Hugging Face's logo

shaheen6
/
PixelTranslationCurriculumCheckpointsPublic

Model card Files
xet
 Community
shaheen6 commited on
Commit
151b60c
·
verified ·
1 Parent(s): acc06cd

Add files using upload-large-folder tool

Browse files
Files changed (50) hide show
  1. contextual/pct100/downstream.complete +0 -0
  2. contextual/pct100/downstream_2x.complete +0 -0
  3. contextual/pct100/downstream_4x.complete +0 -0
  4. contextual/pct100/downstream_da.complete +0 -0
  5. contextual/pct100/finetuning/pixel/Dialect_Combined/checkpoint-1000/trainer_state.json +616 -0
  6. contextual/pct100/finetuning/pixel/Dialect_Combined/checkpoint-2000/trainer_state.json +1216 -0
  7. contextual/pct100/finetuning/pixel/Dialect_Combined/training_complete.json +19 -0
  8. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-3000/trainer_state.json +1816 -0
  9. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-4000/trainer_state.json +2416 -0
  10. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_1x/training_complete.json +19 -0
  11. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_2x/checkpoint-5000/trainer_state.json +3016 -0
  12. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_2x/checkpoint-6000/trainer_state.json +3616 -0
  13. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_2x/training_complete.json +19 -0
  14. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_4x/checkpoint-10000/trainer_state.json +0 -0
  15. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_4x/checkpoint-11000/trainer_state.json +0 -0
  16. contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_4x/training_complete.json +19 -0
  17. contextual/pct100/manifest.json +38 -0
  18. contextual/pct100/msa_training.complete +0 -0
  19. contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/rng_state_1.pth +3 -0
  20. contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/rng_state_2.pth +3 -0
  21. contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/rng_state_3.pth +3 -0
  22. contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/trainer_state.json +0 -0
  23. contextual/pct100/pretraining/pixel/alignment/checkpoint-46000/trainer_state.json +0 -0
  24. contextual/pct100/pretraining/pixel/alignment/final/training_complete.json +21 -0
  25. contextual/pct100/pretraining/pixel/full_epoch1/checkpoint-44000/trainer_state.json +0 -0
  26. contextual/pct100/pretraining/pixel/full_epoch1/checkpoint-46000/trainer_state.json +0 -0
  27. contextual/pct100/pretraining/pixel/full_epoch1/final/training_complete.json +21 -0
  28. contextual/pct100/pretraining/pixel/full_epoch2/checkpoint-44000/trainer_state.json +0 -0
  29. contextual/pct100/pretraining/pixel/full_epoch2/checkpoint-46000/trainer_state.json +0 -0
  30. contextual/pct100/pretraining/pixel/full_epoch2/final/training_complete.json +21 -0
  31. contextual/pct100/renderer_audit.json +38 -0
  32. contextual/pct50/experiment.complete +0 -0
  33. contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/best_model.bin +3 -0
  34. contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-3000/pytorch_model.bin +3 -0
  35. contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-3000/trainer_state.json +1816 -0
  36. contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-4000/pytorch_model.bin +3 -0
  37. contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-4000/trainer_state.json +2416 -0
  38. contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/training_complete.json +19 -0
  39. contextual/pct50/manifest.json +41 -0
  40. contextual/pct50/msa_training.complete +0 -0
  41. contextual/pct50/pretraining/pixel/alignment/checkpoint-20000/trainer_state.json +0 -0
  42. contextual/pct50/pretraining/pixel/alignment/checkpoint-22000/trainer_state.json +0 -0
  43. contextual/pct50/pretraining/pixel/alignment/final/training_complete.json +21 -0
  44. contextual/pct50/pretraining/pixel/full_epoch1/checkpoint-20000/trainer_state.json +0 -0
  45. contextual/pct50/pretraining/pixel/full_epoch1/checkpoint-22000/trainer_state.json +0 -0
  46. contextual/pct50/pretraining/pixel/full_epoch1/final/training_complete.json +21 -0
  47. contextual/pct50/pretraining/pixel/full_epoch2/checkpoint-20000/trainer_state.json +0 -0
  48. contextual/pct50/pretraining/pixel/full_epoch2/checkpoint-22000/trainer_state.json +0 -0
  49. contextual/pct50/pretraining/pixel/full_epoch2/final/training_complete.json +21 -0
  50. contextual/pct50/renderer_audit.json +38 -0
contextual/pct100/downstream.complete ADDED
File without changes
contextual/pct100/downstream_2x.complete ADDED
File without changes
contextual/pct100/downstream_4x.complete ADDED
File without changes
contextual/pct100/downstream_da.complete ADDED
File without changes
contextual/pct100/finetuning/pixel/Dialect_Combined/checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,616 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.43802014892685065,
5
+ "global_step": 1000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 1.3100436681222706e-06,
13
+ "loss": 2.5451,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 2.6200873362445413e-06,
19
+ "loss": 2.4202,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 3.930131004366812e-06,
25
+ "loss": 2.459,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 5.240174672489083e-06,
31
+ "loss": 2.4559,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 6.550218340611354e-06,
37
+ "loss": 2.423,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 7.860262008733624e-06,
43
+ "loss": 2.3773,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 9.170305676855895e-06,
49
+ "loss": 2.2703,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.04,
54
+ "learning_rate": 1.0480349344978165e-05,
55
+ "loss": 2.3069,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.04,
60
+ "learning_rate": 1.1790393013100436e-05,
61
+ "loss": 2.138,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 1.3100436681222708e-05,
67
+ "loss": 2.1507,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.05,
72
+ "learning_rate": 1.4410480349344979e-05,
73
+ "loss": 2.0823,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.05,
78
+ "learning_rate": 1.5720524017467248e-05,
79
+ "loss": 2.0432,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.06,
84
+ "learning_rate": 1.703056768558952e-05,
85
+ "loss": 2.0268,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.06,
90
+ "learning_rate": 1.834061135371179e-05,
91
+ "loss": 2.0528,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.07,
96
+ "learning_rate": 1.965065502183406e-05,
97
+ "loss": 2.0324,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.07,
102
+ "learning_rate": 2.096069868995633e-05,
103
+ "loss": 1.9949,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.07,
108
+ "learning_rate": 2.2270742358078603e-05,
109
+ "loss": 1.9536,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.08,
114
+ "learning_rate": 2.3580786026200872e-05,
115
+ "loss": 1.9369,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.08,
120
+ "learning_rate": 2.4890829694323147e-05,
121
+ "loss": 1.9243,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.09,
126
+ "learning_rate": 2.6200873362445416e-05,
127
+ "loss": 1.9412,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.09,
132
+ "learning_rate": 2.751091703056769e-05,
133
+ "loss": 1.8752,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.1,
138
+ "learning_rate": 2.8820960698689958e-05,
139
+ "loss": 1.9139,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.1,
144
+ "learning_rate": 2.999998245473042e-05,
145
+ "loss": 1.8984,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.11,
150
+ "learning_rate": 2.9997877072044015e-05,
151
+ "loss": 1.8423,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.11,
156
+ "learning_rate": 2.9992263199785727e-05,
157
+ "loss": 1.8393,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.11,
162
+ "learning_rate": 2.998314215122224e-05,
163
+ "loss": 1.7595,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.12,
168
+ "learning_rate": 2.9970516060062453e-05,
169
+ "loss": 1.8361,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.12,
174
+ "learning_rate": 2.9954387879958338e-05,
175
+ "loss": 1.7841,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.13,
180
+ "learning_rate": 2.9934761383813976e-05,
181
+ "loss": 1.7991,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.13,
186
+ "learning_rate": 2.9911641162902978e-05,
187
+ "loss": 1.7774,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.14,
192
+ "learning_rate": 2.98850326257944e-05,
193
+ "loss": 1.8,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.14,
198
+ "learning_rate": 2.985494199708753e-05,
199
+ "loss": 1.787,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.14,
204
+ "learning_rate": 2.982137631595575e-05,
205
+ "loss": 1.7824,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.15,
210
+ "learning_rate": 2.9784343434499845e-05,
211
+ "loss": 1.7173,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.15,
216
+ "learning_rate": 2.9743852015911114e-05,
217
+ "loss": 1.7807,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.16,
222
+ "learning_rate": 2.9699911532444826e-05,
223
+ "loss": 1.7617,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.16,
228
+ "learning_rate": 2.9652532263204296e-05,
229
+ "loss": 1.7639,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.17,
234
+ "learning_rate": 2.9601725291736297e-05,
235
+ "loss": 1.7164,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.17,
240
+ "learning_rate": 2.954750250343826e-05,
241
+ "loss": 1.7242,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.18,
246
+ "learning_rate": 2.9489876582777887e-05,
247
+ "loss": 1.6936,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.18,
252
+ "learning_rate": 2.942886101032584e-05,
253
+ "loss": 1.7375,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.18,
258
+ "learning_rate": 2.9364470059602206e-05,
259
+ "loss": 1.6889,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.19,
264
+ "learning_rate": 2.929671879373746e-05,
265
+ "loss": 1.6968,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.19,
270
+ "learning_rate": 2.92256230619487e-05,
271
+ "loss": 1.6817,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.2,
276
+ "learning_rate": 2.9151199495832017e-05,
277
+ "loss": 1.6604,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.2,
282
+ "learning_rate": 2.9073465505471807e-05,
283
+ "loss": 1.6688,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.21,
288
+ "learning_rate": 2.8992439275367996e-05,
289
+ "loss": 1.6587,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.21,
294
+ "learning_rate": 2.89081397601821e-05,
295
+ "loss": 1.6928,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.21,
300
+ "learning_rate": 2.8820586680303122e-05,
301
+ "loss": 1.7002,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.22,
306
+ "learning_rate": 2.872980051723428e-05,
307
+ "loss": 1.6813,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.22,
312
+ "learning_rate": 2.8635802508801775e-05,
313
+ "loss": 1.6792,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.23,
318
+ "learning_rate": 2.853861464418652e-05,
319
+ "loss": 1.6166,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.23,
324
+ "learning_rate": 2.8438259658780186e-05,
325
+ "loss": 1.6664,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.24,
330
+ "learning_rate": 2.833476102886666e-05,
331
+ "loss": 1.6942,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.24,
336
+ "learning_rate": 2.822814296613017e-05,
337
+ "loss": 1.693,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.25,
342
+ "learning_rate": 2.811843041199139e-05,
343
+ "loss": 1.6742,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.25,
348
+ "learning_rate": 2.800564903177284e-05,
349
+ "loss": 1.6637,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.25,
354
+ "learning_rate": 2.7889825208694915e-05,
355
+ "loss": 1.6899,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.26,
360
+ "learning_rate": 2.7770986037704026e-05,
361
+ "loss": 1.659,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.26,
366
+ "learning_rate": 2.7649159319134192e-05,
367
+ "loss": 1.6435,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.27,
372
+ "learning_rate": 2.7524373552203645e-05,
373
+ "loss": 1.6964,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.27,
378
+ "learning_rate": 2.7396657928347924e-05,
379
+ "loss": 1.6864,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.28,
384
+ "learning_rate": 2.72660423243911e-05,
385
+ "loss": 1.6273,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.28,
390
+ "learning_rate": 2.7132557295556566e-05,
391
+ "loss": 1.6173,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.28,
396
+ "learning_rate": 2.6996234068319198e-05,
397
+ "loss": 1.6577,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.29,
402
+ "learning_rate": 2.6857104533100482e-05,
403
+ "loss": 1.6459,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.29,
408
+ "learning_rate": 2.67152012368083e-05,
409
+ "loss": 1.6777,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.3,
414
+ "learning_rate": 2.657055737522315e-05,
415
+ "loss": 1.606,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.3,
420
+ "learning_rate": 2.642320678523258e-05,
421
+ "loss": 1.6268,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.31,
426
+ "learning_rate": 2.6273183936915657e-05,
427
+ "loss": 1.6301,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.31,
432
+ "learning_rate": 2.6120523925479275e-05,
433
+ "loss": 1.6264,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.32,
438
+ "learning_rate": 2.596526246304826e-05,
439
+ "loss": 1.6058,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.32,
444
+ "learning_rate": 2.5807435870311153e-05,
445
+ "loss": 1.5989,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.32,
450
+ "learning_rate": 2.5647081068023612e-05,
451
+ "loss": 1.6449,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.33,
456
+ "learning_rate": 2.5484235568371455e-05,
457
+ "loss": 1.6501,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.33,
462
+ "learning_rate": 2.5318937466195347e-05,
463
+ "loss": 1.6576,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.34,
468
+ "learning_rate": 2.5151225430079186e-05,
469
+ "loss": 1.6367,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.34,
474
+ "learning_rate": 2.498113869330424e-05,
475
+ "loss": 1.5932,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.35,
480
+ "learning_rate": 2.4808717044671252e-05,
481
+ "loss": 1.6535,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.35,
486
+ "learning_rate": 2.4634000819192524e-05,
487
+ "loss": 1.6104,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.35,
492
+ "learning_rate": 2.4457030888656257e-05,
493
+ "loss": 1.5829,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.36,
498
+ "learning_rate": 2.4277848652065314e-05,
499
+ "loss": 1.5652,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.36,
504
+ "learning_rate": 2.4096496025952643e-05,
505
+ "loss": 1.5931,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.37,
510
+ "learning_rate": 2.3913015434575658e-05,
511
+ "loss": 1.577,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.37,
516
+ "learning_rate": 2.3727449799991824e-05,
517
+ "loss": 1.5854,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.38,
522
+ "learning_rate": 2.3539842532017792e-05,
523
+ "loss": 1.5575,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.38,
528
+ "learning_rate": 2.3350237518074434e-05,
529
+ "loss": 1.6006,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.39,
534
+ "learning_rate": 2.315867911292016e-05,
535
+ "loss": 1.5524,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.39,
540
+ "learning_rate": 2.2965212128274908e-05,
541
+ "loss": 1.6059,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.39,
546
+ "learning_rate": 2.2769881822337202e-05,
547
+ "loss": 1.5913,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.4,
552
+ "learning_rate": 2.2572733889196826e-05,
553
+ "loss": 1.6139,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.4,
558
+ "learning_rate": 2.237381444814548e-05,
559
+ "loss": 1.5775,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.41,
564
+ "learning_rate": 2.2173170032887983e-05,
565
+ "loss": 1.5449,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.41,
570
+ "learning_rate": 2.197084758065653e-05,
571
+ "loss": 1.5664,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.42,
576
+ "learning_rate": 2.1766894421230555e-05,
577
+ "loss": 1.5408,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.42,
582
+ "learning_rate": 2.1561358265864765e-05,
583
+ "loss": 1.5566,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.42,
588
+ "learning_rate": 2.1354287196127925e-05,
589
+ "loss": 1.6271,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.43,
594
+ "learning_rate": 2.1145729652655013e-05,
595
+ "loss": 1.6316,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.43,
600
+ "learning_rate": 2.0935734423815405e-05,
601
+ "loss": 1.5676,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.44,
606
+ "learning_rate": 2.0724350634299685e-05,
607
+ "loss": 1.5719,
608
+ "step": 1000
609
+ }
610
+ ],
611
+ "max_steps": 2283,
612
+ "num_train_epochs": 1,
613
+ "total_flos": 0.0,
614
+ "trial_name": null,
615
+ "trial_params": null
616
+ }
contextual/pct100/finetuning/pixel/Dialect_Combined/checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,1216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.8760402978537013,
5
+ "global_step": 2000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 1.3100436681222706e-06,
13
+ "loss": 2.5451,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 2.6200873362445413e-06,
19
+ "loss": 2.4202,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 3.930131004366812e-06,
25
+ "loss": 2.459,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 5.240174672489083e-06,
31
+ "loss": 2.4559,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 6.550218340611354e-06,
37
+ "loss": 2.423,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 7.860262008733624e-06,
43
+ "loss": 2.3773,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 9.170305676855895e-06,
49
+ "loss": 2.2703,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.04,
54
+ "learning_rate": 1.0480349344978165e-05,
55
+ "loss": 2.3069,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.04,
60
+ "learning_rate": 1.1790393013100436e-05,
61
+ "loss": 2.138,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 1.3100436681222708e-05,
67
+ "loss": 2.1507,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.05,
72
+ "learning_rate": 1.4410480349344979e-05,
73
+ "loss": 2.0823,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.05,
78
+ "learning_rate": 1.5720524017467248e-05,
79
+ "loss": 2.0432,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.06,
84
+ "learning_rate": 1.703056768558952e-05,
85
+ "loss": 2.0268,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.06,
90
+ "learning_rate": 1.834061135371179e-05,
91
+ "loss": 2.0528,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.07,
96
+ "learning_rate": 1.965065502183406e-05,
97
+ "loss": 2.0324,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.07,
102
+ "learning_rate": 2.096069868995633e-05,
103
+ "loss": 1.9949,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.07,
108
+ "learning_rate": 2.2270742358078603e-05,
109
+ "loss": 1.9536,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.08,
114
+ "learning_rate": 2.3580786026200872e-05,
115
+ "loss": 1.9369,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.08,
120
+ "learning_rate": 2.4890829694323147e-05,
121
+ "loss": 1.9243,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.09,
126
+ "learning_rate": 2.6200873362445416e-05,
127
+ "loss": 1.9412,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.09,
132
+ "learning_rate": 2.751091703056769e-05,
133
+ "loss": 1.8752,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.1,
138
+ "learning_rate": 2.8820960698689958e-05,
139
+ "loss": 1.9139,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.1,
144
+ "learning_rate": 2.999998245473042e-05,
145
+ "loss": 1.8984,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.11,
150
+ "learning_rate": 2.9997877072044015e-05,
151
+ "loss": 1.8423,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.11,
156
+ "learning_rate": 2.9992263199785727e-05,
157
+ "loss": 1.8393,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.11,
162
+ "learning_rate": 2.998314215122224e-05,
163
+ "loss": 1.7595,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.12,
168
+ "learning_rate": 2.9970516060062453e-05,
169
+ "loss": 1.8361,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.12,
174
+ "learning_rate": 2.9954387879958338e-05,
175
+ "loss": 1.7841,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.13,
180
+ "learning_rate": 2.9934761383813976e-05,
181
+ "loss": 1.7991,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.13,
186
+ "learning_rate": 2.9911641162902978e-05,
187
+ "loss": 1.7774,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.14,
192
+ "learning_rate": 2.98850326257944e-05,
193
+ "loss": 1.8,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.14,
198
+ "learning_rate": 2.985494199708753e-05,
199
+ "loss": 1.787,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.14,
204
+ "learning_rate": 2.982137631595575e-05,
205
+ "loss": 1.7824,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.15,
210
+ "learning_rate": 2.9784343434499845e-05,
211
+ "loss": 1.7173,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.15,
216
+ "learning_rate": 2.9743852015911114e-05,
217
+ "loss": 1.7807,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.16,
222
+ "learning_rate": 2.9699911532444826e-05,
223
+ "loss": 1.7617,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.16,
228
+ "learning_rate": 2.9652532263204296e-05,
229
+ "loss": 1.7639,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.17,
234
+ "learning_rate": 2.9601725291736297e-05,
235
+ "loss": 1.7164,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.17,
240
+ "learning_rate": 2.954750250343826e-05,
241
+ "loss": 1.7242,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.18,
246
+ "learning_rate": 2.9489876582777887e-05,
247
+ "loss": 1.6936,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.18,
252
+ "learning_rate": 2.942886101032584e-05,
253
+ "loss": 1.7375,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.18,
258
+ "learning_rate": 2.9364470059602206e-05,
259
+ "loss": 1.6889,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.19,
264
+ "learning_rate": 2.929671879373746e-05,
265
+ "loss": 1.6968,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.19,
270
+ "learning_rate": 2.92256230619487e-05,
271
+ "loss": 1.6817,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.2,
276
+ "learning_rate": 2.9151199495832017e-05,
277
+ "loss": 1.6604,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.2,
282
+ "learning_rate": 2.9073465505471807e-05,
283
+ "loss": 1.6688,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.21,
288
+ "learning_rate": 2.8992439275367996e-05,
289
+ "loss": 1.6587,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.21,
294
+ "learning_rate": 2.89081397601821e-05,
295
+ "loss": 1.6928,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.21,
300
+ "learning_rate": 2.8820586680303122e-05,
301
+ "loss": 1.7002,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.22,
306
+ "learning_rate": 2.872980051723428e-05,
307
+ "loss": 1.6813,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.22,
312
+ "learning_rate": 2.8635802508801775e-05,
313
+ "loss": 1.6792,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.23,
318
+ "learning_rate": 2.853861464418652e-05,
319
+ "loss": 1.6166,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.23,
324
+ "learning_rate": 2.8438259658780186e-05,
325
+ "loss": 1.6664,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.24,
330
+ "learning_rate": 2.833476102886666e-05,
331
+ "loss": 1.6942,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.24,
336
+ "learning_rate": 2.822814296613017e-05,
337
+ "loss": 1.693,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.25,
342
+ "learning_rate": 2.811843041199139e-05,
343
+ "loss": 1.6742,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.25,
348
+ "learning_rate": 2.800564903177284e-05,
349
+ "loss": 1.6637,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.25,
354
+ "learning_rate": 2.7889825208694915e-05,
355
+ "loss": 1.6899,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.26,
360
+ "learning_rate": 2.7770986037704026e-05,
361
+ "loss": 1.659,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.26,
366
+ "learning_rate": 2.7649159319134192e-05,
367
+ "loss": 1.6435,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.27,
372
+ "learning_rate": 2.7524373552203645e-05,
373
+ "loss": 1.6964,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.27,
378
+ "learning_rate": 2.7396657928347924e-05,
379
+ "loss": 1.6864,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.28,
384
+ "learning_rate": 2.72660423243911e-05,
385
+ "loss": 1.6273,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.28,
390
+ "learning_rate": 2.7132557295556566e-05,
391
+ "loss": 1.6173,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.28,
396
+ "learning_rate": 2.6996234068319198e-05,
397
+ "loss": 1.6577,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.29,
402
+ "learning_rate": 2.6857104533100482e-05,
403
+ "loss": 1.6459,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.29,
408
+ "learning_rate": 2.67152012368083e-05,
409
+ "loss": 1.6777,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.3,
414
+ "learning_rate": 2.657055737522315e-05,
415
+ "loss": 1.606,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.3,
420
+ "learning_rate": 2.642320678523258e-05,
421
+ "loss": 1.6268,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.31,
426
+ "learning_rate": 2.6273183936915657e-05,
427
+ "loss": 1.6301,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.31,
432
+ "learning_rate": 2.6120523925479275e-05,
433
+ "loss": 1.6264,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.32,
438
+ "learning_rate": 2.596526246304826e-05,
439
+ "loss": 1.6058,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.32,
444
+ "learning_rate": 2.5807435870311153e-05,
445
+ "loss": 1.5989,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.32,
450
+ "learning_rate": 2.5647081068023612e-05,
451
+ "loss": 1.6449,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.33,
456
+ "learning_rate": 2.5484235568371455e-05,
457
+ "loss": 1.6501,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.33,
462
+ "learning_rate": 2.5318937466195347e-05,
463
+ "loss": 1.6576,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.34,
468
+ "learning_rate": 2.5151225430079186e-05,
469
+ "loss": 1.6367,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.34,
474
+ "learning_rate": 2.498113869330424e-05,
475
+ "loss": 1.5932,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.35,
480
+ "learning_rate": 2.4808717044671252e-05,
481
+ "loss": 1.6535,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.35,
486
+ "learning_rate": 2.4634000819192524e-05,
487
+ "loss": 1.6104,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.35,
492
+ "learning_rate": 2.4457030888656257e-05,
493
+ "loss": 1.5829,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.36,
498
+ "learning_rate": 2.4277848652065314e-05,
499
+ "loss": 1.5652,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.36,
504
+ "learning_rate": 2.4096496025952643e-05,
505
+ "loss": 1.5931,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.37,
510
+ "learning_rate": 2.3913015434575658e-05,
511
+ "loss": 1.577,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.37,
516
+ "learning_rate": 2.3727449799991824e-05,
517
+ "loss": 1.5854,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.38,
522
+ "learning_rate": 2.3539842532017792e-05,
523
+ "loss": 1.5575,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.38,
528
+ "learning_rate": 2.3350237518074434e-05,
529
+ "loss": 1.6006,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.39,
534
+ "learning_rate": 2.315867911292016e-05,
535
+ "loss": 1.5524,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.39,
540
+ "learning_rate": 2.2965212128274908e-05,
541
+ "loss": 1.6059,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.39,
546
+ "learning_rate": 2.2769881822337202e-05,
547
+ "loss": 1.5913,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.4,
552
+ "learning_rate": 2.2572733889196826e-05,
553
+ "loss": 1.6139,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.4,
558
+ "learning_rate": 2.237381444814548e-05,
559
+ "loss": 1.5775,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.41,
564
+ "learning_rate": 2.2173170032887983e-05,
565
+ "loss": 1.5449,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.41,
570
+ "learning_rate": 2.197084758065653e-05,
571
+ "loss": 1.5664,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.42,
576
+ "learning_rate": 2.1766894421230555e-05,
577
+ "loss": 1.5408,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.42,
582
+ "learning_rate": 2.1561358265864765e-05,
583
+ "loss": 1.5566,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.42,
588
+ "learning_rate": 2.1354287196127925e-05,
589
+ "loss": 1.6271,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.43,
594
+ "learning_rate": 2.1145729652655013e-05,
595
+ "loss": 1.6316,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.43,
600
+ "learning_rate": 2.0935734423815405e-05,
601
+ "loss": 1.5676,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.44,
606
+ "learning_rate": 2.0724350634299685e-05,
607
+ "loss": 1.5719,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.44,
612
+ "learning_rate": 2.0511627733627788e-05,
613
+ "loss": 1.6086,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.45,
618
+ "learning_rate": 2.0297615484581137e-05,
619
+ "loss": 1.5628,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.45,
624
+ "learning_rate": 2.0082363951561548e-05,
625
+ "loss": 1.5568,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.46,
630
+ "learning_rate": 1.9865923488879518e-05,
631
+ "loss": 1.5568,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.46,
636
+ "learning_rate": 1.964834472897474e-05,
637
+ "loss": 1.588,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.46,
642
+ "learning_rate": 1.94296785705715e-05,
643
+ "loss": 1.5351,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.47,
648
+ "learning_rate": 1.9209976166771868e-05,
649
+ "loss": 1.5219,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.47,
654
+ "learning_rate": 1.898928891308928e-05,
655
+ "loss": 1.6106,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.48,
660
+ "learning_rate": 1.8767668435425485e-05,
661
+ "loss": 1.5856,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.48,
666
+ "learning_rate": 1.8545166577993576e-05,
667
+ "loss": 1.5815,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.49,
672
+ "learning_rate": 1.8321835391189933e-05,
673
+ "loss": 1.5464,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.49,
678
+ "learning_rate": 1.8097727119417963e-05,
679
+ "loss": 1.623,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.49,
684
+ "learning_rate": 1.787289418886643e-05,
685
+ "loss": 1.5216,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.5,
690
+ "learning_rate": 1.764738919524529e-05,
691
+ "loss": 1.5943,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.5,
696
+ "learning_rate": 1.7421264891481804e-05,
697
+ "loss": 1.5247,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.51,
702
+ "learning_rate": 1.719457417537998e-05,
703
+ "loss": 1.5036,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.51,
708
+ "learning_rate": 1.6967370077246014e-05,
709
+ "loss": 1.5497,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.52,
714
+ "learning_rate": 1.6739705747482807e-05,
715
+ "loss": 1.5809,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.52,
720
+ "learning_rate": 1.6511634444156355e-05,
721
+ "loss": 1.559,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.53,
726
+ "learning_rate": 1.628320952053695e-05,
727
+ "loss": 1.6082,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.53,
732
+ "learning_rate": 1.605448441261814e-05,
733
+ "loss": 1.5684,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.53,
738
+ "learning_rate": 1.582551262661633e-05,
739
+ "loss": 1.5227,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.54,
744
+ "learning_rate": 1.5596347726453887e-05,
745
+ "loss": 1.5204,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.54,
750
+ "learning_rate": 1.5367043321228868e-05,
751
+ "loss": 1.596,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.55,
756
+ "learning_rate": 1.5137653052674069e-05,
757
+ "loss": 1.5384,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.55,
762
+ "learning_rate": 1.4908230582608519e-05,
763
+ "loss": 1.5533,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.56,
768
+ "learning_rate": 1.4678829580384207e-05,
769
+ "loss": 1.5127,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.56,
774
+ "learning_rate": 1.4449503710331116e-05,
775
+ "loss": 1.5499,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.57,
780
+ "learning_rate": 1.4220306619203353e-05,
781
+ "loss": 1.5495,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.57,
786
+ "learning_rate": 1.399129192362949e-05,
787
+ "loss": 1.4868,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.57,
792
+ "learning_rate": 1.3762513197569838e-05,
793
+ "loss": 1.5192,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.58,
798
+ "learning_rate": 1.3534023959783825e-05,
799
+ "loss": 1.5397,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.58,
804
+ "learning_rate": 1.330587766131015e-05,
805
+ "loss": 1.5471,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.59,
810
+ "learning_rate": 1.3078127672962925e-05,
811
+ "loss": 1.5532,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.59,
816
+ "learning_rate": 1.2850827272846444e-05,
817
+ "loss": 1.5602,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.6,
822
+ "learning_rate": 1.2624029633891776e-05,
823
+ "loss": 1.538,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.6,
828
+ "learning_rate": 1.2397787811417839e-05,
829
+ "loss": 1.5226,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.6,
834
+ "learning_rate": 1.2172154730720058e-05,
835
+ "loss": 1.5437,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.61,
840
+ "learning_rate": 1.194718317468944e-05,
841
+ "loss": 1.5626,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.61,
846
+ "learning_rate": 1.1722925771464914e-05,
847
+ "loss": 1.5683,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.62,
852
+ "learning_rate": 1.1499434982121934e-05,
853
+ "loss": 1.5669,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.62,
858
+ "learning_rate": 1.1276763088400103e-05,
859
+ "loss": 1.5094,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.63,
864
+ "learning_rate": 1.1054962180472819e-05,
865
+ "loss": 1.5596,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.63,
870
+ "learning_rate": 1.083408414476164e-05,
871
+ "loss": 1.5473,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.64,
876
+ "learning_rate": 1.0614180651798437e-05,
877
+ "loss": 1.5201,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.64,
882
+ "learning_rate": 1.039530314413791e-05,
883
+ "loss": 1.5221,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.64,
888
+ "learning_rate": 1.0177502824323563e-05,
889
+ "loss": 1.5393,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.65,
894
+ "learning_rate": 9.960830642909732e-06,
895
+ "loss": 1.5441,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.65,
900
+ "learning_rate": 9.745337286542634e-06,
901
+ "loss": 1.5203,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.66,
906
+ "learning_rate": 9.53107316610309e-06,
907
+ "loss": 1.5602,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.66,
912
+ "learning_rate": 9.318088404913782e-06,
913
+ "loss": 1.5019,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.67,
918
+ "learning_rate": 9.10643282701382e-06,
919
+ "loss": 1.5203,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.67,
924
+ "learning_rate": 8.896155945503234e-06,
925
+ "loss": 1.4599,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.67,
930
+ "learning_rate": 8.687306950960302e-06,
931
+ "loss": 1.5127,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.68,
936
+ "learning_rate": 8.47993469993424e-06,
937
+ "loss": 1.509,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.68,
942
+ "learning_rate": 8.274087703516112e-06,
943
+ "loss": 1.5272,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.69,
948
+ "learning_rate": 8.069814115990468e-06,
949
+ "loss": 1.51,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.69,
954
+ "learning_rate": 7.86716172357054e-06,
955
+ "loss": 1.4966,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.7,
960
+ "learning_rate": 7.666177933219458e-06,
961
+ "loss": 1.5295,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.7,
966
+ "learning_rate": 7.466909761560261e-06,
967
+ "loss": 1.5158,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.71,
972
+ "learning_rate": 7.269403823877158e-06,
973
+ "loss": 1.5404,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.71,
978
+ "learning_rate": 7.073706323210714e-06,
979
+ "loss": 1.5392,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.71,
984
+ "learning_rate": 6.879863039549449e-06,
985
+ "loss": 1.539,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.72,
990
+ "learning_rate": 6.687919319120444e-06,
991
+ "loss": 1.5249,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.72,
996
+ "learning_rate": 6.497920063781366e-06,
997
+ "loss": 1.5101,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.73,
1002
+ "learning_rate": 6.309909720516454e-06,
1003
+ "loss": 1.5202,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.73,
1008
+ "learning_rate": 6.123932271038969e-06,
1009
+ "loss": 1.5144,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.74,
1014
+ "learning_rate": 5.9400312215024036e-06,
1015
+ "loss": 1.48,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.74,
1020
+ "learning_rate": 5.758249592323022e-06,
1021
+ "loss": 1.5081,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.74,
1026
+ "learning_rate": 5.578629908115953e-06,
1027
+ "loss": 1.5193,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.75,
1032
+ "learning_rate": 5.401214187747328e-06,
1033
+ "loss": 1.5101,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.75,
1038
+ "learning_rate": 5.226043934504695e-06,
1039
+ "loss": 1.5079,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.76,
1044
+ "learning_rate": 5.0531601263880755e-06,
1045
+ "loss": 1.5181,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.76,
1050
+ "learning_rate": 4.88260320652384e-06,
1051
+ "loss": 1.4956,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.77,
1056
+ "learning_rate": 4.714413073703804e-06,
1057
+ "loss": 1.5538,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.77,
1062
+ "learning_rate": 4.548629073051552e-06,
1063
+ "loss": 1.5294,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.78,
1068
+ "learning_rate": 4.38528998681839e-06,
1069
+ "loss": 1.547,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.78,
1074
+ "learning_rate": 4.224434025310873e-06,
1075
+ "loss": 1.4895,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.78,
1080
+ "learning_rate": 4.066098817952203e-06,
1081
+ "loss": 1.5382,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.79,
1086
+ "learning_rate": 3.910321404479495e-06,
1087
+ "loss": 1.5465,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.79,
1092
+ "learning_rate": 3.757138226278949e-06,
1093
+ "loss": 1.5266,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.8,
1098
+ "learning_rate": 3.6065851178610653e-06,
1099
+ "loss": 1.5216,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.8,
1104
+ "learning_rate": 3.4586972984777594e-06,
1105
+ "loss": 1.5313,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.81,
1110
+ "learning_rate": 3.313509363883465e-06,
1111
+ "loss": 1.5268,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.81,
1116
+ "learning_rate": 3.171055278242041e-06,
1117
+ "loss": 1.5157,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.81,
1122
+ "learning_rate": 3.031368366181474e-06,
1123
+ "loss": 1.5528,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.82,
1128
+ "learning_rate": 2.894481304998151e-06,
1129
+ "loss": 1.5244,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.82,
1134
+ "learning_rate": 2.7604261170125895e-06,
1135
+ "loss": 1.4934,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.83,
1140
+ "learning_rate": 2.6292341620783728e-06,
1141
+ "loss": 1.4907,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.83,
1146
+ "learning_rate": 2.5009361302460788e-06,
1147
+ "loss": 1.5115,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.84,
1152
+ "learning_rate": 2.3755620345838614e-06,
1153
+ "loss": 1.5441,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.84,
1158
+ "learning_rate": 2.2531412041564326e-06,
1159
+ "loss": 1.4515,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.85,
1164
+ "learning_rate": 2.1337022771640518e-06,
1165
+ "loss": 1.5211,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.85,
1170
+ "learning_rate": 2.0172731942430956e-06,
1171
+ "loss": 1.5178,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.85,
1176
+ "learning_rate": 1.90388119192987e-06,
1177
+ "loss": 1.5231,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.86,
1182
+ "learning_rate": 1.7935527962890779e-06,
1183
+ "loss": 1.5189,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.86,
1188
+ "learning_rate": 1.6863138167085236e-06,
1189
+ "loss": 1.5274,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.87,
1194
+ "learning_rate": 1.5821893398614535e-06,
1195
+ "loss": 1.5041,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.87,
1200
+ "learning_rate": 1.4812037238379862e-06,
1201
+ "loss": 1.4956,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.88,
1206
+ "learning_rate": 1.383380592446939e-06,
1207
+ "loss": 1.5179,
1208
+ "step": 2000
1209
+ }
1210
+ ],
1211
+ "max_steps": 2283,
1212
+ "num_train_epochs": 1,
1213
+ "total_flos": 0.0,
1214
+ "trial_name": null,
1215
+ "trial_params": null
1216
+ }
contextual/pct100/finetuning/pixel/Dialect_Combined/training_complete.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_path": "/root/pixel_cloud/contextual_100_work/data/contextual/downstream/Dialect_Combined_max64",
3
+ "dataset_rows": 292203,
4
+ "epochs": 1,
5
+ "global_step": 2283,
6
+ "input_checkpoint": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct100/pretraining/pixel/full_epoch2/final/pytorch_model.bin",
7
+ "learning_rate": 3e-05,
8
+ "metrics": {
9
+ "epoch": 1.0,
10
+ "train_loss": 1.6365420714941037,
11
+ "train_runtime": 587.3954,
12
+ "train_samples_per_second": 497.455,
13
+ "train_steps_per_second": 3.887
14
+ },
15
+ "model_type": "pixel",
16
+ "scheduler": "cosine",
17
+ "seed": 42,
18
+ "warmup_ratio": 0.1
19
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,1816 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.657030223390276,
5
+ "global_step": 3000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 6.564551422319475e-07,
13
+ "loss": 1.5971,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 1.312910284463895e-06,
19
+ "loss": 1.5699,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.9693654266958425e-06,
25
+ "loss": 1.5699,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 2.62582056892779e-06,
31
+ "loss": 1.527,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 3.2822757111597377e-06,
37
+ "loss": 1.5937,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 3.938730853391685e-06,
43
+ "loss": 1.6245,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.595185995623632e-06,
49
+ "loss": 1.5741,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 5.25164113785558e-06,
55
+ "loss": 1.5243,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 5.908096280087528e-06,
61
+ "loss": 1.4965,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 6.564551422319475e-06,
67
+ "loss": 1.5052,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 7.221006564551422e-06,
73
+ "loss": 1.5105,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 7.87746170678337e-06,
79
+ "loss": 1.3905,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 8.533916849015316e-06,
85
+ "loss": 1.5286,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 9.190371991247264e-06,
91
+ "loss": 1.4211,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 9.846827133479212e-06,
97
+ "loss": 1.4245,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 1.050328227571116e-05,
103
+ "loss": 1.4384,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 1.1159737417943107e-05,
109
+ "loss": 1.5141,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 1.1816192560175055e-05,
115
+ "loss": 1.4806,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 1.2472647702407003e-05,
121
+ "loss": 1.4234,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.04,
126
+ "learning_rate": 1.312910284463895e-05,
127
+ "loss": 1.3824,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 1.3785557986870897e-05,
133
+ "loss": 1.3962,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 1.4442013129102845e-05,
139
+ "loss": 1.4024,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 1.509846827133479e-05,
145
+ "loss": 1.4596,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.05,
150
+ "learning_rate": 1.575492341356674e-05,
151
+ "loss": 1.3976,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.05,
156
+ "learning_rate": 1.6411378555798686e-05,
157
+ "loss": 1.3971,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 1.7067833698030632e-05,
163
+ "loss": 1.3716,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 1.7724288840262582e-05,
169
+ "loss": 1.3521,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 1.8380743982494528e-05,
175
+ "loss": 1.309,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.06,
180
+ "learning_rate": 1.9037199124726478e-05,
181
+ "loss": 1.3445,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 1.9693654266958424e-05,
187
+ "loss": 1.3534,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 2.0350109409190373e-05,
193
+ "loss": 1.4066,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 2.100656455142232e-05,
199
+ "loss": 1.3413,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.07,
204
+ "learning_rate": 2.1663019693654265e-05,
205
+ "loss": 1.3277,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.07,
210
+ "learning_rate": 2.2319474835886215e-05,
211
+ "loss": 1.3379,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 2.297592997811816e-05,
217
+ "loss": 1.3441,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 2.363238512035011e-05,
223
+ "loss": 1.333,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.08,
228
+ "learning_rate": 2.4288840262582056e-05,
229
+ "loss": 1.2779,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.08,
234
+ "learning_rate": 2.4945295404814006e-05,
235
+ "loss": 1.3564,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 2.5601750547045952e-05,
241
+ "loss": 1.349,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 2.62582056892779e-05,
247
+ "loss": 1.3001,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.09,
252
+ "learning_rate": 2.6914660831509848e-05,
253
+ "loss": 1.3437,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.09,
258
+ "learning_rate": 2.7571115973741794e-05,
259
+ "loss": 1.2933,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.09,
264
+ "learning_rate": 2.8227571115973743e-05,
265
+ "loss": 1.3225,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 2.888402625820569e-05,
271
+ "loss": 1.2689,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 2.954048140043764e-05,
277
+ "loss": 1.3056,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.1,
282
+ "learning_rate": 2.9999960542365528e-05,
283
+ "loss": 1.3204,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.1,
288
+ "learning_rate": 2.9999259079083055e-05,
289
+ "loss": 1.2423,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 2.9997680826676758e-05,
295
+ "loss": 1.3127,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 2.9995225877404156e-05,
301
+ "loss": 1.2385,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.11,
306
+ "learning_rate": 2.9991894374770534e-05,
307
+ "loss": 1.2054,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.11,
312
+ "learning_rate": 2.9987686513520527e-05,
313
+ "loss": 1.3012,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.11,
318
+ "learning_rate": 2.9982602539626753e-05,
319
+ "loss": 1.2569,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 2.997664275027542e-05,
325
+ "loss": 1.2873,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.12,
330
+ "learning_rate": 2.996980749384897e-05,
331
+ "loss": 1.2596,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.12,
336
+ "learning_rate": 2.9962097169905706e-05,
337
+ "loss": 1.2721,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.12,
342
+ "learning_rate": 2.995351222915644e-05,
343
+ "loss": 1.3147,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.12,
348
+ "learning_rate": 2.9944053173438132e-05,
349
+ "loss": 1.2471,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 2.9933720555684573e-05,
355
+ "loss": 1.2009,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.13,
360
+ "learning_rate": 2.992251497989406e-05,
361
+ "loss": 1.234,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.13,
366
+ "learning_rate": 2.9910437101094075e-05,
367
+ "loss": 1.2737,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.13,
372
+ "learning_rate": 2.9897487625303016e-05,
373
+ "loss": 1.2294,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 2.9883667309488905e-05,
379
+ "loss": 1.2725,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.14,
384
+ "learning_rate": 2.9868976961525146e-05,
385
+ "loss": 1.2189,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.14,
390
+ "learning_rate": 2.985341744014332e-05,
391
+ "loss": 1.201,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.14,
396
+ "learning_rate": 2.983698965488296e-05,
397
+ "loss": 1.2314,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.14,
402
+ "learning_rate": 2.9819694566038392e-05,
403
+ "loss": 1.221,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.15,
408
+ "learning_rate": 2.980153318460261e-05,
409
+ "loss": 1.2214,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.15,
414
+ "learning_rate": 2.9782506572208166e-05,
415
+ "loss": 1.2892,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.15,
420
+ "learning_rate": 2.976261584106512e-05,
421
+ "loss": 1.2917,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.15,
426
+ "learning_rate": 2.974186215389601e-05,
427
+ "loss": 1.2192,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 2.9720246723867926e-05,
433
+ "loss": 1.2382,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.16,
438
+ "learning_rate": 2.9697770814521526e-05,
439
+ "loss": 1.2755,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.16,
444
+ "learning_rate": 2.9674435739697234e-05,
445
+ "loss": 1.2814,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.16,
450
+ "learning_rate": 2.9650242863458425e-05,
451
+ "loss": 1.1896,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.16,
456
+ "learning_rate": 2.9625193600011658e-05,
457
+ "loss": 1.2687,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.17,
462
+ "learning_rate": 2.959928941362405e-05,
463
+ "loss": 1.2341,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.17,
468
+ "learning_rate": 2.9572531818537654e-05,
469
+ "loss": 1.2552,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.17,
474
+ "learning_rate": 2.9544922378880957e-05,
475
+ "loss": 1.2113,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.17,
480
+ "learning_rate": 2.951646270857743e-05,
481
+ "loss": 1.2488,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.18,
486
+ "learning_rate": 2.9487154471251217e-05,
487
+ "loss": 1.2324,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.18,
492
+ "learning_rate": 2.9456999380129854e-05,
493
+ "loss": 1.2215,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.18,
498
+ "learning_rate": 2.942599919794413e-05,
499
+ "loss": 1.214,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.18,
504
+ "learning_rate": 2.939415573682507e-05,
505
+ "loss": 1.2048,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.18,
510
+ "learning_rate": 2.9361470858197978e-05,
511
+ "loss": 1.2311,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.19,
516
+ "learning_rate": 2.932794647267363e-05,
517
+ "loss": 1.2172,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.19,
522
+ "learning_rate": 2.929358453993659e-05,
523
+ "loss": 1.2192,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.19,
528
+ "learning_rate": 2.925838706863069e-05,
529
+ "loss": 1.2167,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.19,
534
+ "learning_rate": 2.9222356116241536e-05,
535
+ "loss": 1.1808,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.19,
540
+ "learning_rate": 2.9185493788976324e-05,
541
+ "loss": 1.2055,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.2,
546
+ "learning_rate": 2.914780224164066e-05,
547
+ "loss": 1.2608,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.2,
552
+ "learning_rate": 2.9109283677512624e-05,
553
+ "loss": 1.2,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.2,
558
+ "learning_rate": 2.9069940348213978e-05,
559
+ "loss": 1.2113,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.2,
564
+ "learning_rate": 2.9029774553578535e-05,
565
+ "loss": 1.2165,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.21,
570
+ "learning_rate": 2.8988788641517723e-05,
571
+ "loss": 1.1871,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.21,
576
+ "learning_rate": 2.8946985007883354e-05,
577
+ "loss": 1.2014,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.21,
582
+ "learning_rate": 2.890436609632755e-05,
583
+ "loss": 1.2269,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.21,
588
+ "learning_rate": 2.8860934398159905e-05,
589
+ "loss": 1.2274,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.21,
594
+ "learning_rate": 2.881669245220187e-05,
595
+ "loss": 1.1952,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.22,
600
+ "learning_rate": 2.8771642844638316e-05,
601
+ "loss": 1.2569,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.22,
606
+ "learning_rate": 2.8725788208866394e-05,
607
+ "loss": 1.2208,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.22,
612
+ "learning_rate": 2.867913122534155e-05,
613
+ "loss": 1.2224,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.22,
618
+ "learning_rate": 2.8631674621420895e-05,
619
+ "loss": 1.2134,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.23,
624
+ "learning_rate": 2.8583421171203728e-05,
625
+ "loss": 1.1977,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.23,
630
+ "learning_rate": 2.8534373695369393e-05,
631
+ "loss": 1.1629,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.23,
636
+ "learning_rate": 2.8484535061012404e-05,
637
+ "loss": 1.2527,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.23,
642
+ "learning_rate": 2.8433908181474845e-05,
643
+ "loss": 1.1827,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.23,
648
+ "learning_rate": 2.8382496016176033e-05,
649
+ "loss": 1.1919,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.24,
654
+ "learning_rate": 2.8330301570439588e-05,
655
+ "loss": 1.1684,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.24,
660
+ "learning_rate": 2.8277327895317703e-05,
661
+ "loss": 1.2174,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.24,
666
+ "learning_rate": 2.8223578087412806e-05,
667
+ "loss": 1.2241,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.24,
672
+ "learning_rate": 2.816905528869657e-05,
673
+ "loss": 1.2101,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.25,
678
+ "learning_rate": 2.8113762686326204e-05,
679
+ "loss": 1.2062,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.25,
684
+ "learning_rate": 2.80577035124582e-05,
685
+ "loss": 1.2098,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.25,
690
+ "learning_rate": 2.8000881044059343e-05,
691
+ "loss": 1.164,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.25,
696
+ "learning_rate": 2.7943298602715196e-05,
697
+ "loss": 1.2248,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.25,
702
+ "learning_rate": 2.7884959554435894e-05,
703
+ "loss": 1.1556,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.26,
708
+ "learning_rate": 2.7825867309459432e-05,
709
+ "loss": 1.2187,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.26,
714
+ "learning_rate": 2.7766025322052274e-05,
715
+ "loss": 1.2088,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.26,
720
+ "learning_rate": 2.7705437090307463e-05,
721
+ "loss": 1.2224,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.26,
726
+ "learning_rate": 2.7644106155940102e-05,
727
+ "loss": 1.1823,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.27,
732
+ "learning_rate": 2.7582036104080366e-05,
733
+ "loss": 1.2382,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.27,
738
+ "learning_rate": 2.7519230563063903e-05,
739
+ "loss": 1.2109,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.27,
744
+ "learning_rate": 2.7455693204219735e-05,
745
+ "loss": 1.2101,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.27,
750
+ "learning_rate": 2.7391427741655673e-05,
751
+ "loss": 1.2332,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.27,
756
+ "learning_rate": 2.7326437932041178e-05,
757
+ "loss": 1.1669,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.28,
762
+ "learning_rate": 2.7260727574387795e-05,
763
+ "loss": 1.2106,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.28,
768
+ "learning_rate": 2.7194300509827035e-05,
769
+ "loss": 1.1902,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.28,
774
+ "learning_rate": 2.7127160621385888e-05,
775
+ "loss": 1.2188,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.28,
780
+ "learning_rate": 2.7059311833759813e-05,
781
+ "loss": 1.1779,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.28,
786
+ "learning_rate": 2.6990758113083318e-05,
787
+ "loss": 1.1895,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.29,
792
+ "learning_rate": 2.692150346669812e-05,
793
+ "loss": 1.1877,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.29,
798
+ "learning_rate": 2.6851551942918905e-05,
799
+ "loss": 1.1375,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.29,
804
+ "learning_rate": 2.6780907630796663e-05,
805
+ "loss": 1.2363,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.29,
810
+ "learning_rate": 2.6709574659879674e-05,
811
+ "loss": 1.2,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.3,
816
+ "learning_rate": 2.6637557199972115e-05,
817
+ "loss": 1.1543,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.3,
822
+ "learning_rate": 2.656485946089029e-05,
823
+ "loss": 1.1622,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.3,
828
+ "learning_rate": 2.6491485692216576e-05,
829
+ "loss": 1.1567,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.3,
834
+ "learning_rate": 2.6417440183050987e-05,
835
+ "loss": 1.1625,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.3,
840
+ "learning_rate": 2.6342727261760464e-05,
841
+ "loss": 1.1884,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.31,
846
+ "learning_rate": 2.626735129572585e-05,
847
+ "loss": 1.1959,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.31,
852
+ "learning_rate": 2.61913166910866e-05,
853
+ "loss": 1.1931,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.31,
858
+ "learning_rate": 2.6114627892483217e-05,
859
+ "loss": 1.2012,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.31,
864
+ "learning_rate": 2.603728938279743e-05,
865
+ "loss": 1.1935,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.32,
870
+ "learning_rate": 2.5959305682890154e-05,
871
+ "loss": 1.1837,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.32,
876
+ "learning_rate": 2.5880681351337224e-05,
877
+ "loss": 1.1224,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.32,
882
+ "learning_rate": 2.5801420984162914e-05,
883
+ "loss": 1.1786,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.32,
888
+ "learning_rate": 2.572152921457126e-05,
889
+ "loss": 1.2358,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.32,
894
+ "learning_rate": 2.5641010712675268e-05,
895
+ "loss": 1.2231,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.33,
900
+ "learning_rate": 2.5559870185223863e-05,
901
+ "loss": 1.1609,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.33,
906
+ "learning_rate": 2.5478112375326796e-05,
907
+ "loss": 1.1832,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.33,
912
+ "learning_rate": 2.5395742062177375e-05,
913
+ "loss": 1.1918,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.33,
918
+ "learning_rate": 2.5312764060773073e-05,
919
+ "loss": 1.1982,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.34,
924
+ "learning_rate": 2.5229183221634098e-05,
925
+ "loss": 1.1328,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.34,
930
+ "learning_rate": 2.514500443051982e-05,
931
+ "loss": 1.2386,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.34,
936
+ "learning_rate": 2.5060232608143207e-05,
937
+ "loss": 1.2092,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.34,
942
+ "learning_rate": 2.4974872709883138e-05,
943
+ "loss": 1.138,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.34,
948
+ "learning_rate": 2.488892972549479e-05,
949
+ "loss": 1.1936,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.35,
954
+ "learning_rate": 2.48024086788179e-05,
955
+ "loss": 1.1598,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.35,
960
+ "learning_rate": 2.471531462748315e-05,
961
+ "loss": 1.1555,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.35,
966
+ "learning_rate": 2.4627652662616463e-05,
967
+ "loss": 1.1645,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.35,
972
+ "learning_rate": 2.4539427908541462e-05,
973
+ "loss": 1.154,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.35,
978
+ "learning_rate": 2.4450645522479855e-05,
979
+ "loss": 1.1773,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.36,
984
+ "learning_rate": 2.4361310694250044e-05,
985
+ "loss": 1.1095,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.36,
990
+ "learning_rate": 2.4271428645963676e-05,
991
+ "loss": 1.183,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.36,
996
+ "learning_rate": 2.4181004631720444e-05,
997
+ "loss": 1.1415,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.36,
1002
+ "learning_rate": 2.409004393730091e-05,
1003
+ "loss": 1.1263,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.37,
1008
+ "learning_rate": 2.3998551879857554e-05,
1009
+ "loss": 1.1323,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.37,
1014
+ "learning_rate": 2.3906533807603952e-05,
1015
+ "loss": 1.1679,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.37,
1020
+ "learning_rate": 2.3813995099502108e-05,
1021
+ "loss": 1.2309,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.37,
1026
+ "learning_rate": 2.372094116494809e-05,
1027
+ "loss": 1.2011,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.37,
1032
+ "learning_rate": 2.3627377443455745e-05,
1033
+ "loss": 1.1399,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.38,
1038
+ "learning_rate": 2.3533309404338802e-05,
1039
+ "loss": 1.1185,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.38,
1044
+ "learning_rate": 2.343874254639111e-05,
1045
+ "loss": 1.1908,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.38,
1050
+ "learning_rate": 2.3343682397565216e-05,
1051
+ "loss": 1.1455,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.38,
1056
+ "learning_rate": 2.324813451464924e-05,
1057
+ "loss": 1.1138,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.39,
1062
+ "learning_rate": 2.3152104482942038e-05,
1063
+ "loss": 1.1683,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.39,
1068
+ "learning_rate": 2.3055597915926706e-05,
1069
+ "loss": 1.1528,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.39,
1074
+ "learning_rate": 2.295862045494246e-05,
1075
+ "loss": 1.1433,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.39,
1080
+ "learning_rate": 2.2861177768854833e-05,
1081
+ "loss": 1.2187,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.39,
1086
+ "learning_rate": 2.2763275553724355e-05,
1087
+ "loss": 1.1773,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.4,
1092
+ "learning_rate": 2.2664919532473526e-05,
1093
+ "loss": 1.1288,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.4,
1098
+ "learning_rate": 2.2566115454552325e-05,
1099
+ "loss": 1.1872,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.4,
1104
+ "learning_rate": 2.2466869095602093e-05,
1105
+ "loss": 1.187,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.4,
1110
+ "learning_rate": 2.2367186257117935e-05,
1111
+ "loss": 1.1506,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.41,
1116
+ "learning_rate": 2.2267072766109584e-05,
1117
+ "loss": 1.1548,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.41,
1122
+ "learning_rate": 2.2166534474760776e-05,
1123
+ "loss": 1.1747,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.41,
1128
+ "learning_rate": 2.206557726008717e-05,
1129
+ "loss": 1.1174,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.41,
1134
+ "learning_rate": 2.196420702359278e-05,
1135
+ "loss": 1.2132,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.41,
1140
+ "learning_rate": 2.1862429690925028e-05,
1141
+ "loss": 1.1553,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.42,
1146
+ "learning_rate": 2.176025121152834e-05,
1147
+ "loss": 1.1164,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.42,
1152
+ "learning_rate": 2.1657677558296378e-05,
1153
+ "loss": 1.137,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.42,
1158
+ "learning_rate": 2.155471472722288e-05,
1159
+ "loss": 1.1301,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.42,
1164
+ "learning_rate": 2.145136873705116e-05,
1165
+ "loss": 1.1698,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.42,
1170
+ "learning_rate": 2.134764562892231e-05,
1171
+ "loss": 1.156,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.43,
1176
+ "learning_rate": 2.1243551466022024e-05,
1177
+ "loss": 1.1195,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.43,
1182
+ "learning_rate": 2.113909233322618e-05,
1183
+ "loss": 1.161,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.43,
1188
+ "learning_rate": 2.103427433674517e-05,
1189
+ "loss": 1.1478,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.43,
1194
+ "learning_rate": 2.092910360376693e-05,
1195
+ "loss": 1.1472,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.44,
1200
+ "learning_rate": 2.0823586282098773e-05,
1201
+ "loss": 1.2036,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.44,
1206
+ "learning_rate": 2.0717728539808053e-05,
1207
+ "loss": 1.1643,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.44,
1212
+ "learning_rate": 2.0611536564861552e-05,
1213
+ "loss": 1.16,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.44,
1218
+ "learning_rate": 2.0505016564763803e-05,
1219
+ "loss": 1.1677,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.44,
1224
+ "learning_rate": 2.039817476619421e-05,
1225
+ "loss": 1.1765,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.45,
1230
+ "learning_rate": 2.029101741464306e-05,
1231
+ "loss": 1.1864,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.45,
1236
+ "learning_rate": 2.0183550774046454e-05,
1237
+ "loss": 1.1602,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.45,
1242
+ "learning_rate": 2.007578112642013e-05,
1243
+ "loss": 1.1733,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.45,
1248
+ "learning_rate": 1.9967714771492266e-05,
1249
+ "loss": 1.1544,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.46,
1254
+ "learning_rate": 1.9859358026335186e-05,
1255
+ "loss": 1.1212,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.46,
1260
+ "learning_rate": 1.9750717224996153e-05,
1261
+ "loss": 1.1324,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.46,
1266
+ "learning_rate": 1.964179871812704e-05,
1267
+ "loss": 1.1565,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.46,
1272
+ "learning_rate": 1.953260887261316e-05,
1273
+ "loss": 1.1682,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.46,
1278
+ "learning_rate": 1.9423154071201056e-05,
1279
+ "loss": 1.1407,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.47,
1284
+ "learning_rate": 1.931344071212539e-05,
1285
+ "loss": 1.1275,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.47,
1290
+ "learning_rate": 1.920347520873494e-05,
1291
+ "loss": 1.1789,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.47,
1296
+ "learning_rate": 1.909326398911772e-05,
1297
+ "loss": 1.1529,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.47,
1302
+ "learning_rate": 1.8982813495725202e-05,
1303
+ "loss": 1.1587,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.48,
1308
+ "learning_rate": 1.8872130184995713e-05,
1309
+ "loss": 1.1519,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.48,
1314
+ "learning_rate": 1.8761220526977046e-05,
1315
+ "loss": 1.1309,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.48,
1320
+ "learning_rate": 1.865009100494823e-05,
1321
+ "loss": 1.1494,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.48,
1326
+ "learning_rate": 1.8538748115040556e-05,
1327
+ "loss": 1.1685,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.48,
1332
+ "learning_rate": 1.8427198365857846e-05,
1333
+ "loss": 1.1581,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.49,
1338
+ "learning_rate": 1.8315448278095976e-05,
1339
+ "loss": 1.167,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.49,
1344
+ "learning_rate": 1.820350438416171e-05,
1345
+ "loss": 1.1376,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.49,
1350
+ "learning_rate": 1.809137322779086e-05,
1351
+ "loss": 1.1487,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.49,
1356
+ "learning_rate": 1.797906136366574e-05,
1357
+ "loss": 1.1313,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.49,
1362
+ "learning_rate": 1.7866575357032042e-05,
1363
+ "loss": 1.2109,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.5,
1368
+ "learning_rate": 1.7753921783315033e-05,
1369
+ "loss": 1.162,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.5,
1374
+ "learning_rate": 1.764110722773521e-05,
1375
+ "loss": 1.1593,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.5,
1380
+ "learning_rate": 1.7528138284923326e-05,
1381
+ "loss": 1.177,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.5,
1386
+ "learning_rate": 1.741502155853493e-05,
1387
+ "loss": 1.1137,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.51,
1392
+ "learning_rate": 1.7301763660864328e-05,
1393
+ "loss": 1.1503,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.51,
1398
+ "learning_rate": 1.7188371212458063e-05,
1399
+ "loss": 1.1725,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.51,
1404
+ "learning_rate": 1.7074850841727914e-05,
1405
+ "loss": 1.1668,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.51,
1410
+ "learning_rate": 1.696120918456341e-05,
1411
+ "loss": 1.1573,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.51,
1416
+ "learning_rate": 1.684745288394396e-05,
1417
+ "loss": 1.1691,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.52,
1422
+ "learning_rate": 1.673358858955049e-05,
1423
+ "loss": 1.1176,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.52,
1428
+ "learning_rate": 1.6619622957376762e-05,
1429
+ "loss": 1.1469,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.52,
1434
+ "learning_rate": 1.65055626493403e-05,
1435
+ "loss": 1.1049,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.52,
1440
+ "learning_rate": 1.6391414332892942e-05,
1441
+ "loss": 1.2469,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.53,
1446
+ "learning_rate": 1.6277184680631102e-05,
1447
+ "loss": 1.1681,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.53,
1452
+ "learning_rate": 1.616288036990572e-05,
1453
+ "loss": 1.131,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.53,
1458
+ "learning_rate": 1.6048508082431938e-05,
1459
+ "loss": 1.1285,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.53,
1464
+ "learning_rate": 1.5934074503898504e-05,
1465
+ "loss": 1.1555,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.53,
1470
+ "learning_rate": 1.5819586323576972e-05,
1471
+ "loss": 1.1507,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.54,
1476
+ "learning_rate": 1.5705050233930672e-05,
1477
+ "loss": 1.1126,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.54,
1482
+ "learning_rate": 1.5590472930223484e-05,
1483
+ "loss": 1.166,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.54,
1488
+ "learning_rate": 1.54758611101285e-05,
1489
+ "loss": 1.1631,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.54,
1494
+ "learning_rate": 1.536122147333646e-05,
1495
+ "loss": 1.1394,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.55,
1500
+ "learning_rate": 1.5246560721164158e-05,
1501
+ "loss": 1.1729,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.55,
1506
+ "learning_rate": 1.5131885556162693e-05,
1507
+ "loss": 1.1309,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.55,
1512
+ "learning_rate": 1.5017202681725672e-05,
1513
+ "loss": 1.1462,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.55,
1518
+ "learning_rate": 1.4902518801697365e-05,
1519
+ "loss": 1.1553,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.55,
1524
+ "learning_rate": 1.4787840619980825e-05,
1525
+ "loss": 1.153,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.56,
1530
+ "learning_rate": 1.4673174840146e-05,
1531
+ "loss": 1.1712,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.56,
1536
+ "learning_rate": 1.455852816503788e-05,
1537
+ "loss": 1.1652,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.56,
1542
+ "learning_rate": 1.4443907296384687e-05,
1543
+ "loss": 1.1061,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.56,
1548
+ "learning_rate": 1.4329318934406113e-05,
1549
+ "loss": 1.1369,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.57,
1554
+ "learning_rate": 1.4214769777421652e-05,
1555
+ "loss": 1.1275,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.57,
1560
+ "learning_rate": 1.4100266521459044e-05,
1561
+ "loss": 1.1134,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.57,
1566
+ "learning_rate": 1.3985815859862881e-05,
1567
+ "loss": 1.1047,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.57,
1572
+ "learning_rate": 1.3871424482903308e-05,
1573
+ "loss": 1.0891,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.57,
1578
+ "learning_rate": 1.3757099077384965e-05,
1579
+ "loss": 1.1455,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.58,
1584
+ "learning_rate": 1.3642846326256088e-05,
1585
+ "loss": 1.1312,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.58,
1590
+ "learning_rate": 1.3528672908217888e-05,
1591
+ "loss": 1.1389,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.58,
1596
+ "learning_rate": 1.34145854973341e-05,
1597
+ "loss": 1.1836,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.58,
1602
+ "learning_rate": 1.3300590762640868e-05,
1603
+ "loss": 1.0937,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.58,
1608
+ "learning_rate": 1.3186695367756916e-05,
1609
+ "loss": 1.1978,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.59,
1614
+ "learning_rate": 1.3072905970494003e-05,
1615
+ "loss": 1.151,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.59,
1620
+ "learning_rate": 1.295922922246775e-05,
1621
+ "loss": 1.1011,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.59,
1626
+ "learning_rate": 1.28456717687088e-05,
1627
+ "loss": 1.0861,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.59,
1632
+ "learning_rate": 1.2732240247274395e-05,
1633
+ "loss": 1.1113,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.6,
1638
+ "learning_rate": 1.2618941288860352e-05,
1639
+ "loss": 1.1614,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.6,
1644
+ "learning_rate": 1.250578151641344e-05,
1645
+ "loss": 1.1503,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.6,
1650
+ "learning_rate": 1.239276754474423e-05,
1651
+ "loss": 1.1446,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.6,
1656
+ "learning_rate": 1.2279905980140473e-05,
1657
+ "loss": 1.1023,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.6,
1662
+ "learning_rate": 1.216720341998085e-05,
1663
+ "loss": 1.1615,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.61,
1668
+ "learning_rate": 1.2054666452349389e-05,
1669
+ "loss": 1.0865,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.61,
1674
+ "learning_rate": 1.19423016556503e-05,
1675
+ "loss": 1.127,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.61,
1680
+ "learning_rate": 1.1830115598223478e-05,
1681
+ "loss": 1.2051,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.61,
1686
+ "learning_rate": 1.171811483796051e-05,
1687
+ "loss": 1.0966,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.62,
1692
+ "learning_rate": 1.1606305921921349e-05,
1693
+ "loss": 1.1681,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.62,
1698
+ "learning_rate": 1.1494695385951596e-05,
1699
+ "loss": 1.201,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.62,
1704
+ "learning_rate": 1.1383289754300463e-05,
1705
+ "loss": 1.1575,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.62,
1710
+ "learning_rate": 1.1272095539239358e-05,
1711
+ "loss": 1.1316,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.62,
1716
+ "learning_rate": 1.1161119240681233e-05,
1717
+ "loss": 1.151,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.63,
1722
+ "learning_rate": 1.1050367345800628e-05,
1723
+ "loss": 1.1619,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.63,
1728
+ "learning_rate": 1.0939846328654456e-05,
1729
+ "loss": 1.1264,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.63,
1734
+ "learning_rate": 1.0829562649803551e-05,
1735
+ "loss": 1.1052,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.63,
1740
+ "learning_rate": 1.0719522755935018e-05,
1741
+ "loss": 1.1075,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.64,
1746
+ "learning_rate": 1.0609733079485404e-05,
1747
+ "loss": 1.1597,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.64,
1752
+ "learning_rate": 1.0500200038264657e-05,
1753
+ "loss": 1.1135,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.64,
1758
+ "learning_rate": 1.0390930035080994e-05,
1759
+ "loss": 1.1193,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.64,
1764
+ "learning_rate": 1.0281929457366599e-05,
1765
+ "loss": 1.1912,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.64,
1770
+ "learning_rate": 1.0173204676804284e-05,
1771
+ "loss": 1.1432,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.65,
1776
+ "learning_rate": 1.0064762048954982e-05,
1777
+ "loss": 1.172,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.65,
1782
+ "learning_rate": 9.956607912886247e-06,
1783
+ "loss": 1.1299,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.65,
1788
+ "learning_rate": 9.848748590801712e-06,
1789
+ "loss": 1.0882,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.65,
1794
+ "learning_rate": 9.741190387671512e-06,
1795
+ "loss": 1.1207,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.65,
1800
+ "learning_rate": 9.633939590863725e-06,
1801
+ "loss": 1.1749,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.66,
1806
+ "learning_rate": 9.527002469776827e-06,
1807
+ "loss": 1.0907,
1808
+ "step": 3000
1809
+ }
1810
+ ],
1811
+ "max_steps": 4566,
1812
+ "num_train_epochs": 1,
1813
+ "total_flos": 0.0,
1814
+ "trial_name": null,
1815
+ "trial_params": null
1816
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,2416 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.8760402978537013,
5
+ "global_step": 4000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 6.564551422319475e-07,
13
+ "loss": 1.5971,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 1.312910284463895e-06,
19
+ "loss": 1.5699,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.9693654266958425e-06,
25
+ "loss": 1.5699,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 2.62582056892779e-06,
31
+ "loss": 1.527,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 3.2822757111597377e-06,
37
+ "loss": 1.5937,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 3.938730853391685e-06,
43
+ "loss": 1.6245,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.595185995623632e-06,
49
+ "loss": 1.5741,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 5.25164113785558e-06,
55
+ "loss": 1.5243,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 5.908096280087528e-06,
61
+ "loss": 1.4965,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 6.564551422319475e-06,
67
+ "loss": 1.5052,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 7.221006564551422e-06,
73
+ "loss": 1.5105,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 7.87746170678337e-06,
79
+ "loss": 1.3905,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 8.533916849015316e-06,
85
+ "loss": 1.5286,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 9.190371991247264e-06,
91
+ "loss": 1.4211,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 9.846827133479212e-06,
97
+ "loss": 1.4245,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 1.050328227571116e-05,
103
+ "loss": 1.4384,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 1.1159737417943107e-05,
109
+ "loss": 1.5141,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 1.1816192560175055e-05,
115
+ "loss": 1.4806,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 1.2472647702407003e-05,
121
+ "loss": 1.4234,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.04,
126
+ "learning_rate": 1.312910284463895e-05,
127
+ "loss": 1.3824,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 1.3785557986870897e-05,
133
+ "loss": 1.3962,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 1.4442013129102845e-05,
139
+ "loss": 1.4024,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 1.509846827133479e-05,
145
+ "loss": 1.4596,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.05,
150
+ "learning_rate": 1.575492341356674e-05,
151
+ "loss": 1.3976,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.05,
156
+ "learning_rate": 1.6411378555798686e-05,
157
+ "loss": 1.3971,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 1.7067833698030632e-05,
163
+ "loss": 1.3716,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 1.7724288840262582e-05,
169
+ "loss": 1.3521,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 1.8380743982494528e-05,
175
+ "loss": 1.309,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.06,
180
+ "learning_rate": 1.9037199124726478e-05,
181
+ "loss": 1.3445,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 1.9693654266958424e-05,
187
+ "loss": 1.3534,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 2.0350109409190373e-05,
193
+ "loss": 1.4066,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 2.100656455142232e-05,
199
+ "loss": 1.3413,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.07,
204
+ "learning_rate": 2.1663019693654265e-05,
205
+ "loss": 1.3277,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.07,
210
+ "learning_rate": 2.2319474835886215e-05,
211
+ "loss": 1.3379,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 2.297592997811816e-05,
217
+ "loss": 1.3441,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 2.363238512035011e-05,
223
+ "loss": 1.333,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.08,
228
+ "learning_rate": 2.4288840262582056e-05,
229
+ "loss": 1.2779,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.08,
234
+ "learning_rate": 2.4945295404814006e-05,
235
+ "loss": 1.3564,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 2.5601750547045952e-05,
241
+ "loss": 1.349,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 2.62582056892779e-05,
247
+ "loss": 1.3001,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.09,
252
+ "learning_rate": 2.6914660831509848e-05,
253
+ "loss": 1.3437,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.09,
258
+ "learning_rate": 2.7571115973741794e-05,
259
+ "loss": 1.2933,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.09,
264
+ "learning_rate": 2.8227571115973743e-05,
265
+ "loss": 1.3225,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 2.888402625820569e-05,
271
+ "loss": 1.2689,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 2.954048140043764e-05,
277
+ "loss": 1.3056,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.1,
282
+ "learning_rate": 2.9999960542365528e-05,
283
+ "loss": 1.3204,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.1,
288
+ "learning_rate": 2.9999259079083055e-05,
289
+ "loss": 1.2423,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 2.9997680826676758e-05,
295
+ "loss": 1.3127,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 2.9995225877404156e-05,
301
+ "loss": 1.2385,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.11,
306
+ "learning_rate": 2.9991894374770534e-05,
307
+ "loss": 1.2054,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.11,
312
+ "learning_rate": 2.9987686513520527e-05,
313
+ "loss": 1.3012,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.11,
318
+ "learning_rate": 2.9982602539626753e-05,
319
+ "loss": 1.2569,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 2.997664275027542e-05,
325
+ "loss": 1.2873,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.12,
330
+ "learning_rate": 2.996980749384897e-05,
331
+ "loss": 1.2596,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.12,
336
+ "learning_rate": 2.9962097169905706e-05,
337
+ "loss": 1.2721,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.12,
342
+ "learning_rate": 2.995351222915644e-05,
343
+ "loss": 1.3147,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.12,
348
+ "learning_rate": 2.9944053173438132e-05,
349
+ "loss": 1.2471,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 2.9933720555684573e-05,
355
+ "loss": 1.2009,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.13,
360
+ "learning_rate": 2.992251497989406e-05,
361
+ "loss": 1.234,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.13,
366
+ "learning_rate": 2.9910437101094075e-05,
367
+ "loss": 1.2737,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.13,
372
+ "learning_rate": 2.9897487625303016e-05,
373
+ "loss": 1.2294,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 2.9883667309488905e-05,
379
+ "loss": 1.2725,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.14,
384
+ "learning_rate": 2.9868976961525146e-05,
385
+ "loss": 1.2189,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.14,
390
+ "learning_rate": 2.985341744014332e-05,
391
+ "loss": 1.201,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.14,
396
+ "learning_rate": 2.983698965488296e-05,
397
+ "loss": 1.2314,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.14,
402
+ "learning_rate": 2.9819694566038392e-05,
403
+ "loss": 1.221,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.15,
408
+ "learning_rate": 2.980153318460261e-05,
409
+ "loss": 1.2214,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.15,
414
+ "learning_rate": 2.9782506572208166e-05,
415
+ "loss": 1.2892,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.15,
420
+ "learning_rate": 2.976261584106512e-05,
421
+ "loss": 1.2917,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.15,
426
+ "learning_rate": 2.974186215389601e-05,
427
+ "loss": 1.2192,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 2.9720246723867926e-05,
433
+ "loss": 1.2382,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.16,
438
+ "learning_rate": 2.9697770814521526e-05,
439
+ "loss": 1.2755,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.16,
444
+ "learning_rate": 2.9674435739697234e-05,
445
+ "loss": 1.2814,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.16,
450
+ "learning_rate": 2.9650242863458425e-05,
451
+ "loss": 1.1896,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.16,
456
+ "learning_rate": 2.9625193600011658e-05,
457
+ "loss": 1.2687,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.17,
462
+ "learning_rate": 2.959928941362405e-05,
463
+ "loss": 1.2341,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.17,
468
+ "learning_rate": 2.9572531818537654e-05,
469
+ "loss": 1.2552,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.17,
474
+ "learning_rate": 2.9544922378880957e-05,
475
+ "loss": 1.2113,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.17,
480
+ "learning_rate": 2.951646270857743e-05,
481
+ "loss": 1.2488,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.18,
486
+ "learning_rate": 2.9487154471251217e-05,
487
+ "loss": 1.2324,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.18,
492
+ "learning_rate": 2.9456999380129854e-05,
493
+ "loss": 1.2215,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.18,
498
+ "learning_rate": 2.942599919794413e-05,
499
+ "loss": 1.214,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.18,
504
+ "learning_rate": 2.939415573682507e-05,
505
+ "loss": 1.2048,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.18,
510
+ "learning_rate": 2.9361470858197978e-05,
511
+ "loss": 1.2311,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.19,
516
+ "learning_rate": 2.932794647267363e-05,
517
+ "loss": 1.2172,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.19,
522
+ "learning_rate": 2.929358453993659e-05,
523
+ "loss": 1.2192,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.19,
528
+ "learning_rate": 2.925838706863069e-05,
529
+ "loss": 1.2167,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.19,
534
+ "learning_rate": 2.9222356116241536e-05,
535
+ "loss": 1.1808,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.19,
540
+ "learning_rate": 2.9185493788976324e-05,
541
+ "loss": 1.2055,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.2,
546
+ "learning_rate": 2.914780224164066e-05,
547
+ "loss": 1.2608,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.2,
552
+ "learning_rate": 2.9109283677512624e-05,
553
+ "loss": 1.2,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.2,
558
+ "learning_rate": 2.9069940348213978e-05,
559
+ "loss": 1.2113,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.2,
564
+ "learning_rate": 2.9029774553578535e-05,
565
+ "loss": 1.2165,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.21,
570
+ "learning_rate": 2.8988788641517723e-05,
571
+ "loss": 1.1871,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.21,
576
+ "learning_rate": 2.8946985007883354e-05,
577
+ "loss": 1.2014,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.21,
582
+ "learning_rate": 2.890436609632755e-05,
583
+ "loss": 1.2269,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.21,
588
+ "learning_rate": 2.8860934398159905e-05,
589
+ "loss": 1.2274,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.21,
594
+ "learning_rate": 2.881669245220187e-05,
595
+ "loss": 1.1952,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.22,
600
+ "learning_rate": 2.8771642844638316e-05,
601
+ "loss": 1.2569,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.22,
606
+ "learning_rate": 2.8725788208866394e-05,
607
+ "loss": 1.2208,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.22,
612
+ "learning_rate": 2.867913122534155e-05,
613
+ "loss": 1.2224,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.22,
618
+ "learning_rate": 2.8631674621420895e-05,
619
+ "loss": 1.2134,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.23,
624
+ "learning_rate": 2.8583421171203728e-05,
625
+ "loss": 1.1977,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.23,
630
+ "learning_rate": 2.8534373695369393e-05,
631
+ "loss": 1.1629,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.23,
636
+ "learning_rate": 2.8484535061012404e-05,
637
+ "loss": 1.2527,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.23,
642
+ "learning_rate": 2.8433908181474845e-05,
643
+ "loss": 1.1827,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.23,
648
+ "learning_rate": 2.8382496016176033e-05,
649
+ "loss": 1.1919,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.24,
654
+ "learning_rate": 2.8330301570439588e-05,
655
+ "loss": 1.1684,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.24,
660
+ "learning_rate": 2.8277327895317703e-05,
661
+ "loss": 1.2174,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.24,
666
+ "learning_rate": 2.8223578087412806e-05,
667
+ "loss": 1.2241,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.24,
672
+ "learning_rate": 2.816905528869657e-05,
673
+ "loss": 1.2101,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.25,
678
+ "learning_rate": 2.8113762686326204e-05,
679
+ "loss": 1.2062,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.25,
684
+ "learning_rate": 2.80577035124582e-05,
685
+ "loss": 1.2098,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.25,
690
+ "learning_rate": 2.8000881044059343e-05,
691
+ "loss": 1.164,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.25,
696
+ "learning_rate": 2.7943298602715196e-05,
697
+ "loss": 1.2248,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.25,
702
+ "learning_rate": 2.7884959554435894e-05,
703
+ "loss": 1.1556,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.26,
708
+ "learning_rate": 2.7825867309459432e-05,
709
+ "loss": 1.2187,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.26,
714
+ "learning_rate": 2.7766025322052274e-05,
715
+ "loss": 1.2088,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.26,
720
+ "learning_rate": 2.7705437090307463e-05,
721
+ "loss": 1.2224,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.26,
726
+ "learning_rate": 2.7644106155940102e-05,
727
+ "loss": 1.1823,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.27,
732
+ "learning_rate": 2.7582036104080366e-05,
733
+ "loss": 1.2382,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.27,
738
+ "learning_rate": 2.7519230563063903e-05,
739
+ "loss": 1.2109,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.27,
744
+ "learning_rate": 2.7455693204219735e-05,
745
+ "loss": 1.2101,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.27,
750
+ "learning_rate": 2.7391427741655673e-05,
751
+ "loss": 1.2332,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.27,
756
+ "learning_rate": 2.7326437932041178e-05,
757
+ "loss": 1.1669,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.28,
762
+ "learning_rate": 2.7260727574387795e-05,
763
+ "loss": 1.2106,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.28,
768
+ "learning_rate": 2.7194300509827035e-05,
769
+ "loss": 1.1902,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.28,
774
+ "learning_rate": 2.7127160621385888e-05,
775
+ "loss": 1.2188,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.28,
780
+ "learning_rate": 2.7059311833759813e-05,
781
+ "loss": 1.1779,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.28,
786
+ "learning_rate": 2.6990758113083318e-05,
787
+ "loss": 1.1895,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.29,
792
+ "learning_rate": 2.692150346669812e-05,
793
+ "loss": 1.1877,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.29,
798
+ "learning_rate": 2.6851551942918905e-05,
799
+ "loss": 1.1375,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.29,
804
+ "learning_rate": 2.6780907630796663e-05,
805
+ "loss": 1.2363,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.29,
810
+ "learning_rate": 2.6709574659879674e-05,
811
+ "loss": 1.2,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.3,
816
+ "learning_rate": 2.6637557199972115e-05,
817
+ "loss": 1.1543,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.3,
822
+ "learning_rate": 2.656485946089029e-05,
823
+ "loss": 1.1622,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.3,
828
+ "learning_rate": 2.6491485692216576e-05,
829
+ "loss": 1.1567,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.3,
834
+ "learning_rate": 2.6417440183050987e-05,
835
+ "loss": 1.1625,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.3,
840
+ "learning_rate": 2.6342727261760464e-05,
841
+ "loss": 1.1884,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.31,
846
+ "learning_rate": 2.626735129572585e-05,
847
+ "loss": 1.1959,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.31,
852
+ "learning_rate": 2.61913166910866e-05,
853
+ "loss": 1.1931,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.31,
858
+ "learning_rate": 2.6114627892483217e-05,
859
+ "loss": 1.2012,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.31,
864
+ "learning_rate": 2.603728938279743e-05,
865
+ "loss": 1.1935,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.32,
870
+ "learning_rate": 2.5959305682890154e-05,
871
+ "loss": 1.1837,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.32,
876
+ "learning_rate": 2.5880681351337224e-05,
877
+ "loss": 1.1224,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.32,
882
+ "learning_rate": 2.5801420984162914e-05,
883
+ "loss": 1.1786,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.32,
888
+ "learning_rate": 2.572152921457126e-05,
889
+ "loss": 1.2358,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.32,
894
+ "learning_rate": 2.5641010712675268e-05,
895
+ "loss": 1.2231,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.33,
900
+ "learning_rate": 2.5559870185223863e-05,
901
+ "loss": 1.1609,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.33,
906
+ "learning_rate": 2.5478112375326796e-05,
907
+ "loss": 1.1832,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.33,
912
+ "learning_rate": 2.5395742062177375e-05,
913
+ "loss": 1.1918,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.33,
918
+ "learning_rate": 2.5312764060773073e-05,
919
+ "loss": 1.1982,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.34,
924
+ "learning_rate": 2.5229183221634098e-05,
925
+ "loss": 1.1328,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.34,
930
+ "learning_rate": 2.514500443051982e-05,
931
+ "loss": 1.2386,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.34,
936
+ "learning_rate": 2.5060232608143207e-05,
937
+ "loss": 1.2092,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.34,
942
+ "learning_rate": 2.4974872709883138e-05,
943
+ "loss": 1.138,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.34,
948
+ "learning_rate": 2.488892972549479e-05,
949
+ "loss": 1.1936,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.35,
954
+ "learning_rate": 2.48024086788179e-05,
955
+ "loss": 1.1598,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.35,
960
+ "learning_rate": 2.471531462748315e-05,
961
+ "loss": 1.1555,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.35,
966
+ "learning_rate": 2.4627652662616463e-05,
967
+ "loss": 1.1645,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.35,
972
+ "learning_rate": 2.4539427908541462e-05,
973
+ "loss": 1.154,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.35,
978
+ "learning_rate": 2.4450645522479855e-05,
979
+ "loss": 1.1773,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.36,
984
+ "learning_rate": 2.4361310694250044e-05,
985
+ "loss": 1.1095,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.36,
990
+ "learning_rate": 2.4271428645963676e-05,
991
+ "loss": 1.183,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.36,
996
+ "learning_rate": 2.4181004631720444e-05,
997
+ "loss": 1.1415,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.36,
1002
+ "learning_rate": 2.409004393730091e-05,
1003
+ "loss": 1.1263,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.37,
1008
+ "learning_rate": 2.3998551879857554e-05,
1009
+ "loss": 1.1323,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.37,
1014
+ "learning_rate": 2.3906533807603952e-05,
1015
+ "loss": 1.1679,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.37,
1020
+ "learning_rate": 2.3813995099502108e-05,
1021
+ "loss": 1.2309,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.37,
1026
+ "learning_rate": 2.372094116494809e-05,
1027
+ "loss": 1.2011,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.37,
1032
+ "learning_rate": 2.3627377443455745e-05,
1033
+ "loss": 1.1399,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.38,
1038
+ "learning_rate": 2.3533309404338802e-05,
1039
+ "loss": 1.1185,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.38,
1044
+ "learning_rate": 2.343874254639111e-05,
1045
+ "loss": 1.1908,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.38,
1050
+ "learning_rate": 2.3343682397565216e-05,
1051
+ "loss": 1.1455,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.38,
1056
+ "learning_rate": 2.324813451464924e-05,
1057
+ "loss": 1.1138,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.39,
1062
+ "learning_rate": 2.3152104482942038e-05,
1063
+ "loss": 1.1683,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.39,
1068
+ "learning_rate": 2.3055597915926706e-05,
1069
+ "loss": 1.1528,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.39,
1074
+ "learning_rate": 2.295862045494246e-05,
1075
+ "loss": 1.1433,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.39,
1080
+ "learning_rate": 2.2861177768854833e-05,
1081
+ "loss": 1.2187,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.39,
1086
+ "learning_rate": 2.2763275553724355e-05,
1087
+ "loss": 1.1773,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.4,
1092
+ "learning_rate": 2.2664919532473526e-05,
1093
+ "loss": 1.1288,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.4,
1098
+ "learning_rate": 2.2566115454552325e-05,
1099
+ "loss": 1.1872,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.4,
1104
+ "learning_rate": 2.2466869095602093e-05,
1105
+ "loss": 1.187,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.4,
1110
+ "learning_rate": 2.2367186257117935e-05,
1111
+ "loss": 1.1506,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.41,
1116
+ "learning_rate": 2.2267072766109584e-05,
1117
+ "loss": 1.1548,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.41,
1122
+ "learning_rate": 2.2166534474760776e-05,
1123
+ "loss": 1.1747,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.41,
1128
+ "learning_rate": 2.206557726008717e-05,
1129
+ "loss": 1.1174,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.41,
1134
+ "learning_rate": 2.196420702359278e-05,
1135
+ "loss": 1.2132,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.41,
1140
+ "learning_rate": 2.1862429690925028e-05,
1141
+ "loss": 1.1553,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.42,
1146
+ "learning_rate": 2.176025121152834e-05,
1147
+ "loss": 1.1164,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.42,
1152
+ "learning_rate": 2.1657677558296378e-05,
1153
+ "loss": 1.137,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.42,
1158
+ "learning_rate": 2.155471472722288e-05,
1159
+ "loss": 1.1301,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.42,
1164
+ "learning_rate": 2.145136873705116e-05,
1165
+ "loss": 1.1698,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.42,
1170
+ "learning_rate": 2.134764562892231e-05,
1171
+ "loss": 1.156,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.43,
1176
+ "learning_rate": 2.1243551466022024e-05,
1177
+ "loss": 1.1195,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.43,
1182
+ "learning_rate": 2.113909233322618e-05,
1183
+ "loss": 1.161,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.43,
1188
+ "learning_rate": 2.103427433674517e-05,
1189
+ "loss": 1.1478,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.43,
1194
+ "learning_rate": 2.092910360376693e-05,
1195
+ "loss": 1.1472,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.44,
1200
+ "learning_rate": 2.0823586282098773e-05,
1201
+ "loss": 1.2036,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.44,
1206
+ "learning_rate": 2.0717728539808053e-05,
1207
+ "loss": 1.1643,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.44,
1212
+ "learning_rate": 2.0611536564861552e-05,
1213
+ "loss": 1.16,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.44,
1218
+ "learning_rate": 2.0505016564763803e-05,
1219
+ "loss": 1.1677,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.44,
1224
+ "learning_rate": 2.039817476619421e-05,
1225
+ "loss": 1.1765,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.45,
1230
+ "learning_rate": 2.029101741464306e-05,
1231
+ "loss": 1.1864,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.45,
1236
+ "learning_rate": 2.0183550774046454e-05,
1237
+ "loss": 1.1602,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.45,
1242
+ "learning_rate": 2.007578112642013e-05,
1243
+ "loss": 1.1733,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.45,
1248
+ "learning_rate": 1.9967714771492266e-05,
1249
+ "loss": 1.1544,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.46,
1254
+ "learning_rate": 1.9859358026335186e-05,
1255
+ "loss": 1.1212,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.46,
1260
+ "learning_rate": 1.9750717224996153e-05,
1261
+ "loss": 1.1324,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.46,
1266
+ "learning_rate": 1.964179871812704e-05,
1267
+ "loss": 1.1565,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.46,
1272
+ "learning_rate": 1.953260887261316e-05,
1273
+ "loss": 1.1682,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.46,
1278
+ "learning_rate": 1.9423154071201056e-05,
1279
+ "loss": 1.1407,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.47,
1284
+ "learning_rate": 1.931344071212539e-05,
1285
+ "loss": 1.1275,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.47,
1290
+ "learning_rate": 1.920347520873494e-05,
1291
+ "loss": 1.1789,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.47,
1296
+ "learning_rate": 1.909326398911772e-05,
1297
+ "loss": 1.1529,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.47,
1302
+ "learning_rate": 1.8982813495725202e-05,
1303
+ "loss": 1.1587,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.48,
1308
+ "learning_rate": 1.8872130184995713e-05,
1309
+ "loss": 1.1519,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.48,
1314
+ "learning_rate": 1.8761220526977046e-05,
1315
+ "loss": 1.1309,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.48,
1320
+ "learning_rate": 1.865009100494823e-05,
1321
+ "loss": 1.1494,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.48,
1326
+ "learning_rate": 1.8538748115040556e-05,
1327
+ "loss": 1.1685,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.48,
1332
+ "learning_rate": 1.8427198365857846e-05,
1333
+ "loss": 1.1581,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.49,
1338
+ "learning_rate": 1.8315448278095976e-05,
1339
+ "loss": 1.167,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.49,
1344
+ "learning_rate": 1.820350438416171e-05,
1345
+ "loss": 1.1376,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.49,
1350
+ "learning_rate": 1.809137322779086e-05,
1351
+ "loss": 1.1487,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.49,
1356
+ "learning_rate": 1.797906136366574e-05,
1357
+ "loss": 1.1313,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.49,
1362
+ "learning_rate": 1.7866575357032042e-05,
1363
+ "loss": 1.2109,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.5,
1368
+ "learning_rate": 1.7753921783315033e-05,
1369
+ "loss": 1.162,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.5,
1374
+ "learning_rate": 1.764110722773521e-05,
1375
+ "loss": 1.1593,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.5,
1380
+ "learning_rate": 1.7528138284923326e-05,
1381
+ "loss": 1.177,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.5,
1386
+ "learning_rate": 1.741502155853493e-05,
1387
+ "loss": 1.1137,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.51,
1392
+ "learning_rate": 1.7301763660864328e-05,
1393
+ "loss": 1.1503,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.51,
1398
+ "learning_rate": 1.7188371212458063e-05,
1399
+ "loss": 1.1725,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.51,
1404
+ "learning_rate": 1.7074850841727914e-05,
1405
+ "loss": 1.1668,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.51,
1410
+ "learning_rate": 1.696120918456341e-05,
1411
+ "loss": 1.1573,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.51,
1416
+ "learning_rate": 1.684745288394396e-05,
1417
+ "loss": 1.1691,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.52,
1422
+ "learning_rate": 1.673358858955049e-05,
1423
+ "loss": 1.1176,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.52,
1428
+ "learning_rate": 1.6619622957376762e-05,
1429
+ "loss": 1.1469,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.52,
1434
+ "learning_rate": 1.65055626493403e-05,
1435
+ "loss": 1.1049,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.52,
1440
+ "learning_rate": 1.6391414332892942e-05,
1441
+ "loss": 1.2469,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.53,
1446
+ "learning_rate": 1.6277184680631102e-05,
1447
+ "loss": 1.1681,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.53,
1452
+ "learning_rate": 1.616288036990572e-05,
1453
+ "loss": 1.131,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.53,
1458
+ "learning_rate": 1.6048508082431938e-05,
1459
+ "loss": 1.1285,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.53,
1464
+ "learning_rate": 1.5934074503898504e-05,
1465
+ "loss": 1.1555,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.53,
1470
+ "learning_rate": 1.5819586323576972e-05,
1471
+ "loss": 1.1507,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.54,
1476
+ "learning_rate": 1.5705050233930672e-05,
1477
+ "loss": 1.1126,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.54,
1482
+ "learning_rate": 1.5590472930223484e-05,
1483
+ "loss": 1.166,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.54,
1488
+ "learning_rate": 1.54758611101285e-05,
1489
+ "loss": 1.1631,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.54,
1494
+ "learning_rate": 1.536122147333646e-05,
1495
+ "loss": 1.1394,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.55,
1500
+ "learning_rate": 1.5246560721164158e-05,
1501
+ "loss": 1.1729,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.55,
1506
+ "learning_rate": 1.5131885556162693e-05,
1507
+ "loss": 1.1309,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.55,
1512
+ "learning_rate": 1.5017202681725672e-05,
1513
+ "loss": 1.1462,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.55,
1518
+ "learning_rate": 1.4902518801697365e-05,
1519
+ "loss": 1.1553,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.55,
1524
+ "learning_rate": 1.4787840619980825e-05,
1525
+ "loss": 1.153,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.56,
1530
+ "learning_rate": 1.4673174840146e-05,
1531
+ "loss": 1.1712,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.56,
1536
+ "learning_rate": 1.455852816503788e-05,
1537
+ "loss": 1.1652,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.56,
1542
+ "learning_rate": 1.4443907296384687e-05,
1543
+ "loss": 1.1061,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.56,
1548
+ "learning_rate": 1.4329318934406113e-05,
1549
+ "loss": 1.1369,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.57,
1554
+ "learning_rate": 1.4214769777421652e-05,
1555
+ "loss": 1.1275,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.57,
1560
+ "learning_rate": 1.4100266521459044e-05,
1561
+ "loss": 1.1134,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.57,
1566
+ "learning_rate": 1.3985815859862881e-05,
1567
+ "loss": 1.1047,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.57,
1572
+ "learning_rate": 1.3871424482903308e-05,
1573
+ "loss": 1.0891,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.57,
1578
+ "learning_rate": 1.3757099077384965e-05,
1579
+ "loss": 1.1455,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.58,
1584
+ "learning_rate": 1.3642846326256088e-05,
1585
+ "loss": 1.1312,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.58,
1590
+ "learning_rate": 1.3528672908217888e-05,
1591
+ "loss": 1.1389,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.58,
1596
+ "learning_rate": 1.34145854973341e-05,
1597
+ "loss": 1.1836,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.58,
1602
+ "learning_rate": 1.3300590762640868e-05,
1603
+ "loss": 1.0937,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.58,
1608
+ "learning_rate": 1.3186695367756916e-05,
1609
+ "loss": 1.1978,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.59,
1614
+ "learning_rate": 1.3072905970494003e-05,
1615
+ "loss": 1.151,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.59,
1620
+ "learning_rate": 1.295922922246775e-05,
1621
+ "loss": 1.1011,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.59,
1626
+ "learning_rate": 1.28456717687088e-05,
1627
+ "loss": 1.0861,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.59,
1632
+ "learning_rate": 1.2732240247274395e-05,
1633
+ "loss": 1.1113,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.6,
1638
+ "learning_rate": 1.2618941288860352e-05,
1639
+ "loss": 1.1614,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.6,
1644
+ "learning_rate": 1.250578151641344e-05,
1645
+ "loss": 1.1503,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.6,
1650
+ "learning_rate": 1.239276754474423e-05,
1651
+ "loss": 1.1446,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.6,
1656
+ "learning_rate": 1.2279905980140473e-05,
1657
+ "loss": 1.1023,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.6,
1662
+ "learning_rate": 1.216720341998085e-05,
1663
+ "loss": 1.1615,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.61,
1668
+ "learning_rate": 1.2054666452349389e-05,
1669
+ "loss": 1.0865,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.61,
1674
+ "learning_rate": 1.19423016556503e-05,
1675
+ "loss": 1.127,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.61,
1680
+ "learning_rate": 1.1830115598223478e-05,
1681
+ "loss": 1.2051,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.61,
1686
+ "learning_rate": 1.171811483796051e-05,
1687
+ "loss": 1.0966,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.62,
1692
+ "learning_rate": 1.1606305921921349e-05,
1693
+ "loss": 1.1681,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.62,
1698
+ "learning_rate": 1.1494695385951596e-05,
1699
+ "loss": 1.201,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.62,
1704
+ "learning_rate": 1.1383289754300463e-05,
1705
+ "loss": 1.1575,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.62,
1710
+ "learning_rate": 1.1272095539239358e-05,
1711
+ "loss": 1.1316,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.62,
1716
+ "learning_rate": 1.1161119240681233e-05,
1717
+ "loss": 1.151,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.63,
1722
+ "learning_rate": 1.1050367345800628e-05,
1723
+ "loss": 1.1619,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.63,
1728
+ "learning_rate": 1.0939846328654456e-05,
1729
+ "loss": 1.1264,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.63,
1734
+ "learning_rate": 1.0829562649803551e-05,
1735
+ "loss": 1.1052,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.63,
1740
+ "learning_rate": 1.0719522755935018e-05,
1741
+ "loss": 1.1075,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.64,
1746
+ "learning_rate": 1.0609733079485404e-05,
1747
+ "loss": 1.1597,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.64,
1752
+ "learning_rate": 1.0500200038264657e-05,
1753
+ "loss": 1.1135,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.64,
1758
+ "learning_rate": 1.0390930035080994e-05,
1759
+ "loss": 1.1193,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.64,
1764
+ "learning_rate": 1.0281929457366599e-05,
1765
+ "loss": 1.1912,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.64,
1770
+ "learning_rate": 1.0173204676804284e-05,
1771
+ "loss": 1.1432,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.65,
1776
+ "learning_rate": 1.0064762048954982e-05,
1777
+ "loss": 1.172,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.65,
1782
+ "learning_rate": 9.956607912886247e-06,
1783
+ "loss": 1.1299,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.65,
1788
+ "learning_rate": 9.848748590801712e-06,
1789
+ "loss": 1.0882,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.65,
1794
+ "learning_rate": 9.741190387671512e-06,
1795
+ "loss": 1.1207,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.65,
1800
+ "learning_rate": 9.633939590863725e-06,
1801
+ "loss": 1.1749,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.66,
1806
+ "learning_rate": 9.527002469776827e-06,
1807
+ "loss": 1.0907,
1808
+ "step": 3000
1809
+ },
1810
+ {
1811
+ "epoch": 0.66,
1812
+ "learning_rate": 9.420385275473244e-06,
1813
+ "loss": 1.1059,
1814
+ "step": 3010
1815
+ },
1816
+ {
1817
+ "epoch": 0.66,
1818
+ "learning_rate": 9.314094240313913e-06,
1819
+ "loss": 1.1428,
1820
+ "step": 3020
1821
+ },
1822
+ {
1823
+ "epoch": 0.66,
1824
+ "learning_rate": 9.208135577593974e-06,
1825
+ "loss": 1.1886,
1826
+ "step": 3030
1827
+ },
1828
+ {
1829
+ "epoch": 0.67,
1830
+ "learning_rate": 9.102515481179567e-06,
1831
+ "loss": 1.1088,
1832
+ "step": 3040
1833
+ },
1834
+ {
1835
+ "epoch": 0.67,
1836
+ "learning_rate": 8.997240125145793e-06,
1837
+ "loss": 1.1211,
1838
+ "step": 3050
1839
+ },
1840
+ {
1841
+ "epoch": 0.67,
1842
+ "learning_rate": 8.892315663415763e-06,
1843
+ "loss": 1.0975,
1844
+ "step": 3060
1845
+ },
1846
+ {
1847
+ "epoch": 0.67,
1848
+ "learning_rate": 8.787748229400886e-06,
1849
+ "loss": 1.1041,
1850
+ "step": 3070
1851
+ },
1852
+ {
1853
+ "epoch": 0.67,
1854
+ "learning_rate": 8.683543935642351e-06,
1855
+ "loss": 1.0865,
1856
+ "step": 3080
1857
+ },
1858
+ {
1859
+ "epoch": 0.68,
1860
+ "learning_rate": 8.579708873453808e-06,
1861
+ "loss": 1.1124,
1862
+ "step": 3090
1863
+ },
1864
+ {
1865
+ "epoch": 0.68,
1866
+ "learning_rate": 8.476249112565269e-06,
1867
+ "loss": 1.1762,
1868
+ "step": 3100
1869
+ },
1870
+ {
1871
+ "epoch": 0.68,
1872
+ "learning_rate": 8.37317070076834e-06,
1873
+ "loss": 1.1522,
1874
+ "step": 3110
1875
+ },
1876
+ {
1877
+ "epoch": 0.68,
1878
+ "learning_rate": 8.270479663562674e-06,
1879
+ "loss": 1.1455,
1880
+ "step": 3120
1881
+ },
1882
+ {
1883
+ "epoch": 0.69,
1884
+ "learning_rate": 8.168182003803752e-06,
1885
+ "loss": 1.0856,
1886
+ "step": 3130
1887
+ },
1888
+ {
1889
+ "epoch": 0.69,
1890
+ "learning_rate": 8.066283701351982e-06,
1891
+ "loss": 1.1308,
1892
+ "step": 3140
1893
+ },
1894
+ {
1895
+ "epoch": 0.69,
1896
+ "learning_rate": 7.964790712723124e-06,
1897
+ "loss": 1.1232,
1898
+ "step": 3150
1899
+ },
1900
+ {
1901
+ "epoch": 0.69,
1902
+ "learning_rate": 7.863708970740155e-06,
1903
+ "loss": 1.066,
1904
+ "step": 3160
1905
+ },
1906
+ {
1907
+ "epoch": 0.69,
1908
+ "learning_rate": 7.763044384186388e-06,
1909
+ "loss": 1.1462,
1910
+ "step": 3170
1911
+ },
1912
+ {
1913
+ "epoch": 0.7,
1914
+ "learning_rate": 7.662802837460131e-06,
1915
+ "loss": 1.1295,
1916
+ "step": 3180
1917
+ },
1918
+ {
1919
+ "epoch": 0.7,
1920
+ "learning_rate": 7.5629901902306696e-06,
1921
+ "loss": 1.1225,
1922
+ "step": 3190
1923
+ },
1924
+ {
1925
+ "epoch": 0.7,
1926
+ "learning_rate": 7.463612277095793e-06,
1927
+ "loss": 1.1331,
1928
+ "step": 3200
1929
+ },
1930
+ {
1931
+ "epoch": 0.7,
1932
+ "learning_rate": 7.364674907240651e-06,
1933
+ "loss": 1.1902,
1934
+ "step": 3210
1935
+ },
1936
+ {
1937
+ "epoch": 0.71,
1938
+ "learning_rate": 7.2661838640982405e-06,
1939
+ "loss": 1.1444,
1940
+ "step": 3220
1941
+ },
1942
+ {
1943
+ "epoch": 0.71,
1944
+ "learning_rate": 7.16814490501131e-06,
1945
+ "loss": 1.1374,
1946
+ "step": 3230
1947
+ },
1948
+ {
1949
+ "epoch": 0.71,
1950
+ "learning_rate": 7.070563760895802e-06,
1951
+ "loss": 1.1353,
1952
+ "step": 3240
1953
+ },
1954
+ {
1955
+ "epoch": 0.71,
1956
+ "learning_rate": 6.97344613590587e-06,
1957
+ "loss": 1.1184,
1958
+ "step": 3250
1959
+ },
1960
+ {
1961
+ "epoch": 0.71,
1962
+ "learning_rate": 6.876797707100399e-06,
1963
+ "loss": 1.1029,
1964
+ "step": 3260
1965
+ },
1966
+ {
1967
+ "epoch": 0.72,
1968
+ "learning_rate": 6.780624124111224e-06,
1969
+ "loss": 1.1621,
1970
+ "step": 3270
1971
+ },
1972
+ {
1973
+ "epoch": 0.72,
1974
+ "learning_rate": 6.684931008812788e-06,
1975
+ "loss": 1.1299,
1976
+ "step": 3280
1977
+ },
1978
+ {
1979
+ "epoch": 0.72,
1980
+ "learning_rate": 6.589723954993589e-06,
1981
+ "loss": 1.0911,
1982
+ "step": 3290
1983
+ },
1984
+ {
1985
+ "epoch": 0.72,
1986
+ "learning_rate": 6.495008528029127e-06,
1987
+ "loss": 1.1545,
1988
+ "step": 3300
1989
+ },
1990
+ {
1991
+ "epoch": 0.72,
1992
+ "learning_rate": 6.4007902645566565e-06,
1993
+ "loss": 1.1036,
1994
+ "step": 3310
1995
+ },
1996
+ {
1997
+ "epoch": 0.73,
1998
+ "learning_rate": 6.30707467215145e-06,
1999
+ "loss": 1.101,
2000
+ "step": 3320
2001
+ },
2002
+ {
2003
+ "epoch": 0.73,
2004
+ "learning_rate": 6.213867229004918e-06,
2005
+ "loss": 1.1241,
2006
+ "step": 3330
2007
+ },
2008
+ {
2009
+ "epoch": 0.73,
2010
+ "learning_rate": 6.121173383604346e-06,
2011
+ "loss": 1.1364,
2012
+ "step": 3340
2013
+ },
2014
+ {
2015
+ "epoch": 0.73,
2016
+ "learning_rate": 6.028998554414411e-06,
2017
+ "loss": 1.1363,
2018
+ "step": 3350
2019
+ },
2020
+ {
2021
+ "epoch": 0.74,
2022
+ "learning_rate": 5.9373481295604464e-06,
2023
+ "loss": 1.1153,
2024
+ "step": 3360
2025
+ },
2026
+ {
2027
+ "epoch": 0.74,
2028
+ "learning_rate": 5.8462274665134505e-06,
2029
+ "loss": 1.0876,
2030
+ "step": 3370
2031
+ },
2032
+ {
2033
+ "epoch": 0.74,
2034
+ "learning_rate": 5.755641891776954e-06,
2035
+ "loss": 1.1324,
2036
+ "step": 3380
2037
+ },
2038
+ {
2039
+ "epoch": 0.74,
2040
+ "learning_rate": 5.665596700575622e-06,
2041
+ "loss": 1.0916,
2042
+ "step": 3390
2043
+ },
2044
+ {
2045
+ "epoch": 0.74,
2046
+ "learning_rate": 5.576097156545748e-06,
2047
+ "loss": 1.096,
2048
+ "step": 3400
2049
+ },
2050
+ {
2051
+ "epoch": 0.75,
2052
+ "learning_rate": 5.487148491427516e-06,
2053
+ "loss": 1.1398,
2054
+ "step": 3410
2055
+ },
2056
+ {
2057
+ "epoch": 0.75,
2058
+ "learning_rate": 5.39875590475926e-06,
2059
+ "loss": 1.1753,
2060
+ "step": 3420
2061
+ },
2062
+ {
2063
+ "epoch": 0.75,
2064
+ "learning_rate": 5.310924563573426e-06,
2065
+ "loss": 1.1579,
2066
+ "step": 3430
2067
+ },
2068
+ {
2069
+ "epoch": 0.75,
2070
+ "learning_rate": 5.223659602094603e-06,
2071
+ "loss": 1.1328,
2072
+ "step": 3440
2073
+ },
2074
+ {
2075
+ "epoch": 0.76,
2076
+ "learning_rate": 5.136966121439367e-06,
2077
+ "loss": 1.1426,
2078
+ "step": 3450
2079
+ },
2080
+ {
2081
+ "epoch": 0.76,
2082
+ "learning_rate": 5.0508491893180994e-06,
2083
+ "loss": 1.1243,
2084
+ "step": 3460
2085
+ },
2086
+ {
2087
+ "epoch": 0.76,
2088
+ "learning_rate": 4.965313839738755e-06,
2089
+ "loss": 1.1673,
2090
+ "step": 3470
2091
+ },
2092
+ {
2093
+ "epoch": 0.76,
2094
+ "learning_rate": 4.880365072712579e-06,
2095
+ "loss": 1.142,
2096
+ "step": 3480
2097
+ },
2098
+ {
2099
+ "epoch": 0.76,
2100
+ "learning_rate": 4.7960078539618575e-06,
2101
+ "loss": 1.1312,
2102
+ "step": 3490
2103
+ },
2104
+ {
2105
+ "epoch": 0.77,
2106
+ "learning_rate": 4.712247114629627e-06,
2107
+ "loss": 1.1428,
2108
+ "step": 3500
2109
+ },
2110
+ {
2111
+ "epoch": 0.77,
2112
+ "learning_rate": 4.629087750991422e-06,
2113
+ "loss": 1.1255,
2114
+ "step": 3510
2115
+ },
2116
+ {
2117
+ "epoch": 0.77,
2118
+ "learning_rate": 4.546534624169054e-06,
2119
+ "loss": 1.1348,
2120
+ "step": 3520
2121
+ },
2122
+ {
2123
+ "epoch": 0.77,
2124
+ "learning_rate": 4.464592559846474e-06,
2125
+ "loss": 1.1474,
2126
+ "step": 3530
2127
+ },
2128
+ {
2129
+ "epoch": 0.78,
2130
+ "learning_rate": 4.383266347987664e-06,
2131
+ "loss": 1.0874,
2132
+ "step": 3540
2133
+ },
2134
+ {
2135
+ "epoch": 0.78,
2136
+ "learning_rate": 4.302560742556651e-06,
2137
+ "loss": 1.1492,
2138
+ "step": 3550
2139
+ },
2140
+ {
2141
+ "epoch": 0.78,
2142
+ "learning_rate": 4.2224804612395985e-06,
2143
+ "loss": 1.1539,
2144
+ "step": 3560
2145
+ },
2146
+ {
2147
+ "epoch": 0.78,
2148
+ "learning_rate": 4.143030185169045e-06,
2149
+ "loss": 1.1288,
2150
+ "step": 3570
2151
+ },
2152
+ {
2153
+ "epoch": 0.78,
2154
+ "learning_rate": 4.064214558650262e-06,
2155
+ "loss": 1.1366,
2156
+ "step": 3580
2157
+ },
2158
+ {
2159
+ "epoch": 0.79,
2160
+ "learning_rate": 3.986038188889751e-06,
2161
+ "loss": 1.141,
2162
+ "step": 3590
2163
+ },
2164
+ {
2165
+ "epoch": 0.79,
2166
+ "learning_rate": 3.908505645725963e-06,
2167
+ "loss": 1.1404,
2168
+ "step": 3600
2169
+ },
2170
+ {
2171
+ "epoch": 0.79,
2172
+ "learning_rate": 3.831621461362134e-06,
2173
+ "loss": 1.115,
2174
+ "step": 3610
2175
+ },
2176
+ {
2177
+ "epoch": 0.79,
2178
+ "learning_rate": 3.7553901301013687e-06,
2179
+ "loss": 1.0857,
2180
+ "step": 3620
2181
+ },
2182
+ {
2183
+ "epoch": 0.8,
2184
+ "learning_rate": 3.6798161080839265e-06,
2185
+ "loss": 1.1158,
2186
+ "step": 3630
2187
+ },
2188
+ {
2189
+ "epoch": 0.8,
2190
+ "learning_rate": 3.604903813026713e-06,
2191
+ "loss": 1.1022,
2192
+ "step": 3640
2193
+ },
2194
+ {
2195
+ "epoch": 0.8,
2196
+ "learning_rate": 3.530657623965068e-06,
2197
+ "loss": 1.1388,
2198
+ "step": 3650
2199
+ },
2200
+ {
2201
+ "epoch": 0.8,
2202
+ "learning_rate": 3.4570818809967768e-06,
2203
+ "loss": 1.1296,
2204
+ "step": 3660
2205
+ },
2206
+ {
2207
+ "epoch": 0.8,
2208
+ "learning_rate": 3.3841808850283597e-06,
2209
+ "loss": 1.1591,
2210
+ "step": 3670
2211
+ },
2212
+ {
2213
+ "epoch": 0.81,
2214
+ "learning_rate": 3.311958897523671e-06,
2215
+ "loss": 1.1397,
2216
+ "step": 3680
2217
+ },
2218
+ {
2219
+ "epoch": 0.81,
2220
+ "learning_rate": 3.2404201402547893e-06,
2221
+ "loss": 1.1468,
2222
+ "step": 3690
2223
+ },
2224
+ {
2225
+ "epoch": 0.81,
2226
+ "learning_rate": 3.1695687950552216e-06,
2227
+ "loss": 1.0949,
2228
+ "step": 3700
2229
+ },
2230
+ {
2231
+ "epoch": 0.81,
2232
+ "learning_rate": 3.099409003575473e-06,
2233
+ "loss": 1.1348,
2234
+ "step": 3710
2235
+ },
2236
+ {
2237
+ "epoch": 0.81,
2238
+ "learning_rate": 3.029944867040928e-06,
2239
+ "loss": 1.1241,
2240
+ "step": 3720
2241
+ },
2242
+ {
2243
+ "epoch": 0.82,
2244
+ "learning_rate": 2.9611804460121195e-06,
2245
+ "loss": 1.1145,
2246
+ "step": 3730
2247
+ },
2248
+ {
2249
+ "epoch": 0.82,
2250
+ "learning_rate": 2.8931197601473634e-06,
2251
+ "loss": 1.11,
2252
+ "step": 3740
2253
+ },
2254
+ {
2255
+ "epoch": 0.82,
2256
+ "learning_rate": 2.8257667879677762e-06,
2257
+ "loss": 1.138,
2258
+ "step": 3750
2259
+ },
2260
+ {
2261
+ "epoch": 0.82,
2262
+ "learning_rate": 2.7591254666247455e-06,
2263
+ "loss": 1.1738,
2264
+ "step": 3760
2265
+ },
2266
+ {
2267
+ "epoch": 0.83,
2268
+ "learning_rate": 2.6931996916697284e-06,
2269
+ "loss": 1.1612,
2270
+ "step": 3770
2271
+ },
2272
+ {
2273
+ "epoch": 0.83,
2274
+ "learning_rate": 2.627993316826584e-06,
2275
+ "loss": 1.144,
2276
+ "step": 3780
2277
+ },
2278
+ {
2279
+ "epoch": 0.83,
2280
+ "learning_rate": 2.563510153766262e-06,
2281
+ "loss": 1.1366,
2282
+ "step": 3790
2283
+ },
2284
+ {
2285
+ "epoch": 0.83,
2286
+ "learning_rate": 2.499753971884036e-06,
2287
+ "loss": 1.1093,
2288
+ "step": 3800
2289
+ },
2290
+ {
2291
+ "epoch": 0.83,
2292
+ "learning_rate": 2.4367284980791046e-06,
2293
+ "loss": 1.1113,
2294
+ "step": 3810
2295
+ },
2296
+ {
2297
+ "epoch": 0.84,
2298
+ "learning_rate": 2.3744374165367833e-06,
2299
+ "loss": 1.1389,
2300
+ "step": 3820
2301
+ },
2302
+ {
2303
+ "epoch": 0.84,
2304
+ "learning_rate": 2.3128843685131194e-06,
2305
+ "loss": 1.0978,
2306
+ "step": 3830
2307
+ },
2308
+ {
2309
+ "epoch": 0.84,
2310
+ "learning_rate": 2.2520729521220402e-06,
2311
+ "loss": 1.0949,
2312
+ "step": 3840
2313
+ },
2314
+ {
2315
+ "epoch": 0.84,
2316
+ "learning_rate": 2.1920067221250385e-06,
2317
+ "loss": 1.1312,
2318
+ "step": 3850
2319
+ },
2320
+ {
2321
+ "epoch": 0.85,
2322
+ "learning_rate": 2.132689189723349e-06,
2323
+ "loss": 1.169,
2324
+ "step": 3860
2325
+ },
2326
+ {
2327
+ "epoch": 0.85,
2328
+ "learning_rate": 2.0741238223527426e-06,
2329
+ "loss": 1.1062,
2330
+ "step": 3870
2331
+ },
2332
+ {
2333
+ "epoch": 0.85,
2334
+ "learning_rate": 2.0163140434807864e-06,
2335
+ "loss": 1.1209,
2336
+ "step": 3880
2337
+ },
2338
+ {
2339
+ "epoch": 0.85,
2340
+ "learning_rate": 1.959263232406766e-06,
2341
+ "loss": 1.1179,
2342
+ "step": 3890
2343
+ },
2344
+ {
2345
+ "epoch": 0.85,
2346
+ "learning_rate": 1.9029747240641077e-06,
2347
+ "loss": 1.1256,
2348
+ "step": 3900
2349
+ },
2350
+ {
2351
+ "epoch": 0.86,
2352
+ "learning_rate": 1.847451808825479e-06,
2353
+ "loss": 1.1185,
2354
+ "step": 3910
2355
+ },
2356
+ {
2357
+ "epoch": 0.86,
2358
+ "learning_rate": 1.792697732310401e-06,
2359
+ "loss": 1.1355,
2360
+ "step": 3920
2361
+ },
2362
+ {
2363
+ "epoch": 0.86,
2364
+ "learning_rate": 1.7387156951955596e-06,
2365
+ "loss": 1.1461,
2366
+ "step": 3930
2367
+ },
2368
+ {
2369
+ "epoch": 0.86,
2370
+ "learning_rate": 1.6855088530276902e-06,
2371
+ "loss": 1.1348,
2372
+ "step": 3940
2373
+ },
2374
+ {
2375
+ "epoch": 0.87,
2376
+ "learning_rate": 1.6330803160391228e-06,
2377
+ "loss": 1.1323,
2378
+ "step": 3950
2379
+ },
2380
+ {
2381
+ "epoch": 0.87,
2382
+ "learning_rate": 1.5814331489659756e-06,
2383
+ "loss": 1.1788,
2384
+ "step": 3960
2385
+ },
2386
+ {
2387
+ "epoch": 0.87,
2388
+ "learning_rate": 1.53057037086899e-06,
2389
+ "loss": 1.1394,
2390
+ "step": 3970
2391
+ },
2392
+ {
2393
+ "epoch": 0.87,
2394
+ "learning_rate": 1.4804949549570818e-06,
2395
+ "loss": 1.1737,
2396
+ "step": 3980
2397
+ },
2398
+ {
2399
+ "epoch": 0.87,
2400
+ "learning_rate": 1.4312098284135e-06,
2401
+ "loss": 1.1482,
2402
+ "step": 3990
2403
+ },
2404
+ {
2405
+ "epoch": 0.88,
2406
+ "learning_rate": 1.3827178722247508e-06,
2407
+ "loss": 1.0941,
2408
+ "step": 4000
2409
+ }
2410
+ ],
2411
+ "max_steps": 4566,
2412
+ "num_train_epochs": 1,
2413
+ "total_flos": 0.0,
2414
+ "trial_name": null,
2415
+ "trial_params": null
2416
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_1x/training_complete.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_path": "/root/pixel_cloud/contextual_100_work/data/contextual/downstream/MSA_Dialect_Combined_1x_max64",
3
+ "dataset_rows": 584406,
4
+ "epochs": 1,
5
+ "global_step": 4566,
6
+ "input_checkpoint": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct100/pretraining/pixel/full_epoch2/final/pytorch_model.bin",
7
+ "learning_rate": 3e-05,
8
+ "metrics": {
9
+ "epoch": 1.0,
10
+ "train_loss": 1.1852035445181153,
11
+ "train_runtime": 1226.3263,
12
+ "train_samples_per_second": 476.55,
13
+ "train_steps_per_second": 3.723
14
+ },
15
+ "model_type": "pixel",
16
+ "scheduler": "cosine",
17
+ "seed": 42,
18
+ "warmup_ratio": 0.1
19
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_2x/checkpoint-5000/trainer_state.json ADDED
@@ -0,0 +1,3016 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.7300335815447511,
5
+ "global_step": 5000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.3795620437956206e-07,
13
+ "loss": 1.2883,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 8.759124087591241e-07,
19
+ "loss": 1.3105,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.0,
24
+ "learning_rate": 1.3138686131386862e-06,
25
+ "loss": 1.3972,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 1.7518248175182482e-06,
31
+ "loss": 1.2908,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 2.1897810218978103e-06,
37
+ "loss": 1.2576,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 2.6277372262773724e-06,
43
+ "loss": 1.2685,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.01,
48
+ "learning_rate": 3.0656934306569344e-06,
49
+ "loss": 1.3179,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.01,
54
+ "learning_rate": 3.5036496350364965e-06,
55
+ "loss": 1.289,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.01,
60
+ "learning_rate": 3.9416058394160585e-06,
61
+ "loss": 1.2187,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.01,
66
+ "learning_rate": 4.379562043795621e-06,
67
+ "loss": 1.2577,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 4.817518248175183e-06,
73
+ "loss": 1.2545,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.02,
78
+ "learning_rate": 5.255474452554745e-06,
79
+ "loss": 1.2564,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.02,
84
+ "learning_rate": 5.693430656934307e-06,
85
+ "loss": 1.3263,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.02,
90
+ "learning_rate": 6.131386861313869e-06,
91
+ "loss": 1.2394,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.02,
96
+ "learning_rate": 6.569343065693431e-06,
97
+ "loss": 1.2536,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.02,
102
+ "learning_rate": 7.007299270072993e-06,
103
+ "loss": 1.221,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.02,
108
+ "learning_rate": 7.445255474452555e-06,
109
+ "loss": 1.2291,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.03,
114
+ "learning_rate": 7.883211678832117e-06,
115
+ "loss": 1.256,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.03,
120
+ "learning_rate": 8.321167883211678e-06,
121
+ "loss": 1.2402,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.03,
126
+ "learning_rate": 8.759124087591241e-06,
127
+ "loss": 1.2466,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.03,
132
+ "learning_rate": 9.197080291970804e-06,
133
+ "loss": 1.2007,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.03,
138
+ "learning_rate": 9.635036496350365e-06,
139
+ "loss": 1.2417,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.03,
144
+ "learning_rate": 1.0072992700729927e-05,
145
+ "loss": 1.2404,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.04,
150
+ "learning_rate": 1.051094890510949e-05,
151
+ "loss": 1.2551,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.04,
156
+ "learning_rate": 1.094890510948905e-05,
157
+ "loss": 1.169,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.04,
162
+ "learning_rate": 1.1386861313868614e-05,
163
+ "loss": 1.1858,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.04,
168
+ "learning_rate": 1.1824817518248176e-05,
169
+ "loss": 1.1528,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.04,
174
+ "learning_rate": 1.2262773722627738e-05,
175
+ "loss": 1.1979,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.04,
180
+ "learning_rate": 1.2700729927007299e-05,
181
+ "loss": 1.2189,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.04,
186
+ "learning_rate": 1.3138686131386862e-05,
187
+ "loss": 1.177,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.05,
192
+ "learning_rate": 1.3576642335766423e-05,
193
+ "loss": 1.1802,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.05,
198
+ "learning_rate": 1.4014598540145986e-05,
199
+ "loss": 1.1951,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.05,
204
+ "learning_rate": 1.4452554744525549e-05,
205
+ "loss": 1.15,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.05,
210
+ "learning_rate": 1.489051094890511e-05,
211
+ "loss": 1.2175,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.05,
216
+ "learning_rate": 1.5328467153284673e-05,
217
+ "loss": 1.221,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.05,
222
+ "learning_rate": 1.5766423357664234e-05,
223
+ "loss": 1.1714,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.05,
228
+ "learning_rate": 1.6204379562043795e-05,
229
+ "loss": 1.1593,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.06,
234
+ "learning_rate": 1.6642335766423357e-05,
235
+ "loss": 1.1302,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.06,
240
+ "learning_rate": 1.7080291970802918e-05,
241
+ "loss": 1.1517,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.06,
246
+ "learning_rate": 1.7518248175182482e-05,
247
+ "loss": 1.1623,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.06,
252
+ "learning_rate": 1.7956204379562047e-05,
253
+ "loss": 1.1689,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.06,
258
+ "learning_rate": 1.8394160583941608e-05,
259
+ "loss": 1.1682,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.06,
264
+ "learning_rate": 1.883211678832117e-05,
265
+ "loss": 1.1459,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.06,
270
+ "learning_rate": 1.927007299270073e-05,
271
+ "loss": 1.183,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.07,
276
+ "learning_rate": 1.9708029197080292e-05,
277
+ "loss": 1.1587,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.07,
282
+ "learning_rate": 2.0145985401459853e-05,
283
+ "loss": 1.1588,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.07,
288
+ "learning_rate": 2.0583941605839418e-05,
289
+ "loss": 1.1047,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.07,
294
+ "learning_rate": 2.102189781021898e-05,
295
+ "loss": 1.1835,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.07,
300
+ "learning_rate": 2.145985401459854e-05,
301
+ "loss": 1.0676,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.07,
306
+ "learning_rate": 2.18978102189781e-05,
307
+ "loss": 1.135,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.07,
312
+ "learning_rate": 2.2335766423357663e-05,
313
+ "loss": 1.1218,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.08,
318
+ "learning_rate": 2.2773722627737227e-05,
319
+ "loss": 1.1111,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.08,
324
+ "learning_rate": 2.3211678832116792e-05,
325
+ "loss": 1.1311,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.08,
330
+ "learning_rate": 2.3649635036496353e-05,
331
+ "loss": 1.119,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.08,
336
+ "learning_rate": 2.4087591240875914e-05,
337
+ "loss": 1.1116,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.08,
342
+ "learning_rate": 2.4525547445255475e-05,
343
+ "loss": 1.1061,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.08,
348
+ "learning_rate": 2.4963503649635037e-05,
349
+ "loss": 1.0921,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.08,
354
+ "learning_rate": 2.5401459854014598e-05,
355
+ "loss": 1.1249,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.09,
360
+ "learning_rate": 2.583941605839416e-05,
361
+ "loss": 1.0942,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.09,
366
+ "learning_rate": 2.6277372262773724e-05,
367
+ "loss": 1.1066,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.09,
372
+ "learning_rate": 2.6715328467153285e-05,
373
+ "loss": 1.0921,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.09,
378
+ "learning_rate": 2.7153284671532846e-05,
379
+ "loss": 1.0947,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.09,
384
+ "learning_rate": 2.759124087591241e-05,
385
+ "loss": 1.0746,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.09,
390
+ "learning_rate": 2.8029197080291972e-05,
391
+ "loss": 1.1277,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.09,
396
+ "learning_rate": 2.8467153284671533e-05,
397
+ "loss": 1.1284,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.1,
402
+ "learning_rate": 2.8905109489051098e-05,
403
+ "loss": 1.0711,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.1,
408
+ "learning_rate": 2.934306569343066e-05,
409
+ "loss": 1.1079,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.1,
414
+ "learning_rate": 2.978102189781022e-05,
415
+ "loss": 1.0963,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.1,
420
+ "learning_rate": 2.9999951294778548e-05,
421
+ "loss": 1.1275,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.1,
426
+ "learning_rate": 2.9999561654904685e-05,
427
+ "loss": 1.13,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.1,
432
+ "learning_rate": 2.9998782385278275e-05,
433
+ "loss": 1.1415,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.11,
438
+ "learning_rate": 2.9997613506141683e-05,
439
+ "loss": 1.0811,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.11,
444
+ "learning_rate": 2.9996055047857808e-05,
445
+ "loss": 1.0885,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.11,
450
+ "learning_rate": 2.999410705090927e-05,
451
+ "loss": 1.112,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.11,
456
+ "learning_rate": 2.999176956589739e-05,
457
+ "loss": 1.0727,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.11,
462
+ "learning_rate": 2.9989042653540863e-05,
463
+ "loss": 1.1176,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.11,
468
+ "learning_rate": 2.9985926384674167e-05,
469
+ "loss": 1.1499,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.11,
474
+ "learning_rate": 2.9982420840245734e-05,
475
+ "loss": 1.082,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.12,
480
+ "learning_rate": 2.9978526111315856e-05,
481
+ "loss": 1.1111,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.12,
486
+ "learning_rate": 2.997424229905431e-05,
487
+ "loss": 1.1424,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.12,
492
+ "learning_rate": 2.9969569514737728e-05,
493
+ "loss": 1.1018,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.12,
498
+ "learning_rate": 2.9964507879746708e-05,
499
+ "loss": 1.0874,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.12,
504
+ "learning_rate": 2.9959057525562664e-05,
505
+ "loss": 1.106,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.12,
510
+ "learning_rate": 2.9953218593764413e-05,
511
+ "loss": 1.1178,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.12,
516
+ "learning_rate": 2.994699123602449e-05,
517
+ "loss": 1.1031,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.13,
522
+ "learning_rate": 2.9940375614105206e-05,
523
+ "loss": 1.0894,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.13,
528
+ "learning_rate": 2.9933371899854472e-05,
529
+ "loss": 1.1483,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.13,
534
+ "learning_rate": 2.992598027520129e-05,
535
+ "loss": 1.0965,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.13,
540
+ "learning_rate": 2.991820093215107e-05,
541
+ "loss": 1.1053,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.13,
546
+ "learning_rate": 2.9910034072780608e-05,
547
+ "loss": 1.0736,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.13,
552
+ "learning_rate": 2.990147990923287e-05,
553
+ "loss": 1.0897,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.13,
558
+ "learning_rate": 2.9892538663711463e-05,
559
+ "loss": 1.0914,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.14,
564
+ "learning_rate": 2.988321056847485e-05,
565
+ "loss": 1.0944,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.14,
570
+ "learning_rate": 2.987349586583035e-05,
571
+ "loss": 1.076,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.14,
576
+ "learning_rate": 2.9863394808127805e-05,
577
+ "loss": 1.0562,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.14,
582
+ "learning_rate": 2.985290765775307e-05,
583
+ "loss": 1.1115,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.14,
588
+ "learning_rate": 2.9842034687121153e-05,
589
+ "loss": 1.0514,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.14,
594
+ "learning_rate": 2.9830776178669162e-05,
595
+ "loss": 1.0988,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.14,
600
+ "learning_rate": 2.9819132424848973e-05,
601
+ "loss": 1.0696,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.15,
606
+ "learning_rate": 2.9807103728119618e-05,
607
+ "loss": 1.0758,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.15,
612
+ "learning_rate": 2.9794690400939434e-05,
613
+ "loss": 1.0612,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.15,
618
+ "learning_rate": 2.9781892765757952e-05,
619
+ "loss": 1.0428,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.15,
624
+ "learning_rate": 2.976871115500752e-05,
625
+ "loss": 1.08,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.15,
630
+ "learning_rate": 2.9755145911094648e-05,
631
+ "loss": 1.0708,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.15,
636
+ "learning_rate": 2.9741197386391156e-05,
637
+ "loss": 1.1058,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.15,
642
+ "learning_rate": 2.972686594322497e-05,
643
+ "loss": 1.0646,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.16,
648
+ "learning_rate": 2.971215195387075e-05,
649
+ "loss": 1.0603,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.16,
654
+ "learning_rate": 2.96970558005402e-05,
655
+ "loss": 1.0867,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.16,
660
+ "learning_rate": 2.9681577875372146e-05,
661
+ "loss": 1.1372,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.16,
666
+ "learning_rate": 2.966571858042235e-05,
667
+ "loss": 1.0973,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.16,
672
+ "learning_rate": 2.964947832765306e-05,
673
+ "loss": 1.1167,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.16,
678
+ "learning_rate": 2.9632857538922312e-05,
679
+ "loss": 1.084,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.16,
684
+ "learning_rate": 2.9615856645972982e-05,
685
+ "loss": 1.0874,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.17,
690
+ "learning_rate": 2.959847609042155e-05,
691
+ "loss": 1.0702,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.17,
696
+ "learning_rate": 2.9580716323746648e-05,
697
+ "loss": 1.0416,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.17,
702
+ "learning_rate": 2.9562577807277328e-05,
703
+ "loss": 1.0661,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.17,
708
+ "learning_rate": 2.9544061012181073e-05,
709
+ "loss": 1.1111,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.17,
714
+ "learning_rate": 2.9525166419451546e-05,
715
+ "loss": 1.1008,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.17,
720
+ "learning_rate": 2.9505894519896136e-05,
721
+ "loss": 1.0829,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.18,
726
+ "learning_rate": 2.948624581412316e-05,
727
+ "loss": 1.0617,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.18,
732
+ "learning_rate": 2.94662208125289e-05,
733
+ "loss": 1.0606,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.18,
738
+ "learning_rate": 2.9445820035284313e-05,
739
+ "loss": 1.0683,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.18,
744
+ "learning_rate": 2.9425044012321538e-05,
745
+ "loss": 1.0838,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.18,
750
+ "learning_rate": 2.940389328332012e-05,
751
+ "loss": 1.0626,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.18,
756
+ "learning_rate": 2.9382368397693015e-05,
757
+ "loss": 1.0625,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.18,
762
+ "learning_rate": 2.936046991457227e-05,
763
+ "loss": 1.0158,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.19,
768
+ "learning_rate": 2.933819840279456e-05,
769
+ "loss": 1.0347,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.19,
774
+ "learning_rate": 2.9315554440886357e-05,
775
+ "loss": 1.0357,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.19,
780
+ "learning_rate": 2.9292538617048938e-05,
781
+ "loss": 1.0939,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.19,
786
+ "learning_rate": 2.9269151529143092e-05,
787
+ "loss": 1.0294,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.19,
792
+ "learning_rate": 2.9245393784673586e-05,
793
+ "loss": 1.1038,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.19,
798
+ "learning_rate": 2.92212660007734e-05,
799
+ "loss": 1.0597,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.19,
804
+ "learning_rate": 2.9196768804187677e-05,
805
+ "loss": 1.0171,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.2,
810
+ "learning_rate": 2.917190283125746e-05,
811
+ "loss": 1.0761,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.2,
816
+ "learning_rate": 2.9146668727903143e-05,
817
+ "loss": 1.0433,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.2,
822
+ "learning_rate": 2.9121067149607717e-05,
823
+ "loss": 1.0268,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.2,
828
+ "learning_rate": 2.9095098761399722e-05,
829
+ "loss": 1.0745,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.2,
834
+ "learning_rate": 2.906876423783598e-05,
835
+ "loss": 1.0593,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.2,
840
+ "learning_rate": 2.9042064262984078e-05,
841
+ "loss": 1.0437,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.2,
846
+ "learning_rate": 2.9014999530404588e-05,
847
+ "loss": 1.0326,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.21,
852
+ "learning_rate": 2.898757074313306e-05,
853
+ "loss": 1.0709,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.21,
858
+ "learning_rate": 2.8959778613661753e-05,
859
+ "loss": 1.0558,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.21,
864
+ "learning_rate": 2.8931623863921136e-05,
865
+ "loss": 1.0621,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.21,
870
+ "learning_rate": 2.8903107225261125e-05,
871
+ "loss": 0.9988,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.21,
876
+ "learning_rate": 2.8874229438432085e-05,
877
+ "loss": 1.0354,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.21,
882
+ "learning_rate": 2.8844991253565612e-05,
883
+ "loss": 1.0501,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.21,
888
+ "learning_rate": 2.8815393430155006e-05,
889
+ "loss": 0.9936,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.22,
894
+ "learning_rate": 2.8785436737035577e-05,
895
+ "loss": 1.0557,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.22,
900
+ "learning_rate": 2.8755121952364665e-05,
901
+ "loss": 1.0624,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.22,
906
+ "learning_rate": 2.8724449863601413e-05,
907
+ "loss": 1.0811,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.22,
912
+ "learning_rate": 2.8693421267486333e-05,
913
+ "loss": 0.9973,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.22,
918
+ "learning_rate": 2.8662036970020586e-05,
919
+ "loss": 1.0887,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.22,
924
+ "learning_rate": 2.8630297786445075e-05,
925
+ "loss": 1.0397,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.22,
930
+ "learning_rate": 2.859820454121924e-05,
931
+ "loss": 1.0791,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.23,
936
+ "learning_rate": 2.856575806799965e-05,
937
+ "loss": 1.0323,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.23,
942
+ "learning_rate": 2.853295920961837e-05,
943
+ "loss": 1.0451,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.23,
948
+ "learning_rate": 2.8499808818061025e-05,
949
+ "loss": 1.0755,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.23,
954
+ "learning_rate": 2.846630775444471e-05,
955
+ "loss": 1.0316,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.23,
960
+ "learning_rate": 2.843245688899559e-05,
961
+ "loss": 1.0297,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.23,
966
+ "learning_rate": 2.839825710102632e-05,
967
+ "loss": 1.0021,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.24,
972
+ "learning_rate": 2.836370927891319e-05,
973
+ "loss": 1.0469,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.24,
978
+ "learning_rate": 2.8328814320073046e-05,
979
+ "loss": 1.002,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.24,
984
+ "learning_rate": 2.829357313094e-05,
985
+ "loss": 1.0441,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.24,
990
+ "learning_rate": 2.8257986626941845e-05,
991
+ "loss": 1.101,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.24,
996
+ "learning_rate": 2.822205573247632e-05,
997
+ "loss": 1.069,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.24,
1002
+ "learning_rate": 2.8185781380887062e-05,
1003
+ "loss": 1.0309,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.24,
1008
+ "learning_rate": 2.8149164514439382e-05,
1009
+ "loss": 1.0329,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.25,
1014
+ "learning_rate": 2.8112206084295792e-05,
1015
+ "loss": 1.0377,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.25,
1020
+ "learning_rate": 2.807490705049127e-05,
1021
+ "loss": 1.1056,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.25,
1026
+ "learning_rate": 2.8037268381908354e-05,
1027
+ "loss": 1.0733,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.25,
1032
+ "learning_rate": 2.7999291056251956e-05,
1033
+ "loss": 1.0134,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.25,
1038
+ "learning_rate": 2.7960976060023973e-05,
1039
+ "loss": 1.0467,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.25,
1044
+ "learning_rate": 2.792232438849766e-05,
1045
+ "loss": 1.0289,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.25,
1050
+ "learning_rate": 2.7883337045691765e-05,
1051
+ "loss": 0.9964,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.26,
1056
+ "learning_rate": 2.7844015044344466e-05,
1057
+ "loss": 1.0221,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.26,
1062
+ "learning_rate": 2.7804359405887057e-05,
1063
+ "loss": 1.027,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.26,
1068
+ "learning_rate": 2.7764371160417414e-05,
1069
+ "loss": 1.0273,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.26,
1074
+ "learning_rate": 2.7724051346673245e-05,
1075
+ "loss": 1.0745,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.26,
1080
+ "learning_rate": 2.768340101200508e-05,
1081
+ "loss": 1.0127,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.26,
1086
+ "learning_rate": 2.7642421212349106e-05,
1087
+ "loss": 1.0355,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.26,
1092
+ "learning_rate": 2.7601113012199715e-05,
1093
+ "loss": 0.9789,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.27,
1098
+ "learning_rate": 2.7559477484581847e-05,
1099
+ "loss": 1.0704,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.27,
1104
+ "learning_rate": 2.7517515711023125e-05,
1105
+ "loss": 1.0902,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.27,
1110
+ "learning_rate": 2.747522878152577e-05,
1111
+ "loss": 1.0068,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.27,
1116
+ "learning_rate": 2.7432617794538276e-05,
1117
+ "loss": 1.0375,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.27,
1122
+ "learning_rate": 2.7389683856926868e-05,
1123
+ "loss": 1.0262,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.27,
1128
+ "learning_rate": 2.734642808394678e-05,
1129
+ "loss": 1.0245,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.27,
1134
+ "learning_rate": 2.7302851599213247e-05,
1135
+ "loss": 0.997,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.28,
1140
+ "learning_rate": 2.725895553467235e-05,
1141
+ "loss": 1.0468,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.28,
1146
+ "learning_rate": 2.7214741030571592e-05,
1147
+ "loss": 1.026,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.28,
1152
+ "learning_rate": 2.7170209235430294e-05,
1153
+ "loss": 1.0155,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.28,
1158
+ "learning_rate": 2.712536130600974e-05,
1159
+ "loss": 1.0718,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.28,
1164
+ "learning_rate": 2.7080198407283158e-05,
1165
+ "loss": 1.045,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.28,
1170
+ "learning_rate": 2.7034721712405437e-05,
1171
+ "loss": 1.0383,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.28,
1176
+ "learning_rate": 2.6988932402682644e-05,
1177
+ "loss": 1.0196,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.29,
1182
+ "learning_rate": 2.694283166754137e-05,
1183
+ "loss": 1.0087,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.29,
1188
+ "learning_rate": 2.6896420704497824e-05,
1189
+ "loss": 1.0493,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.29,
1194
+ "learning_rate": 2.684970071912669e-05,
1195
+ "loss": 1.0588,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.29,
1200
+ "learning_rate": 2.6802672925029866e-05,
1201
+ "loss": 1.0106,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.29,
1206
+ "learning_rate": 2.6755338543804898e-05,
1207
+ "loss": 1.0489,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.29,
1212
+ "learning_rate": 2.6707698805013265e-05,
1213
+ "loss": 1.0583,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.29,
1218
+ "learning_rate": 2.6659754946148446e-05,
1219
+ "loss": 1.0133,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.3,
1224
+ "learning_rate": 2.6611508212603755e-05,
1225
+ "loss": 1.0838,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.3,
1230
+ "learning_rate": 2.6562959857640006e-05,
1231
+ "loss": 1.0223,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.3,
1236
+ "learning_rate": 2.6514111142352958e-05,
1237
+ "loss": 1.0647,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.3,
1242
+ "learning_rate": 2.6464963335640543e-05,
1243
+ "loss": 1.0354,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.3,
1248
+ "learning_rate": 2.6415517714169922e-05,
1249
+ "loss": 1.0518,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.3,
1254
+ "learning_rate": 2.6365775562344316e-05,
1255
+ "loss": 1.0197,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.31,
1260
+ "learning_rate": 2.6315738172269635e-05,
1261
+ "loss": 1.0214,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.31,
1266
+ "learning_rate": 2.6265406843720924e-05,
1267
+ "loss": 0.9732,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.31,
1272
+ "learning_rate": 2.62147828841086e-05,
1273
+ "loss": 1.0811,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.31,
1278
+ "learning_rate": 2.6163867608444473e-05,
1279
+ "loss": 1.031,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.31,
1284
+ "learning_rate": 2.6112662339307617e-05,
1285
+ "loss": 1.0148,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.31,
1290
+ "learning_rate": 2.6061168406809985e-05,
1291
+ "loss": 1.0187,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.31,
1296
+ "learning_rate": 2.6009387148561884e-05,
1297
+ "loss": 1.0356,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.32,
1302
+ "learning_rate": 2.5957319909637208e-05,
1303
+ "loss": 1.0386,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.32,
1308
+ "learning_rate": 2.590496804253851e-05,
1309
+ "loss": 1.0803,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.32,
1314
+ "learning_rate": 2.5852332907161862e-05,
1315
+ "loss": 1.0175,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.32,
1320
+ "learning_rate": 2.5799415870761545e-05,
1321
+ "loss": 1.0571,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.32,
1326
+ "learning_rate": 2.57462183079145e-05,
1327
+ "loss": 1.0155,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.32,
1332
+ "learning_rate": 2.5692741600484676e-05,
1333
+ "loss": 1.0745,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.32,
1338
+ "learning_rate": 2.563898713758708e-05,
1339
+ "loss": 1.0834,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.33,
1344
+ "learning_rate": 2.5584956315551716e-05,
1345
+ "loss": 1.0132,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.33,
1350
+ "learning_rate": 2.5530650537887327e-05,
1351
+ "loss": 1.0205,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.33,
1356
+ "learning_rate": 2.5476071215244915e-05,
1357
+ "loss": 1.0325,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.33,
1362
+ "learning_rate": 2.5421219765381107e-05,
1363
+ "loss": 1.0503,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.33,
1368
+ "learning_rate": 2.536609761312134e-05,
1369
+ "loss": 1.0507,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.33,
1374
+ "learning_rate": 2.5310706190322806e-05,
1375
+ "loss": 1.0492,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.33,
1380
+ "learning_rate": 2.5255046935837332e-05,
1381
+ "loss": 1.0016,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.34,
1386
+ "learning_rate": 2.5199121295473926e-05,
1387
+ "loss": 1.0152,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.34,
1392
+ "learning_rate": 2.5142930721961277e-05,
1393
+ "loss": 1.044,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.34,
1398
+ "learning_rate": 2.5086476674909983e-05,
1399
+ "loss": 1.0151,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.34,
1404
+ "learning_rate": 2.502976062077466e-05,
1405
+ "loss": 1.0248,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.34,
1410
+ "learning_rate": 2.4972784032815837e-05,
1411
+ "loss": 0.9813,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.34,
1416
+ "learning_rate": 2.4915548391061688e-05,
1417
+ "loss": 1.0249,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.34,
1422
+ "learning_rate": 2.4858055182269586e-05,
1423
+ "loss": 1.001,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.35,
1428
+ "learning_rate": 2.48003058998875e-05,
1429
+ "loss": 1.0554,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.35,
1434
+ "learning_rate": 2.474230204401516e-05,
1435
+ "loss": 1.0022,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.35,
1440
+ "learning_rate": 2.468404512136514e-05,
1441
+ "loss": 1.0518,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.35,
1446
+ "learning_rate": 2.4625536645223687e-05,
1447
+ "loss": 1.0541,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.35,
1452
+ "learning_rate": 2.4566778135411417e-05,
1453
+ "loss": 1.0201,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.35,
1458
+ "learning_rate": 2.4507771118243838e-05,
1459
+ "loss": 0.9828,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.35,
1464
+ "learning_rate": 2.4448517126491717e-05,
1465
+ "loss": 1.0663,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.36,
1470
+ "learning_rate": 2.438901769934124e-05,
1471
+ "loss": 1.0088,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.36,
1476
+ "learning_rate": 2.4329274382354043e-05,
1477
+ "loss": 1.0139,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.36,
1482
+ "learning_rate": 2.4269288727427066e-05,
1483
+ "loss": 1.0488,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.36,
1488
+ "learning_rate": 2.420906229275224e-05,
1489
+ "loss": 1.0568,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.36,
1494
+ "learning_rate": 2.4148596642775997e-05,
1495
+ "loss": 1.0521,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.36,
1500
+ "learning_rate": 2.408789334815866e-05,
1501
+ "loss": 0.9795,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.37,
1506
+ "learning_rate": 2.402695398573362e-05,
1507
+ "loss": 1.0708,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.37,
1512
+ "learning_rate": 2.3965780138466375e-05,
1513
+ "loss": 1.0154,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.37,
1518
+ "learning_rate": 2.3904373395413435e-05,
1519
+ "loss": 1.0395,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.37,
1524
+ "learning_rate": 2.3842735351681016e-05,
1525
+ "loss": 1.0331,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.37,
1530
+ "learning_rate": 2.3780867608383624e-05,
1531
+ "loss": 0.9829,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.37,
1536
+ "learning_rate": 2.3718771772602456e-05,
1537
+ "loss": 1.0569,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.37,
1542
+ "learning_rate": 2.365644945734366e-05,
1543
+ "loss": 1.0729,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.38,
1548
+ "learning_rate": 2.3593902281496424e-05,
1549
+ "loss": 1.0369,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.38,
1554
+ "learning_rate": 2.3531131869790948e-05,
1555
+ "loss": 0.9883,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.38,
1560
+ "learning_rate": 2.34681398527562e-05,
1561
+ "loss": 1.0359,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.38,
1566
+ "learning_rate": 2.3404927866677607e-05,
1567
+ "loss": 1.0336,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.38,
1572
+ "learning_rate": 2.3341497553554512e-05,
1573
+ "loss": 0.9892,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.38,
1578
+ "learning_rate": 2.3277850561057554e-05,
1579
+ "loss": 1.0212,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.38,
1584
+ "learning_rate": 2.3213988542485838e-05,
1585
+ "loss": 1.072,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.39,
1590
+ "learning_rate": 2.3149913156724005e-05,
1591
+ "loss": 0.9956,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.39,
1596
+ "learning_rate": 2.3085626068199145e-05,
1597
+ "loss": 1.0685,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.39,
1602
+ "learning_rate": 2.3021128946837557e-05,
1603
+ "loss": 1.0248,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.39,
1608
+ "learning_rate": 2.2956423468021352e-05,
1609
+ "loss": 1.017,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.39,
1614
+ "learning_rate": 2.2891511312544972e-05,
1615
+ "loss": 1.029,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.39,
1620
+ "learning_rate": 2.2826394166571485e-05,
1621
+ "loss": 1.0368,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.39,
1626
+ "learning_rate": 2.2761073721588825e-05,
1627
+ "loss": 0.961,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.4,
1632
+ "learning_rate": 2.2695551674365836e-05,
1633
+ "loss": 1.0061,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.4,
1638
+ "learning_rate": 2.2629829726908186e-05,
1639
+ "loss": 1.0595,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.4,
1644
+ "learning_rate": 2.2563909586414174e-05,
1645
+ "loss": 1.0482,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.4,
1650
+ "learning_rate": 2.2497792965230372e-05,
1651
+ "loss": 0.9802,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.4,
1656
+ "learning_rate": 2.2431481580807163e-05,
1657
+ "loss": 1.0083,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.4,
1662
+ "learning_rate": 2.2364977155654095e-05,
1663
+ "loss": 1.0171,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.4,
1668
+ "learning_rate": 2.2298281417295167e-05,
1669
+ "loss": 0.9965,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.41,
1674
+ "learning_rate": 2.2231396098223944e-05,
1675
+ "loss": 1.0747,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.41,
1680
+ "learning_rate": 2.2164322935858552e-05,
1681
+ "loss": 1.0038,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.41,
1686
+ "learning_rate": 2.2097063672496555e-05,
1687
+ "loss": 0.9683,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.41,
1692
+ "learning_rate": 2.2029620055269677e-05,
1693
+ "loss": 1.0157,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.41,
1698
+ "learning_rate": 2.196199383609845e-05,
1699
+ "loss": 1.0479,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.41,
1704
+ "learning_rate": 2.1894186771646672e-05,
1705
+ "loss": 0.9811,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.41,
1710
+ "learning_rate": 2.1826200623275802e-05,
1711
+ "loss": 1.0051,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.42,
1716
+ "learning_rate": 2.1758037156999185e-05,
1717
+ "loss": 1.0145,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.42,
1722
+ "learning_rate": 2.1689698143436208e-05,
1723
+ "loss": 0.9983,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.42,
1728
+ "learning_rate": 2.1621185357766267e-05,
1729
+ "loss": 1.0197,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.42,
1734
+ "learning_rate": 2.155250057968269e-05,
1735
+ "loss": 1.0619,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.42,
1740
+ "learning_rate": 2.1483645593346487e-05,
1741
+ "loss": 0.9677,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.42,
1746
+ "learning_rate": 2.1414622187340015e-05,
1747
+ "loss": 1.066,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.42,
1752
+ "learning_rate": 2.1345432154620506e-05,
1753
+ "loss": 1.0523,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.43,
1758
+ "learning_rate": 2.1276077292473502e-05,
1759
+ "loss": 1.0226,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.43,
1764
+ "learning_rate": 2.1206559402466176e-05,
1765
+ "loss": 1.0658,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.43,
1770
+ "learning_rate": 2.113688029040051e-05,
1771
+ "loss": 1.0519,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.43,
1776
+ "learning_rate": 2.106704176626641e-05,
1777
+ "loss": 1.0363,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.43,
1782
+ "learning_rate": 2.099704564419469e-05,
1783
+ "loss": 1.0182,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.43,
1788
+ "learning_rate": 2.0926893742409923e-05,
1789
+ "loss": 1.0752,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.44,
1794
+ "learning_rate": 2.0856587883183247e-05,
1795
+ "loss": 1.0286,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.44,
1800
+ "learning_rate": 2.078612989278499e-05,
1801
+ "loss": 1.0012,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.44,
1806
+ "learning_rate": 2.0715521601437265e-05,
1807
+ "loss": 1.0053,
1808
+ "step": 3000
1809
+ },
1810
+ {
1811
+ "epoch": 0.44,
1812
+ "learning_rate": 2.0644764843266413e-05,
1813
+ "loss": 1.0562,
1814
+ "step": 3010
1815
+ },
1816
+ {
1817
+ "epoch": 0.44,
1818
+ "learning_rate": 2.0573861456255354e-05,
1819
+ "loss": 0.9871,
1820
+ "step": 3020
1821
+ },
1822
+ {
1823
+ "epoch": 0.44,
1824
+ "learning_rate": 2.050281328219585e-05,
1825
+ "loss": 1.0166,
1826
+ "step": 3030
1827
+ },
1828
+ {
1829
+ "epoch": 0.44,
1830
+ "learning_rate": 2.0431622166640667e-05,
1831
+ "loss": 1.0534,
1832
+ "step": 3040
1833
+ },
1834
+ {
1835
+ "epoch": 0.45,
1836
+ "learning_rate": 2.0360289958855634e-05,
1837
+ "loss": 1.032,
1838
+ "step": 3050
1839
+ },
1840
+ {
1841
+ "epoch": 0.45,
1842
+ "learning_rate": 2.028881851177159e-05,
1843
+ "loss": 0.9949,
1844
+ "step": 3060
1845
+ },
1846
+ {
1847
+ "epoch": 0.45,
1848
+ "learning_rate": 2.021720968193628e-05,
1849
+ "loss": 0.9932,
1850
+ "step": 3070
1851
+ },
1852
+ {
1853
+ "epoch": 0.45,
1854
+ "learning_rate": 2.0145465329466092e-05,
1855
+ "loss": 0.9909,
1856
+ "step": 3080
1857
+ },
1858
+ {
1859
+ "epoch": 0.45,
1860
+ "learning_rate": 2.007358731799779e-05,
1861
+ "loss": 0.9935,
1862
+ "step": 3090
1863
+ },
1864
+ {
1865
+ "epoch": 0.45,
1866
+ "learning_rate": 2.0001577514640046e-05,
1867
+ "loss": 1.0379,
1868
+ "step": 3100
1869
+ },
1870
+ {
1871
+ "epoch": 0.45,
1872
+ "learning_rate": 1.9929437789924987e-05,
1873
+ "loss": 1.0241,
1874
+ "step": 3110
1875
+ },
1876
+ {
1877
+ "epoch": 0.46,
1878
+ "learning_rate": 1.9857170017759572e-05,
1879
+ "loss": 0.9976,
1880
+ "step": 3120
1881
+ },
1882
+ {
1883
+ "epoch": 0.46,
1884
+ "learning_rate": 1.978477607537695e-05,
1885
+ "loss": 1.0481,
1886
+ "step": 3130
1887
+ },
1888
+ {
1889
+ "epoch": 0.46,
1890
+ "learning_rate": 1.9712257843287656e-05,
1891
+ "loss": 1.0236,
1892
+ "step": 3140
1893
+ },
1894
+ {
1895
+ "epoch": 0.46,
1896
+ "learning_rate": 1.9639617205230798e-05,
1897
+ "loss": 1.0249,
1898
+ "step": 3150
1899
+ },
1900
+ {
1901
+ "epoch": 0.46,
1902
+ "learning_rate": 1.9566856048125102e-05,
1903
+ "loss": 1.0197,
1904
+ "step": 3160
1905
+ },
1906
+ {
1907
+ "epoch": 0.46,
1908
+ "learning_rate": 1.9493976262019916e-05,
1909
+ "loss": 1.0039,
1910
+ "step": 3170
1911
+ },
1912
+ {
1913
+ "epoch": 0.46,
1914
+ "learning_rate": 1.9420979740046095e-05,
1915
+ "loss": 1.041,
1916
+ "step": 3180
1917
+ },
1918
+ {
1919
+ "epoch": 0.47,
1920
+ "learning_rate": 1.9347868378366837e-05,
1921
+ "loss": 1.0347,
1922
+ "step": 3190
1923
+ },
1924
+ {
1925
+ "epoch": 0.47,
1926
+ "learning_rate": 1.9274644076128428e-05,
1927
+ "loss": 1.016,
1928
+ "step": 3200
1929
+ },
1930
+ {
1931
+ "epoch": 0.47,
1932
+ "learning_rate": 1.9201308735410895e-05,
1933
+ "loss": 1.041,
1934
+ "step": 3210
1935
+ },
1936
+ {
1937
+ "epoch": 0.47,
1938
+ "learning_rate": 1.9127864261178623e-05,
1939
+ "loss": 0.9939,
1940
+ "step": 3220
1941
+ },
1942
+ {
1943
+ "epoch": 0.47,
1944
+ "learning_rate": 1.9054312561230855e-05,
1945
+ "loss": 1.0674,
1946
+ "step": 3230
1947
+ },
1948
+ {
1949
+ "epoch": 0.47,
1950
+ "learning_rate": 1.8980655546152123e-05,
1951
+ "loss": 1.0005,
1952
+ "step": 3240
1953
+ },
1954
+ {
1955
+ "epoch": 0.47,
1956
+ "learning_rate": 1.8906895129262655e-05,
1957
+ "loss": 1.0059,
1958
+ "step": 3250
1959
+ },
1960
+ {
1961
+ "epoch": 0.48,
1962
+ "learning_rate": 1.8833033226568636e-05,
1963
+ "loss": 0.986,
1964
+ "step": 3260
1965
+ },
1966
+ {
1967
+ "epoch": 0.48,
1968
+ "learning_rate": 1.875907175671246e-05,
1969
+ "loss": 1.0224,
1970
+ "step": 3270
1971
+ },
1972
+ {
1973
+ "epoch": 0.48,
1974
+ "learning_rate": 1.8685012640922886e-05,
1975
+ "loss": 0.9967,
1976
+ "step": 3280
1977
+ },
1978
+ {
1979
+ "epoch": 0.48,
1980
+ "learning_rate": 1.8610857802965127e-05,
1981
+ "loss": 0.9802,
1982
+ "step": 3290
1983
+ },
1984
+ {
1985
+ "epoch": 0.48,
1986
+ "learning_rate": 1.853660916909089e-05,
1987
+ "loss": 1.0024,
1988
+ "step": 3300
1989
+ },
1990
+ {
1991
+ "epoch": 0.48,
1992
+ "learning_rate": 1.8462268667988316e-05,
1993
+ "loss": 1.0104,
1994
+ "step": 3310
1995
+ },
1996
+ {
1997
+ "epoch": 0.48,
1998
+ "learning_rate": 1.8387838230731915e-05,
1999
+ "loss": 0.9954,
2000
+ "step": 3320
2001
+ },
2002
+ {
2003
+ "epoch": 0.49,
2004
+ "learning_rate": 1.8313319790732373e-05,
2005
+ "loss": 0.9586,
2006
+ "step": 3330
2007
+ },
2008
+ {
2009
+ "epoch": 0.49,
2010
+ "learning_rate": 1.823871528368635e-05,
2011
+ "loss": 0.968,
2012
+ "step": 3340
2013
+ },
2014
+ {
2015
+ "epoch": 0.49,
2016
+ "learning_rate": 1.8164026647526183e-05,
2017
+ "loss": 0.9949,
2018
+ "step": 3350
2019
+ },
2020
+ {
2021
+ "epoch": 0.49,
2022
+ "learning_rate": 1.808925582236956e-05,
2023
+ "loss": 1.0205,
2024
+ "step": 3360
2025
+ },
2026
+ {
2027
+ "epoch": 0.49,
2028
+ "learning_rate": 1.8014404750469113e-05,
2029
+ "loss": 1.0236,
2030
+ "step": 3370
2031
+ },
2032
+ {
2033
+ "epoch": 0.49,
2034
+ "learning_rate": 1.7939475376161963e-05,
2035
+ "loss": 1.0316,
2036
+ "step": 3380
2037
+ },
2038
+ {
2039
+ "epoch": 0.49,
2040
+ "learning_rate": 1.7864469645819234e-05,
2041
+ "loss": 1.0432,
2042
+ "step": 3390
2043
+ },
2044
+ {
2045
+ "epoch": 0.5,
2046
+ "learning_rate": 1.778938950779547e-05,
2047
+ "loss": 1.0138,
2048
+ "step": 3400
2049
+ },
2050
+ {
2051
+ "epoch": 0.5,
2052
+ "learning_rate": 1.7714236912378038e-05,
2053
+ "loss": 1.0157,
2054
+ "step": 3410
2055
+ },
2056
+ {
2057
+ "epoch": 0.5,
2058
+ "learning_rate": 1.7639013811736458e-05,
2059
+ "loss": 1.0072,
2060
+ "step": 3420
2061
+ },
2062
+ {
2063
+ "epoch": 0.5,
2064
+ "learning_rate": 1.7563722159871712e-05,
2065
+ "loss": 0.996,
2066
+ "step": 3430
2067
+ },
2068
+ {
2069
+ "epoch": 0.5,
2070
+ "learning_rate": 1.7488363912565466e-05,
2071
+ "loss": 1.0315,
2072
+ "step": 3440
2073
+ },
2074
+ {
2075
+ "epoch": 0.5,
2076
+ "learning_rate": 1.7412941027329267e-05,
2077
+ "loss": 1.0605,
2078
+ "step": 3450
2079
+ },
2080
+ {
2081
+ "epoch": 0.51,
2082
+ "learning_rate": 1.733745546335372e-05,
2083
+ "loss": 1.0237,
2084
+ "step": 3460
2085
+ },
2086
+ {
2087
+ "epoch": 0.51,
2088
+ "learning_rate": 1.7261909181457554e-05,
2089
+ "loss": 1.012,
2090
+ "step": 3470
2091
+ },
2092
+ {
2093
+ "epoch": 0.51,
2094
+ "learning_rate": 1.718630414403674e-05,
2095
+ "loss": 0.987,
2096
+ "step": 3480
2097
+ },
2098
+ {
2099
+ "epoch": 0.51,
2100
+ "learning_rate": 1.7110642315013454e-05,
2101
+ "loss": 1.0672,
2102
+ "step": 3490
2103
+ },
2104
+ {
2105
+ "epoch": 0.51,
2106
+ "learning_rate": 1.7034925659785127e-05,
2107
+ "loss": 0.975,
2108
+ "step": 3500
2109
+ },
2110
+ {
2111
+ "epoch": 0.51,
2112
+ "learning_rate": 1.695915614517333e-05,
2113
+ "loss": 1.0022,
2114
+ "step": 3510
2115
+ },
2116
+ {
2117
+ "epoch": 0.51,
2118
+ "learning_rate": 1.6883335739372743e-05,
2119
+ "loss": 1.0443,
2120
+ "step": 3520
2121
+ },
2122
+ {
2123
+ "epoch": 0.52,
2124
+ "learning_rate": 1.680746641189998e-05,
2125
+ "loss": 0.9779,
2126
+ "step": 3530
2127
+ },
2128
+ {
2129
+ "epoch": 0.52,
2130
+ "learning_rate": 1.6731550133542443e-05,
2131
+ "loss": 0.9595,
2132
+ "step": 3540
2133
+ },
2134
+ {
2135
+ "epoch": 0.52,
2136
+ "learning_rate": 1.665558887630716e-05,
2137
+ "loss": 1.0116,
2138
+ "step": 3550
2139
+ },
2140
+ {
2141
+ "epoch": 0.52,
2142
+ "learning_rate": 1.65795846133695e-05,
2143
+ "loss": 1.0014,
2144
+ "step": 3560
2145
+ },
2146
+ {
2147
+ "epoch": 0.52,
2148
+ "learning_rate": 1.650353931902198e-05,
2149
+ "loss": 1.0118,
2150
+ "step": 3570
2151
+ },
2152
+ {
2153
+ "epoch": 0.52,
2154
+ "learning_rate": 1.6427454968622937e-05,
2155
+ "loss": 1.031,
2156
+ "step": 3580
2157
+ },
2158
+ {
2159
+ "epoch": 0.52,
2160
+ "learning_rate": 1.6351333538545236e-05,
2161
+ "loss": 0.9675,
2162
+ "step": 3590
2163
+ },
2164
+ {
2165
+ "epoch": 0.53,
2166
+ "learning_rate": 1.6275177006124916e-05,
2167
+ "loss": 1.0421,
2168
+ "step": 3600
2169
+ },
2170
+ {
2171
+ "epoch": 0.53,
2172
+ "learning_rate": 1.619898734960986e-05,
2173
+ "loss": 1.0175,
2174
+ "step": 3610
2175
+ },
2176
+ {
2177
+ "epoch": 0.53,
2178
+ "learning_rate": 1.6122766548108364e-05,
2179
+ "loss": 0.9907,
2180
+ "step": 3620
2181
+ },
2182
+ {
2183
+ "epoch": 0.53,
2184
+ "learning_rate": 1.6046516581537754e-05,
2185
+ "loss": 0.9671,
2186
+ "step": 3630
2187
+ },
2188
+ {
2189
+ "epoch": 0.53,
2190
+ "learning_rate": 1.5970239430572958e-05,
2191
+ "loss": 1.0066,
2192
+ "step": 3640
2193
+ },
2194
+ {
2195
+ "epoch": 0.53,
2196
+ "learning_rate": 1.5893937076595036e-05,
2197
+ "loss": 0.9498,
2198
+ "step": 3650
2199
+ },
2200
+ {
2201
+ "epoch": 0.53,
2202
+ "learning_rate": 1.581761150163973e-05,
2203
+ "loss": 1.02,
2204
+ "step": 3660
2205
+ },
2206
+ {
2207
+ "epoch": 0.54,
2208
+ "learning_rate": 1.5741264688345972e-05,
2209
+ "loss": 0.9901,
2210
+ "step": 3670
2211
+ },
2212
+ {
2213
+ "epoch": 0.54,
2214
+ "learning_rate": 1.5664898619904384e-05,
2215
+ "loss": 1.0409,
2216
+ "step": 3680
2217
+ },
2218
+ {
2219
+ "epoch": 0.54,
2220
+ "learning_rate": 1.5588515280005754e-05,
2221
+ "loss": 1.0504,
2222
+ "step": 3690
2223
+ },
2224
+ {
2225
+ "epoch": 0.54,
2226
+ "learning_rate": 1.551211665278952e-05,
2227
+ "loss": 1.0119,
2228
+ "step": 3700
2229
+ },
2230
+ {
2231
+ "epoch": 0.54,
2232
+ "learning_rate": 1.5435704722792224e-05,
2233
+ "loss": 1.0011,
2234
+ "step": 3710
2235
+ },
2236
+ {
2237
+ "epoch": 0.54,
2238
+ "learning_rate": 1.5359281474895966e-05,
2239
+ "loss": 0.9896,
2240
+ "step": 3720
2241
+ },
2242
+ {
2243
+ "epoch": 0.54,
2244
+ "learning_rate": 1.5282848894276837e-05,
2245
+ "loss": 1.0349,
2246
+ "step": 3730
2247
+ },
2248
+ {
2249
+ "epoch": 0.55,
2250
+ "learning_rate": 1.5206408966353343e-05,
2251
+ "loss": 0.9609,
2252
+ "step": 3740
2253
+ },
2254
+ {
2255
+ "epoch": 0.55,
2256
+ "learning_rate": 1.5129963676734874e-05,
2257
+ "loss": 1.0007,
2258
+ "step": 3750
2259
+ },
2260
+ {
2261
+ "epoch": 0.55,
2262
+ "learning_rate": 1.5053515011170068e-05,
2263
+ "loss": 1.0384,
2264
+ "step": 3760
2265
+ },
2266
+ {
2267
+ "epoch": 0.55,
2268
+ "learning_rate": 1.4977064955495265e-05,
2269
+ "loss": 1.0187,
2270
+ "step": 3770
2271
+ },
2272
+ {
2273
+ "epoch": 0.55,
2274
+ "learning_rate": 1.4900615495582916e-05,
2275
+ "loss": 0.9817,
2276
+ "step": 3780
2277
+ },
2278
+ {
2279
+ "epoch": 0.55,
2280
+ "learning_rate": 1.4824168617290009e-05,
2281
+ "loss": 0.9856,
2282
+ "step": 3790
2283
+ },
2284
+ {
2285
+ "epoch": 0.55,
2286
+ "learning_rate": 1.4747726306406444e-05,
2287
+ "loss": 1.01,
2288
+ "step": 3800
2289
+ },
2290
+ {
2291
+ "epoch": 0.56,
2292
+ "learning_rate": 1.4671290548603503e-05,
2293
+ "loss": 1.0264,
2294
+ "step": 3810
2295
+ },
2296
+ {
2297
+ "epoch": 0.56,
2298
+ "learning_rate": 1.4594863329382235e-05,
2299
+ "loss": 1.0105,
2300
+ "step": 3820
2301
+ },
2302
+ {
2303
+ "epoch": 0.56,
2304
+ "learning_rate": 1.4518446634021887e-05,
2305
+ "loss": 1.0075,
2306
+ "step": 3830
2307
+ },
2308
+ {
2309
+ "epoch": 0.56,
2310
+ "learning_rate": 1.444204244752834e-05,
2311
+ "loss": 1.0036,
2312
+ "step": 3840
2313
+ },
2314
+ {
2315
+ "epoch": 0.56,
2316
+ "learning_rate": 1.436565275458254e-05,
2317
+ "loss": 1.0064,
2318
+ "step": 3850
2319
+ },
2320
+ {
2321
+ "epoch": 0.56,
2322
+ "learning_rate": 1.428927953948896e-05,
2323
+ "loss": 0.9885,
2324
+ "step": 3860
2325
+ },
2326
+ {
2327
+ "epoch": 0.57,
2328
+ "learning_rate": 1.4212924786124033e-05,
2329
+ "loss": 0.9896,
2330
+ "step": 3870
2331
+ },
2332
+ {
2333
+ "epoch": 0.57,
2334
+ "learning_rate": 1.4136590477884626e-05,
2335
+ "loss": 1.0143,
2336
+ "step": 3880
2337
+ },
2338
+ {
2339
+ "epoch": 0.57,
2340
+ "learning_rate": 1.4060278597636528e-05,
2341
+ "loss": 1.0107,
2342
+ "step": 3890
2343
+ },
2344
+ {
2345
+ "epoch": 0.57,
2346
+ "learning_rate": 1.3983991127662936e-05,
2347
+ "loss": 0.968,
2348
+ "step": 3900
2349
+ },
2350
+ {
2351
+ "epoch": 0.57,
2352
+ "learning_rate": 1.3907730049612967e-05,
2353
+ "loss": 1.0048,
2354
+ "step": 3910
2355
+ },
2356
+ {
2357
+ "epoch": 0.57,
2358
+ "learning_rate": 1.3831497344450171e-05,
2359
+ "loss": 1.0245,
2360
+ "step": 3920
2361
+ },
2362
+ {
2363
+ "epoch": 0.57,
2364
+ "learning_rate": 1.3755294992401086e-05,
2365
+ "loss": 1.04,
2366
+ "step": 3930
2367
+ },
2368
+ {
2369
+ "epoch": 0.58,
2370
+ "learning_rate": 1.367912497290379e-05,
2371
+ "loss": 1.0232,
2372
+ "step": 3940
2373
+ },
2374
+ {
2375
+ "epoch": 0.58,
2376
+ "learning_rate": 1.3602989264556506e-05,
2377
+ "loss": 1.0247,
2378
+ "step": 3950
2379
+ },
2380
+ {
2381
+ "epoch": 0.58,
2382
+ "learning_rate": 1.3526889845066166e-05,
2383
+ "loss": 1.0201,
2384
+ "step": 3960
2385
+ },
2386
+ {
2387
+ "epoch": 0.58,
2388
+ "learning_rate": 1.345082869119707e-05,
2389
+ "loss": 0.9729,
2390
+ "step": 3970
2391
+ },
2392
+ {
2393
+ "epoch": 0.58,
2394
+ "learning_rate": 1.3374807778719532e-05,
2395
+ "loss": 0.9981,
2396
+ "step": 3980
2397
+ },
2398
+ {
2399
+ "epoch": 0.58,
2400
+ "learning_rate": 1.3298829082358541e-05,
2401
+ "loss": 0.9969,
2402
+ "step": 3990
2403
+ },
2404
+ {
2405
+ "epoch": 0.58,
2406
+ "learning_rate": 1.3222894575742489e-05,
2407
+ "loss": 1.0115,
2408
+ "step": 4000
2409
+ },
2410
+ {
2411
+ "epoch": 0.59,
2412
+ "learning_rate": 1.3147006231351884e-05,
2413
+ "loss": 1.0155,
2414
+ "step": 4010
2415
+ },
2416
+ {
2417
+ "epoch": 0.59,
2418
+ "learning_rate": 1.307116602046812e-05,
2419
+ "loss": 1.0244,
2420
+ "step": 4020
2421
+ },
2422
+ {
2423
+ "epoch": 0.59,
2424
+ "learning_rate": 1.2995375913122272e-05,
2425
+ "loss": 0.9805,
2426
+ "step": 4030
2427
+ },
2428
+ {
2429
+ "epoch": 0.59,
2430
+ "learning_rate": 1.291963787804393e-05,
2431
+ "loss": 1.0217,
2432
+ "step": 4040
2433
+ },
2434
+ {
2435
+ "epoch": 0.59,
2436
+ "learning_rate": 1.2843953882610043e-05,
2437
+ "loss": 1.0406,
2438
+ "step": 4050
2439
+ },
2440
+ {
2441
+ "epoch": 0.59,
2442
+ "learning_rate": 1.2768325892793819e-05,
2443
+ "loss": 1.056,
2444
+ "step": 4060
2445
+ },
2446
+ {
2447
+ "epoch": 0.59,
2448
+ "learning_rate": 1.2692755873113666e-05,
2449
+ "loss": 0.9892,
2450
+ "step": 4070
2451
+ },
2452
+ {
2453
+ "epoch": 0.6,
2454
+ "learning_rate": 1.2617245786582157e-05,
2455
+ "loss": 0.9933,
2456
+ "step": 4080
2457
+ },
2458
+ {
2459
+ "epoch": 0.6,
2460
+ "learning_rate": 1.2541797594655031e-05,
2461
+ "loss": 1.0152,
2462
+ "step": 4090
2463
+ },
2464
+ {
2465
+ "epoch": 0.6,
2466
+ "learning_rate": 1.2466413257180254e-05,
2467
+ "loss": 1.0015,
2468
+ "step": 4100
2469
+ },
2470
+ {
2471
+ "epoch": 0.6,
2472
+ "learning_rate": 1.2391094732347098e-05,
2473
+ "loss": 1.0352,
2474
+ "step": 4110
2475
+ },
2476
+ {
2477
+ "epoch": 0.6,
2478
+ "learning_rate": 1.2315843976635276e-05,
2479
+ "loss": 1.0022,
2480
+ "step": 4120
2481
+ },
2482
+ {
2483
+ "epoch": 0.6,
2484
+ "learning_rate": 1.2240662944764146e-05,
2485
+ "loss": 0.9675,
2486
+ "step": 4130
2487
+ },
2488
+ {
2489
+ "epoch": 0.6,
2490
+ "learning_rate": 1.2165553589641895e-05,
2491
+ "loss": 0.9959,
2492
+ "step": 4140
2493
+ },
2494
+ {
2495
+ "epoch": 0.61,
2496
+ "learning_rate": 1.2090517862314831e-05,
2497
+ "loss": 0.9988,
2498
+ "step": 4150
2499
+ },
2500
+ {
2501
+ "epoch": 0.61,
2502
+ "learning_rate": 1.2015557711916713e-05,
2503
+ "loss": 1.0156,
2504
+ "step": 4160
2505
+ },
2506
+ {
2507
+ "epoch": 0.61,
2508
+ "learning_rate": 1.1940675085618088e-05,
2509
+ "loss": 1.0343,
2510
+ "step": 4170
2511
+ },
2512
+ {
2513
+ "epoch": 0.61,
2514
+ "learning_rate": 1.1865871928575759e-05,
2515
+ "loss": 1.0788,
2516
+ "step": 4180
2517
+ },
2518
+ {
2519
+ "epoch": 0.61,
2520
+ "learning_rate": 1.1791150183882203e-05,
2521
+ "loss": 1.0417,
2522
+ "step": 4190
2523
+ },
2524
+ {
2525
+ "epoch": 0.61,
2526
+ "learning_rate": 1.1716511792515136e-05,
2527
+ "loss": 1.0103,
2528
+ "step": 4200
2529
+ },
2530
+ {
2531
+ "epoch": 0.61,
2532
+ "learning_rate": 1.1641958693287074e-05,
2533
+ "loss": 1.0284,
2534
+ "step": 4210
2535
+ },
2536
+ {
2537
+ "epoch": 0.62,
2538
+ "learning_rate": 1.1567492822794997e-05,
2539
+ "loss": 1.0103,
2540
+ "step": 4220
2541
+ },
2542
+ {
2543
+ "epoch": 0.62,
2544
+ "learning_rate": 1.1493116115370005e-05,
2545
+ "loss": 0.9962,
2546
+ "step": 4230
2547
+ },
2548
+ {
2549
+ "epoch": 0.62,
2550
+ "learning_rate": 1.1418830503027092e-05,
2551
+ "loss": 0.9914,
2552
+ "step": 4240
2553
+ },
2554
+ {
2555
+ "epoch": 0.62,
2556
+ "learning_rate": 1.1344637915414975e-05,
2557
+ "loss": 1.0264,
2558
+ "step": 4250
2559
+ },
2560
+ {
2561
+ "epoch": 0.62,
2562
+ "learning_rate": 1.1270540279765934e-05,
2563
+ "loss": 1.0131,
2564
+ "step": 4260
2565
+ },
2566
+ {
2567
+ "epoch": 0.62,
2568
+ "learning_rate": 1.119653952084579e-05,
2569
+ "loss": 1.0106,
2570
+ "step": 4270
2571
+ },
2572
+ {
2573
+ "epoch": 0.62,
2574
+ "learning_rate": 1.112263756090387e-05,
2575
+ "loss": 1.0005,
2576
+ "step": 4280
2577
+ },
2578
+ {
2579
+ "epoch": 0.63,
2580
+ "learning_rate": 1.1048836319623097e-05,
2581
+ "loss": 1.0377,
2582
+ "step": 4290
2583
+ },
2584
+ {
2585
+ "epoch": 0.63,
2586
+ "learning_rate": 1.0975137714070112e-05,
2587
+ "loss": 0.9819,
2588
+ "step": 4300
2589
+ },
2590
+ {
2591
+ "epoch": 0.63,
2592
+ "learning_rate": 1.0901543658645505e-05,
2593
+ "loss": 1.0004,
2594
+ "step": 4310
2595
+ },
2596
+ {
2597
+ "epoch": 0.63,
2598
+ "learning_rate": 1.082805606503404e-05,
2599
+ "loss": 0.9984,
2600
+ "step": 4320
2601
+ },
2602
+ {
2603
+ "epoch": 0.63,
2604
+ "learning_rate": 1.0754676842155023e-05,
2605
+ "loss": 0.9955,
2606
+ "step": 4330
2607
+ },
2608
+ {
2609
+ "epoch": 0.63,
2610
+ "learning_rate": 1.068140789611273e-05,
2611
+ "loss": 1.0401,
2612
+ "step": 4340
2613
+ },
2614
+ {
2615
+ "epoch": 0.64,
2616
+ "learning_rate": 1.0608251130146855e-05,
2617
+ "loss": 0.9783,
2618
+ "step": 4350
2619
+ },
2620
+ {
2621
+ "epoch": 0.64,
2622
+ "learning_rate": 1.053520844458312e-05,
2623
+ "loss": 1.0094,
2624
+ "step": 4360
2625
+ },
2626
+ {
2627
+ "epoch": 0.64,
2628
+ "learning_rate": 1.0462281736783865e-05,
2629
+ "loss": 1.018,
2630
+ "step": 4370
2631
+ },
2632
+ {
2633
+ "epoch": 0.64,
2634
+ "learning_rate": 1.0389472901098789e-05,
2635
+ "loss": 1.0326,
2636
+ "step": 4380
2637
+ },
2638
+ {
2639
+ "epoch": 0.64,
2640
+ "learning_rate": 1.031678382881573e-05,
2641
+ "loss": 1.0061,
2642
+ "step": 4390
2643
+ },
2644
+ {
2645
+ "epoch": 0.64,
2646
+ "learning_rate": 1.0244216408111565e-05,
2647
+ "loss": 1.027,
2648
+ "step": 4400
2649
+ },
2650
+ {
2651
+ "epoch": 0.64,
2652
+ "learning_rate": 1.0171772524003114e-05,
2653
+ "loss": 1.0247,
2654
+ "step": 4410
2655
+ },
2656
+ {
2657
+ "epoch": 0.65,
2658
+ "learning_rate": 1.0099454058298211e-05,
2659
+ "loss": 1.0366,
2660
+ "step": 4420
2661
+ },
2662
+ {
2663
+ "epoch": 0.65,
2664
+ "learning_rate": 1.0027262889546819e-05,
2665
+ "loss": 0.996,
2666
+ "step": 4430
2667
+ },
2668
+ {
2669
+ "epoch": 0.65,
2670
+ "learning_rate": 9.955200892992209e-06,
2671
+ "loss": 0.9731,
2672
+ "step": 4440
2673
+ },
2674
+ {
2675
+ "epoch": 0.65,
2676
+ "learning_rate": 9.883269940522289e-06,
2677
+ "loss": 1.0416,
2678
+ "step": 4450
2679
+ },
2680
+ {
2681
+ "epoch": 0.65,
2682
+ "learning_rate": 9.811471900620935e-06,
2683
+ "loss": 1.0342,
2684
+ "step": 4460
2685
+ },
2686
+ {
2687
+ "epoch": 0.65,
2688
+ "learning_rate": 9.739808638319485e-06,
2689
+ "loss": 1.0043,
2690
+ "step": 4470
2691
+ },
2692
+ {
2693
+ "epoch": 0.65,
2694
+ "learning_rate": 9.668282015148282e-06,
2695
+ "loss": 0.9578,
2696
+ "step": 4480
2697
+ },
2698
+ {
2699
+ "epoch": 0.66,
2700
+ "learning_rate": 9.596893889088311e-06,
2701
+ "loss": 0.9867,
2702
+ "step": 4490
2703
+ },
2704
+ {
2705
+ "epoch": 0.66,
2706
+ "learning_rate": 9.525646114522974e-06,
2707
+ "loss": 0.9794,
2708
+ "step": 4500
2709
+ },
2710
+ {
2711
+ "epoch": 0.66,
2712
+ "learning_rate": 9.454540542189862e-06,
2713
+ "loss": 1.0271,
2714
+ "step": 4510
2715
+ },
2716
+ {
2717
+ "epoch": 0.66,
2718
+ "learning_rate": 9.383579019132725e-06,
2719
+ "loss": 1.0165,
2720
+ "step": 4520
2721
+ },
2722
+ {
2723
+ "epoch": 0.66,
2724
+ "learning_rate": 9.31276338865347e-06,
2725
+ "loss": 0.9486,
2726
+ "step": 4530
2727
+ },
2728
+ {
2729
+ "epoch": 0.66,
2730
+ "learning_rate": 9.242095490264293e-06,
2731
+ "loss": 0.9823,
2732
+ "step": 4540
2733
+ },
2734
+ {
2735
+ "epoch": 0.66,
2736
+ "learning_rate": 9.171577159639897e-06,
2737
+ "loss": 1.0255,
2738
+ "step": 4550
2739
+ },
2740
+ {
2741
+ "epoch": 0.67,
2742
+ "learning_rate": 9.101210228569787e-06,
2743
+ "loss": 0.9583,
2744
+ "step": 4560
2745
+ },
2746
+ {
2747
+ "epoch": 0.67,
2748
+ "learning_rate": 9.030996524910705e-06,
2749
+ "loss": 1.0067,
2750
+ "step": 4570
2751
+ },
2752
+ {
2753
+ "epoch": 0.67,
2754
+ "learning_rate": 8.960937872539146e-06,
2755
+ "loss": 1.0157,
2756
+ "step": 4580
2757
+ },
2758
+ {
2759
+ "epoch": 0.67,
2760
+ "learning_rate": 8.891036091303997e-06,
2761
+ "loss": 1.0111,
2762
+ "step": 4590
2763
+ },
2764
+ {
2765
+ "epoch": 0.67,
2766
+ "learning_rate": 8.821292996979227e-06,
2767
+ "loss": 0.985,
2768
+ "step": 4600
2769
+ },
2770
+ {
2771
+ "epoch": 0.67,
2772
+ "learning_rate": 8.751710401216751e-06,
2773
+ "loss": 0.9972,
2774
+ "step": 4610
2775
+ },
2776
+ {
2777
+ "epoch": 0.67,
2778
+ "learning_rate": 8.68229011149935e-06,
2779
+ "loss": 1.0101,
2780
+ "step": 4620
2781
+ },
2782
+ {
2783
+ "epoch": 0.68,
2784
+ "learning_rate": 8.613033931093765e-06,
2785
+ "loss": 1.0335,
2786
+ "step": 4630
2787
+ },
2788
+ {
2789
+ "epoch": 0.68,
2790
+ "learning_rate": 8.543943659003778e-06,
2791
+ "loss": 0.9619,
2792
+ "step": 4640
2793
+ },
2794
+ {
2795
+ "epoch": 0.68,
2796
+ "learning_rate": 8.475021089923549e-06,
2797
+ "loss": 0.9699,
2798
+ "step": 4650
2799
+ },
2800
+ {
2801
+ "epoch": 0.68,
2802
+ "learning_rate": 8.406268014190972e-06,
2803
+ "loss": 1.0082,
2804
+ "step": 4660
2805
+ },
2806
+ {
2807
+ "epoch": 0.68,
2808
+ "learning_rate": 8.337686217741144e-06,
2809
+ "loss": 1.0069,
2810
+ "step": 4670
2811
+ },
2812
+ {
2813
+ "epoch": 0.68,
2814
+ "learning_rate": 8.269277482060041e-06,
2815
+ "loss": 0.9989,
2816
+ "step": 4680
2817
+ },
2818
+ {
2819
+ "epoch": 0.68,
2820
+ "learning_rate": 8.201043584138162e-06,
2821
+ "loss": 1.024,
2822
+ "step": 4690
2823
+ },
2824
+ {
2825
+ "epoch": 0.69,
2826
+ "learning_rate": 8.13298629642443e-06,
2827
+ "loss": 0.9841,
2828
+ "step": 4700
2829
+ },
2830
+ {
2831
+ "epoch": 0.69,
2832
+ "learning_rate": 8.0651073867801e-06,
2833
+ "loss": 1.0588,
2834
+ "step": 4710
2835
+ },
2836
+ {
2837
+ "epoch": 0.69,
2838
+ "learning_rate": 7.997408618432906e-06,
2839
+ "loss": 1.0106,
2840
+ "step": 4720
2841
+ },
2842
+ {
2843
+ "epoch": 0.69,
2844
+ "learning_rate": 7.929891749931181e-06,
2845
+ "loss": 1.0286,
2846
+ "step": 4730
2847
+ },
2848
+ {
2849
+ "epoch": 0.69,
2850
+ "learning_rate": 7.862558535098233e-06,
2851
+ "loss": 0.9935,
2852
+ "step": 4740
2853
+ },
2854
+ {
2855
+ "epoch": 0.69,
2856
+ "learning_rate": 7.795410722986768e-06,
2857
+ "loss": 1.0016,
2858
+ "step": 4750
2859
+ },
2860
+ {
2861
+ "epoch": 0.69,
2862
+ "learning_rate": 7.728450057833437e-06,
2863
+ "loss": 1.0029,
2864
+ "step": 4760
2865
+ },
2866
+ {
2867
+ "epoch": 0.7,
2868
+ "learning_rate": 7.66167827901358e-06,
2869
+ "loss": 0.963,
2870
+ "step": 4770
2871
+ },
2872
+ {
2873
+ "epoch": 0.7,
2874
+ "learning_rate": 7.5950971209959854e-06,
2875
+ "loss": 0.9964,
2876
+ "step": 4780
2877
+ },
2878
+ {
2879
+ "epoch": 0.7,
2880
+ "learning_rate": 7.528708313297875e-06,
2881
+ "loss": 0.9823,
2882
+ "step": 4790
2883
+ },
2884
+ {
2885
+ "epoch": 0.7,
2886
+ "learning_rate": 7.4625135804399504e-06,
2887
+ "loss": 1.003,
2888
+ "step": 4800
2889
+ },
2890
+ {
2891
+ "epoch": 0.7,
2892
+ "learning_rate": 7.3965146419016245e-06,
2893
+ "loss": 1.0053,
2894
+ "step": 4810
2895
+ },
2896
+ {
2897
+ "epoch": 0.7,
2898
+ "learning_rate": 7.330713212076333e-06,
2899
+ "loss": 1.0341,
2900
+ "step": 4820
2901
+ },
2902
+ {
2903
+ "epoch": 0.71,
2904
+ "learning_rate": 7.265111000227018e-06,
2905
+ "loss": 0.9492,
2906
+ "step": 4830
2907
+ },
2908
+ {
2909
+ "epoch": 0.71,
2910
+ "learning_rate": 7.199709710441719e-06,
2911
+ "loss": 1.0086,
2912
+ "step": 4840
2913
+ },
2914
+ {
2915
+ "epoch": 0.71,
2916
+ "learning_rate": 7.134511041589295e-06,
2917
+ "loss": 1.0015,
2918
+ "step": 4850
2919
+ },
2920
+ {
2921
+ "epoch": 0.71,
2922
+ "learning_rate": 7.069516687275326e-06,
2923
+ "loss": 0.9316,
2924
+ "step": 4860
2925
+ },
2926
+ {
2927
+ "epoch": 0.71,
2928
+ "learning_rate": 7.004728335798092e-06,
2929
+ "loss": 1.0404,
2930
+ "step": 4870
2931
+ },
2932
+ {
2933
+ "epoch": 0.71,
2934
+ "learning_rate": 6.940147670104741e-06,
2935
+ "loss": 1.032,
2936
+ "step": 4880
2937
+ },
2938
+ {
2939
+ "epoch": 0.71,
2940
+ "learning_rate": 6.875776367747534e-06,
2941
+ "loss": 1.0138,
2942
+ "step": 4890
2943
+ },
2944
+ {
2945
+ "epoch": 0.72,
2946
+ "learning_rate": 6.811616100840315e-06,
2947
+ "loss": 1.0512,
2948
+ "step": 4900
2949
+ },
2950
+ {
2951
+ "epoch": 0.72,
2952
+ "learning_rate": 6.74766853601505e-06,
2953
+ "loss": 0.9984,
2954
+ "step": 4910
2955
+ },
2956
+ {
2957
+ "epoch": 0.72,
2958
+ "learning_rate": 6.683935334378546e-06,
2959
+ "loss": 1.0459,
2960
+ "step": 4920
2961
+ },
2962
+ {
2963
+ "epoch": 0.72,
2964
+ "learning_rate": 6.620418151469274e-06,
2965
+ "loss": 0.9889,
2966
+ "step": 4930
2967
+ },
2968
+ {
2969
+ "epoch": 0.72,
2970
+ "learning_rate": 6.5571186372144095e-06,
2971
+ "loss": 1.0057,
2972
+ "step": 4940
2973
+ },
2974
+ {
2975
+ "epoch": 0.72,
2976
+ "learning_rate": 6.494038435886937e-06,
2977
+ "loss": 1.0167,
2978
+ "step": 4950
2979
+ },
2980
+ {
2981
+ "epoch": 0.72,
2982
+ "learning_rate": 6.431179186062962e-06,
2983
+ "loss": 0.9865,
2984
+ "step": 4960
2985
+ },
2986
+ {
2987
+ "epoch": 0.73,
2988
+ "learning_rate": 6.368542520579131e-06,
2989
+ "loss": 0.9901,
2990
+ "step": 4970
2991
+ },
2992
+ {
2993
+ "epoch": 0.73,
2994
+ "learning_rate": 6.306130066490215e-06,
2995
+ "loss": 1.0113,
2996
+ "step": 4980
2997
+ },
2998
+ {
2999
+ "epoch": 0.73,
3000
+ "learning_rate": 6.2439434450268675e-06,
3001
+ "loss": 0.9662,
3002
+ "step": 4990
3003
+ },
3004
+ {
3005
+ "epoch": 0.73,
3006
+ "learning_rate": 6.181984271553485e-06,
3007
+ "loss": 0.9991,
3008
+ "step": 5000
3009
+ }
3010
+ ],
3011
+ "max_steps": 6849,
3012
+ "num_train_epochs": 1,
3013
+ "total_flos": 0.0,
3014
+ "trial_name": null,
3015
+ "trial_params": null
3016
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_2x/checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,3616 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.8760402978537013,
5
+ "global_step": 6000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.3795620437956206e-07,
13
+ "loss": 1.2883,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 8.759124087591241e-07,
19
+ "loss": 1.3105,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.0,
24
+ "learning_rate": 1.3138686131386862e-06,
25
+ "loss": 1.3972,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 1.7518248175182482e-06,
31
+ "loss": 1.2908,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 2.1897810218978103e-06,
37
+ "loss": 1.2576,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 2.6277372262773724e-06,
43
+ "loss": 1.2685,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.01,
48
+ "learning_rate": 3.0656934306569344e-06,
49
+ "loss": 1.3179,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.01,
54
+ "learning_rate": 3.5036496350364965e-06,
55
+ "loss": 1.289,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.01,
60
+ "learning_rate": 3.9416058394160585e-06,
61
+ "loss": 1.2187,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.01,
66
+ "learning_rate": 4.379562043795621e-06,
67
+ "loss": 1.2577,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 4.817518248175183e-06,
73
+ "loss": 1.2545,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.02,
78
+ "learning_rate": 5.255474452554745e-06,
79
+ "loss": 1.2564,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.02,
84
+ "learning_rate": 5.693430656934307e-06,
85
+ "loss": 1.3263,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.02,
90
+ "learning_rate": 6.131386861313869e-06,
91
+ "loss": 1.2394,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.02,
96
+ "learning_rate": 6.569343065693431e-06,
97
+ "loss": 1.2536,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.02,
102
+ "learning_rate": 7.007299270072993e-06,
103
+ "loss": 1.221,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.02,
108
+ "learning_rate": 7.445255474452555e-06,
109
+ "loss": 1.2291,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.03,
114
+ "learning_rate": 7.883211678832117e-06,
115
+ "loss": 1.256,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.03,
120
+ "learning_rate": 8.321167883211678e-06,
121
+ "loss": 1.2402,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.03,
126
+ "learning_rate": 8.759124087591241e-06,
127
+ "loss": 1.2466,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.03,
132
+ "learning_rate": 9.197080291970804e-06,
133
+ "loss": 1.2007,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.03,
138
+ "learning_rate": 9.635036496350365e-06,
139
+ "loss": 1.2417,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.03,
144
+ "learning_rate": 1.0072992700729927e-05,
145
+ "loss": 1.2404,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.04,
150
+ "learning_rate": 1.051094890510949e-05,
151
+ "loss": 1.2551,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.04,
156
+ "learning_rate": 1.094890510948905e-05,
157
+ "loss": 1.169,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.04,
162
+ "learning_rate": 1.1386861313868614e-05,
163
+ "loss": 1.1858,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.04,
168
+ "learning_rate": 1.1824817518248176e-05,
169
+ "loss": 1.1528,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.04,
174
+ "learning_rate": 1.2262773722627738e-05,
175
+ "loss": 1.1979,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.04,
180
+ "learning_rate": 1.2700729927007299e-05,
181
+ "loss": 1.2189,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.04,
186
+ "learning_rate": 1.3138686131386862e-05,
187
+ "loss": 1.177,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.05,
192
+ "learning_rate": 1.3576642335766423e-05,
193
+ "loss": 1.1802,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.05,
198
+ "learning_rate": 1.4014598540145986e-05,
199
+ "loss": 1.1951,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.05,
204
+ "learning_rate": 1.4452554744525549e-05,
205
+ "loss": 1.15,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.05,
210
+ "learning_rate": 1.489051094890511e-05,
211
+ "loss": 1.2175,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.05,
216
+ "learning_rate": 1.5328467153284673e-05,
217
+ "loss": 1.221,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.05,
222
+ "learning_rate": 1.5766423357664234e-05,
223
+ "loss": 1.1714,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.05,
228
+ "learning_rate": 1.6204379562043795e-05,
229
+ "loss": 1.1593,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.06,
234
+ "learning_rate": 1.6642335766423357e-05,
235
+ "loss": 1.1302,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.06,
240
+ "learning_rate": 1.7080291970802918e-05,
241
+ "loss": 1.1517,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.06,
246
+ "learning_rate": 1.7518248175182482e-05,
247
+ "loss": 1.1623,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.06,
252
+ "learning_rate": 1.7956204379562047e-05,
253
+ "loss": 1.1689,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.06,
258
+ "learning_rate": 1.8394160583941608e-05,
259
+ "loss": 1.1682,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.06,
264
+ "learning_rate": 1.883211678832117e-05,
265
+ "loss": 1.1459,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.06,
270
+ "learning_rate": 1.927007299270073e-05,
271
+ "loss": 1.183,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.07,
276
+ "learning_rate": 1.9708029197080292e-05,
277
+ "loss": 1.1587,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.07,
282
+ "learning_rate": 2.0145985401459853e-05,
283
+ "loss": 1.1588,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.07,
288
+ "learning_rate": 2.0583941605839418e-05,
289
+ "loss": 1.1047,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.07,
294
+ "learning_rate": 2.102189781021898e-05,
295
+ "loss": 1.1835,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.07,
300
+ "learning_rate": 2.145985401459854e-05,
301
+ "loss": 1.0676,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.07,
306
+ "learning_rate": 2.18978102189781e-05,
307
+ "loss": 1.135,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.07,
312
+ "learning_rate": 2.2335766423357663e-05,
313
+ "loss": 1.1218,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.08,
318
+ "learning_rate": 2.2773722627737227e-05,
319
+ "loss": 1.1111,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.08,
324
+ "learning_rate": 2.3211678832116792e-05,
325
+ "loss": 1.1311,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.08,
330
+ "learning_rate": 2.3649635036496353e-05,
331
+ "loss": 1.119,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.08,
336
+ "learning_rate": 2.4087591240875914e-05,
337
+ "loss": 1.1116,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.08,
342
+ "learning_rate": 2.4525547445255475e-05,
343
+ "loss": 1.1061,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.08,
348
+ "learning_rate": 2.4963503649635037e-05,
349
+ "loss": 1.0921,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.08,
354
+ "learning_rate": 2.5401459854014598e-05,
355
+ "loss": 1.1249,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.09,
360
+ "learning_rate": 2.583941605839416e-05,
361
+ "loss": 1.0942,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.09,
366
+ "learning_rate": 2.6277372262773724e-05,
367
+ "loss": 1.1066,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.09,
372
+ "learning_rate": 2.6715328467153285e-05,
373
+ "loss": 1.0921,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.09,
378
+ "learning_rate": 2.7153284671532846e-05,
379
+ "loss": 1.0947,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.09,
384
+ "learning_rate": 2.759124087591241e-05,
385
+ "loss": 1.0746,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.09,
390
+ "learning_rate": 2.8029197080291972e-05,
391
+ "loss": 1.1277,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.09,
396
+ "learning_rate": 2.8467153284671533e-05,
397
+ "loss": 1.1284,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.1,
402
+ "learning_rate": 2.8905109489051098e-05,
403
+ "loss": 1.0711,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.1,
408
+ "learning_rate": 2.934306569343066e-05,
409
+ "loss": 1.1079,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.1,
414
+ "learning_rate": 2.978102189781022e-05,
415
+ "loss": 1.0963,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.1,
420
+ "learning_rate": 2.9999951294778548e-05,
421
+ "loss": 1.1275,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.1,
426
+ "learning_rate": 2.9999561654904685e-05,
427
+ "loss": 1.13,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.1,
432
+ "learning_rate": 2.9998782385278275e-05,
433
+ "loss": 1.1415,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.11,
438
+ "learning_rate": 2.9997613506141683e-05,
439
+ "loss": 1.0811,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.11,
444
+ "learning_rate": 2.9996055047857808e-05,
445
+ "loss": 1.0885,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.11,
450
+ "learning_rate": 2.999410705090927e-05,
451
+ "loss": 1.112,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.11,
456
+ "learning_rate": 2.999176956589739e-05,
457
+ "loss": 1.0727,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.11,
462
+ "learning_rate": 2.9989042653540863e-05,
463
+ "loss": 1.1176,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.11,
468
+ "learning_rate": 2.9985926384674167e-05,
469
+ "loss": 1.1499,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.11,
474
+ "learning_rate": 2.9982420840245734e-05,
475
+ "loss": 1.082,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.12,
480
+ "learning_rate": 2.9978526111315856e-05,
481
+ "loss": 1.1111,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.12,
486
+ "learning_rate": 2.997424229905431e-05,
487
+ "loss": 1.1424,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.12,
492
+ "learning_rate": 2.9969569514737728e-05,
493
+ "loss": 1.1018,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.12,
498
+ "learning_rate": 2.9964507879746708e-05,
499
+ "loss": 1.0874,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.12,
504
+ "learning_rate": 2.9959057525562664e-05,
505
+ "loss": 1.106,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.12,
510
+ "learning_rate": 2.9953218593764413e-05,
511
+ "loss": 1.1178,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.12,
516
+ "learning_rate": 2.994699123602449e-05,
517
+ "loss": 1.1031,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.13,
522
+ "learning_rate": 2.9940375614105206e-05,
523
+ "loss": 1.0894,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.13,
528
+ "learning_rate": 2.9933371899854472e-05,
529
+ "loss": 1.1483,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.13,
534
+ "learning_rate": 2.992598027520129e-05,
535
+ "loss": 1.0965,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.13,
540
+ "learning_rate": 2.991820093215107e-05,
541
+ "loss": 1.1053,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.13,
546
+ "learning_rate": 2.9910034072780608e-05,
547
+ "loss": 1.0736,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.13,
552
+ "learning_rate": 2.990147990923287e-05,
553
+ "loss": 1.0897,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.13,
558
+ "learning_rate": 2.9892538663711463e-05,
559
+ "loss": 1.0914,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.14,
564
+ "learning_rate": 2.988321056847485e-05,
565
+ "loss": 1.0944,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.14,
570
+ "learning_rate": 2.987349586583035e-05,
571
+ "loss": 1.076,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.14,
576
+ "learning_rate": 2.9863394808127805e-05,
577
+ "loss": 1.0562,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.14,
582
+ "learning_rate": 2.985290765775307e-05,
583
+ "loss": 1.1115,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.14,
588
+ "learning_rate": 2.9842034687121153e-05,
589
+ "loss": 1.0514,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.14,
594
+ "learning_rate": 2.9830776178669162e-05,
595
+ "loss": 1.0988,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.14,
600
+ "learning_rate": 2.9819132424848973e-05,
601
+ "loss": 1.0696,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.15,
606
+ "learning_rate": 2.9807103728119618e-05,
607
+ "loss": 1.0758,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.15,
612
+ "learning_rate": 2.9794690400939434e-05,
613
+ "loss": 1.0612,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.15,
618
+ "learning_rate": 2.9781892765757952e-05,
619
+ "loss": 1.0428,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.15,
624
+ "learning_rate": 2.976871115500752e-05,
625
+ "loss": 1.08,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.15,
630
+ "learning_rate": 2.9755145911094648e-05,
631
+ "loss": 1.0708,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.15,
636
+ "learning_rate": 2.9741197386391156e-05,
637
+ "loss": 1.1058,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.15,
642
+ "learning_rate": 2.972686594322497e-05,
643
+ "loss": 1.0646,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.16,
648
+ "learning_rate": 2.971215195387075e-05,
649
+ "loss": 1.0603,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.16,
654
+ "learning_rate": 2.96970558005402e-05,
655
+ "loss": 1.0867,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.16,
660
+ "learning_rate": 2.9681577875372146e-05,
661
+ "loss": 1.1372,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.16,
666
+ "learning_rate": 2.966571858042235e-05,
667
+ "loss": 1.0973,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.16,
672
+ "learning_rate": 2.964947832765306e-05,
673
+ "loss": 1.1167,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.16,
678
+ "learning_rate": 2.9632857538922312e-05,
679
+ "loss": 1.084,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.16,
684
+ "learning_rate": 2.9615856645972982e-05,
685
+ "loss": 1.0874,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.17,
690
+ "learning_rate": 2.959847609042155e-05,
691
+ "loss": 1.0702,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.17,
696
+ "learning_rate": 2.9580716323746648e-05,
697
+ "loss": 1.0416,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.17,
702
+ "learning_rate": 2.9562577807277328e-05,
703
+ "loss": 1.0661,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.17,
708
+ "learning_rate": 2.9544061012181073e-05,
709
+ "loss": 1.1111,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.17,
714
+ "learning_rate": 2.9525166419451546e-05,
715
+ "loss": 1.1008,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.17,
720
+ "learning_rate": 2.9505894519896136e-05,
721
+ "loss": 1.0829,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.18,
726
+ "learning_rate": 2.948624581412316e-05,
727
+ "loss": 1.0617,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.18,
732
+ "learning_rate": 2.94662208125289e-05,
733
+ "loss": 1.0606,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.18,
738
+ "learning_rate": 2.9445820035284313e-05,
739
+ "loss": 1.0683,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.18,
744
+ "learning_rate": 2.9425044012321538e-05,
745
+ "loss": 1.0838,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.18,
750
+ "learning_rate": 2.940389328332012e-05,
751
+ "loss": 1.0626,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.18,
756
+ "learning_rate": 2.9382368397693015e-05,
757
+ "loss": 1.0625,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.18,
762
+ "learning_rate": 2.936046991457227e-05,
763
+ "loss": 1.0158,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.19,
768
+ "learning_rate": 2.933819840279456e-05,
769
+ "loss": 1.0347,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.19,
774
+ "learning_rate": 2.9315554440886357e-05,
775
+ "loss": 1.0357,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.19,
780
+ "learning_rate": 2.9292538617048938e-05,
781
+ "loss": 1.0939,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.19,
786
+ "learning_rate": 2.9269151529143092e-05,
787
+ "loss": 1.0294,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.19,
792
+ "learning_rate": 2.9245393784673586e-05,
793
+ "loss": 1.1038,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.19,
798
+ "learning_rate": 2.92212660007734e-05,
799
+ "loss": 1.0597,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.19,
804
+ "learning_rate": 2.9196768804187677e-05,
805
+ "loss": 1.0171,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.2,
810
+ "learning_rate": 2.917190283125746e-05,
811
+ "loss": 1.0761,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.2,
816
+ "learning_rate": 2.9146668727903143e-05,
817
+ "loss": 1.0433,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.2,
822
+ "learning_rate": 2.9121067149607717e-05,
823
+ "loss": 1.0268,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.2,
828
+ "learning_rate": 2.9095098761399722e-05,
829
+ "loss": 1.0745,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.2,
834
+ "learning_rate": 2.906876423783598e-05,
835
+ "loss": 1.0593,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.2,
840
+ "learning_rate": 2.9042064262984078e-05,
841
+ "loss": 1.0437,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.2,
846
+ "learning_rate": 2.9014999530404588e-05,
847
+ "loss": 1.0326,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.21,
852
+ "learning_rate": 2.898757074313306e-05,
853
+ "loss": 1.0709,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.21,
858
+ "learning_rate": 2.8959778613661753e-05,
859
+ "loss": 1.0558,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.21,
864
+ "learning_rate": 2.8931623863921136e-05,
865
+ "loss": 1.0621,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.21,
870
+ "learning_rate": 2.8903107225261125e-05,
871
+ "loss": 0.9988,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.21,
876
+ "learning_rate": 2.8874229438432085e-05,
877
+ "loss": 1.0354,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.21,
882
+ "learning_rate": 2.8844991253565612e-05,
883
+ "loss": 1.0501,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.21,
888
+ "learning_rate": 2.8815393430155006e-05,
889
+ "loss": 0.9936,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.22,
894
+ "learning_rate": 2.8785436737035577e-05,
895
+ "loss": 1.0557,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.22,
900
+ "learning_rate": 2.8755121952364665e-05,
901
+ "loss": 1.0624,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.22,
906
+ "learning_rate": 2.8724449863601413e-05,
907
+ "loss": 1.0811,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.22,
912
+ "learning_rate": 2.8693421267486333e-05,
913
+ "loss": 0.9973,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.22,
918
+ "learning_rate": 2.8662036970020586e-05,
919
+ "loss": 1.0887,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.22,
924
+ "learning_rate": 2.8630297786445075e-05,
925
+ "loss": 1.0397,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.22,
930
+ "learning_rate": 2.859820454121924e-05,
931
+ "loss": 1.0791,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.23,
936
+ "learning_rate": 2.856575806799965e-05,
937
+ "loss": 1.0323,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.23,
942
+ "learning_rate": 2.853295920961837e-05,
943
+ "loss": 1.0451,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.23,
948
+ "learning_rate": 2.8499808818061025e-05,
949
+ "loss": 1.0755,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.23,
954
+ "learning_rate": 2.846630775444471e-05,
955
+ "loss": 1.0316,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.23,
960
+ "learning_rate": 2.843245688899559e-05,
961
+ "loss": 1.0297,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.23,
966
+ "learning_rate": 2.839825710102632e-05,
967
+ "loss": 1.0021,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.24,
972
+ "learning_rate": 2.836370927891319e-05,
973
+ "loss": 1.0469,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.24,
978
+ "learning_rate": 2.8328814320073046e-05,
979
+ "loss": 1.002,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.24,
984
+ "learning_rate": 2.829357313094e-05,
985
+ "loss": 1.0441,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.24,
990
+ "learning_rate": 2.8257986626941845e-05,
991
+ "loss": 1.101,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.24,
996
+ "learning_rate": 2.822205573247632e-05,
997
+ "loss": 1.069,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.24,
1002
+ "learning_rate": 2.8185781380887062e-05,
1003
+ "loss": 1.0309,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.24,
1008
+ "learning_rate": 2.8149164514439382e-05,
1009
+ "loss": 1.0329,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.25,
1014
+ "learning_rate": 2.8112206084295792e-05,
1015
+ "loss": 1.0377,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.25,
1020
+ "learning_rate": 2.807490705049127e-05,
1021
+ "loss": 1.1056,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.25,
1026
+ "learning_rate": 2.8037268381908354e-05,
1027
+ "loss": 1.0733,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.25,
1032
+ "learning_rate": 2.7999291056251956e-05,
1033
+ "loss": 1.0134,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.25,
1038
+ "learning_rate": 2.7960976060023973e-05,
1039
+ "loss": 1.0467,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.25,
1044
+ "learning_rate": 2.792232438849766e-05,
1045
+ "loss": 1.0289,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.25,
1050
+ "learning_rate": 2.7883337045691765e-05,
1051
+ "loss": 0.9964,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.26,
1056
+ "learning_rate": 2.7844015044344466e-05,
1057
+ "loss": 1.0221,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.26,
1062
+ "learning_rate": 2.7804359405887057e-05,
1063
+ "loss": 1.027,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.26,
1068
+ "learning_rate": 2.7764371160417414e-05,
1069
+ "loss": 1.0273,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.26,
1074
+ "learning_rate": 2.7724051346673245e-05,
1075
+ "loss": 1.0745,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.26,
1080
+ "learning_rate": 2.768340101200508e-05,
1081
+ "loss": 1.0127,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.26,
1086
+ "learning_rate": 2.7642421212349106e-05,
1087
+ "loss": 1.0355,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.26,
1092
+ "learning_rate": 2.7601113012199715e-05,
1093
+ "loss": 0.9789,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.27,
1098
+ "learning_rate": 2.7559477484581847e-05,
1099
+ "loss": 1.0704,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.27,
1104
+ "learning_rate": 2.7517515711023125e-05,
1105
+ "loss": 1.0902,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.27,
1110
+ "learning_rate": 2.747522878152577e-05,
1111
+ "loss": 1.0068,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.27,
1116
+ "learning_rate": 2.7432617794538276e-05,
1117
+ "loss": 1.0375,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.27,
1122
+ "learning_rate": 2.7389683856926868e-05,
1123
+ "loss": 1.0262,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.27,
1128
+ "learning_rate": 2.734642808394678e-05,
1129
+ "loss": 1.0245,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.27,
1134
+ "learning_rate": 2.7302851599213247e-05,
1135
+ "loss": 0.997,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.28,
1140
+ "learning_rate": 2.725895553467235e-05,
1141
+ "loss": 1.0468,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.28,
1146
+ "learning_rate": 2.7214741030571592e-05,
1147
+ "loss": 1.026,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.28,
1152
+ "learning_rate": 2.7170209235430294e-05,
1153
+ "loss": 1.0155,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.28,
1158
+ "learning_rate": 2.712536130600974e-05,
1159
+ "loss": 1.0718,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.28,
1164
+ "learning_rate": 2.7080198407283158e-05,
1165
+ "loss": 1.045,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.28,
1170
+ "learning_rate": 2.7034721712405437e-05,
1171
+ "loss": 1.0383,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.28,
1176
+ "learning_rate": 2.6988932402682644e-05,
1177
+ "loss": 1.0196,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.29,
1182
+ "learning_rate": 2.694283166754137e-05,
1183
+ "loss": 1.0087,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.29,
1188
+ "learning_rate": 2.6896420704497824e-05,
1189
+ "loss": 1.0493,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.29,
1194
+ "learning_rate": 2.684970071912669e-05,
1195
+ "loss": 1.0588,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.29,
1200
+ "learning_rate": 2.6802672925029866e-05,
1201
+ "loss": 1.0106,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.29,
1206
+ "learning_rate": 2.6755338543804898e-05,
1207
+ "loss": 1.0489,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.29,
1212
+ "learning_rate": 2.6707698805013265e-05,
1213
+ "loss": 1.0583,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.29,
1218
+ "learning_rate": 2.6659754946148446e-05,
1219
+ "loss": 1.0133,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.3,
1224
+ "learning_rate": 2.6611508212603755e-05,
1225
+ "loss": 1.0838,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.3,
1230
+ "learning_rate": 2.6562959857640006e-05,
1231
+ "loss": 1.0223,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.3,
1236
+ "learning_rate": 2.6514111142352958e-05,
1237
+ "loss": 1.0647,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.3,
1242
+ "learning_rate": 2.6464963335640543e-05,
1243
+ "loss": 1.0354,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.3,
1248
+ "learning_rate": 2.6415517714169922e-05,
1249
+ "loss": 1.0518,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.3,
1254
+ "learning_rate": 2.6365775562344316e-05,
1255
+ "loss": 1.0197,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.31,
1260
+ "learning_rate": 2.6315738172269635e-05,
1261
+ "loss": 1.0214,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.31,
1266
+ "learning_rate": 2.6265406843720924e-05,
1267
+ "loss": 0.9732,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.31,
1272
+ "learning_rate": 2.62147828841086e-05,
1273
+ "loss": 1.0811,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.31,
1278
+ "learning_rate": 2.6163867608444473e-05,
1279
+ "loss": 1.031,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.31,
1284
+ "learning_rate": 2.6112662339307617e-05,
1285
+ "loss": 1.0148,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.31,
1290
+ "learning_rate": 2.6061168406809985e-05,
1291
+ "loss": 1.0187,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.31,
1296
+ "learning_rate": 2.6009387148561884e-05,
1297
+ "loss": 1.0356,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.32,
1302
+ "learning_rate": 2.5957319909637208e-05,
1303
+ "loss": 1.0386,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.32,
1308
+ "learning_rate": 2.590496804253851e-05,
1309
+ "loss": 1.0803,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.32,
1314
+ "learning_rate": 2.5852332907161862e-05,
1315
+ "loss": 1.0175,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.32,
1320
+ "learning_rate": 2.5799415870761545e-05,
1321
+ "loss": 1.0571,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.32,
1326
+ "learning_rate": 2.57462183079145e-05,
1327
+ "loss": 1.0155,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.32,
1332
+ "learning_rate": 2.5692741600484676e-05,
1333
+ "loss": 1.0745,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.32,
1338
+ "learning_rate": 2.563898713758708e-05,
1339
+ "loss": 1.0834,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.33,
1344
+ "learning_rate": 2.5584956315551716e-05,
1345
+ "loss": 1.0132,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.33,
1350
+ "learning_rate": 2.5530650537887327e-05,
1351
+ "loss": 1.0205,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.33,
1356
+ "learning_rate": 2.5476071215244915e-05,
1357
+ "loss": 1.0325,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.33,
1362
+ "learning_rate": 2.5421219765381107e-05,
1363
+ "loss": 1.0503,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.33,
1368
+ "learning_rate": 2.536609761312134e-05,
1369
+ "loss": 1.0507,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.33,
1374
+ "learning_rate": 2.5310706190322806e-05,
1375
+ "loss": 1.0492,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.33,
1380
+ "learning_rate": 2.5255046935837332e-05,
1381
+ "loss": 1.0016,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.34,
1386
+ "learning_rate": 2.5199121295473926e-05,
1387
+ "loss": 1.0152,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.34,
1392
+ "learning_rate": 2.5142930721961277e-05,
1393
+ "loss": 1.044,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.34,
1398
+ "learning_rate": 2.5086476674909983e-05,
1399
+ "loss": 1.0151,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.34,
1404
+ "learning_rate": 2.502976062077466e-05,
1405
+ "loss": 1.0248,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.34,
1410
+ "learning_rate": 2.4972784032815837e-05,
1411
+ "loss": 0.9813,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.34,
1416
+ "learning_rate": 2.4915548391061688e-05,
1417
+ "loss": 1.0249,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.34,
1422
+ "learning_rate": 2.4858055182269586e-05,
1423
+ "loss": 1.001,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.35,
1428
+ "learning_rate": 2.48003058998875e-05,
1429
+ "loss": 1.0554,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.35,
1434
+ "learning_rate": 2.474230204401516e-05,
1435
+ "loss": 1.0022,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.35,
1440
+ "learning_rate": 2.468404512136514e-05,
1441
+ "loss": 1.0518,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.35,
1446
+ "learning_rate": 2.4625536645223687e-05,
1447
+ "loss": 1.0541,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.35,
1452
+ "learning_rate": 2.4566778135411417e-05,
1453
+ "loss": 1.0201,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.35,
1458
+ "learning_rate": 2.4507771118243838e-05,
1459
+ "loss": 0.9828,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.35,
1464
+ "learning_rate": 2.4448517126491717e-05,
1465
+ "loss": 1.0663,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.36,
1470
+ "learning_rate": 2.438901769934124e-05,
1471
+ "loss": 1.0088,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.36,
1476
+ "learning_rate": 2.4329274382354043e-05,
1477
+ "loss": 1.0139,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.36,
1482
+ "learning_rate": 2.4269288727427066e-05,
1483
+ "loss": 1.0488,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.36,
1488
+ "learning_rate": 2.420906229275224e-05,
1489
+ "loss": 1.0568,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.36,
1494
+ "learning_rate": 2.4148596642775997e-05,
1495
+ "loss": 1.0521,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.36,
1500
+ "learning_rate": 2.408789334815866e-05,
1501
+ "loss": 0.9795,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.37,
1506
+ "learning_rate": 2.402695398573362e-05,
1507
+ "loss": 1.0708,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.37,
1512
+ "learning_rate": 2.3965780138466375e-05,
1513
+ "loss": 1.0154,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.37,
1518
+ "learning_rate": 2.3904373395413435e-05,
1519
+ "loss": 1.0395,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.37,
1524
+ "learning_rate": 2.3842735351681016e-05,
1525
+ "loss": 1.0331,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.37,
1530
+ "learning_rate": 2.3780867608383624e-05,
1531
+ "loss": 0.9829,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.37,
1536
+ "learning_rate": 2.3718771772602456e-05,
1537
+ "loss": 1.0569,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.37,
1542
+ "learning_rate": 2.365644945734366e-05,
1543
+ "loss": 1.0729,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.38,
1548
+ "learning_rate": 2.3593902281496424e-05,
1549
+ "loss": 1.0369,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.38,
1554
+ "learning_rate": 2.3531131869790948e-05,
1555
+ "loss": 0.9883,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.38,
1560
+ "learning_rate": 2.34681398527562e-05,
1561
+ "loss": 1.0359,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.38,
1566
+ "learning_rate": 2.3404927866677607e-05,
1567
+ "loss": 1.0336,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.38,
1572
+ "learning_rate": 2.3341497553554512e-05,
1573
+ "loss": 0.9892,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.38,
1578
+ "learning_rate": 2.3277850561057554e-05,
1579
+ "loss": 1.0212,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.38,
1584
+ "learning_rate": 2.3213988542485838e-05,
1585
+ "loss": 1.072,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.39,
1590
+ "learning_rate": 2.3149913156724005e-05,
1591
+ "loss": 0.9956,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.39,
1596
+ "learning_rate": 2.3085626068199145e-05,
1597
+ "loss": 1.0685,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.39,
1602
+ "learning_rate": 2.3021128946837557e-05,
1603
+ "loss": 1.0248,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.39,
1608
+ "learning_rate": 2.2956423468021352e-05,
1609
+ "loss": 1.017,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.39,
1614
+ "learning_rate": 2.2891511312544972e-05,
1615
+ "loss": 1.029,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.39,
1620
+ "learning_rate": 2.2826394166571485e-05,
1621
+ "loss": 1.0368,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.39,
1626
+ "learning_rate": 2.2761073721588825e-05,
1627
+ "loss": 0.961,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.4,
1632
+ "learning_rate": 2.2695551674365836e-05,
1633
+ "loss": 1.0061,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.4,
1638
+ "learning_rate": 2.2629829726908186e-05,
1639
+ "loss": 1.0595,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.4,
1644
+ "learning_rate": 2.2563909586414174e-05,
1645
+ "loss": 1.0482,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.4,
1650
+ "learning_rate": 2.2497792965230372e-05,
1651
+ "loss": 0.9802,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.4,
1656
+ "learning_rate": 2.2431481580807163e-05,
1657
+ "loss": 1.0083,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.4,
1662
+ "learning_rate": 2.2364977155654095e-05,
1663
+ "loss": 1.0171,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.4,
1668
+ "learning_rate": 2.2298281417295167e-05,
1669
+ "loss": 0.9965,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.41,
1674
+ "learning_rate": 2.2231396098223944e-05,
1675
+ "loss": 1.0747,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.41,
1680
+ "learning_rate": 2.2164322935858552e-05,
1681
+ "loss": 1.0038,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.41,
1686
+ "learning_rate": 2.2097063672496555e-05,
1687
+ "loss": 0.9683,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.41,
1692
+ "learning_rate": 2.2029620055269677e-05,
1693
+ "loss": 1.0157,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.41,
1698
+ "learning_rate": 2.196199383609845e-05,
1699
+ "loss": 1.0479,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.41,
1704
+ "learning_rate": 2.1894186771646672e-05,
1705
+ "loss": 0.9811,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.41,
1710
+ "learning_rate": 2.1826200623275802e-05,
1711
+ "loss": 1.0051,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.42,
1716
+ "learning_rate": 2.1758037156999185e-05,
1717
+ "loss": 1.0145,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.42,
1722
+ "learning_rate": 2.1689698143436208e-05,
1723
+ "loss": 0.9983,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.42,
1728
+ "learning_rate": 2.1621185357766267e-05,
1729
+ "loss": 1.0197,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.42,
1734
+ "learning_rate": 2.155250057968269e-05,
1735
+ "loss": 1.0619,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.42,
1740
+ "learning_rate": 2.1483645593346487e-05,
1741
+ "loss": 0.9677,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.42,
1746
+ "learning_rate": 2.1414622187340015e-05,
1747
+ "loss": 1.066,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.42,
1752
+ "learning_rate": 2.1345432154620506e-05,
1753
+ "loss": 1.0523,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.43,
1758
+ "learning_rate": 2.1276077292473502e-05,
1759
+ "loss": 1.0226,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.43,
1764
+ "learning_rate": 2.1206559402466176e-05,
1765
+ "loss": 1.0658,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.43,
1770
+ "learning_rate": 2.113688029040051e-05,
1771
+ "loss": 1.0519,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.43,
1776
+ "learning_rate": 2.106704176626641e-05,
1777
+ "loss": 1.0363,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.43,
1782
+ "learning_rate": 2.099704564419469e-05,
1783
+ "loss": 1.0182,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.43,
1788
+ "learning_rate": 2.0926893742409923e-05,
1789
+ "loss": 1.0752,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.44,
1794
+ "learning_rate": 2.0856587883183247e-05,
1795
+ "loss": 1.0286,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.44,
1800
+ "learning_rate": 2.078612989278499e-05,
1801
+ "loss": 1.0012,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.44,
1806
+ "learning_rate": 2.0715521601437265e-05,
1807
+ "loss": 1.0053,
1808
+ "step": 3000
1809
+ },
1810
+ {
1811
+ "epoch": 0.44,
1812
+ "learning_rate": 2.0644764843266413e-05,
1813
+ "loss": 1.0562,
1814
+ "step": 3010
1815
+ },
1816
+ {
1817
+ "epoch": 0.44,
1818
+ "learning_rate": 2.0573861456255354e-05,
1819
+ "loss": 0.9871,
1820
+ "step": 3020
1821
+ },
1822
+ {
1823
+ "epoch": 0.44,
1824
+ "learning_rate": 2.050281328219585e-05,
1825
+ "loss": 1.0166,
1826
+ "step": 3030
1827
+ },
1828
+ {
1829
+ "epoch": 0.44,
1830
+ "learning_rate": 2.0431622166640667e-05,
1831
+ "loss": 1.0534,
1832
+ "step": 3040
1833
+ },
1834
+ {
1835
+ "epoch": 0.45,
1836
+ "learning_rate": 2.0360289958855634e-05,
1837
+ "loss": 1.032,
1838
+ "step": 3050
1839
+ },
1840
+ {
1841
+ "epoch": 0.45,
1842
+ "learning_rate": 2.028881851177159e-05,
1843
+ "loss": 0.9949,
1844
+ "step": 3060
1845
+ },
1846
+ {
1847
+ "epoch": 0.45,
1848
+ "learning_rate": 2.021720968193628e-05,
1849
+ "loss": 0.9932,
1850
+ "step": 3070
1851
+ },
1852
+ {
1853
+ "epoch": 0.45,
1854
+ "learning_rate": 2.0145465329466092e-05,
1855
+ "loss": 0.9909,
1856
+ "step": 3080
1857
+ },
1858
+ {
1859
+ "epoch": 0.45,
1860
+ "learning_rate": 2.007358731799779e-05,
1861
+ "loss": 0.9935,
1862
+ "step": 3090
1863
+ },
1864
+ {
1865
+ "epoch": 0.45,
1866
+ "learning_rate": 2.0001577514640046e-05,
1867
+ "loss": 1.0379,
1868
+ "step": 3100
1869
+ },
1870
+ {
1871
+ "epoch": 0.45,
1872
+ "learning_rate": 1.9929437789924987e-05,
1873
+ "loss": 1.0241,
1874
+ "step": 3110
1875
+ },
1876
+ {
1877
+ "epoch": 0.46,
1878
+ "learning_rate": 1.9857170017759572e-05,
1879
+ "loss": 0.9976,
1880
+ "step": 3120
1881
+ },
1882
+ {
1883
+ "epoch": 0.46,
1884
+ "learning_rate": 1.978477607537695e-05,
1885
+ "loss": 1.0481,
1886
+ "step": 3130
1887
+ },
1888
+ {
1889
+ "epoch": 0.46,
1890
+ "learning_rate": 1.9712257843287656e-05,
1891
+ "loss": 1.0236,
1892
+ "step": 3140
1893
+ },
1894
+ {
1895
+ "epoch": 0.46,
1896
+ "learning_rate": 1.9639617205230798e-05,
1897
+ "loss": 1.0249,
1898
+ "step": 3150
1899
+ },
1900
+ {
1901
+ "epoch": 0.46,
1902
+ "learning_rate": 1.9566856048125102e-05,
1903
+ "loss": 1.0197,
1904
+ "step": 3160
1905
+ },
1906
+ {
1907
+ "epoch": 0.46,
1908
+ "learning_rate": 1.9493976262019916e-05,
1909
+ "loss": 1.0039,
1910
+ "step": 3170
1911
+ },
1912
+ {
1913
+ "epoch": 0.46,
1914
+ "learning_rate": 1.9420979740046095e-05,
1915
+ "loss": 1.041,
1916
+ "step": 3180
1917
+ },
1918
+ {
1919
+ "epoch": 0.47,
1920
+ "learning_rate": 1.9347868378366837e-05,
1921
+ "loss": 1.0347,
1922
+ "step": 3190
1923
+ },
1924
+ {
1925
+ "epoch": 0.47,
1926
+ "learning_rate": 1.9274644076128428e-05,
1927
+ "loss": 1.016,
1928
+ "step": 3200
1929
+ },
1930
+ {
1931
+ "epoch": 0.47,
1932
+ "learning_rate": 1.9201308735410895e-05,
1933
+ "loss": 1.041,
1934
+ "step": 3210
1935
+ },
1936
+ {
1937
+ "epoch": 0.47,
1938
+ "learning_rate": 1.9127864261178623e-05,
1939
+ "loss": 0.9939,
1940
+ "step": 3220
1941
+ },
1942
+ {
1943
+ "epoch": 0.47,
1944
+ "learning_rate": 1.9054312561230855e-05,
1945
+ "loss": 1.0674,
1946
+ "step": 3230
1947
+ },
1948
+ {
1949
+ "epoch": 0.47,
1950
+ "learning_rate": 1.8980655546152123e-05,
1951
+ "loss": 1.0005,
1952
+ "step": 3240
1953
+ },
1954
+ {
1955
+ "epoch": 0.47,
1956
+ "learning_rate": 1.8906895129262655e-05,
1957
+ "loss": 1.0059,
1958
+ "step": 3250
1959
+ },
1960
+ {
1961
+ "epoch": 0.48,
1962
+ "learning_rate": 1.8833033226568636e-05,
1963
+ "loss": 0.986,
1964
+ "step": 3260
1965
+ },
1966
+ {
1967
+ "epoch": 0.48,
1968
+ "learning_rate": 1.875907175671246e-05,
1969
+ "loss": 1.0224,
1970
+ "step": 3270
1971
+ },
1972
+ {
1973
+ "epoch": 0.48,
1974
+ "learning_rate": 1.8685012640922886e-05,
1975
+ "loss": 0.9967,
1976
+ "step": 3280
1977
+ },
1978
+ {
1979
+ "epoch": 0.48,
1980
+ "learning_rate": 1.8610857802965127e-05,
1981
+ "loss": 0.9802,
1982
+ "step": 3290
1983
+ },
1984
+ {
1985
+ "epoch": 0.48,
1986
+ "learning_rate": 1.853660916909089e-05,
1987
+ "loss": 1.0024,
1988
+ "step": 3300
1989
+ },
1990
+ {
1991
+ "epoch": 0.48,
1992
+ "learning_rate": 1.8462268667988316e-05,
1993
+ "loss": 1.0104,
1994
+ "step": 3310
1995
+ },
1996
+ {
1997
+ "epoch": 0.48,
1998
+ "learning_rate": 1.8387838230731915e-05,
1999
+ "loss": 0.9954,
2000
+ "step": 3320
2001
+ },
2002
+ {
2003
+ "epoch": 0.49,
2004
+ "learning_rate": 1.8313319790732373e-05,
2005
+ "loss": 0.9586,
2006
+ "step": 3330
2007
+ },
2008
+ {
2009
+ "epoch": 0.49,
2010
+ "learning_rate": 1.823871528368635e-05,
2011
+ "loss": 0.968,
2012
+ "step": 3340
2013
+ },
2014
+ {
2015
+ "epoch": 0.49,
2016
+ "learning_rate": 1.8164026647526183e-05,
2017
+ "loss": 0.9949,
2018
+ "step": 3350
2019
+ },
2020
+ {
2021
+ "epoch": 0.49,
2022
+ "learning_rate": 1.808925582236956e-05,
2023
+ "loss": 1.0205,
2024
+ "step": 3360
2025
+ },
2026
+ {
2027
+ "epoch": 0.49,
2028
+ "learning_rate": 1.8014404750469113e-05,
2029
+ "loss": 1.0236,
2030
+ "step": 3370
2031
+ },
2032
+ {
2033
+ "epoch": 0.49,
2034
+ "learning_rate": 1.7939475376161963e-05,
2035
+ "loss": 1.0316,
2036
+ "step": 3380
2037
+ },
2038
+ {
2039
+ "epoch": 0.49,
2040
+ "learning_rate": 1.7864469645819234e-05,
2041
+ "loss": 1.0432,
2042
+ "step": 3390
2043
+ },
2044
+ {
2045
+ "epoch": 0.5,
2046
+ "learning_rate": 1.778938950779547e-05,
2047
+ "loss": 1.0138,
2048
+ "step": 3400
2049
+ },
2050
+ {
2051
+ "epoch": 0.5,
2052
+ "learning_rate": 1.7714236912378038e-05,
2053
+ "loss": 1.0157,
2054
+ "step": 3410
2055
+ },
2056
+ {
2057
+ "epoch": 0.5,
2058
+ "learning_rate": 1.7639013811736458e-05,
2059
+ "loss": 1.0072,
2060
+ "step": 3420
2061
+ },
2062
+ {
2063
+ "epoch": 0.5,
2064
+ "learning_rate": 1.7563722159871712e-05,
2065
+ "loss": 0.996,
2066
+ "step": 3430
2067
+ },
2068
+ {
2069
+ "epoch": 0.5,
2070
+ "learning_rate": 1.7488363912565466e-05,
2071
+ "loss": 1.0315,
2072
+ "step": 3440
2073
+ },
2074
+ {
2075
+ "epoch": 0.5,
2076
+ "learning_rate": 1.7412941027329267e-05,
2077
+ "loss": 1.0605,
2078
+ "step": 3450
2079
+ },
2080
+ {
2081
+ "epoch": 0.51,
2082
+ "learning_rate": 1.733745546335372e-05,
2083
+ "loss": 1.0237,
2084
+ "step": 3460
2085
+ },
2086
+ {
2087
+ "epoch": 0.51,
2088
+ "learning_rate": 1.7261909181457554e-05,
2089
+ "loss": 1.012,
2090
+ "step": 3470
2091
+ },
2092
+ {
2093
+ "epoch": 0.51,
2094
+ "learning_rate": 1.718630414403674e-05,
2095
+ "loss": 0.987,
2096
+ "step": 3480
2097
+ },
2098
+ {
2099
+ "epoch": 0.51,
2100
+ "learning_rate": 1.7110642315013454e-05,
2101
+ "loss": 1.0672,
2102
+ "step": 3490
2103
+ },
2104
+ {
2105
+ "epoch": 0.51,
2106
+ "learning_rate": 1.7034925659785127e-05,
2107
+ "loss": 0.975,
2108
+ "step": 3500
2109
+ },
2110
+ {
2111
+ "epoch": 0.51,
2112
+ "learning_rate": 1.695915614517333e-05,
2113
+ "loss": 1.0022,
2114
+ "step": 3510
2115
+ },
2116
+ {
2117
+ "epoch": 0.51,
2118
+ "learning_rate": 1.6883335739372743e-05,
2119
+ "loss": 1.0443,
2120
+ "step": 3520
2121
+ },
2122
+ {
2123
+ "epoch": 0.52,
2124
+ "learning_rate": 1.680746641189998e-05,
2125
+ "loss": 0.9779,
2126
+ "step": 3530
2127
+ },
2128
+ {
2129
+ "epoch": 0.52,
2130
+ "learning_rate": 1.6731550133542443e-05,
2131
+ "loss": 0.9595,
2132
+ "step": 3540
2133
+ },
2134
+ {
2135
+ "epoch": 0.52,
2136
+ "learning_rate": 1.665558887630716e-05,
2137
+ "loss": 1.0116,
2138
+ "step": 3550
2139
+ },
2140
+ {
2141
+ "epoch": 0.52,
2142
+ "learning_rate": 1.65795846133695e-05,
2143
+ "loss": 1.0014,
2144
+ "step": 3560
2145
+ },
2146
+ {
2147
+ "epoch": 0.52,
2148
+ "learning_rate": 1.650353931902198e-05,
2149
+ "loss": 1.0118,
2150
+ "step": 3570
2151
+ },
2152
+ {
2153
+ "epoch": 0.52,
2154
+ "learning_rate": 1.6427454968622937e-05,
2155
+ "loss": 1.031,
2156
+ "step": 3580
2157
+ },
2158
+ {
2159
+ "epoch": 0.52,
2160
+ "learning_rate": 1.6351333538545236e-05,
2161
+ "loss": 0.9675,
2162
+ "step": 3590
2163
+ },
2164
+ {
2165
+ "epoch": 0.53,
2166
+ "learning_rate": 1.6275177006124916e-05,
2167
+ "loss": 1.0421,
2168
+ "step": 3600
2169
+ },
2170
+ {
2171
+ "epoch": 0.53,
2172
+ "learning_rate": 1.619898734960986e-05,
2173
+ "loss": 1.0175,
2174
+ "step": 3610
2175
+ },
2176
+ {
2177
+ "epoch": 0.53,
2178
+ "learning_rate": 1.6122766548108364e-05,
2179
+ "loss": 0.9907,
2180
+ "step": 3620
2181
+ },
2182
+ {
2183
+ "epoch": 0.53,
2184
+ "learning_rate": 1.6046516581537754e-05,
2185
+ "loss": 0.9671,
2186
+ "step": 3630
2187
+ },
2188
+ {
2189
+ "epoch": 0.53,
2190
+ "learning_rate": 1.5970239430572958e-05,
2191
+ "loss": 1.0066,
2192
+ "step": 3640
2193
+ },
2194
+ {
2195
+ "epoch": 0.53,
2196
+ "learning_rate": 1.5893937076595036e-05,
2197
+ "loss": 0.9498,
2198
+ "step": 3650
2199
+ },
2200
+ {
2201
+ "epoch": 0.53,
2202
+ "learning_rate": 1.581761150163973e-05,
2203
+ "loss": 1.02,
2204
+ "step": 3660
2205
+ },
2206
+ {
2207
+ "epoch": 0.54,
2208
+ "learning_rate": 1.5741264688345972e-05,
2209
+ "loss": 0.9901,
2210
+ "step": 3670
2211
+ },
2212
+ {
2213
+ "epoch": 0.54,
2214
+ "learning_rate": 1.5664898619904384e-05,
2215
+ "loss": 1.0409,
2216
+ "step": 3680
2217
+ },
2218
+ {
2219
+ "epoch": 0.54,
2220
+ "learning_rate": 1.5588515280005754e-05,
2221
+ "loss": 1.0504,
2222
+ "step": 3690
2223
+ },
2224
+ {
2225
+ "epoch": 0.54,
2226
+ "learning_rate": 1.551211665278952e-05,
2227
+ "loss": 1.0119,
2228
+ "step": 3700
2229
+ },
2230
+ {
2231
+ "epoch": 0.54,
2232
+ "learning_rate": 1.5435704722792224e-05,
2233
+ "loss": 1.0011,
2234
+ "step": 3710
2235
+ },
2236
+ {
2237
+ "epoch": 0.54,
2238
+ "learning_rate": 1.5359281474895966e-05,
2239
+ "loss": 0.9896,
2240
+ "step": 3720
2241
+ },
2242
+ {
2243
+ "epoch": 0.54,
2244
+ "learning_rate": 1.5282848894276837e-05,
2245
+ "loss": 1.0349,
2246
+ "step": 3730
2247
+ },
2248
+ {
2249
+ "epoch": 0.55,
2250
+ "learning_rate": 1.5206408966353343e-05,
2251
+ "loss": 0.9609,
2252
+ "step": 3740
2253
+ },
2254
+ {
2255
+ "epoch": 0.55,
2256
+ "learning_rate": 1.5129963676734874e-05,
2257
+ "loss": 1.0007,
2258
+ "step": 3750
2259
+ },
2260
+ {
2261
+ "epoch": 0.55,
2262
+ "learning_rate": 1.5053515011170068e-05,
2263
+ "loss": 1.0384,
2264
+ "step": 3760
2265
+ },
2266
+ {
2267
+ "epoch": 0.55,
2268
+ "learning_rate": 1.4977064955495265e-05,
2269
+ "loss": 1.0187,
2270
+ "step": 3770
2271
+ },
2272
+ {
2273
+ "epoch": 0.55,
2274
+ "learning_rate": 1.4900615495582916e-05,
2275
+ "loss": 0.9817,
2276
+ "step": 3780
2277
+ },
2278
+ {
2279
+ "epoch": 0.55,
2280
+ "learning_rate": 1.4824168617290009e-05,
2281
+ "loss": 0.9856,
2282
+ "step": 3790
2283
+ },
2284
+ {
2285
+ "epoch": 0.55,
2286
+ "learning_rate": 1.4747726306406444e-05,
2287
+ "loss": 1.01,
2288
+ "step": 3800
2289
+ },
2290
+ {
2291
+ "epoch": 0.56,
2292
+ "learning_rate": 1.4671290548603503e-05,
2293
+ "loss": 1.0264,
2294
+ "step": 3810
2295
+ },
2296
+ {
2297
+ "epoch": 0.56,
2298
+ "learning_rate": 1.4594863329382235e-05,
2299
+ "loss": 1.0105,
2300
+ "step": 3820
2301
+ },
2302
+ {
2303
+ "epoch": 0.56,
2304
+ "learning_rate": 1.4518446634021887e-05,
2305
+ "loss": 1.0075,
2306
+ "step": 3830
2307
+ },
2308
+ {
2309
+ "epoch": 0.56,
2310
+ "learning_rate": 1.444204244752834e-05,
2311
+ "loss": 1.0036,
2312
+ "step": 3840
2313
+ },
2314
+ {
2315
+ "epoch": 0.56,
2316
+ "learning_rate": 1.436565275458254e-05,
2317
+ "loss": 1.0064,
2318
+ "step": 3850
2319
+ },
2320
+ {
2321
+ "epoch": 0.56,
2322
+ "learning_rate": 1.428927953948896e-05,
2323
+ "loss": 0.9885,
2324
+ "step": 3860
2325
+ },
2326
+ {
2327
+ "epoch": 0.57,
2328
+ "learning_rate": 1.4212924786124033e-05,
2329
+ "loss": 0.9896,
2330
+ "step": 3870
2331
+ },
2332
+ {
2333
+ "epoch": 0.57,
2334
+ "learning_rate": 1.4136590477884626e-05,
2335
+ "loss": 1.0143,
2336
+ "step": 3880
2337
+ },
2338
+ {
2339
+ "epoch": 0.57,
2340
+ "learning_rate": 1.4060278597636528e-05,
2341
+ "loss": 1.0107,
2342
+ "step": 3890
2343
+ },
2344
+ {
2345
+ "epoch": 0.57,
2346
+ "learning_rate": 1.3983991127662936e-05,
2347
+ "loss": 0.968,
2348
+ "step": 3900
2349
+ },
2350
+ {
2351
+ "epoch": 0.57,
2352
+ "learning_rate": 1.3907730049612967e-05,
2353
+ "loss": 1.0048,
2354
+ "step": 3910
2355
+ },
2356
+ {
2357
+ "epoch": 0.57,
2358
+ "learning_rate": 1.3831497344450171e-05,
2359
+ "loss": 1.0245,
2360
+ "step": 3920
2361
+ },
2362
+ {
2363
+ "epoch": 0.57,
2364
+ "learning_rate": 1.3755294992401086e-05,
2365
+ "loss": 1.04,
2366
+ "step": 3930
2367
+ },
2368
+ {
2369
+ "epoch": 0.58,
2370
+ "learning_rate": 1.367912497290379e-05,
2371
+ "loss": 1.0232,
2372
+ "step": 3940
2373
+ },
2374
+ {
2375
+ "epoch": 0.58,
2376
+ "learning_rate": 1.3602989264556506e-05,
2377
+ "loss": 1.0247,
2378
+ "step": 3950
2379
+ },
2380
+ {
2381
+ "epoch": 0.58,
2382
+ "learning_rate": 1.3526889845066166e-05,
2383
+ "loss": 1.0201,
2384
+ "step": 3960
2385
+ },
2386
+ {
2387
+ "epoch": 0.58,
2388
+ "learning_rate": 1.345082869119707e-05,
2389
+ "loss": 0.9729,
2390
+ "step": 3970
2391
+ },
2392
+ {
2393
+ "epoch": 0.58,
2394
+ "learning_rate": 1.3374807778719532e-05,
2395
+ "loss": 0.9981,
2396
+ "step": 3980
2397
+ },
2398
+ {
2399
+ "epoch": 0.58,
2400
+ "learning_rate": 1.3298829082358541e-05,
2401
+ "loss": 0.9969,
2402
+ "step": 3990
2403
+ },
2404
+ {
2405
+ "epoch": 0.58,
2406
+ "learning_rate": 1.3222894575742489e-05,
2407
+ "loss": 1.0115,
2408
+ "step": 4000
2409
+ },
2410
+ {
2411
+ "epoch": 0.59,
2412
+ "learning_rate": 1.3147006231351884e-05,
2413
+ "loss": 1.0155,
2414
+ "step": 4010
2415
+ },
2416
+ {
2417
+ "epoch": 0.59,
2418
+ "learning_rate": 1.307116602046812e-05,
2419
+ "loss": 1.0244,
2420
+ "step": 4020
2421
+ },
2422
+ {
2423
+ "epoch": 0.59,
2424
+ "learning_rate": 1.2995375913122272e-05,
2425
+ "loss": 0.9805,
2426
+ "step": 4030
2427
+ },
2428
+ {
2429
+ "epoch": 0.59,
2430
+ "learning_rate": 1.291963787804393e-05,
2431
+ "loss": 1.0217,
2432
+ "step": 4040
2433
+ },
2434
+ {
2435
+ "epoch": 0.59,
2436
+ "learning_rate": 1.2843953882610043e-05,
2437
+ "loss": 1.0406,
2438
+ "step": 4050
2439
+ },
2440
+ {
2441
+ "epoch": 0.59,
2442
+ "learning_rate": 1.2768325892793819e-05,
2443
+ "loss": 1.056,
2444
+ "step": 4060
2445
+ },
2446
+ {
2447
+ "epoch": 0.59,
2448
+ "learning_rate": 1.2692755873113666e-05,
2449
+ "loss": 0.9892,
2450
+ "step": 4070
2451
+ },
2452
+ {
2453
+ "epoch": 0.6,
2454
+ "learning_rate": 1.2617245786582157e-05,
2455
+ "loss": 0.9933,
2456
+ "step": 4080
2457
+ },
2458
+ {
2459
+ "epoch": 0.6,
2460
+ "learning_rate": 1.2541797594655031e-05,
2461
+ "loss": 1.0152,
2462
+ "step": 4090
2463
+ },
2464
+ {
2465
+ "epoch": 0.6,
2466
+ "learning_rate": 1.2466413257180254e-05,
2467
+ "loss": 1.0015,
2468
+ "step": 4100
2469
+ },
2470
+ {
2471
+ "epoch": 0.6,
2472
+ "learning_rate": 1.2391094732347098e-05,
2473
+ "loss": 1.0352,
2474
+ "step": 4110
2475
+ },
2476
+ {
2477
+ "epoch": 0.6,
2478
+ "learning_rate": 1.2315843976635276e-05,
2479
+ "loss": 1.0022,
2480
+ "step": 4120
2481
+ },
2482
+ {
2483
+ "epoch": 0.6,
2484
+ "learning_rate": 1.2240662944764146e-05,
2485
+ "loss": 0.9675,
2486
+ "step": 4130
2487
+ },
2488
+ {
2489
+ "epoch": 0.6,
2490
+ "learning_rate": 1.2165553589641895e-05,
2491
+ "loss": 0.9959,
2492
+ "step": 4140
2493
+ },
2494
+ {
2495
+ "epoch": 0.61,
2496
+ "learning_rate": 1.2090517862314831e-05,
2497
+ "loss": 0.9988,
2498
+ "step": 4150
2499
+ },
2500
+ {
2501
+ "epoch": 0.61,
2502
+ "learning_rate": 1.2015557711916713e-05,
2503
+ "loss": 1.0156,
2504
+ "step": 4160
2505
+ },
2506
+ {
2507
+ "epoch": 0.61,
2508
+ "learning_rate": 1.1940675085618088e-05,
2509
+ "loss": 1.0343,
2510
+ "step": 4170
2511
+ },
2512
+ {
2513
+ "epoch": 0.61,
2514
+ "learning_rate": 1.1865871928575759e-05,
2515
+ "loss": 1.0788,
2516
+ "step": 4180
2517
+ },
2518
+ {
2519
+ "epoch": 0.61,
2520
+ "learning_rate": 1.1791150183882203e-05,
2521
+ "loss": 1.0417,
2522
+ "step": 4190
2523
+ },
2524
+ {
2525
+ "epoch": 0.61,
2526
+ "learning_rate": 1.1716511792515136e-05,
2527
+ "loss": 1.0103,
2528
+ "step": 4200
2529
+ },
2530
+ {
2531
+ "epoch": 0.61,
2532
+ "learning_rate": 1.1641958693287074e-05,
2533
+ "loss": 1.0284,
2534
+ "step": 4210
2535
+ },
2536
+ {
2537
+ "epoch": 0.62,
2538
+ "learning_rate": 1.1567492822794997e-05,
2539
+ "loss": 1.0103,
2540
+ "step": 4220
2541
+ },
2542
+ {
2543
+ "epoch": 0.62,
2544
+ "learning_rate": 1.1493116115370005e-05,
2545
+ "loss": 0.9962,
2546
+ "step": 4230
2547
+ },
2548
+ {
2549
+ "epoch": 0.62,
2550
+ "learning_rate": 1.1418830503027092e-05,
2551
+ "loss": 0.9914,
2552
+ "step": 4240
2553
+ },
2554
+ {
2555
+ "epoch": 0.62,
2556
+ "learning_rate": 1.1344637915414975e-05,
2557
+ "loss": 1.0264,
2558
+ "step": 4250
2559
+ },
2560
+ {
2561
+ "epoch": 0.62,
2562
+ "learning_rate": 1.1270540279765934e-05,
2563
+ "loss": 1.0131,
2564
+ "step": 4260
2565
+ },
2566
+ {
2567
+ "epoch": 0.62,
2568
+ "learning_rate": 1.119653952084579e-05,
2569
+ "loss": 1.0106,
2570
+ "step": 4270
2571
+ },
2572
+ {
2573
+ "epoch": 0.62,
2574
+ "learning_rate": 1.112263756090387e-05,
2575
+ "loss": 1.0005,
2576
+ "step": 4280
2577
+ },
2578
+ {
2579
+ "epoch": 0.63,
2580
+ "learning_rate": 1.1048836319623097e-05,
2581
+ "loss": 1.0377,
2582
+ "step": 4290
2583
+ },
2584
+ {
2585
+ "epoch": 0.63,
2586
+ "learning_rate": 1.0975137714070112e-05,
2587
+ "loss": 0.9819,
2588
+ "step": 4300
2589
+ },
2590
+ {
2591
+ "epoch": 0.63,
2592
+ "learning_rate": 1.0901543658645505e-05,
2593
+ "loss": 1.0004,
2594
+ "step": 4310
2595
+ },
2596
+ {
2597
+ "epoch": 0.63,
2598
+ "learning_rate": 1.082805606503404e-05,
2599
+ "loss": 0.9984,
2600
+ "step": 4320
2601
+ },
2602
+ {
2603
+ "epoch": 0.63,
2604
+ "learning_rate": 1.0754676842155023e-05,
2605
+ "loss": 0.9955,
2606
+ "step": 4330
2607
+ },
2608
+ {
2609
+ "epoch": 0.63,
2610
+ "learning_rate": 1.068140789611273e-05,
2611
+ "loss": 1.0401,
2612
+ "step": 4340
2613
+ },
2614
+ {
2615
+ "epoch": 0.64,
2616
+ "learning_rate": 1.0608251130146855e-05,
2617
+ "loss": 0.9783,
2618
+ "step": 4350
2619
+ },
2620
+ {
2621
+ "epoch": 0.64,
2622
+ "learning_rate": 1.053520844458312e-05,
2623
+ "loss": 1.0094,
2624
+ "step": 4360
2625
+ },
2626
+ {
2627
+ "epoch": 0.64,
2628
+ "learning_rate": 1.0462281736783865e-05,
2629
+ "loss": 1.018,
2630
+ "step": 4370
2631
+ },
2632
+ {
2633
+ "epoch": 0.64,
2634
+ "learning_rate": 1.0389472901098789e-05,
2635
+ "loss": 1.0326,
2636
+ "step": 4380
2637
+ },
2638
+ {
2639
+ "epoch": 0.64,
2640
+ "learning_rate": 1.031678382881573e-05,
2641
+ "loss": 1.0061,
2642
+ "step": 4390
2643
+ },
2644
+ {
2645
+ "epoch": 0.64,
2646
+ "learning_rate": 1.0244216408111565e-05,
2647
+ "loss": 1.027,
2648
+ "step": 4400
2649
+ },
2650
+ {
2651
+ "epoch": 0.64,
2652
+ "learning_rate": 1.0171772524003114e-05,
2653
+ "loss": 1.0247,
2654
+ "step": 4410
2655
+ },
2656
+ {
2657
+ "epoch": 0.65,
2658
+ "learning_rate": 1.0099454058298211e-05,
2659
+ "loss": 1.0366,
2660
+ "step": 4420
2661
+ },
2662
+ {
2663
+ "epoch": 0.65,
2664
+ "learning_rate": 1.0027262889546819e-05,
2665
+ "loss": 0.996,
2666
+ "step": 4430
2667
+ },
2668
+ {
2669
+ "epoch": 0.65,
2670
+ "learning_rate": 9.955200892992209e-06,
2671
+ "loss": 0.9731,
2672
+ "step": 4440
2673
+ },
2674
+ {
2675
+ "epoch": 0.65,
2676
+ "learning_rate": 9.883269940522289e-06,
2677
+ "loss": 1.0416,
2678
+ "step": 4450
2679
+ },
2680
+ {
2681
+ "epoch": 0.65,
2682
+ "learning_rate": 9.811471900620935e-06,
2683
+ "loss": 1.0342,
2684
+ "step": 4460
2685
+ },
2686
+ {
2687
+ "epoch": 0.65,
2688
+ "learning_rate": 9.739808638319485e-06,
2689
+ "loss": 1.0043,
2690
+ "step": 4470
2691
+ },
2692
+ {
2693
+ "epoch": 0.65,
2694
+ "learning_rate": 9.668282015148282e-06,
2695
+ "loss": 0.9578,
2696
+ "step": 4480
2697
+ },
2698
+ {
2699
+ "epoch": 0.66,
2700
+ "learning_rate": 9.596893889088311e-06,
2701
+ "loss": 0.9867,
2702
+ "step": 4490
2703
+ },
2704
+ {
2705
+ "epoch": 0.66,
2706
+ "learning_rate": 9.525646114522974e-06,
2707
+ "loss": 0.9794,
2708
+ "step": 4500
2709
+ },
2710
+ {
2711
+ "epoch": 0.66,
2712
+ "learning_rate": 9.454540542189862e-06,
2713
+ "loss": 1.0271,
2714
+ "step": 4510
2715
+ },
2716
+ {
2717
+ "epoch": 0.66,
2718
+ "learning_rate": 9.383579019132725e-06,
2719
+ "loss": 1.0165,
2720
+ "step": 4520
2721
+ },
2722
+ {
2723
+ "epoch": 0.66,
2724
+ "learning_rate": 9.31276338865347e-06,
2725
+ "loss": 0.9486,
2726
+ "step": 4530
2727
+ },
2728
+ {
2729
+ "epoch": 0.66,
2730
+ "learning_rate": 9.242095490264293e-06,
2731
+ "loss": 0.9823,
2732
+ "step": 4540
2733
+ },
2734
+ {
2735
+ "epoch": 0.66,
2736
+ "learning_rate": 9.171577159639897e-06,
2737
+ "loss": 1.0255,
2738
+ "step": 4550
2739
+ },
2740
+ {
2741
+ "epoch": 0.67,
2742
+ "learning_rate": 9.101210228569787e-06,
2743
+ "loss": 0.9583,
2744
+ "step": 4560
2745
+ },
2746
+ {
2747
+ "epoch": 0.67,
2748
+ "learning_rate": 9.030996524910705e-06,
2749
+ "loss": 1.0067,
2750
+ "step": 4570
2751
+ },
2752
+ {
2753
+ "epoch": 0.67,
2754
+ "learning_rate": 8.960937872539146e-06,
2755
+ "loss": 1.0157,
2756
+ "step": 4580
2757
+ },
2758
+ {
2759
+ "epoch": 0.67,
2760
+ "learning_rate": 8.891036091303997e-06,
2761
+ "loss": 1.0111,
2762
+ "step": 4590
2763
+ },
2764
+ {
2765
+ "epoch": 0.67,
2766
+ "learning_rate": 8.821292996979227e-06,
2767
+ "loss": 0.985,
2768
+ "step": 4600
2769
+ },
2770
+ {
2771
+ "epoch": 0.67,
2772
+ "learning_rate": 8.751710401216751e-06,
2773
+ "loss": 0.9972,
2774
+ "step": 4610
2775
+ },
2776
+ {
2777
+ "epoch": 0.67,
2778
+ "learning_rate": 8.68229011149935e-06,
2779
+ "loss": 1.0101,
2780
+ "step": 4620
2781
+ },
2782
+ {
2783
+ "epoch": 0.68,
2784
+ "learning_rate": 8.613033931093765e-06,
2785
+ "loss": 1.0335,
2786
+ "step": 4630
2787
+ },
2788
+ {
2789
+ "epoch": 0.68,
2790
+ "learning_rate": 8.543943659003778e-06,
2791
+ "loss": 0.9619,
2792
+ "step": 4640
2793
+ },
2794
+ {
2795
+ "epoch": 0.68,
2796
+ "learning_rate": 8.475021089923549e-06,
2797
+ "loss": 0.9699,
2798
+ "step": 4650
2799
+ },
2800
+ {
2801
+ "epoch": 0.68,
2802
+ "learning_rate": 8.406268014190972e-06,
2803
+ "loss": 1.0082,
2804
+ "step": 4660
2805
+ },
2806
+ {
2807
+ "epoch": 0.68,
2808
+ "learning_rate": 8.337686217741144e-06,
2809
+ "loss": 1.0069,
2810
+ "step": 4670
2811
+ },
2812
+ {
2813
+ "epoch": 0.68,
2814
+ "learning_rate": 8.269277482060041e-06,
2815
+ "loss": 0.9989,
2816
+ "step": 4680
2817
+ },
2818
+ {
2819
+ "epoch": 0.68,
2820
+ "learning_rate": 8.201043584138162e-06,
2821
+ "loss": 1.024,
2822
+ "step": 4690
2823
+ },
2824
+ {
2825
+ "epoch": 0.69,
2826
+ "learning_rate": 8.13298629642443e-06,
2827
+ "loss": 0.9841,
2828
+ "step": 4700
2829
+ },
2830
+ {
2831
+ "epoch": 0.69,
2832
+ "learning_rate": 8.0651073867801e-06,
2833
+ "loss": 1.0588,
2834
+ "step": 4710
2835
+ },
2836
+ {
2837
+ "epoch": 0.69,
2838
+ "learning_rate": 7.997408618432906e-06,
2839
+ "loss": 1.0106,
2840
+ "step": 4720
2841
+ },
2842
+ {
2843
+ "epoch": 0.69,
2844
+ "learning_rate": 7.929891749931181e-06,
2845
+ "loss": 1.0286,
2846
+ "step": 4730
2847
+ },
2848
+ {
2849
+ "epoch": 0.69,
2850
+ "learning_rate": 7.862558535098233e-06,
2851
+ "loss": 0.9935,
2852
+ "step": 4740
2853
+ },
2854
+ {
2855
+ "epoch": 0.69,
2856
+ "learning_rate": 7.795410722986768e-06,
2857
+ "loss": 1.0016,
2858
+ "step": 4750
2859
+ },
2860
+ {
2861
+ "epoch": 0.69,
2862
+ "learning_rate": 7.728450057833437e-06,
2863
+ "loss": 1.0029,
2864
+ "step": 4760
2865
+ },
2866
+ {
2867
+ "epoch": 0.7,
2868
+ "learning_rate": 7.66167827901358e-06,
2869
+ "loss": 0.963,
2870
+ "step": 4770
2871
+ },
2872
+ {
2873
+ "epoch": 0.7,
2874
+ "learning_rate": 7.5950971209959854e-06,
2875
+ "loss": 0.9964,
2876
+ "step": 4780
2877
+ },
2878
+ {
2879
+ "epoch": 0.7,
2880
+ "learning_rate": 7.528708313297875e-06,
2881
+ "loss": 0.9823,
2882
+ "step": 4790
2883
+ },
2884
+ {
2885
+ "epoch": 0.7,
2886
+ "learning_rate": 7.4625135804399504e-06,
2887
+ "loss": 1.003,
2888
+ "step": 4800
2889
+ },
2890
+ {
2891
+ "epoch": 0.7,
2892
+ "learning_rate": 7.3965146419016245e-06,
2893
+ "loss": 1.0053,
2894
+ "step": 4810
2895
+ },
2896
+ {
2897
+ "epoch": 0.7,
2898
+ "learning_rate": 7.330713212076333e-06,
2899
+ "loss": 1.0341,
2900
+ "step": 4820
2901
+ },
2902
+ {
2903
+ "epoch": 0.71,
2904
+ "learning_rate": 7.265111000227018e-06,
2905
+ "loss": 0.9492,
2906
+ "step": 4830
2907
+ },
2908
+ {
2909
+ "epoch": 0.71,
2910
+ "learning_rate": 7.199709710441719e-06,
2911
+ "loss": 1.0086,
2912
+ "step": 4840
2913
+ },
2914
+ {
2915
+ "epoch": 0.71,
2916
+ "learning_rate": 7.134511041589295e-06,
2917
+ "loss": 1.0015,
2918
+ "step": 4850
2919
+ },
2920
+ {
2921
+ "epoch": 0.71,
2922
+ "learning_rate": 7.069516687275326e-06,
2923
+ "loss": 0.9316,
2924
+ "step": 4860
2925
+ },
2926
+ {
2927
+ "epoch": 0.71,
2928
+ "learning_rate": 7.004728335798092e-06,
2929
+ "loss": 1.0404,
2930
+ "step": 4870
2931
+ },
2932
+ {
2933
+ "epoch": 0.71,
2934
+ "learning_rate": 6.940147670104741e-06,
2935
+ "loss": 1.032,
2936
+ "step": 4880
2937
+ },
2938
+ {
2939
+ "epoch": 0.71,
2940
+ "learning_rate": 6.875776367747534e-06,
2941
+ "loss": 1.0138,
2942
+ "step": 4890
2943
+ },
2944
+ {
2945
+ "epoch": 0.72,
2946
+ "learning_rate": 6.811616100840315e-06,
2947
+ "loss": 1.0512,
2948
+ "step": 4900
2949
+ },
2950
+ {
2951
+ "epoch": 0.72,
2952
+ "learning_rate": 6.74766853601505e-06,
2953
+ "loss": 0.9984,
2954
+ "step": 4910
2955
+ },
2956
+ {
2957
+ "epoch": 0.72,
2958
+ "learning_rate": 6.683935334378546e-06,
2959
+ "loss": 1.0459,
2960
+ "step": 4920
2961
+ },
2962
+ {
2963
+ "epoch": 0.72,
2964
+ "learning_rate": 6.620418151469274e-06,
2965
+ "loss": 0.9889,
2966
+ "step": 4930
2967
+ },
2968
+ {
2969
+ "epoch": 0.72,
2970
+ "learning_rate": 6.5571186372144095e-06,
2971
+ "loss": 1.0057,
2972
+ "step": 4940
2973
+ },
2974
+ {
2975
+ "epoch": 0.72,
2976
+ "learning_rate": 6.494038435886937e-06,
2977
+ "loss": 1.0167,
2978
+ "step": 4950
2979
+ },
2980
+ {
2981
+ "epoch": 0.72,
2982
+ "learning_rate": 6.431179186062962e-06,
2983
+ "loss": 0.9865,
2984
+ "step": 4960
2985
+ },
2986
+ {
2987
+ "epoch": 0.73,
2988
+ "learning_rate": 6.368542520579131e-06,
2989
+ "loss": 0.9901,
2990
+ "step": 4970
2991
+ },
2992
+ {
2993
+ "epoch": 0.73,
2994
+ "learning_rate": 6.306130066490215e-06,
2995
+ "loss": 1.0113,
2996
+ "step": 4980
2997
+ },
2998
+ {
2999
+ "epoch": 0.73,
3000
+ "learning_rate": 6.2439434450268675e-06,
3001
+ "loss": 0.9662,
3002
+ "step": 4990
3003
+ },
3004
+ {
3005
+ "epoch": 0.73,
3006
+ "learning_rate": 6.181984271553485e-06,
3007
+ "loss": 0.9991,
3008
+ "step": 5000
3009
+ },
3010
+ {
3011
+ "epoch": 0.73,
3012
+ "learning_rate": 6.120254155526274e-06,
3013
+ "loss": 1.0123,
3014
+ "step": 5010
3015
+ },
3016
+ {
3017
+ "epoch": 0.73,
3018
+ "learning_rate": 6.058754700451399e-06,
3019
+ "loss": 1.0799,
3020
+ "step": 5020
3021
+ },
3022
+ {
3023
+ "epoch": 0.73,
3024
+ "learning_rate": 5.997487503843386e-06,
3025
+ "loss": 1.0039,
3026
+ "step": 5030
3027
+ },
3028
+ {
3029
+ "epoch": 0.74,
3030
+ "learning_rate": 5.936454157183583e-06,
3031
+ "loss": 1.0017,
3032
+ "step": 5040
3033
+ },
3034
+ {
3035
+ "epoch": 0.74,
3036
+ "learning_rate": 5.875656245878837e-06,
3037
+ "loss": 0.9983,
3038
+ "step": 5050
3039
+ },
3040
+ {
3041
+ "epoch": 0.74,
3042
+ "learning_rate": 5.815095349220319e-06,
3043
+ "loss": 1.0257,
3044
+ "step": 5060
3045
+ },
3046
+ {
3047
+ "epoch": 0.74,
3048
+ "learning_rate": 5.754773040342461e-06,
3049
+ "loss": 1.0044,
3050
+ "step": 5070
3051
+ },
3052
+ {
3053
+ "epoch": 0.74,
3054
+ "learning_rate": 5.694690886182152e-06,
3055
+ "loss": 1.0179,
3056
+ "step": 5080
3057
+ },
3058
+ {
3059
+ "epoch": 0.74,
3060
+ "learning_rate": 5.634850447437987e-06,
3061
+ "loss": 0.9734,
3062
+ "step": 5090
3063
+ },
3064
+ {
3065
+ "epoch": 0.74,
3066
+ "learning_rate": 5.57525327852975e-06,
3067
+ "loss": 1.0234,
3068
+ "step": 5100
3069
+ },
3070
+ {
3071
+ "epoch": 0.75,
3072
+ "learning_rate": 5.51590092755802e-06,
3073
+ "loss": 0.9776,
3074
+ "step": 5110
3075
+ },
3076
+ {
3077
+ "epoch": 0.75,
3078
+ "learning_rate": 5.4567949362639755e-06,
3079
+ "loss": 1.024,
3080
+ "step": 5120
3081
+ },
3082
+ {
3083
+ "epoch": 0.75,
3084
+ "learning_rate": 5.397936839989335e-06,
3085
+ "loss": 0.9682,
3086
+ "step": 5130
3087
+ },
3088
+ {
3089
+ "epoch": 0.75,
3090
+ "learning_rate": 5.339328167636477e-06,
3091
+ "loss": 0.9607,
3092
+ "step": 5140
3093
+ },
3094
+ {
3095
+ "epoch": 0.75,
3096
+ "learning_rate": 5.280970441628731e-06,
3097
+ "loss": 1.0421,
3098
+ "step": 5150
3099
+ },
3100
+ {
3101
+ "epoch": 0.75,
3102
+ "learning_rate": 5.222865177870807e-06,
3103
+ "loss": 0.9709,
3104
+ "step": 5160
3105
+ },
3106
+ {
3107
+ "epoch": 0.75,
3108
+ "learning_rate": 5.1650138857094544e-06,
3109
+ "loss": 1.0699,
3110
+ "step": 5170
3111
+ },
3112
+ {
3113
+ "epoch": 0.76,
3114
+ "learning_rate": 5.107418067894232e-06,
3115
+ "loss": 1.012,
3116
+ "step": 5180
3117
+ },
3118
+ {
3119
+ "epoch": 0.76,
3120
+ "learning_rate": 5.050079220538482e-06,
3121
+ "loss": 1.027,
3122
+ "step": 5190
3123
+ },
3124
+ {
3125
+ "epoch": 0.76,
3126
+ "learning_rate": 4.9929988330804484e-06,
3127
+ "loss": 1.0105,
3128
+ "step": 5200
3129
+ },
3130
+ {
3131
+ "epoch": 0.76,
3132
+ "learning_rate": 4.936178388244612e-06,
3133
+ "loss": 1.0257,
3134
+ "step": 5210
3135
+ },
3136
+ {
3137
+ "epoch": 0.76,
3138
+ "learning_rate": 4.879619362003166e-06,
3139
+ "loss": 1.0078,
3140
+ "step": 5220
3141
+ },
3142
+ {
3143
+ "epoch": 0.76,
3144
+ "learning_rate": 4.823323223537669e-06,
3145
+ "loss": 0.9772,
3146
+ "step": 5230
3147
+ },
3148
+ {
3149
+ "epoch": 0.77,
3150
+ "learning_rate": 4.767291435200894e-06,
3151
+ "loss": 0.9974,
3152
+ "step": 5240
3153
+ },
3154
+ {
3155
+ "epoch": 0.77,
3156
+ "learning_rate": 4.711525452478817e-06,
3157
+ "loss": 0.9691,
3158
+ "step": 5250
3159
+ },
3160
+ {
3161
+ "epoch": 0.77,
3162
+ "learning_rate": 4.656026723952851e-06,
3163
+ "loss": 1.0158,
3164
+ "step": 5260
3165
+ },
3166
+ {
3167
+ "epoch": 0.77,
3168
+ "learning_rate": 4.600796691262176e-06,
3169
+ "loss": 1.0188,
3170
+ "step": 5270
3171
+ },
3172
+ {
3173
+ "epoch": 0.77,
3174
+ "learning_rate": 4.545836789066326e-06,
3175
+ "loss": 1.016,
3176
+ "step": 5280
3177
+ },
3178
+ {
3179
+ "epoch": 0.77,
3180
+ "learning_rate": 4.4911484450078795e-06,
3181
+ "loss": 0.9551,
3182
+ "step": 5290
3183
+ },
3184
+ {
3185
+ "epoch": 0.77,
3186
+ "learning_rate": 4.436733079675421e-06,
3187
+ "loss": 1.0459,
3188
+ "step": 5300
3189
+ },
3190
+ {
3191
+ "epoch": 0.78,
3192
+ "learning_rate": 4.382592106566613e-06,
3193
+ "loss": 0.9958,
3194
+ "step": 5310
3195
+ },
3196
+ {
3197
+ "epoch": 0.78,
3198
+ "learning_rate": 4.328726932051481e-06,
3199
+ "loss": 0.9862,
3200
+ "step": 5320
3201
+ },
3202
+ {
3203
+ "epoch": 0.78,
3204
+ "learning_rate": 4.2751389553358984e-06,
3205
+ "loss": 1.0188,
3206
+ "step": 5330
3207
+ },
3208
+ {
3209
+ "epoch": 0.78,
3210
+ "learning_rate": 4.221829568425205e-06,
3211
+ "loss": 1.014,
3212
+ "step": 5340
3213
+ },
3214
+ {
3215
+ "epoch": 0.78,
3216
+ "learning_rate": 4.168800156088088e-06,
3217
+ "loss": 1.0026,
3218
+ "step": 5350
3219
+ },
3220
+ {
3221
+ "epoch": 0.78,
3222
+ "learning_rate": 4.11605209582059e-06,
3223
+ "loss": 0.9917,
3224
+ "step": 5360
3225
+ },
3226
+ {
3227
+ "epoch": 0.78,
3228
+ "learning_rate": 4.063586757810337e-06,
3229
+ "loss": 1.0121,
3230
+ "step": 5370
3231
+ },
3232
+ {
3233
+ "epoch": 0.79,
3234
+ "learning_rate": 4.011405504900918e-06,
3235
+ "loss": 1.0027,
3236
+ "step": 5380
3237
+ },
3238
+ {
3239
+ "epoch": 0.79,
3240
+ "learning_rate": 3.959509692556528e-06,
3241
+ "loss": 0.9739,
3242
+ "step": 5390
3243
+ },
3244
+ {
3245
+ "epoch": 0.79,
3246
+ "learning_rate": 3.907900668826731e-06,
3247
+ "loss": 0.9491,
3248
+ "step": 5400
3249
+ },
3250
+ {
3251
+ "epoch": 0.79,
3252
+ "learning_rate": 3.8565797743114466e-06,
3253
+ "loss": 1.0175,
3254
+ "step": 5410
3255
+ },
3256
+ {
3257
+ "epoch": 0.79,
3258
+ "learning_rate": 3.8055483421261296e-06,
3259
+ "loss": 1.0162,
3260
+ "step": 5420
3261
+ },
3262
+ {
3263
+ "epoch": 0.79,
3264
+ "learning_rate": 3.7548076978671304e-06,
3265
+ "loss": 0.9831,
3266
+ "step": 5430
3267
+ },
3268
+ {
3269
+ "epoch": 0.79,
3270
+ "learning_rate": 3.704359159577286e-06,
3271
+ "loss": 1.0251,
3272
+ "step": 5440
3273
+ },
3274
+ {
3275
+ "epoch": 0.8,
3276
+ "learning_rate": 3.654204037711644e-06,
3277
+ "loss": 1.0904,
3278
+ "step": 5450
3279
+ },
3280
+ {
3281
+ "epoch": 0.8,
3282
+ "learning_rate": 3.604343635103475e-06,
3283
+ "loss": 1.017,
3284
+ "step": 5460
3285
+ },
3286
+ {
3287
+ "epoch": 0.8,
3288
+ "learning_rate": 3.5547792469303744e-06,
3289
+ "loss": 0.9493,
3290
+ "step": 5470
3291
+ },
3292
+ {
3293
+ "epoch": 0.8,
3294
+ "learning_rate": 3.505512160680653e-06,
3295
+ "loss": 0.9584,
3296
+ "step": 5480
3297
+ },
3298
+ {
3299
+ "epoch": 0.8,
3300
+ "learning_rate": 3.4565436561198867e-06,
3301
+ "loss": 0.9801,
3302
+ "step": 5490
3303
+ },
3304
+ {
3305
+ "epoch": 0.8,
3306
+ "learning_rate": 3.4078750052576677e-06,
3307
+ "loss": 1.0025,
3308
+ "step": 5500
3309
+ },
3310
+ {
3311
+ "epoch": 0.8,
3312
+ "learning_rate": 3.35950747231457e-06,
3313
+ "loss": 1.0691,
3314
+ "step": 5510
3315
+ },
3316
+ {
3317
+ "epoch": 0.81,
3318
+ "learning_rate": 3.311442313689297e-06,
3319
+ "loss": 0.9631,
3320
+ "step": 5520
3321
+ },
3322
+ {
3323
+ "epoch": 0.81,
3324
+ "learning_rate": 3.2636807779260626e-06,
3325
+ "loss": 0.9863,
3326
+ "step": 5530
3327
+ },
3328
+ {
3329
+ "epoch": 0.81,
3330
+ "learning_rate": 3.216224105682137e-06,
3331
+ "loss": 0.9778,
3332
+ "step": 5540
3333
+ },
3334
+ {
3335
+ "epoch": 0.81,
3336
+ "learning_rate": 3.169073529695658e-06,
3337
+ "loss": 1.0446,
3338
+ "step": 5550
3339
+ },
3340
+ {
3341
+ "epoch": 0.81,
3342
+ "learning_rate": 3.1222302747535564e-06,
3343
+ "loss": 1.0131,
3344
+ "step": 5560
3345
+ },
3346
+ {
3347
+ "epoch": 0.81,
3348
+ "learning_rate": 3.075695557659782e-06,
3349
+ "loss": 0.9859,
3350
+ "step": 5570
3351
+ },
3352
+ {
3353
+ "epoch": 0.81,
3354
+ "learning_rate": 3.0294705872036793e-06,
3355
+ "loss": 1.0158,
3356
+ "step": 5580
3357
+ },
3358
+ {
3359
+ "epoch": 0.82,
3360
+ "learning_rate": 2.983556564128594e-06,
3361
+ "loss": 0.9944,
3362
+ "step": 5590
3363
+ },
3364
+ {
3365
+ "epoch": 0.82,
3366
+ "learning_rate": 2.937954681100678e-06,
3367
+ "loss": 0.9501,
3368
+ "step": 5600
3369
+ },
3370
+ {
3371
+ "epoch": 0.82,
3372
+ "learning_rate": 2.8926661226779e-06,
3373
+ "loss": 1.0188,
3374
+ "step": 5610
3375
+ },
3376
+ {
3377
+ "epoch": 0.82,
3378
+ "learning_rate": 2.8476920652792987e-06,
3379
+ "loss": 0.9919,
3380
+ "step": 5620
3381
+ },
3382
+ {
3383
+ "epoch": 0.82,
3384
+ "learning_rate": 2.8030336771543965e-06,
3385
+ "loss": 0.9613,
3386
+ "step": 5630
3387
+ },
3388
+ {
3389
+ "epoch": 0.82,
3390
+ "learning_rate": 2.7586921183528857e-06,
3391
+ "loss": 1.0496,
3392
+ "step": 5640
3393
+ },
3394
+ {
3395
+ "epoch": 0.82,
3396
+ "learning_rate": 2.7146685406944567e-06,
3397
+ "loss": 0.9877,
3398
+ "step": 5650
3399
+ },
3400
+ {
3401
+ "epoch": 0.83,
3402
+ "learning_rate": 2.6709640877389073e-06,
3403
+ "loss": 1.0,
3404
+ "step": 5660
3405
+ },
3406
+ {
3407
+ "epoch": 0.83,
3408
+ "learning_rate": 2.6275798947564277e-06,
3409
+ "loss": 0.9826,
3410
+ "step": 5670
3411
+ },
3412
+ {
3413
+ "epoch": 0.83,
3414
+ "learning_rate": 2.584517088698107e-06,
3415
+ "loss": 1.0358,
3416
+ "step": 5680
3417
+ },
3418
+ {
3419
+ "epoch": 0.83,
3420
+ "learning_rate": 2.5417767881666716e-06,
3421
+ "loss": 1.0195,
3422
+ "step": 5690
3423
+ },
3424
+ {
3425
+ "epoch": 0.83,
3426
+ "learning_rate": 2.4993601033874053e-06,
3427
+ "loss": 0.9495,
3428
+ "step": 5700
3429
+ },
3430
+ {
3431
+ "epoch": 0.83,
3432
+ "learning_rate": 2.457268136179341e-06,
3433
+ "loss": 1.0138,
3434
+ "step": 5710
3435
+ },
3436
+ {
3437
+ "epoch": 0.84,
3438
+ "learning_rate": 2.4155019799266016e-06,
3439
+ "loss": 0.9828,
3440
+ "step": 5720
3441
+ },
3442
+ {
3443
+ "epoch": 0.84,
3444
+ "learning_rate": 2.3740627195500503e-06,
3445
+ "loss": 0.981,
3446
+ "step": 5730
3447
+ },
3448
+ {
3449
+ "epoch": 0.84,
3450
+ "learning_rate": 2.3329514314790514e-06,
3451
+ "loss": 0.9832,
3452
+ "step": 5740
3453
+ },
3454
+ {
3455
+ "epoch": 0.84,
3456
+ "learning_rate": 2.292169183623545e-06,
3457
+ "loss": 0.9748,
3458
+ "step": 5750
3459
+ },
3460
+ {
3461
+ "epoch": 0.84,
3462
+ "learning_rate": 2.251717035346307e-06,
3463
+ "loss": 0.9994,
3464
+ "step": 5760
3465
+ },
3466
+ {
3467
+ "epoch": 0.84,
3468
+ "learning_rate": 2.2115960374353987e-06,
3469
+ "loss": 1.0136,
3470
+ "step": 5770
3471
+ },
3472
+ {
3473
+ "epoch": 0.84,
3474
+ "learning_rate": 2.1718072320769234e-06,
3475
+ "loss": 0.9981,
3476
+ "step": 5780
3477
+ },
3478
+ {
3479
+ "epoch": 0.85,
3480
+ "learning_rate": 2.1323516528279002e-06,
3481
+ "loss": 0.99,
3482
+ "step": 5790
3483
+ },
3484
+ {
3485
+ "epoch": 0.85,
3486
+ "learning_rate": 2.0932303245894587e-06,
3487
+ "loss": 1.0173,
3488
+ "step": 5800
3489
+ },
3490
+ {
3491
+ "epoch": 0.85,
3492
+ "learning_rate": 2.0544442635801823e-06,
3493
+ "loss": 1.0113,
3494
+ "step": 5810
3495
+ },
3496
+ {
3497
+ "epoch": 0.85,
3498
+ "learning_rate": 2.0159944773097494e-06,
3499
+ "loss": 0.9995,
3500
+ "step": 5820
3501
+ },
3502
+ {
3503
+ "epoch": 0.85,
3504
+ "learning_rate": 1.9778819645527173e-06,
3505
+ "loss": 1.0165,
3506
+ "step": 5830
3507
+ },
3508
+ {
3509
+ "epoch": 0.85,
3510
+ "learning_rate": 1.9401077153226158e-06,
3511
+ "loss": 1.0047,
3512
+ "step": 5840
3513
+ },
3514
+ {
3515
+ "epoch": 0.85,
3516
+ "learning_rate": 1.9026727108462167e-06,
3517
+ "loss": 0.9809,
3518
+ "step": 5850
3519
+ },
3520
+ {
3521
+ "epoch": 0.86,
3522
+ "learning_rate": 1.865577923538025e-06,
3523
+ "loss": 0.9962,
3524
+ "step": 5860
3525
+ },
3526
+ {
3527
+ "epoch": 0.86,
3528
+ "learning_rate": 1.8288243169750674e-06,
3529
+ "loss": 0.9608,
3530
+ "step": 5870
3531
+ },
3532
+ {
3533
+ "epoch": 0.86,
3534
+ "learning_rate": 1.7924128458718115e-06,
3535
+ "loss": 1.0048,
3536
+ "step": 5880
3537
+ },
3538
+ {
3539
+ "epoch": 0.86,
3540
+ "learning_rate": 1.7563444560553972e-06,
3541
+ "loss": 0.9685,
3542
+ "step": 5890
3543
+ },
3544
+ {
3545
+ "epoch": 0.86,
3546
+ "learning_rate": 1.720620084441053e-06,
3547
+ "loss": 1.0268,
3548
+ "step": 5900
3549
+ },
3550
+ {
3551
+ "epoch": 0.86,
3552
+ "learning_rate": 1.6852406590077762e-06,
3553
+ "loss": 0.9855,
3554
+ "step": 5910
3555
+ },
3556
+ {
3557
+ "epoch": 0.86,
3558
+ "learning_rate": 1.650207098774203e-06,
3559
+ "loss": 0.9621,
3560
+ "step": 5920
3561
+ },
3562
+ {
3563
+ "epoch": 0.87,
3564
+ "learning_rate": 1.615520313774756e-06,
3565
+ "loss": 1.03,
3566
+ "step": 5930
3567
+ },
3568
+ {
3569
+ "epoch": 0.87,
3570
+ "learning_rate": 1.5811812050359986e-06,
3571
+ "loss": 0.9722,
3572
+ "step": 5940
3573
+ },
3574
+ {
3575
+ "epoch": 0.87,
3576
+ "learning_rate": 1.5471906645532203e-06,
3577
+ "loss": 1.03,
3578
+ "step": 5950
3579
+ },
3580
+ {
3581
+ "epoch": 0.87,
3582
+ "learning_rate": 1.5135495752672911e-06,
3583
+ "loss": 1.0527,
3584
+ "step": 5960
3585
+ },
3586
+ {
3587
+ "epoch": 0.87,
3588
+ "learning_rate": 1.4802588110416914e-06,
3589
+ "loss": 1.0058,
3590
+ "step": 5970
3591
+ },
3592
+ {
3593
+ "epoch": 0.87,
3594
+ "learning_rate": 1.4473192366398502e-06,
3595
+ "loss": 1.0255,
3596
+ "step": 5980
3597
+ },
3598
+ {
3599
+ "epoch": 0.87,
3600
+ "learning_rate": 1.4147317077026423e-06,
3601
+ "loss": 1.012,
3602
+ "step": 5990
3603
+ },
3604
+ {
3605
+ "epoch": 0.88,
3606
+ "learning_rate": 1.3824970707262042e-06,
3607
+ "loss": 0.9735,
3608
+ "step": 6000
3609
+ }
3610
+ ],
3611
+ "max_steps": 6849,
3612
+ "num_train_epochs": 1,
3613
+ "total_flos": 0.0,
3614
+ "trial_name": null,
3615
+ "trial_params": null
3616
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_2x/training_complete.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_path": "/root/pixel_cloud/contextual_100_work/data/contextual/downstream/MSA_Dialect_Combined_2x_max64",
3
+ "dataset_rows": 876609,
4
+ "epochs": 1,
5
+ "global_step": 6849,
6
+ "input_checkpoint": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct100/pretraining/pixel/full_epoch2/final/pytorch_model.bin",
7
+ "learning_rate": 3e-05,
8
+ "metrics": {
9
+ "epoch": 1.0,
10
+ "train_loss": 1.036986277353713,
11
+ "train_runtime": 1855.6998,
12
+ "train_samples_per_second": 472.387,
13
+ "train_steps_per_second": 3.691
14
+ },
15
+ "model_type": "pixel",
16
+ "scheduler": "cosine",
17
+ "seed": 42,
18
+ "warmup_ratio": 0.1
19
+ }
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_4x/checkpoint-10000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_4x/checkpoint-11000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/finetuning/pixel/MSA_Dialect_Combined_4x/training_complete.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_path": "/root/pixel_cloud/contextual_100_work/data/contextual/downstream/MSA_Dialect_Combined_4x_max64",
3
+ "dataset_rows": 1461015,
4
+ "epochs": 1,
5
+ "global_step": 11414,
6
+ "input_checkpoint": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct100/pretraining/pixel/full_epoch2/final/pytorch_model.bin",
7
+ "learning_rate": 3e-05,
8
+ "metrics": {
9
+ "epoch": 1.0,
10
+ "train_loss": 0.9203722805740413,
11
+ "train_runtime": 3126.7364,
12
+ "train_samples_per_second": 467.265,
13
+ "train_steps_per_second": 3.65
14
+ },
15
+ "model_type": "pixel",
16
+ "scheduler": "cosine",
17
+ "seed": 42,
18
+ "warmup_ratio": 0.1
19
+ }
contextual/pct100/manifest.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checked_train_indices": [
3
+ 0,
4
+ 1,
5
+ 3434577,
6
+ 12021020,
7
+ 24042039,
8
+ 24042040
9
+ ],
10
+ "inspected_word_images": 67,
11
+ "percentage": 100,
12
+ "raw_dataset": "/workspace/datasets/pixel_translation_MSA_filtered_32",
13
+ "raw_train_rows": 24042041,
14
+ "rendered_dataset": "/root/pixel_cloud/contextual_100_work/data/contextual/pct100/msa",
15
+ "rendered_train_rows": 24042041,
16
+ "renderer": {
17
+ "arabic_shaping": "Pango shapes the pair with join-causing boundary sentinels; every observed pair/join-state combination maps to one cached patch",
18
+ "bigram_overlap": 1,
19
+ "canonical_key": "ordered grapheme pair plus joins_previous and joins_following",
20
+ "direction_policy": "bigram patches follow physical canvas x-order; RTL tokens reverse logical bigram order while Pango controls order inside each patch",
21
+ "dpi": 120,
22
+ "font_file": "/root/pixel_cloud/code/pixel_translation/PixelSmolLM/rendering/fonts/GoNotoCurrent.ttf",
23
+ "font_size": 10,
24
+ "image_height": 24,
25
+ "image_width": 384,
26
+ "patch_extraction": "render each bigram in a deterministic synthetic Arabic joining context selected by its two external connection states",
27
+ "patch_size": 24,
28
+ "patches_per_word": 16,
29
+ "shaping_mode": "canonical",
30
+ "single_grapheme_policy": "one unigram patch",
31
+ "strategy": "overlapping_grapheme_bigram_patches",
32
+ "unicode_unit": "extended grapheme cluster (regex \\X)",
33
+ "version": 1,
34
+ "word_segmentation": "same punctuation-aware segmentation as PixelSmolLM"
35
+ },
36
+ "schema_version": 1,
37
+ "valid": true
38
+ }
contextual/pct100/msa_training.complete ADDED
File without changes
contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e9aa441491b9ca89e796944520fa1db332a67c0a1a920be83edd2d96d741716d
3
+ size 15024
contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1ee3434533b24fb771504fa8cceb5c2ea25fe0de1641128feaceccc65afe6ed
3
+ size 15024
contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93b4a44be1335173d2e3120bd0d1e6346f3e832d8935752c70ce1e98f017fa87
3
+ size 15024
contextual/pct100/pretraining/pixel/alignment/checkpoint-44000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/pretraining/pixel/alignment/checkpoint-46000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/pretraining/pixel/alignment/final/training_complete.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epochs": 1,
3
+ "global_step": 46957,
4
+ "initial_weights": null,
5
+ "learning_rate": 0.0003,
6
+ "metrics": {
7
+ "epoch": 1.0,
8
+ "train_loss": 2.0845835111762523,
9
+ "train_runtime": 25715.3295,
10
+ "train_samples_per_second": 934.93,
11
+ "train_steps_per_second": 1.826
12
+ },
13
+ "model_type": "pixel",
14
+ "percentage": 100,
15
+ "scheduler": "cosine",
16
+ "seed": 42,
17
+ "stage": "alignment",
18
+ "subset_rows": 24042041,
19
+ "total_rows": 24042041,
20
+ "warmup_ratio": 0.1
21
+ }
contextual/pct100/pretraining/pixel/full_epoch1/checkpoint-44000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/pretraining/pixel/full_epoch1/checkpoint-46000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/pretraining/pixel/full_epoch1/final/training_complete.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epochs": 1,
3
+ "global_step": 46957,
4
+ "initial_weights": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct100/pretraining/pixel/alignment/final/pytorch_model.bin",
5
+ "learning_rate": 0.0003,
6
+ "metrics": {
7
+ "epoch": 1.0,
8
+ "train_loss": 0.9602556531960436,
9
+ "train_runtime": 30723.4856,
10
+ "train_samples_per_second": 782.53,
11
+ "train_steps_per_second": 1.528
12
+ },
13
+ "model_type": "pixel",
14
+ "percentage": 100,
15
+ "scheduler": "cosine",
16
+ "seed": 42,
17
+ "stage": "full",
18
+ "subset_rows": 24042041,
19
+ "total_rows": 24042041,
20
+ "warmup_ratio": 0.1
21
+ }
contextual/pct100/pretraining/pixel/full_epoch2/checkpoint-44000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/pretraining/pixel/full_epoch2/checkpoint-46000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct100/pretraining/pixel/full_epoch2/final/training_complete.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epochs": 1,
3
+ "global_step": 46957,
4
+ "initial_weights": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct100/pretraining/pixel/full_epoch1/final/pytorch_model.bin",
5
+ "learning_rate": 0.0003,
6
+ "metrics": {
7
+ "epoch": 1.0,
8
+ "train_loss": 0.8417160780981193,
9
+ "train_runtime": 30673.4432,
10
+ "train_samples_per_second": 783.806,
11
+ "train_steps_per_second": 1.531
12
+ },
13
+ "model_type": "pixel",
14
+ "percentage": 100,
15
+ "scheduler": "cosine",
16
+ "seed": 42,
17
+ "stage": "full",
18
+ "subset_rows": 24042041,
19
+ "total_rows": 24042041,
20
+ "warmup_ratio": 0.1
21
+ }
contextual/pct100/renderer_audit.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "blank_patch_fraction": 0.0,
3
+ "blank_patches": 0,
4
+ "dataset": "/workspace/datasets/pixel_translation_MSA_filtered_32",
5
+ "ink_touches_horizontal_edge_fraction": 0.04563407977903112,
6
+ "ink_touches_horizontal_edge_patches": 186395,
7
+ "logical_extent_over_24px_fraction": 0.021954645265080154,
8
+ "logical_extent_over_24px_patches": 89675,
9
+ "rendered_patches": 4084557,
10
+ "renderer": {
11
+ "arabic_shaping": "Pango shapes the pair with join-causing boundary sentinels; every observed pair/join-state combination maps to one cached patch",
12
+ "bigram_overlap": 1,
13
+ "canonical_key": "ordered grapheme pair plus joins_previous and joins_following",
14
+ "direction_policy": "bigram patches follow physical canvas x-order; RTL tokens reverse logical bigram order while Pango controls order inside each patch",
15
+ "dpi": 120,
16
+ "font_file": "/root/pixel_cloud/code/pixel_translation/PixelSmolLM/rendering/fonts/GoNotoCurrent.ttf",
17
+ "font_size": 10,
18
+ "image_height": 24,
19
+ "image_width": 384,
20
+ "patch_extraction": "render each bigram in a deterministic synthetic Arabic joining context selected by its two external connection states",
21
+ "patch_size": 24,
22
+ "patches_per_word": 16,
23
+ "shaping_mode": "canonical",
24
+ "single_grapheme_policy": "one unigram patch",
25
+ "strategy": "overlapping_grapheme_bigram_patches",
26
+ "unicode_unit": "extended grapheme cluster (regex \\X)",
27
+ "version": 1,
28
+ "word_segmentation": "same punctuation-aware segmentation as PixelSmolLM"
29
+ },
30
+ "sample_rows": 100000,
31
+ "sample_selection": "train rows [0, 100000)",
32
+ "sample_words_after_32_word_cap": 1228012,
33
+ "unique_rendered_units": 4421,
34
+ "unknown_glyph_fraction": 0.0,
35
+ "unknown_glyph_patches": 0,
36
+ "words_over_16_bigrams": 244,
37
+ "words_over_16_bigrams_fraction": 0.00019869512675771897
38
+ }
contextual/pct50/experiment.complete ADDED
File without changes
contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/best_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4c69a628bccef84f6aac68737d01b5817c9030c24022604269989f72237bfb5
3
+ size 1942238590
contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-3000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbd2d4f6a13357559e462879e5fa09c07012449378296e74153191ac98b218f2
3
+ size 1942236130
contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,1816 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.657030223390276,
5
+ "global_step": 3000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 6.564551422319475e-07,
13
+ "loss": 1.6933,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 1.312910284463895e-06,
19
+ "loss": 1.6626,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.9693654266958425e-06,
25
+ "loss": 1.6637,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 2.62582056892779e-06,
31
+ "loss": 1.6255,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 3.2822757111597377e-06,
37
+ "loss": 1.6863,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 3.938730853391685e-06,
43
+ "loss": 1.7109,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.595185995623632e-06,
49
+ "loss": 1.6576,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 5.25164113785558e-06,
55
+ "loss": 1.6035,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 5.908096280087528e-06,
61
+ "loss": 1.576,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 6.564551422319475e-06,
67
+ "loss": 1.6065,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 7.221006564551422e-06,
73
+ "loss": 1.5913,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 7.87746170678337e-06,
79
+ "loss": 1.4796,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 8.533916849015316e-06,
85
+ "loss": 1.5987,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 9.190371991247264e-06,
91
+ "loss": 1.4971,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 9.846827133479212e-06,
97
+ "loss": 1.5054,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 1.050328227571116e-05,
103
+ "loss": 1.5211,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 1.1159737417943107e-05,
109
+ "loss": 1.5883,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 1.1816192560175055e-05,
115
+ "loss": 1.5497,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 1.2472647702407003e-05,
121
+ "loss": 1.5075,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.04,
126
+ "learning_rate": 1.312910284463895e-05,
127
+ "loss": 1.4538,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 1.3785557986870897e-05,
133
+ "loss": 1.4766,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 1.4442013129102845e-05,
139
+ "loss": 1.4861,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 1.509846827133479e-05,
145
+ "loss": 1.5375,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.05,
150
+ "learning_rate": 1.575492341356674e-05,
151
+ "loss": 1.4774,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.05,
156
+ "learning_rate": 1.6411378555798686e-05,
157
+ "loss": 1.4756,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 1.7067833698030632e-05,
163
+ "loss": 1.4494,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 1.7724288840262582e-05,
169
+ "loss": 1.4283,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 1.8380743982494528e-05,
175
+ "loss": 1.382,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.06,
180
+ "learning_rate": 1.9037199124726478e-05,
181
+ "loss": 1.4212,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 1.9693654266958424e-05,
187
+ "loss": 1.4249,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 2.0350109409190373e-05,
193
+ "loss": 1.4774,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 2.100656455142232e-05,
199
+ "loss": 1.4058,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.07,
204
+ "learning_rate": 2.1663019693654265e-05,
205
+ "loss": 1.4106,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.07,
210
+ "learning_rate": 2.2319474835886215e-05,
211
+ "loss": 1.4115,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 2.297592997811816e-05,
217
+ "loss": 1.4148,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 2.363238512035011e-05,
223
+ "loss": 1.4041,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.08,
228
+ "learning_rate": 2.4288840262582056e-05,
229
+ "loss": 1.354,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.08,
234
+ "learning_rate": 2.4945295404814006e-05,
235
+ "loss": 1.4267,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 2.5601750547045952e-05,
241
+ "loss": 1.4255,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 2.62582056892779e-05,
247
+ "loss": 1.3738,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.09,
252
+ "learning_rate": 2.6914660831509848e-05,
253
+ "loss": 1.4188,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.09,
258
+ "learning_rate": 2.7571115973741794e-05,
259
+ "loss": 1.3703,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.09,
264
+ "learning_rate": 2.8227571115973743e-05,
265
+ "loss": 1.3946,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 2.888402625820569e-05,
271
+ "loss": 1.3408,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 2.954048140043764e-05,
277
+ "loss": 1.3751,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.1,
282
+ "learning_rate": 2.9999960542365528e-05,
283
+ "loss": 1.3858,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.1,
288
+ "learning_rate": 2.9999259079083055e-05,
289
+ "loss": 1.3235,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 2.9997680826676758e-05,
295
+ "loss": 1.394,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 2.9995225877404156e-05,
301
+ "loss": 1.3088,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.11,
306
+ "learning_rate": 2.9991894374770534e-05,
307
+ "loss": 1.2879,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.11,
312
+ "learning_rate": 2.9987686513520527e-05,
313
+ "loss": 1.3771,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.11,
318
+ "learning_rate": 2.9982602539626753e-05,
319
+ "loss": 1.3274,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 2.997664275027542e-05,
325
+ "loss": 1.364,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.12,
330
+ "learning_rate": 2.996980749384897e-05,
331
+ "loss": 1.3429,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.12,
336
+ "learning_rate": 2.9962097169905706e-05,
337
+ "loss": 1.3442,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.12,
342
+ "learning_rate": 2.995351222915644e-05,
343
+ "loss": 1.3834,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.12,
348
+ "learning_rate": 2.9944053173438132e-05,
349
+ "loss": 1.3202,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 2.9933720555684573e-05,
355
+ "loss": 1.2823,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.13,
360
+ "learning_rate": 2.992251497989406e-05,
361
+ "loss": 1.3072,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.13,
366
+ "learning_rate": 2.9910437101094075e-05,
367
+ "loss": 1.3526,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.13,
372
+ "learning_rate": 2.9897487625303016e-05,
373
+ "loss": 1.2923,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 2.9883667309488905e-05,
379
+ "loss": 1.3429,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.14,
384
+ "learning_rate": 2.9868976961525146e-05,
385
+ "loss": 1.2892,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.14,
390
+ "learning_rate": 2.985341744014332e-05,
391
+ "loss": 1.2687,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.14,
396
+ "learning_rate": 2.983698965488296e-05,
397
+ "loss": 1.3028,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.14,
402
+ "learning_rate": 2.9819694566038392e-05,
403
+ "loss": 1.2969,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.15,
408
+ "learning_rate": 2.980153318460261e-05,
409
+ "loss": 1.2961,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.15,
414
+ "learning_rate": 2.9782506572208166e-05,
415
+ "loss": 1.3652,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.15,
420
+ "learning_rate": 2.976261584106512e-05,
421
+ "loss": 1.3636,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.15,
426
+ "learning_rate": 2.974186215389601e-05,
427
+ "loss": 1.3004,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 2.9720246723867926e-05,
433
+ "loss": 1.319,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.16,
438
+ "learning_rate": 2.9697770814521526e-05,
439
+ "loss": 1.3539,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.16,
444
+ "learning_rate": 2.9674435739697234e-05,
445
+ "loss": 1.3482,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.16,
450
+ "learning_rate": 2.9650242863458425e-05,
451
+ "loss": 1.2532,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.16,
456
+ "learning_rate": 2.9625193600011658e-05,
457
+ "loss": 1.3419,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.17,
462
+ "learning_rate": 2.959928941362405e-05,
463
+ "loss": 1.315,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.17,
468
+ "learning_rate": 2.9572531818537654e-05,
469
+ "loss": 1.3297,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.17,
474
+ "learning_rate": 2.9544922378880957e-05,
475
+ "loss": 1.2816,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.17,
480
+ "learning_rate": 2.951646270857743e-05,
481
+ "loss": 1.3217,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.18,
486
+ "learning_rate": 2.9487154471251217e-05,
487
+ "loss": 1.3086,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.18,
492
+ "learning_rate": 2.9456999380129854e-05,
493
+ "loss": 1.2999,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.18,
498
+ "learning_rate": 2.942599919794413e-05,
499
+ "loss": 1.2864,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.18,
504
+ "learning_rate": 2.939415573682507e-05,
505
+ "loss": 1.2831,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.18,
510
+ "learning_rate": 2.9361470858197978e-05,
511
+ "loss": 1.3005,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.19,
516
+ "learning_rate": 2.932794647267363e-05,
517
+ "loss": 1.2977,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.19,
522
+ "learning_rate": 2.929358453993659e-05,
523
+ "loss": 1.2909,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.19,
528
+ "learning_rate": 2.925838706863069e-05,
529
+ "loss": 1.3001,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.19,
534
+ "learning_rate": 2.9222356116241536e-05,
535
+ "loss": 1.2478,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.19,
540
+ "learning_rate": 2.9185493788976324e-05,
541
+ "loss": 1.2797,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.2,
546
+ "learning_rate": 2.914780224164066e-05,
547
+ "loss": 1.3308,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.2,
552
+ "learning_rate": 2.9109283677512624e-05,
553
+ "loss": 1.277,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.2,
558
+ "learning_rate": 2.9069940348213978e-05,
559
+ "loss": 1.2825,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.2,
564
+ "learning_rate": 2.9029774553578535e-05,
565
+ "loss": 1.2899,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.21,
570
+ "learning_rate": 2.8988788641517723e-05,
571
+ "loss": 1.2567,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.21,
576
+ "learning_rate": 2.8946985007883354e-05,
577
+ "loss": 1.2643,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.21,
582
+ "learning_rate": 2.890436609632755e-05,
583
+ "loss": 1.3001,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.21,
588
+ "learning_rate": 2.8860934398159905e-05,
589
+ "loss": 1.2995,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.21,
594
+ "learning_rate": 2.881669245220187e-05,
595
+ "loss": 1.2811,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.22,
600
+ "learning_rate": 2.8771642844638316e-05,
601
+ "loss": 1.3262,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.22,
606
+ "learning_rate": 2.8725788208866394e-05,
607
+ "loss": 1.2916,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.22,
612
+ "learning_rate": 2.867913122534155e-05,
613
+ "loss": 1.2962,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.22,
618
+ "learning_rate": 2.8631674621420895e-05,
619
+ "loss": 1.2852,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.23,
624
+ "learning_rate": 2.8583421171203728e-05,
625
+ "loss": 1.2704,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.23,
630
+ "learning_rate": 2.8534373695369393e-05,
631
+ "loss": 1.2264,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.23,
636
+ "learning_rate": 2.8484535061012404e-05,
637
+ "loss": 1.321,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.23,
642
+ "learning_rate": 2.8433908181474845e-05,
643
+ "loss": 1.2498,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.23,
648
+ "learning_rate": 2.8382496016176033e-05,
649
+ "loss": 1.2612,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.24,
654
+ "learning_rate": 2.8330301570439588e-05,
655
+ "loss": 1.2424,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.24,
660
+ "learning_rate": 2.8277327895317703e-05,
661
+ "loss": 1.2828,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.24,
666
+ "learning_rate": 2.8223578087412806e-05,
667
+ "loss": 1.2996,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.24,
672
+ "learning_rate": 2.816905528869657e-05,
673
+ "loss": 1.2903,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.25,
678
+ "learning_rate": 2.8113762686326204e-05,
679
+ "loss": 1.268,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.25,
684
+ "learning_rate": 2.80577035124582e-05,
685
+ "loss": 1.2798,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.25,
690
+ "learning_rate": 2.8000881044059343e-05,
691
+ "loss": 1.2449,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.25,
696
+ "learning_rate": 2.7943298602715196e-05,
697
+ "loss": 1.2877,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.25,
702
+ "learning_rate": 2.7884959554435894e-05,
703
+ "loss": 1.2238,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.26,
708
+ "learning_rate": 2.7825867309459432e-05,
709
+ "loss": 1.2947,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.26,
714
+ "learning_rate": 2.7766025322052274e-05,
715
+ "loss": 1.2752,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.26,
720
+ "learning_rate": 2.7705437090307463e-05,
721
+ "loss": 1.2855,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.26,
726
+ "learning_rate": 2.7644106155940102e-05,
727
+ "loss": 1.2525,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.27,
732
+ "learning_rate": 2.7582036104080366e-05,
733
+ "loss": 1.312,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.27,
738
+ "learning_rate": 2.7519230563063903e-05,
739
+ "loss": 1.2819,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.27,
744
+ "learning_rate": 2.7455693204219735e-05,
745
+ "loss": 1.2797,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.27,
750
+ "learning_rate": 2.7391427741655673e-05,
751
+ "loss": 1.3005,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.27,
756
+ "learning_rate": 2.7326437932041178e-05,
757
+ "loss": 1.238,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.28,
762
+ "learning_rate": 2.7260727574387795e-05,
763
+ "loss": 1.2775,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.28,
768
+ "learning_rate": 2.7194300509827035e-05,
769
+ "loss": 1.2597,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.28,
774
+ "learning_rate": 2.7127160621385888e-05,
775
+ "loss": 1.2922,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.28,
780
+ "learning_rate": 2.7059311833759813e-05,
781
+ "loss": 1.2436,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.28,
786
+ "learning_rate": 2.6990758113083318e-05,
787
+ "loss": 1.2664,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.29,
792
+ "learning_rate": 2.692150346669812e-05,
793
+ "loss": 1.2589,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.29,
798
+ "learning_rate": 2.6851551942918905e-05,
799
+ "loss": 1.2152,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.29,
804
+ "learning_rate": 2.6780907630796663e-05,
805
+ "loss": 1.3032,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.29,
810
+ "learning_rate": 2.6709574659879674e-05,
811
+ "loss": 1.2787,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.3,
816
+ "learning_rate": 2.6637557199972115e-05,
817
+ "loss": 1.2237,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.3,
822
+ "learning_rate": 2.656485946089029e-05,
823
+ "loss": 1.23,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.3,
828
+ "learning_rate": 2.6491485692216576e-05,
829
+ "loss": 1.2218,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.3,
834
+ "learning_rate": 2.6417440183050987e-05,
835
+ "loss": 1.2336,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.3,
840
+ "learning_rate": 2.6342727261760464e-05,
841
+ "loss": 1.2625,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.31,
846
+ "learning_rate": 2.626735129572585e-05,
847
+ "loss": 1.2652,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.31,
852
+ "learning_rate": 2.61913166910866e-05,
853
+ "loss": 1.2637,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.31,
858
+ "learning_rate": 2.6114627892483217e-05,
859
+ "loss": 1.2746,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.31,
864
+ "learning_rate": 2.603728938279743e-05,
865
+ "loss": 1.2665,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.32,
870
+ "learning_rate": 2.5959305682890154e-05,
871
+ "loss": 1.2488,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.32,
876
+ "learning_rate": 2.5880681351337224e-05,
877
+ "loss": 1.1899,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.32,
882
+ "learning_rate": 2.5801420984162914e-05,
883
+ "loss": 1.2592,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.32,
888
+ "learning_rate": 2.572152921457126e-05,
889
+ "loss": 1.2979,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.32,
894
+ "learning_rate": 2.5641010712675268e-05,
895
+ "loss": 1.2936,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.33,
900
+ "learning_rate": 2.5559870185223863e-05,
901
+ "loss": 1.2341,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.33,
906
+ "learning_rate": 2.5478112375326796e-05,
907
+ "loss": 1.2528,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.33,
912
+ "learning_rate": 2.5395742062177375e-05,
913
+ "loss": 1.2545,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.33,
918
+ "learning_rate": 2.5312764060773073e-05,
919
+ "loss": 1.2641,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.34,
924
+ "learning_rate": 2.5229183221634098e-05,
925
+ "loss": 1.2035,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.34,
930
+ "learning_rate": 2.514500443051982e-05,
931
+ "loss": 1.3151,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.34,
936
+ "learning_rate": 2.5060232608143207e-05,
937
+ "loss": 1.2686,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.34,
942
+ "learning_rate": 2.4974872709883138e-05,
943
+ "loss": 1.1989,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.34,
948
+ "learning_rate": 2.488892972549479e-05,
949
+ "loss": 1.2634,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.35,
954
+ "learning_rate": 2.48024086788179e-05,
955
+ "loss": 1.2348,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.35,
960
+ "learning_rate": 2.471531462748315e-05,
961
+ "loss": 1.2217,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.35,
966
+ "learning_rate": 2.4627652662616463e-05,
967
+ "loss": 1.2214,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.35,
972
+ "learning_rate": 2.4539427908541462e-05,
973
+ "loss": 1.2326,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.35,
978
+ "learning_rate": 2.4450645522479855e-05,
979
+ "loss": 1.248,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.36,
984
+ "learning_rate": 2.4361310694250044e-05,
985
+ "loss": 1.1883,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.36,
990
+ "learning_rate": 2.4271428645963676e-05,
991
+ "loss": 1.2515,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.36,
996
+ "learning_rate": 2.4181004631720444e-05,
997
+ "loss": 1.2135,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.36,
1002
+ "learning_rate": 2.409004393730091e-05,
1003
+ "loss": 1.1927,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.37,
1008
+ "learning_rate": 2.3998551879857554e-05,
1009
+ "loss": 1.2098,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.37,
1014
+ "learning_rate": 2.3906533807603952e-05,
1015
+ "loss": 1.2371,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.37,
1020
+ "learning_rate": 2.3813995099502108e-05,
1021
+ "loss": 1.2927,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.37,
1026
+ "learning_rate": 2.372094116494809e-05,
1027
+ "loss": 1.2772,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.37,
1032
+ "learning_rate": 2.3627377443455745e-05,
1033
+ "loss": 1.2128,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.38,
1038
+ "learning_rate": 2.3533309404338802e-05,
1039
+ "loss": 1.1945,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.38,
1044
+ "learning_rate": 2.343874254639111e-05,
1045
+ "loss": 1.252,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.38,
1050
+ "learning_rate": 2.3343682397565216e-05,
1051
+ "loss": 1.2167,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.38,
1056
+ "learning_rate": 2.324813451464924e-05,
1057
+ "loss": 1.1846,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.39,
1062
+ "learning_rate": 2.3152104482942038e-05,
1063
+ "loss": 1.2399,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.39,
1068
+ "learning_rate": 2.3055597915926706e-05,
1069
+ "loss": 1.2176,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.39,
1074
+ "learning_rate": 2.295862045494246e-05,
1075
+ "loss": 1.2142,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.39,
1080
+ "learning_rate": 2.2861177768854833e-05,
1081
+ "loss": 1.2858,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.39,
1086
+ "learning_rate": 2.2763275553724355e-05,
1087
+ "loss": 1.246,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.4,
1092
+ "learning_rate": 2.2664919532473526e-05,
1093
+ "loss": 1.194,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.4,
1098
+ "learning_rate": 2.2566115454552325e-05,
1099
+ "loss": 1.2541,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.4,
1104
+ "learning_rate": 2.2466869095602093e-05,
1105
+ "loss": 1.2588,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.4,
1110
+ "learning_rate": 2.2367186257117935e-05,
1111
+ "loss": 1.2168,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.41,
1116
+ "learning_rate": 2.2267072766109584e-05,
1117
+ "loss": 1.2172,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.41,
1122
+ "learning_rate": 2.2166534474760776e-05,
1123
+ "loss": 1.2479,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.41,
1128
+ "learning_rate": 2.206557726008717e-05,
1129
+ "loss": 1.194,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.41,
1134
+ "learning_rate": 2.196420702359278e-05,
1135
+ "loss": 1.2847,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.41,
1140
+ "learning_rate": 2.1862429690925028e-05,
1141
+ "loss": 1.2212,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.42,
1146
+ "learning_rate": 2.176025121152834e-05,
1147
+ "loss": 1.1853,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.42,
1152
+ "learning_rate": 2.1657677558296378e-05,
1153
+ "loss": 1.2051,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.42,
1158
+ "learning_rate": 2.155471472722288e-05,
1159
+ "loss": 1.2055,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.42,
1164
+ "learning_rate": 2.145136873705116e-05,
1165
+ "loss": 1.2395,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.42,
1170
+ "learning_rate": 2.134764562892231e-05,
1171
+ "loss": 1.2289,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.43,
1176
+ "learning_rate": 2.1243551466022024e-05,
1177
+ "loss": 1.1809,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.43,
1182
+ "learning_rate": 2.113909233322618e-05,
1183
+ "loss": 1.2361,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.43,
1188
+ "learning_rate": 2.103427433674517e-05,
1189
+ "loss": 1.2121,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.43,
1194
+ "learning_rate": 2.092910360376693e-05,
1195
+ "loss": 1.2141,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.44,
1200
+ "learning_rate": 2.0823586282098773e-05,
1201
+ "loss": 1.2639,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.44,
1206
+ "learning_rate": 2.0717728539808053e-05,
1207
+ "loss": 1.2365,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.44,
1212
+ "learning_rate": 2.0611536564861552e-05,
1213
+ "loss": 1.234,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.44,
1218
+ "learning_rate": 2.0505016564763803e-05,
1219
+ "loss": 1.237,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.44,
1224
+ "learning_rate": 2.039817476619421e-05,
1225
+ "loss": 1.2521,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.45,
1230
+ "learning_rate": 2.029101741464306e-05,
1231
+ "loss": 1.2471,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.45,
1236
+ "learning_rate": 2.0183550774046454e-05,
1237
+ "loss": 1.2258,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.45,
1242
+ "learning_rate": 2.007578112642013e-05,
1243
+ "loss": 1.2378,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.45,
1248
+ "learning_rate": 1.9967714771492266e-05,
1249
+ "loss": 1.2271,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.46,
1254
+ "learning_rate": 1.9859358026335186e-05,
1255
+ "loss": 1.1904,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.46,
1260
+ "learning_rate": 1.9750717224996153e-05,
1261
+ "loss": 1.1951,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.46,
1266
+ "learning_rate": 1.964179871812704e-05,
1267
+ "loss": 1.2221,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.46,
1272
+ "learning_rate": 1.953260887261316e-05,
1273
+ "loss": 1.2414,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.46,
1278
+ "learning_rate": 1.9423154071201056e-05,
1279
+ "loss": 1.2194,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.47,
1284
+ "learning_rate": 1.931344071212539e-05,
1285
+ "loss": 1.1928,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.47,
1290
+ "learning_rate": 1.920347520873494e-05,
1291
+ "loss": 1.2564,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.47,
1296
+ "learning_rate": 1.909326398911772e-05,
1297
+ "loss": 1.2236,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.47,
1302
+ "learning_rate": 1.8982813495725202e-05,
1303
+ "loss": 1.234,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.48,
1308
+ "learning_rate": 1.8872130184995713e-05,
1309
+ "loss": 1.2259,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.48,
1314
+ "learning_rate": 1.8761220526977046e-05,
1315
+ "loss": 1.1991,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.48,
1320
+ "learning_rate": 1.865009100494823e-05,
1321
+ "loss": 1.2216,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.48,
1326
+ "learning_rate": 1.8538748115040556e-05,
1327
+ "loss": 1.2395,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.48,
1332
+ "learning_rate": 1.8427198365857846e-05,
1333
+ "loss": 1.2205,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.49,
1338
+ "learning_rate": 1.8315448278095976e-05,
1339
+ "loss": 1.2481,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.49,
1344
+ "learning_rate": 1.820350438416171e-05,
1345
+ "loss": 1.2101,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.49,
1350
+ "learning_rate": 1.809137322779086e-05,
1351
+ "loss": 1.2099,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.49,
1356
+ "learning_rate": 1.797906136366574e-05,
1357
+ "loss": 1.2017,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.49,
1362
+ "learning_rate": 1.7866575357032042e-05,
1363
+ "loss": 1.2732,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.5,
1368
+ "learning_rate": 1.7753921783315033e-05,
1369
+ "loss": 1.2328,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.5,
1374
+ "learning_rate": 1.764110722773521e-05,
1375
+ "loss": 1.2229,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.5,
1380
+ "learning_rate": 1.7528138284923326e-05,
1381
+ "loss": 1.2477,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.5,
1386
+ "learning_rate": 1.741502155853493e-05,
1387
+ "loss": 1.1843,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.51,
1392
+ "learning_rate": 1.7301763660864328e-05,
1393
+ "loss": 1.2243,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.51,
1398
+ "learning_rate": 1.7188371212458063e-05,
1399
+ "loss": 1.2417,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.51,
1404
+ "learning_rate": 1.7074850841727914e-05,
1405
+ "loss": 1.2306,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.51,
1410
+ "learning_rate": 1.696120918456341e-05,
1411
+ "loss": 1.2198,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.51,
1416
+ "learning_rate": 1.684745288394396e-05,
1417
+ "loss": 1.233,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.52,
1422
+ "learning_rate": 1.673358858955049e-05,
1423
+ "loss": 1.1856,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.52,
1428
+ "learning_rate": 1.6619622957376762e-05,
1429
+ "loss": 1.22,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.52,
1434
+ "learning_rate": 1.65055626493403e-05,
1435
+ "loss": 1.1693,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.52,
1440
+ "learning_rate": 1.6391414332892942e-05,
1441
+ "loss": 1.3195,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.53,
1446
+ "learning_rate": 1.6277184680631102e-05,
1447
+ "loss": 1.2458,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.53,
1452
+ "learning_rate": 1.616288036990572e-05,
1453
+ "loss": 1.2039,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.53,
1458
+ "learning_rate": 1.6048508082431938e-05,
1459
+ "loss": 1.1957,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.53,
1464
+ "learning_rate": 1.5934074503898504e-05,
1465
+ "loss": 1.2281,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.53,
1470
+ "learning_rate": 1.5819586323576972e-05,
1471
+ "loss": 1.2234,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.54,
1476
+ "learning_rate": 1.5705050233930672e-05,
1477
+ "loss": 1.1819,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.54,
1482
+ "learning_rate": 1.5590472930223484e-05,
1483
+ "loss": 1.2269,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.54,
1488
+ "learning_rate": 1.54758611101285e-05,
1489
+ "loss": 1.2223,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.54,
1494
+ "learning_rate": 1.536122147333646e-05,
1495
+ "loss": 1.2082,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.55,
1500
+ "learning_rate": 1.5246560721164158e-05,
1501
+ "loss": 1.2458,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.55,
1506
+ "learning_rate": 1.5131885556162693e-05,
1507
+ "loss": 1.19,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.55,
1512
+ "learning_rate": 1.5017202681725672e-05,
1513
+ "loss": 1.2061,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.55,
1518
+ "learning_rate": 1.4902518801697365e-05,
1519
+ "loss": 1.218,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.55,
1524
+ "learning_rate": 1.4787840619980825e-05,
1525
+ "loss": 1.2255,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.56,
1530
+ "learning_rate": 1.4673174840146e-05,
1531
+ "loss": 1.2308,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.56,
1536
+ "learning_rate": 1.455852816503788e-05,
1537
+ "loss": 1.2326,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.56,
1542
+ "learning_rate": 1.4443907296384687e-05,
1543
+ "loss": 1.1673,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.56,
1548
+ "learning_rate": 1.4329318934406113e-05,
1549
+ "loss": 1.2031,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.57,
1554
+ "learning_rate": 1.4214769777421652e-05,
1555
+ "loss": 1.1931,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.57,
1560
+ "learning_rate": 1.4100266521459044e-05,
1561
+ "loss": 1.1841,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.57,
1566
+ "learning_rate": 1.3985815859862881e-05,
1567
+ "loss": 1.1648,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.57,
1572
+ "learning_rate": 1.3871424482903308e-05,
1573
+ "loss": 1.152,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.57,
1578
+ "learning_rate": 1.3757099077384965e-05,
1579
+ "loss": 1.2081,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.58,
1584
+ "learning_rate": 1.3642846326256088e-05,
1585
+ "loss": 1.2008,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.58,
1590
+ "learning_rate": 1.3528672908217888e-05,
1591
+ "loss": 1.206,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.58,
1596
+ "learning_rate": 1.34145854973341e-05,
1597
+ "loss": 1.2553,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.58,
1602
+ "learning_rate": 1.3300590762640868e-05,
1603
+ "loss": 1.1694,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.58,
1608
+ "learning_rate": 1.3186695367756916e-05,
1609
+ "loss": 1.2611,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.59,
1614
+ "learning_rate": 1.3072905970494003e-05,
1615
+ "loss": 1.2169,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.59,
1620
+ "learning_rate": 1.295922922246775e-05,
1621
+ "loss": 1.1643,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.59,
1626
+ "learning_rate": 1.28456717687088e-05,
1627
+ "loss": 1.1531,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.59,
1632
+ "learning_rate": 1.2732240247274395e-05,
1633
+ "loss": 1.1715,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.6,
1638
+ "learning_rate": 1.2618941288860352e-05,
1639
+ "loss": 1.2318,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.6,
1644
+ "learning_rate": 1.250578151641344e-05,
1645
+ "loss": 1.221,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.6,
1650
+ "learning_rate": 1.239276754474423e-05,
1651
+ "loss": 1.225,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.6,
1656
+ "learning_rate": 1.2279905980140473e-05,
1657
+ "loss": 1.1759,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.6,
1662
+ "learning_rate": 1.216720341998085e-05,
1663
+ "loss": 1.2252,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.61,
1668
+ "learning_rate": 1.2054666452349389e-05,
1669
+ "loss": 1.1474,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.61,
1674
+ "learning_rate": 1.19423016556503e-05,
1675
+ "loss": 1.1946,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.61,
1680
+ "learning_rate": 1.1830115598223478e-05,
1681
+ "loss": 1.2803,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.61,
1686
+ "learning_rate": 1.171811483796051e-05,
1687
+ "loss": 1.1657,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.62,
1692
+ "learning_rate": 1.1606305921921349e-05,
1693
+ "loss": 1.2366,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.62,
1698
+ "learning_rate": 1.1494695385951596e-05,
1699
+ "loss": 1.2607,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.62,
1704
+ "learning_rate": 1.1383289754300463e-05,
1705
+ "loss": 1.2157,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.62,
1710
+ "learning_rate": 1.1272095539239358e-05,
1711
+ "loss": 1.2014,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.62,
1716
+ "learning_rate": 1.1161119240681233e-05,
1717
+ "loss": 1.2222,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.63,
1722
+ "learning_rate": 1.1050367345800628e-05,
1723
+ "loss": 1.2167,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.63,
1728
+ "learning_rate": 1.0939846328654456e-05,
1729
+ "loss": 1.1899,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.63,
1734
+ "learning_rate": 1.0829562649803551e-05,
1735
+ "loss": 1.1716,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.63,
1740
+ "learning_rate": 1.0719522755935018e-05,
1741
+ "loss": 1.1764,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.64,
1746
+ "learning_rate": 1.0609733079485404e-05,
1747
+ "loss": 1.224,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.64,
1752
+ "learning_rate": 1.0500200038264657e-05,
1753
+ "loss": 1.1807,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.64,
1758
+ "learning_rate": 1.0390930035080994e-05,
1759
+ "loss": 1.1854,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.64,
1764
+ "learning_rate": 1.0281929457366599e-05,
1765
+ "loss": 1.257,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.64,
1770
+ "learning_rate": 1.0173204676804284e-05,
1771
+ "loss": 1.2168,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.65,
1776
+ "learning_rate": 1.0064762048954982e-05,
1777
+ "loss": 1.2345,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.65,
1782
+ "learning_rate": 9.956607912886247e-06,
1783
+ "loss": 1.1904,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.65,
1788
+ "learning_rate": 9.848748590801712e-06,
1789
+ "loss": 1.1636,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.65,
1794
+ "learning_rate": 9.741190387671512e-06,
1795
+ "loss": 1.1864,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.65,
1800
+ "learning_rate": 9.633939590863725e-06,
1801
+ "loss": 1.2384,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.66,
1806
+ "learning_rate": 9.527002469776827e-06,
1807
+ "loss": 1.1556,
1808
+ "step": 3000
1809
+ }
1810
+ ],
1811
+ "max_steps": 4566,
1812
+ "num_train_epochs": 1,
1813
+ "total_flos": 0.0,
1814
+ "trial_name": null,
1815
+ "trial_params": null
1816
+ }
contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-4000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b85346bd68b7ceeb8a70626225fb65145986c2b681863b1e3242a8b56beacc02
3
+ size 1942236130
contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,2416 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.8760402978537013,
5
+ "global_step": 4000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 6.564551422319475e-07,
13
+ "loss": 1.6933,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 1.312910284463895e-06,
19
+ "loss": 1.6626,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.9693654266958425e-06,
25
+ "loss": 1.6637,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 2.62582056892779e-06,
31
+ "loss": 1.6255,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 3.2822757111597377e-06,
37
+ "loss": 1.6863,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 3.938730853391685e-06,
43
+ "loss": 1.7109,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.595185995623632e-06,
49
+ "loss": 1.6576,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 5.25164113785558e-06,
55
+ "loss": 1.6035,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 5.908096280087528e-06,
61
+ "loss": 1.576,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 6.564551422319475e-06,
67
+ "loss": 1.6065,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 7.221006564551422e-06,
73
+ "loss": 1.5913,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 7.87746170678337e-06,
79
+ "loss": 1.4796,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 8.533916849015316e-06,
85
+ "loss": 1.5987,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 9.190371991247264e-06,
91
+ "loss": 1.4971,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 9.846827133479212e-06,
97
+ "loss": 1.5054,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 1.050328227571116e-05,
103
+ "loss": 1.5211,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 1.1159737417943107e-05,
109
+ "loss": 1.5883,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 1.1816192560175055e-05,
115
+ "loss": 1.5497,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 1.2472647702407003e-05,
121
+ "loss": 1.5075,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.04,
126
+ "learning_rate": 1.312910284463895e-05,
127
+ "loss": 1.4538,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 1.3785557986870897e-05,
133
+ "loss": 1.4766,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 1.4442013129102845e-05,
139
+ "loss": 1.4861,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 1.509846827133479e-05,
145
+ "loss": 1.5375,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.05,
150
+ "learning_rate": 1.575492341356674e-05,
151
+ "loss": 1.4774,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.05,
156
+ "learning_rate": 1.6411378555798686e-05,
157
+ "loss": 1.4756,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 1.7067833698030632e-05,
163
+ "loss": 1.4494,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 1.7724288840262582e-05,
169
+ "loss": 1.4283,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 1.8380743982494528e-05,
175
+ "loss": 1.382,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.06,
180
+ "learning_rate": 1.9037199124726478e-05,
181
+ "loss": 1.4212,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 1.9693654266958424e-05,
187
+ "loss": 1.4249,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 2.0350109409190373e-05,
193
+ "loss": 1.4774,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 2.100656455142232e-05,
199
+ "loss": 1.4058,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.07,
204
+ "learning_rate": 2.1663019693654265e-05,
205
+ "loss": 1.4106,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.07,
210
+ "learning_rate": 2.2319474835886215e-05,
211
+ "loss": 1.4115,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 2.297592997811816e-05,
217
+ "loss": 1.4148,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 2.363238512035011e-05,
223
+ "loss": 1.4041,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.08,
228
+ "learning_rate": 2.4288840262582056e-05,
229
+ "loss": 1.354,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.08,
234
+ "learning_rate": 2.4945295404814006e-05,
235
+ "loss": 1.4267,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 2.5601750547045952e-05,
241
+ "loss": 1.4255,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 2.62582056892779e-05,
247
+ "loss": 1.3738,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.09,
252
+ "learning_rate": 2.6914660831509848e-05,
253
+ "loss": 1.4188,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.09,
258
+ "learning_rate": 2.7571115973741794e-05,
259
+ "loss": 1.3703,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.09,
264
+ "learning_rate": 2.8227571115973743e-05,
265
+ "loss": 1.3946,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 2.888402625820569e-05,
271
+ "loss": 1.3408,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 2.954048140043764e-05,
277
+ "loss": 1.3751,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.1,
282
+ "learning_rate": 2.9999960542365528e-05,
283
+ "loss": 1.3858,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.1,
288
+ "learning_rate": 2.9999259079083055e-05,
289
+ "loss": 1.3235,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 2.9997680826676758e-05,
295
+ "loss": 1.394,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 2.9995225877404156e-05,
301
+ "loss": 1.3088,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.11,
306
+ "learning_rate": 2.9991894374770534e-05,
307
+ "loss": 1.2879,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.11,
312
+ "learning_rate": 2.9987686513520527e-05,
313
+ "loss": 1.3771,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.11,
318
+ "learning_rate": 2.9982602539626753e-05,
319
+ "loss": 1.3274,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 2.997664275027542e-05,
325
+ "loss": 1.364,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.12,
330
+ "learning_rate": 2.996980749384897e-05,
331
+ "loss": 1.3429,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.12,
336
+ "learning_rate": 2.9962097169905706e-05,
337
+ "loss": 1.3442,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.12,
342
+ "learning_rate": 2.995351222915644e-05,
343
+ "loss": 1.3834,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.12,
348
+ "learning_rate": 2.9944053173438132e-05,
349
+ "loss": 1.3202,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 2.9933720555684573e-05,
355
+ "loss": 1.2823,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.13,
360
+ "learning_rate": 2.992251497989406e-05,
361
+ "loss": 1.3072,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.13,
366
+ "learning_rate": 2.9910437101094075e-05,
367
+ "loss": 1.3526,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.13,
372
+ "learning_rate": 2.9897487625303016e-05,
373
+ "loss": 1.2923,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 2.9883667309488905e-05,
379
+ "loss": 1.3429,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.14,
384
+ "learning_rate": 2.9868976961525146e-05,
385
+ "loss": 1.2892,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.14,
390
+ "learning_rate": 2.985341744014332e-05,
391
+ "loss": 1.2687,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.14,
396
+ "learning_rate": 2.983698965488296e-05,
397
+ "loss": 1.3028,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.14,
402
+ "learning_rate": 2.9819694566038392e-05,
403
+ "loss": 1.2969,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.15,
408
+ "learning_rate": 2.980153318460261e-05,
409
+ "loss": 1.2961,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.15,
414
+ "learning_rate": 2.9782506572208166e-05,
415
+ "loss": 1.3652,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.15,
420
+ "learning_rate": 2.976261584106512e-05,
421
+ "loss": 1.3636,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.15,
426
+ "learning_rate": 2.974186215389601e-05,
427
+ "loss": 1.3004,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 2.9720246723867926e-05,
433
+ "loss": 1.319,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.16,
438
+ "learning_rate": 2.9697770814521526e-05,
439
+ "loss": 1.3539,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.16,
444
+ "learning_rate": 2.9674435739697234e-05,
445
+ "loss": 1.3482,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.16,
450
+ "learning_rate": 2.9650242863458425e-05,
451
+ "loss": 1.2532,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.16,
456
+ "learning_rate": 2.9625193600011658e-05,
457
+ "loss": 1.3419,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.17,
462
+ "learning_rate": 2.959928941362405e-05,
463
+ "loss": 1.315,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.17,
468
+ "learning_rate": 2.9572531818537654e-05,
469
+ "loss": 1.3297,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.17,
474
+ "learning_rate": 2.9544922378880957e-05,
475
+ "loss": 1.2816,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.17,
480
+ "learning_rate": 2.951646270857743e-05,
481
+ "loss": 1.3217,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.18,
486
+ "learning_rate": 2.9487154471251217e-05,
487
+ "loss": 1.3086,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.18,
492
+ "learning_rate": 2.9456999380129854e-05,
493
+ "loss": 1.2999,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.18,
498
+ "learning_rate": 2.942599919794413e-05,
499
+ "loss": 1.2864,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.18,
504
+ "learning_rate": 2.939415573682507e-05,
505
+ "loss": 1.2831,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.18,
510
+ "learning_rate": 2.9361470858197978e-05,
511
+ "loss": 1.3005,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.19,
516
+ "learning_rate": 2.932794647267363e-05,
517
+ "loss": 1.2977,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.19,
522
+ "learning_rate": 2.929358453993659e-05,
523
+ "loss": 1.2909,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.19,
528
+ "learning_rate": 2.925838706863069e-05,
529
+ "loss": 1.3001,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.19,
534
+ "learning_rate": 2.9222356116241536e-05,
535
+ "loss": 1.2478,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.19,
540
+ "learning_rate": 2.9185493788976324e-05,
541
+ "loss": 1.2797,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.2,
546
+ "learning_rate": 2.914780224164066e-05,
547
+ "loss": 1.3308,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.2,
552
+ "learning_rate": 2.9109283677512624e-05,
553
+ "loss": 1.277,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.2,
558
+ "learning_rate": 2.9069940348213978e-05,
559
+ "loss": 1.2825,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.2,
564
+ "learning_rate": 2.9029774553578535e-05,
565
+ "loss": 1.2899,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.21,
570
+ "learning_rate": 2.8988788641517723e-05,
571
+ "loss": 1.2567,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.21,
576
+ "learning_rate": 2.8946985007883354e-05,
577
+ "loss": 1.2643,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.21,
582
+ "learning_rate": 2.890436609632755e-05,
583
+ "loss": 1.3001,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.21,
588
+ "learning_rate": 2.8860934398159905e-05,
589
+ "loss": 1.2995,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.21,
594
+ "learning_rate": 2.881669245220187e-05,
595
+ "loss": 1.2811,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.22,
600
+ "learning_rate": 2.8771642844638316e-05,
601
+ "loss": 1.3262,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.22,
606
+ "learning_rate": 2.8725788208866394e-05,
607
+ "loss": 1.2916,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.22,
612
+ "learning_rate": 2.867913122534155e-05,
613
+ "loss": 1.2962,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.22,
618
+ "learning_rate": 2.8631674621420895e-05,
619
+ "loss": 1.2852,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.23,
624
+ "learning_rate": 2.8583421171203728e-05,
625
+ "loss": 1.2704,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.23,
630
+ "learning_rate": 2.8534373695369393e-05,
631
+ "loss": 1.2264,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.23,
636
+ "learning_rate": 2.8484535061012404e-05,
637
+ "loss": 1.321,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.23,
642
+ "learning_rate": 2.8433908181474845e-05,
643
+ "loss": 1.2498,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.23,
648
+ "learning_rate": 2.8382496016176033e-05,
649
+ "loss": 1.2612,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.24,
654
+ "learning_rate": 2.8330301570439588e-05,
655
+ "loss": 1.2424,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.24,
660
+ "learning_rate": 2.8277327895317703e-05,
661
+ "loss": 1.2828,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.24,
666
+ "learning_rate": 2.8223578087412806e-05,
667
+ "loss": 1.2996,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.24,
672
+ "learning_rate": 2.816905528869657e-05,
673
+ "loss": 1.2903,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.25,
678
+ "learning_rate": 2.8113762686326204e-05,
679
+ "loss": 1.268,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.25,
684
+ "learning_rate": 2.80577035124582e-05,
685
+ "loss": 1.2798,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.25,
690
+ "learning_rate": 2.8000881044059343e-05,
691
+ "loss": 1.2449,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.25,
696
+ "learning_rate": 2.7943298602715196e-05,
697
+ "loss": 1.2877,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.25,
702
+ "learning_rate": 2.7884959554435894e-05,
703
+ "loss": 1.2238,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.26,
708
+ "learning_rate": 2.7825867309459432e-05,
709
+ "loss": 1.2947,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.26,
714
+ "learning_rate": 2.7766025322052274e-05,
715
+ "loss": 1.2752,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.26,
720
+ "learning_rate": 2.7705437090307463e-05,
721
+ "loss": 1.2855,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.26,
726
+ "learning_rate": 2.7644106155940102e-05,
727
+ "loss": 1.2525,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.27,
732
+ "learning_rate": 2.7582036104080366e-05,
733
+ "loss": 1.312,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.27,
738
+ "learning_rate": 2.7519230563063903e-05,
739
+ "loss": 1.2819,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.27,
744
+ "learning_rate": 2.7455693204219735e-05,
745
+ "loss": 1.2797,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.27,
750
+ "learning_rate": 2.7391427741655673e-05,
751
+ "loss": 1.3005,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.27,
756
+ "learning_rate": 2.7326437932041178e-05,
757
+ "loss": 1.238,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.28,
762
+ "learning_rate": 2.7260727574387795e-05,
763
+ "loss": 1.2775,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.28,
768
+ "learning_rate": 2.7194300509827035e-05,
769
+ "loss": 1.2597,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.28,
774
+ "learning_rate": 2.7127160621385888e-05,
775
+ "loss": 1.2922,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.28,
780
+ "learning_rate": 2.7059311833759813e-05,
781
+ "loss": 1.2436,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.28,
786
+ "learning_rate": 2.6990758113083318e-05,
787
+ "loss": 1.2664,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.29,
792
+ "learning_rate": 2.692150346669812e-05,
793
+ "loss": 1.2589,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.29,
798
+ "learning_rate": 2.6851551942918905e-05,
799
+ "loss": 1.2152,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.29,
804
+ "learning_rate": 2.6780907630796663e-05,
805
+ "loss": 1.3032,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.29,
810
+ "learning_rate": 2.6709574659879674e-05,
811
+ "loss": 1.2787,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.3,
816
+ "learning_rate": 2.6637557199972115e-05,
817
+ "loss": 1.2237,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.3,
822
+ "learning_rate": 2.656485946089029e-05,
823
+ "loss": 1.23,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.3,
828
+ "learning_rate": 2.6491485692216576e-05,
829
+ "loss": 1.2218,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.3,
834
+ "learning_rate": 2.6417440183050987e-05,
835
+ "loss": 1.2336,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.3,
840
+ "learning_rate": 2.6342727261760464e-05,
841
+ "loss": 1.2625,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.31,
846
+ "learning_rate": 2.626735129572585e-05,
847
+ "loss": 1.2652,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.31,
852
+ "learning_rate": 2.61913166910866e-05,
853
+ "loss": 1.2637,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.31,
858
+ "learning_rate": 2.6114627892483217e-05,
859
+ "loss": 1.2746,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.31,
864
+ "learning_rate": 2.603728938279743e-05,
865
+ "loss": 1.2665,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.32,
870
+ "learning_rate": 2.5959305682890154e-05,
871
+ "loss": 1.2488,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.32,
876
+ "learning_rate": 2.5880681351337224e-05,
877
+ "loss": 1.1899,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.32,
882
+ "learning_rate": 2.5801420984162914e-05,
883
+ "loss": 1.2592,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.32,
888
+ "learning_rate": 2.572152921457126e-05,
889
+ "loss": 1.2979,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.32,
894
+ "learning_rate": 2.5641010712675268e-05,
895
+ "loss": 1.2936,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.33,
900
+ "learning_rate": 2.5559870185223863e-05,
901
+ "loss": 1.2341,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.33,
906
+ "learning_rate": 2.5478112375326796e-05,
907
+ "loss": 1.2528,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.33,
912
+ "learning_rate": 2.5395742062177375e-05,
913
+ "loss": 1.2545,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.33,
918
+ "learning_rate": 2.5312764060773073e-05,
919
+ "loss": 1.2641,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.34,
924
+ "learning_rate": 2.5229183221634098e-05,
925
+ "loss": 1.2035,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.34,
930
+ "learning_rate": 2.514500443051982e-05,
931
+ "loss": 1.3151,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.34,
936
+ "learning_rate": 2.5060232608143207e-05,
937
+ "loss": 1.2686,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.34,
942
+ "learning_rate": 2.4974872709883138e-05,
943
+ "loss": 1.1989,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.34,
948
+ "learning_rate": 2.488892972549479e-05,
949
+ "loss": 1.2634,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.35,
954
+ "learning_rate": 2.48024086788179e-05,
955
+ "loss": 1.2348,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.35,
960
+ "learning_rate": 2.471531462748315e-05,
961
+ "loss": 1.2217,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.35,
966
+ "learning_rate": 2.4627652662616463e-05,
967
+ "loss": 1.2214,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.35,
972
+ "learning_rate": 2.4539427908541462e-05,
973
+ "loss": 1.2326,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.35,
978
+ "learning_rate": 2.4450645522479855e-05,
979
+ "loss": 1.248,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.36,
984
+ "learning_rate": 2.4361310694250044e-05,
985
+ "loss": 1.1883,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.36,
990
+ "learning_rate": 2.4271428645963676e-05,
991
+ "loss": 1.2515,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.36,
996
+ "learning_rate": 2.4181004631720444e-05,
997
+ "loss": 1.2135,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.36,
1002
+ "learning_rate": 2.409004393730091e-05,
1003
+ "loss": 1.1927,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.37,
1008
+ "learning_rate": 2.3998551879857554e-05,
1009
+ "loss": 1.2098,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.37,
1014
+ "learning_rate": 2.3906533807603952e-05,
1015
+ "loss": 1.2371,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.37,
1020
+ "learning_rate": 2.3813995099502108e-05,
1021
+ "loss": 1.2927,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.37,
1026
+ "learning_rate": 2.372094116494809e-05,
1027
+ "loss": 1.2772,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.37,
1032
+ "learning_rate": 2.3627377443455745e-05,
1033
+ "loss": 1.2128,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.38,
1038
+ "learning_rate": 2.3533309404338802e-05,
1039
+ "loss": 1.1945,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.38,
1044
+ "learning_rate": 2.343874254639111e-05,
1045
+ "loss": 1.252,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.38,
1050
+ "learning_rate": 2.3343682397565216e-05,
1051
+ "loss": 1.2167,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.38,
1056
+ "learning_rate": 2.324813451464924e-05,
1057
+ "loss": 1.1846,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.39,
1062
+ "learning_rate": 2.3152104482942038e-05,
1063
+ "loss": 1.2399,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.39,
1068
+ "learning_rate": 2.3055597915926706e-05,
1069
+ "loss": 1.2176,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.39,
1074
+ "learning_rate": 2.295862045494246e-05,
1075
+ "loss": 1.2142,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.39,
1080
+ "learning_rate": 2.2861177768854833e-05,
1081
+ "loss": 1.2858,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.39,
1086
+ "learning_rate": 2.2763275553724355e-05,
1087
+ "loss": 1.246,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.4,
1092
+ "learning_rate": 2.2664919532473526e-05,
1093
+ "loss": 1.194,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.4,
1098
+ "learning_rate": 2.2566115454552325e-05,
1099
+ "loss": 1.2541,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.4,
1104
+ "learning_rate": 2.2466869095602093e-05,
1105
+ "loss": 1.2588,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.4,
1110
+ "learning_rate": 2.2367186257117935e-05,
1111
+ "loss": 1.2168,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.41,
1116
+ "learning_rate": 2.2267072766109584e-05,
1117
+ "loss": 1.2172,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.41,
1122
+ "learning_rate": 2.2166534474760776e-05,
1123
+ "loss": 1.2479,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.41,
1128
+ "learning_rate": 2.206557726008717e-05,
1129
+ "loss": 1.194,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.41,
1134
+ "learning_rate": 2.196420702359278e-05,
1135
+ "loss": 1.2847,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.41,
1140
+ "learning_rate": 2.1862429690925028e-05,
1141
+ "loss": 1.2212,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.42,
1146
+ "learning_rate": 2.176025121152834e-05,
1147
+ "loss": 1.1853,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.42,
1152
+ "learning_rate": 2.1657677558296378e-05,
1153
+ "loss": 1.2051,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.42,
1158
+ "learning_rate": 2.155471472722288e-05,
1159
+ "loss": 1.2055,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.42,
1164
+ "learning_rate": 2.145136873705116e-05,
1165
+ "loss": 1.2395,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.42,
1170
+ "learning_rate": 2.134764562892231e-05,
1171
+ "loss": 1.2289,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.43,
1176
+ "learning_rate": 2.1243551466022024e-05,
1177
+ "loss": 1.1809,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.43,
1182
+ "learning_rate": 2.113909233322618e-05,
1183
+ "loss": 1.2361,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.43,
1188
+ "learning_rate": 2.103427433674517e-05,
1189
+ "loss": 1.2121,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.43,
1194
+ "learning_rate": 2.092910360376693e-05,
1195
+ "loss": 1.2141,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.44,
1200
+ "learning_rate": 2.0823586282098773e-05,
1201
+ "loss": 1.2639,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.44,
1206
+ "learning_rate": 2.0717728539808053e-05,
1207
+ "loss": 1.2365,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.44,
1212
+ "learning_rate": 2.0611536564861552e-05,
1213
+ "loss": 1.234,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.44,
1218
+ "learning_rate": 2.0505016564763803e-05,
1219
+ "loss": 1.237,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.44,
1224
+ "learning_rate": 2.039817476619421e-05,
1225
+ "loss": 1.2521,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.45,
1230
+ "learning_rate": 2.029101741464306e-05,
1231
+ "loss": 1.2471,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.45,
1236
+ "learning_rate": 2.0183550774046454e-05,
1237
+ "loss": 1.2258,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.45,
1242
+ "learning_rate": 2.007578112642013e-05,
1243
+ "loss": 1.2378,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.45,
1248
+ "learning_rate": 1.9967714771492266e-05,
1249
+ "loss": 1.2271,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.46,
1254
+ "learning_rate": 1.9859358026335186e-05,
1255
+ "loss": 1.1904,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.46,
1260
+ "learning_rate": 1.9750717224996153e-05,
1261
+ "loss": 1.1951,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.46,
1266
+ "learning_rate": 1.964179871812704e-05,
1267
+ "loss": 1.2221,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.46,
1272
+ "learning_rate": 1.953260887261316e-05,
1273
+ "loss": 1.2414,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.46,
1278
+ "learning_rate": 1.9423154071201056e-05,
1279
+ "loss": 1.2194,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.47,
1284
+ "learning_rate": 1.931344071212539e-05,
1285
+ "loss": 1.1928,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.47,
1290
+ "learning_rate": 1.920347520873494e-05,
1291
+ "loss": 1.2564,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.47,
1296
+ "learning_rate": 1.909326398911772e-05,
1297
+ "loss": 1.2236,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.47,
1302
+ "learning_rate": 1.8982813495725202e-05,
1303
+ "loss": 1.234,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.48,
1308
+ "learning_rate": 1.8872130184995713e-05,
1309
+ "loss": 1.2259,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.48,
1314
+ "learning_rate": 1.8761220526977046e-05,
1315
+ "loss": 1.1991,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.48,
1320
+ "learning_rate": 1.865009100494823e-05,
1321
+ "loss": 1.2216,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.48,
1326
+ "learning_rate": 1.8538748115040556e-05,
1327
+ "loss": 1.2395,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.48,
1332
+ "learning_rate": 1.8427198365857846e-05,
1333
+ "loss": 1.2205,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.49,
1338
+ "learning_rate": 1.8315448278095976e-05,
1339
+ "loss": 1.2481,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.49,
1344
+ "learning_rate": 1.820350438416171e-05,
1345
+ "loss": 1.2101,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.49,
1350
+ "learning_rate": 1.809137322779086e-05,
1351
+ "loss": 1.2099,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.49,
1356
+ "learning_rate": 1.797906136366574e-05,
1357
+ "loss": 1.2017,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.49,
1362
+ "learning_rate": 1.7866575357032042e-05,
1363
+ "loss": 1.2732,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.5,
1368
+ "learning_rate": 1.7753921783315033e-05,
1369
+ "loss": 1.2328,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.5,
1374
+ "learning_rate": 1.764110722773521e-05,
1375
+ "loss": 1.2229,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.5,
1380
+ "learning_rate": 1.7528138284923326e-05,
1381
+ "loss": 1.2477,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.5,
1386
+ "learning_rate": 1.741502155853493e-05,
1387
+ "loss": 1.1843,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.51,
1392
+ "learning_rate": 1.7301763660864328e-05,
1393
+ "loss": 1.2243,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.51,
1398
+ "learning_rate": 1.7188371212458063e-05,
1399
+ "loss": 1.2417,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.51,
1404
+ "learning_rate": 1.7074850841727914e-05,
1405
+ "loss": 1.2306,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.51,
1410
+ "learning_rate": 1.696120918456341e-05,
1411
+ "loss": 1.2198,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.51,
1416
+ "learning_rate": 1.684745288394396e-05,
1417
+ "loss": 1.233,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.52,
1422
+ "learning_rate": 1.673358858955049e-05,
1423
+ "loss": 1.1856,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.52,
1428
+ "learning_rate": 1.6619622957376762e-05,
1429
+ "loss": 1.22,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.52,
1434
+ "learning_rate": 1.65055626493403e-05,
1435
+ "loss": 1.1693,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.52,
1440
+ "learning_rate": 1.6391414332892942e-05,
1441
+ "loss": 1.3195,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.53,
1446
+ "learning_rate": 1.6277184680631102e-05,
1447
+ "loss": 1.2458,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.53,
1452
+ "learning_rate": 1.616288036990572e-05,
1453
+ "loss": 1.2039,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.53,
1458
+ "learning_rate": 1.6048508082431938e-05,
1459
+ "loss": 1.1957,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.53,
1464
+ "learning_rate": 1.5934074503898504e-05,
1465
+ "loss": 1.2281,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.53,
1470
+ "learning_rate": 1.5819586323576972e-05,
1471
+ "loss": 1.2234,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.54,
1476
+ "learning_rate": 1.5705050233930672e-05,
1477
+ "loss": 1.1819,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.54,
1482
+ "learning_rate": 1.5590472930223484e-05,
1483
+ "loss": 1.2269,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.54,
1488
+ "learning_rate": 1.54758611101285e-05,
1489
+ "loss": 1.2223,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.54,
1494
+ "learning_rate": 1.536122147333646e-05,
1495
+ "loss": 1.2082,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.55,
1500
+ "learning_rate": 1.5246560721164158e-05,
1501
+ "loss": 1.2458,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.55,
1506
+ "learning_rate": 1.5131885556162693e-05,
1507
+ "loss": 1.19,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.55,
1512
+ "learning_rate": 1.5017202681725672e-05,
1513
+ "loss": 1.2061,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.55,
1518
+ "learning_rate": 1.4902518801697365e-05,
1519
+ "loss": 1.218,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.55,
1524
+ "learning_rate": 1.4787840619980825e-05,
1525
+ "loss": 1.2255,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.56,
1530
+ "learning_rate": 1.4673174840146e-05,
1531
+ "loss": 1.2308,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.56,
1536
+ "learning_rate": 1.455852816503788e-05,
1537
+ "loss": 1.2326,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.56,
1542
+ "learning_rate": 1.4443907296384687e-05,
1543
+ "loss": 1.1673,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.56,
1548
+ "learning_rate": 1.4329318934406113e-05,
1549
+ "loss": 1.2031,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.57,
1554
+ "learning_rate": 1.4214769777421652e-05,
1555
+ "loss": 1.1931,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.57,
1560
+ "learning_rate": 1.4100266521459044e-05,
1561
+ "loss": 1.1841,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.57,
1566
+ "learning_rate": 1.3985815859862881e-05,
1567
+ "loss": 1.1648,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.57,
1572
+ "learning_rate": 1.3871424482903308e-05,
1573
+ "loss": 1.152,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.57,
1578
+ "learning_rate": 1.3757099077384965e-05,
1579
+ "loss": 1.2081,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.58,
1584
+ "learning_rate": 1.3642846326256088e-05,
1585
+ "loss": 1.2008,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.58,
1590
+ "learning_rate": 1.3528672908217888e-05,
1591
+ "loss": 1.206,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.58,
1596
+ "learning_rate": 1.34145854973341e-05,
1597
+ "loss": 1.2553,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.58,
1602
+ "learning_rate": 1.3300590762640868e-05,
1603
+ "loss": 1.1694,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.58,
1608
+ "learning_rate": 1.3186695367756916e-05,
1609
+ "loss": 1.2611,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.59,
1614
+ "learning_rate": 1.3072905970494003e-05,
1615
+ "loss": 1.2169,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.59,
1620
+ "learning_rate": 1.295922922246775e-05,
1621
+ "loss": 1.1643,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.59,
1626
+ "learning_rate": 1.28456717687088e-05,
1627
+ "loss": 1.1531,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.59,
1632
+ "learning_rate": 1.2732240247274395e-05,
1633
+ "loss": 1.1715,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.6,
1638
+ "learning_rate": 1.2618941288860352e-05,
1639
+ "loss": 1.2318,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.6,
1644
+ "learning_rate": 1.250578151641344e-05,
1645
+ "loss": 1.221,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.6,
1650
+ "learning_rate": 1.239276754474423e-05,
1651
+ "loss": 1.225,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.6,
1656
+ "learning_rate": 1.2279905980140473e-05,
1657
+ "loss": 1.1759,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.6,
1662
+ "learning_rate": 1.216720341998085e-05,
1663
+ "loss": 1.2252,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.61,
1668
+ "learning_rate": 1.2054666452349389e-05,
1669
+ "loss": 1.1474,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.61,
1674
+ "learning_rate": 1.19423016556503e-05,
1675
+ "loss": 1.1946,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.61,
1680
+ "learning_rate": 1.1830115598223478e-05,
1681
+ "loss": 1.2803,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.61,
1686
+ "learning_rate": 1.171811483796051e-05,
1687
+ "loss": 1.1657,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.62,
1692
+ "learning_rate": 1.1606305921921349e-05,
1693
+ "loss": 1.2366,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.62,
1698
+ "learning_rate": 1.1494695385951596e-05,
1699
+ "loss": 1.2607,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.62,
1704
+ "learning_rate": 1.1383289754300463e-05,
1705
+ "loss": 1.2157,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.62,
1710
+ "learning_rate": 1.1272095539239358e-05,
1711
+ "loss": 1.2014,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.62,
1716
+ "learning_rate": 1.1161119240681233e-05,
1717
+ "loss": 1.2222,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.63,
1722
+ "learning_rate": 1.1050367345800628e-05,
1723
+ "loss": 1.2167,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.63,
1728
+ "learning_rate": 1.0939846328654456e-05,
1729
+ "loss": 1.1899,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.63,
1734
+ "learning_rate": 1.0829562649803551e-05,
1735
+ "loss": 1.1716,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.63,
1740
+ "learning_rate": 1.0719522755935018e-05,
1741
+ "loss": 1.1764,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.64,
1746
+ "learning_rate": 1.0609733079485404e-05,
1747
+ "loss": 1.224,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.64,
1752
+ "learning_rate": 1.0500200038264657e-05,
1753
+ "loss": 1.1807,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.64,
1758
+ "learning_rate": 1.0390930035080994e-05,
1759
+ "loss": 1.1854,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.64,
1764
+ "learning_rate": 1.0281929457366599e-05,
1765
+ "loss": 1.257,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.64,
1770
+ "learning_rate": 1.0173204676804284e-05,
1771
+ "loss": 1.2168,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.65,
1776
+ "learning_rate": 1.0064762048954982e-05,
1777
+ "loss": 1.2345,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.65,
1782
+ "learning_rate": 9.956607912886247e-06,
1783
+ "loss": 1.1904,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.65,
1788
+ "learning_rate": 9.848748590801712e-06,
1789
+ "loss": 1.1636,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.65,
1794
+ "learning_rate": 9.741190387671512e-06,
1795
+ "loss": 1.1864,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.65,
1800
+ "learning_rate": 9.633939590863725e-06,
1801
+ "loss": 1.2384,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.66,
1806
+ "learning_rate": 9.527002469776827e-06,
1807
+ "loss": 1.1556,
1808
+ "step": 3000
1809
+ },
1810
+ {
1811
+ "epoch": 0.66,
1812
+ "learning_rate": 9.420385275473244e-06,
1813
+ "loss": 1.1727,
1814
+ "step": 3010
1815
+ },
1816
+ {
1817
+ "epoch": 0.66,
1818
+ "learning_rate": 9.314094240313913e-06,
1819
+ "loss": 1.2124,
1820
+ "step": 3020
1821
+ },
1822
+ {
1823
+ "epoch": 0.66,
1824
+ "learning_rate": 9.208135577593974e-06,
1825
+ "loss": 1.2529,
1826
+ "step": 3030
1827
+ },
1828
+ {
1829
+ "epoch": 0.67,
1830
+ "learning_rate": 9.102515481179567e-06,
1831
+ "loss": 1.183,
1832
+ "step": 3040
1833
+ },
1834
+ {
1835
+ "epoch": 0.67,
1836
+ "learning_rate": 8.997240125145793e-06,
1837
+ "loss": 1.2023,
1838
+ "step": 3050
1839
+ },
1840
+ {
1841
+ "epoch": 0.67,
1842
+ "learning_rate": 8.892315663415763e-06,
1843
+ "loss": 1.173,
1844
+ "step": 3060
1845
+ },
1846
+ {
1847
+ "epoch": 0.67,
1848
+ "learning_rate": 8.787748229400886e-06,
1849
+ "loss": 1.178,
1850
+ "step": 3070
1851
+ },
1852
+ {
1853
+ "epoch": 0.67,
1854
+ "learning_rate": 8.683543935642351e-06,
1855
+ "loss": 1.1542,
1856
+ "step": 3080
1857
+ },
1858
+ {
1859
+ "epoch": 0.68,
1860
+ "learning_rate": 8.579708873453808e-06,
1861
+ "loss": 1.1853,
1862
+ "step": 3090
1863
+ },
1864
+ {
1865
+ "epoch": 0.68,
1866
+ "learning_rate": 8.476249112565269e-06,
1867
+ "loss": 1.2455,
1868
+ "step": 3100
1869
+ },
1870
+ {
1871
+ "epoch": 0.68,
1872
+ "learning_rate": 8.37317070076834e-06,
1873
+ "loss": 1.217,
1874
+ "step": 3110
1875
+ },
1876
+ {
1877
+ "epoch": 0.68,
1878
+ "learning_rate": 8.270479663562674e-06,
1879
+ "loss": 1.2092,
1880
+ "step": 3120
1881
+ },
1882
+ {
1883
+ "epoch": 0.69,
1884
+ "learning_rate": 8.168182003803752e-06,
1885
+ "loss": 1.1525,
1886
+ "step": 3130
1887
+ },
1888
+ {
1889
+ "epoch": 0.69,
1890
+ "learning_rate": 8.066283701351982e-06,
1891
+ "loss": 1.1955,
1892
+ "step": 3140
1893
+ },
1894
+ {
1895
+ "epoch": 0.69,
1896
+ "learning_rate": 7.964790712723124e-06,
1897
+ "loss": 1.1945,
1898
+ "step": 3150
1899
+ },
1900
+ {
1901
+ "epoch": 0.69,
1902
+ "learning_rate": 7.863708970740155e-06,
1903
+ "loss": 1.1492,
1904
+ "step": 3160
1905
+ },
1906
+ {
1907
+ "epoch": 0.69,
1908
+ "learning_rate": 7.763044384186388e-06,
1909
+ "loss": 1.2082,
1910
+ "step": 3170
1911
+ },
1912
+ {
1913
+ "epoch": 0.7,
1914
+ "learning_rate": 7.662802837460131e-06,
1915
+ "loss": 1.2047,
1916
+ "step": 3180
1917
+ },
1918
+ {
1919
+ "epoch": 0.7,
1920
+ "learning_rate": 7.5629901902306696e-06,
1921
+ "loss": 1.1821,
1922
+ "step": 3190
1923
+ },
1924
+ {
1925
+ "epoch": 0.7,
1926
+ "learning_rate": 7.463612277095793e-06,
1927
+ "loss": 1.2059,
1928
+ "step": 3200
1929
+ },
1930
+ {
1931
+ "epoch": 0.7,
1932
+ "learning_rate": 7.364674907240651e-06,
1933
+ "loss": 1.255,
1934
+ "step": 3210
1935
+ },
1936
+ {
1937
+ "epoch": 0.71,
1938
+ "learning_rate": 7.2661838640982405e-06,
1939
+ "loss": 1.2092,
1940
+ "step": 3220
1941
+ },
1942
+ {
1943
+ "epoch": 0.71,
1944
+ "learning_rate": 7.16814490501131e-06,
1945
+ "loss": 1.1946,
1946
+ "step": 3230
1947
+ },
1948
+ {
1949
+ "epoch": 0.71,
1950
+ "learning_rate": 7.070563760895802e-06,
1951
+ "loss": 1.2,
1952
+ "step": 3240
1953
+ },
1954
+ {
1955
+ "epoch": 0.71,
1956
+ "learning_rate": 6.97344613590587e-06,
1957
+ "loss": 1.1783,
1958
+ "step": 3250
1959
+ },
1960
+ {
1961
+ "epoch": 0.71,
1962
+ "learning_rate": 6.876797707100399e-06,
1963
+ "loss": 1.1666,
1964
+ "step": 3260
1965
+ },
1966
+ {
1967
+ "epoch": 0.72,
1968
+ "learning_rate": 6.780624124111224e-06,
1969
+ "loss": 1.2223,
1970
+ "step": 3270
1971
+ },
1972
+ {
1973
+ "epoch": 0.72,
1974
+ "learning_rate": 6.684931008812788e-06,
1975
+ "loss": 1.1982,
1976
+ "step": 3280
1977
+ },
1978
+ {
1979
+ "epoch": 0.72,
1980
+ "learning_rate": 6.589723954993589e-06,
1981
+ "loss": 1.16,
1982
+ "step": 3290
1983
+ },
1984
+ {
1985
+ "epoch": 0.72,
1986
+ "learning_rate": 6.495008528029127e-06,
1987
+ "loss": 1.2192,
1988
+ "step": 3300
1989
+ },
1990
+ {
1991
+ "epoch": 0.72,
1992
+ "learning_rate": 6.4007902645566565e-06,
1993
+ "loss": 1.1691,
1994
+ "step": 3310
1995
+ },
1996
+ {
1997
+ "epoch": 0.73,
1998
+ "learning_rate": 6.30707467215145e-06,
1999
+ "loss": 1.1704,
2000
+ "step": 3320
2001
+ },
2002
+ {
2003
+ "epoch": 0.73,
2004
+ "learning_rate": 6.213867229004918e-06,
2005
+ "loss": 1.177,
2006
+ "step": 3330
2007
+ },
2008
+ {
2009
+ "epoch": 0.73,
2010
+ "learning_rate": 6.121173383604346e-06,
2011
+ "loss": 1.2047,
2012
+ "step": 3340
2013
+ },
2014
+ {
2015
+ "epoch": 0.73,
2016
+ "learning_rate": 6.028998554414411e-06,
2017
+ "loss": 1.197,
2018
+ "step": 3350
2019
+ },
2020
+ {
2021
+ "epoch": 0.74,
2022
+ "learning_rate": 5.9373481295604464e-06,
2023
+ "loss": 1.1881,
2024
+ "step": 3360
2025
+ },
2026
+ {
2027
+ "epoch": 0.74,
2028
+ "learning_rate": 5.8462274665134505e-06,
2029
+ "loss": 1.1489,
2030
+ "step": 3370
2031
+ },
2032
+ {
2033
+ "epoch": 0.74,
2034
+ "learning_rate": 5.755641891776954e-06,
2035
+ "loss": 1.206,
2036
+ "step": 3380
2037
+ },
2038
+ {
2039
+ "epoch": 0.74,
2040
+ "learning_rate": 5.665596700575622e-06,
2041
+ "loss": 1.1685,
2042
+ "step": 3390
2043
+ },
2044
+ {
2045
+ "epoch": 0.74,
2046
+ "learning_rate": 5.576097156545748e-06,
2047
+ "loss": 1.1586,
2048
+ "step": 3400
2049
+ },
2050
+ {
2051
+ "epoch": 0.75,
2052
+ "learning_rate": 5.487148491427516e-06,
2053
+ "loss": 1.2015,
2054
+ "step": 3410
2055
+ },
2056
+ {
2057
+ "epoch": 0.75,
2058
+ "learning_rate": 5.39875590475926e-06,
2059
+ "loss": 1.2439,
2060
+ "step": 3420
2061
+ },
2062
+ {
2063
+ "epoch": 0.75,
2064
+ "learning_rate": 5.310924563573426e-06,
2065
+ "loss": 1.2239,
2066
+ "step": 3430
2067
+ },
2068
+ {
2069
+ "epoch": 0.75,
2070
+ "learning_rate": 5.223659602094603e-06,
2071
+ "loss": 1.1954,
2072
+ "step": 3440
2073
+ },
2074
+ {
2075
+ "epoch": 0.76,
2076
+ "learning_rate": 5.136966121439367e-06,
2077
+ "loss": 1.2042,
2078
+ "step": 3450
2079
+ },
2080
+ {
2081
+ "epoch": 0.76,
2082
+ "learning_rate": 5.0508491893180994e-06,
2083
+ "loss": 1.1919,
2084
+ "step": 3460
2085
+ },
2086
+ {
2087
+ "epoch": 0.76,
2088
+ "learning_rate": 4.965313839738755e-06,
2089
+ "loss": 1.2301,
2090
+ "step": 3470
2091
+ },
2092
+ {
2093
+ "epoch": 0.76,
2094
+ "learning_rate": 4.880365072712579e-06,
2095
+ "loss": 1.2049,
2096
+ "step": 3480
2097
+ },
2098
+ {
2099
+ "epoch": 0.76,
2100
+ "learning_rate": 4.7960078539618575e-06,
2101
+ "loss": 1.2065,
2102
+ "step": 3490
2103
+ },
2104
+ {
2105
+ "epoch": 0.77,
2106
+ "learning_rate": 4.712247114629627e-06,
2107
+ "loss": 1.2094,
2108
+ "step": 3500
2109
+ },
2110
+ {
2111
+ "epoch": 0.77,
2112
+ "learning_rate": 4.629087750991422e-06,
2113
+ "loss": 1.197,
2114
+ "step": 3510
2115
+ },
2116
+ {
2117
+ "epoch": 0.77,
2118
+ "learning_rate": 4.546534624169054e-06,
2119
+ "loss": 1.2022,
2120
+ "step": 3520
2121
+ },
2122
+ {
2123
+ "epoch": 0.77,
2124
+ "learning_rate": 4.464592559846474e-06,
2125
+ "loss": 1.2126,
2126
+ "step": 3530
2127
+ },
2128
+ {
2129
+ "epoch": 0.78,
2130
+ "learning_rate": 4.383266347987664e-06,
2131
+ "loss": 1.1537,
2132
+ "step": 3540
2133
+ },
2134
+ {
2135
+ "epoch": 0.78,
2136
+ "learning_rate": 4.302560742556651e-06,
2137
+ "loss": 1.2184,
2138
+ "step": 3550
2139
+ },
2140
+ {
2141
+ "epoch": 0.78,
2142
+ "learning_rate": 4.2224804612395985e-06,
2143
+ "loss": 1.2281,
2144
+ "step": 3560
2145
+ },
2146
+ {
2147
+ "epoch": 0.78,
2148
+ "learning_rate": 4.143030185169045e-06,
2149
+ "loss": 1.1872,
2150
+ "step": 3570
2151
+ },
2152
+ {
2153
+ "epoch": 0.78,
2154
+ "learning_rate": 4.064214558650262e-06,
2155
+ "loss": 1.2056,
2156
+ "step": 3580
2157
+ },
2158
+ {
2159
+ "epoch": 0.79,
2160
+ "learning_rate": 3.986038188889751e-06,
2161
+ "loss": 1.2057,
2162
+ "step": 3590
2163
+ },
2164
+ {
2165
+ "epoch": 0.79,
2166
+ "learning_rate": 3.908505645725963e-06,
2167
+ "loss": 1.2098,
2168
+ "step": 3600
2169
+ },
2170
+ {
2171
+ "epoch": 0.79,
2172
+ "learning_rate": 3.831621461362134e-06,
2173
+ "loss": 1.1819,
2174
+ "step": 3610
2175
+ },
2176
+ {
2177
+ "epoch": 0.79,
2178
+ "learning_rate": 3.7553901301013687e-06,
2179
+ "loss": 1.1491,
2180
+ "step": 3620
2181
+ },
2182
+ {
2183
+ "epoch": 0.8,
2184
+ "learning_rate": 3.6798161080839265e-06,
2185
+ "loss": 1.187,
2186
+ "step": 3630
2187
+ },
2188
+ {
2189
+ "epoch": 0.8,
2190
+ "learning_rate": 3.604903813026713e-06,
2191
+ "loss": 1.1732,
2192
+ "step": 3640
2193
+ },
2194
+ {
2195
+ "epoch": 0.8,
2196
+ "learning_rate": 3.530657623965068e-06,
2197
+ "loss": 1.1956,
2198
+ "step": 3650
2199
+ },
2200
+ {
2201
+ "epoch": 0.8,
2202
+ "learning_rate": 3.4570818809967768e-06,
2203
+ "loss": 1.1907,
2204
+ "step": 3660
2205
+ },
2206
+ {
2207
+ "epoch": 0.8,
2208
+ "learning_rate": 3.3841808850283597e-06,
2209
+ "loss": 1.2285,
2210
+ "step": 3670
2211
+ },
2212
+ {
2213
+ "epoch": 0.81,
2214
+ "learning_rate": 3.311958897523671e-06,
2215
+ "loss": 1.2041,
2216
+ "step": 3680
2217
+ },
2218
+ {
2219
+ "epoch": 0.81,
2220
+ "learning_rate": 3.2404201402547893e-06,
2221
+ "loss": 1.2125,
2222
+ "step": 3690
2223
+ },
2224
+ {
2225
+ "epoch": 0.81,
2226
+ "learning_rate": 3.1695687950552216e-06,
2227
+ "loss": 1.1567,
2228
+ "step": 3700
2229
+ },
2230
+ {
2231
+ "epoch": 0.81,
2232
+ "learning_rate": 3.099409003575473e-06,
2233
+ "loss": 1.1988,
2234
+ "step": 3710
2235
+ },
2236
+ {
2237
+ "epoch": 0.81,
2238
+ "learning_rate": 3.029944867040928e-06,
2239
+ "loss": 1.1917,
2240
+ "step": 3720
2241
+ },
2242
+ {
2243
+ "epoch": 0.82,
2244
+ "learning_rate": 2.9611804460121195e-06,
2245
+ "loss": 1.1768,
2246
+ "step": 3730
2247
+ },
2248
+ {
2249
+ "epoch": 0.82,
2250
+ "learning_rate": 2.8931197601473634e-06,
2251
+ "loss": 1.1814,
2252
+ "step": 3740
2253
+ },
2254
+ {
2255
+ "epoch": 0.82,
2256
+ "learning_rate": 2.8257667879677762e-06,
2257
+ "loss": 1.2024,
2258
+ "step": 3750
2259
+ },
2260
+ {
2261
+ "epoch": 0.82,
2262
+ "learning_rate": 2.7591254666247455e-06,
2263
+ "loss": 1.2349,
2264
+ "step": 3760
2265
+ },
2266
+ {
2267
+ "epoch": 0.83,
2268
+ "learning_rate": 2.6931996916697284e-06,
2269
+ "loss": 1.2297,
2270
+ "step": 3770
2271
+ },
2272
+ {
2273
+ "epoch": 0.83,
2274
+ "learning_rate": 2.627993316826584e-06,
2275
+ "loss": 1.2128,
2276
+ "step": 3780
2277
+ },
2278
+ {
2279
+ "epoch": 0.83,
2280
+ "learning_rate": 2.563510153766262e-06,
2281
+ "loss": 1.1999,
2282
+ "step": 3790
2283
+ },
2284
+ {
2285
+ "epoch": 0.83,
2286
+ "learning_rate": 2.499753971884036e-06,
2287
+ "loss": 1.1829,
2288
+ "step": 3800
2289
+ },
2290
+ {
2291
+ "epoch": 0.83,
2292
+ "learning_rate": 2.4367284980791046e-06,
2293
+ "loss": 1.166,
2294
+ "step": 3810
2295
+ },
2296
+ {
2297
+ "epoch": 0.84,
2298
+ "learning_rate": 2.3744374165367833e-06,
2299
+ "loss": 1.2094,
2300
+ "step": 3820
2301
+ },
2302
+ {
2303
+ "epoch": 0.84,
2304
+ "learning_rate": 2.3128843685131194e-06,
2305
+ "loss": 1.1759,
2306
+ "step": 3830
2307
+ },
2308
+ {
2309
+ "epoch": 0.84,
2310
+ "learning_rate": 2.2520729521220402e-06,
2311
+ "loss": 1.1667,
2312
+ "step": 3840
2313
+ },
2314
+ {
2315
+ "epoch": 0.84,
2316
+ "learning_rate": 2.1920067221250385e-06,
2317
+ "loss": 1.2021,
2318
+ "step": 3850
2319
+ },
2320
+ {
2321
+ "epoch": 0.85,
2322
+ "learning_rate": 2.132689189723349e-06,
2323
+ "loss": 1.2265,
2324
+ "step": 3860
2325
+ },
2326
+ {
2327
+ "epoch": 0.85,
2328
+ "learning_rate": 2.0741238223527426e-06,
2329
+ "loss": 1.1781,
2330
+ "step": 3870
2331
+ },
2332
+ {
2333
+ "epoch": 0.85,
2334
+ "learning_rate": 2.0163140434807864e-06,
2335
+ "loss": 1.1841,
2336
+ "step": 3880
2337
+ },
2338
+ {
2339
+ "epoch": 0.85,
2340
+ "learning_rate": 1.959263232406766e-06,
2341
+ "loss": 1.1788,
2342
+ "step": 3890
2343
+ },
2344
+ {
2345
+ "epoch": 0.85,
2346
+ "learning_rate": 1.9029747240641077e-06,
2347
+ "loss": 1.191,
2348
+ "step": 3900
2349
+ },
2350
+ {
2351
+ "epoch": 0.86,
2352
+ "learning_rate": 1.847451808825479e-06,
2353
+ "loss": 1.1811,
2354
+ "step": 3910
2355
+ },
2356
+ {
2357
+ "epoch": 0.86,
2358
+ "learning_rate": 1.792697732310401e-06,
2359
+ "loss": 1.2047,
2360
+ "step": 3920
2361
+ },
2362
+ {
2363
+ "epoch": 0.86,
2364
+ "learning_rate": 1.7387156951955596e-06,
2365
+ "loss": 1.2186,
2366
+ "step": 3930
2367
+ },
2368
+ {
2369
+ "epoch": 0.86,
2370
+ "learning_rate": 1.6855088530276902e-06,
2371
+ "loss": 1.1985,
2372
+ "step": 3940
2373
+ },
2374
+ {
2375
+ "epoch": 0.87,
2376
+ "learning_rate": 1.6330803160391228e-06,
2377
+ "loss": 1.201,
2378
+ "step": 3950
2379
+ },
2380
+ {
2381
+ "epoch": 0.87,
2382
+ "learning_rate": 1.5814331489659756e-06,
2383
+ "loss": 1.2301,
2384
+ "step": 3960
2385
+ },
2386
+ {
2387
+ "epoch": 0.87,
2388
+ "learning_rate": 1.53057037086899e-06,
2389
+ "loss": 1.2073,
2390
+ "step": 3970
2391
+ },
2392
+ {
2393
+ "epoch": 0.87,
2394
+ "learning_rate": 1.4804949549570818e-06,
2395
+ "loss": 1.2449,
2396
+ "step": 3980
2397
+ },
2398
+ {
2399
+ "epoch": 0.87,
2400
+ "learning_rate": 1.4312098284135e-06,
2401
+ "loss": 1.2077,
2402
+ "step": 3990
2403
+ },
2404
+ {
2405
+ "epoch": 0.88,
2406
+ "learning_rate": 1.3827178722247508e-06,
2407
+ "loss": 1.1627,
2408
+ "step": 4000
2409
+ }
2410
+ ],
2411
+ "max_steps": 4566,
2412
+ "num_train_epochs": 1,
2413
+ "total_flos": 0.0,
2414
+ "trial_name": null,
2415
+ "trial_params": null
2416
+ }
contextual/pct50/finetuning/pixel/MSA_Dialect_Combined_1x/training_complete.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_path": "/root/pixel_cloud/contextual_100_work/data/contextual/downstream/MSA_Dialect_Combined_1x_max64",
3
+ "dataset_rows": 584406,
4
+ "epochs": 1,
5
+ "global_step": 4566,
6
+ "input_checkpoint": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct50/pretraining/pixel/full_epoch2/final/pytorch_model.bin",
7
+ "learning_rate": 3e-05,
8
+ "metrics": {
9
+ "epoch": 1.0,
10
+ "train_loss": 1.2548768560039438,
11
+ "train_runtime": 1217.8261,
12
+ "train_samples_per_second": 479.876,
13
+ "train_steps_per_second": 3.749
14
+ },
15
+ "model_type": "pixel",
16
+ "scheduler": "cosine",
17
+ "seed": 42,
18
+ "warmup_ratio": 0.1
19
+ }
contextual/pct50/manifest.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backing_percentage": 50,
3
+ "checked_train_indices": [
4
+ 0,
5
+ 1,
6
+ 1717288,
7
+ 6010510,
8
+ 12021018,
9
+ 12021019
10
+ ],
11
+ "inspected_word_images": 59,
12
+ "raw_dataset": "/workspace/datasets/pixel_translation_MSA_filtered_32",
13
+ "raw_train_rows": 24042041,
14
+ "rendered_backing_rows": 12021020,
15
+ "rendered_dataset": "/root/pixel_cloud/contextual_100_work/data/contextual/pct50/msa",
16
+ "renderer": {
17
+ "arabic_shaping": "Pango shapes the pair with join-causing boundary sentinels; every observed pair/join-state combination maps to one cached patch",
18
+ "bigram_overlap": 1,
19
+ "canonical_key": "ordered grapheme pair plus joins_previous and joins_following",
20
+ "direction_policy": "bigram patches follow physical canvas x-order; RTL tokens reverse logical bigram order while Pango controls order inside each patch",
21
+ "dpi": 120,
22
+ "font_file": "/root/pixel_cloud/code/pixel_translation/PixelSmolLM/rendering/fonts/GoNotoCurrent.ttf",
23
+ "font_size": 10,
24
+ "image_height": 24,
25
+ "image_width": 384,
26
+ "patch_extraction": "render each bigram in a deterministic synthetic Arabic joining context selected by its two external connection states",
27
+ "patch_size": 24,
28
+ "patches_per_word": 16,
29
+ "shaping_mode": "canonical",
30
+ "single_grapheme_policy": "one unigram patch",
31
+ "strategy": "overlapping_grapheme_bigram_patches",
32
+ "unicode_unit": "extended grapheme cluster (regex \\X)",
33
+ "version": 1,
34
+ "word_segmentation": "same punctuation-aware segmentation as PixelSmolLM"
35
+ },
36
+ "requested_percentage": 50,
37
+ "requested_train_rows": 12021020,
38
+ "schema_version": 1,
39
+ "selection": "prefix_of_already_shuffled_filtered_MSA",
40
+ "valid": true
41
+ }
contextual/pct50/msa_training.complete ADDED
File without changes
contextual/pct50/pretraining/pixel/alignment/checkpoint-20000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct50/pretraining/pixel/alignment/checkpoint-22000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct50/pretraining/pixel/alignment/final/training_complete.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epochs": 1,
3
+ "global_step": 23479,
4
+ "initial_weights": null,
5
+ "learning_rate": 0.0003,
6
+ "metrics": {
7
+ "epoch": 1.0,
8
+ "train_loss": 2.845592947772691,
9
+ "train_runtime": 12834.8112,
10
+ "train_samples_per_second": 936.595,
11
+ "train_steps_per_second": 1.829
12
+ },
13
+ "model_type": "pixel",
14
+ "percentage": 50,
15
+ "scheduler": "cosine",
16
+ "seed": 42,
17
+ "stage": "alignment",
18
+ "subset_rows": 12021020,
19
+ "total_rows": 24042041,
20
+ "warmup_ratio": 0.1
21
+ }
contextual/pct50/pretraining/pixel/full_epoch1/checkpoint-20000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct50/pretraining/pixel/full_epoch1/checkpoint-22000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct50/pretraining/pixel/full_epoch1/final/training_complete.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epochs": 1,
3
+ "global_step": 23479,
4
+ "initial_weights": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct50/pretraining/pixel/alignment/final/pytorch_model.bin",
5
+ "learning_rate": 0.0003,
6
+ "metrics": {
7
+ "epoch": 1.0,
8
+ "train_loss": 1.0701919350582858,
9
+ "train_runtime": 15323.287,
10
+ "train_samples_per_second": 784.494,
11
+ "train_steps_per_second": 1.532
12
+ },
13
+ "model_type": "pixel",
14
+ "percentage": 50,
15
+ "scheduler": "cosine",
16
+ "seed": 42,
17
+ "stage": "full",
18
+ "subset_rows": 12021020,
19
+ "total_rows": 24042041,
20
+ "warmup_ratio": 0.1
21
+ }
contextual/pct50/pretraining/pixel/full_epoch2/checkpoint-20000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct50/pretraining/pixel/full_epoch2/checkpoint-22000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
contextual/pct50/pretraining/pixel/full_epoch2/final/training_complete.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epochs": 1,
3
+ "global_step": 23479,
4
+ "initial_weights": "/root/pixel_cloud/contextual_100_work/checkpoints/contextual/pct50/pretraining/pixel/full_epoch1/final/pytorch_model.bin",
5
+ "learning_rate": 0.0003,
6
+ "metrics": {
7
+ "epoch": 1.0,
8
+ "train_loss": 0.8899728605727828,
9
+ "train_runtime": 15329.4025,
10
+ "train_samples_per_second": 784.181,
11
+ "train_steps_per_second": 1.532
12
+ },
13
+ "model_type": "pixel",
14
+ "percentage": 50,
15
+ "scheduler": "cosine",
16
+ "seed": 42,
17
+ "stage": "full",
18
+ "subset_rows": 12021020,
19
+ "total_rows": 24042041,
20
+ "warmup_ratio": 0.1
21
+ }
contextual/pct50/renderer_audit.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "blank_patch_fraction": 0.0,
3
+ "blank_patches": 0,
4
+ "dataset": "/workspace/datasets/pixel_translation_MSA_filtered_32",
5
+ "ink_touches_horizontal_edge_fraction": 0.04563407977903112,
6
+ "ink_touches_horizontal_edge_patches": 186395,
7
+ "logical_extent_over_24px_fraction": 0.021954645265080154,
8
+ "logical_extent_over_24px_patches": 89675,
9
+ "rendered_patches": 4084557,
10
+ "renderer": {
11
+ "arabic_shaping": "Pango shapes the pair with join-causing boundary sentinels; every observed pair/join-state combination maps to one cached patch",
12
+ "bigram_overlap": 1,
13
+ "canonical_key": "ordered grapheme pair plus joins_previous and joins_following",
14
+ "direction_policy": "bigram patches follow physical canvas x-order; RTL tokens reverse logical bigram order while Pango controls order inside each patch",
15
+ "dpi": 120,
16
+ "font_file": "/root/pixel_cloud/code/pixel_translation/PixelSmolLM/rendering/fonts/GoNotoCurrent.ttf",
17
+ "font_size": 10,
18
+ "image_height": 24,
19
+ "image_width": 384,
20
+ "patch_extraction": "render each bigram in a deterministic synthetic Arabic joining context selected by its two external connection states",
21
+ "patch_size": 24,
22
+ "patches_per_word": 16,
23
+ "shaping_mode": "canonical",
24
+ "single_grapheme_policy": "one unigram patch",
25
+ "strategy": "overlapping_grapheme_bigram_patches",
26
+ "unicode_unit": "extended grapheme cluster (regex \\X)",
27
+ "version": 1,
28
+ "word_segmentation": "same punctuation-aware segmentation as PixelSmolLM"
29
+ },
30
+ "sample_rows": 100000,
31
+ "sample_selection": "train rows [0, 100000)",
32
+ "sample_words_after_32_word_cap": 1228012,
33
+ "unique_rendered_units": 4421,
34
+ "unknown_glyph_fraction": 0.0,
35
+ "unknown_glyph_patches": 0,
36
+ "words_over_16_bigrams": 244,
37
+ "words_over_16_bigrams_fraction": 0.00019869512675771897
38
+ }