Hugging Face's logo

ManojAlexender
/
Final_layer_LM

Text Classification
Transformers
Safetensors
roberta
Model card Files
xet
Community
ManojAlexender commited on
Commit
d7c40f8
·
verified ·
1 Parent(s): cbde079

Upload folder using huggingface_hub

Browse files
Files changed (38) hide show
  1. checkpoint-1000/config.json +38 -0
  2. checkpoint-1000/model.safetensors +3 -0
  3. checkpoint-1000/optimizer.pt +3 -0
  4. checkpoint-1000/rng_state.pth +3 -0
  5. checkpoint-1000/scheduler.pt +3 -0
  6. checkpoint-1000/trainer_state.json +645 -0
  7. checkpoint-1000/training_args.bin +3 -0
  8. checkpoint-1500/config.json +38 -0
  9. checkpoint-1500/model.safetensors +3 -0
  10. checkpoint-1500/optimizer.pt +3 -0
  11. checkpoint-1500/rng_state.pth +3 -0
  12. checkpoint-1500/scheduler.pt +3 -0
  13. checkpoint-1500/trainer_state.json +957 -0
  14. checkpoint-1500/training_args.bin +3 -0
  15. checkpoint-500/config.json +38 -0
  16. checkpoint-500/model.safetensors +3 -0
  17. checkpoint-500/optimizer.pt +3 -0
  18. checkpoint-500/rng_state.pth +3 -0
  19. checkpoint-500/scheduler.pt +3 -0
  20. checkpoint-500/trainer_state.json +333 -0
  21. checkpoint-500/training_args.bin +3 -0
  22. checkpoint-5500/config.json +38 -0
  23. checkpoint-5500/model.safetensors +3 -0
  24. checkpoint-5500/optimizer.pt +3 -0
  25. checkpoint-5500/rng_state.pth +3 -0
  26. checkpoint-5500/scheduler.pt +3 -0
  27. checkpoint-5500/trainer_state.json +3453 -0
  28. checkpoint-5500/training_args.bin +3 -0
  29. checkpoint-6000/config.json +38 -0
  30. checkpoint-6000/model.safetensors +3 -0
  31. checkpoint-6000/optimizer.pt +3 -0
  32. checkpoint-6000/rng_state.pth +3 -0
  33. checkpoint-6000/scheduler.pt +3 -0
  34. checkpoint-6000/trainer_state.json +3765 -0
  35. checkpoint-6000/training_args.bin +3 -0
  36. config.json +38 -0
  37. model.safetensors +3 -0
  38. training_args.bin +3 -0
checkpoint-1000/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base-finetuned-8M-MLM-500K/",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
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": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "LABEL_0": 0,
22
+ "LABEL_1": 1,
23
+ "LABEL_2": 2
24
+ },
25
+ "layer_norm_eps": 1e-05,
26
+ "max_position_embeddings": 514,
27
+ "model_type": "roberta",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 1,
31
+ "position_embedding_type": "absolute",
32
+ "problem_type": "single_label_classification",
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.37.0",
35
+ "type_vocab_size": 1,
36
+ "use_cache": true,
37
+ "vocab_size": 50265
38
+ }
checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26b148310157f24bf9080b0549d5bec80d7827bfb2c5da5a10f46ca428213f60
3
+ size 498615900
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6d7277e52d048ae0677bb0dc8c30558596ad2b744e294082aca014282de5a26
3
+ size 997351610
checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f60afc0309a2bae6ed3bdc111c999443aa93913f7dc62a094fcae984727fe64a
3
+ size 14244
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4e51d0e89ead22d7ab0bdcb953f900fa848c4855a92f9316a89a491e479239f
3
+ size 1064
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,645 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9186971783638,
3
+ "best_model_checkpoint": "./Finetuned_Classification_model_200k/checkpoint-500",
4
+ "epoch": 0.16,
5
+ "eval_steps": 500,
6
+ "global_step": 1000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 1e-05,
14
+ "loss": 1.037,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 2e-05,
20
+ "loss": 0.907,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 3e-05,
26
+ "loss": 0.7159,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4e-05,
32
+ "loss": 0.5758,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.3454,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.997326203208557e-05,
44
+ "loss": 0.1851,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 4.994652406417113e-05,
50
+ "loss": 0.2821,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 4.9919786096256686e-05,
56
+ "loss": 0.1741,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 4.9893048128342245e-05,
62
+ "loss": 0.1749,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.986631016042781e-05,
68
+ "loss": 0.2407,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 4.983957219251337e-05,
74
+ "loss": 0.2612,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 4.981283422459893e-05,
80
+ "loss": 0.1677,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 4.9786096256684495e-05,
86
+ "loss": 0.1123,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 4.975935828877006e-05,
92
+ "loss": 0.1368,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 4.973262032085561e-05,
98
+ "loss": 0.1449,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.03,
103
+ "learning_rate": 4.970588235294118e-05,
104
+ "loss": 0.152,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.03,
109
+ "learning_rate": 4.967914438502674e-05,
110
+ "loss": 0.2504,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 4.9652406417112304e-05,
116
+ "loss": 0.2578,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 4.962566844919786e-05,
122
+ "loss": 0.1289,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 4.959893048128342e-05,
128
+ "loss": 0.1263,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 4.957219251336899e-05,
134
+ "loss": 0.1031,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.04,
139
+ "learning_rate": 4.9545454545454553e-05,
140
+ "loss": 0.2453,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.04,
145
+ "learning_rate": 4.9518716577540106e-05,
146
+ "loss": 0.0675,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.04,
151
+ "learning_rate": 4.949197860962567e-05,
152
+ "loss": 0.1807,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.04,
157
+ "learning_rate": 4.946524064171123e-05,
158
+ "loss": 0.1493,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.04,
163
+ "learning_rate": 4.9438502673796796e-05,
164
+ "loss": 0.109,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 4.9411764705882355e-05,
170
+ "loss": 0.2148,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 4.9385026737967914e-05,
176
+ "loss": 0.2574,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.05,
181
+ "learning_rate": 4.935828877005348e-05,
182
+ "loss": 0.2198,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.05,
187
+ "learning_rate": 4.933155080213904e-05,
188
+ "loss": 0.1574,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.05,
193
+ "learning_rate": 4.93048128342246e-05,
194
+ "loss": 0.0866,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.05,
199
+ "learning_rate": 4.9278074866310164e-05,
200
+ "loss": 0.1371,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.05,
205
+ "learning_rate": 4.925133689839572e-05,
206
+ "loss": 0.2074,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.05,
211
+ "learning_rate": 4.922459893048129e-05,
212
+ "loss": 0.3628,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.06,
217
+ "learning_rate": 4.919786096256685e-05,
218
+ "loss": 0.1888,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.06,
223
+ "learning_rate": 4.917112299465241e-05,
224
+ "loss": 0.0934,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.06,
229
+ "learning_rate": 4.914438502673797e-05,
230
+ "loss": 0.1593,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.06,
235
+ "learning_rate": 4.911764705882353e-05,
236
+ "loss": 0.0721,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.06,
241
+ "learning_rate": 4.909090909090909e-05,
242
+ "loss": 0.1139,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.06,
247
+ "learning_rate": 4.906417112299466e-05,
248
+ "loss": 0.1118,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.07,
253
+ "learning_rate": 4.9037433155080216e-05,
254
+ "loss": 0.218,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.07,
259
+ "learning_rate": 4.901069518716578e-05,
260
+ "loss": 0.126,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.07,
265
+ "learning_rate": 4.898395721925134e-05,
266
+ "loss": 0.1057,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.07,
271
+ "learning_rate": 4.89572192513369e-05,
272
+ "loss": 0.1825,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.07,
277
+ "learning_rate": 4.8930481283422465e-05,
278
+ "loss": 0.1379,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.07,
283
+ "learning_rate": 4.8903743315508024e-05,
284
+ "loss": 0.1311,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.08,
289
+ "learning_rate": 4.8877005347593584e-05,
290
+ "loss": 0.0722,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.08,
295
+ "learning_rate": 4.885026737967915e-05,
296
+ "loss": 0.1286,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.08,
301
+ "learning_rate": 4.882352941176471e-05,
302
+ "loss": 0.1534,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.08,
307
+ "learning_rate": 4.879679144385027e-05,
308
+ "loss": 0.1332,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.08,
313
+ "eval_accuracy": 0.8395015105740181,
314
+ "eval_f1": 0.8361086180158485,
315
+ "eval_loss": 0.9186971783638,
316
+ "eval_precision": 0.8701532825604937,
317
+ "eval_recall": 0.8395015105740181,
318
+ "eval_runtime": 10.5139,
319
+ "eval_samples_per_second": 251.857,
320
+ "eval_steps_per_second": 3.995,
321
+ "step": 500
322
+ },
323
+ {
324
+ "epoch": 0.08,
325
+ "learning_rate": 4.8770053475935826e-05,
326
+ "loss": 0.1317,
327
+ "step": 510
328
+ },
329
+ {
330
+ "epoch": 0.08,
331
+ "learning_rate": 4.874331550802139e-05,
332
+ "loss": 0.0683,
333
+ "step": 520
334
+ },
335
+ {
336
+ "epoch": 0.08,
337
+ "learning_rate": 4.871657754010696e-05,
338
+ "loss": 0.0474,
339
+ "step": 530
340
+ },
341
+ {
342
+ "epoch": 0.09,
343
+ "learning_rate": 4.868983957219252e-05,
344
+ "loss": 0.2302,
345
+ "step": 540
346
+ },
347
+ {
348
+ "epoch": 0.09,
349
+ "learning_rate": 4.8663101604278076e-05,
350
+ "loss": 0.1788,
351
+ "step": 550
352
+ },
353
+ {
354
+ "epoch": 0.09,
355
+ "learning_rate": 4.863636363636364e-05,
356
+ "loss": 0.0917,
357
+ "step": 560
358
+ },
359
+ {
360
+ "epoch": 0.09,
361
+ "learning_rate": 4.86096256684492e-05,
362
+ "loss": 0.2001,
363
+ "step": 570
364
+ },
365
+ {
366
+ "epoch": 0.09,
367
+ "learning_rate": 4.858288770053476e-05,
368
+ "loss": 0.3409,
369
+ "step": 580
370
+ },
371
+ {
372
+ "epoch": 0.09,
373
+ "learning_rate": 4.855614973262032e-05,
374
+ "loss": 0.2044,
375
+ "step": 590
376
+ },
377
+ {
378
+ "epoch": 0.1,
379
+ "learning_rate": 4.8529411764705885e-05,
380
+ "loss": 0.1292,
381
+ "step": 600
382
+ },
383
+ {
384
+ "epoch": 0.1,
385
+ "learning_rate": 4.850267379679145e-05,
386
+ "loss": 0.1422,
387
+ "step": 610
388
+ },
389
+ {
390
+ "epoch": 0.1,
391
+ "learning_rate": 4.8475935828877e-05,
392
+ "loss": 0.1165,
393
+ "step": 620
394
+ },
395
+ {
396
+ "epoch": 0.1,
397
+ "learning_rate": 4.844919786096257e-05,
398
+ "loss": 0.1115,
399
+ "step": 630
400
+ },
401
+ {
402
+ "epoch": 0.1,
403
+ "learning_rate": 4.8422459893048135e-05,
404
+ "loss": 0.0957,
405
+ "step": 640
406
+ },
407
+ {
408
+ "epoch": 0.1,
409
+ "learning_rate": 4.8395721925133694e-05,
410
+ "loss": 0.1791,
411
+ "step": 650
412
+ },
413
+ {
414
+ "epoch": 0.11,
415
+ "learning_rate": 4.836898395721925e-05,
416
+ "loss": 0.0691,
417
+ "step": 660
418
+ },
419
+ {
420
+ "epoch": 0.11,
421
+ "learning_rate": 4.834224598930481e-05,
422
+ "loss": 0.2829,
423
+ "step": 670
424
+ },
425
+ {
426
+ "epoch": 0.11,
427
+ "learning_rate": 4.831550802139038e-05,
428
+ "loss": 0.2635,
429
+ "step": 680
430
+ },
431
+ {
432
+ "epoch": 0.11,
433
+ "learning_rate": 4.828877005347594e-05,
434
+ "loss": 0.1722,
435
+ "step": 690
436
+ },
437
+ {
438
+ "epoch": 0.11,
439
+ "learning_rate": 4.8262032085561496e-05,
440
+ "loss": 0.0974,
441
+ "step": 700
442
+ },
443
+ {
444
+ "epoch": 0.11,
445
+ "learning_rate": 4.823529411764706e-05,
446
+ "loss": 0.0888,
447
+ "step": 710
448
+ },
449
+ {
450
+ "epoch": 0.12,
451
+ "learning_rate": 4.820855614973262e-05,
452
+ "loss": 0.3148,
453
+ "step": 720
454
+ },
455
+ {
456
+ "epoch": 0.12,
457
+ "learning_rate": 4.8181818181818186e-05,
458
+ "loss": 0.3556,
459
+ "step": 730
460
+ },
461
+ {
462
+ "epoch": 0.12,
463
+ "learning_rate": 4.8155080213903745e-05,
464
+ "loss": 0.1788,
465
+ "step": 740
466
+ },
467
+ {
468
+ "epoch": 0.12,
469
+ "learning_rate": 4.8128342245989304e-05,
470
+ "loss": 0.4376,
471
+ "step": 750
472
+ },
473
+ {
474
+ "epoch": 0.12,
475
+ "learning_rate": 4.810160427807487e-05,
476
+ "loss": 0.2413,
477
+ "step": 760
478
+ },
479
+ {
480
+ "epoch": 0.12,
481
+ "learning_rate": 4.807486631016043e-05,
482
+ "loss": 0.1802,
483
+ "step": 770
484
+ },
485
+ {
486
+ "epoch": 0.12,
487
+ "learning_rate": 4.804812834224599e-05,
488
+ "loss": 0.1364,
489
+ "step": 780
490
+ },
491
+ {
492
+ "epoch": 0.13,
493
+ "learning_rate": 4.8021390374331554e-05,
494
+ "loss": 0.0667,
495
+ "step": 790
496
+ },
497
+ {
498
+ "epoch": 0.13,
499
+ "learning_rate": 4.799465240641711e-05,
500
+ "loss": 0.1409,
501
+ "step": 800
502
+ },
503
+ {
504
+ "epoch": 0.13,
505
+ "learning_rate": 4.796791443850268e-05,
506
+ "loss": 0.0897,
507
+ "step": 810
508
+ },
509
+ {
510
+ "epoch": 0.13,
511
+ "learning_rate": 4.794117647058824e-05,
512
+ "loss": 0.0871,
513
+ "step": 820
514
+ },
515
+ {
516
+ "epoch": 0.13,
517
+ "learning_rate": 4.79144385026738e-05,
518
+ "loss": 0.0552,
519
+ "step": 830
520
+ },
521
+ {
522
+ "epoch": 0.13,
523
+ "learning_rate": 4.788770053475936e-05,
524
+ "loss": 0.2569,
525
+ "step": 840
526
+ },
527
+ {
528
+ "epoch": 0.14,
529
+ "learning_rate": 4.786096256684492e-05,
530
+ "loss": 0.2877,
531
+ "step": 850
532
+ },
533
+ {
534
+ "epoch": 0.14,
535
+ "learning_rate": 4.783422459893048e-05,
536
+ "loss": 0.1365,
537
+ "step": 860
538
+ },
539
+ {
540
+ "epoch": 0.14,
541
+ "learning_rate": 4.7807486631016047e-05,
542
+ "loss": 0.1604,
543
+ "step": 870
544
+ },
545
+ {
546
+ "epoch": 0.14,
547
+ "learning_rate": 4.7780748663101606e-05,
548
+ "loss": 0.2179,
549
+ "step": 880
550
+ },
551
+ {
552
+ "epoch": 0.14,
553
+ "learning_rate": 4.7754010695187165e-05,
554
+ "loss": 0.0969,
555
+ "step": 890
556
+ },
557
+ {
558
+ "epoch": 0.14,
559
+ "learning_rate": 4.772727272727273e-05,
560
+ "loss": 0.0953,
561
+ "step": 900
562
+ },
563
+ {
564
+ "epoch": 0.15,
565
+ "learning_rate": 4.770053475935829e-05,
566
+ "loss": 0.0803,
567
+ "step": 910
568
+ },
569
+ {
570
+ "epoch": 0.15,
571
+ "learning_rate": 4.7673796791443855e-05,
572
+ "loss": 0.1573,
573
+ "step": 920
574
+ },
575
+ {
576
+ "epoch": 0.15,
577
+ "learning_rate": 4.7647058823529414e-05,
578
+ "loss": 0.1388,
579
+ "step": 930
580
+ },
581
+ {
582
+ "epoch": 0.15,
583
+ "learning_rate": 4.7620320855614973e-05,
584
+ "loss": 0.1559,
585
+ "step": 940
586
+ },
587
+ {
588
+ "epoch": 0.15,
589
+ "learning_rate": 4.759358288770054e-05,
590
+ "loss": 0.2411,
591
+ "step": 950
592
+ },
593
+ {
594
+ "epoch": 0.15,
595
+ "learning_rate": 4.75668449197861e-05,
596
+ "loss": 0.1119,
597
+ "step": 960
598
+ },
599
+ {
600
+ "epoch": 0.16,
601
+ "learning_rate": 4.754010695187166e-05,
602
+ "loss": 0.1416,
603
+ "step": 970
604
+ },
605
+ {
606
+ "epoch": 0.16,
607
+ "learning_rate": 4.751336898395722e-05,
608
+ "loss": 0.1073,
609
+ "step": 980
610
+ },
611
+ {
612
+ "epoch": 0.16,
613
+ "learning_rate": 4.748663101604278e-05,
614
+ "loss": 0.1799,
615
+ "step": 990
616
+ },
617
+ {
618
+ "epoch": 0.16,
619
+ "learning_rate": 4.745989304812835e-05,
620
+ "loss": 0.127,
621
+ "step": 1000
622
+ },
623
+ {
624
+ "epoch": 0.16,
625
+ "eval_accuracy": 0.847809667673716,
626
+ "eval_f1": 0.8460260531471735,
627
+ "eval_loss": 1.017073392868042,
628
+ "eval_precision": 0.8647086198841812,
629
+ "eval_recall": 0.847809667673716,
630
+ "eval_runtime": 10.5379,
631
+ "eval_samples_per_second": 251.284,
632
+ "eval_steps_per_second": 3.986,
633
+ "step": 1000
634
+ }
635
+ ],
636
+ "logging_steps": 10,
637
+ "max_steps": 18750,
638
+ "num_input_tokens_seen": 0,
639
+ "num_train_epochs": 3,
640
+ "save_steps": 500,
641
+ "total_flos": 8419629367296000.0,
642
+ "train_batch_size": 16,
643
+ "trial_name": null,
644
+ "trial_params": null
645
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ea61cfa8a59fe23f079dded341db3c34a77c19986a2b3885296af3c7226c95
3
+ size 4664
checkpoint-1500/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base-finetuned-8M-MLM-500K/",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
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": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "LABEL_0": 0,
22
+ "LABEL_1": 1,
23
+ "LABEL_2": 2
24
+ },
25
+ "layer_norm_eps": 1e-05,
26
+ "max_position_embeddings": 514,
27
+ "model_type": "roberta",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 1,
31
+ "position_embedding_type": "absolute",
32
+ "problem_type": "single_label_classification",
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.37.0",
35
+ "type_vocab_size": 1,
36
+ "use_cache": true,
37
+ "vocab_size": 50265
38
+ }
checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:927dea851bffd43fca2b975d3c66517f985016930c2a473fa192fb3f8d53d9b1
3
+ size 498615900
checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:848678280e2d5f906bd6add4d8d6aa552ff212b75c2d94a4a3fb0935a2dd143b
3
+ size 997351610
checkpoint-1500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab69a922df6e90ce8de4fa69bd1800deae1772daffd7e006f9395bdd00797153
3
+ size 14244
checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b65f70c38e7c6c7dfbd4b6e2d83ca9d022dc9b106d8d58369ba8118bd51685f
3
+ size 1064
checkpoint-1500/trainer_state.json ADDED
@@ -0,0 +1,957 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9186971783638,
3
+ "best_model_checkpoint": "./Finetuned_Classification_model_200k/checkpoint-500",
4
+ "epoch": 0.24,
5
+ "eval_steps": 500,
6
+ "global_step": 1500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 1e-05,
14
+ "loss": 1.037,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 2e-05,
20
+ "loss": 0.907,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 3e-05,
26
+ "loss": 0.7159,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4e-05,
32
+ "loss": 0.5758,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.3454,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.997326203208557e-05,
44
+ "loss": 0.1851,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 4.994652406417113e-05,
50
+ "loss": 0.2821,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 4.9919786096256686e-05,
56
+ "loss": 0.1741,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 4.9893048128342245e-05,
62
+ "loss": 0.1749,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.986631016042781e-05,
68
+ "loss": 0.2407,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 4.983957219251337e-05,
74
+ "loss": 0.2612,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 4.981283422459893e-05,
80
+ "loss": 0.1677,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 4.9786096256684495e-05,
86
+ "loss": 0.1123,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 4.975935828877006e-05,
92
+ "loss": 0.1368,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 4.973262032085561e-05,
98
+ "loss": 0.1449,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.03,
103
+ "learning_rate": 4.970588235294118e-05,
104
+ "loss": 0.152,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.03,
109
+ "learning_rate": 4.967914438502674e-05,
110
+ "loss": 0.2504,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 4.9652406417112304e-05,
116
+ "loss": 0.2578,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 4.962566844919786e-05,
122
+ "loss": 0.1289,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 4.959893048128342e-05,
128
+ "loss": 0.1263,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 4.957219251336899e-05,
134
+ "loss": 0.1031,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.04,
139
+ "learning_rate": 4.9545454545454553e-05,
140
+ "loss": 0.2453,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.04,
145
+ "learning_rate": 4.9518716577540106e-05,
146
+ "loss": 0.0675,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.04,
151
+ "learning_rate": 4.949197860962567e-05,
152
+ "loss": 0.1807,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.04,
157
+ "learning_rate": 4.946524064171123e-05,
158
+ "loss": 0.1493,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.04,
163
+ "learning_rate": 4.9438502673796796e-05,
164
+ "loss": 0.109,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 4.9411764705882355e-05,
170
+ "loss": 0.2148,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 4.9385026737967914e-05,
176
+ "loss": 0.2574,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.05,
181
+ "learning_rate": 4.935828877005348e-05,
182
+ "loss": 0.2198,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.05,
187
+ "learning_rate": 4.933155080213904e-05,
188
+ "loss": 0.1574,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.05,
193
+ "learning_rate": 4.93048128342246e-05,
194
+ "loss": 0.0866,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.05,
199
+ "learning_rate": 4.9278074866310164e-05,
200
+ "loss": 0.1371,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.05,
205
+ "learning_rate": 4.925133689839572e-05,
206
+ "loss": 0.2074,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.05,
211
+ "learning_rate": 4.922459893048129e-05,
212
+ "loss": 0.3628,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.06,
217
+ "learning_rate": 4.919786096256685e-05,
218
+ "loss": 0.1888,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.06,
223
+ "learning_rate": 4.917112299465241e-05,
224
+ "loss": 0.0934,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.06,
229
+ "learning_rate": 4.914438502673797e-05,
230
+ "loss": 0.1593,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.06,
235
+ "learning_rate": 4.911764705882353e-05,
236
+ "loss": 0.0721,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.06,
241
+ "learning_rate": 4.909090909090909e-05,
242
+ "loss": 0.1139,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.06,
247
+ "learning_rate": 4.906417112299466e-05,
248
+ "loss": 0.1118,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.07,
253
+ "learning_rate": 4.9037433155080216e-05,
254
+ "loss": 0.218,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.07,
259
+ "learning_rate": 4.901069518716578e-05,
260
+ "loss": 0.126,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.07,
265
+ "learning_rate": 4.898395721925134e-05,
266
+ "loss": 0.1057,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.07,
271
+ "learning_rate": 4.89572192513369e-05,
272
+ "loss": 0.1825,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.07,
277
+ "learning_rate": 4.8930481283422465e-05,
278
+ "loss": 0.1379,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.07,
283
+ "learning_rate": 4.8903743315508024e-05,
284
+ "loss": 0.1311,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.08,
289
+ "learning_rate": 4.8877005347593584e-05,
290
+ "loss": 0.0722,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.08,
295
+ "learning_rate": 4.885026737967915e-05,
296
+ "loss": 0.1286,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.08,
301
+ "learning_rate": 4.882352941176471e-05,
302
+ "loss": 0.1534,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.08,
307
+ "learning_rate": 4.879679144385027e-05,
308
+ "loss": 0.1332,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.08,
313
+ "eval_accuracy": 0.8395015105740181,
314
+ "eval_f1": 0.8361086180158485,
315
+ "eval_loss": 0.9186971783638,
316
+ "eval_precision": 0.8701532825604937,
317
+ "eval_recall": 0.8395015105740181,
318
+ "eval_runtime": 10.5139,
319
+ "eval_samples_per_second": 251.857,
320
+ "eval_steps_per_second": 3.995,
321
+ "step": 500
322
+ },
323
+ {
324
+ "epoch": 0.08,
325
+ "learning_rate": 4.8770053475935826e-05,
326
+ "loss": 0.1317,
327
+ "step": 510
328
+ },
329
+ {
330
+ "epoch": 0.08,
331
+ "learning_rate": 4.874331550802139e-05,
332
+ "loss": 0.0683,
333
+ "step": 520
334
+ },
335
+ {
336
+ "epoch": 0.08,
337
+ "learning_rate": 4.871657754010696e-05,
338
+ "loss": 0.0474,
339
+ "step": 530
340
+ },
341
+ {
342
+ "epoch": 0.09,
343
+ "learning_rate": 4.868983957219252e-05,
344
+ "loss": 0.2302,
345
+ "step": 540
346
+ },
347
+ {
348
+ "epoch": 0.09,
349
+ "learning_rate": 4.8663101604278076e-05,
350
+ "loss": 0.1788,
351
+ "step": 550
352
+ },
353
+ {
354
+ "epoch": 0.09,
355
+ "learning_rate": 4.863636363636364e-05,
356
+ "loss": 0.0917,
357
+ "step": 560
358
+ },
359
+ {
360
+ "epoch": 0.09,
361
+ "learning_rate": 4.86096256684492e-05,
362
+ "loss": 0.2001,
363
+ "step": 570
364
+ },
365
+ {
366
+ "epoch": 0.09,
367
+ "learning_rate": 4.858288770053476e-05,
368
+ "loss": 0.3409,
369
+ "step": 580
370
+ },
371
+ {
372
+ "epoch": 0.09,
373
+ "learning_rate": 4.855614973262032e-05,
374
+ "loss": 0.2044,
375
+ "step": 590
376
+ },
377
+ {
378
+ "epoch": 0.1,
379
+ "learning_rate": 4.8529411764705885e-05,
380
+ "loss": 0.1292,
381
+ "step": 600
382
+ },
383
+ {
384
+ "epoch": 0.1,
385
+ "learning_rate": 4.850267379679145e-05,
386
+ "loss": 0.1422,
387
+ "step": 610
388
+ },
389
+ {
390
+ "epoch": 0.1,
391
+ "learning_rate": 4.8475935828877e-05,
392
+ "loss": 0.1165,
393
+ "step": 620
394
+ },
395
+ {
396
+ "epoch": 0.1,
397
+ "learning_rate": 4.844919786096257e-05,
398
+ "loss": 0.1115,
399
+ "step": 630
400
+ },
401
+ {
402
+ "epoch": 0.1,
403
+ "learning_rate": 4.8422459893048135e-05,
404
+ "loss": 0.0957,
405
+ "step": 640
406
+ },
407
+ {
408
+ "epoch": 0.1,
409
+ "learning_rate": 4.8395721925133694e-05,
410
+ "loss": 0.1791,
411
+ "step": 650
412
+ },
413
+ {
414
+ "epoch": 0.11,
415
+ "learning_rate": 4.836898395721925e-05,
416
+ "loss": 0.0691,
417
+ "step": 660
418
+ },
419
+ {
420
+ "epoch": 0.11,
421
+ "learning_rate": 4.834224598930481e-05,
422
+ "loss": 0.2829,
423
+ "step": 670
424
+ },
425
+ {
426
+ "epoch": 0.11,
427
+ "learning_rate": 4.831550802139038e-05,
428
+ "loss": 0.2635,
429
+ "step": 680
430
+ },
431
+ {
432
+ "epoch": 0.11,
433
+ "learning_rate": 4.828877005347594e-05,
434
+ "loss": 0.1722,
435
+ "step": 690
436
+ },
437
+ {
438
+ "epoch": 0.11,
439
+ "learning_rate": 4.8262032085561496e-05,
440
+ "loss": 0.0974,
441
+ "step": 700
442
+ },
443
+ {
444
+ "epoch": 0.11,
445
+ "learning_rate": 4.823529411764706e-05,
446
+ "loss": 0.0888,
447
+ "step": 710
448
+ },
449
+ {
450
+ "epoch": 0.12,
451
+ "learning_rate": 4.820855614973262e-05,
452
+ "loss": 0.3148,
453
+ "step": 720
454
+ },
455
+ {
456
+ "epoch": 0.12,
457
+ "learning_rate": 4.8181818181818186e-05,
458
+ "loss": 0.3556,
459
+ "step": 730
460
+ },
461
+ {
462
+ "epoch": 0.12,
463
+ "learning_rate": 4.8155080213903745e-05,
464
+ "loss": 0.1788,
465
+ "step": 740
466
+ },
467
+ {
468
+ "epoch": 0.12,
469
+ "learning_rate": 4.8128342245989304e-05,
470
+ "loss": 0.4376,
471
+ "step": 750
472
+ },
473
+ {
474
+ "epoch": 0.12,
475
+ "learning_rate": 4.810160427807487e-05,
476
+ "loss": 0.2413,
477
+ "step": 760
478
+ },
479
+ {
480
+ "epoch": 0.12,
481
+ "learning_rate": 4.807486631016043e-05,
482
+ "loss": 0.1802,
483
+ "step": 770
484
+ },
485
+ {
486
+ "epoch": 0.12,
487
+ "learning_rate": 4.804812834224599e-05,
488
+ "loss": 0.1364,
489
+ "step": 780
490
+ },
491
+ {
492
+ "epoch": 0.13,
493
+ "learning_rate": 4.8021390374331554e-05,
494
+ "loss": 0.0667,
495
+ "step": 790
496
+ },
497
+ {
498
+ "epoch": 0.13,
499
+ "learning_rate": 4.799465240641711e-05,
500
+ "loss": 0.1409,
501
+ "step": 800
502
+ },
503
+ {
504
+ "epoch": 0.13,
505
+ "learning_rate": 4.796791443850268e-05,
506
+ "loss": 0.0897,
507
+ "step": 810
508
+ },
509
+ {
510
+ "epoch": 0.13,
511
+ "learning_rate": 4.794117647058824e-05,
512
+ "loss": 0.0871,
513
+ "step": 820
514
+ },
515
+ {
516
+ "epoch": 0.13,
517
+ "learning_rate": 4.79144385026738e-05,
518
+ "loss": 0.0552,
519
+ "step": 830
520
+ },
521
+ {
522
+ "epoch": 0.13,
523
+ "learning_rate": 4.788770053475936e-05,
524
+ "loss": 0.2569,
525
+ "step": 840
526
+ },
527
+ {
528
+ "epoch": 0.14,
529
+ "learning_rate": 4.786096256684492e-05,
530
+ "loss": 0.2877,
531
+ "step": 850
532
+ },
533
+ {
534
+ "epoch": 0.14,
535
+ "learning_rate": 4.783422459893048e-05,
536
+ "loss": 0.1365,
537
+ "step": 860
538
+ },
539
+ {
540
+ "epoch": 0.14,
541
+ "learning_rate": 4.7807486631016047e-05,
542
+ "loss": 0.1604,
543
+ "step": 870
544
+ },
545
+ {
546
+ "epoch": 0.14,
547
+ "learning_rate": 4.7780748663101606e-05,
548
+ "loss": 0.2179,
549
+ "step": 880
550
+ },
551
+ {
552
+ "epoch": 0.14,
553
+ "learning_rate": 4.7754010695187165e-05,
554
+ "loss": 0.0969,
555
+ "step": 890
556
+ },
557
+ {
558
+ "epoch": 0.14,
559
+ "learning_rate": 4.772727272727273e-05,
560
+ "loss": 0.0953,
561
+ "step": 900
562
+ },
563
+ {
564
+ "epoch": 0.15,
565
+ "learning_rate": 4.770053475935829e-05,
566
+ "loss": 0.0803,
567
+ "step": 910
568
+ },
569
+ {
570
+ "epoch": 0.15,
571
+ "learning_rate": 4.7673796791443855e-05,
572
+ "loss": 0.1573,
573
+ "step": 920
574
+ },
575
+ {
576
+ "epoch": 0.15,
577
+ "learning_rate": 4.7647058823529414e-05,
578
+ "loss": 0.1388,
579
+ "step": 930
580
+ },
581
+ {
582
+ "epoch": 0.15,
583
+ "learning_rate": 4.7620320855614973e-05,
584
+ "loss": 0.1559,
585
+ "step": 940
586
+ },
587
+ {
588
+ "epoch": 0.15,
589
+ "learning_rate": 4.759358288770054e-05,
590
+ "loss": 0.2411,
591
+ "step": 950
592
+ },
593
+ {
594
+ "epoch": 0.15,
595
+ "learning_rate": 4.75668449197861e-05,
596
+ "loss": 0.1119,
597
+ "step": 960
598
+ },
599
+ {
600
+ "epoch": 0.16,
601
+ "learning_rate": 4.754010695187166e-05,
602
+ "loss": 0.1416,
603
+ "step": 970
604
+ },
605
+ {
606
+ "epoch": 0.16,
607
+ "learning_rate": 4.751336898395722e-05,
608
+ "loss": 0.1073,
609
+ "step": 980
610
+ },
611
+ {
612
+ "epoch": 0.16,
613
+ "learning_rate": 4.748663101604278e-05,
614
+ "loss": 0.1799,
615
+ "step": 990
616
+ },
617
+ {
618
+ "epoch": 0.16,
619
+ "learning_rate": 4.745989304812835e-05,
620
+ "loss": 0.127,
621
+ "step": 1000
622
+ },
623
+ {
624
+ "epoch": 0.16,
625
+ "eval_accuracy": 0.847809667673716,
626
+ "eval_f1": 0.8460260531471735,
627
+ "eval_loss": 1.017073392868042,
628
+ "eval_precision": 0.8647086198841812,
629
+ "eval_recall": 0.847809667673716,
630
+ "eval_runtime": 10.5379,
631
+ "eval_samples_per_second": 251.284,
632
+ "eval_steps_per_second": 3.986,
633
+ "step": 1000
634
+ },
635
+ {
636
+ "epoch": 0.16,
637
+ "learning_rate": 4.74331550802139e-05,
638
+ "loss": 0.1774,
639
+ "step": 1010
640
+ },
641
+ {
642
+ "epoch": 0.16,
643
+ "learning_rate": 4.7406417112299466e-05,
644
+ "loss": 0.0869,
645
+ "step": 1020
646
+ },
647
+ {
648
+ "epoch": 0.16,
649
+ "learning_rate": 4.737967914438503e-05,
650
+ "loss": 0.0686,
651
+ "step": 1030
652
+ },
653
+ {
654
+ "epoch": 0.17,
655
+ "learning_rate": 4.735294117647059e-05,
656
+ "loss": 0.124,
657
+ "step": 1040
658
+ },
659
+ {
660
+ "epoch": 0.17,
661
+ "learning_rate": 4.732620320855615e-05,
662
+ "loss": 0.0993,
663
+ "step": 1050
664
+ },
665
+ {
666
+ "epoch": 0.17,
667
+ "learning_rate": 4.7299465240641716e-05,
668
+ "loss": 0.2451,
669
+ "step": 1060
670
+ },
671
+ {
672
+ "epoch": 0.17,
673
+ "learning_rate": 4.7272727272727275e-05,
674
+ "loss": 0.1607,
675
+ "step": 1070
676
+ },
677
+ {
678
+ "epoch": 0.17,
679
+ "learning_rate": 4.724598930481284e-05,
680
+ "loss": 0.1289,
681
+ "step": 1080
682
+ },
683
+ {
684
+ "epoch": 0.17,
685
+ "learning_rate": 4.721925133689839e-05,
686
+ "loss": 0.1813,
687
+ "step": 1090
688
+ },
689
+ {
690
+ "epoch": 0.18,
691
+ "learning_rate": 4.719251336898396e-05,
692
+ "loss": 0.1628,
693
+ "step": 1100
694
+ },
695
+ {
696
+ "epoch": 0.18,
697
+ "learning_rate": 4.7165775401069524e-05,
698
+ "loss": 0.3152,
699
+ "step": 1110
700
+ },
701
+ {
702
+ "epoch": 0.18,
703
+ "learning_rate": 4.7139037433155083e-05,
704
+ "loss": 0.3835,
705
+ "step": 1120
706
+ },
707
+ {
708
+ "epoch": 0.18,
709
+ "learning_rate": 4.711229946524064e-05,
710
+ "loss": 0.1713,
711
+ "step": 1130
712
+ },
713
+ {
714
+ "epoch": 0.18,
715
+ "learning_rate": 4.70855614973262e-05,
716
+ "loss": 0.1812,
717
+ "step": 1140
718
+ },
719
+ {
720
+ "epoch": 0.18,
721
+ "learning_rate": 4.705882352941177e-05,
722
+ "loss": 0.1928,
723
+ "step": 1150
724
+ },
725
+ {
726
+ "epoch": 0.19,
727
+ "learning_rate": 4.703208556149733e-05,
728
+ "loss": 0.2963,
729
+ "step": 1160
730
+ },
731
+ {
732
+ "epoch": 0.19,
733
+ "learning_rate": 4.7005347593582885e-05,
734
+ "loss": 0.2073,
735
+ "step": 1170
736
+ },
737
+ {
738
+ "epoch": 0.19,
739
+ "learning_rate": 4.697860962566845e-05,
740
+ "loss": 0.2299,
741
+ "step": 1180
742
+ },
743
+ {
744
+ "epoch": 0.19,
745
+ "learning_rate": 4.695187165775402e-05,
746
+ "loss": 0.1832,
747
+ "step": 1190
748
+ },
749
+ {
750
+ "epoch": 0.19,
751
+ "learning_rate": 4.6925133689839576e-05,
752
+ "loss": 0.1844,
753
+ "step": 1200
754
+ },
755
+ {
756
+ "epoch": 0.19,
757
+ "learning_rate": 4.6898395721925135e-05,
758
+ "loss": 0.4959,
759
+ "step": 1210
760
+ },
761
+ {
762
+ "epoch": 0.2,
763
+ "learning_rate": 4.6871657754010694e-05,
764
+ "loss": 0.2516,
765
+ "step": 1220
766
+ },
767
+ {
768
+ "epoch": 0.2,
769
+ "learning_rate": 4.684491978609626e-05,
770
+ "loss": 0.1754,
771
+ "step": 1230
772
+ },
773
+ {
774
+ "epoch": 0.2,
775
+ "learning_rate": 4.681818181818182e-05,
776
+ "loss": 0.2554,
777
+ "step": 1240
778
+ },
779
+ {
780
+ "epoch": 0.2,
781
+ "learning_rate": 4.679144385026738e-05,
782
+ "loss": 0.1976,
783
+ "step": 1250
784
+ },
785
+ {
786
+ "epoch": 0.2,
787
+ "learning_rate": 4.6764705882352944e-05,
788
+ "loss": 0.1906,
789
+ "step": 1260
790
+ },
791
+ {
792
+ "epoch": 0.2,
793
+ "learning_rate": 4.673796791443851e-05,
794
+ "loss": 0.0806,
795
+ "step": 1270
796
+ },
797
+ {
798
+ "epoch": 0.2,
799
+ "learning_rate": 4.671122994652407e-05,
800
+ "loss": 0.1229,
801
+ "step": 1280
802
+ },
803
+ {
804
+ "epoch": 0.21,
805
+ "learning_rate": 4.668449197860963e-05,
806
+ "loss": 0.1222,
807
+ "step": 1290
808
+ },
809
+ {
810
+ "epoch": 0.21,
811
+ "learning_rate": 4.665775401069519e-05,
812
+ "loss": 0.1049,
813
+ "step": 1300
814
+ },
815
+ {
816
+ "epoch": 0.21,
817
+ "learning_rate": 4.663101604278075e-05,
818
+ "loss": 0.1548,
819
+ "step": 1310
820
+ },
821
+ {
822
+ "epoch": 0.21,
823
+ "learning_rate": 4.660427807486631e-05,
824
+ "loss": 0.1077,
825
+ "step": 1320
826
+ },
827
+ {
828
+ "epoch": 0.21,
829
+ "learning_rate": 4.657754010695187e-05,
830
+ "loss": 0.1949,
831
+ "step": 1330
832
+ },
833
+ {
834
+ "epoch": 0.21,
835
+ "learning_rate": 4.6550802139037436e-05,
836
+ "loss": 0.1349,
837
+ "step": 1340
838
+ },
839
+ {
840
+ "epoch": 0.22,
841
+ "learning_rate": 4.6524064171123e-05,
842
+ "loss": 0.223,
843
+ "step": 1350
844
+ },
845
+ {
846
+ "epoch": 0.22,
847
+ "learning_rate": 4.6497326203208555e-05,
848
+ "loss": 0.1132,
849
+ "step": 1360
850
+ },
851
+ {
852
+ "epoch": 0.22,
853
+ "learning_rate": 4.647058823529412e-05,
854
+ "loss": 0.123,
855
+ "step": 1370
856
+ },
857
+ {
858
+ "epoch": 0.22,
859
+ "learning_rate": 4.644385026737968e-05,
860
+ "loss": 0.1741,
861
+ "step": 1380
862
+ },
863
+ {
864
+ "epoch": 0.22,
865
+ "learning_rate": 4.6417112299465245e-05,
866
+ "loss": 0.0913,
867
+ "step": 1390
868
+ },
869
+ {
870
+ "epoch": 0.22,
871
+ "learning_rate": 4.6390374331550804e-05,
872
+ "loss": 0.1428,
873
+ "step": 1400
874
+ },
875
+ {
876
+ "epoch": 0.23,
877
+ "learning_rate": 4.636363636363636e-05,
878
+ "loss": 0.1171,
879
+ "step": 1410
880
+ },
881
+ {
882
+ "epoch": 0.23,
883
+ "learning_rate": 4.633689839572193e-05,
884
+ "loss": 0.1773,
885
+ "step": 1420
886
+ },
887
+ {
888
+ "epoch": 0.23,
889
+ "learning_rate": 4.631016042780749e-05,
890
+ "loss": 0.1372,
891
+ "step": 1430
892
+ },
893
+ {
894
+ "epoch": 0.23,
895
+ "learning_rate": 4.628342245989305e-05,
896
+ "loss": 0.2797,
897
+ "step": 1440
898
+ },
899
+ {
900
+ "epoch": 0.23,
901
+ "learning_rate": 4.625668449197861e-05,
902
+ "loss": 0.2813,
903
+ "step": 1450
904
+ },
905
+ {
906
+ "epoch": 0.23,
907
+ "learning_rate": 4.622994652406417e-05,
908
+ "loss": 0.2861,
909
+ "step": 1460
910
+ },
911
+ {
912
+ "epoch": 0.24,
913
+ "learning_rate": 4.620320855614974e-05,
914
+ "loss": 0.0677,
915
+ "step": 1470
916
+ },
917
+ {
918
+ "epoch": 0.24,
919
+ "learning_rate": 4.61764705882353e-05,
920
+ "loss": 0.104,
921
+ "step": 1480
922
+ },
923
+ {
924
+ "epoch": 0.24,
925
+ "learning_rate": 4.6149732620320856e-05,
926
+ "loss": 0.1533,
927
+ "step": 1490
928
+ },
929
+ {
930
+ "epoch": 0.24,
931
+ "learning_rate": 4.612299465240642e-05,
932
+ "loss": 0.1186,
933
+ "step": 1500
934
+ },
935
+ {
936
+ "epoch": 0.24,
937
+ "eval_accuracy": 0.8376132930513596,
938
+ "eval_f1": 0.8363352339205664,
939
+ "eval_loss": 1.0860812664031982,
940
+ "eval_precision": 0.8484990166173995,
941
+ "eval_recall": 0.8376132930513596,
942
+ "eval_runtime": 10.5226,
943
+ "eval_samples_per_second": 251.648,
944
+ "eval_steps_per_second": 3.991,
945
+ "step": 1500
946
+ }
947
+ ],
948
+ "logging_steps": 10,
949
+ "max_steps": 18750,
950
+ "num_input_tokens_seen": 0,
951
+ "num_train_epochs": 3,
952
+ "save_steps": 500,
953
+ "total_flos": 1.2629444050944e+16,
954
+ "train_batch_size": 16,
955
+ "trial_name": null,
956
+ "trial_params": null
957
+ }
checkpoint-1500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ea61cfa8a59fe23f079dded341db3c34a77c19986a2b3885296af3c7226c95
3
+ size 4664
checkpoint-500/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base-finetuned-8M-MLM-500K/",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
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": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "LABEL_0": 0,
22
+ "LABEL_1": 1,
23
+ "LABEL_2": 2
24
+ },
25
+ "layer_norm_eps": 1e-05,
26
+ "max_position_embeddings": 514,
27
+ "model_type": "roberta",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 1,
31
+ "position_embedding_type": "absolute",
32
+ "problem_type": "single_label_classification",
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.37.0",
35
+ "type_vocab_size": 1,
36
+ "use_cache": true,
37
+ "vocab_size": 50265
38
+ }
checkpoint-500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1881c8805f63e122ad3cce42ddbfd8665810c0f7a369d3ec30445d5336712e0f
3
+ size 498615900
checkpoint-500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98551cd9ea2cccd8e7a9d18dbd32a2e9731ab6a0db593f42ffc8e2b79e02cead
3
+ size 997351610
checkpoint-500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbb2f160b2f27fe124e3876ff8b3640607b085b9d8395054b7d815186c5b8d27
3
+ size 14244
checkpoint-500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0f9f0a54ec1655fdfd1d707ad20becc7b52483773204d0fc4d1d01359a3fb1e
3
+ size 1064
checkpoint-500/trainer_state.json ADDED
@@ -0,0 +1,333 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9186971783638,
3
+ "best_model_checkpoint": "./Finetuned_Classification_model_200k/checkpoint-500",
4
+ "epoch": 0.08,
5
+ "eval_steps": 500,
6
+ "global_step": 500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 1e-05,
14
+ "loss": 1.037,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 2e-05,
20
+ "loss": 0.907,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 3e-05,
26
+ "loss": 0.7159,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4e-05,
32
+ "loss": 0.5758,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.3454,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.997326203208557e-05,
44
+ "loss": 0.1851,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 4.994652406417113e-05,
50
+ "loss": 0.2821,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 4.9919786096256686e-05,
56
+ "loss": 0.1741,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 4.9893048128342245e-05,
62
+ "loss": 0.1749,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.986631016042781e-05,
68
+ "loss": 0.2407,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 4.983957219251337e-05,
74
+ "loss": 0.2612,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 4.981283422459893e-05,
80
+ "loss": 0.1677,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 4.9786096256684495e-05,
86
+ "loss": 0.1123,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 4.975935828877006e-05,
92
+ "loss": 0.1368,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 4.973262032085561e-05,
98
+ "loss": 0.1449,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.03,
103
+ "learning_rate": 4.970588235294118e-05,
104
+ "loss": 0.152,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.03,
109
+ "learning_rate": 4.967914438502674e-05,
110
+ "loss": 0.2504,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 4.9652406417112304e-05,
116
+ "loss": 0.2578,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 4.962566844919786e-05,
122
+ "loss": 0.1289,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 4.959893048128342e-05,
128
+ "loss": 0.1263,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 4.957219251336899e-05,
134
+ "loss": 0.1031,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.04,
139
+ "learning_rate": 4.9545454545454553e-05,
140
+ "loss": 0.2453,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.04,
145
+ "learning_rate": 4.9518716577540106e-05,
146
+ "loss": 0.0675,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.04,
151
+ "learning_rate": 4.949197860962567e-05,
152
+ "loss": 0.1807,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.04,
157
+ "learning_rate": 4.946524064171123e-05,
158
+ "loss": 0.1493,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.04,
163
+ "learning_rate": 4.9438502673796796e-05,
164
+ "loss": 0.109,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 4.9411764705882355e-05,
170
+ "loss": 0.2148,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 4.9385026737967914e-05,
176
+ "loss": 0.2574,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.05,
181
+ "learning_rate": 4.935828877005348e-05,
182
+ "loss": 0.2198,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.05,
187
+ "learning_rate": 4.933155080213904e-05,
188
+ "loss": 0.1574,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.05,
193
+ "learning_rate": 4.93048128342246e-05,
194
+ "loss": 0.0866,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.05,
199
+ "learning_rate": 4.9278074866310164e-05,
200
+ "loss": 0.1371,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.05,
205
+ "learning_rate": 4.925133689839572e-05,
206
+ "loss": 0.2074,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.05,
211
+ "learning_rate": 4.922459893048129e-05,
212
+ "loss": 0.3628,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.06,
217
+ "learning_rate": 4.919786096256685e-05,
218
+ "loss": 0.1888,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.06,
223
+ "learning_rate": 4.917112299465241e-05,
224
+ "loss": 0.0934,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.06,
229
+ "learning_rate": 4.914438502673797e-05,
230
+ "loss": 0.1593,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.06,
235
+ "learning_rate": 4.911764705882353e-05,
236
+ "loss": 0.0721,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.06,
241
+ "learning_rate": 4.909090909090909e-05,
242
+ "loss": 0.1139,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.06,
247
+ "learning_rate": 4.906417112299466e-05,
248
+ "loss": 0.1118,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.07,
253
+ "learning_rate": 4.9037433155080216e-05,
254
+ "loss": 0.218,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.07,
259
+ "learning_rate": 4.901069518716578e-05,
260
+ "loss": 0.126,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.07,
265
+ "learning_rate": 4.898395721925134e-05,
266
+ "loss": 0.1057,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.07,
271
+ "learning_rate": 4.89572192513369e-05,
272
+ "loss": 0.1825,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.07,
277
+ "learning_rate": 4.8930481283422465e-05,
278
+ "loss": 0.1379,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.07,
283
+ "learning_rate": 4.8903743315508024e-05,
284
+ "loss": 0.1311,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.08,
289
+ "learning_rate": 4.8877005347593584e-05,
290
+ "loss": 0.0722,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.08,
295
+ "learning_rate": 4.885026737967915e-05,
296
+ "loss": 0.1286,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.08,
301
+ "learning_rate": 4.882352941176471e-05,
302
+ "loss": 0.1534,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.08,
307
+ "learning_rate": 4.879679144385027e-05,
308
+ "loss": 0.1332,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.08,
313
+ "eval_accuracy": 0.8395015105740181,
314
+ "eval_f1": 0.8361086180158485,
315
+ "eval_loss": 0.9186971783638,
316
+ "eval_precision": 0.8701532825604937,
317
+ "eval_recall": 0.8395015105740181,
318
+ "eval_runtime": 10.5139,
319
+ "eval_samples_per_second": 251.857,
320
+ "eval_steps_per_second": 3.995,
321
+ "step": 500
322
+ }
323
+ ],
324
+ "logging_steps": 10,
325
+ "max_steps": 18750,
326
+ "num_input_tokens_seen": 0,
327
+ "num_train_epochs": 3,
328
+ "save_steps": 500,
329
+ "total_flos": 4209814683648000.0,
330
+ "train_batch_size": 16,
331
+ "trial_name": null,
332
+ "trial_params": null
333
+ }
checkpoint-500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ea61cfa8a59fe23f079dded341db3c34a77c19986a2b3885296af3c7226c95
3
+ size 4664
checkpoint-5500/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base-finetuned-8M-MLM-500K/",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
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": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "LABEL_0": 0,
22
+ "LABEL_1": 1,
23
+ "LABEL_2": 2
24
+ },
25
+ "layer_norm_eps": 1e-05,
26
+ "max_position_embeddings": 514,
27
+ "model_type": "roberta",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 1,
31
+ "position_embedding_type": "absolute",
32
+ "problem_type": "single_label_classification",
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.37.0",
35
+ "type_vocab_size": 1,
36
+ "use_cache": true,
37
+ "vocab_size": 50265
38
+ }
checkpoint-5500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbf96e62ad7d1873c25dff8775d36da1898ef548cf24d8f248600b0225177a79
3
+ size 498615900
checkpoint-5500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f0c2927b3cb22f9dabba616c516bf45ca43eaafd2a8dd3f3d61e82c62af9d22
3
+ size 997351610
checkpoint-5500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5c01ea585b4712252347c2d000ee91ad3cde0d78d20295a4f0d76299fc993fa
3
+ size 14244
checkpoint-5500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df6985210f80dfebeafa38f837c8d9948625668d44407cf249bd929fa6224f38
3
+ size 1064
checkpoint-5500/trainer_state.json ADDED
@@ -0,0 +1,3453 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9186971783638,
3
+ "best_model_checkpoint": "./Finetuned_Classification_model_200k/checkpoint-500",
4
+ "epoch": 0.88,
5
+ "eval_steps": 500,
6
+ "global_step": 5500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 1e-05,
14
+ "loss": 1.037,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 2e-05,
20
+ "loss": 0.907,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 3e-05,
26
+ "loss": 0.7159,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4e-05,
32
+ "loss": 0.5758,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.3454,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.997326203208557e-05,
44
+ "loss": 0.1851,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 4.994652406417113e-05,
50
+ "loss": 0.2821,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 4.9919786096256686e-05,
56
+ "loss": 0.1741,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 4.9893048128342245e-05,
62
+ "loss": 0.1749,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.986631016042781e-05,
68
+ "loss": 0.2407,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 4.983957219251337e-05,
74
+ "loss": 0.2612,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 4.981283422459893e-05,
80
+ "loss": 0.1677,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 4.9786096256684495e-05,
86
+ "loss": 0.1123,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 4.975935828877006e-05,
92
+ "loss": 0.1368,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 4.973262032085561e-05,
98
+ "loss": 0.1449,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.03,
103
+ "learning_rate": 4.970588235294118e-05,
104
+ "loss": 0.152,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.03,
109
+ "learning_rate": 4.967914438502674e-05,
110
+ "loss": 0.2504,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 4.9652406417112304e-05,
116
+ "loss": 0.2578,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 4.962566844919786e-05,
122
+ "loss": 0.1289,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 4.959893048128342e-05,
128
+ "loss": 0.1263,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 4.957219251336899e-05,
134
+ "loss": 0.1031,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.04,
139
+ "learning_rate": 4.9545454545454553e-05,
140
+ "loss": 0.2453,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.04,
145
+ "learning_rate": 4.9518716577540106e-05,
146
+ "loss": 0.0675,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.04,
151
+ "learning_rate": 4.949197860962567e-05,
152
+ "loss": 0.1807,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.04,
157
+ "learning_rate": 4.946524064171123e-05,
158
+ "loss": 0.1493,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.04,
163
+ "learning_rate": 4.9438502673796796e-05,
164
+ "loss": 0.109,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 4.9411764705882355e-05,
170
+ "loss": 0.2148,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 4.9385026737967914e-05,
176
+ "loss": 0.2574,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.05,
181
+ "learning_rate": 4.935828877005348e-05,
182
+ "loss": 0.2198,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.05,
187
+ "learning_rate": 4.933155080213904e-05,
188
+ "loss": 0.1574,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.05,
193
+ "learning_rate": 4.93048128342246e-05,
194
+ "loss": 0.0866,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.05,
199
+ "learning_rate": 4.9278074866310164e-05,
200
+ "loss": 0.1371,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.05,
205
+ "learning_rate": 4.925133689839572e-05,
206
+ "loss": 0.2074,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.05,
211
+ "learning_rate": 4.922459893048129e-05,
212
+ "loss": 0.3628,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.06,
217
+ "learning_rate": 4.919786096256685e-05,
218
+ "loss": 0.1888,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.06,
223
+ "learning_rate": 4.917112299465241e-05,
224
+ "loss": 0.0934,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.06,
229
+ "learning_rate": 4.914438502673797e-05,
230
+ "loss": 0.1593,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.06,
235
+ "learning_rate": 4.911764705882353e-05,
236
+ "loss": 0.0721,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.06,
241
+ "learning_rate": 4.909090909090909e-05,
242
+ "loss": 0.1139,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.06,
247
+ "learning_rate": 4.906417112299466e-05,
248
+ "loss": 0.1118,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.07,
253
+ "learning_rate": 4.9037433155080216e-05,
254
+ "loss": 0.218,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.07,
259
+ "learning_rate": 4.901069518716578e-05,
260
+ "loss": 0.126,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.07,
265
+ "learning_rate": 4.898395721925134e-05,
266
+ "loss": 0.1057,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.07,
271
+ "learning_rate": 4.89572192513369e-05,
272
+ "loss": 0.1825,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.07,
277
+ "learning_rate": 4.8930481283422465e-05,
278
+ "loss": 0.1379,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.07,
283
+ "learning_rate": 4.8903743315508024e-05,
284
+ "loss": 0.1311,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.08,
289
+ "learning_rate": 4.8877005347593584e-05,
290
+ "loss": 0.0722,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.08,
295
+ "learning_rate": 4.885026737967915e-05,
296
+ "loss": 0.1286,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.08,
301
+ "learning_rate": 4.882352941176471e-05,
302
+ "loss": 0.1534,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.08,
307
+ "learning_rate": 4.879679144385027e-05,
308
+ "loss": 0.1332,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.08,
313
+ "eval_accuracy": 0.8395015105740181,
314
+ "eval_f1": 0.8361086180158485,
315
+ "eval_loss": 0.9186971783638,
316
+ "eval_precision": 0.8701532825604937,
317
+ "eval_recall": 0.8395015105740181,
318
+ "eval_runtime": 10.5139,
319
+ "eval_samples_per_second": 251.857,
320
+ "eval_steps_per_second": 3.995,
321
+ "step": 500
322
+ },
323
+ {
324
+ "epoch": 0.08,
325
+ "learning_rate": 4.8770053475935826e-05,
326
+ "loss": 0.1317,
327
+ "step": 510
328
+ },
329
+ {
330
+ "epoch": 0.08,
331
+ "learning_rate": 4.874331550802139e-05,
332
+ "loss": 0.0683,
333
+ "step": 520
334
+ },
335
+ {
336
+ "epoch": 0.08,
337
+ "learning_rate": 4.871657754010696e-05,
338
+ "loss": 0.0474,
339
+ "step": 530
340
+ },
341
+ {
342
+ "epoch": 0.09,
343
+ "learning_rate": 4.868983957219252e-05,
344
+ "loss": 0.2302,
345
+ "step": 540
346
+ },
347
+ {
348
+ "epoch": 0.09,
349
+ "learning_rate": 4.8663101604278076e-05,
350
+ "loss": 0.1788,
351
+ "step": 550
352
+ },
353
+ {
354
+ "epoch": 0.09,
355
+ "learning_rate": 4.863636363636364e-05,
356
+ "loss": 0.0917,
357
+ "step": 560
358
+ },
359
+ {
360
+ "epoch": 0.09,
361
+ "learning_rate": 4.86096256684492e-05,
362
+ "loss": 0.2001,
363
+ "step": 570
364
+ },
365
+ {
366
+ "epoch": 0.09,
367
+ "learning_rate": 4.858288770053476e-05,
368
+ "loss": 0.3409,
369
+ "step": 580
370
+ },
371
+ {
372
+ "epoch": 0.09,
373
+ "learning_rate": 4.855614973262032e-05,
374
+ "loss": 0.2044,
375
+ "step": 590
376
+ },
377
+ {
378
+ "epoch": 0.1,
379
+ "learning_rate": 4.8529411764705885e-05,
380
+ "loss": 0.1292,
381
+ "step": 600
382
+ },
383
+ {
384
+ "epoch": 0.1,
385
+ "learning_rate": 4.850267379679145e-05,
386
+ "loss": 0.1422,
387
+ "step": 610
388
+ },
389
+ {
390
+ "epoch": 0.1,
391
+ "learning_rate": 4.8475935828877e-05,
392
+ "loss": 0.1165,
393
+ "step": 620
394
+ },
395
+ {
396
+ "epoch": 0.1,
397
+ "learning_rate": 4.844919786096257e-05,
398
+ "loss": 0.1115,
399
+ "step": 630
400
+ },
401
+ {
402
+ "epoch": 0.1,
403
+ "learning_rate": 4.8422459893048135e-05,
404
+ "loss": 0.0957,
405
+ "step": 640
406
+ },
407
+ {
408
+ "epoch": 0.1,
409
+ "learning_rate": 4.8395721925133694e-05,
410
+ "loss": 0.1791,
411
+ "step": 650
412
+ },
413
+ {
414
+ "epoch": 0.11,
415
+ "learning_rate": 4.836898395721925e-05,
416
+ "loss": 0.0691,
417
+ "step": 660
418
+ },
419
+ {
420
+ "epoch": 0.11,
421
+ "learning_rate": 4.834224598930481e-05,
422
+ "loss": 0.2829,
423
+ "step": 670
424
+ },
425
+ {
426
+ "epoch": 0.11,
427
+ "learning_rate": 4.831550802139038e-05,
428
+ "loss": 0.2635,
429
+ "step": 680
430
+ },
431
+ {
432
+ "epoch": 0.11,
433
+ "learning_rate": 4.828877005347594e-05,
434
+ "loss": 0.1722,
435
+ "step": 690
436
+ },
437
+ {
438
+ "epoch": 0.11,
439
+ "learning_rate": 4.8262032085561496e-05,
440
+ "loss": 0.0974,
441
+ "step": 700
442
+ },
443
+ {
444
+ "epoch": 0.11,
445
+ "learning_rate": 4.823529411764706e-05,
446
+ "loss": 0.0888,
447
+ "step": 710
448
+ },
449
+ {
450
+ "epoch": 0.12,
451
+ "learning_rate": 4.820855614973262e-05,
452
+ "loss": 0.3148,
453
+ "step": 720
454
+ },
455
+ {
456
+ "epoch": 0.12,
457
+ "learning_rate": 4.8181818181818186e-05,
458
+ "loss": 0.3556,
459
+ "step": 730
460
+ },
461
+ {
462
+ "epoch": 0.12,
463
+ "learning_rate": 4.8155080213903745e-05,
464
+ "loss": 0.1788,
465
+ "step": 740
466
+ },
467
+ {
468
+ "epoch": 0.12,
469
+ "learning_rate": 4.8128342245989304e-05,
470
+ "loss": 0.4376,
471
+ "step": 750
472
+ },
473
+ {
474
+ "epoch": 0.12,
475
+ "learning_rate": 4.810160427807487e-05,
476
+ "loss": 0.2413,
477
+ "step": 760
478
+ },
479
+ {
480
+ "epoch": 0.12,
481
+ "learning_rate": 4.807486631016043e-05,
482
+ "loss": 0.1802,
483
+ "step": 770
484
+ },
485
+ {
486
+ "epoch": 0.12,
487
+ "learning_rate": 4.804812834224599e-05,
488
+ "loss": 0.1364,
489
+ "step": 780
490
+ },
491
+ {
492
+ "epoch": 0.13,
493
+ "learning_rate": 4.8021390374331554e-05,
494
+ "loss": 0.0667,
495
+ "step": 790
496
+ },
497
+ {
498
+ "epoch": 0.13,
499
+ "learning_rate": 4.799465240641711e-05,
500
+ "loss": 0.1409,
501
+ "step": 800
502
+ },
503
+ {
504
+ "epoch": 0.13,
505
+ "learning_rate": 4.796791443850268e-05,
506
+ "loss": 0.0897,
507
+ "step": 810
508
+ },
509
+ {
510
+ "epoch": 0.13,
511
+ "learning_rate": 4.794117647058824e-05,
512
+ "loss": 0.0871,
513
+ "step": 820
514
+ },
515
+ {
516
+ "epoch": 0.13,
517
+ "learning_rate": 4.79144385026738e-05,
518
+ "loss": 0.0552,
519
+ "step": 830
520
+ },
521
+ {
522
+ "epoch": 0.13,
523
+ "learning_rate": 4.788770053475936e-05,
524
+ "loss": 0.2569,
525
+ "step": 840
526
+ },
527
+ {
528
+ "epoch": 0.14,
529
+ "learning_rate": 4.786096256684492e-05,
530
+ "loss": 0.2877,
531
+ "step": 850
532
+ },
533
+ {
534
+ "epoch": 0.14,
535
+ "learning_rate": 4.783422459893048e-05,
536
+ "loss": 0.1365,
537
+ "step": 860
538
+ },
539
+ {
540
+ "epoch": 0.14,
541
+ "learning_rate": 4.7807486631016047e-05,
542
+ "loss": 0.1604,
543
+ "step": 870
544
+ },
545
+ {
546
+ "epoch": 0.14,
547
+ "learning_rate": 4.7780748663101606e-05,
548
+ "loss": 0.2179,
549
+ "step": 880
550
+ },
551
+ {
552
+ "epoch": 0.14,
553
+ "learning_rate": 4.7754010695187165e-05,
554
+ "loss": 0.0969,
555
+ "step": 890
556
+ },
557
+ {
558
+ "epoch": 0.14,
559
+ "learning_rate": 4.772727272727273e-05,
560
+ "loss": 0.0953,
561
+ "step": 900
562
+ },
563
+ {
564
+ "epoch": 0.15,
565
+ "learning_rate": 4.770053475935829e-05,
566
+ "loss": 0.0803,
567
+ "step": 910
568
+ },
569
+ {
570
+ "epoch": 0.15,
571
+ "learning_rate": 4.7673796791443855e-05,
572
+ "loss": 0.1573,
573
+ "step": 920
574
+ },
575
+ {
576
+ "epoch": 0.15,
577
+ "learning_rate": 4.7647058823529414e-05,
578
+ "loss": 0.1388,
579
+ "step": 930
580
+ },
581
+ {
582
+ "epoch": 0.15,
583
+ "learning_rate": 4.7620320855614973e-05,
584
+ "loss": 0.1559,
585
+ "step": 940
586
+ },
587
+ {
588
+ "epoch": 0.15,
589
+ "learning_rate": 4.759358288770054e-05,
590
+ "loss": 0.2411,
591
+ "step": 950
592
+ },
593
+ {
594
+ "epoch": 0.15,
595
+ "learning_rate": 4.75668449197861e-05,
596
+ "loss": 0.1119,
597
+ "step": 960
598
+ },
599
+ {
600
+ "epoch": 0.16,
601
+ "learning_rate": 4.754010695187166e-05,
602
+ "loss": 0.1416,
603
+ "step": 970
604
+ },
605
+ {
606
+ "epoch": 0.16,
607
+ "learning_rate": 4.751336898395722e-05,
608
+ "loss": 0.1073,
609
+ "step": 980
610
+ },
611
+ {
612
+ "epoch": 0.16,
613
+ "learning_rate": 4.748663101604278e-05,
614
+ "loss": 0.1799,
615
+ "step": 990
616
+ },
617
+ {
618
+ "epoch": 0.16,
619
+ "learning_rate": 4.745989304812835e-05,
620
+ "loss": 0.127,
621
+ "step": 1000
622
+ },
623
+ {
624
+ "epoch": 0.16,
625
+ "eval_accuracy": 0.847809667673716,
626
+ "eval_f1": 0.8460260531471735,
627
+ "eval_loss": 1.017073392868042,
628
+ "eval_precision": 0.8647086198841812,
629
+ "eval_recall": 0.847809667673716,
630
+ "eval_runtime": 10.5379,
631
+ "eval_samples_per_second": 251.284,
632
+ "eval_steps_per_second": 3.986,
633
+ "step": 1000
634
+ },
635
+ {
636
+ "epoch": 0.16,
637
+ "learning_rate": 4.74331550802139e-05,
638
+ "loss": 0.1774,
639
+ "step": 1010
640
+ },
641
+ {
642
+ "epoch": 0.16,
643
+ "learning_rate": 4.7406417112299466e-05,
644
+ "loss": 0.0869,
645
+ "step": 1020
646
+ },
647
+ {
648
+ "epoch": 0.16,
649
+ "learning_rate": 4.737967914438503e-05,
650
+ "loss": 0.0686,
651
+ "step": 1030
652
+ },
653
+ {
654
+ "epoch": 0.17,
655
+ "learning_rate": 4.735294117647059e-05,
656
+ "loss": 0.124,
657
+ "step": 1040
658
+ },
659
+ {
660
+ "epoch": 0.17,
661
+ "learning_rate": 4.732620320855615e-05,
662
+ "loss": 0.0993,
663
+ "step": 1050
664
+ },
665
+ {
666
+ "epoch": 0.17,
667
+ "learning_rate": 4.7299465240641716e-05,
668
+ "loss": 0.2451,
669
+ "step": 1060
670
+ },
671
+ {
672
+ "epoch": 0.17,
673
+ "learning_rate": 4.7272727272727275e-05,
674
+ "loss": 0.1607,
675
+ "step": 1070
676
+ },
677
+ {
678
+ "epoch": 0.17,
679
+ "learning_rate": 4.724598930481284e-05,
680
+ "loss": 0.1289,
681
+ "step": 1080
682
+ },
683
+ {
684
+ "epoch": 0.17,
685
+ "learning_rate": 4.721925133689839e-05,
686
+ "loss": 0.1813,
687
+ "step": 1090
688
+ },
689
+ {
690
+ "epoch": 0.18,
691
+ "learning_rate": 4.719251336898396e-05,
692
+ "loss": 0.1628,
693
+ "step": 1100
694
+ },
695
+ {
696
+ "epoch": 0.18,
697
+ "learning_rate": 4.7165775401069524e-05,
698
+ "loss": 0.3152,
699
+ "step": 1110
700
+ },
701
+ {
702
+ "epoch": 0.18,
703
+ "learning_rate": 4.7139037433155083e-05,
704
+ "loss": 0.3835,
705
+ "step": 1120
706
+ },
707
+ {
708
+ "epoch": 0.18,
709
+ "learning_rate": 4.711229946524064e-05,
710
+ "loss": 0.1713,
711
+ "step": 1130
712
+ },
713
+ {
714
+ "epoch": 0.18,
715
+ "learning_rate": 4.70855614973262e-05,
716
+ "loss": 0.1812,
717
+ "step": 1140
718
+ },
719
+ {
720
+ "epoch": 0.18,
721
+ "learning_rate": 4.705882352941177e-05,
722
+ "loss": 0.1928,
723
+ "step": 1150
724
+ },
725
+ {
726
+ "epoch": 0.19,
727
+ "learning_rate": 4.703208556149733e-05,
728
+ "loss": 0.2963,
729
+ "step": 1160
730
+ },
731
+ {
732
+ "epoch": 0.19,
733
+ "learning_rate": 4.7005347593582885e-05,
734
+ "loss": 0.2073,
735
+ "step": 1170
736
+ },
737
+ {
738
+ "epoch": 0.19,
739
+ "learning_rate": 4.697860962566845e-05,
740
+ "loss": 0.2299,
741
+ "step": 1180
742
+ },
743
+ {
744
+ "epoch": 0.19,
745
+ "learning_rate": 4.695187165775402e-05,
746
+ "loss": 0.1832,
747
+ "step": 1190
748
+ },
749
+ {
750
+ "epoch": 0.19,
751
+ "learning_rate": 4.6925133689839576e-05,
752
+ "loss": 0.1844,
753
+ "step": 1200
754
+ },
755
+ {
756
+ "epoch": 0.19,
757
+ "learning_rate": 4.6898395721925135e-05,
758
+ "loss": 0.4959,
759
+ "step": 1210
760
+ },
761
+ {
762
+ "epoch": 0.2,
763
+ "learning_rate": 4.6871657754010694e-05,
764
+ "loss": 0.2516,
765
+ "step": 1220
766
+ },
767
+ {
768
+ "epoch": 0.2,
769
+ "learning_rate": 4.684491978609626e-05,
770
+ "loss": 0.1754,
771
+ "step": 1230
772
+ },
773
+ {
774
+ "epoch": 0.2,
775
+ "learning_rate": 4.681818181818182e-05,
776
+ "loss": 0.2554,
777
+ "step": 1240
778
+ },
779
+ {
780
+ "epoch": 0.2,
781
+ "learning_rate": 4.679144385026738e-05,
782
+ "loss": 0.1976,
783
+ "step": 1250
784
+ },
785
+ {
786
+ "epoch": 0.2,
787
+ "learning_rate": 4.6764705882352944e-05,
788
+ "loss": 0.1906,
789
+ "step": 1260
790
+ },
791
+ {
792
+ "epoch": 0.2,
793
+ "learning_rate": 4.673796791443851e-05,
794
+ "loss": 0.0806,
795
+ "step": 1270
796
+ },
797
+ {
798
+ "epoch": 0.2,
799
+ "learning_rate": 4.671122994652407e-05,
800
+ "loss": 0.1229,
801
+ "step": 1280
802
+ },
803
+ {
804
+ "epoch": 0.21,
805
+ "learning_rate": 4.668449197860963e-05,
806
+ "loss": 0.1222,
807
+ "step": 1290
808
+ },
809
+ {
810
+ "epoch": 0.21,
811
+ "learning_rate": 4.665775401069519e-05,
812
+ "loss": 0.1049,
813
+ "step": 1300
814
+ },
815
+ {
816
+ "epoch": 0.21,
817
+ "learning_rate": 4.663101604278075e-05,
818
+ "loss": 0.1548,
819
+ "step": 1310
820
+ },
821
+ {
822
+ "epoch": 0.21,
823
+ "learning_rate": 4.660427807486631e-05,
824
+ "loss": 0.1077,
825
+ "step": 1320
826
+ },
827
+ {
828
+ "epoch": 0.21,
829
+ "learning_rate": 4.657754010695187e-05,
830
+ "loss": 0.1949,
831
+ "step": 1330
832
+ },
833
+ {
834
+ "epoch": 0.21,
835
+ "learning_rate": 4.6550802139037436e-05,
836
+ "loss": 0.1349,
837
+ "step": 1340
838
+ },
839
+ {
840
+ "epoch": 0.22,
841
+ "learning_rate": 4.6524064171123e-05,
842
+ "loss": 0.223,
843
+ "step": 1350
844
+ },
845
+ {
846
+ "epoch": 0.22,
847
+ "learning_rate": 4.6497326203208555e-05,
848
+ "loss": 0.1132,
849
+ "step": 1360
850
+ },
851
+ {
852
+ "epoch": 0.22,
853
+ "learning_rate": 4.647058823529412e-05,
854
+ "loss": 0.123,
855
+ "step": 1370
856
+ },
857
+ {
858
+ "epoch": 0.22,
859
+ "learning_rate": 4.644385026737968e-05,
860
+ "loss": 0.1741,
861
+ "step": 1380
862
+ },
863
+ {
864
+ "epoch": 0.22,
865
+ "learning_rate": 4.6417112299465245e-05,
866
+ "loss": 0.0913,
867
+ "step": 1390
868
+ },
869
+ {
870
+ "epoch": 0.22,
871
+ "learning_rate": 4.6390374331550804e-05,
872
+ "loss": 0.1428,
873
+ "step": 1400
874
+ },
875
+ {
876
+ "epoch": 0.23,
877
+ "learning_rate": 4.636363636363636e-05,
878
+ "loss": 0.1171,
879
+ "step": 1410
880
+ },
881
+ {
882
+ "epoch": 0.23,
883
+ "learning_rate": 4.633689839572193e-05,
884
+ "loss": 0.1773,
885
+ "step": 1420
886
+ },
887
+ {
888
+ "epoch": 0.23,
889
+ "learning_rate": 4.631016042780749e-05,
890
+ "loss": 0.1372,
891
+ "step": 1430
892
+ },
893
+ {
894
+ "epoch": 0.23,
895
+ "learning_rate": 4.628342245989305e-05,
896
+ "loss": 0.2797,
897
+ "step": 1440
898
+ },
899
+ {
900
+ "epoch": 0.23,
901
+ "learning_rate": 4.625668449197861e-05,
902
+ "loss": 0.2813,
903
+ "step": 1450
904
+ },
905
+ {
906
+ "epoch": 0.23,
907
+ "learning_rate": 4.622994652406417e-05,
908
+ "loss": 0.2861,
909
+ "step": 1460
910
+ },
911
+ {
912
+ "epoch": 0.24,
913
+ "learning_rate": 4.620320855614974e-05,
914
+ "loss": 0.0677,
915
+ "step": 1470
916
+ },
917
+ {
918
+ "epoch": 0.24,
919
+ "learning_rate": 4.61764705882353e-05,
920
+ "loss": 0.104,
921
+ "step": 1480
922
+ },
923
+ {
924
+ "epoch": 0.24,
925
+ "learning_rate": 4.6149732620320856e-05,
926
+ "loss": 0.1533,
927
+ "step": 1490
928
+ },
929
+ {
930
+ "epoch": 0.24,
931
+ "learning_rate": 4.612299465240642e-05,
932
+ "loss": 0.1186,
933
+ "step": 1500
934
+ },
935
+ {
936
+ "epoch": 0.24,
937
+ "eval_accuracy": 0.8376132930513596,
938
+ "eval_f1": 0.8363352339205664,
939
+ "eval_loss": 1.0860812664031982,
940
+ "eval_precision": 0.8484990166173995,
941
+ "eval_recall": 0.8376132930513596,
942
+ "eval_runtime": 10.5226,
943
+ "eval_samples_per_second": 251.648,
944
+ "eval_steps_per_second": 3.991,
945
+ "step": 1500
946
+ },
947
+ {
948
+ "epoch": 0.24,
949
+ "learning_rate": 4.609625668449198e-05,
950
+ "loss": 0.0683,
951
+ "step": 1510
952
+ },
953
+ {
954
+ "epoch": 0.24,
955
+ "learning_rate": 4.606951871657754e-05,
956
+ "loss": 0.1767,
957
+ "step": 1520
958
+ },
959
+ {
960
+ "epoch": 0.24,
961
+ "learning_rate": 4.6042780748663106e-05,
962
+ "loss": 0.0559,
963
+ "step": 1530
964
+ },
965
+ {
966
+ "epoch": 0.25,
967
+ "learning_rate": 4.6016042780748665e-05,
968
+ "loss": 0.1056,
969
+ "step": 1540
970
+ },
971
+ {
972
+ "epoch": 0.25,
973
+ "learning_rate": 4.598930481283423e-05,
974
+ "loss": 0.1607,
975
+ "step": 1550
976
+ },
977
+ {
978
+ "epoch": 0.25,
979
+ "learning_rate": 4.596256684491979e-05,
980
+ "loss": 0.053,
981
+ "step": 1560
982
+ },
983
+ {
984
+ "epoch": 0.25,
985
+ "learning_rate": 4.593582887700535e-05,
986
+ "loss": 0.2743,
987
+ "step": 1570
988
+ },
989
+ {
990
+ "epoch": 0.25,
991
+ "learning_rate": 4.5909090909090914e-05,
992
+ "loss": 0.1569,
993
+ "step": 1580
994
+ },
995
+ {
996
+ "epoch": 0.25,
997
+ "learning_rate": 4.588235294117647e-05,
998
+ "loss": 0.0648,
999
+ "step": 1590
1000
+ },
1001
+ {
1002
+ "epoch": 0.26,
1003
+ "learning_rate": 4.585561497326203e-05,
1004
+ "loss": 0.132,
1005
+ "step": 1600
1006
+ },
1007
+ {
1008
+ "epoch": 0.26,
1009
+ "learning_rate": 4.58288770053476e-05,
1010
+ "loss": 0.1115,
1011
+ "step": 1610
1012
+ },
1013
+ {
1014
+ "epoch": 0.26,
1015
+ "learning_rate": 4.580213903743316e-05,
1016
+ "loss": 0.2035,
1017
+ "step": 1620
1018
+ },
1019
+ {
1020
+ "epoch": 0.26,
1021
+ "learning_rate": 4.5775401069518716e-05,
1022
+ "loss": 0.1193,
1023
+ "step": 1630
1024
+ },
1025
+ {
1026
+ "epoch": 0.26,
1027
+ "learning_rate": 4.5748663101604275e-05,
1028
+ "loss": 0.2867,
1029
+ "step": 1640
1030
+ },
1031
+ {
1032
+ "epoch": 0.26,
1033
+ "learning_rate": 4.572192513368984e-05,
1034
+ "loss": 0.2423,
1035
+ "step": 1650
1036
+ },
1037
+ {
1038
+ "epoch": 0.27,
1039
+ "learning_rate": 4.569518716577541e-05,
1040
+ "loss": 0.0959,
1041
+ "step": 1660
1042
+ },
1043
+ {
1044
+ "epoch": 0.27,
1045
+ "learning_rate": 4.5668449197860966e-05,
1046
+ "loss": 0.062,
1047
+ "step": 1670
1048
+ },
1049
+ {
1050
+ "epoch": 0.27,
1051
+ "learning_rate": 4.5641711229946525e-05,
1052
+ "loss": 0.3052,
1053
+ "step": 1680
1054
+ },
1055
+ {
1056
+ "epoch": 0.27,
1057
+ "learning_rate": 4.561497326203209e-05,
1058
+ "loss": 0.0814,
1059
+ "step": 1690
1060
+ },
1061
+ {
1062
+ "epoch": 0.27,
1063
+ "learning_rate": 4.558823529411765e-05,
1064
+ "loss": 0.4406,
1065
+ "step": 1700
1066
+ },
1067
+ {
1068
+ "epoch": 0.27,
1069
+ "learning_rate": 4.556149732620321e-05,
1070
+ "loss": 0.4316,
1071
+ "step": 1710
1072
+ },
1073
+ {
1074
+ "epoch": 0.28,
1075
+ "learning_rate": 4.553475935828877e-05,
1076
+ "loss": 0.2551,
1077
+ "step": 1720
1078
+ },
1079
+ {
1080
+ "epoch": 0.28,
1081
+ "learning_rate": 4.5508021390374334e-05,
1082
+ "loss": 0.4731,
1083
+ "step": 1730
1084
+ },
1085
+ {
1086
+ "epoch": 0.28,
1087
+ "learning_rate": 4.54812834224599e-05,
1088
+ "loss": 0.3468,
1089
+ "step": 1740
1090
+ },
1091
+ {
1092
+ "epoch": 0.28,
1093
+ "learning_rate": 4.545454545454546e-05,
1094
+ "loss": 0.4397,
1095
+ "step": 1750
1096
+ },
1097
+ {
1098
+ "epoch": 0.28,
1099
+ "learning_rate": 4.542780748663102e-05,
1100
+ "loss": 0.2163,
1101
+ "step": 1760
1102
+ },
1103
+ {
1104
+ "epoch": 0.28,
1105
+ "learning_rate": 4.5401069518716583e-05,
1106
+ "loss": 0.1675,
1107
+ "step": 1770
1108
+ },
1109
+ {
1110
+ "epoch": 0.28,
1111
+ "learning_rate": 4.537433155080214e-05,
1112
+ "loss": 0.124,
1113
+ "step": 1780
1114
+ },
1115
+ {
1116
+ "epoch": 0.29,
1117
+ "learning_rate": 4.53475935828877e-05,
1118
+ "loss": 0.192,
1119
+ "step": 1790
1120
+ },
1121
+ {
1122
+ "epoch": 0.29,
1123
+ "learning_rate": 4.532085561497326e-05,
1124
+ "loss": 0.2626,
1125
+ "step": 1800
1126
+ },
1127
+ {
1128
+ "epoch": 0.29,
1129
+ "learning_rate": 4.5294117647058826e-05,
1130
+ "loss": 0.233,
1131
+ "step": 1810
1132
+ },
1133
+ {
1134
+ "epoch": 0.29,
1135
+ "learning_rate": 4.526737967914439e-05,
1136
+ "loss": 0.2005,
1137
+ "step": 1820
1138
+ },
1139
+ {
1140
+ "epoch": 0.29,
1141
+ "learning_rate": 4.5240641711229944e-05,
1142
+ "loss": 0.1611,
1143
+ "step": 1830
1144
+ },
1145
+ {
1146
+ "epoch": 0.29,
1147
+ "learning_rate": 4.521390374331551e-05,
1148
+ "loss": 0.1215,
1149
+ "step": 1840
1150
+ },
1151
+ {
1152
+ "epoch": 0.3,
1153
+ "learning_rate": 4.518716577540107e-05,
1154
+ "loss": 0.1504,
1155
+ "step": 1850
1156
+ },
1157
+ {
1158
+ "epoch": 0.3,
1159
+ "learning_rate": 4.5160427807486635e-05,
1160
+ "loss": 0.1317,
1161
+ "step": 1860
1162
+ },
1163
+ {
1164
+ "epoch": 0.3,
1165
+ "learning_rate": 4.5133689839572194e-05,
1166
+ "loss": 0.1624,
1167
+ "step": 1870
1168
+ },
1169
+ {
1170
+ "epoch": 0.3,
1171
+ "learning_rate": 4.510695187165775e-05,
1172
+ "loss": 0.1815,
1173
+ "step": 1880
1174
+ },
1175
+ {
1176
+ "epoch": 0.3,
1177
+ "learning_rate": 4.508021390374332e-05,
1178
+ "loss": 0.2438,
1179
+ "step": 1890
1180
+ },
1181
+ {
1182
+ "epoch": 0.3,
1183
+ "learning_rate": 4.5053475935828885e-05,
1184
+ "loss": 0.1908,
1185
+ "step": 1900
1186
+ },
1187
+ {
1188
+ "epoch": 0.31,
1189
+ "learning_rate": 4.502673796791444e-05,
1190
+ "loss": 0.2333,
1191
+ "step": 1910
1192
+ },
1193
+ {
1194
+ "epoch": 0.31,
1195
+ "learning_rate": 4.5e-05,
1196
+ "loss": 0.2385,
1197
+ "step": 1920
1198
+ },
1199
+ {
1200
+ "epoch": 0.31,
1201
+ "learning_rate": 4.497326203208556e-05,
1202
+ "loss": 0.1889,
1203
+ "step": 1930
1204
+ },
1205
+ {
1206
+ "epoch": 0.31,
1207
+ "learning_rate": 4.494652406417113e-05,
1208
+ "loss": 0.1202,
1209
+ "step": 1940
1210
+ },
1211
+ {
1212
+ "epoch": 0.31,
1213
+ "learning_rate": 4.491978609625669e-05,
1214
+ "loss": 0.3926,
1215
+ "step": 1950
1216
+ },
1217
+ {
1218
+ "epoch": 0.31,
1219
+ "learning_rate": 4.4893048128342246e-05,
1220
+ "loss": 0.3656,
1221
+ "step": 1960
1222
+ },
1223
+ {
1224
+ "epoch": 0.32,
1225
+ "learning_rate": 4.486631016042781e-05,
1226
+ "loss": 0.1626,
1227
+ "step": 1970
1228
+ },
1229
+ {
1230
+ "epoch": 0.32,
1231
+ "learning_rate": 4.483957219251337e-05,
1232
+ "loss": 0.1318,
1233
+ "step": 1980
1234
+ },
1235
+ {
1236
+ "epoch": 0.32,
1237
+ "learning_rate": 4.481283422459893e-05,
1238
+ "loss": 0.2922,
1239
+ "step": 1990
1240
+ },
1241
+ {
1242
+ "epoch": 0.32,
1243
+ "learning_rate": 4.4786096256684495e-05,
1244
+ "loss": 0.3545,
1245
+ "step": 2000
1246
+ },
1247
+ {
1248
+ "epoch": 0.32,
1249
+ "eval_accuracy": 0.8413897280966768,
1250
+ "eval_f1": 0.8394373029347564,
1251
+ "eval_loss": 2.3635265827178955,
1252
+ "eval_precision": 0.8588437332105303,
1253
+ "eval_recall": 0.8413897280966768,
1254
+ "eval_runtime": 10.5481,
1255
+ "eval_samples_per_second": 251.039,
1256
+ "eval_steps_per_second": 3.982,
1257
+ "step": 2000
1258
+ },
1259
+ {
1260
+ "epoch": 0.32,
1261
+ "learning_rate": 4.4759358288770054e-05,
1262
+ "loss": 0.1883,
1263
+ "step": 2010
1264
+ },
1265
+ {
1266
+ "epoch": 0.32,
1267
+ "learning_rate": 4.473262032085562e-05,
1268
+ "loss": 0.4851,
1269
+ "step": 2020
1270
+ },
1271
+ {
1272
+ "epoch": 0.32,
1273
+ "learning_rate": 4.470588235294118e-05,
1274
+ "loss": 0.4886,
1275
+ "step": 2030
1276
+ },
1277
+ {
1278
+ "epoch": 0.33,
1279
+ "learning_rate": 4.467914438502674e-05,
1280
+ "loss": 0.2792,
1281
+ "step": 2040
1282
+ },
1283
+ {
1284
+ "epoch": 0.33,
1285
+ "learning_rate": 4.4652406417112304e-05,
1286
+ "loss": 0.2475,
1287
+ "step": 2050
1288
+ },
1289
+ {
1290
+ "epoch": 0.33,
1291
+ "learning_rate": 4.462566844919786e-05,
1292
+ "loss": 0.1728,
1293
+ "step": 2060
1294
+ },
1295
+ {
1296
+ "epoch": 0.33,
1297
+ "learning_rate": 4.459893048128342e-05,
1298
+ "loss": 0.0873,
1299
+ "step": 2070
1300
+ },
1301
+ {
1302
+ "epoch": 0.33,
1303
+ "learning_rate": 4.457219251336899e-05,
1304
+ "loss": 0.1553,
1305
+ "step": 2080
1306
+ },
1307
+ {
1308
+ "epoch": 0.33,
1309
+ "learning_rate": 4.454545454545455e-05,
1310
+ "loss": 0.3127,
1311
+ "step": 2090
1312
+ },
1313
+ {
1314
+ "epoch": 0.34,
1315
+ "learning_rate": 4.4518716577540106e-05,
1316
+ "loss": 0.4534,
1317
+ "step": 2100
1318
+ },
1319
+ {
1320
+ "epoch": 0.34,
1321
+ "learning_rate": 4.449197860962567e-05,
1322
+ "loss": 0.2118,
1323
+ "step": 2110
1324
+ },
1325
+ {
1326
+ "epoch": 0.34,
1327
+ "learning_rate": 4.446524064171123e-05,
1328
+ "loss": 0.3214,
1329
+ "step": 2120
1330
+ },
1331
+ {
1332
+ "epoch": 0.34,
1333
+ "learning_rate": 4.44385026737968e-05,
1334
+ "loss": 0.1867,
1335
+ "step": 2130
1336
+ },
1337
+ {
1338
+ "epoch": 0.34,
1339
+ "learning_rate": 4.4411764705882356e-05,
1340
+ "loss": 0.3318,
1341
+ "step": 2140
1342
+ },
1343
+ {
1344
+ "epoch": 0.34,
1345
+ "learning_rate": 4.4385026737967915e-05,
1346
+ "loss": 0.164,
1347
+ "step": 2150
1348
+ },
1349
+ {
1350
+ "epoch": 0.35,
1351
+ "learning_rate": 4.435828877005348e-05,
1352
+ "loss": 0.2003,
1353
+ "step": 2160
1354
+ },
1355
+ {
1356
+ "epoch": 0.35,
1357
+ "learning_rate": 4.433155080213904e-05,
1358
+ "loss": 0.1854,
1359
+ "step": 2170
1360
+ },
1361
+ {
1362
+ "epoch": 0.35,
1363
+ "learning_rate": 4.43048128342246e-05,
1364
+ "loss": 0.2871,
1365
+ "step": 2180
1366
+ },
1367
+ {
1368
+ "epoch": 0.35,
1369
+ "learning_rate": 4.4278074866310165e-05,
1370
+ "loss": 0.306,
1371
+ "step": 2190
1372
+ },
1373
+ {
1374
+ "epoch": 0.35,
1375
+ "learning_rate": 4.4251336898395724e-05,
1376
+ "loss": 0.462,
1377
+ "step": 2200
1378
+ },
1379
+ {
1380
+ "epoch": 0.35,
1381
+ "learning_rate": 4.422459893048129e-05,
1382
+ "loss": 0.3016,
1383
+ "step": 2210
1384
+ },
1385
+ {
1386
+ "epoch": 0.36,
1387
+ "learning_rate": 4.419786096256684e-05,
1388
+ "loss": 0.2322,
1389
+ "step": 2220
1390
+ },
1391
+ {
1392
+ "epoch": 0.36,
1393
+ "learning_rate": 4.417112299465241e-05,
1394
+ "loss": 0.141,
1395
+ "step": 2230
1396
+ },
1397
+ {
1398
+ "epoch": 0.36,
1399
+ "learning_rate": 4.414438502673797e-05,
1400
+ "loss": 0.1501,
1401
+ "step": 2240
1402
+ },
1403
+ {
1404
+ "epoch": 0.36,
1405
+ "learning_rate": 4.411764705882353e-05,
1406
+ "loss": 0.2379,
1407
+ "step": 2250
1408
+ },
1409
+ {
1410
+ "epoch": 0.36,
1411
+ "learning_rate": 4.409090909090909e-05,
1412
+ "loss": 0.2233,
1413
+ "step": 2260
1414
+ },
1415
+ {
1416
+ "epoch": 0.36,
1417
+ "learning_rate": 4.406417112299465e-05,
1418
+ "loss": 0.4372,
1419
+ "step": 2270
1420
+ },
1421
+ {
1422
+ "epoch": 0.36,
1423
+ "learning_rate": 4.4037433155080216e-05,
1424
+ "loss": 0.5062,
1425
+ "step": 2280
1426
+ },
1427
+ {
1428
+ "epoch": 0.37,
1429
+ "learning_rate": 4.401069518716578e-05,
1430
+ "loss": 0.1763,
1431
+ "step": 2290
1432
+ },
1433
+ {
1434
+ "epoch": 0.37,
1435
+ "learning_rate": 4.3983957219251334e-05,
1436
+ "loss": 0.1486,
1437
+ "step": 2300
1438
+ },
1439
+ {
1440
+ "epoch": 0.37,
1441
+ "learning_rate": 4.39572192513369e-05,
1442
+ "loss": 0.2072,
1443
+ "step": 2310
1444
+ },
1445
+ {
1446
+ "epoch": 0.37,
1447
+ "learning_rate": 4.3930481283422466e-05,
1448
+ "loss": 0.0901,
1449
+ "step": 2320
1450
+ },
1451
+ {
1452
+ "epoch": 0.37,
1453
+ "learning_rate": 4.3903743315508025e-05,
1454
+ "loss": 0.351,
1455
+ "step": 2330
1456
+ },
1457
+ {
1458
+ "epoch": 0.37,
1459
+ "learning_rate": 4.3877005347593584e-05,
1460
+ "loss": 0.2836,
1461
+ "step": 2340
1462
+ },
1463
+ {
1464
+ "epoch": 0.38,
1465
+ "learning_rate": 4.385026737967914e-05,
1466
+ "loss": 0.3663,
1467
+ "step": 2350
1468
+ },
1469
+ {
1470
+ "epoch": 0.38,
1471
+ "learning_rate": 4.382352941176471e-05,
1472
+ "loss": 0.4,
1473
+ "step": 2360
1474
+ },
1475
+ {
1476
+ "epoch": 0.38,
1477
+ "learning_rate": 4.3796791443850275e-05,
1478
+ "loss": 0.3752,
1479
+ "step": 2370
1480
+ },
1481
+ {
1482
+ "epoch": 0.38,
1483
+ "learning_rate": 4.377005347593583e-05,
1484
+ "loss": 0.22,
1485
+ "step": 2380
1486
+ },
1487
+ {
1488
+ "epoch": 0.38,
1489
+ "learning_rate": 4.374331550802139e-05,
1490
+ "loss": 0.3702,
1491
+ "step": 2390
1492
+ },
1493
+ {
1494
+ "epoch": 0.38,
1495
+ "learning_rate": 4.371657754010696e-05,
1496
+ "loss": 0.1035,
1497
+ "step": 2400
1498
+ },
1499
+ {
1500
+ "epoch": 0.39,
1501
+ "learning_rate": 4.368983957219252e-05,
1502
+ "loss": 0.1581,
1503
+ "step": 2410
1504
+ },
1505
+ {
1506
+ "epoch": 0.39,
1507
+ "learning_rate": 4.3663101604278077e-05,
1508
+ "loss": 0.2575,
1509
+ "step": 2420
1510
+ },
1511
+ {
1512
+ "epoch": 0.39,
1513
+ "learning_rate": 4.3636363636363636e-05,
1514
+ "loss": 0.1866,
1515
+ "step": 2430
1516
+ },
1517
+ {
1518
+ "epoch": 0.39,
1519
+ "learning_rate": 4.36096256684492e-05,
1520
+ "loss": 0.3816,
1521
+ "step": 2440
1522
+ },
1523
+ {
1524
+ "epoch": 0.39,
1525
+ "learning_rate": 4.358288770053476e-05,
1526
+ "loss": 0.4613,
1527
+ "step": 2450
1528
+ },
1529
+ {
1530
+ "epoch": 0.39,
1531
+ "learning_rate": 4.355614973262032e-05,
1532
+ "loss": 0.1567,
1533
+ "step": 2460
1534
+ },
1535
+ {
1536
+ "epoch": 0.4,
1537
+ "learning_rate": 4.3529411764705885e-05,
1538
+ "loss": 0.1287,
1539
+ "step": 2470
1540
+ },
1541
+ {
1542
+ "epoch": 0.4,
1543
+ "learning_rate": 4.3502673796791444e-05,
1544
+ "loss": 0.187,
1545
+ "step": 2480
1546
+ },
1547
+ {
1548
+ "epoch": 0.4,
1549
+ "learning_rate": 4.347593582887701e-05,
1550
+ "loss": 0.1354,
1551
+ "step": 2490
1552
+ },
1553
+ {
1554
+ "epoch": 0.4,
1555
+ "learning_rate": 4.344919786096257e-05,
1556
+ "loss": 0.0646,
1557
+ "step": 2500
1558
+ },
1559
+ {
1560
+ "epoch": 0.4,
1561
+ "eval_accuracy": 0.849320241691843,
1562
+ "eval_f1": 0.8472401124107073,
1563
+ "eval_loss": 1.3920754194259644,
1564
+ "eval_precision": 0.869443045077059,
1565
+ "eval_recall": 0.849320241691843,
1566
+ "eval_runtime": 10.5377,
1567
+ "eval_samples_per_second": 251.287,
1568
+ "eval_steps_per_second": 3.986,
1569
+ "step": 2500
1570
+ },
1571
+ {
1572
+ "epoch": 0.4,
1573
+ "learning_rate": 4.342245989304813e-05,
1574
+ "loss": 0.1651,
1575
+ "step": 2510
1576
+ },
1577
+ {
1578
+ "epoch": 0.4,
1579
+ "learning_rate": 4.3395721925133694e-05,
1580
+ "loss": 0.0982,
1581
+ "step": 2520
1582
+ },
1583
+ {
1584
+ "epoch": 0.4,
1585
+ "learning_rate": 4.336898395721925e-05,
1586
+ "loss": 0.2102,
1587
+ "step": 2530
1588
+ },
1589
+ {
1590
+ "epoch": 0.41,
1591
+ "learning_rate": 4.334224598930481e-05,
1592
+ "loss": 0.2102,
1593
+ "step": 2540
1594
+ },
1595
+ {
1596
+ "epoch": 0.41,
1597
+ "learning_rate": 4.331550802139038e-05,
1598
+ "loss": 0.2738,
1599
+ "step": 2550
1600
+ },
1601
+ {
1602
+ "epoch": 0.41,
1603
+ "learning_rate": 4.328877005347594e-05,
1604
+ "loss": 0.0848,
1605
+ "step": 2560
1606
+ },
1607
+ {
1608
+ "epoch": 0.41,
1609
+ "learning_rate": 4.3262032085561496e-05,
1610
+ "loss": 0.2072,
1611
+ "step": 2570
1612
+ },
1613
+ {
1614
+ "epoch": 0.41,
1615
+ "learning_rate": 4.323529411764706e-05,
1616
+ "loss": 0.6024,
1617
+ "step": 2580
1618
+ },
1619
+ {
1620
+ "epoch": 0.41,
1621
+ "learning_rate": 4.320855614973262e-05,
1622
+ "loss": 0.5888,
1623
+ "step": 2590
1624
+ },
1625
+ {
1626
+ "epoch": 0.42,
1627
+ "learning_rate": 4.318181818181819e-05,
1628
+ "loss": 0.1546,
1629
+ "step": 2600
1630
+ },
1631
+ {
1632
+ "epoch": 0.42,
1633
+ "learning_rate": 4.3155080213903746e-05,
1634
+ "loss": 0.2223,
1635
+ "step": 2610
1636
+ },
1637
+ {
1638
+ "epoch": 0.42,
1639
+ "learning_rate": 4.3128342245989305e-05,
1640
+ "loss": 0.2447,
1641
+ "step": 2620
1642
+ },
1643
+ {
1644
+ "epoch": 0.42,
1645
+ "learning_rate": 4.310160427807487e-05,
1646
+ "loss": 0.1098,
1647
+ "step": 2630
1648
+ },
1649
+ {
1650
+ "epoch": 0.42,
1651
+ "learning_rate": 4.307486631016043e-05,
1652
+ "loss": 0.3385,
1653
+ "step": 2640
1654
+ },
1655
+ {
1656
+ "epoch": 0.42,
1657
+ "learning_rate": 4.304812834224599e-05,
1658
+ "loss": 0.1067,
1659
+ "step": 2650
1660
+ },
1661
+ {
1662
+ "epoch": 0.43,
1663
+ "learning_rate": 4.3021390374331554e-05,
1664
+ "loss": 0.2155,
1665
+ "step": 2660
1666
+ },
1667
+ {
1668
+ "epoch": 0.43,
1669
+ "learning_rate": 4.2994652406417113e-05,
1670
+ "loss": 0.345,
1671
+ "step": 2670
1672
+ },
1673
+ {
1674
+ "epoch": 0.43,
1675
+ "learning_rate": 4.296791443850268e-05,
1676
+ "loss": 0.2231,
1677
+ "step": 2680
1678
+ },
1679
+ {
1680
+ "epoch": 0.43,
1681
+ "learning_rate": 4.294117647058823e-05,
1682
+ "loss": 0.2863,
1683
+ "step": 2690
1684
+ },
1685
+ {
1686
+ "epoch": 0.43,
1687
+ "learning_rate": 4.29144385026738e-05,
1688
+ "loss": 0.211,
1689
+ "step": 2700
1690
+ },
1691
+ {
1692
+ "epoch": 0.43,
1693
+ "learning_rate": 4.288770053475936e-05,
1694
+ "loss": 0.1859,
1695
+ "step": 2710
1696
+ },
1697
+ {
1698
+ "epoch": 0.44,
1699
+ "learning_rate": 4.286096256684492e-05,
1700
+ "loss": 0.4434,
1701
+ "step": 2720
1702
+ },
1703
+ {
1704
+ "epoch": 0.44,
1705
+ "learning_rate": 4.283422459893048e-05,
1706
+ "loss": 0.2067,
1707
+ "step": 2730
1708
+ },
1709
+ {
1710
+ "epoch": 0.44,
1711
+ "learning_rate": 4.280748663101605e-05,
1712
+ "loss": 0.2656,
1713
+ "step": 2740
1714
+ },
1715
+ {
1716
+ "epoch": 0.44,
1717
+ "learning_rate": 4.2780748663101606e-05,
1718
+ "loss": 0.3386,
1719
+ "step": 2750
1720
+ },
1721
+ {
1722
+ "epoch": 0.44,
1723
+ "learning_rate": 4.275401069518717e-05,
1724
+ "loss": 0.6852,
1725
+ "step": 2760
1726
+ },
1727
+ {
1728
+ "epoch": 0.44,
1729
+ "learning_rate": 4.2727272727272724e-05,
1730
+ "loss": 0.2283,
1731
+ "step": 2770
1732
+ },
1733
+ {
1734
+ "epoch": 0.44,
1735
+ "learning_rate": 4.270053475935829e-05,
1736
+ "loss": 0.2341,
1737
+ "step": 2780
1738
+ },
1739
+ {
1740
+ "epoch": 0.45,
1741
+ "learning_rate": 4.2673796791443856e-05,
1742
+ "loss": 0.268,
1743
+ "step": 2790
1744
+ },
1745
+ {
1746
+ "epoch": 0.45,
1747
+ "learning_rate": 4.2647058823529415e-05,
1748
+ "loss": 0.2749,
1749
+ "step": 2800
1750
+ },
1751
+ {
1752
+ "epoch": 0.45,
1753
+ "learning_rate": 4.2620320855614974e-05,
1754
+ "loss": 0.2868,
1755
+ "step": 2810
1756
+ },
1757
+ {
1758
+ "epoch": 0.45,
1759
+ "learning_rate": 4.259358288770054e-05,
1760
+ "loss": 0.2351,
1761
+ "step": 2820
1762
+ },
1763
+ {
1764
+ "epoch": 0.45,
1765
+ "learning_rate": 4.25668449197861e-05,
1766
+ "loss": 0.156,
1767
+ "step": 2830
1768
+ },
1769
+ {
1770
+ "epoch": 0.45,
1771
+ "learning_rate": 4.254010695187166e-05,
1772
+ "loss": 0.1897,
1773
+ "step": 2840
1774
+ },
1775
+ {
1776
+ "epoch": 0.46,
1777
+ "learning_rate": 4.251336898395722e-05,
1778
+ "loss": 0.2117,
1779
+ "step": 2850
1780
+ },
1781
+ {
1782
+ "epoch": 0.46,
1783
+ "learning_rate": 4.248663101604278e-05,
1784
+ "loss": 0.2102,
1785
+ "step": 2860
1786
+ },
1787
+ {
1788
+ "epoch": 0.46,
1789
+ "learning_rate": 4.245989304812835e-05,
1790
+ "loss": 0.3609,
1791
+ "step": 2870
1792
+ },
1793
+ {
1794
+ "epoch": 0.46,
1795
+ "learning_rate": 4.243315508021391e-05,
1796
+ "loss": 0.2409,
1797
+ "step": 2880
1798
+ },
1799
+ {
1800
+ "epoch": 0.46,
1801
+ "learning_rate": 4.2406417112299466e-05,
1802
+ "loss": 0.5209,
1803
+ "step": 2890
1804
+ },
1805
+ {
1806
+ "epoch": 0.46,
1807
+ "learning_rate": 4.2379679144385025e-05,
1808
+ "loss": 0.3209,
1809
+ "step": 2900
1810
+ },
1811
+ {
1812
+ "epoch": 0.47,
1813
+ "learning_rate": 4.235294117647059e-05,
1814
+ "loss": 0.3782,
1815
+ "step": 2910
1816
+ },
1817
+ {
1818
+ "epoch": 0.47,
1819
+ "learning_rate": 4.232620320855615e-05,
1820
+ "loss": 0.1297,
1821
+ "step": 2920
1822
+ },
1823
+ {
1824
+ "epoch": 0.47,
1825
+ "learning_rate": 4.229946524064171e-05,
1826
+ "loss": 0.365,
1827
+ "step": 2930
1828
+ },
1829
+ {
1830
+ "epoch": 0.47,
1831
+ "learning_rate": 4.2272727272727275e-05,
1832
+ "loss": 0.4153,
1833
+ "step": 2940
1834
+ },
1835
+ {
1836
+ "epoch": 0.47,
1837
+ "learning_rate": 4.224598930481284e-05,
1838
+ "loss": 0.1112,
1839
+ "step": 2950
1840
+ },
1841
+ {
1842
+ "epoch": 0.47,
1843
+ "learning_rate": 4.221925133689839e-05,
1844
+ "loss": 0.1813,
1845
+ "step": 2960
1846
+ },
1847
+ {
1848
+ "epoch": 0.48,
1849
+ "learning_rate": 4.219251336898396e-05,
1850
+ "loss": 0.1488,
1851
+ "step": 2970
1852
+ },
1853
+ {
1854
+ "epoch": 0.48,
1855
+ "learning_rate": 4.216577540106952e-05,
1856
+ "loss": 0.1983,
1857
+ "step": 2980
1858
+ },
1859
+ {
1860
+ "epoch": 0.48,
1861
+ "learning_rate": 4.2139037433155084e-05,
1862
+ "loss": 0.3848,
1863
+ "step": 2990
1864
+ },
1865
+ {
1866
+ "epoch": 0.48,
1867
+ "learning_rate": 4.211229946524064e-05,
1868
+ "loss": 0.3005,
1869
+ "step": 3000
1870
+ },
1871
+ {
1872
+ "epoch": 0.48,
1873
+ "eval_accuracy": 0.8425226586102719,
1874
+ "eval_f1": 0.8406242841944506,
1875
+ "eval_loss": 2.313152313232422,
1876
+ "eval_precision": 0.8596586910099404,
1877
+ "eval_recall": 0.8425226586102719,
1878
+ "eval_runtime": 10.5381,
1879
+ "eval_samples_per_second": 251.278,
1880
+ "eval_steps_per_second": 3.986,
1881
+ "step": 3000
1882
+ },
1883
+ {
1884
+ "epoch": 0.48,
1885
+ "learning_rate": 4.20855614973262e-05,
1886
+ "loss": 0.3137,
1887
+ "step": 3010
1888
+ },
1889
+ {
1890
+ "epoch": 0.48,
1891
+ "learning_rate": 4.205882352941177e-05,
1892
+ "loss": 0.3145,
1893
+ "step": 3020
1894
+ },
1895
+ {
1896
+ "epoch": 0.48,
1897
+ "learning_rate": 4.2032085561497334e-05,
1898
+ "loss": 0.1838,
1899
+ "step": 3030
1900
+ },
1901
+ {
1902
+ "epoch": 0.49,
1903
+ "learning_rate": 4.2005347593582886e-05,
1904
+ "loss": 0.328,
1905
+ "step": 3040
1906
+ },
1907
+ {
1908
+ "epoch": 0.49,
1909
+ "learning_rate": 4.197860962566845e-05,
1910
+ "loss": 0.2712,
1911
+ "step": 3050
1912
+ },
1913
+ {
1914
+ "epoch": 0.49,
1915
+ "learning_rate": 4.195187165775401e-05,
1916
+ "loss": 0.2833,
1917
+ "step": 3060
1918
+ },
1919
+ {
1920
+ "epoch": 0.49,
1921
+ "learning_rate": 4.1925133689839577e-05,
1922
+ "loss": 0.425,
1923
+ "step": 3070
1924
+ },
1925
+ {
1926
+ "epoch": 0.49,
1927
+ "learning_rate": 4.1898395721925136e-05,
1928
+ "loss": 0.227,
1929
+ "step": 3080
1930
+ },
1931
+ {
1932
+ "epoch": 0.49,
1933
+ "learning_rate": 4.1871657754010695e-05,
1934
+ "loss": 0.299,
1935
+ "step": 3090
1936
+ },
1937
+ {
1938
+ "epoch": 0.5,
1939
+ "learning_rate": 4.184491978609626e-05,
1940
+ "loss": 0.3831,
1941
+ "step": 3100
1942
+ },
1943
+ {
1944
+ "epoch": 0.5,
1945
+ "learning_rate": 4.181818181818182e-05,
1946
+ "loss": 0.1673,
1947
+ "step": 3110
1948
+ },
1949
+ {
1950
+ "epoch": 0.5,
1951
+ "learning_rate": 4.179144385026738e-05,
1952
+ "loss": 0.2316,
1953
+ "step": 3120
1954
+ },
1955
+ {
1956
+ "epoch": 0.5,
1957
+ "learning_rate": 4.1764705882352944e-05,
1958
+ "loss": 0.3018,
1959
+ "step": 3130
1960
+ },
1961
+ {
1962
+ "epoch": 0.5,
1963
+ "learning_rate": 4.17379679144385e-05,
1964
+ "loss": 0.1735,
1965
+ "step": 3140
1966
+ },
1967
+ {
1968
+ "epoch": 0.5,
1969
+ "learning_rate": 4.171122994652407e-05,
1970
+ "loss": 0.3896,
1971
+ "step": 3150
1972
+ },
1973
+ {
1974
+ "epoch": 0.51,
1975
+ "learning_rate": 4.168449197860963e-05,
1976
+ "loss": 0.166,
1977
+ "step": 3160
1978
+ },
1979
+ {
1980
+ "epoch": 0.51,
1981
+ "learning_rate": 4.165775401069519e-05,
1982
+ "loss": 0.1794,
1983
+ "step": 3170
1984
+ },
1985
+ {
1986
+ "epoch": 0.51,
1987
+ "learning_rate": 4.163101604278075e-05,
1988
+ "loss": 0.2399,
1989
+ "step": 3180
1990
+ },
1991
+ {
1992
+ "epoch": 0.51,
1993
+ "learning_rate": 4.160427807486631e-05,
1994
+ "loss": 0.331,
1995
+ "step": 3190
1996
+ },
1997
+ {
1998
+ "epoch": 0.51,
1999
+ "learning_rate": 4.157754010695187e-05,
2000
+ "loss": 0.4568,
2001
+ "step": 3200
2002
+ },
2003
+ {
2004
+ "epoch": 0.51,
2005
+ "learning_rate": 4.155080213903744e-05,
2006
+ "loss": 0.4554,
2007
+ "step": 3210
2008
+ },
2009
+ {
2010
+ "epoch": 0.52,
2011
+ "learning_rate": 4.1524064171122996e-05,
2012
+ "loss": 0.3125,
2013
+ "step": 3220
2014
+ },
2015
+ {
2016
+ "epoch": 0.52,
2017
+ "learning_rate": 4.149732620320856e-05,
2018
+ "loss": 0.2818,
2019
+ "step": 3230
2020
+ },
2021
+ {
2022
+ "epoch": 0.52,
2023
+ "learning_rate": 4.147058823529412e-05,
2024
+ "loss": 0.4162,
2025
+ "step": 3240
2026
+ },
2027
+ {
2028
+ "epoch": 0.52,
2029
+ "learning_rate": 4.144385026737968e-05,
2030
+ "loss": 0.4248,
2031
+ "step": 3250
2032
+ },
2033
+ {
2034
+ "epoch": 0.52,
2035
+ "learning_rate": 4.1417112299465246e-05,
2036
+ "loss": 0.3776,
2037
+ "step": 3260
2038
+ },
2039
+ {
2040
+ "epoch": 0.52,
2041
+ "learning_rate": 4.1390374331550805e-05,
2042
+ "loss": 0.3877,
2043
+ "step": 3270
2044
+ },
2045
+ {
2046
+ "epoch": 0.52,
2047
+ "learning_rate": 4.1363636363636364e-05,
2048
+ "loss": 0.0185,
2049
+ "step": 3280
2050
+ },
2051
+ {
2052
+ "epoch": 0.53,
2053
+ "learning_rate": 4.133689839572193e-05,
2054
+ "loss": 0.1879,
2055
+ "step": 3290
2056
+ },
2057
+ {
2058
+ "epoch": 0.53,
2059
+ "learning_rate": 4.131016042780749e-05,
2060
+ "loss": 0.1577,
2061
+ "step": 3300
2062
+ },
2063
+ {
2064
+ "epoch": 0.53,
2065
+ "learning_rate": 4.128342245989305e-05,
2066
+ "loss": 0.342,
2067
+ "step": 3310
2068
+ },
2069
+ {
2070
+ "epoch": 0.53,
2071
+ "learning_rate": 4.125668449197861e-05,
2072
+ "loss": 0.2412,
2073
+ "step": 3320
2074
+ },
2075
+ {
2076
+ "epoch": 0.53,
2077
+ "learning_rate": 4.122994652406417e-05,
2078
+ "loss": 0.2567,
2079
+ "step": 3330
2080
+ },
2081
+ {
2082
+ "epoch": 0.53,
2083
+ "learning_rate": 4.120320855614974e-05,
2084
+ "loss": 0.2748,
2085
+ "step": 3340
2086
+ },
2087
+ {
2088
+ "epoch": 0.54,
2089
+ "learning_rate": 4.11764705882353e-05,
2090
+ "loss": 0.1134,
2091
+ "step": 3350
2092
+ },
2093
+ {
2094
+ "epoch": 0.54,
2095
+ "learning_rate": 4.1149732620320856e-05,
2096
+ "loss": 0.3265,
2097
+ "step": 3360
2098
+ },
2099
+ {
2100
+ "epoch": 0.54,
2101
+ "learning_rate": 4.112299465240642e-05,
2102
+ "loss": 0.2037,
2103
+ "step": 3370
2104
+ },
2105
+ {
2106
+ "epoch": 0.54,
2107
+ "learning_rate": 4.109625668449198e-05,
2108
+ "loss": 0.2894,
2109
+ "step": 3380
2110
+ },
2111
+ {
2112
+ "epoch": 0.54,
2113
+ "learning_rate": 4.106951871657754e-05,
2114
+ "loss": 0.3418,
2115
+ "step": 3390
2116
+ },
2117
+ {
2118
+ "epoch": 0.54,
2119
+ "learning_rate": 4.10427807486631e-05,
2120
+ "loss": 0.3356,
2121
+ "step": 3400
2122
+ },
2123
+ {
2124
+ "epoch": 0.55,
2125
+ "learning_rate": 4.1016042780748665e-05,
2126
+ "loss": 0.228,
2127
+ "step": 3410
2128
+ },
2129
+ {
2130
+ "epoch": 0.55,
2131
+ "learning_rate": 4.098930481283423e-05,
2132
+ "loss": 0.3992,
2133
+ "step": 3420
2134
+ },
2135
+ {
2136
+ "epoch": 0.55,
2137
+ "learning_rate": 4.096256684491978e-05,
2138
+ "loss": 0.2551,
2139
+ "step": 3430
2140
+ },
2141
+ {
2142
+ "epoch": 0.55,
2143
+ "learning_rate": 4.093582887700535e-05,
2144
+ "loss": 0.184,
2145
+ "step": 3440
2146
+ },
2147
+ {
2148
+ "epoch": 0.55,
2149
+ "learning_rate": 4.0909090909090915e-05,
2150
+ "loss": 0.3881,
2151
+ "step": 3450
2152
+ },
2153
+ {
2154
+ "epoch": 0.55,
2155
+ "learning_rate": 4.0882352941176474e-05,
2156
+ "loss": 0.3422,
2157
+ "step": 3460
2158
+ },
2159
+ {
2160
+ "epoch": 0.56,
2161
+ "learning_rate": 4.085561497326203e-05,
2162
+ "loss": 0.2786,
2163
+ "step": 3470
2164
+ },
2165
+ {
2166
+ "epoch": 0.56,
2167
+ "learning_rate": 4.082887700534759e-05,
2168
+ "loss": 0.2854,
2169
+ "step": 3480
2170
+ },
2171
+ {
2172
+ "epoch": 0.56,
2173
+ "learning_rate": 4.080213903743316e-05,
2174
+ "loss": 0.4252,
2175
+ "step": 3490
2176
+ },
2177
+ {
2178
+ "epoch": 0.56,
2179
+ "learning_rate": 4.0775401069518723e-05,
2180
+ "loss": 0.1091,
2181
+ "step": 3500
2182
+ },
2183
+ {
2184
+ "epoch": 0.56,
2185
+ "eval_accuracy": 0.8459214501510574,
2186
+ "eval_f1": 0.843780419456653,
2187
+ "eval_loss": 2.0897836685180664,
2188
+ "eval_precision": 0.8659851431247721,
2189
+ "eval_recall": 0.8459214501510574,
2190
+ "eval_runtime": 10.5263,
2191
+ "eval_samples_per_second": 251.559,
2192
+ "eval_steps_per_second": 3.99,
2193
+ "step": 3500
2194
+ },
2195
+ {
2196
+ "epoch": 0.56,
2197
+ "learning_rate": 4.0748663101604276e-05,
2198
+ "loss": 0.3208,
2199
+ "step": 3510
2200
+ },
2201
+ {
2202
+ "epoch": 0.56,
2203
+ "learning_rate": 4.072192513368984e-05,
2204
+ "loss": 0.0881,
2205
+ "step": 3520
2206
+ },
2207
+ {
2208
+ "epoch": 0.56,
2209
+ "learning_rate": 4.06951871657754e-05,
2210
+ "loss": 0.1677,
2211
+ "step": 3530
2212
+ },
2213
+ {
2214
+ "epoch": 0.57,
2215
+ "learning_rate": 4.0668449197860966e-05,
2216
+ "loss": 0.049,
2217
+ "step": 3540
2218
+ },
2219
+ {
2220
+ "epoch": 0.57,
2221
+ "learning_rate": 4.0641711229946525e-05,
2222
+ "loss": 0.1629,
2223
+ "step": 3550
2224
+ },
2225
+ {
2226
+ "epoch": 0.57,
2227
+ "learning_rate": 4.0614973262032084e-05,
2228
+ "loss": 0.2262,
2229
+ "step": 3560
2230
+ },
2231
+ {
2232
+ "epoch": 0.57,
2233
+ "learning_rate": 4.058823529411765e-05,
2234
+ "loss": 0.2134,
2235
+ "step": 3570
2236
+ },
2237
+ {
2238
+ "epoch": 0.57,
2239
+ "learning_rate": 4.0561497326203216e-05,
2240
+ "loss": 0.3984,
2241
+ "step": 3580
2242
+ },
2243
+ {
2244
+ "epoch": 0.57,
2245
+ "learning_rate": 4.053475935828877e-05,
2246
+ "loss": 0.1527,
2247
+ "step": 3590
2248
+ },
2249
+ {
2250
+ "epoch": 0.58,
2251
+ "learning_rate": 4.0508021390374334e-05,
2252
+ "loss": 0.1532,
2253
+ "step": 3600
2254
+ },
2255
+ {
2256
+ "epoch": 0.58,
2257
+ "learning_rate": 4.048128342245989e-05,
2258
+ "loss": 0.1918,
2259
+ "step": 3610
2260
+ },
2261
+ {
2262
+ "epoch": 0.58,
2263
+ "learning_rate": 4.045454545454546e-05,
2264
+ "loss": 0.2594,
2265
+ "step": 3620
2266
+ },
2267
+ {
2268
+ "epoch": 0.58,
2269
+ "learning_rate": 4.042780748663102e-05,
2270
+ "loss": 0.1803,
2271
+ "step": 3630
2272
+ },
2273
+ {
2274
+ "epoch": 0.58,
2275
+ "learning_rate": 4.040106951871658e-05,
2276
+ "loss": 0.2573,
2277
+ "step": 3640
2278
+ },
2279
+ {
2280
+ "epoch": 0.58,
2281
+ "learning_rate": 4.037433155080214e-05,
2282
+ "loss": 0.2865,
2283
+ "step": 3650
2284
+ },
2285
+ {
2286
+ "epoch": 0.59,
2287
+ "learning_rate": 4.03475935828877e-05,
2288
+ "loss": 0.053,
2289
+ "step": 3660
2290
+ },
2291
+ {
2292
+ "epoch": 0.59,
2293
+ "learning_rate": 4.032085561497326e-05,
2294
+ "loss": 0.4603,
2295
+ "step": 3670
2296
+ },
2297
+ {
2298
+ "epoch": 0.59,
2299
+ "learning_rate": 4.029411764705883e-05,
2300
+ "loss": 0.1409,
2301
+ "step": 3680
2302
+ },
2303
+ {
2304
+ "epoch": 0.59,
2305
+ "learning_rate": 4.0267379679144386e-05,
2306
+ "loss": 0.2257,
2307
+ "step": 3690
2308
+ },
2309
+ {
2310
+ "epoch": 0.59,
2311
+ "learning_rate": 4.024064171122995e-05,
2312
+ "loss": 0.2827,
2313
+ "step": 3700
2314
+ },
2315
+ {
2316
+ "epoch": 0.59,
2317
+ "learning_rate": 4.021390374331551e-05,
2318
+ "loss": 0.2166,
2319
+ "step": 3710
2320
+ },
2321
+ {
2322
+ "epoch": 0.6,
2323
+ "learning_rate": 4.018716577540107e-05,
2324
+ "loss": 0.2091,
2325
+ "step": 3720
2326
+ },
2327
+ {
2328
+ "epoch": 0.6,
2329
+ "learning_rate": 4.0160427807486636e-05,
2330
+ "loss": 0.4851,
2331
+ "step": 3730
2332
+ },
2333
+ {
2334
+ "epoch": 0.6,
2335
+ "learning_rate": 4.0133689839572195e-05,
2336
+ "loss": 0.3742,
2337
+ "step": 3740
2338
+ },
2339
+ {
2340
+ "epoch": 0.6,
2341
+ "learning_rate": 4.0106951871657754e-05,
2342
+ "loss": 0.492,
2343
+ "step": 3750
2344
+ },
2345
+ {
2346
+ "epoch": 0.6,
2347
+ "learning_rate": 4.008021390374332e-05,
2348
+ "loss": 0.5621,
2349
+ "step": 3760
2350
+ },
2351
+ {
2352
+ "epoch": 0.6,
2353
+ "learning_rate": 4.005347593582888e-05,
2354
+ "loss": 0.2801,
2355
+ "step": 3770
2356
+ },
2357
+ {
2358
+ "epoch": 0.6,
2359
+ "learning_rate": 4.002673796791444e-05,
2360
+ "loss": 0.2604,
2361
+ "step": 3780
2362
+ },
2363
+ {
2364
+ "epoch": 0.61,
2365
+ "learning_rate": 4e-05,
2366
+ "loss": 0.3608,
2367
+ "step": 3790
2368
+ },
2369
+ {
2370
+ "epoch": 0.61,
2371
+ "learning_rate": 3.997326203208556e-05,
2372
+ "loss": 0.167,
2373
+ "step": 3800
2374
+ },
2375
+ {
2376
+ "epoch": 0.61,
2377
+ "learning_rate": 3.994652406417113e-05,
2378
+ "loss": 0.4472,
2379
+ "step": 3810
2380
+ },
2381
+ {
2382
+ "epoch": 0.61,
2383
+ "learning_rate": 3.991978609625669e-05,
2384
+ "loss": 0.3022,
2385
+ "step": 3820
2386
+ },
2387
+ {
2388
+ "epoch": 0.61,
2389
+ "learning_rate": 3.9893048128342246e-05,
2390
+ "loss": 0.3179,
2391
+ "step": 3830
2392
+ },
2393
+ {
2394
+ "epoch": 0.61,
2395
+ "learning_rate": 3.986631016042781e-05,
2396
+ "loss": 0.2939,
2397
+ "step": 3840
2398
+ },
2399
+ {
2400
+ "epoch": 0.62,
2401
+ "learning_rate": 3.983957219251337e-05,
2402
+ "loss": 0.013,
2403
+ "step": 3850
2404
+ },
2405
+ {
2406
+ "epoch": 0.62,
2407
+ "learning_rate": 3.981283422459893e-05,
2408
+ "loss": 0.1864,
2409
+ "step": 3860
2410
+ },
2411
+ {
2412
+ "epoch": 0.62,
2413
+ "learning_rate": 3.9786096256684496e-05,
2414
+ "loss": 0.3368,
2415
+ "step": 3870
2416
+ },
2417
+ {
2418
+ "epoch": 0.62,
2419
+ "learning_rate": 3.9759358288770055e-05,
2420
+ "loss": 0.1275,
2421
+ "step": 3880
2422
+ },
2423
+ {
2424
+ "epoch": 0.62,
2425
+ "learning_rate": 3.973262032085562e-05,
2426
+ "loss": 0.1441,
2427
+ "step": 3890
2428
+ },
2429
+ {
2430
+ "epoch": 0.62,
2431
+ "learning_rate": 3.970588235294117e-05,
2432
+ "loss": 0.358,
2433
+ "step": 3900
2434
+ },
2435
+ {
2436
+ "epoch": 0.63,
2437
+ "learning_rate": 3.967914438502674e-05,
2438
+ "loss": 0.1918,
2439
+ "step": 3910
2440
+ },
2441
+ {
2442
+ "epoch": 0.63,
2443
+ "learning_rate": 3.9652406417112305e-05,
2444
+ "loss": 0.2604,
2445
+ "step": 3920
2446
+ },
2447
+ {
2448
+ "epoch": 0.63,
2449
+ "learning_rate": 3.9625668449197864e-05,
2450
+ "loss": 0.0787,
2451
+ "step": 3930
2452
+ },
2453
+ {
2454
+ "epoch": 0.63,
2455
+ "learning_rate": 3.959893048128342e-05,
2456
+ "loss": 0.2841,
2457
+ "step": 3940
2458
+ },
2459
+ {
2460
+ "epoch": 0.63,
2461
+ "learning_rate": 3.957219251336899e-05,
2462
+ "loss": 0.4061,
2463
+ "step": 3950
2464
+ },
2465
+ {
2466
+ "epoch": 0.63,
2467
+ "learning_rate": 3.954545454545455e-05,
2468
+ "loss": 0.3767,
2469
+ "step": 3960
2470
+ },
2471
+ {
2472
+ "epoch": 0.64,
2473
+ "learning_rate": 3.951871657754011e-05,
2474
+ "loss": 0.1036,
2475
+ "step": 3970
2476
+ },
2477
+ {
2478
+ "epoch": 0.64,
2479
+ "learning_rate": 3.9491978609625666e-05,
2480
+ "loss": 0.38,
2481
+ "step": 3980
2482
+ },
2483
+ {
2484
+ "epoch": 0.64,
2485
+ "learning_rate": 3.946524064171123e-05,
2486
+ "loss": 0.1721,
2487
+ "step": 3990
2488
+ },
2489
+ {
2490
+ "epoch": 0.64,
2491
+ "learning_rate": 3.94385026737968e-05,
2492
+ "loss": 0.2176,
2493
+ "step": 4000
2494
+ },
2495
+ {
2496
+ "epoch": 0.64,
2497
+ "eval_accuracy": 0.847809667673716,
2498
+ "eval_f1": 0.8461258728150335,
2499
+ "eval_loss": 2.1692090034484863,
2500
+ "eval_precision": 0.8637304048866307,
2501
+ "eval_recall": 0.847809667673716,
2502
+ "eval_runtime": 10.5307,
2503
+ "eval_samples_per_second": 251.455,
2504
+ "eval_steps_per_second": 3.988,
2505
+ "step": 4000
2506
+ },
2507
+ {
2508
+ "epoch": 0.64,
2509
+ "learning_rate": 3.9411764705882356e-05,
2510
+ "loss": 0.1731,
2511
+ "step": 4010
2512
+ },
2513
+ {
2514
+ "epoch": 0.64,
2515
+ "learning_rate": 3.9385026737967915e-05,
2516
+ "loss": 0.0994,
2517
+ "step": 4020
2518
+ },
2519
+ {
2520
+ "epoch": 0.64,
2521
+ "learning_rate": 3.9358288770053474e-05,
2522
+ "loss": 0.3668,
2523
+ "step": 4030
2524
+ },
2525
+ {
2526
+ "epoch": 0.65,
2527
+ "learning_rate": 3.933155080213904e-05,
2528
+ "loss": 0.6184,
2529
+ "step": 4040
2530
+ },
2531
+ {
2532
+ "epoch": 0.65,
2533
+ "learning_rate": 3.93048128342246e-05,
2534
+ "loss": 0.1739,
2535
+ "step": 4050
2536
+ },
2537
+ {
2538
+ "epoch": 0.65,
2539
+ "learning_rate": 3.927807486631016e-05,
2540
+ "loss": 0.3273,
2541
+ "step": 4060
2542
+ },
2543
+ {
2544
+ "epoch": 0.65,
2545
+ "learning_rate": 3.9251336898395724e-05,
2546
+ "loss": 0.2285,
2547
+ "step": 4070
2548
+ },
2549
+ {
2550
+ "epoch": 0.65,
2551
+ "learning_rate": 3.922459893048129e-05,
2552
+ "loss": 0.2541,
2553
+ "step": 4080
2554
+ },
2555
+ {
2556
+ "epoch": 0.65,
2557
+ "learning_rate": 3.919786096256685e-05,
2558
+ "loss": 0.3116,
2559
+ "step": 4090
2560
+ },
2561
+ {
2562
+ "epoch": 0.66,
2563
+ "learning_rate": 3.917112299465241e-05,
2564
+ "loss": 0.4674,
2565
+ "step": 4100
2566
+ },
2567
+ {
2568
+ "epoch": 0.66,
2569
+ "learning_rate": 3.914438502673797e-05,
2570
+ "loss": 0.2002,
2571
+ "step": 4110
2572
+ },
2573
+ {
2574
+ "epoch": 0.66,
2575
+ "learning_rate": 3.911764705882353e-05,
2576
+ "loss": 0.3206,
2577
+ "step": 4120
2578
+ },
2579
+ {
2580
+ "epoch": 0.66,
2581
+ "learning_rate": 3.909090909090909e-05,
2582
+ "loss": 0.4748,
2583
+ "step": 4130
2584
+ },
2585
+ {
2586
+ "epoch": 0.66,
2587
+ "learning_rate": 3.906417112299465e-05,
2588
+ "loss": 0.1869,
2589
+ "step": 4140
2590
+ },
2591
+ {
2592
+ "epoch": 0.66,
2593
+ "learning_rate": 3.903743315508022e-05,
2594
+ "loss": 0.1935,
2595
+ "step": 4150
2596
+ },
2597
+ {
2598
+ "epoch": 0.67,
2599
+ "learning_rate": 3.901069518716578e-05,
2600
+ "loss": 0.2528,
2601
+ "step": 4160
2602
+ },
2603
+ {
2604
+ "epoch": 0.67,
2605
+ "learning_rate": 3.8983957219251335e-05,
2606
+ "loss": 0.275,
2607
+ "step": 4170
2608
+ },
2609
+ {
2610
+ "epoch": 0.67,
2611
+ "learning_rate": 3.89572192513369e-05,
2612
+ "loss": 0.2676,
2613
+ "step": 4180
2614
+ },
2615
+ {
2616
+ "epoch": 0.67,
2617
+ "learning_rate": 3.893048128342246e-05,
2618
+ "loss": 0.3452,
2619
+ "step": 4190
2620
+ },
2621
+ {
2622
+ "epoch": 0.67,
2623
+ "learning_rate": 3.8903743315508025e-05,
2624
+ "loss": 0.2249,
2625
+ "step": 4200
2626
+ },
2627
+ {
2628
+ "epoch": 0.67,
2629
+ "learning_rate": 3.8877005347593584e-05,
2630
+ "loss": 0.3211,
2631
+ "step": 4210
2632
+ },
2633
+ {
2634
+ "epoch": 0.68,
2635
+ "learning_rate": 3.8850267379679143e-05,
2636
+ "loss": 0.2125,
2637
+ "step": 4220
2638
+ },
2639
+ {
2640
+ "epoch": 0.68,
2641
+ "learning_rate": 3.882352941176471e-05,
2642
+ "loss": 0.2449,
2643
+ "step": 4230
2644
+ },
2645
+ {
2646
+ "epoch": 0.68,
2647
+ "learning_rate": 3.879679144385027e-05,
2648
+ "loss": 0.1108,
2649
+ "step": 4240
2650
+ },
2651
+ {
2652
+ "epoch": 0.68,
2653
+ "learning_rate": 3.877005347593583e-05,
2654
+ "loss": 0.3127,
2655
+ "step": 4250
2656
+ },
2657
+ {
2658
+ "epoch": 0.68,
2659
+ "learning_rate": 3.874331550802139e-05,
2660
+ "loss": 0.1238,
2661
+ "step": 4260
2662
+ },
2663
+ {
2664
+ "epoch": 0.68,
2665
+ "learning_rate": 3.871657754010695e-05,
2666
+ "loss": 0.2596,
2667
+ "step": 4270
2668
+ },
2669
+ {
2670
+ "epoch": 0.68,
2671
+ "learning_rate": 3.868983957219252e-05,
2672
+ "loss": 0.36,
2673
+ "step": 4280
2674
+ },
2675
+ {
2676
+ "epoch": 0.69,
2677
+ "learning_rate": 3.866310160427808e-05,
2678
+ "loss": 0.2548,
2679
+ "step": 4290
2680
+ },
2681
+ {
2682
+ "epoch": 0.69,
2683
+ "learning_rate": 3.8636363636363636e-05,
2684
+ "loss": 0.1942,
2685
+ "step": 4300
2686
+ },
2687
+ {
2688
+ "epoch": 0.69,
2689
+ "learning_rate": 3.86096256684492e-05,
2690
+ "loss": 0.1626,
2691
+ "step": 4310
2692
+ },
2693
+ {
2694
+ "epoch": 0.69,
2695
+ "learning_rate": 3.858288770053476e-05,
2696
+ "loss": 0.1887,
2697
+ "step": 4320
2698
+ },
2699
+ {
2700
+ "epoch": 0.69,
2701
+ "learning_rate": 3.855614973262032e-05,
2702
+ "loss": 0.3538,
2703
+ "step": 4330
2704
+ },
2705
+ {
2706
+ "epoch": 0.69,
2707
+ "learning_rate": 3.8529411764705886e-05,
2708
+ "loss": 0.2309,
2709
+ "step": 4340
2710
+ },
2711
+ {
2712
+ "epoch": 0.7,
2713
+ "learning_rate": 3.8502673796791445e-05,
2714
+ "loss": 0.147,
2715
+ "step": 4350
2716
+ },
2717
+ {
2718
+ "epoch": 0.7,
2719
+ "learning_rate": 3.847593582887701e-05,
2720
+ "loss": 0.1799,
2721
+ "step": 4360
2722
+ },
2723
+ {
2724
+ "epoch": 0.7,
2725
+ "learning_rate": 3.844919786096257e-05,
2726
+ "loss": 0.2291,
2727
+ "step": 4370
2728
+ },
2729
+ {
2730
+ "epoch": 0.7,
2731
+ "learning_rate": 3.842245989304813e-05,
2732
+ "loss": 0.39,
2733
+ "step": 4380
2734
+ },
2735
+ {
2736
+ "epoch": 0.7,
2737
+ "learning_rate": 3.8395721925133694e-05,
2738
+ "loss": 0.0407,
2739
+ "step": 4390
2740
+ },
2741
+ {
2742
+ "epoch": 0.7,
2743
+ "learning_rate": 3.8368983957219254e-05,
2744
+ "loss": 0.2575,
2745
+ "step": 4400
2746
+ },
2747
+ {
2748
+ "epoch": 0.71,
2749
+ "learning_rate": 3.834224598930481e-05,
2750
+ "loss": 0.1994,
2751
+ "step": 4410
2752
+ },
2753
+ {
2754
+ "epoch": 0.71,
2755
+ "learning_rate": 3.831550802139038e-05,
2756
+ "loss": 0.1931,
2757
+ "step": 4420
2758
+ },
2759
+ {
2760
+ "epoch": 0.71,
2761
+ "learning_rate": 3.828877005347594e-05,
2762
+ "loss": 0.4772,
2763
+ "step": 4430
2764
+ },
2765
+ {
2766
+ "epoch": 0.71,
2767
+ "learning_rate": 3.82620320855615e-05,
2768
+ "loss": 0.406,
2769
+ "step": 4440
2770
+ },
2771
+ {
2772
+ "epoch": 0.71,
2773
+ "learning_rate": 3.8235294117647055e-05,
2774
+ "loss": 0.3865,
2775
+ "step": 4450
2776
+ },
2777
+ {
2778
+ "epoch": 0.71,
2779
+ "learning_rate": 3.820855614973262e-05,
2780
+ "loss": 0.2322,
2781
+ "step": 4460
2782
+ },
2783
+ {
2784
+ "epoch": 0.72,
2785
+ "learning_rate": 3.818181818181819e-05,
2786
+ "loss": 0.3505,
2787
+ "step": 4470
2788
+ },
2789
+ {
2790
+ "epoch": 0.72,
2791
+ "learning_rate": 3.8155080213903746e-05,
2792
+ "loss": 0.2413,
2793
+ "step": 4480
2794
+ },
2795
+ {
2796
+ "epoch": 0.72,
2797
+ "learning_rate": 3.8128342245989305e-05,
2798
+ "loss": 0.1657,
2799
+ "step": 4490
2800
+ },
2801
+ {
2802
+ "epoch": 0.72,
2803
+ "learning_rate": 3.810160427807487e-05,
2804
+ "loss": 0.1567,
2805
+ "step": 4500
2806
+ },
2807
+ {
2808
+ "epoch": 0.72,
2809
+ "eval_accuracy": 0.8462990936555891,
2810
+ "eval_f1": 0.8442050087782637,
2811
+ "eval_loss": 2.184410333633423,
2812
+ "eval_precision": 0.8659758356860412,
2813
+ "eval_recall": 0.8462990936555891,
2814
+ "eval_runtime": 10.5461,
2815
+ "eval_samples_per_second": 251.088,
2816
+ "eval_steps_per_second": 3.983,
2817
+ "step": 4500
2818
+ },
2819
+ {
2820
+ "epoch": 0.72,
2821
+ "learning_rate": 3.807486631016043e-05,
2822
+ "loss": 0.2413,
2823
+ "step": 4510
2824
+ },
2825
+ {
2826
+ "epoch": 0.72,
2827
+ "learning_rate": 3.804812834224599e-05,
2828
+ "loss": 0.5433,
2829
+ "step": 4520
2830
+ },
2831
+ {
2832
+ "epoch": 0.72,
2833
+ "learning_rate": 3.802139037433155e-05,
2834
+ "loss": 0.326,
2835
+ "step": 4530
2836
+ },
2837
+ {
2838
+ "epoch": 0.73,
2839
+ "learning_rate": 3.7994652406417114e-05,
2840
+ "loss": 0.3755,
2841
+ "step": 4540
2842
+ },
2843
+ {
2844
+ "epoch": 0.73,
2845
+ "learning_rate": 3.796791443850268e-05,
2846
+ "loss": 0.2589,
2847
+ "step": 4550
2848
+ },
2849
+ {
2850
+ "epoch": 0.73,
2851
+ "learning_rate": 3.794117647058824e-05,
2852
+ "loss": 0.2539,
2853
+ "step": 4560
2854
+ },
2855
+ {
2856
+ "epoch": 0.73,
2857
+ "learning_rate": 3.79144385026738e-05,
2858
+ "loss": 0.2227,
2859
+ "step": 4570
2860
+ },
2861
+ {
2862
+ "epoch": 0.73,
2863
+ "learning_rate": 3.7887700534759364e-05,
2864
+ "loss": 0.1897,
2865
+ "step": 4580
2866
+ },
2867
+ {
2868
+ "epoch": 0.73,
2869
+ "learning_rate": 3.786096256684492e-05,
2870
+ "loss": 0.2462,
2871
+ "step": 4590
2872
+ },
2873
+ {
2874
+ "epoch": 0.74,
2875
+ "learning_rate": 3.783422459893048e-05,
2876
+ "loss": 0.1308,
2877
+ "step": 4600
2878
+ },
2879
+ {
2880
+ "epoch": 0.74,
2881
+ "learning_rate": 3.780748663101604e-05,
2882
+ "loss": 0.2567,
2883
+ "step": 4610
2884
+ },
2885
+ {
2886
+ "epoch": 0.74,
2887
+ "learning_rate": 3.7780748663101607e-05,
2888
+ "loss": 0.3583,
2889
+ "step": 4620
2890
+ },
2891
+ {
2892
+ "epoch": 0.74,
2893
+ "learning_rate": 3.775401069518717e-05,
2894
+ "loss": 0.3253,
2895
+ "step": 4630
2896
+ },
2897
+ {
2898
+ "epoch": 0.74,
2899
+ "learning_rate": 3.7727272727272725e-05,
2900
+ "loss": 0.2606,
2901
+ "step": 4640
2902
+ },
2903
+ {
2904
+ "epoch": 0.74,
2905
+ "learning_rate": 3.770053475935829e-05,
2906
+ "loss": 0.2173,
2907
+ "step": 4650
2908
+ },
2909
+ {
2910
+ "epoch": 0.75,
2911
+ "learning_rate": 3.767379679144385e-05,
2912
+ "loss": 0.2203,
2913
+ "step": 4660
2914
+ },
2915
+ {
2916
+ "epoch": 0.75,
2917
+ "learning_rate": 3.7647058823529415e-05,
2918
+ "loss": 0.059,
2919
+ "step": 4670
2920
+ },
2921
+ {
2922
+ "epoch": 0.75,
2923
+ "learning_rate": 3.7620320855614974e-05,
2924
+ "loss": 0.3312,
2925
+ "step": 4680
2926
+ },
2927
+ {
2928
+ "epoch": 0.75,
2929
+ "learning_rate": 3.759358288770053e-05,
2930
+ "loss": 0.1597,
2931
+ "step": 4690
2932
+ },
2933
+ {
2934
+ "epoch": 0.75,
2935
+ "learning_rate": 3.75668449197861e-05,
2936
+ "loss": 0.3424,
2937
+ "step": 4700
2938
+ },
2939
+ {
2940
+ "epoch": 0.75,
2941
+ "learning_rate": 3.7540106951871665e-05,
2942
+ "loss": 0.2294,
2943
+ "step": 4710
2944
+ },
2945
+ {
2946
+ "epoch": 0.76,
2947
+ "learning_rate": 3.751336898395722e-05,
2948
+ "loss": 0.1233,
2949
+ "step": 4720
2950
+ },
2951
+ {
2952
+ "epoch": 0.76,
2953
+ "learning_rate": 3.748663101604278e-05,
2954
+ "loss": 0.1573,
2955
+ "step": 4730
2956
+ },
2957
+ {
2958
+ "epoch": 0.76,
2959
+ "learning_rate": 3.745989304812834e-05,
2960
+ "loss": 0.0886,
2961
+ "step": 4740
2962
+ },
2963
+ {
2964
+ "epoch": 0.76,
2965
+ "learning_rate": 3.743315508021391e-05,
2966
+ "loss": 0.1607,
2967
+ "step": 4750
2968
+ },
2969
+ {
2970
+ "epoch": 0.76,
2971
+ "learning_rate": 3.740641711229947e-05,
2972
+ "loss": 0.2734,
2973
+ "step": 4760
2974
+ },
2975
+ {
2976
+ "epoch": 0.76,
2977
+ "learning_rate": 3.7379679144385026e-05,
2978
+ "loss": 0.1011,
2979
+ "step": 4770
2980
+ },
2981
+ {
2982
+ "epoch": 0.76,
2983
+ "learning_rate": 3.735294117647059e-05,
2984
+ "loss": 0.395,
2985
+ "step": 4780
2986
+ },
2987
+ {
2988
+ "epoch": 0.77,
2989
+ "learning_rate": 3.732620320855615e-05,
2990
+ "loss": 0.225,
2991
+ "step": 4790
2992
+ },
2993
+ {
2994
+ "epoch": 0.77,
2995
+ "learning_rate": 3.729946524064171e-05,
2996
+ "loss": 0.3241,
2997
+ "step": 4800
2998
+ },
2999
+ {
3000
+ "epoch": 0.77,
3001
+ "learning_rate": 3.7272727272727276e-05,
3002
+ "loss": 0.3954,
3003
+ "step": 4810
3004
+ },
3005
+ {
3006
+ "epoch": 0.77,
3007
+ "learning_rate": 3.7245989304812835e-05,
3008
+ "loss": 0.1394,
3009
+ "step": 4820
3010
+ },
3011
+ {
3012
+ "epoch": 0.77,
3013
+ "learning_rate": 3.72192513368984e-05,
3014
+ "loss": 0.2615,
3015
+ "step": 4830
3016
+ },
3017
+ {
3018
+ "epoch": 0.77,
3019
+ "learning_rate": 3.719251336898396e-05,
3020
+ "loss": 0.1806,
3021
+ "step": 4840
3022
+ },
3023
+ {
3024
+ "epoch": 0.78,
3025
+ "learning_rate": 3.716577540106952e-05,
3026
+ "loss": 0.2261,
3027
+ "step": 4850
3028
+ },
3029
+ {
3030
+ "epoch": 0.78,
3031
+ "learning_rate": 3.7139037433155084e-05,
3032
+ "loss": 0.3843,
3033
+ "step": 4860
3034
+ },
3035
+ {
3036
+ "epoch": 0.78,
3037
+ "learning_rate": 3.7112299465240643e-05,
3038
+ "loss": 0.3096,
3039
+ "step": 4870
3040
+ },
3041
+ {
3042
+ "epoch": 0.78,
3043
+ "learning_rate": 3.70855614973262e-05,
3044
+ "loss": 0.2883,
3045
+ "step": 4880
3046
+ },
3047
+ {
3048
+ "epoch": 0.78,
3049
+ "learning_rate": 3.705882352941177e-05,
3050
+ "loss": 0.3197,
3051
+ "step": 4890
3052
+ },
3053
+ {
3054
+ "epoch": 0.78,
3055
+ "learning_rate": 3.703208556149733e-05,
3056
+ "loss": 0.5345,
3057
+ "step": 4900
3058
+ },
3059
+ {
3060
+ "epoch": 0.79,
3061
+ "learning_rate": 3.700534759358289e-05,
3062
+ "loss": 0.3752,
3063
+ "step": 4910
3064
+ },
3065
+ {
3066
+ "epoch": 0.79,
3067
+ "learning_rate": 3.697860962566845e-05,
3068
+ "loss": 0.2157,
3069
+ "step": 4920
3070
+ },
3071
+ {
3072
+ "epoch": 0.79,
3073
+ "learning_rate": 3.695187165775401e-05,
3074
+ "loss": 0.1772,
3075
+ "step": 4930
3076
+ },
3077
+ {
3078
+ "epoch": 0.79,
3079
+ "learning_rate": 3.692513368983958e-05,
3080
+ "loss": 0.2575,
3081
+ "step": 4940
3082
+ },
3083
+ {
3084
+ "epoch": 0.79,
3085
+ "learning_rate": 3.6898395721925136e-05,
3086
+ "loss": 0.2138,
3087
+ "step": 4950
3088
+ },
3089
+ {
3090
+ "epoch": 0.79,
3091
+ "learning_rate": 3.6871657754010695e-05,
3092
+ "loss": 0.2703,
3093
+ "step": 4960
3094
+ },
3095
+ {
3096
+ "epoch": 0.8,
3097
+ "learning_rate": 3.684491978609626e-05,
3098
+ "loss": 0.4287,
3099
+ "step": 4970
3100
+ },
3101
+ {
3102
+ "epoch": 0.8,
3103
+ "learning_rate": 3.681818181818182e-05,
3104
+ "loss": 0.1396,
3105
+ "step": 4980
3106
+ },
3107
+ {
3108
+ "epoch": 0.8,
3109
+ "learning_rate": 3.679144385026738e-05,
3110
+ "loss": 0.3636,
3111
+ "step": 4990
3112
+ },
3113
+ {
3114
+ "epoch": 0.8,
3115
+ "learning_rate": 3.6764705882352945e-05,
3116
+ "loss": 0.3738,
3117
+ "step": 5000
3118
+ },
3119
+ {
3120
+ "epoch": 0.8,
3121
+ "eval_accuracy": 0.8466767371601208,
3122
+ "eval_f1": 0.8446831969931973,
3123
+ "eval_loss": 2.3107666969299316,
3124
+ "eval_precision": 0.8654388277401478,
3125
+ "eval_recall": 0.8466767371601208,
3126
+ "eval_runtime": 10.5417,
3127
+ "eval_samples_per_second": 251.194,
3128
+ "eval_steps_per_second": 3.984,
3129
+ "step": 5000
3130
+ },
3131
+ {
3132
+ "epoch": 0.8,
3133
+ "learning_rate": 3.6737967914438504e-05,
3134
+ "loss": 0.3446,
3135
+ "step": 5010
3136
+ },
3137
+ {
3138
+ "epoch": 0.8,
3139
+ "learning_rate": 3.671122994652407e-05,
3140
+ "loss": 0.4857,
3141
+ "step": 5020
3142
+ },
3143
+ {
3144
+ "epoch": 0.8,
3145
+ "learning_rate": 3.668449197860963e-05,
3146
+ "loss": 0.2666,
3147
+ "step": 5030
3148
+ },
3149
+ {
3150
+ "epoch": 0.81,
3151
+ "learning_rate": 3.665775401069519e-05,
3152
+ "loss": 0.4162,
3153
+ "step": 5040
3154
+ },
3155
+ {
3156
+ "epoch": 0.81,
3157
+ "learning_rate": 3.6631016042780753e-05,
3158
+ "loss": 0.1304,
3159
+ "step": 5050
3160
+ },
3161
+ {
3162
+ "epoch": 0.81,
3163
+ "learning_rate": 3.660427807486631e-05,
3164
+ "loss": 0.1575,
3165
+ "step": 5060
3166
+ },
3167
+ {
3168
+ "epoch": 0.81,
3169
+ "learning_rate": 3.657754010695187e-05,
3170
+ "loss": 0.1837,
3171
+ "step": 5070
3172
+ },
3173
+ {
3174
+ "epoch": 0.81,
3175
+ "learning_rate": 3.655080213903743e-05,
3176
+ "loss": 0.4065,
3177
+ "step": 5080
3178
+ },
3179
+ {
3180
+ "epoch": 0.81,
3181
+ "learning_rate": 3.6524064171122996e-05,
3182
+ "loss": 0.3935,
3183
+ "step": 5090
3184
+ },
3185
+ {
3186
+ "epoch": 0.82,
3187
+ "learning_rate": 3.649732620320856e-05,
3188
+ "loss": 0.1707,
3189
+ "step": 5100
3190
+ },
3191
+ {
3192
+ "epoch": 0.82,
3193
+ "learning_rate": 3.6470588235294114e-05,
3194
+ "loss": 0.0913,
3195
+ "step": 5110
3196
+ },
3197
+ {
3198
+ "epoch": 0.82,
3199
+ "learning_rate": 3.644385026737968e-05,
3200
+ "loss": 0.2487,
3201
+ "step": 5120
3202
+ },
3203
+ {
3204
+ "epoch": 0.82,
3205
+ "learning_rate": 3.6417112299465246e-05,
3206
+ "loss": 0.092,
3207
+ "step": 5130
3208
+ },
3209
+ {
3210
+ "epoch": 0.82,
3211
+ "learning_rate": 3.6390374331550805e-05,
3212
+ "loss": 0.2974,
3213
+ "step": 5140
3214
+ },
3215
+ {
3216
+ "epoch": 0.82,
3217
+ "learning_rate": 3.6363636363636364e-05,
3218
+ "loss": 0.2422,
3219
+ "step": 5150
3220
+ },
3221
+ {
3222
+ "epoch": 0.83,
3223
+ "learning_rate": 3.633689839572192e-05,
3224
+ "loss": 0.2698,
3225
+ "step": 5160
3226
+ },
3227
+ {
3228
+ "epoch": 0.83,
3229
+ "learning_rate": 3.631016042780749e-05,
3230
+ "loss": 0.2058,
3231
+ "step": 5170
3232
+ },
3233
+ {
3234
+ "epoch": 0.83,
3235
+ "learning_rate": 3.6283422459893055e-05,
3236
+ "loss": 0.3773,
3237
+ "step": 5180
3238
+ },
3239
+ {
3240
+ "epoch": 0.83,
3241
+ "learning_rate": 3.625668449197861e-05,
3242
+ "loss": 0.3958,
3243
+ "step": 5190
3244
+ },
3245
+ {
3246
+ "epoch": 0.83,
3247
+ "learning_rate": 3.622994652406417e-05,
3248
+ "loss": 0.2811,
3249
+ "step": 5200
3250
+ },
3251
+ {
3252
+ "epoch": 0.83,
3253
+ "learning_rate": 3.620320855614974e-05,
3254
+ "loss": 0.2158,
3255
+ "step": 5210
3256
+ },
3257
+ {
3258
+ "epoch": 0.84,
3259
+ "learning_rate": 3.61764705882353e-05,
3260
+ "loss": 0.2284,
3261
+ "step": 5220
3262
+ },
3263
+ {
3264
+ "epoch": 0.84,
3265
+ "learning_rate": 3.614973262032086e-05,
3266
+ "loss": 0.1471,
3267
+ "step": 5230
3268
+ },
3269
+ {
3270
+ "epoch": 0.84,
3271
+ "learning_rate": 3.6122994652406416e-05,
3272
+ "loss": 0.2005,
3273
+ "step": 5240
3274
+ },
3275
+ {
3276
+ "epoch": 0.84,
3277
+ "learning_rate": 3.609625668449198e-05,
3278
+ "loss": 0.1275,
3279
+ "step": 5250
3280
+ },
3281
+ {
3282
+ "epoch": 0.84,
3283
+ "learning_rate": 3.606951871657754e-05,
3284
+ "loss": 0.2067,
3285
+ "step": 5260
3286
+ },
3287
+ {
3288
+ "epoch": 0.84,
3289
+ "learning_rate": 3.60427807486631e-05,
3290
+ "loss": 0.5021,
3291
+ "step": 5270
3292
+ },
3293
+ {
3294
+ "epoch": 0.84,
3295
+ "learning_rate": 3.6016042780748665e-05,
3296
+ "loss": 0.4771,
3297
+ "step": 5280
3298
+ },
3299
+ {
3300
+ "epoch": 0.85,
3301
+ "learning_rate": 3.5989304812834225e-05,
3302
+ "loss": 0.4451,
3303
+ "step": 5290
3304
+ },
3305
+ {
3306
+ "epoch": 0.85,
3307
+ "learning_rate": 3.596256684491979e-05,
3308
+ "loss": 0.1325,
3309
+ "step": 5300
3310
+ },
3311
+ {
3312
+ "epoch": 0.85,
3313
+ "learning_rate": 3.593582887700535e-05,
3314
+ "loss": 0.2466,
3315
+ "step": 5310
3316
+ },
3317
+ {
3318
+ "epoch": 0.85,
3319
+ "learning_rate": 3.590909090909091e-05,
3320
+ "loss": 0.3195,
3321
+ "step": 5320
3322
+ },
3323
+ {
3324
+ "epoch": 0.85,
3325
+ "learning_rate": 3.5882352941176474e-05,
3326
+ "loss": 0.2085,
3327
+ "step": 5330
3328
+ },
3329
+ {
3330
+ "epoch": 0.85,
3331
+ "learning_rate": 3.585561497326203e-05,
3332
+ "loss": 0.1185,
3333
+ "step": 5340
3334
+ },
3335
+ {
3336
+ "epoch": 0.86,
3337
+ "learning_rate": 3.582887700534759e-05,
3338
+ "loss": 0.5152,
3339
+ "step": 5350
3340
+ },
3341
+ {
3342
+ "epoch": 0.86,
3343
+ "learning_rate": 3.580213903743316e-05,
3344
+ "loss": 0.186,
3345
+ "step": 5360
3346
+ },
3347
+ {
3348
+ "epoch": 0.86,
3349
+ "learning_rate": 3.577540106951872e-05,
3350
+ "loss": 0.2386,
3351
+ "step": 5370
3352
+ },
3353
+ {
3354
+ "epoch": 0.86,
3355
+ "learning_rate": 3.5748663101604276e-05,
3356
+ "loss": 0.4468,
3357
+ "step": 5380
3358
+ },
3359
+ {
3360
+ "epoch": 0.86,
3361
+ "learning_rate": 3.572192513368984e-05,
3362
+ "loss": 0.1996,
3363
+ "step": 5390
3364
+ },
3365
+ {
3366
+ "epoch": 0.86,
3367
+ "learning_rate": 3.56951871657754e-05,
3368
+ "loss": 0.1182,
3369
+ "step": 5400
3370
+ },
3371
+ {
3372
+ "epoch": 0.87,
3373
+ "learning_rate": 3.566844919786097e-05,
3374
+ "loss": 0.157,
3375
+ "step": 5410
3376
+ },
3377
+ {
3378
+ "epoch": 0.87,
3379
+ "learning_rate": 3.5641711229946526e-05,
3380
+ "loss": 0.1696,
3381
+ "step": 5420
3382
+ },
3383
+ {
3384
+ "epoch": 0.87,
3385
+ "learning_rate": 3.5614973262032085e-05,
3386
+ "loss": 0.2641,
3387
+ "step": 5430
3388
+ },
3389
+ {
3390
+ "epoch": 0.87,
3391
+ "learning_rate": 3.558823529411765e-05,
3392
+ "loss": 0.2224,
3393
+ "step": 5440
3394
+ },
3395
+ {
3396
+ "epoch": 0.87,
3397
+ "learning_rate": 3.556149732620321e-05,
3398
+ "loss": 0.2676,
3399
+ "step": 5450
3400
+ },
3401
+ {
3402
+ "epoch": 0.87,
3403
+ "learning_rate": 3.553475935828877e-05,
3404
+ "loss": 0.213,
3405
+ "step": 5460
3406
+ },
3407
+ {
3408
+ "epoch": 0.88,
3409
+ "learning_rate": 3.5508021390374335e-05,
3410
+ "loss": 0.3497,
3411
+ "step": 5470
3412
+ },
3413
+ {
3414
+ "epoch": 0.88,
3415
+ "learning_rate": 3.5481283422459894e-05,
3416
+ "loss": 0.3077,
3417
+ "step": 5480
3418
+ },
3419
+ {
3420
+ "epoch": 0.88,
3421
+ "learning_rate": 3.545454545454546e-05,
3422
+ "loss": 0.4294,
3423
+ "step": 5490
3424
+ },
3425
+ {
3426
+ "epoch": 0.88,
3427
+ "learning_rate": 3.542780748663101e-05,
3428
+ "loss": 0.1212,
3429
+ "step": 5500
3430
+ },
3431
+ {
3432
+ "epoch": 0.88,
3433
+ "eval_accuracy": 0.8444108761329305,
3434
+ "eval_f1": 0.8425221036853814,
3435
+ "eval_loss": 2.348144054412842,
3436
+ "eval_precision": 0.8617668418744113,
3437
+ "eval_recall": 0.8444108761329305,
3438
+ "eval_runtime": 10.5364,
3439
+ "eval_samples_per_second": 251.32,
3440
+ "eval_steps_per_second": 3.986,
3441
+ "step": 5500
3442
+ }
3443
+ ],
3444
+ "logging_steps": 10,
3445
+ "max_steps": 18750,
3446
+ "num_input_tokens_seen": 0,
3447
+ "num_train_epochs": 3,
3448
+ "save_steps": 500,
3449
+ "total_flos": 4.6307961520128e+16,
3450
+ "train_batch_size": 16,
3451
+ "trial_name": null,
3452
+ "trial_params": null
3453
+ }
checkpoint-5500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ea61cfa8a59fe23f079dded341db3c34a77c19986a2b3885296af3c7226c95
3
+ size 4664
checkpoint-6000/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base-finetuned-8M-MLM-500K/",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
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": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "LABEL_0": 0,
22
+ "LABEL_1": 1,
23
+ "LABEL_2": 2
24
+ },
25
+ "layer_norm_eps": 1e-05,
26
+ "max_position_embeddings": 514,
27
+ "model_type": "roberta",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 1,
31
+ "position_embedding_type": "absolute",
32
+ "problem_type": "single_label_classification",
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.37.0",
35
+ "type_vocab_size": 1,
36
+ "use_cache": true,
37
+ "vocab_size": 50265
38
+ }
checkpoint-6000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c2ce8036ff5e291017bf4b5df5619f3af6d8cf9b48a4fdee8eb15c79a54010f
3
+ size 498615900
checkpoint-6000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cee5fb01d9edce0a1a3b31b55e00e23218ca6728a7771bb22672370f845e28b8
3
+ size 997351610
checkpoint-6000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d0b62444b26de032123da080ab0a2aae128c5f2a535e9214dacc1183ec0e62a
3
+ size 14244
checkpoint-6000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d54d514a35751f0be018531b9a58c85756f186c5f52f1fd615f36ce5699a49f8
3
+ size 1064
checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,3765 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9186971783638,
3
+ "best_model_checkpoint": "./Finetuned_Classification_model_200k/checkpoint-500",
4
+ "epoch": 0.96,
5
+ "eval_steps": 500,
6
+ "global_step": 6000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 1e-05,
14
+ "loss": 1.037,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 2e-05,
20
+ "loss": 0.907,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 3e-05,
26
+ "loss": 0.7159,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4e-05,
32
+ "loss": 0.5758,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.3454,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.997326203208557e-05,
44
+ "loss": 0.1851,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 4.994652406417113e-05,
50
+ "loss": 0.2821,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 4.9919786096256686e-05,
56
+ "loss": 0.1741,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 4.9893048128342245e-05,
62
+ "loss": 0.1749,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.986631016042781e-05,
68
+ "loss": 0.2407,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 4.983957219251337e-05,
74
+ "loss": 0.2612,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 4.981283422459893e-05,
80
+ "loss": 0.1677,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 4.9786096256684495e-05,
86
+ "loss": 0.1123,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 4.975935828877006e-05,
92
+ "loss": 0.1368,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 4.973262032085561e-05,
98
+ "loss": 0.1449,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.03,
103
+ "learning_rate": 4.970588235294118e-05,
104
+ "loss": 0.152,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.03,
109
+ "learning_rate": 4.967914438502674e-05,
110
+ "loss": 0.2504,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 4.9652406417112304e-05,
116
+ "loss": 0.2578,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 4.962566844919786e-05,
122
+ "loss": 0.1289,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 4.959893048128342e-05,
128
+ "loss": 0.1263,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 4.957219251336899e-05,
134
+ "loss": 0.1031,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.04,
139
+ "learning_rate": 4.9545454545454553e-05,
140
+ "loss": 0.2453,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.04,
145
+ "learning_rate": 4.9518716577540106e-05,
146
+ "loss": 0.0675,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.04,
151
+ "learning_rate": 4.949197860962567e-05,
152
+ "loss": 0.1807,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.04,
157
+ "learning_rate": 4.946524064171123e-05,
158
+ "loss": 0.1493,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.04,
163
+ "learning_rate": 4.9438502673796796e-05,
164
+ "loss": 0.109,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 4.9411764705882355e-05,
170
+ "loss": 0.2148,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 4.9385026737967914e-05,
176
+ "loss": 0.2574,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.05,
181
+ "learning_rate": 4.935828877005348e-05,
182
+ "loss": 0.2198,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.05,
187
+ "learning_rate": 4.933155080213904e-05,
188
+ "loss": 0.1574,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.05,
193
+ "learning_rate": 4.93048128342246e-05,
194
+ "loss": 0.0866,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.05,
199
+ "learning_rate": 4.9278074866310164e-05,
200
+ "loss": 0.1371,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.05,
205
+ "learning_rate": 4.925133689839572e-05,
206
+ "loss": 0.2074,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.05,
211
+ "learning_rate": 4.922459893048129e-05,
212
+ "loss": 0.3628,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.06,
217
+ "learning_rate": 4.919786096256685e-05,
218
+ "loss": 0.1888,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.06,
223
+ "learning_rate": 4.917112299465241e-05,
224
+ "loss": 0.0934,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.06,
229
+ "learning_rate": 4.914438502673797e-05,
230
+ "loss": 0.1593,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.06,
235
+ "learning_rate": 4.911764705882353e-05,
236
+ "loss": 0.0721,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.06,
241
+ "learning_rate": 4.909090909090909e-05,
242
+ "loss": 0.1139,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.06,
247
+ "learning_rate": 4.906417112299466e-05,
248
+ "loss": 0.1118,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.07,
253
+ "learning_rate": 4.9037433155080216e-05,
254
+ "loss": 0.218,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.07,
259
+ "learning_rate": 4.901069518716578e-05,
260
+ "loss": 0.126,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.07,
265
+ "learning_rate": 4.898395721925134e-05,
266
+ "loss": 0.1057,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.07,
271
+ "learning_rate": 4.89572192513369e-05,
272
+ "loss": 0.1825,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.07,
277
+ "learning_rate": 4.8930481283422465e-05,
278
+ "loss": 0.1379,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.07,
283
+ "learning_rate": 4.8903743315508024e-05,
284
+ "loss": 0.1311,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.08,
289
+ "learning_rate": 4.8877005347593584e-05,
290
+ "loss": 0.0722,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.08,
295
+ "learning_rate": 4.885026737967915e-05,
296
+ "loss": 0.1286,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.08,
301
+ "learning_rate": 4.882352941176471e-05,
302
+ "loss": 0.1534,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.08,
307
+ "learning_rate": 4.879679144385027e-05,
308
+ "loss": 0.1332,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.08,
313
+ "eval_accuracy": 0.8395015105740181,
314
+ "eval_f1": 0.8361086180158485,
315
+ "eval_loss": 0.9186971783638,
316
+ "eval_precision": 0.8701532825604937,
317
+ "eval_recall": 0.8395015105740181,
318
+ "eval_runtime": 10.5139,
319
+ "eval_samples_per_second": 251.857,
320
+ "eval_steps_per_second": 3.995,
321
+ "step": 500
322
+ },
323
+ {
324
+ "epoch": 0.08,
325
+ "learning_rate": 4.8770053475935826e-05,
326
+ "loss": 0.1317,
327
+ "step": 510
328
+ },
329
+ {
330
+ "epoch": 0.08,
331
+ "learning_rate": 4.874331550802139e-05,
332
+ "loss": 0.0683,
333
+ "step": 520
334
+ },
335
+ {
336
+ "epoch": 0.08,
337
+ "learning_rate": 4.871657754010696e-05,
338
+ "loss": 0.0474,
339
+ "step": 530
340
+ },
341
+ {
342
+ "epoch": 0.09,
343
+ "learning_rate": 4.868983957219252e-05,
344
+ "loss": 0.2302,
345
+ "step": 540
346
+ },
347
+ {
348
+ "epoch": 0.09,
349
+ "learning_rate": 4.8663101604278076e-05,
350
+ "loss": 0.1788,
351
+ "step": 550
352
+ },
353
+ {
354
+ "epoch": 0.09,
355
+ "learning_rate": 4.863636363636364e-05,
356
+ "loss": 0.0917,
357
+ "step": 560
358
+ },
359
+ {
360
+ "epoch": 0.09,
361
+ "learning_rate": 4.86096256684492e-05,
362
+ "loss": 0.2001,
363
+ "step": 570
364
+ },
365
+ {
366
+ "epoch": 0.09,
367
+ "learning_rate": 4.858288770053476e-05,
368
+ "loss": 0.3409,
369
+ "step": 580
370
+ },
371
+ {
372
+ "epoch": 0.09,
373
+ "learning_rate": 4.855614973262032e-05,
374
+ "loss": 0.2044,
375
+ "step": 590
376
+ },
377
+ {
378
+ "epoch": 0.1,
379
+ "learning_rate": 4.8529411764705885e-05,
380
+ "loss": 0.1292,
381
+ "step": 600
382
+ },
383
+ {
384
+ "epoch": 0.1,
385
+ "learning_rate": 4.850267379679145e-05,
386
+ "loss": 0.1422,
387
+ "step": 610
388
+ },
389
+ {
390
+ "epoch": 0.1,
391
+ "learning_rate": 4.8475935828877e-05,
392
+ "loss": 0.1165,
393
+ "step": 620
394
+ },
395
+ {
396
+ "epoch": 0.1,
397
+ "learning_rate": 4.844919786096257e-05,
398
+ "loss": 0.1115,
399
+ "step": 630
400
+ },
401
+ {
402
+ "epoch": 0.1,
403
+ "learning_rate": 4.8422459893048135e-05,
404
+ "loss": 0.0957,
405
+ "step": 640
406
+ },
407
+ {
408
+ "epoch": 0.1,
409
+ "learning_rate": 4.8395721925133694e-05,
410
+ "loss": 0.1791,
411
+ "step": 650
412
+ },
413
+ {
414
+ "epoch": 0.11,
415
+ "learning_rate": 4.836898395721925e-05,
416
+ "loss": 0.0691,
417
+ "step": 660
418
+ },
419
+ {
420
+ "epoch": 0.11,
421
+ "learning_rate": 4.834224598930481e-05,
422
+ "loss": 0.2829,
423
+ "step": 670
424
+ },
425
+ {
426
+ "epoch": 0.11,
427
+ "learning_rate": 4.831550802139038e-05,
428
+ "loss": 0.2635,
429
+ "step": 680
430
+ },
431
+ {
432
+ "epoch": 0.11,
433
+ "learning_rate": 4.828877005347594e-05,
434
+ "loss": 0.1722,
435
+ "step": 690
436
+ },
437
+ {
438
+ "epoch": 0.11,
439
+ "learning_rate": 4.8262032085561496e-05,
440
+ "loss": 0.0974,
441
+ "step": 700
442
+ },
443
+ {
444
+ "epoch": 0.11,
445
+ "learning_rate": 4.823529411764706e-05,
446
+ "loss": 0.0888,
447
+ "step": 710
448
+ },
449
+ {
450
+ "epoch": 0.12,
451
+ "learning_rate": 4.820855614973262e-05,
452
+ "loss": 0.3148,
453
+ "step": 720
454
+ },
455
+ {
456
+ "epoch": 0.12,
457
+ "learning_rate": 4.8181818181818186e-05,
458
+ "loss": 0.3556,
459
+ "step": 730
460
+ },
461
+ {
462
+ "epoch": 0.12,
463
+ "learning_rate": 4.8155080213903745e-05,
464
+ "loss": 0.1788,
465
+ "step": 740
466
+ },
467
+ {
468
+ "epoch": 0.12,
469
+ "learning_rate": 4.8128342245989304e-05,
470
+ "loss": 0.4376,
471
+ "step": 750
472
+ },
473
+ {
474
+ "epoch": 0.12,
475
+ "learning_rate": 4.810160427807487e-05,
476
+ "loss": 0.2413,
477
+ "step": 760
478
+ },
479
+ {
480
+ "epoch": 0.12,
481
+ "learning_rate": 4.807486631016043e-05,
482
+ "loss": 0.1802,
483
+ "step": 770
484
+ },
485
+ {
486
+ "epoch": 0.12,
487
+ "learning_rate": 4.804812834224599e-05,
488
+ "loss": 0.1364,
489
+ "step": 780
490
+ },
491
+ {
492
+ "epoch": 0.13,
493
+ "learning_rate": 4.8021390374331554e-05,
494
+ "loss": 0.0667,
495
+ "step": 790
496
+ },
497
+ {
498
+ "epoch": 0.13,
499
+ "learning_rate": 4.799465240641711e-05,
500
+ "loss": 0.1409,
501
+ "step": 800
502
+ },
503
+ {
504
+ "epoch": 0.13,
505
+ "learning_rate": 4.796791443850268e-05,
506
+ "loss": 0.0897,
507
+ "step": 810
508
+ },
509
+ {
510
+ "epoch": 0.13,
511
+ "learning_rate": 4.794117647058824e-05,
512
+ "loss": 0.0871,
513
+ "step": 820
514
+ },
515
+ {
516
+ "epoch": 0.13,
517
+ "learning_rate": 4.79144385026738e-05,
518
+ "loss": 0.0552,
519
+ "step": 830
520
+ },
521
+ {
522
+ "epoch": 0.13,
523
+ "learning_rate": 4.788770053475936e-05,
524
+ "loss": 0.2569,
525
+ "step": 840
526
+ },
527
+ {
528
+ "epoch": 0.14,
529
+ "learning_rate": 4.786096256684492e-05,
530
+ "loss": 0.2877,
531
+ "step": 850
532
+ },
533
+ {
534
+ "epoch": 0.14,
535
+ "learning_rate": 4.783422459893048e-05,
536
+ "loss": 0.1365,
537
+ "step": 860
538
+ },
539
+ {
540
+ "epoch": 0.14,
541
+ "learning_rate": 4.7807486631016047e-05,
542
+ "loss": 0.1604,
543
+ "step": 870
544
+ },
545
+ {
546
+ "epoch": 0.14,
547
+ "learning_rate": 4.7780748663101606e-05,
548
+ "loss": 0.2179,
549
+ "step": 880
550
+ },
551
+ {
552
+ "epoch": 0.14,
553
+ "learning_rate": 4.7754010695187165e-05,
554
+ "loss": 0.0969,
555
+ "step": 890
556
+ },
557
+ {
558
+ "epoch": 0.14,
559
+ "learning_rate": 4.772727272727273e-05,
560
+ "loss": 0.0953,
561
+ "step": 900
562
+ },
563
+ {
564
+ "epoch": 0.15,
565
+ "learning_rate": 4.770053475935829e-05,
566
+ "loss": 0.0803,
567
+ "step": 910
568
+ },
569
+ {
570
+ "epoch": 0.15,
571
+ "learning_rate": 4.7673796791443855e-05,
572
+ "loss": 0.1573,
573
+ "step": 920
574
+ },
575
+ {
576
+ "epoch": 0.15,
577
+ "learning_rate": 4.7647058823529414e-05,
578
+ "loss": 0.1388,
579
+ "step": 930
580
+ },
581
+ {
582
+ "epoch": 0.15,
583
+ "learning_rate": 4.7620320855614973e-05,
584
+ "loss": 0.1559,
585
+ "step": 940
586
+ },
587
+ {
588
+ "epoch": 0.15,
589
+ "learning_rate": 4.759358288770054e-05,
590
+ "loss": 0.2411,
591
+ "step": 950
592
+ },
593
+ {
594
+ "epoch": 0.15,
595
+ "learning_rate": 4.75668449197861e-05,
596
+ "loss": 0.1119,
597
+ "step": 960
598
+ },
599
+ {
600
+ "epoch": 0.16,
601
+ "learning_rate": 4.754010695187166e-05,
602
+ "loss": 0.1416,
603
+ "step": 970
604
+ },
605
+ {
606
+ "epoch": 0.16,
607
+ "learning_rate": 4.751336898395722e-05,
608
+ "loss": 0.1073,
609
+ "step": 980
610
+ },
611
+ {
612
+ "epoch": 0.16,
613
+ "learning_rate": 4.748663101604278e-05,
614
+ "loss": 0.1799,
615
+ "step": 990
616
+ },
617
+ {
618
+ "epoch": 0.16,
619
+ "learning_rate": 4.745989304812835e-05,
620
+ "loss": 0.127,
621
+ "step": 1000
622
+ },
623
+ {
624
+ "epoch": 0.16,
625
+ "eval_accuracy": 0.847809667673716,
626
+ "eval_f1": 0.8460260531471735,
627
+ "eval_loss": 1.017073392868042,
628
+ "eval_precision": 0.8647086198841812,
629
+ "eval_recall": 0.847809667673716,
630
+ "eval_runtime": 10.5379,
631
+ "eval_samples_per_second": 251.284,
632
+ "eval_steps_per_second": 3.986,
633
+ "step": 1000
634
+ },
635
+ {
636
+ "epoch": 0.16,
637
+ "learning_rate": 4.74331550802139e-05,
638
+ "loss": 0.1774,
639
+ "step": 1010
640
+ },
641
+ {
642
+ "epoch": 0.16,
643
+ "learning_rate": 4.7406417112299466e-05,
644
+ "loss": 0.0869,
645
+ "step": 1020
646
+ },
647
+ {
648
+ "epoch": 0.16,
649
+ "learning_rate": 4.737967914438503e-05,
650
+ "loss": 0.0686,
651
+ "step": 1030
652
+ },
653
+ {
654
+ "epoch": 0.17,
655
+ "learning_rate": 4.735294117647059e-05,
656
+ "loss": 0.124,
657
+ "step": 1040
658
+ },
659
+ {
660
+ "epoch": 0.17,
661
+ "learning_rate": 4.732620320855615e-05,
662
+ "loss": 0.0993,
663
+ "step": 1050
664
+ },
665
+ {
666
+ "epoch": 0.17,
667
+ "learning_rate": 4.7299465240641716e-05,
668
+ "loss": 0.2451,
669
+ "step": 1060
670
+ },
671
+ {
672
+ "epoch": 0.17,
673
+ "learning_rate": 4.7272727272727275e-05,
674
+ "loss": 0.1607,
675
+ "step": 1070
676
+ },
677
+ {
678
+ "epoch": 0.17,
679
+ "learning_rate": 4.724598930481284e-05,
680
+ "loss": 0.1289,
681
+ "step": 1080
682
+ },
683
+ {
684
+ "epoch": 0.17,
685
+ "learning_rate": 4.721925133689839e-05,
686
+ "loss": 0.1813,
687
+ "step": 1090
688
+ },
689
+ {
690
+ "epoch": 0.18,
691
+ "learning_rate": 4.719251336898396e-05,
692
+ "loss": 0.1628,
693
+ "step": 1100
694
+ },
695
+ {
696
+ "epoch": 0.18,
697
+ "learning_rate": 4.7165775401069524e-05,
698
+ "loss": 0.3152,
699
+ "step": 1110
700
+ },
701
+ {
702
+ "epoch": 0.18,
703
+ "learning_rate": 4.7139037433155083e-05,
704
+ "loss": 0.3835,
705
+ "step": 1120
706
+ },
707
+ {
708
+ "epoch": 0.18,
709
+ "learning_rate": 4.711229946524064e-05,
710
+ "loss": 0.1713,
711
+ "step": 1130
712
+ },
713
+ {
714
+ "epoch": 0.18,
715
+ "learning_rate": 4.70855614973262e-05,
716
+ "loss": 0.1812,
717
+ "step": 1140
718
+ },
719
+ {
720
+ "epoch": 0.18,
721
+ "learning_rate": 4.705882352941177e-05,
722
+ "loss": 0.1928,
723
+ "step": 1150
724
+ },
725
+ {
726
+ "epoch": 0.19,
727
+ "learning_rate": 4.703208556149733e-05,
728
+ "loss": 0.2963,
729
+ "step": 1160
730
+ },
731
+ {
732
+ "epoch": 0.19,
733
+ "learning_rate": 4.7005347593582885e-05,
734
+ "loss": 0.2073,
735
+ "step": 1170
736
+ },
737
+ {
738
+ "epoch": 0.19,
739
+ "learning_rate": 4.697860962566845e-05,
740
+ "loss": 0.2299,
741
+ "step": 1180
742
+ },
743
+ {
744
+ "epoch": 0.19,
745
+ "learning_rate": 4.695187165775402e-05,
746
+ "loss": 0.1832,
747
+ "step": 1190
748
+ },
749
+ {
750
+ "epoch": 0.19,
751
+ "learning_rate": 4.6925133689839576e-05,
752
+ "loss": 0.1844,
753
+ "step": 1200
754
+ },
755
+ {
756
+ "epoch": 0.19,
757
+ "learning_rate": 4.6898395721925135e-05,
758
+ "loss": 0.4959,
759
+ "step": 1210
760
+ },
761
+ {
762
+ "epoch": 0.2,
763
+ "learning_rate": 4.6871657754010694e-05,
764
+ "loss": 0.2516,
765
+ "step": 1220
766
+ },
767
+ {
768
+ "epoch": 0.2,
769
+ "learning_rate": 4.684491978609626e-05,
770
+ "loss": 0.1754,
771
+ "step": 1230
772
+ },
773
+ {
774
+ "epoch": 0.2,
775
+ "learning_rate": 4.681818181818182e-05,
776
+ "loss": 0.2554,
777
+ "step": 1240
778
+ },
779
+ {
780
+ "epoch": 0.2,
781
+ "learning_rate": 4.679144385026738e-05,
782
+ "loss": 0.1976,
783
+ "step": 1250
784
+ },
785
+ {
786
+ "epoch": 0.2,
787
+ "learning_rate": 4.6764705882352944e-05,
788
+ "loss": 0.1906,
789
+ "step": 1260
790
+ },
791
+ {
792
+ "epoch": 0.2,
793
+ "learning_rate": 4.673796791443851e-05,
794
+ "loss": 0.0806,
795
+ "step": 1270
796
+ },
797
+ {
798
+ "epoch": 0.2,
799
+ "learning_rate": 4.671122994652407e-05,
800
+ "loss": 0.1229,
801
+ "step": 1280
802
+ },
803
+ {
804
+ "epoch": 0.21,
805
+ "learning_rate": 4.668449197860963e-05,
806
+ "loss": 0.1222,
807
+ "step": 1290
808
+ },
809
+ {
810
+ "epoch": 0.21,
811
+ "learning_rate": 4.665775401069519e-05,
812
+ "loss": 0.1049,
813
+ "step": 1300
814
+ },
815
+ {
816
+ "epoch": 0.21,
817
+ "learning_rate": 4.663101604278075e-05,
818
+ "loss": 0.1548,
819
+ "step": 1310
820
+ },
821
+ {
822
+ "epoch": 0.21,
823
+ "learning_rate": 4.660427807486631e-05,
824
+ "loss": 0.1077,
825
+ "step": 1320
826
+ },
827
+ {
828
+ "epoch": 0.21,
829
+ "learning_rate": 4.657754010695187e-05,
830
+ "loss": 0.1949,
831
+ "step": 1330
832
+ },
833
+ {
834
+ "epoch": 0.21,
835
+ "learning_rate": 4.6550802139037436e-05,
836
+ "loss": 0.1349,
837
+ "step": 1340
838
+ },
839
+ {
840
+ "epoch": 0.22,
841
+ "learning_rate": 4.6524064171123e-05,
842
+ "loss": 0.223,
843
+ "step": 1350
844
+ },
845
+ {
846
+ "epoch": 0.22,
847
+ "learning_rate": 4.6497326203208555e-05,
848
+ "loss": 0.1132,
849
+ "step": 1360
850
+ },
851
+ {
852
+ "epoch": 0.22,
853
+ "learning_rate": 4.647058823529412e-05,
854
+ "loss": 0.123,
855
+ "step": 1370
856
+ },
857
+ {
858
+ "epoch": 0.22,
859
+ "learning_rate": 4.644385026737968e-05,
860
+ "loss": 0.1741,
861
+ "step": 1380
862
+ },
863
+ {
864
+ "epoch": 0.22,
865
+ "learning_rate": 4.6417112299465245e-05,
866
+ "loss": 0.0913,
867
+ "step": 1390
868
+ },
869
+ {
870
+ "epoch": 0.22,
871
+ "learning_rate": 4.6390374331550804e-05,
872
+ "loss": 0.1428,
873
+ "step": 1400
874
+ },
875
+ {
876
+ "epoch": 0.23,
877
+ "learning_rate": 4.636363636363636e-05,
878
+ "loss": 0.1171,
879
+ "step": 1410
880
+ },
881
+ {
882
+ "epoch": 0.23,
883
+ "learning_rate": 4.633689839572193e-05,
884
+ "loss": 0.1773,
885
+ "step": 1420
886
+ },
887
+ {
888
+ "epoch": 0.23,
889
+ "learning_rate": 4.631016042780749e-05,
890
+ "loss": 0.1372,
891
+ "step": 1430
892
+ },
893
+ {
894
+ "epoch": 0.23,
895
+ "learning_rate": 4.628342245989305e-05,
896
+ "loss": 0.2797,
897
+ "step": 1440
898
+ },
899
+ {
900
+ "epoch": 0.23,
901
+ "learning_rate": 4.625668449197861e-05,
902
+ "loss": 0.2813,
903
+ "step": 1450
904
+ },
905
+ {
906
+ "epoch": 0.23,
907
+ "learning_rate": 4.622994652406417e-05,
908
+ "loss": 0.2861,
909
+ "step": 1460
910
+ },
911
+ {
912
+ "epoch": 0.24,
913
+ "learning_rate": 4.620320855614974e-05,
914
+ "loss": 0.0677,
915
+ "step": 1470
916
+ },
917
+ {
918
+ "epoch": 0.24,
919
+ "learning_rate": 4.61764705882353e-05,
920
+ "loss": 0.104,
921
+ "step": 1480
922
+ },
923
+ {
924
+ "epoch": 0.24,
925
+ "learning_rate": 4.6149732620320856e-05,
926
+ "loss": 0.1533,
927
+ "step": 1490
928
+ },
929
+ {
930
+ "epoch": 0.24,
931
+ "learning_rate": 4.612299465240642e-05,
932
+ "loss": 0.1186,
933
+ "step": 1500
934
+ },
935
+ {
936
+ "epoch": 0.24,
937
+ "eval_accuracy": 0.8376132930513596,
938
+ "eval_f1": 0.8363352339205664,
939
+ "eval_loss": 1.0860812664031982,
940
+ "eval_precision": 0.8484990166173995,
941
+ "eval_recall": 0.8376132930513596,
942
+ "eval_runtime": 10.5226,
943
+ "eval_samples_per_second": 251.648,
944
+ "eval_steps_per_second": 3.991,
945
+ "step": 1500
946
+ },
947
+ {
948
+ "epoch": 0.24,
949
+ "learning_rate": 4.609625668449198e-05,
950
+ "loss": 0.0683,
951
+ "step": 1510
952
+ },
953
+ {
954
+ "epoch": 0.24,
955
+ "learning_rate": 4.606951871657754e-05,
956
+ "loss": 0.1767,
957
+ "step": 1520
958
+ },
959
+ {
960
+ "epoch": 0.24,
961
+ "learning_rate": 4.6042780748663106e-05,
962
+ "loss": 0.0559,
963
+ "step": 1530
964
+ },
965
+ {
966
+ "epoch": 0.25,
967
+ "learning_rate": 4.6016042780748665e-05,
968
+ "loss": 0.1056,
969
+ "step": 1540
970
+ },
971
+ {
972
+ "epoch": 0.25,
973
+ "learning_rate": 4.598930481283423e-05,
974
+ "loss": 0.1607,
975
+ "step": 1550
976
+ },
977
+ {
978
+ "epoch": 0.25,
979
+ "learning_rate": 4.596256684491979e-05,
980
+ "loss": 0.053,
981
+ "step": 1560
982
+ },
983
+ {
984
+ "epoch": 0.25,
985
+ "learning_rate": 4.593582887700535e-05,
986
+ "loss": 0.2743,
987
+ "step": 1570
988
+ },
989
+ {
990
+ "epoch": 0.25,
991
+ "learning_rate": 4.5909090909090914e-05,
992
+ "loss": 0.1569,
993
+ "step": 1580
994
+ },
995
+ {
996
+ "epoch": 0.25,
997
+ "learning_rate": 4.588235294117647e-05,
998
+ "loss": 0.0648,
999
+ "step": 1590
1000
+ },
1001
+ {
1002
+ "epoch": 0.26,
1003
+ "learning_rate": 4.585561497326203e-05,
1004
+ "loss": 0.132,
1005
+ "step": 1600
1006
+ },
1007
+ {
1008
+ "epoch": 0.26,
1009
+ "learning_rate": 4.58288770053476e-05,
1010
+ "loss": 0.1115,
1011
+ "step": 1610
1012
+ },
1013
+ {
1014
+ "epoch": 0.26,
1015
+ "learning_rate": 4.580213903743316e-05,
1016
+ "loss": 0.2035,
1017
+ "step": 1620
1018
+ },
1019
+ {
1020
+ "epoch": 0.26,
1021
+ "learning_rate": 4.5775401069518716e-05,
1022
+ "loss": 0.1193,
1023
+ "step": 1630
1024
+ },
1025
+ {
1026
+ "epoch": 0.26,
1027
+ "learning_rate": 4.5748663101604275e-05,
1028
+ "loss": 0.2867,
1029
+ "step": 1640
1030
+ },
1031
+ {
1032
+ "epoch": 0.26,
1033
+ "learning_rate": 4.572192513368984e-05,
1034
+ "loss": 0.2423,
1035
+ "step": 1650
1036
+ },
1037
+ {
1038
+ "epoch": 0.27,
1039
+ "learning_rate": 4.569518716577541e-05,
1040
+ "loss": 0.0959,
1041
+ "step": 1660
1042
+ },
1043
+ {
1044
+ "epoch": 0.27,
1045
+ "learning_rate": 4.5668449197860966e-05,
1046
+ "loss": 0.062,
1047
+ "step": 1670
1048
+ },
1049
+ {
1050
+ "epoch": 0.27,
1051
+ "learning_rate": 4.5641711229946525e-05,
1052
+ "loss": 0.3052,
1053
+ "step": 1680
1054
+ },
1055
+ {
1056
+ "epoch": 0.27,
1057
+ "learning_rate": 4.561497326203209e-05,
1058
+ "loss": 0.0814,
1059
+ "step": 1690
1060
+ },
1061
+ {
1062
+ "epoch": 0.27,
1063
+ "learning_rate": 4.558823529411765e-05,
1064
+ "loss": 0.4406,
1065
+ "step": 1700
1066
+ },
1067
+ {
1068
+ "epoch": 0.27,
1069
+ "learning_rate": 4.556149732620321e-05,
1070
+ "loss": 0.4316,
1071
+ "step": 1710
1072
+ },
1073
+ {
1074
+ "epoch": 0.28,
1075
+ "learning_rate": 4.553475935828877e-05,
1076
+ "loss": 0.2551,
1077
+ "step": 1720
1078
+ },
1079
+ {
1080
+ "epoch": 0.28,
1081
+ "learning_rate": 4.5508021390374334e-05,
1082
+ "loss": 0.4731,
1083
+ "step": 1730
1084
+ },
1085
+ {
1086
+ "epoch": 0.28,
1087
+ "learning_rate": 4.54812834224599e-05,
1088
+ "loss": 0.3468,
1089
+ "step": 1740
1090
+ },
1091
+ {
1092
+ "epoch": 0.28,
1093
+ "learning_rate": 4.545454545454546e-05,
1094
+ "loss": 0.4397,
1095
+ "step": 1750
1096
+ },
1097
+ {
1098
+ "epoch": 0.28,
1099
+ "learning_rate": 4.542780748663102e-05,
1100
+ "loss": 0.2163,
1101
+ "step": 1760
1102
+ },
1103
+ {
1104
+ "epoch": 0.28,
1105
+ "learning_rate": 4.5401069518716583e-05,
1106
+ "loss": 0.1675,
1107
+ "step": 1770
1108
+ },
1109
+ {
1110
+ "epoch": 0.28,
1111
+ "learning_rate": 4.537433155080214e-05,
1112
+ "loss": 0.124,
1113
+ "step": 1780
1114
+ },
1115
+ {
1116
+ "epoch": 0.29,
1117
+ "learning_rate": 4.53475935828877e-05,
1118
+ "loss": 0.192,
1119
+ "step": 1790
1120
+ },
1121
+ {
1122
+ "epoch": 0.29,
1123
+ "learning_rate": 4.532085561497326e-05,
1124
+ "loss": 0.2626,
1125
+ "step": 1800
1126
+ },
1127
+ {
1128
+ "epoch": 0.29,
1129
+ "learning_rate": 4.5294117647058826e-05,
1130
+ "loss": 0.233,
1131
+ "step": 1810
1132
+ },
1133
+ {
1134
+ "epoch": 0.29,
1135
+ "learning_rate": 4.526737967914439e-05,
1136
+ "loss": 0.2005,
1137
+ "step": 1820
1138
+ },
1139
+ {
1140
+ "epoch": 0.29,
1141
+ "learning_rate": 4.5240641711229944e-05,
1142
+ "loss": 0.1611,
1143
+ "step": 1830
1144
+ },
1145
+ {
1146
+ "epoch": 0.29,
1147
+ "learning_rate": 4.521390374331551e-05,
1148
+ "loss": 0.1215,
1149
+ "step": 1840
1150
+ },
1151
+ {
1152
+ "epoch": 0.3,
1153
+ "learning_rate": 4.518716577540107e-05,
1154
+ "loss": 0.1504,
1155
+ "step": 1850
1156
+ },
1157
+ {
1158
+ "epoch": 0.3,
1159
+ "learning_rate": 4.5160427807486635e-05,
1160
+ "loss": 0.1317,
1161
+ "step": 1860
1162
+ },
1163
+ {
1164
+ "epoch": 0.3,
1165
+ "learning_rate": 4.5133689839572194e-05,
1166
+ "loss": 0.1624,
1167
+ "step": 1870
1168
+ },
1169
+ {
1170
+ "epoch": 0.3,
1171
+ "learning_rate": 4.510695187165775e-05,
1172
+ "loss": 0.1815,
1173
+ "step": 1880
1174
+ },
1175
+ {
1176
+ "epoch": 0.3,
1177
+ "learning_rate": 4.508021390374332e-05,
1178
+ "loss": 0.2438,
1179
+ "step": 1890
1180
+ },
1181
+ {
1182
+ "epoch": 0.3,
1183
+ "learning_rate": 4.5053475935828885e-05,
1184
+ "loss": 0.1908,
1185
+ "step": 1900
1186
+ },
1187
+ {
1188
+ "epoch": 0.31,
1189
+ "learning_rate": 4.502673796791444e-05,
1190
+ "loss": 0.2333,
1191
+ "step": 1910
1192
+ },
1193
+ {
1194
+ "epoch": 0.31,
1195
+ "learning_rate": 4.5e-05,
1196
+ "loss": 0.2385,
1197
+ "step": 1920
1198
+ },
1199
+ {
1200
+ "epoch": 0.31,
1201
+ "learning_rate": 4.497326203208556e-05,
1202
+ "loss": 0.1889,
1203
+ "step": 1930
1204
+ },
1205
+ {
1206
+ "epoch": 0.31,
1207
+ "learning_rate": 4.494652406417113e-05,
1208
+ "loss": 0.1202,
1209
+ "step": 1940
1210
+ },
1211
+ {
1212
+ "epoch": 0.31,
1213
+ "learning_rate": 4.491978609625669e-05,
1214
+ "loss": 0.3926,
1215
+ "step": 1950
1216
+ },
1217
+ {
1218
+ "epoch": 0.31,
1219
+ "learning_rate": 4.4893048128342246e-05,
1220
+ "loss": 0.3656,
1221
+ "step": 1960
1222
+ },
1223
+ {
1224
+ "epoch": 0.32,
1225
+ "learning_rate": 4.486631016042781e-05,
1226
+ "loss": 0.1626,
1227
+ "step": 1970
1228
+ },
1229
+ {
1230
+ "epoch": 0.32,
1231
+ "learning_rate": 4.483957219251337e-05,
1232
+ "loss": 0.1318,
1233
+ "step": 1980
1234
+ },
1235
+ {
1236
+ "epoch": 0.32,
1237
+ "learning_rate": 4.481283422459893e-05,
1238
+ "loss": 0.2922,
1239
+ "step": 1990
1240
+ },
1241
+ {
1242
+ "epoch": 0.32,
1243
+ "learning_rate": 4.4786096256684495e-05,
1244
+ "loss": 0.3545,
1245
+ "step": 2000
1246
+ },
1247
+ {
1248
+ "epoch": 0.32,
1249
+ "eval_accuracy": 0.8413897280966768,
1250
+ "eval_f1": 0.8394373029347564,
1251
+ "eval_loss": 2.3635265827178955,
1252
+ "eval_precision": 0.8588437332105303,
1253
+ "eval_recall": 0.8413897280966768,
1254
+ "eval_runtime": 10.5481,
1255
+ "eval_samples_per_second": 251.039,
1256
+ "eval_steps_per_second": 3.982,
1257
+ "step": 2000
1258
+ },
1259
+ {
1260
+ "epoch": 0.32,
1261
+ "learning_rate": 4.4759358288770054e-05,
1262
+ "loss": 0.1883,
1263
+ "step": 2010
1264
+ },
1265
+ {
1266
+ "epoch": 0.32,
1267
+ "learning_rate": 4.473262032085562e-05,
1268
+ "loss": 0.4851,
1269
+ "step": 2020
1270
+ },
1271
+ {
1272
+ "epoch": 0.32,
1273
+ "learning_rate": 4.470588235294118e-05,
1274
+ "loss": 0.4886,
1275
+ "step": 2030
1276
+ },
1277
+ {
1278
+ "epoch": 0.33,
1279
+ "learning_rate": 4.467914438502674e-05,
1280
+ "loss": 0.2792,
1281
+ "step": 2040
1282
+ },
1283
+ {
1284
+ "epoch": 0.33,
1285
+ "learning_rate": 4.4652406417112304e-05,
1286
+ "loss": 0.2475,
1287
+ "step": 2050
1288
+ },
1289
+ {
1290
+ "epoch": 0.33,
1291
+ "learning_rate": 4.462566844919786e-05,
1292
+ "loss": 0.1728,
1293
+ "step": 2060
1294
+ },
1295
+ {
1296
+ "epoch": 0.33,
1297
+ "learning_rate": 4.459893048128342e-05,
1298
+ "loss": 0.0873,
1299
+ "step": 2070
1300
+ },
1301
+ {
1302
+ "epoch": 0.33,
1303
+ "learning_rate": 4.457219251336899e-05,
1304
+ "loss": 0.1553,
1305
+ "step": 2080
1306
+ },
1307
+ {
1308
+ "epoch": 0.33,
1309
+ "learning_rate": 4.454545454545455e-05,
1310
+ "loss": 0.3127,
1311
+ "step": 2090
1312
+ },
1313
+ {
1314
+ "epoch": 0.34,
1315
+ "learning_rate": 4.4518716577540106e-05,
1316
+ "loss": 0.4534,
1317
+ "step": 2100
1318
+ },
1319
+ {
1320
+ "epoch": 0.34,
1321
+ "learning_rate": 4.449197860962567e-05,
1322
+ "loss": 0.2118,
1323
+ "step": 2110
1324
+ },
1325
+ {
1326
+ "epoch": 0.34,
1327
+ "learning_rate": 4.446524064171123e-05,
1328
+ "loss": 0.3214,
1329
+ "step": 2120
1330
+ },
1331
+ {
1332
+ "epoch": 0.34,
1333
+ "learning_rate": 4.44385026737968e-05,
1334
+ "loss": 0.1867,
1335
+ "step": 2130
1336
+ },
1337
+ {
1338
+ "epoch": 0.34,
1339
+ "learning_rate": 4.4411764705882356e-05,
1340
+ "loss": 0.3318,
1341
+ "step": 2140
1342
+ },
1343
+ {
1344
+ "epoch": 0.34,
1345
+ "learning_rate": 4.4385026737967915e-05,
1346
+ "loss": 0.164,
1347
+ "step": 2150
1348
+ },
1349
+ {
1350
+ "epoch": 0.35,
1351
+ "learning_rate": 4.435828877005348e-05,
1352
+ "loss": 0.2003,
1353
+ "step": 2160
1354
+ },
1355
+ {
1356
+ "epoch": 0.35,
1357
+ "learning_rate": 4.433155080213904e-05,
1358
+ "loss": 0.1854,
1359
+ "step": 2170
1360
+ },
1361
+ {
1362
+ "epoch": 0.35,
1363
+ "learning_rate": 4.43048128342246e-05,
1364
+ "loss": 0.2871,
1365
+ "step": 2180
1366
+ },
1367
+ {
1368
+ "epoch": 0.35,
1369
+ "learning_rate": 4.4278074866310165e-05,
1370
+ "loss": 0.306,
1371
+ "step": 2190
1372
+ },
1373
+ {
1374
+ "epoch": 0.35,
1375
+ "learning_rate": 4.4251336898395724e-05,
1376
+ "loss": 0.462,
1377
+ "step": 2200
1378
+ },
1379
+ {
1380
+ "epoch": 0.35,
1381
+ "learning_rate": 4.422459893048129e-05,
1382
+ "loss": 0.3016,
1383
+ "step": 2210
1384
+ },
1385
+ {
1386
+ "epoch": 0.36,
1387
+ "learning_rate": 4.419786096256684e-05,
1388
+ "loss": 0.2322,
1389
+ "step": 2220
1390
+ },
1391
+ {
1392
+ "epoch": 0.36,
1393
+ "learning_rate": 4.417112299465241e-05,
1394
+ "loss": 0.141,
1395
+ "step": 2230
1396
+ },
1397
+ {
1398
+ "epoch": 0.36,
1399
+ "learning_rate": 4.414438502673797e-05,
1400
+ "loss": 0.1501,
1401
+ "step": 2240
1402
+ },
1403
+ {
1404
+ "epoch": 0.36,
1405
+ "learning_rate": 4.411764705882353e-05,
1406
+ "loss": 0.2379,
1407
+ "step": 2250
1408
+ },
1409
+ {
1410
+ "epoch": 0.36,
1411
+ "learning_rate": 4.409090909090909e-05,
1412
+ "loss": 0.2233,
1413
+ "step": 2260
1414
+ },
1415
+ {
1416
+ "epoch": 0.36,
1417
+ "learning_rate": 4.406417112299465e-05,
1418
+ "loss": 0.4372,
1419
+ "step": 2270
1420
+ },
1421
+ {
1422
+ "epoch": 0.36,
1423
+ "learning_rate": 4.4037433155080216e-05,
1424
+ "loss": 0.5062,
1425
+ "step": 2280
1426
+ },
1427
+ {
1428
+ "epoch": 0.37,
1429
+ "learning_rate": 4.401069518716578e-05,
1430
+ "loss": 0.1763,
1431
+ "step": 2290
1432
+ },
1433
+ {
1434
+ "epoch": 0.37,
1435
+ "learning_rate": 4.3983957219251334e-05,
1436
+ "loss": 0.1486,
1437
+ "step": 2300
1438
+ },
1439
+ {
1440
+ "epoch": 0.37,
1441
+ "learning_rate": 4.39572192513369e-05,
1442
+ "loss": 0.2072,
1443
+ "step": 2310
1444
+ },
1445
+ {
1446
+ "epoch": 0.37,
1447
+ "learning_rate": 4.3930481283422466e-05,
1448
+ "loss": 0.0901,
1449
+ "step": 2320
1450
+ },
1451
+ {
1452
+ "epoch": 0.37,
1453
+ "learning_rate": 4.3903743315508025e-05,
1454
+ "loss": 0.351,
1455
+ "step": 2330
1456
+ },
1457
+ {
1458
+ "epoch": 0.37,
1459
+ "learning_rate": 4.3877005347593584e-05,
1460
+ "loss": 0.2836,
1461
+ "step": 2340
1462
+ },
1463
+ {
1464
+ "epoch": 0.38,
1465
+ "learning_rate": 4.385026737967914e-05,
1466
+ "loss": 0.3663,
1467
+ "step": 2350
1468
+ },
1469
+ {
1470
+ "epoch": 0.38,
1471
+ "learning_rate": 4.382352941176471e-05,
1472
+ "loss": 0.4,
1473
+ "step": 2360
1474
+ },
1475
+ {
1476
+ "epoch": 0.38,
1477
+ "learning_rate": 4.3796791443850275e-05,
1478
+ "loss": 0.3752,
1479
+ "step": 2370
1480
+ },
1481
+ {
1482
+ "epoch": 0.38,
1483
+ "learning_rate": 4.377005347593583e-05,
1484
+ "loss": 0.22,
1485
+ "step": 2380
1486
+ },
1487
+ {
1488
+ "epoch": 0.38,
1489
+ "learning_rate": 4.374331550802139e-05,
1490
+ "loss": 0.3702,
1491
+ "step": 2390
1492
+ },
1493
+ {
1494
+ "epoch": 0.38,
1495
+ "learning_rate": 4.371657754010696e-05,
1496
+ "loss": 0.1035,
1497
+ "step": 2400
1498
+ },
1499
+ {
1500
+ "epoch": 0.39,
1501
+ "learning_rate": 4.368983957219252e-05,
1502
+ "loss": 0.1581,
1503
+ "step": 2410
1504
+ },
1505
+ {
1506
+ "epoch": 0.39,
1507
+ "learning_rate": 4.3663101604278077e-05,
1508
+ "loss": 0.2575,
1509
+ "step": 2420
1510
+ },
1511
+ {
1512
+ "epoch": 0.39,
1513
+ "learning_rate": 4.3636363636363636e-05,
1514
+ "loss": 0.1866,
1515
+ "step": 2430
1516
+ },
1517
+ {
1518
+ "epoch": 0.39,
1519
+ "learning_rate": 4.36096256684492e-05,
1520
+ "loss": 0.3816,
1521
+ "step": 2440
1522
+ },
1523
+ {
1524
+ "epoch": 0.39,
1525
+ "learning_rate": 4.358288770053476e-05,
1526
+ "loss": 0.4613,
1527
+ "step": 2450
1528
+ },
1529
+ {
1530
+ "epoch": 0.39,
1531
+ "learning_rate": 4.355614973262032e-05,
1532
+ "loss": 0.1567,
1533
+ "step": 2460
1534
+ },
1535
+ {
1536
+ "epoch": 0.4,
1537
+ "learning_rate": 4.3529411764705885e-05,
1538
+ "loss": 0.1287,
1539
+ "step": 2470
1540
+ },
1541
+ {
1542
+ "epoch": 0.4,
1543
+ "learning_rate": 4.3502673796791444e-05,
1544
+ "loss": 0.187,
1545
+ "step": 2480
1546
+ },
1547
+ {
1548
+ "epoch": 0.4,
1549
+ "learning_rate": 4.347593582887701e-05,
1550
+ "loss": 0.1354,
1551
+ "step": 2490
1552
+ },
1553
+ {
1554
+ "epoch": 0.4,
1555
+ "learning_rate": 4.344919786096257e-05,
1556
+ "loss": 0.0646,
1557
+ "step": 2500
1558
+ },
1559
+ {
1560
+ "epoch": 0.4,
1561
+ "eval_accuracy": 0.849320241691843,
1562
+ "eval_f1": 0.8472401124107073,
1563
+ "eval_loss": 1.3920754194259644,
1564
+ "eval_precision": 0.869443045077059,
1565
+ "eval_recall": 0.849320241691843,
1566
+ "eval_runtime": 10.5377,
1567
+ "eval_samples_per_second": 251.287,
1568
+ "eval_steps_per_second": 3.986,
1569
+ "step": 2500
1570
+ },
1571
+ {
1572
+ "epoch": 0.4,
1573
+ "learning_rate": 4.342245989304813e-05,
1574
+ "loss": 0.1651,
1575
+ "step": 2510
1576
+ },
1577
+ {
1578
+ "epoch": 0.4,
1579
+ "learning_rate": 4.3395721925133694e-05,
1580
+ "loss": 0.0982,
1581
+ "step": 2520
1582
+ },
1583
+ {
1584
+ "epoch": 0.4,
1585
+ "learning_rate": 4.336898395721925e-05,
1586
+ "loss": 0.2102,
1587
+ "step": 2530
1588
+ },
1589
+ {
1590
+ "epoch": 0.41,
1591
+ "learning_rate": 4.334224598930481e-05,
1592
+ "loss": 0.2102,
1593
+ "step": 2540
1594
+ },
1595
+ {
1596
+ "epoch": 0.41,
1597
+ "learning_rate": 4.331550802139038e-05,
1598
+ "loss": 0.2738,
1599
+ "step": 2550
1600
+ },
1601
+ {
1602
+ "epoch": 0.41,
1603
+ "learning_rate": 4.328877005347594e-05,
1604
+ "loss": 0.0848,
1605
+ "step": 2560
1606
+ },
1607
+ {
1608
+ "epoch": 0.41,
1609
+ "learning_rate": 4.3262032085561496e-05,
1610
+ "loss": 0.2072,
1611
+ "step": 2570
1612
+ },
1613
+ {
1614
+ "epoch": 0.41,
1615
+ "learning_rate": 4.323529411764706e-05,
1616
+ "loss": 0.6024,
1617
+ "step": 2580
1618
+ },
1619
+ {
1620
+ "epoch": 0.41,
1621
+ "learning_rate": 4.320855614973262e-05,
1622
+ "loss": 0.5888,
1623
+ "step": 2590
1624
+ },
1625
+ {
1626
+ "epoch": 0.42,
1627
+ "learning_rate": 4.318181818181819e-05,
1628
+ "loss": 0.1546,
1629
+ "step": 2600
1630
+ },
1631
+ {
1632
+ "epoch": 0.42,
1633
+ "learning_rate": 4.3155080213903746e-05,
1634
+ "loss": 0.2223,
1635
+ "step": 2610
1636
+ },
1637
+ {
1638
+ "epoch": 0.42,
1639
+ "learning_rate": 4.3128342245989305e-05,
1640
+ "loss": 0.2447,
1641
+ "step": 2620
1642
+ },
1643
+ {
1644
+ "epoch": 0.42,
1645
+ "learning_rate": 4.310160427807487e-05,
1646
+ "loss": 0.1098,
1647
+ "step": 2630
1648
+ },
1649
+ {
1650
+ "epoch": 0.42,
1651
+ "learning_rate": 4.307486631016043e-05,
1652
+ "loss": 0.3385,
1653
+ "step": 2640
1654
+ },
1655
+ {
1656
+ "epoch": 0.42,
1657
+ "learning_rate": 4.304812834224599e-05,
1658
+ "loss": 0.1067,
1659
+ "step": 2650
1660
+ },
1661
+ {
1662
+ "epoch": 0.43,
1663
+ "learning_rate": 4.3021390374331554e-05,
1664
+ "loss": 0.2155,
1665
+ "step": 2660
1666
+ },
1667
+ {
1668
+ "epoch": 0.43,
1669
+ "learning_rate": 4.2994652406417113e-05,
1670
+ "loss": 0.345,
1671
+ "step": 2670
1672
+ },
1673
+ {
1674
+ "epoch": 0.43,
1675
+ "learning_rate": 4.296791443850268e-05,
1676
+ "loss": 0.2231,
1677
+ "step": 2680
1678
+ },
1679
+ {
1680
+ "epoch": 0.43,
1681
+ "learning_rate": 4.294117647058823e-05,
1682
+ "loss": 0.2863,
1683
+ "step": 2690
1684
+ },
1685
+ {
1686
+ "epoch": 0.43,
1687
+ "learning_rate": 4.29144385026738e-05,
1688
+ "loss": 0.211,
1689
+ "step": 2700
1690
+ },
1691
+ {
1692
+ "epoch": 0.43,
1693
+ "learning_rate": 4.288770053475936e-05,
1694
+ "loss": 0.1859,
1695
+ "step": 2710
1696
+ },
1697
+ {
1698
+ "epoch": 0.44,
1699
+ "learning_rate": 4.286096256684492e-05,
1700
+ "loss": 0.4434,
1701
+ "step": 2720
1702
+ },
1703
+ {
1704
+ "epoch": 0.44,
1705
+ "learning_rate": 4.283422459893048e-05,
1706
+ "loss": 0.2067,
1707
+ "step": 2730
1708
+ },
1709
+ {
1710
+ "epoch": 0.44,
1711
+ "learning_rate": 4.280748663101605e-05,
1712
+ "loss": 0.2656,
1713
+ "step": 2740
1714
+ },
1715
+ {
1716
+ "epoch": 0.44,
1717
+ "learning_rate": 4.2780748663101606e-05,
1718
+ "loss": 0.3386,
1719
+ "step": 2750
1720
+ },
1721
+ {
1722
+ "epoch": 0.44,
1723
+ "learning_rate": 4.275401069518717e-05,
1724
+ "loss": 0.6852,
1725
+ "step": 2760
1726
+ },
1727
+ {
1728
+ "epoch": 0.44,
1729
+ "learning_rate": 4.2727272727272724e-05,
1730
+ "loss": 0.2283,
1731
+ "step": 2770
1732
+ },
1733
+ {
1734
+ "epoch": 0.44,
1735
+ "learning_rate": 4.270053475935829e-05,
1736
+ "loss": 0.2341,
1737
+ "step": 2780
1738
+ },
1739
+ {
1740
+ "epoch": 0.45,
1741
+ "learning_rate": 4.2673796791443856e-05,
1742
+ "loss": 0.268,
1743
+ "step": 2790
1744
+ },
1745
+ {
1746
+ "epoch": 0.45,
1747
+ "learning_rate": 4.2647058823529415e-05,
1748
+ "loss": 0.2749,
1749
+ "step": 2800
1750
+ },
1751
+ {
1752
+ "epoch": 0.45,
1753
+ "learning_rate": 4.2620320855614974e-05,
1754
+ "loss": 0.2868,
1755
+ "step": 2810
1756
+ },
1757
+ {
1758
+ "epoch": 0.45,
1759
+ "learning_rate": 4.259358288770054e-05,
1760
+ "loss": 0.2351,
1761
+ "step": 2820
1762
+ },
1763
+ {
1764
+ "epoch": 0.45,
1765
+ "learning_rate": 4.25668449197861e-05,
1766
+ "loss": 0.156,
1767
+ "step": 2830
1768
+ },
1769
+ {
1770
+ "epoch": 0.45,
1771
+ "learning_rate": 4.254010695187166e-05,
1772
+ "loss": 0.1897,
1773
+ "step": 2840
1774
+ },
1775
+ {
1776
+ "epoch": 0.46,
1777
+ "learning_rate": 4.251336898395722e-05,
1778
+ "loss": 0.2117,
1779
+ "step": 2850
1780
+ },
1781
+ {
1782
+ "epoch": 0.46,
1783
+ "learning_rate": 4.248663101604278e-05,
1784
+ "loss": 0.2102,
1785
+ "step": 2860
1786
+ },
1787
+ {
1788
+ "epoch": 0.46,
1789
+ "learning_rate": 4.245989304812835e-05,
1790
+ "loss": 0.3609,
1791
+ "step": 2870
1792
+ },
1793
+ {
1794
+ "epoch": 0.46,
1795
+ "learning_rate": 4.243315508021391e-05,
1796
+ "loss": 0.2409,
1797
+ "step": 2880
1798
+ },
1799
+ {
1800
+ "epoch": 0.46,
1801
+ "learning_rate": 4.2406417112299466e-05,
1802
+ "loss": 0.5209,
1803
+ "step": 2890
1804
+ },
1805
+ {
1806
+ "epoch": 0.46,
1807
+ "learning_rate": 4.2379679144385025e-05,
1808
+ "loss": 0.3209,
1809
+ "step": 2900
1810
+ },
1811
+ {
1812
+ "epoch": 0.47,
1813
+ "learning_rate": 4.235294117647059e-05,
1814
+ "loss": 0.3782,
1815
+ "step": 2910
1816
+ },
1817
+ {
1818
+ "epoch": 0.47,
1819
+ "learning_rate": 4.232620320855615e-05,
1820
+ "loss": 0.1297,
1821
+ "step": 2920
1822
+ },
1823
+ {
1824
+ "epoch": 0.47,
1825
+ "learning_rate": 4.229946524064171e-05,
1826
+ "loss": 0.365,
1827
+ "step": 2930
1828
+ },
1829
+ {
1830
+ "epoch": 0.47,
1831
+ "learning_rate": 4.2272727272727275e-05,
1832
+ "loss": 0.4153,
1833
+ "step": 2940
1834
+ },
1835
+ {
1836
+ "epoch": 0.47,
1837
+ "learning_rate": 4.224598930481284e-05,
1838
+ "loss": 0.1112,
1839
+ "step": 2950
1840
+ },
1841
+ {
1842
+ "epoch": 0.47,
1843
+ "learning_rate": 4.221925133689839e-05,
1844
+ "loss": 0.1813,
1845
+ "step": 2960
1846
+ },
1847
+ {
1848
+ "epoch": 0.48,
1849
+ "learning_rate": 4.219251336898396e-05,
1850
+ "loss": 0.1488,
1851
+ "step": 2970
1852
+ },
1853
+ {
1854
+ "epoch": 0.48,
1855
+ "learning_rate": 4.216577540106952e-05,
1856
+ "loss": 0.1983,
1857
+ "step": 2980
1858
+ },
1859
+ {
1860
+ "epoch": 0.48,
1861
+ "learning_rate": 4.2139037433155084e-05,
1862
+ "loss": 0.3848,
1863
+ "step": 2990
1864
+ },
1865
+ {
1866
+ "epoch": 0.48,
1867
+ "learning_rate": 4.211229946524064e-05,
1868
+ "loss": 0.3005,
1869
+ "step": 3000
1870
+ },
1871
+ {
1872
+ "epoch": 0.48,
1873
+ "eval_accuracy": 0.8425226586102719,
1874
+ "eval_f1": 0.8406242841944506,
1875
+ "eval_loss": 2.313152313232422,
1876
+ "eval_precision": 0.8596586910099404,
1877
+ "eval_recall": 0.8425226586102719,
1878
+ "eval_runtime": 10.5381,
1879
+ "eval_samples_per_second": 251.278,
1880
+ "eval_steps_per_second": 3.986,
1881
+ "step": 3000
1882
+ },
1883
+ {
1884
+ "epoch": 0.48,
1885
+ "learning_rate": 4.20855614973262e-05,
1886
+ "loss": 0.3137,
1887
+ "step": 3010
1888
+ },
1889
+ {
1890
+ "epoch": 0.48,
1891
+ "learning_rate": 4.205882352941177e-05,
1892
+ "loss": 0.3145,
1893
+ "step": 3020
1894
+ },
1895
+ {
1896
+ "epoch": 0.48,
1897
+ "learning_rate": 4.2032085561497334e-05,
1898
+ "loss": 0.1838,
1899
+ "step": 3030
1900
+ },
1901
+ {
1902
+ "epoch": 0.49,
1903
+ "learning_rate": 4.2005347593582886e-05,
1904
+ "loss": 0.328,
1905
+ "step": 3040
1906
+ },
1907
+ {
1908
+ "epoch": 0.49,
1909
+ "learning_rate": 4.197860962566845e-05,
1910
+ "loss": 0.2712,
1911
+ "step": 3050
1912
+ },
1913
+ {
1914
+ "epoch": 0.49,
1915
+ "learning_rate": 4.195187165775401e-05,
1916
+ "loss": 0.2833,
1917
+ "step": 3060
1918
+ },
1919
+ {
1920
+ "epoch": 0.49,
1921
+ "learning_rate": 4.1925133689839577e-05,
1922
+ "loss": 0.425,
1923
+ "step": 3070
1924
+ },
1925
+ {
1926
+ "epoch": 0.49,
1927
+ "learning_rate": 4.1898395721925136e-05,
1928
+ "loss": 0.227,
1929
+ "step": 3080
1930
+ },
1931
+ {
1932
+ "epoch": 0.49,
1933
+ "learning_rate": 4.1871657754010695e-05,
1934
+ "loss": 0.299,
1935
+ "step": 3090
1936
+ },
1937
+ {
1938
+ "epoch": 0.5,
1939
+ "learning_rate": 4.184491978609626e-05,
1940
+ "loss": 0.3831,
1941
+ "step": 3100
1942
+ },
1943
+ {
1944
+ "epoch": 0.5,
1945
+ "learning_rate": 4.181818181818182e-05,
1946
+ "loss": 0.1673,
1947
+ "step": 3110
1948
+ },
1949
+ {
1950
+ "epoch": 0.5,
1951
+ "learning_rate": 4.179144385026738e-05,
1952
+ "loss": 0.2316,
1953
+ "step": 3120
1954
+ },
1955
+ {
1956
+ "epoch": 0.5,
1957
+ "learning_rate": 4.1764705882352944e-05,
1958
+ "loss": 0.3018,
1959
+ "step": 3130
1960
+ },
1961
+ {
1962
+ "epoch": 0.5,
1963
+ "learning_rate": 4.17379679144385e-05,
1964
+ "loss": 0.1735,
1965
+ "step": 3140
1966
+ },
1967
+ {
1968
+ "epoch": 0.5,
1969
+ "learning_rate": 4.171122994652407e-05,
1970
+ "loss": 0.3896,
1971
+ "step": 3150
1972
+ },
1973
+ {
1974
+ "epoch": 0.51,
1975
+ "learning_rate": 4.168449197860963e-05,
1976
+ "loss": 0.166,
1977
+ "step": 3160
1978
+ },
1979
+ {
1980
+ "epoch": 0.51,
1981
+ "learning_rate": 4.165775401069519e-05,
1982
+ "loss": 0.1794,
1983
+ "step": 3170
1984
+ },
1985
+ {
1986
+ "epoch": 0.51,
1987
+ "learning_rate": 4.163101604278075e-05,
1988
+ "loss": 0.2399,
1989
+ "step": 3180
1990
+ },
1991
+ {
1992
+ "epoch": 0.51,
1993
+ "learning_rate": 4.160427807486631e-05,
1994
+ "loss": 0.331,
1995
+ "step": 3190
1996
+ },
1997
+ {
1998
+ "epoch": 0.51,
1999
+ "learning_rate": 4.157754010695187e-05,
2000
+ "loss": 0.4568,
2001
+ "step": 3200
2002
+ },
2003
+ {
2004
+ "epoch": 0.51,
2005
+ "learning_rate": 4.155080213903744e-05,
2006
+ "loss": 0.4554,
2007
+ "step": 3210
2008
+ },
2009
+ {
2010
+ "epoch": 0.52,
2011
+ "learning_rate": 4.1524064171122996e-05,
2012
+ "loss": 0.3125,
2013
+ "step": 3220
2014
+ },
2015
+ {
2016
+ "epoch": 0.52,
2017
+ "learning_rate": 4.149732620320856e-05,
2018
+ "loss": 0.2818,
2019
+ "step": 3230
2020
+ },
2021
+ {
2022
+ "epoch": 0.52,
2023
+ "learning_rate": 4.147058823529412e-05,
2024
+ "loss": 0.4162,
2025
+ "step": 3240
2026
+ },
2027
+ {
2028
+ "epoch": 0.52,
2029
+ "learning_rate": 4.144385026737968e-05,
2030
+ "loss": 0.4248,
2031
+ "step": 3250
2032
+ },
2033
+ {
2034
+ "epoch": 0.52,
2035
+ "learning_rate": 4.1417112299465246e-05,
2036
+ "loss": 0.3776,
2037
+ "step": 3260
2038
+ },
2039
+ {
2040
+ "epoch": 0.52,
2041
+ "learning_rate": 4.1390374331550805e-05,
2042
+ "loss": 0.3877,
2043
+ "step": 3270
2044
+ },
2045
+ {
2046
+ "epoch": 0.52,
2047
+ "learning_rate": 4.1363636363636364e-05,
2048
+ "loss": 0.0185,
2049
+ "step": 3280
2050
+ },
2051
+ {
2052
+ "epoch": 0.53,
2053
+ "learning_rate": 4.133689839572193e-05,
2054
+ "loss": 0.1879,
2055
+ "step": 3290
2056
+ },
2057
+ {
2058
+ "epoch": 0.53,
2059
+ "learning_rate": 4.131016042780749e-05,
2060
+ "loss": 0.1577,
2061
+ "step": 3300
2062
+ },
2063
+ {
2064
+ "epoch": 0.53,
2065
+ "learning_rate": 4.128342245989305e-05,
2066
+ "loss": 0.342,
2067
+ "step": 3310
2068
+ },
2069
+ {
2070
+ "epoch": 0.53,
2071
+ "learning_rate": 4.125668449197861e-05,
2072
+ "loss": 0.2412,
2073
+ "step": 3320
2074
+ },
2075
+ {
2076
+ "epoch": 0.53,
2077
+ "learning_rate": 4.122994652406417e-05,
2078
+ "loss": 0.2567,
2079
+ "step": 3330
2080
+ },
2081
+ {
2082
+ "epoch": 0.53,
2083
+ "learning_rate": 4.120320855614974e-05,
2084
+ "loss": 0.2748,
2085
+ "step": 3340
2086
+ },
2087
+ {
2088
+ "epoch": 0.54,
2089
+ "learning_rate": 4.11764705882353e-05,
2090
+ "loss": 0.1134,
2091
+ "step": 3350
2092
+ },
2093
+ {
2094
+ "epoch": 0.54,
2095
+ "learning_rate": 4.1149732620320856e-05,
2096
+ "loss": 0.3265,
2097
+ "step": 3360
2098
+ },
2099
+ {
2100
+ "epoch": 0.54,
2101
+ "learning_rate": 4.112299465240642e-05,
2102
+ "loss": 0.2037,
2103
+ "step": 3370
2104
+ },
2105
+ {
2106
+ "epoch": 0.54,
2107
+ "learning_rate": 4.109625668449198e-05,
2108
+ "loss": 0.2894,
2109
+ "step": 3380
2110
+ },
2111
+ {
2112
+ "epoch": 0.54,
2113
+ "learning_rate": 4.106951871657754e-05,
2114
+ "loss": 0.3418,
2115
+ "step": 3390
2116
+ },
2117
+ {
2118
+ "epoch": 0.54,
2119
+ "learning_rate": 4.10427807486631e-05,
2120
+ "loss": 0.3356,
2121
+ "step": 3400
2122
+ },
2123
+ {
2124
+ "epoch": 0.55,
2125
+ "learning_rate": 4.1016042780748665e-05,
2126
+ "loss": 0.228,
2127
+ "step": 3410
2128
+ },
2129
+ {
2130
+ "epoch": 0.55,
2131
+ "learning_rate": 4.098930481283423e-05,
2132
+ "loss": 0.3992,
2133
+ "step": 3420
2134
+ },
2135
+ {
2136
+ "epoch": 0.55,
2137
+ "learning_rate": 4.096256684491978e-05,
2138
+ "loss": 0.2551,
2139
+ "step": 3430
2140
+ },
2141
+ {
2142
+ "epoch": 0.55,
2143
+ "learning_rate": 4.093582887700535e-05,
2144
+ "loss": 0.184,
2145
+ "step": 3440
2146
+ },
2147
+ {
2148
+ "epoch": 0.55,
2149
+ "learning_rate": 4.0909090909090915e-05,
2150
+ "loss": 0.3881,
2151
+ "step": 3450
2152
+ },
2153
+ {
2154
+ "epoch": 0.55,
2155
+ "learning_rate": 4.0882352941176474e-05,
2156
+ "loss": 0.3422,
2157
+ "step": 3460
2158
+ },
2159
+ {
2160
+ "epoch": 0.56,
2161
+ "learning_rate": 4.085561497326203e-05,
2162
+ "loss": 0.2786,
2163
+ "step": 3470
2164
+ },
2165
+ {
2166
+ "epoch": 0.56,
2167
+ "learning_rate": 4.082887700534759e-05,
2168
+ "loss": 0.2854,
2169
+ "step": 3480
2170
+ },
2171
+ {
2172
+ "epoch": 0.56,
2173
+ "learning_rate": 4.080213903743316e-05,
2174
+ "loss": 0.4252,
2175
+ "step": 3490
2176
+ },
2177
+ {
2178
+ "epoch": 0.56,
2179
+ "learning_rate": 4.0775401069518723e-05,
2180
+ "loss": 0.1091,
2181
+ "step": 3500
2182
+ },
2183
+ {
2184
+ "epoch": 0.56,
2185
+ "eval_accuracy": 0.8459214501510574,
2186
+ "eval_f1": 0.843780419456653,
2187
+ "eval_loss": 2.0897836685180664,
2188
+ "eval_precision": 0.8659851431247721,
2189
+ "eval_recall": 0.8459214501510574,
2190
+ "eval_runtime": 10.5263,
2191
+ "eval_samples_per_second": 251.559,
2192
+ "eval_steps_per_second": 3.99,
2193
+ "step": 3500
2194
+ },
2195
+ {
2196
+ "epoch": 0.56,
2197
+ "learning_rate": 4.0748663101604276e-05,
2198
+ "loss": 0.3208,
2199
+ "step": 3510
2200
+ },
2201
+ {
2202
+ "epoch": 0.56,
2203
+ "learning_rate": 4.072192513368984e-05,
2204
+ "loss": 0.0881,
2205
+ "step": 3520
2206
+ },
2207
+ {
2208
+ "epoch": 0.56,
2209
+ "learning_rate": 4.06951871657754e-05,
2210
+ "loss": 0.1677,
2211
+ "step": 3530
2212
+ },
2213
+ {
2214
+ "epoch": 0.57,
2215
+ "learning_rate": 4.0668449197860966e-05,
2216
+ "loss": 0.049,
2217
+ "step": 3540
2218
+ },
2219
+ {
2220
+ "epoch": 0.57,
2221
+ "learning_rate": 4.0641711229946525e-05,
2222
+ "loss": 0.1629,
2223
+ "step": 3550
2224
+ },
2225
+ {
2226
+ "epoch": 0.57,
2227
+ "learning_rate": 4.0614973262032084e-05,
2228
+ "loss": 0.2262,
2229
+ "step": 3560
2230
+ },
2231
+ {
2232
+ "epoch": 0.57,
2233
+ "learning_rate": 4.058823529411765e-05,
2234
+ "loss": 0.2134,
2235
+ "step": 3570
2236
+ },
2237
+ {
2238
+ "epoch": 0.57,
2239
+ "learning_rate": 4.0561497326203216e-05,
2240
+ "loss": 0.3984,
2241
+ "step": 3580
2242
+ },
2243
+ {
2244
+ "epoch": 0.57,
2245
+ "learning_rate": 4.053475935828877e-05,
2246
+ "loss": 0.1527,
2247
+ "step": 3590
2248
+ },
2249
+ {
2250
+ "epoch": 0.58,
2251
+ "learning_rate": 4.0508021390374334e-05,
2252
+ "loss": 0.1532,
2253
+ "step": 3600
2254
+ },
2255
+ {
2256
+ "epoch": 0.58,
2257
+ "learning_rate": 4.048128342245989e-05,
2258
+ "loss": 0.1918,
2259
+ "step": 3610
2260
+ },
2261
+ {
2262
+ "epoch": 0.58,
2263
+ "learning_rate": 4.045454545454546e-05,
2264
+ "loss": 0.2594,
2265
+ "step": 3620
2266
+ },
2267
+ {
2268
+ "epoch": 0.58,
2269
+ "learning_rate": 4.042780748663102e-05,
2270
+ "loss": 0.1803,
2271
+ "step": 3630
2272
+ },
2273
+ {
2274
+ "epoch": 0.58,
2275
+ "learning_rate": 4.040106951871658e-05,
2276
+ "loss": 0.2573,
2277
+ "step": 3640
2278
+ },
2279
+ {
2280
+ "epoch": 0.58,
2281
+ "learning_rate": 4.037433155080214e-05,
2282
+ "loss": 0.2865,
2283
+ "step": 3650
2284
+ },
2285
+ {
2286
+ "epoch": 0.59,
2287
+ "learning_rate": 4.03475935828877e-05,
2288
+ "loss": 0.053,
2289
+ "step": 3660
2290
+ },
2291
+ {
2292
+ "epoch": 0.59,
2293
+ "learning_rate": 4.032085561497326e-05,
2294
+ "loss": 0.4603,
2295
+ "step": 3670
2296
+ },
2297
+ {
2298
+ "epoch": 0.59,
2299
+ "learning_rate": 4.029411764705883e-05,
2300
+ "loss": 0.1409,
2301
+ "step": 3680
2302
+ },
2303
+ {
2304
+ "epoch": 0.59,
2305
+ "learning_rate": 4.0267379679144386e-05,
2306
+ "loss": 0.2257,
2307
+ "step": 3690
2308
+ },
2309
+ {
2310
+ "epoch": 0.59,
2311
+ "learning_rate": 4.024064171122995e-05,
2312
+ "loss": 0.2827,
2313
+ "step": 3700
2314
+ },
2315
+ {
2316
+ "epoch": 0.59,
2317
+ "learning_rate": 4.021390374331551e-05,
2318
+ "loss": 0.2166,
2319
+ "step": 3710
2320
+ },
2321
+ {
2322
+ "epoch": 0.6,
2323
+ "learning_rate": 4.018716577540107e-05,
2324
+ "loss": 0.2091,
2325
+ "step": 3720
2326
+ },
2327
+ {
2328
+ "epoch": 0.6,
2329
+ "learning_rate": 4.0160427807486636e-05,
2330
+ "loss": 0.4851,
2331
+ "step": 3730
2332
+ },
2333
+ {
2334
+ "epoch": 0.6,
2335
+ "learning_rate": 4.0133689839572195e-05,
2336
+ "loss": 0.3742,
2337
+ "step": 3740
2338
+ },
2339
+ {
2340
+ "epoch": 0.6,
2341
+ "learning_rate": 4.0106951871657754e-05,
2342
+ "loss": 0.492,
2343
+ "step": 3750
2344
+ },
2345
+ {
2346
+ "epoch": 0.6,
2347
+ "learning_rate": 4.008021390374332e-05,
2348
+ "loss": 0.5621,
2349
+ "step": 3760
2350
+ },
2351
+ {
2352
+ "epoch": 0.6,
2353
+ "learning_rate": 4.005347593582888e-05,
2354
+ "loss": 0.2801,
2355
+ "step": 3770
2356
+ },
2357
+ {
2358
+ "epoch": 0.6,
2359
+ "learning_rate": 4.002673796791444e-05,
2360
+ "loss": 0.2604,
2361
+ "step": 3780
2362
+ },
2363
+ {
2364
+ "epoch": 0.61,
2365
+ "learning_rate": 4e-05,
2366
+ "loss": 0.3608,
2367
+ "step": 3790
2368
+ },
2369
+ {
2370
+ "epoch": 0.61,
2371
+ "learning_rate": 3.997326203208556e-05,
2372
+ "loss": 0.167,
2373
+ "step": 3800
2374
+ },
2375
+ {
2376
+ "epoch": 0.61,
2377
+ "learning_rate": 3.994652406417113e-05,
2378
+ "loss": 0.4472,
2379
+ "step": 3810
2380
+ },
2381
+ {
2382
+ "epoch": 0.61,
2383
+ "learning_rate": 3.991978609625669e-05,
2384
+ "loss": 0.3022,
2385
+ "step": 3820
2386
+ },
2387
+ {
2388
+ "epoch": 0.61,
2389
+ "learning_rate": 3.9893048128342246e-05,
2390
+ "loss": 0.3179,
2391
+ "step": 3830
2392
+ },
2393
+ {
2394
+ "epoch": 0.61,
2395
+ "learning_rate": 3.986631016042781e-05,
2396
+ "loss": 0.2939,
2397
+ "step": 3840
2398
+ },
2399
+ {
2400
+ "epoch": 0.62,
2401
+ "learning_rate": 3.983957219251337e-05,
2402
+ "loss": 0.013,
2403
+ "step": 3850
2404
+ },
2405
+ {
2406
+ "epoch": 0.62,
2407
+ "learning_rate": 3.981283422459893e-05,
2408
+ "loss": 0.1864,
2409
+ "step": 3860
2410
+ },
2411
+ {
2412
+ "epoch": 0.62,
2413
+ "learning_rate": 3.9786096256684496e-05,
2414
+ "loss": 0.3368,
2415
+ "step": 3870
2416
+ },
2417
+ {
2418
+ "epoch": 0.62,
2419
+ "learning_rate": 3.9759358288770055e-05,
2420
+ "loss": 0.1275,
2421
+ "step": 3880
2422
+ },
2423
+ {
2424
+ "epoch": 0.62,
2425
+ "learning_rate": 3.973262032085562e-05,
2426
+ "loss": 0.1441,
2427
+ "step": 3890
2428
+ },
2429
+ {
2430
+ "epoch": 0.62,
2431
+ "learning_rate": 3.970588235294117e-05,
2432
+ "loss": 0.358,
2433
+ "step": 3900
2434
+ },
2435
+ {
2436
+ "epoch": 0.63,
2437
+ "learning_rate": 3.967914438502674e-05,
2438
+ "loss": 0.1918,
2439
+ "step": 3910
2440
+ },
2441
+ {
2442
+ "epoch": 0.63,
2443
+ "learning_rate": 3.9652406417112305e-05,
2444
+ "loss": 0.2604,
2445
+ "step": 3920
2446
+ },
2447
+ {
2448
+ "epoch": 0.63,
2449
+ "learning_rate": 3.9625668449197864e-05,
2450
+ "loss": 0.0787,
2451
+ "step": 3930
2452
+ },
2453
+ {
2454
+ "epoch": 0.63,
2455
+ "learning_rate": 3.959893048128342e-05,
2456
+ "loss": 0.2841,
2457
+ "step": 3940
2458
+ },
2459
+ {
2460
+ "epoch": 0.63,
2461
+ "learning_rate": 3.957219251336899e-05,
2462
+ "loss": 0.4061,
2463
+ "step": 3950
2464
+ },
2465
+ {
2466
+ "epoch": 0.63,
2467
+ "learning_rate": 3.954545454545455e-05,
2468
+ "loss": 0.3767,
2469
+ "step": 3960
2470
+ },
2471
+ {
2472
+ "epoch": 0.64,
2473
+ "learning_rate": 3.951871657754011e-05,
2474
+ "loss": 0.1036,
2475
+ "step": 3970
2476
+ },
2477
+ {
2478
+ "epoch": 0.64,
2479
+ "learning_rate": 3.9491978609625666e-05,
2480
+ "loss": 0.38,
2481
+ "step": 3980
2482
+ },
2483
+ {
2484
+ "epoch": 0.64,
2485
+ "learning_rate": 3.946524064171123e-05,
2486
+ "loss": 0.1721,
2487
+ "step": 3990
2488
+ },
2489
+ {
2490
+ "epoch": 0.64,
2491
+ "learning_rate": 3.94385026737968e-05,
2492
+ "loss": 0.2176,
2493
+ "step": 4000
2494
+ },
2495
+ {
2496
+ "epoch": 0.64,
2497
+ "eval_accuracy": 0.847809667673716,
2498
+ "eval_f1": 0.8461258728150335,
2499
+ "eval_loss": 2.1692090034484863,
2500
+ "eval_precision": 0.8637304048866307,
2501
+ "eval_recall": 0.847809667673716,
2502
+ "eval_runtime": 10.5307,
2503
+ "eval_samples_per_second": 251.455,
2504
+ "eval_steps_per_second": 3.988,
2505
+ "step": 4000
2506
+ },
2507
+ {
2508
+ "epoch": 0.64,
2509
+ "learning_rate": 3.9411764705882356e-05,
2510
+ "loss": 0.1731,
2511
+ "step": 4010
2512
+ },
2513
+ {
2514
+ "epoch": 0.64,
2515
+ "learning_rate": 3.9385026737967915e-05,
2516
+ "loss": 0.0994,
2517
+ "step": 4020
2518
+ },
2519
+ {
2520
+ "epoch": 0.64,
2521
+ "learning_rate": 3.9358288770053474e-05,
2522
+ "loss": 0.3668,
2523
+ "step": 4030
2524
+ },
2525
+ {
2526
+ "epoch": 0.65,
2527
+ "learning_rate": 3.933155080213904e-05,
2528
+ "loss": 0.6184,
2529
+ "step": 4040
2530
+ },
2531
+ {
2532
+ "epoch": 0.65,
2533
+ "learning_rate": 3.93048128342246e-05,
2534
+ "loss": 0.1739,
2535
+ "step": 4050
2536
+ },
2537
+ {
2538
+ "epoch": 0.65,
2539
+ "learning_rate": 3.927807486631016e-05,
2540
+ "loss": 0.3273,
2541
+ "step": 4060
2542
+ },
2543
+ {
2544
+ "epoch": 0.65,
2545
+ "learning_rate": 3.9251336898395724e-05,
2546
+ "loss": 0.2285,
2547
+ "step": 4070
2548
+ },
2549
+ {
2550
+ "epoch": 0.65,
2551
+ "learning_rate": 3.922459893048129e-05,
2552
+ "loss": 0.2541,
2553
+ "step": 4080
2554
+ },
2555
+ {
2556
+ "epoch": 0.65,
2557
+ "learning_rate": 3.919786096256685e-05,
2558
+ "loss": 0.3116,
2559
+ "step": 4090
2560
+ },
2561
+ {
2562
+ "epoch": 0.66,
2563
+ "learning_rate": 3.917112299465241e-05,
2564
+ "loss": 0.4674,
2565
+ "step": 4100
2566
+ },
2567
+ {
2568
+ "epoch": 0.66,
2569
+ "learning_rate": 3.914438502673797e-05,
2570
+ "loss": 0.2002,
2571
+ "step": 4110
2572
+ },
2573
+ {
2574
+ "epoch": 0.66,
2575
+ "learning_rate": 3.911764705882353e-05,
2576
+ "loss": 0.3206,
2577
+ "step": 4120
2578
+ },
2579
+ {
2580
+ "epoch": 0.66,
2581
+ "learning_rate": 3.909090909090909e-05,
2582
+ "loss": 0.4748,
2583
+ "step": 4130
2584
+ },
2585
+ {
2586
+ "epoch": 0.66,
2587
+ "learning_rate": 3.906417112299465e-05,
2588
+ "loss": 0.1869,
2589
+ "step": 4140
2590
+ },
2591
+ {
2592
+ "epoch": 0.66,
2593
+ "learning_rate": 3.903743315508022e-05,
2594
+ "loss": 0.1935,
2595
+ "step": 4150
2596
+ },
2597
+ {
2598
+ "epoch": 0.67,
2599
+ "learning_rate": 3.901069518716578e-05,
2600
+ "loss": 0.2528,
2601
+ "step": 4160
2602
+ },
2603
+ {
2604
+ "epoch": 0.67,
2605
+ "learning_rate": 3.8983957219251335e-05,
2606
+ "loss": 0.275,
2607
+ "step": 4170
2608
+ },
2609
+ {
2610
+ "epoch": 0.67,
2611
+ "learning_rate": 3.89572192513369e-05,
2612
+ "loss": 0.2676,
2613
+ "step": 4180
2614
+ },
2615
+ {
2616
+ "epoch": 0.67,
2617
+ "learning_rate": 3.893048128342246e-05,
2618
+ "loss": 0.3452,
2619
+ "step": 4190
2620
+ },
2621
+ {
2622
+ "epoch": 0.67,
2623
+ "learning_rate": 3.8903743315508025e-05,
2624
+ "loss": 0.2249,
2625
+ "step": 4200
2626
+ },
2627
+ {
2628
+ "epoch": 0.67,
2629
+ "learning_rate": 3.8877005347593584e-05,
2630
+ "loss": 0.3211,
2631
+ "step": 4210
2632
+ },
2633
+ {
2634
+ "epoch": 0.68,
2635
+ "learning_rate": 3.8850267379679143e-05,
2636
+ "loss": 0.2125,
2637
+ "step": 4220
2638
+ },
2639
+ {
2640
+ "epoch": 0.68,
2641
+ "learning_rate": 3.882352941176471e-05,
2642
+ "loss": 0.2449,
2643
+ "step": 4230
2644
+ },
2645
+ {
2646
+ "epoch": 0.68,
2647
+ "learning_rate": 3.879679144385027e-05,
2648
+ "loss": 0.1108,
2649
+ "step": 4240
2650
+ },
2651
+ {
2652
+ "epoch": 0.68,
2653
+ "learning_rate": 3.877005347593583e-05,
2654
+ "loss": 0.3127,
2655
+ "step": 4250
2656
+ },
2657
+ {
2658
+ "epoch": 0.68,
2659
+ "learning_rate": 3.874331550802139e-05,
2660
+ "loss": 0.1238,
2661
+ "step": 4260
2662
+ },
2663
+ {
2664
+ "epoch": 0.68,
2665
+ "learning_rate": 3.871657754010695e-05,
2666
+ "loss": 0.2596,
2667
+ "step": 4270
2668
+ },
2669
+ {
2670
+ "epoch": 0.68,
2671
+ "learning_rate": 3.868983957219252e-05,
2672
+ "loss": 0.36,
2673
+ "step": 4280
2674
+ },
2675
+ {
2676
+ "epoch": 0.69,
2677
+ "learning_rate": 3.866310160427808e-05,
2678
+ "loss": 0.2548,
2679
+ "step": 4290
2680
+ },
2681
+ {
2682
+ "epoch": 0.69,
2683
+ "learning_rate": 3.8636363636363636e-05,
2684
+ "loss": 0.1942,
2685
+ "step": 4300
2686
+ },
2687
+ {
2688
+ "epoch": 0.69,
2689
+ "learning_rate": 3.86096256684492e-05,
2690
+ "loss": 0.1626,
2691
+ "step": 4310
2692
+ },
2693
+ {
2694
+ "epoch": 0.69,
2695
+ "learning_rate": 3.858288770053476e-05,
2696
+ "loss": 0.1887,
2697
+ "step": 4320
2698
+ },
2699
+ {
2700
+ "epoch": 0.69,
2701
+ "learning_rate": 3.855614973262032e-05,
2702
+ "loss": 0.3538,
2703
+ "step": 4330
2704
+ },
2705
+ {
2706
+ "epoch": 0.69,
2707
+ "learning_rate": 3.8529411764705886e-05,
2708
+ "loss": 0.2309,
2709
+ "step": 4340
2710
+ },
2711
+ {
2712
+ "epoch": 0.7,
2713
+ "learning_rate": 3.8502673796791445e-05,
2714
+ "loss": 0.147,
2715
+ "step": 4350
2716
+ },
2717
+ {
2718
+ "epoch": 0.7,
2719
+ "learning_rate": 3.847593582887701e-05,
2720
+ "loss": 0.1799,
2721
+ "step": 4360
2722
+ },
2723
+ {
2724
+ "epoch": 0.7,
2725
+ "learning_rate": 3.844919786096257e-05,
2726
+ "loss": 0.2291,
2727
+ "step": 4370
2728
+ },
2729
+ {
2730
+ "epoch": 0.7,
2731
+ "learning_rate": 3.842245989304813e-05,
2732
+ "loss": 0.39,
2733
+ "step": 4380
2734
+ },
2735
+ {
2736
+ "epoch": 0.7,
2737
+ "learning_rate": 3.8395721925133694e-05,
2738
+ "loss": 0.0407,
2739
+ "step": 4390
2740
+ },
2741
+ {
2742
+ "epoch": 0.7,
2743
+ "learning_rate": 3.8368983957219254e-05,
2744
+ "loss": 0.2575,
2745
+ "step": 4400
2746
+ },
2747
+ {
2748
+ "epoch": 0.71,
2749
+ "learning_rate": 3.834224598930481e-05,
2750
+ "loss": 0.1994,
2751
+ "step": 4410
2752
+ },
2753
+ {
2754
+ "epoch": 0.71,
2755
+ "learning_rate": 3.831550802139038e-05,
2756
+ "loss": 0.1931,
2757
+ "step": 4420
2758
+ },
2759
+ {
2760
+ "epoch": 0.71,
2761
+ "learning_rate": 3.828877005347594e-05,
2762
+ "loss": 0.4772,
2763
+ "step": 4430
2764
+ },
2765
+ {
2766
+ "epoch": 0.71,
2767
+ "learning_rate": 3.82620320855615e-05,
2768
+ "loss": 0.406,
2769
+ "step": 4440
2770
+ },
2771
+ {
2772
+ "epoch": 0.71,
2773
+ "learning_rate": 3.8235294117647055e-05,
2774
+ "loss": 0.3865,
2775
+ "step": 4450
2776
+ },
2777
+ {
2778
+ "epoch": 0.71,
2779
+ "learning_rate": 3.820855614973262e-05,
2780
+ "loss": 0.2322,
2781
+ "step": 4460
2782
+ },
2783
+ {
2784
+ "epoch": 0.72,
2785
+ "learning_rate": 3.818181818181819e-05,
2786
+ "loss": 0.3505,
2787
+ "step": 4470
2788
+ },
2789
+ {
2790
+ "epoch": 0.72,
2791
+ "learning_rate": 3.8155080213903746e-05,
2792
+ "loss": 0.2413,
2793
+ "step": 4480
2794
+ },
2795
+ {
2796
+ "epoch": 0.72,
2797
+ "learning_rate": 3.8128342245989305e-05,
2798
+ "loss": 0.1657,
2799
+ "step": 4490
2800
+ },
2801
+ {
2802
+ "epoch": 0.72,
2803
+ "learning_rate": 3.810160427807487e-05,
2804
+ "loss": 0.1567,
2805
+ "step": 4500
2806
+ },
2807
+ {
2808
+ "epoch": 0.72,
2809
+ "eval_accuracy": 0.8462990936555891,
2810
+ "eval_f1": 0.8442050087782637,
2811
+ "eval_loss": 2.184410333633423,
2812
+ "eval_precision": 0.8659758356860412,
2813
+ "eval_recall": 0.8462990936555891,
2814
+ "eval_runtime": 10.5461,
2815
+ "eval_samples_per_second": 251.088,
2816
+ "eval_steps_per_second": 3.983,
2817
+ "step": 4500
2818
+ },
2819
+ {
2820
+ "epoch": 0.72,
2821
+ "learning_rate": 3.807486631016043e-05,
2822
+ "loss": 0.2413,
2823
+ "step": 4510
2824
+ },
2825
+ {
2826
+ "epoch": 0.72,
2827
+ "learning_rate": 3.804812834224599e-05,
2828
+ "loss": 0.5433,
2829
+ "step": 4520
2830
+ },
2831
+ {
2832
+ "epoch": 0.72,
2833
+ "learning_rate": 3.802139037433155e-05,
2834
+ "loss": 0.326,
2835
+ "step": 4530
2836
+ },
2837
+ {
2838
+ "epoch": 0.73,
2839
+ "learning_rate": 3.7994652406417114e-05,
2840
+ "loss": 0.3755,
2841
+ "step": 4540
2842
+ },
2843
+ {
2844
+ "epoch": 0.73,
2845
+ "learning_rate": 3.796791443850268e-05,
2846
+ "loss": 0.2589,
2847
+ "step": 4550
2848
+ },
2849
+ {
2850
+ "epoch": 0.73,
2851
+ "learning_rate": 3.794117647058824e-05,
2852
+ "loss": 0.2539,
2853
+ "step": 4560
2854
+ },
2855
+ {
2856
+ "epoch": 0.73,
2857
+ "learning_rate": 3.79144385026738e-05,
2858
+ "loss": 0.2227,
2859
+ "step": 4570
2860
+ },
2861
+ {
2862
+ "epoch": 0.73,
2863
+ "learning_rate": 3.7887700534759364e-05,
2864
+ "loss": 0.1897,
2865
+ "step": 4580
2866
+ },
2867
+ {
2868
+ "epoch": 0.73,
2869
+ "learning_rate": 3.786096256684492e-05,
2870
+ "loss": 0.2462,
2871
+ "step": 4590
2872
+ },
2873
+ {
2874
+ "epoch": 0.74,
2875
+ "learning_rate": 3.783422459893048e-05,
2876
+ "loss": 0.1308,
2877
+ "step": 4600
2878
+ },
2879
+ {
2880
+ "epoch": 0.74,
2881
+ "learning_rate": 3.780748663101604e-05,
2882
+ "loss": 0.2567,
2883
+ "step": 4610
2884
+ },
2885
+ {
2886
+ "epoch": 0.74,
2887
+ "learning_rate": 3.7780748663101607e-05,
2888
+ "loss": 0.3583,
2889
+ "step": 4620
2890
+ },
2891
+ {
2892
+ "epoch": 0.74,
2893
+ "learning_rate": 3.775401069518717e-05,
2894
+ "loss": 0.3253,
2895
+ "step": 4630
2896
+ },
2897
+ {
2898
+ "epoch": 0.74,
2899
+ "learning_rate": 3.7727272727272725e-05,
2900
+ "loss": 0.2606,
2901
+ "step": 4640
2902
+ },
2903
+ {
2904
+ "epoch": 0.74,
2905
+ "learning_rate": 3.770053475935829e-05,
2906
+ "loss": 0.2173,
2907
+ "step": 4650
2908
+ },
2909
+ {
2910
+ "epoch": 0.75,
2911
+ "learning_rate": 3.767379679144385e-05,
2912
+ "loss": 0.2203,
2913
+ "step": 4660
2914
+ },
2915
+ {
2916
+ "epoch": 0.75,
2917
+ "learning_rate": 3.7647058823529415e-05,
2918
+ "loss": 0.059,
2919
+ "step": 4670
2920
+ },
2921
+ {
2922
+ "epoch": 0.75,
2923
+ "learning_rate": 3.7620320855614974e-05,
2924
+ "loss": 0.3312,
2925
+ "step": 4680
2926
+ },
2927
+ {
2928
+ "epoch": 0.75,
2929
+ "learning_rate": 3.759358288770053e-05,
2930
+ "loss": 0.1597,
2931
+ "step": 4690
2932
+ },
2933
+ {
2934
+ "epoch": 0.75,
2935
+ "learning_rate": 3.75668449197861e-05,
2936
+ "loss": 0.3424,
2937
+ "step": 4700
2938
+ },
2939
+ {
2940
+ "epoch": 0.75,
2941
+ "learning_rate": 3.7540106951871665e-05,
2942
+ "loss": 0.2294,
2943
+ "step": 4710
2944
+ },
2945
+ {
2946
+ "epoch": 0.76,
2947
+ "learning_rate": 3.751336898395722e-05,
2948
+ "loss": 0.1233,
2949
+ "step": 4720
2950
+ },
2951
+ {
2952
+ "epoch": 0.76,
2953
+ "learning_rate": 3.748663101604278e-05,
2954
+ "loss": 0.1573,
2955
+ "step": 4730
2956
+ },
2957
+ {
2958
+ "epoch": 0.76,
2959
+ "learning_rate": 3.745989304812834e-05,
2960
+ "loss": 0.0886,
2961
+ "step": 4740
2962
+ },
2963
+ {
2964
+ "epoch": 0.76,
2965
+ "learning_rate": 3.743315508021391e-05,
2966
+ "loss": 0.1607,
2967
+ "step": 4750
2968
+ },
2969
+ {
2970
+ "epoch": 0.76,
2971
+ "learning_rate": 3.740641711229947e-05,
2972
+ "loss": 0.2734,
2973
+ "step": 4760
2974
+ },
2975
+ {
2976
+ "epoch": 0.76,
2977
+ "learning_rate": 3.7379679144385026e-05,
2978
+ "loss": 0.1011,
2979
+ "step": 4770
2980
+ },
2981
+ {
2982
+ "epoch": 0.76,
2983
+ "learning_rate": 3.735294117647059e-05,
2984
+ "loss": 0.395,
2985
+ "step": 4780
2986
+ },
2987
+ {
2988
+ "epoch": 0.77,
2989
+ "learning_rate": 3.732620320855615e-05,
2990
+ "loss": 0.225,
2991
+ "step": 4790
2992
+ },
2993
+ {
2994
+ "epoch": 0.77,
2995
+ "learning_rate": 3.729946524064171e-05,
2996
+ "loss": 0.3241,
2997
+ "step": 4800
2998
+ },
2999
+ {
3000
+ "epoch": 0.77,
3001
+ "learning_rate": 3.7272727272727276e-05,
3002
+ "loss": 0.3954,
3003
+ "step": 4810
3004
+ },
3005
+ {
3006
+ "epoch": 0.77,
3007
+ "learning_rate": 3.7245989304812835e-05,
3008
+ "loss": 0.1394,
3009
+ "step": 4820
3010
+ },
3011
+ {
3012
+ "epoch": 0.77,
3013
+ "learning_rate": 3.72192513368984e-05,
3014
+ "loss": 0.2615,
3015
+ "step": 4830
3016
+ },
3017
+ {
3018
+ "epoch": 0.77,
3019
+ "learning_rate": 3.719251336898396e-05,
3020
+ "loss": 0.1806,
3021
+ "step": 4840
3022
+ },
3023
+ {
3024
+ "epoch": 0.78,
3025
+ "learning_rate": 3.716577540106952e-05,
3026
+ "loss": 0.2261,
3027
+ "step": 4850
3028
+ },
3029
+ {
3030
+ "epoch": 0.78,
3031
+ "learning_rate": 3.7139037433155084e-05,
3032
+ "loss": 0.3843,
3033
+ "step": 4860
3034
+ },
3035
+ {
3036
+ "epoch": 0.78,
3037
+ "learning_rate": 3.7112299465240643e-05,
3038
+ "loss": 0.3096,
3039
+ "step": 4870
3040
+ },
3041
+ {
3042
+ "epoch": 0.78,
3043
+ "learning_rate": 3.70855614973262e-05,
3044
+ "loss": 0.2883,
3045
+ "step": 4880
3046
+ },
3047
+ {
3048
+ "epoch": 0.78,
3049
+ "learning_rate": 3.705882352941177e-05,
3050
+ "loss": 0.3197,
3051
+ "step": 4890
3052
+ },
3053
+ {
3054
+ "epoch": 0.78,
3055
+ "learning_rate": 3.703208556149733e-05,
3056
+ "loss": 0.5345,
3057
+ "step": 4900
3058
+ },
3059
+ {
3060
+ "epoch": 0.79,
3061
+ "learning_rate": 3.700534759358289e-05,
3062
+ "loss": 0.3752,
3063
+ "step": 4910
3064
+ },
3065
+ {
3066
+ "epoch": 0.79,
3067
+ "learning_rate": 3.697860962566845e-05,
3068
+ "loss": 0.2157,
3069
+ "step": 4920
3070
+ },
3071
+ {
3072
+ "epoch": 0.79,
3073
+ "learning_rate": 3.695187165775401e-05,
3074
+ "loss": 0.1772,
3075
+ "step": 4930
3076
+ },
3077
+ {
3078
+ "epoch": 0.79,
3079
+ "learning_rate": 3.692513368983958e-05,
3080
+ "loss": 0.2575,
3081
+ "step": 4940
3082
+ },
3083
+ {
3084
+ "epoch": 0.79,
3085
+ "learning_rate": 3.6898395721925136e-05,
3086
+ "loss": 0.2138,
3087
+ "step": 4950
3088
+ },
3089
+ {
3090
+ "epoch": 0.79,
3091
+ "learning_rate": 3.6871657754010695e-05,
3092
+ "loss": 0.2703,
3093
+ "step": 4960
3094
+ },
3095
+ {
3096
+ "epoch": 0.8,
3097
+ "learning_rate": 3.684491978609626e-05,
3098
+ "loss": 0.4287,
3099
+ "step": 4970
3100
+ },
3101
+ {
3102
+ "epoch": 0.8,
3103
+ "learning_rate": 3.681818181818182e-05,
3104
+ "loss": 0.1396,
3105
+ "step": 4980
3106
+ },
3107
+ {
3108
+ "epoch": 0.8,
3109
+ "learning_rate": 3.679144385026738e-05,
3110
+ "loss": 0.3636,
3111
+ "step": 4990
3112
+ },
3113
+ {
3114
+ "epoch": 0.8,
3115
+ "learning_rate": 3.6764705882352945e-05,
3116
+ "loss": 0.3738,
3117
+ "step": 5000
3118
+ },
3119
+ {
3120
+ "epoch": 0.8,
3121
+ "eval_accuracy": 0.8466767371601208,
3122
+ "eval_f1": 0.8446831969931973,
3123
+ "eval_loss": 2.3107666969299316,
3124
+ "eval_precision": 0.8654388277401478,
3125
+ "eval_recall": 0.8466767371601208,
3126
+ "eval_runtime": 10.5417,
3127
+ "eval_samples_per_second": 251.194,
3128
+ "eval_steps_per_second": 3.984,
3129
+ "step": 5000
3130
+ },
3131
+ {
3132
+ "epoch": 0.8,
3133
+ "learning_rate": 3.6737967914438504e-05,
3134
+ "loss": 0.3446,
3135
+ "step": 5010
3136
+ },
3137
+ {
3138
+ "epoch": 0.8,
3139
+ "learning_rate": 3.671122994652407e-05,
3140
+ "loss": 0.4857,
3141
+ "step": 5020
3142
+ },
3143
+ {
3144
+ "epoch": 0.8,
3145
+ "learning_rate": 3.668449197860963e-05,
3146
+ "loss": 0.2666,
3147
+ "step": 5030
3148
+ },
3149
+ {
3150
+ "epoch": 0.81,
3151
+ "learning_rate": 3.665775401069519e-05,
3152
+ "loss": 0.4162,
3153
+ "step": 5040
3154
+ },
3155
+ {
3156
+ "epoch": 0.81,
3157
+ "learning_rate": 3.6631016042780753e-05,
3158
+ "loss": 0.1304,
3159
+ "step": 5050
3160
+ },
3161
+ {
3162
+ "epoch": 0.81,
3163
+ "learning_rate": 3.660427807486631e-05,
3164
+ "loss": 0.1575,
3165
+ "step": 5060
3166
+ },
3167
+ {
3168
+ "epoch": 0.81,
3169
+ "learning_rate": 3.657754010695187e-05,
3170
+ "loss": 0.1837,
3171
+ "step": 5070
3172
+ },
3173
+ {
3174
+ "epoch": 0.81,
3175
+ "learning_rate": 3.655080213903743e-05,
3176
+ "loss": 0.4065,
3177
+ "step": 5080
3178
+ },
3179
+ {
3180
+ "epoch": 0.81,
3181
+ "learning_rate": 3.6524064171122996e-05,
3182
+ "loss": 0.3935,
3183
+ "step": 5090
3184
+ },
3185
+ {
3186
+ "epoch": 0.82,
3187
+ "learning_rate": 3.649732620320856e-05,
3188
+ "loss": 0.1707,
3189
+ "step": 5100
3190
+ },
3191
+ {
3192
+ "epoch": 0.82,
3193
+ "learning_rate": 3.6470588235294114e-05,
3194
+ "loss": 0.0913,
3195
+ "step": 5110
3196
+ },
3197
+ {
3198
+ "epoch": 0.82,
3199
+ "learning_rate": 3.644385026737968e-05,
3200
+ "loss": 0.2487,
3201
+ "step": 5120
3202
+ },
3203
+ {
3204
+ "epoch": 0.82,
3205
+ "learning_rate": 3.6417112299465246e-05,
3206
+ "loss": 0.092,
3207
+ "step": 5130
3208
+ },
3209
+ {
3210
+ "epoch": 0.82,
3211
+ "learning_rate": 3.6390374331550805e-05,
3212
+ "loss": 0.2974,
3213
+ "step": 5140
3214
+ },
3215
+ {
3216
+ "epoch": 0.82,
3217
+ "learning_rate": 3.6363636363636364e-05,
3218
+ "loss": 0.2422,
3219
+ "step": 5150
3220
+ },
3221
+ {
3222
+ "epoch": 0.83,
3223
+ "learning_rate": 3.633689839572192e-05,
3224
+ "loss": 0.2698,
3225
+ "step": 5160
3226
+ },
3227
+ {
3228
+ "epoch": 0.83,
3229
+ "learning_rate": 3.631016042780749e-05,
3230
+ "loss": 0.2058,
3231
+ "step": 5170
3232
+ },
3233
+ {
3234
+ "epoch": 0.83,
3235
+ "learning_rate": 3.6283422459893055e-05,
3236
+ "loss": 0.3773,
3237
+ "step": 5180
3238
+ },
3239
+ {
3240
+ "epoch": 0.83,
3241
+ "learning_rate": 3.625668449197861e-05,
3242
+ "loss": 0.3958,
3243
+ "step": 5190
3244
+ },
3245
+ {
3246
+ "epoch": 0.83,
3247
+ "learning_rate": 3.622994652406417e-05,
3248
+ "loss": 0.2811,
3249
+ "step": 5200
3250
+ },
3251
+ {
3252
+ "epoch": 0.83,
3253
+ "learning_rate": 3.620320855614974e-05,
3254
+ "loss": 0.2158,
3255
+ "step": 5210
3256
+ },
3257
+ {
3258
+ "epoch": 0.84,
3259
+ "learning_rate": 3.61764705882353e-05,
3260
+ "loss": 0.2284,
3261
+ "step": 5220
3262
+ },
3263
+ {
3264
+ "epoch": 0.84,
3265
+ "learning_rate": 3.614973262032086e-05,
3266
+ "loss": 0.1471,
3267
+ "step": 5230
3268
+ },
3269
+ {
3270
+ "epoch": 0.84,
3271
+ "learning_rate": 3.6122994652406416e-05,
3272
+ "loss": 0.2005,
3273
+ "step": 5240
3274
+ },
3275
+ {
3276
+ "epoch": 0.84,
3277
+ "learning_rate": 3.609625668449198e-05,
3278
+ "loss": 0.1275,
3279
+ "step": 5250
3280
+ },
3281
+ {
3282
+ "epoch": 0.84,
3283
+ "learning_rate": 3.606951871657754e-05,
3284
+ "loss": 0.2067,
3285
+ "step": 5260
3286
+ },
3287
+ {
3288
+ "epoch": 0.84,
3289
+ "learning_rate": 3.60427807486631e-05,
3290
+ "loss": 0.5021,
3291
+ "step": 5270
3292
+ },
3293
+ {
3294
+ "epoch": 0.84,
3295
+ "learning_rate": 3.6016042780748665e-05,
3296
+ "loss": 0.4771,
3297
+ "step": 5280
3298
+ },
3299
+ {
3300
+ "epoch": 0.85,
3301
+ "learning_rate": 3.5989304812834225e-05,
3302
+ "loss": 0.4451,
3303
+ "step": 5290
3304
+ },
3305
+ {
3306
+ "epoch": 0.85,
3307
+ "learning_rate": 3.596256684491979e-05,
3308
+ "loss": 0.1325,
3309
+ "step": 5300
3310
+ },
3311
+ {
3312
+ "epoch": 0.85,
3313
+ "learning_rate": 3.593582887700535e-05,
3314
+ "loss": 0.2466,
3315
+ "step": 5310
3316
+ },
3317
+ {
3318
+ "epoch": 0.85,
3319
+ "learning_rate": 3.590909090909091e-05,
3320
+ "loss": 0.3195,
3321
+ "step": 5320
3322
+ },
3323
+ {
3324
+ "epoch": 0.85,
3325
+ "learning_rate": 3.5882352941176474e-05,
3326
+ "loss": 0.2085,
3327
+ "step": 5330
3328
+ },
3329
+ {
3330
+ "epoch": 0.85,
3331
+ "learning_rate": 3.585561497326203e-05,
3332
+ "loss": 0.1185,
3333
+ "step": 5340
3334
+ },
3335
+ {
3336
+ "epoch": 0.86,
3337
+ "learning_rate": 3.582887700534759e-05,
3338
+ "loss": 0.5152,
3339
+ "step": 5350
3340
+ },
3341
+ {
3342
+ "epoch": 0.86,
3343
+ "learning_rate": 3.580213903743316e-05,
3344
+ "loss": 0.186,
3345
+ "step": 5360
3346
+ },
3347
+ {
3348
+ "epoch": 0.86,
3349
+ "learning_rate": 3.577540106951872e-05,
3350
+ "loss": 0.2386,
3351
+ "step": 5370
3352
+ },
3353
+ {
3354
+ "epoch": 0.86,
3355
+ "learning_rate": 3.5748663101604276e-05,
3356
+ "loss": 0.4468,
3357
+ "step": 5380
3358
+ },
3359
+ {
3360
+ "epoch": 0.86,
3361
+ "learning_rate": 3.572192513368984e-05,
3362
+ "loss": 0.1996,
3363
+ "step": 5390
3364
+ },
3365
+ {
3366
+ "epoch": 0.86,
3367
+ "learning_rate": 3.56951871657754e-05,
3368
+ "loss": 0.1182,
3369
+ "step": 5400
3370
+ },
3371
+ {
3372
+ "epoch": 0.87,
3373
+ "learning_rate": 3.566844919786097e-05,
3374
+ "loss": 0.157,
3375
+ "step": 5410
3376
+ },
3377
+ {
3378
+ "epoch": 0.87,
3379
+ "learning_rate": 3.5641711229946526e-05,
3380
+ "loss": 0.1696,
3381
+ "step": 5420
3382
+ },
3383
+ {
3384
+ "epoch": 0.87,
3385
+ "learning_rate": 3.5614973262032085e-05,
3386
+ "loss": 0.2641,
3387
+ "step": 5430
3388
+ },
3389
+ {
3390
+ "epoch": 0.87,
3391
+ "learning_rate": 3.558823529411765e-05,
3392
+ "loss": 0.2224,
3393
+ "step": 5440
3394
+ },
3395
+ {
3396
+ "epoch": 0.87,
3397
+ "learning_rate": 3.556149732620321e-05,
3398
+ "loss": 0.2676,
3399
+ "step": 5450
3400
+ },
3401
+ {
3402
+ "epoch": 0.87,
3403
+ "learning_rate": 3.553475935828877e-05,
3404
+ "loss": 0.213,
3405
+ "step": 5460
3406
+ },
3407
+ {
3408
+ "epoch": 0.88,
3409
+ "learning_rate": 3.5508021390374335e-05,
3410
+ "loss": 0.3497,
3411
+ "step": 5470
3412
+ },
3413
+ {
3414
+ "epoch": 0.88,
3415
+ "learning_rate": 3.5481283422459894e-05,
3416
+ "loss": 0.3077,
3417
+ "step": 5480
3418
+ },
3419
+ {
3420
+ "epoch": 0.88,
3421
+ "learning_rate": 3.545454545454546e-05,
3422
+ "loss": 0.4294,
3423
+ "step": 5490
3424
+ },
3425
+ {
3426
+ "epoch": 0.88,
3427
+ "learning_rate": 3.542780748663101e-05,
3428
+ "loss": 0.1212,
3429
+ "step": 5500
3430
+ },
3431
+ {
3432
+ "epoch": 0.88,
3433
+ "eval_accuracy": 0.8444108761329305,
3434
+ "eval_f1": 0.8425221036853814,
3435
+ "eval_loss": 2.348144054412842,
3436
+ "eval_precision": 0.8617668418744113,
3437
+ "eval_recall": 0.8444108761329305,
3438
+ "eval_runtime": 10.5364,
3439
+ "eval_samples_per_second": 251.32,
3440
+ "eval_steps_per_second": 3.986,
3441
+ "step": 5500
3442
+ },
3443
+ {
3444
+ "epoch": 0.88,
3445
+ "learning_rate": 3.540106951871658e-05,
3446
+ "loss": 0.2518,
3447
+ "step": 5510
3448
+ },
3449
+ {
3450
+ "epoch": 0.88,
3451
+ "learning_rate": 3.537433155080214e-05,
3452
+ "loss": 0.2727,
3453
+ "step": 5520
3454
+ },
3455
+ {
3456
+ "epoch": 0.88,
3457
+ "learning_rate": 3.53475935828877e-05,
3458
+ "loss": 0.2366,
3459
+ "step": 5530
3460
+ },
3461
+ {
3462
+ "epoch": 0.89,
3463
+ "learning_rate": 3.532085561497326e-05,
3464
+ "loss": 0.3752,
3465
+ "step": 5540
3466
+ },
3467
+ {
3468
+ "epoch": 0.89,
3469
+ "learning_rate": 3.529411764705883e-05,
3470
+ "loss": 0.3085,
3471
+ "step": 5550
3472
+ },
3473
+ {
3474
+ "epoch": 0.89,
3475
+ "learning_rate": 3.5267379679144386e-05,
3476
+ "loss": 0.0496,
3477
+ "step": 5560
3478
+ },
3479
+ {
3480
+ "epoch": 0.89,
3481
+ "learning_rate": 3.524064171122995e-05,
3482
+ "loss": 0.3345,
3483
+ "step": 5570
3484
+ },
3485
+ {
3486
+ "epoch": 0.89,
3487
+ "learning_rate": 3.5213903743315504e-05,
3488
+ "loss": 0.393,
3489
+ "step": 5580
3490
+ },
3491
+ {
3492
+ "epoch": 0.89,
3493
+ "learning_rate": 3.518716577540107e-05,
3494
+ "loss": 0.2174,
3495
+ "step": 5590
3496
+ },
3497
+ {
3498
+ "epoch": 0.9,
3499
+ "learning_rate": 3.5160427807486636e-05,
3500
+ "loss": 0.2527,
3501
+ "step": 5600
3502
+ },
3503
+ {
3504
+ "epoch": 0.9,
3505
+ "learning_rate": 3.5133689839572195e-05,
3506
+ "loss": 0.1707,
3507
+ "step": 5610
3508
+ },
3509
+ {
3510
+ "epoch": 0.9,
3511
+ "learning_rate": 3.5106951871657754e-05,
3512
+ "loss": 0.227,
3513
+ "step": 5620
3514
+ },
3515
+ {
3516
+ "epoch": 0.9,
3517
+ "learning_rate": 3.508021390374332e-05,
3518
+ "loss": 0.2253,
3519
+ "step": 5630
3520
+ },
3521
+ {
3522
+ "epoch": 0.9,
3523
+ "learning_rate": 3.505347593582888e-05,
3524
+ "loss": 0.2014,
3525
+ "step": 5640
3526
+ },
3527
+ {
3528
+ "epoch": 0.9,
3529
+ "learning_rate": 3.5026737967914445e-05,
3530
+ "loss": 0.2545,
3531
+ "step": 5650
3532
+ },
3533
+ {
3534
+ "epoch": 0.91,
3535
+ "learning_rate": 3.5e-05,
3536
+ "loss": 0.2213,
3537
+ "step": 5660
3538
+ },
3539
+ {
3540
+ "epoch": 0.91,
3541
+ "learning_rate": 3.497326203208556e-05,
3542
+ "loss": 0.172,
3543
+ "step": 5670
3544
+ },
3545
+ {
3546
+ "epoch": 0.91,
3547
+ "learning_rate": 3.494652406417113e-05,
3548
+ "loss": 0.3737,
3549
+ "step": 5680
3550
+ },
3551
+ {
3552
+ "epoch": 0.91,
3553
+ "learning_rate": 3.491978609625669e-05,
3554
+ "loss": 0.1347,
3555
+ "step": 5690
3556
+ },
3557
+ {
3558
+ "epoch": 0.91,
3559
+ "learning_rate": 3.489304812834225e-05,
3560
+ "loss": 0.2867,
3561
+ "step": 5700
3562
+ },
3563
+ {
3564
+ "epoch": 0.91,
3565
+ "learning_rate": 3.4866310160427806e-05,
3566
+ "loss": 0.2276,
3567
+ "step": 5710
3568
+ },
3569
+ {
3570
+ "epoch": 0.92,
3571
+ "learning_rate": 3.483957219251337e-05,
3572
+ "loss": 0.3448,
3573
+ "step": 5720
3574
+ },
3575
+ {
3576
+ "epoch": 0.92,
3577
+ "learning_rate": 3.481283422459893e-05,
3578
+ "loss": 0.0974,
3579
+ "step": 5730
3580
+ },
3581
+ {
3582
+ "epoch": 0.92,
3583
+ "learning_rate": 3.478609625668449e-05,
3584
+ "loss": 0.1956,
3585
+ "step": 5740
3586
+ },
3587
+ {
3588
+ "epoch": 0.92,
3589
+ "learning_rate": 3.4759358288770055e-05,
3590
+ "loss": 0.1278,
3591
+ "step": 5750
3592
+ },
3593
+ {
3594
+ "epoch": 0.92,
3595
+ "learning_rate": 3.473262032085562e-05,
3596
+ "loss": 0.2225,
3597
+ "step": 5760
3598
+ },
3599
+ {
3600
+ "epoch": 0.92,
3601
+ "learning_rate": 3.470588235294118e-05,
3602
+ "loss": 0.3499,
3603
+ "step": 5770
3604
+ },
3605
+ {
3606
+ "epoch": 0.92,
3607
+ "learning_rate": 3.467914438502674e-05,
3608
+ "loss": 0.2516,
3609
+ "step": 5780
3610
+ },
3611
+ {
3612
+ "epoch": 0.93,
3613
+ "learning_rate": 3.46524064171123e-05,
3614
+ "loss": 0.158,
3615
+ "step": 5790
3616
+ },
3617
+ {
3618
+ "epoch": 0.93,
3619
+ "learning_rate": 3.4625668449197864e-05,
3620
+ "loss": 0.2245,
3621
+ "step": 5800
3622
+ },
3623
+ {
3624
+ "epoch": 0.93,
3625
+ "learning_rate": 3.459893048128342e-05,
3626
+ "loss": 0.3029,
3627
+ "step": 5810
3628
+ },
3629
+ {
3630
+ "epoch": 0.93,
3631
+ "learning_rate": 3.457219251336898e-05,
3632
+ "loss": 0.3581,
3633
+ "step": 5820
3634
+ },
3635
+ {
3636
+ "epoch": 0.93,
3637
+ "learning_rate": 3.454545454545455e-05,
3638
+ "loss": 0.2423,
3639
+ "step": 5830
3640
+ },
3641
+ {
3642
+ "epoch": 0.93,
3643
+ "learning_rate": 3.4518716577540114e-05,
3644
+ "loss": 0.3618,
3645
+ "step": 5840
3646
+ },
3647
+ {
3648
+ "epoch": 0.94,
3649
+ "learning_rate": 3.4491978609625666e-05,
3650
+ "loss": 0.1969,
3651
+ "step": 5850
3652
+ },
3653
+ {
3654
+ "epoch": 0.94,
3655
+ "learning_rate": 3.446524064171123e-05,
3656
+ "loss": 0.282,
3657
+ "step": 5860
3658
+ },
3659
+ {
3660
+ "epoch": 0.94,
3661
+ "learning_rate": 3.443850267379679e-05,
3662
+ "loss": 0.2575,
3663
+ "step": 5870
3664
+ },
3665
+ {
3666
+ "epoch": 0.94,
3667
+ "learning_rate": 3.441176470588236e-05,
3668
+ "loss": 0.3495,
3669
+ "step": 5880
3670
+ },
3671
+ {
3672
+ "epoch": 0.94,
3673
+ "learning_rate": 3.4385026737967916e-05,
3674
+ "loss": 2.8462,
3675
+ "step": 5890
3676
+ },
3677
+ {
3678
+ "epoch": 0.94,
3679
+ "learning_rate": 3.4358288770053475e-05,
3680
+ "loss": 3.0621,
3681
+ "step": 5900
3682
+ },
3683
+ {
3684
+ "epoch": 0.95,
3685
+ "learning_rate": 3.433155080213904e-05,
3686
+ "loss": 1.8662,
3687
+ "step": 5910
3688
+ },
3689
+ {
3690
+ "epoch": 0.95,
3691
+ "learning_rate": 3.4304812834224606e-05,
3692
+ "loss": 0.7081,
3693
+ "step": 5920
3694
+ },
3695
+ {
3696
+ "epoch": 0.95,
3697
+ "learning_rate": 3.427807486631016e-05,
3698
+ "loss": 0.7017,
3699
+ "step": 5930
3700
+ },
3701
+ {
3702
+ "epoch": 0.95,
3703
+ "learning_rate": 3.4251336898395724e-05,
3704
+ "loss": 0.6809,
3705
+ "step": 5940
3706
+ },
3707
+ {
3708
+ "epoch": 0.95,
3709
+ "learning_rate": 3.4224598930481284e-05,
3710
+ "loss": 0.5177,
3711
+ "step": 5950
3712
+ },
3713
+ {
3714
+ "epoch": 0.95,
3715
+ "learning_rate": 3.419786096256685e-05,
3716
+ "loss": 0.3095,
3717
+ "step": 5960
3718
+ },
3719
+ {
3720
+ "epoch": 0.96,
3721
+ "learning_rate": 3.417112299465241e-05,
3722
+ "loss": 0.1414,
3723
+ "step": 5970
3724
+ },
3725
+ {
3726
+ "epoch": 0.96,
3727
+ "learning_rate": 3.414438502673797e-05,
3728
+ "loss": 0.0596,
3729
+ "step": 5980
3730
+ },
3731
+ {
3732
+ "epoch": 0.96,
3733
+ "learning_rate": 3.411764705882353e-05,
3734
+ "loss": 0.2019,
3735
+ "step": 5990
3736
+ },
3737
+ {
3738
+ "epoch": 0.96,
3739
+ "learning_rate": 3.409090909090909e-05,
3740
+ "loss": 0.308,
3741
+ "step": 6000
3742
+ },
3743
+ {
3744
+ "epoch": 0.96,
3745
+ "eval_accuracy": 0.8444108761329305,
3746
+ "eval_f1": 0.8423328431236531,
3747
+ "eval_loss": 1.9854190349578857,
3748
+ "eval_precision": 0.8635785518138459,
3749
+ "eval_recall": 0.8444108761329305,
3750
+ "eval_runtime": 10.5454,
3751
+ "eval_samples_per_second": 251.106,
3752
+ "eval_steps_per_second": 3.983,
3753
+ "step": 6000
3754
+ }
3755
+ ],
3756
+ "logging_steps": 10,
3757
+ "max_steps": 18750,
3758
+ "num_input_tokens_seen": 0,
3759
+ "num_train_epochs": 3,
3760
+ "save_steps": 500,
3761
+ "total_flos": 5.0517776203776e+16,
3762
+ "train_batch_size": 16,
3763
+ "trial_name": null,
3764
+ "trial_params": null
3765
+ }
checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ea61cfa8a59fe23f079dded341db3c34a77c19986a2b3885296af3c7226c95
3
+ size 4664
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base-finetuned-8M-MLM-500K/",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
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": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "LABEL_0": 0,
22
+ "LABEL_1": 1,
23
+ "LABEL_2": 2
24
+ },
25
+ "layer_norm_eps": 1e-05,
26
+ "max_position_embeddings": 514,
27
+ "model_type": "roberta",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 1,
31
+ "position_embedding_type": "absolute",
32
+ "problem_type": "single_label_classification",
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.37.0",
35
+ "type_vocab_size": 1,
36
+ "use_cache": true,
37
+ "vocab_size": 50265
38
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c2ce8036ff5e291017bf4b5df5619f3af6d8cf9b48a4fdee8eb15c79a54010f
3
+ size 498615900
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ea61cfa8a59fe23f079dded341db3c34a77c19986a2b3885296af3c7226c95
3
+ size 4664