ "precision": 0.9205939558459445,
277
+ "recall": 0.9277123735194599,
278
+ "support": 30257.0
279
+ },
280
+ "eval_runtime": 4.7288,
281
+ "eval_samples_per_second": 17.129,
282
+ "eval_steps_per_second": 2.326,
283
+ "eval_weighted avg": {
284
+ "f1-score": 0.9468700903559945,
285
+ "precision": 0.9469030129054358,
286
+ "recall": 0.9468552731599299,
287
+ "support": 30257.0
288
+ },
289
+ "step": 567
290
+ },
291
+ {
292
+ "epoch": 8.0,
293
+ "eval_B": {
294
+ "f1-score": 0.8803630363036303,
295
+ "precision": 0.8563402889245586,
296
+ "recall": 0.9057724957555179,
297
+ "support": 1178.0
298
+ },
299
+ "eval_I": {
300
+ "f1-score": 0.9580651967507121,
301
+ "precision": 0.9550928116947994,
302
+ "recall": 0.9610561405365363,
303
+ "support": 18899.0
304
+ },
305
+ "eval_O": {
306
+ "f1-score": 0.9208882720333101,
307
+ "precision": 0.9294576746047628,
308
+ "recall": 0.912475442043222,
309
+ "support": 10180.0
310
+ },
311
+ "eval_accuracy": 0.9425587467362925,
312
+ "eval_loss": 0.22973769903182983,
313
+ "eval_macro avg": {
314
+ "f1-score": 0.9197721683625509,
315
+ "precision": 0.9136302584080402,
316
+ "recall": 0.9264346927784253,
317
+ "support": 30257.0
318
+ },
319
+ "eval_runtime": 4.7191,
320
+ "eval_samples_per_second": 17.164,
321
+ "eval_steps_per_second": 2.331,
322
+ "eval_weighted avg": {
323
+ "f1-score": 0.9425317916335554,
324
+ "precision": 0.942623096673485,
325
+ "recall": 0.9425587467362925,
326
+ "support": 30257.0
327
+ },
328
+ "step": 648
329
+ },
330
+ {
331
+ "epoch": 9.0,
332
+ "eval_B": {
333
+ "f1-score": 0.8885191347753745,
334
+ "precision": 0.8711256117455138,
335
+ "recall": 0.9066213921901528,
336
+ "support": 1178.0
337
+ },
338
+ "eval_I": {
339
+ "f1-score": 0.9620326487625066,
340
+ "precision": 0.9574445783763954,
341
+ "recall": 0.9666649029049156,
342
+ "support": 18899.0
343
+ },
344
+ "eval_O": {
345
+ "f1-score": 0.9275707898658718,
346
+ "precision": 0.9382914572864322,
347
+ "recall": 0.9170923379174852,
348
+ "support": 10180.0
349
+ },
350
+ "eval_accuracy": 0.9476484780381399,
351
+ "eval_loss": 0.24401217699050903,
352
+ "eval_macro avg": {
353
+ "f1-score": 0.9260408578012509,
354
+ "precision": 0.9222872158027805,
355
+ "recall": 0.9301262110041845,
356
+ "support": 30257.0
357
+ },
358
+ "eval_runtime": 4.701,
359
+ "eval_samples_per_second": 17.23,
360
+ "eval_steps_per_second": 2.34,
361
+ "eval_weighted avg": {
362
+ "f1-score": 0.94757580760031,
363
+ "precision": 0.9476398219436029,
364
+ "recall": 0.9476484780381399,
365
+ "support": 30257.0
366
+ },
367
+ "step": 729
368
+ },
369
+ {
370
+ "epoch": 10.0,
371
+ "eval_B": {
372
+ "f1-score": 0.875959595959596,
373
+ "precision": 0.8357748650732459,
374
+ "recall": 0.9202037351443124,
375
+ "support": 1178.0
376
+ },
377
+ "eval_I": {
378
+ "f1-score": 0.955652945166171,
379
+ "precision": 0.9424749163879599,
380
+ "recall": 0.9692047198264458,
381
+ "support": 18899.0
382
+ },
383
+ "eval_O": {
384
+ "f1-score": 0.9127632580563309,
385
+ "precision": 0.9441469816272966,
386
+ "recall": 0.8833988212180747,
387
+ "support": 10180.0
388
+ },
389
+ "eval_accuracy": 0.9384274713289487,
390
+ "eval_loss": 0.29471203684806824,
391
+ "eval_macro avg": {
392
+ "f1-score": 0.9147919330606994,
393
+ "precision": 0.9074655876961675,
394
+ "recall": 0.9242690920629443,
395
+ "support": 30257.0
396
+ },
397
+ "eval_runtime": 4.6985,
398
+ "eval_samples_per_second": 17.24,
399
+ "eval_steps_per_second": 2.341,
400
+ "eval_weighted avg": {
401
+ "f1-score": 0.9381199518045186,
402
+ "precision": 0.9388833165494999,
403
+ "recall": 0.9384274713289487,
404
+ "support": 30257.0
405
+ },
406
+ "step": 810
407
+ }
408
+ ],
409
+ "logging_steps": 500,
410
+ "max_steps": 4050,
411
+ "num_input_tokens_seen": 0,
412
+ "num_train_epochs": 50,
413
+ "save_steps": 500,
414
+ "total_flos": 1433401103358000.0,
415
+ "train_batch_size": 4,
416
+ "trial_name": null,
417
+ "trial_params": null
418
+ }
checkpoint-810/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9f81281de988dbf2616f1c96e725c00f6e4415b289b670a7cb21136eb66e95f
3
+ size 4856
checkpoint-810/vocab.json ADDED
The diff for this file is too large to render. See raw diff