AleBurzio commited on
Commit
bb2f015
1 Parent(s): 314af00

End of training

Browse files
Files changed (3) hide show
  1. all_results.json +8 -0
  2. train_results.json +8 -0
  3. trainer_state.json +1177 -0
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "train_loss": 0.8035268472507596,
4
+ "train_runtime": 399.0968,
5
+ "train_samples": 383,
6
+ "train_samples_per_second": 4.798,
7
+ "train_steps_per_second": 2.405
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "train_loss": 0.8035268472507596,
4
+ "train_runtime": 399.0968,
5
+ "train_samples": 383,
6
+ "train_samples_per_second": 4.798,
7
+ "train_steps_per_second": 2.405
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1177 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "global_step": 960,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 4.999665344773905e-05,
13
+ "loss": 2.2777,
14
+ "step": 5
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 4.9986614686909146e-05,
19
+ "loss": 2.1303,
20
+ "step": 10
21
+ },
22
+ {
23
+ "epoch": 0.08,
24
+ "learning_rate": 4.996988640512931e-05,
25
+ "loss": 2.143,
26
+ "step": 15
27
+ },
28
+ {
29
+ "epoch": 0.1,
30
+ "learning_rate": 4.994647308096509e-05,
31
+ "loss": 2.2047,
32
+ "step": 20
33
+ },
34
+ {
35
+ "epoch": 0.13,
36
+ "learning_rate": 4.991638098272951e-05,
37
+ "loss": 1.9869,
38
+ "step": 25
39
+ },
40
+ {
41
+ "epoch": 0.16,
42
+ "learning_rate": 4.987961816680492e-05,
43
+ "loss": 2.1792,
44
+ "step": 30
45
+ },
46
+ {
47
+ "epoch": 0.18,
48
+ "learning_rate": 4.983619447548611e-05,
49
+ "loss": 2.0121,
50
+ "step": 35
51
+ },
52
+ {
53
+ "epoch": 0.21,
54
+ "learning_rate": 4.9786121534345265e-05,
55
+ "loss": 1.9919,
56
+ "step": 40
57
+ },
58
+ {
59
+ "epoch": 0.23,
60
+ "learning_rate": 4.972941274911953e-05,
61
+ "loss": 1.9453,
62
+ "step": 45
63
+ },
64
+ {
65
+ "epoch": 0.26,
66
+ "learning_rate": 4.966608330212198e-05,
67
+ "loss": 2.0629,
68
+ "step": 50
69
+ },
70
+ {
71
+ "epoch": 0.29,
72
+ "learning_rate": 4.9596150148176936e-05,
73
+ "loss": 1.8407,
74
+ "step": 55
75
+ },
76
+ {
77
+ "epoch": 0.31,
78
+ "learning_rate": 4.951963201008076e-05,
79
+ "loss": 1.9516,
80
+ "step": 60
81
+ },
82
+ {
83
+ "epoch": 0.34,
84
+ "learning_rate": 4.94365493735893e-05,
85
+ "loss": 1.8641,
86
+ "step": 65
87
+ },
88
+ {
89
+ "epoch": 0.36,
90
+ "learning_rate": 4.934692448193334e-05,
91
+ "loss": 1.8404,
92
+ "step": 70
93
+ },
94
+ {
95
+ "epoch": 0.39,
96
+ "learning_rate": 4.9250781329863606e-05,
97
+ "loss": 1.9995,
98
+ "step": 75
99
+ },
100
+ {
101
+ "epoch": 0.42,
102
+ "learning_rate": 4.914814565722671e-05,
103
+ "loss": 1.822,
104
+ "step": 80
105
+ },
106
+ {
107
+ "epoch": 0.44,
108
+ "learning_rate": 4.903904494207405e-05,
109
+ "loss": 1.9633,
110
+ "step": 85
111
+ },
112
+ {
113
+ "epoch": 0.47,
114
+ "learning_rate": 4.892350839330522e-05,
115
+ "loss": 1.8818,
116
+ "step": 90
117
+ },
118
+ {
119
+ "epoch": 0.49,
120
+ "learning_rate": 4.880156694284811e-05,
121
+ "loss": 1.7518,
122
+ "step": 95
123
+ },
124
+ {
125
+ "epoch": 0.52,
126
+ "learning_rate": 4.867325323737765e-05,
127
+ "loss": 1.8622,
128
+ "step": 100
129
+ },
130
+ {
131
+ "epoch": 0.55,
132
+ "learning_rate": 4.853860162957552e-05,
133
+ "loss": 1.8353,
134
+ "step": 105
135
+ },
136
+ {
137
+ "epoch": 0.57,
138
+ "learning_rate": 4.839764816893315e-05,
139
+ "loss": 1.9588,
140
+ "step": 110
141
+ },
142
+ {
143
+ "epoch": 0.6,
144
+ "learning_rate": 4.8250430592100306e-05,
145
+ "loss": 1.9896,
146
+ "step": 115
147
+ },
148
+ {
149
+ "epoch": 0.62,
150
+ "learning_rate": 4.8096988312782174e-05,
151
+ "loss": 1.8516,
152
+ "step": 120
153
+ },
154
+ {
155
+ "epoch": 0.65,
156
+ "learning_rate": 4.793736241118728e-05,
157
+ "loss": 1.7615,
158
+ "step": 125
159
+ },
160
+ {
161
+ "epoch": 0.68,
162
+ "learning_rate": 4.7771595623029394e-05,
163
+ "loss": 1.7356,
164
+ "step": 130
165
+ },
166
+ {
167
+ "epoch": 0.7,
168
+ "learning_rate": 4.759973232808609e-05,
169
+ "loss": 1.9084,
170
+ "step": 135
171
+ },
172
+ {
173
+ "epoch": 0.73,
174
+ "learning_rate": 4.742181853831721e-05,
175
+ "loss": 1.8052,
176
+ "step": 140
177
+ },
178
+ {
179
+ "epoch": 0.76,
180
+ "learning_rate": 4.7237901885546404e-05,
181
+ "loss": 1.9753,
182
+ "step": 145
183
+ },
184
+ {
185
+ "epoch": 0.78,
186
+ "learning_rate": 4.7048031608708876e-05,
187
+ "loss": 1.8283,
188
+ "step": 150
189
+ },
190
+ {
191
+ "epoch": 0.81,
192
+ "learning_rate": 4.6852258540668974e-05,
193
+ "loss": 1.8031,
194
+ "step": 155
195
+ },
196
+ {
197
+ "epoch": 0.83,
198
+ "learning_rate": 4.665063509461097e-05,
199
+ "loss": 1.842,
200
+ "step": 160
201
+ },
202
+ {
203
+ "epoch": 0.86,
204
+ "learning_rate": 4.6443215250006806e-05,
205
+ "loss": 1.8729,
206
+ "step": 165
207
+ },
208
+ {
209
+ "epoch": 0.89,
210
+ "learning_rate": 4.6230054538164474e-05,
211
+ "loss": 1.824,
212
+ "step": 170
213
+ },
214
+ {
215
+ "epoch": 0.91,
216
+ "learning_rate": 4.601121002736095e-05,
217
+ "loss": 1.5677,
218
+ "step": 175
219
+ },
220
+ {
221
+ "epoch": 0.94,
222
+ "learning_rate": 4.5786740307563636e-05,
223
+ "loss": 1.8157,
224
+ "step": 180
225
+ },
226
+ {
227
+ "epoch": 0.96,
228
+ "learning_rate": 4.555670547474438e-05,
229
+ "loss": 1.7408,
230
+ "step": 185
231
+ },
232
+ {
233
+ "epoch": 0.99,
234
+ "learning_rate": 4.5321167114790385e-05,
235
+ "loss": 1.8033,
236
+ "step": 190
237
+ },
238
+ {
239
+ "epoch": 1.02,
240
+ "learning_rate": 4.508018828701612e-05,
241
+ "loss": 1.3534,
242
+ "step": 195
243
+ },
244
+ {
245
+ "epoch": 1.04,
246
+ "learning_rate": 4.4833833507280884e-05,
247
+ "loss": 1.167,
248
+ "step": 200
249
+ },
250
+ {
251
+ "epoch": 1.07,
252
+ "learning_rate": 4.458216873071626e-05,
253
+ "loss": 1.0725,
254
+ "step": 205
255
+ },
256
+ {
257
+ "epoch": 1.09,
258
+ "learning_rate": 4.4325261334068426e-05,
259
+ "loss": 1.0958,
260
+ "step": 210
261
+ },
262
+ {
263
+ "epoch": 1.12,
264
+ "learning_rate": 4.406318009765971e-05,
265
+ "loss": 1.223,
266
+ "step": 215
267
+ },
268
+ {
269
+ "epoch": 1.15,
270
+ "learning_rate": 4.379599518697444e-05,
271
+ "loss": 1.1112,
272
+ "step": 220
273
+ },
274
+ {
275
+ "epoch": 1.17,
276
+ "learning_rate": 4.352377813387398e-05,
277
+ "loss": 1.0701,
278
+ "step": 225
279
+ },
280
+ {
281
+ "epoch": 1.2,
282
+ "learning_rate": 4.324660181744589e-05,
283
+ "loss": 1.1594,
284
+ "step": 230
285
+ },
286
+ {
287
+ "epoch": 1.22,
288
+ "learning_rate": 4.296454044449245e-05,
289
+ "loss": 1.1602,
290
+ "step": 235
291
+ },
292
+ {
293
+ "epoch": 1.25,
294
+ "learning_rate": 4.267766952966369e-05,
295
+ "loss": 1.1109,
296
+ "step": 240
297
+ },
298
+ {
299
+ "epoch": 1.28,
300
+ "learning_rate": 4.238606587524029e-05,
301
+ "loss": 1.0719,
302
+ "step": 245
303
+ },
304
+ {
305
+ "epoch": 1.3,
306
+ "learning_rate": 4.208980755057178e-05,
307
+ "loss": 1.1497,
308
+ "step": 250
309
+ },
310
+ {
311
+ "epoch": 1.33,
312
+ "learning_rate": 4.178897387117546e-05,
313
+ "loss": 1.2594,
314
+ "step": 255
315
+ },
316
+ {
317
+ "epoch": 1.35,
318
+ "learning_rate": 4.148364537750172e-05,
319
+ "loss": 1.1576,
320
+ "step": 260
321
+ },
322
+ {
323
+ "epoch": 1.38,
324
+ "learning_rate": 4.117390381337144e-05,
325
+ "loss": 1.1553,
326
+ "step": 265
327
+ },
328
+ {
329
+ "epoch": 1.41,
330
+ "learning_rate": 4.085983210409114e-05,
331
+ "loss": 1.1414,
332
+ "step": 270
333
+ },
334
+ {
335
+ "epoch": 1.43,
336
+ "learning_rate": 4.054151433425194e-05,
337
+ "loss": 1.241,
338
+ "step": 275
339
+ },
340
+ {
341
+ "epoch": 1.46,
342
+ "learning_rate": 4.021903572521802e-05,
343
+ "loss": 1.1757,
344
+ "step": 280
345
+ },
346
+ {
347
+ "epoch": 1.48,
348
+ "learning_rate": 3.9892482612310836e-05,
349
+ "loss": 1.1875,
350
+ "step": 285
351
+ },
352
+ {
353
+ "epoch": 1.51,
354
+ "learning_rate": 3.956194242169506e-05,
355
+ "loss": 1.0741,
356
+ "step": 290
357
+ },
358
+ {
359
+ "epoch": 1.54,
360
+ "learning_rate": 3.922750364697246e-05,
361
+ "loss": 1.1855,
362
+ "step": 295
363
+ },
364
+ {
365
+ "epoch": 1.56,
366
+ "learning_rate": 3.888925582549006e-05,
367
+ "loss": 1.1681,
368
+ "step": 300
369
+ },
370
+ {
371
+ "epoch": 1.59,
372
+ "learning_rate": 3.854728951436879e-05,
373
+ "loss": 1.1153,
374
+ "step": 305
375
+ },
376
+ {
377
+ "epoch": 1.61,
378
+ "learning_rate": 3.82016962662592e-05,
379
+ "loss": 1.2223,
380
+ "step": 310
381
+ },
382
+ {
383
+ "epoch": 1.64,
384
+ "learning_rate": 3.785256860483054e-05,
385
+ "loss": 1.1862,
386
+ "step": 315
387
+ },
388
+ {
389
+ "epoch": 1.67,
390
+ "learning_rate": 3.7500000000000003e-05,
391
+ "loss": 1.2807,
392
+ "step": 320
393
+ },
394
+ {
395
+ "epoch": 1.69,
396
+ "learning_rate": 3.7144084842908505e-05,
397
+ "loss": 1.1803,
398
+ "step": 325
399
+ },
400
+ {
401
+ "epoch": 1.72,
402
+ "learning_rate": 3.678491842064995e-05,
403
+ "loss": 1.2763,
404
+ "step": 330
405
+ },
406
+ {
407
+ "epoch": 1.74,
408
+ "learning_rate": 3.6422596890760517e-05,
409
+ "loss": 1.2246,
410
+ "step": 335
411
+ },
412
+ {
413
+ "epoch": 1.77,
414
+ "learning_rate": 3.6057217255475034e-05,
415
+ "loss": 1.0887,
416
+ "step": 340
417
+ },
418
+ {
419
+ "epoch": 1.8,
420
+ "learning_rate": 3.568887733575706e-05,
421
+ "loss": 1.1806,
422
+ "step": 345
423
+ },
424
+ {
425
+ "epoch": 1.82,
426
+ "learning_rate": 3.5317675745109866e-05,
427
+ "loss": 1.0842,
428
+ "step": 350
429
+ },
430
+ {
431
+ "epoch": 1.85,
432
+ "learning_rate": 3.494371186317528e-05,
433
+ "loss": 1.0559,
434
+ "step": 355
435
+ },
436
+ {
437
+ "epoch": 1.88,
438
+ "learning_rate": 3.456708580912725e-05,
439
+ "loss": 1.0297,
440
+ "step": 360
441
+ },
442
+ {
443
+ "epoch": 1.9,
444
+ "learning_rate": 3.418789841486759e-05,
445
+ "loss": 1.0726,
446
+ "step": 365
447
+ },
448
+ {
449
+ "epoch": 1.93,
450
+ "learning_rate": 3.380625119803084e-05,
451
+ "loss": 1.1097,
452
+ "step": 370
453
+ },
454
+ {
455
+ "epoch": 1.95,
456
+ "learning_rate": 3.34222463348055e-05,
457
+ "loss": 1.0985,
458
+ "step": 375
459
+ },
460
+ {
461
+ "epoch": 1.98,
462
+ "learning_rate": 3.303598663257904e-05,
463
+ "loss": 1.182,
464
+ "step": 380
465
+ },
466
+ {
467
+ "epoch": 2.01,
468
+ "learning_rate": 3.264757550241384e-05,
469
+ "loss": 1.0829,
470
+ "step": 385
471
+ },
472
+ {
473
+ "epoch": 2.03,
474
+ "learning_rate": 3.225711693136156e-05,
475
+ "loss": 0.6117,
476
+ "step": 390
477
+ },
478
+ {
479
+ "epoch": 2.06,
480
+ "learning_rate": 3.186471545462331e-05,
481
+ "loss": 0.53,
482
+ "step": 395
483
+ },
484
+ {
485
+ "epoch": 2.08,
486
+ "learning_rate": 3.147047612756302e-05,
487
+ "loss": 0.6219,
488
+ "step": 400
489
+ },
490
+ {
491
+ "epoch": 2.11,
492
+ "learning_rate": 3.10745044975816e-05,
493
+ "loss": 0.528,
494
+ "step": 405
495
+ },
496
+ {
497
+ "epoch": 2.14,
498
+ "learning_rate": 3.0676906575859334e-05,
499
+ "loss": 0.5044,
500
+ "step": 410
501
+ },
502
+ {
503
+ "epoch": 2.16,
504
+ "learning_rate": 3.0277788808974128e-05,
505
+ "loss": 0.6492,
506
+ "step": 415
507
+ },
508
+ {
509
+ "epoch": 2.19,
510
+ "learning_rate": 2.9877258050403212e-05,
511
+ "loss": 0.5539,
512
+ "step": 420
513
+ },
514
+ {
515
+ "epoch": 2.21,
516
+ "learning_rate": 2.9475421531915827e-05,
517
+ "loss": 0.6158,
518
+ "step": 425
519
+ },
520
+ {
521
+ "epoch": 2.24,
522
+ "learning_rate": 2.9072386834864724e-05,
523
+ "loss": 0.5606,
524
+ "step": 430
525
+ },
526
+ {
527
+ "epoch": 2.27,
528
+ "learning_rate": 2.8668261861384045e-05,
529
+ "loss": 0.4933,
530
+ "step": 435
531
+ },
532
+ {
533
+ "epoch": 2.29,
534
+ "learning_rate": 2.8263154805501297e-05,
535
+ "loss": 0.5449,
536
+ "step": 440
537
+ },
538
+ {
539
+ "epoch": 2.32,
540
+ "learning_rate": 2.7857174124171165e-05,
541
+ "loss": 0.5829,
542
+ "step": 445
543
+ },
544
+ {
545
+ "epoch": 2.34,
546
+ "learning_rate": 2.7450428508239024e-05,
547
+ "loss": 0.5175,
548
+ "step": 450
549
+ },
550
+ {
551
+ "epoch": 2.37,
552
+ "learning_rate": 2.704302685334171e-05,
553
+ "loss": 0.5473,
554
+ "step": 455
555
+ },
556
+ {
557
+ "epoch": 2.4,
558
+ "learning_rate": 2.663507823075358e-05,
559
+ "loss": 0.6025,
560
+ "step": 460
561
+ },
562
+ {
563
+ "epoch": 2.42,
564
+ "learning_rate": 2.6226691858185454e-05,
565
+ "loss": 0.5818,
566
+ "step": 465
567
+ },
568
+ {
569
+ "epoch": 2.45,
570
+ "learning_rate": 2.5817977070544407e-05,
571
+ "loss": 0.5363,
572
+ "step": 470
573
+ },
574
+ {
575
+ "epoch": 2.47,
576
+ "learning_rate": 2.5409043290662173e-05,
577
+ "loss": 0.5645,
578
+ "step": 475
579
+ },
580
+ {
581
+ "epoch": 2.5,
582
+ "learning_rate": 2.5e-05,
583
+ "loss": 0.5195,
584
+ "step": 480
585
+ },
586
+ {
587
+ "epoch": 2.53,
588
+ "learning_rate": 2.459095670933783e-05,
589
+ "loss": 0.5496,
590
+ "step": 485
591
+ },
592
+ {
593
+ "epoch": 2.55,
594
+ "learning_rate": 2.41820229294556e-05,
595
+ "loss": 0.5588,
596
+ "step": 490
597
+ },
598
+ {
599
+ "epoch": 2.58,
600
+ "learning_rate": 2.3773308141814552e-05,
601
+ "loss": 0.5454,
602
+ "step": 495
603
+ },
604
+ {
605
+ "epoch": 2.6,
606
+ "learning_rate": 2.3364921769246423e-05,
607
+ "loss": 0.5319,
608
+ "step": 500
609
+ },
610
+ {
611
+ "epoch": 2.63,
612
+ "learning_rate": 2.2956973146658303e-05,
613
+ "loss": 0.5125,
614
+ "step": 505
615
+ },
616
+ {
617
+ "epoch": 2.66,
618
+ "learning_rate": 2.2549571491760986e-05,
619
+ "loss": 0.5035,
620
+ "step": 510
621
+ },
622
+ {
623
+ "epoch": 2.68,
624
+ "learning_rate": 2.2142825875828838e-05,
625
+ "loss": 0.5413,
626
+ "step": 515
627
+ },
628
+ {
629
+ "epoch": 2.71,
630
+ "learning_rate": 2.173684519449872e-05,
631
+ "loss": 0.5031,
632
+ "step": 520
633
+ },
634
+ {
635
+ "epoch": 2.73,
636
+ "learning_rate": 2.1331738138615958e-05,
637
+ "loss": 0.6249,
638
+ "step": 525
639
+ },
640
+ {
641
+ "epoch": 2.76,
642
+ "learning_rate": 2.0927613165135285e-05,
643
+ "loss": 0.5193,
644
+ "step": 530
645
+ },
646
+ {
647
+ "epoch": 2.79,
648
+ "learning_rate": 2.052457846808419e-05,
649
+ "loss": 0.5967,
650
+ "step": 535
651
+ },
652
+ {
653
+ "epoch": 2.81,
654
+ "learning_rate": 2.0122741949596797e-05,
655
+ "loss": 0.5391,
656
+ "step": 540
657
+ },
658
+ {
659
+ "epoch": 2.84,
660
+ "learning_rate": 1.972221119102587e-05,
661
+ "loss": 0.5601,
662
+ "step": 545
663
+ },
664
+ {
665
+ "epoch": 2.86,
666
+ "learning_rate": 1.932309342414067e-05,
667
+ "loss": 0.4991,
668
+ "step": 550
669
+ },
670
+ {
671
+ "epoch": 2.89,
672
+ "learning_rate": 1.8925495502418406e-05,
673
+ "loss": 0.6078,
674
+ "step": 555
675
+ },
676
+ {
677
+ "epoch": 2.92,
678
+ "learning_rate": 1.852952387243698e-05,
679
+ "loss": 0.5349,
680
+ "step": 560
681
+ },
682
+ {
683
+ "epoch": 2.94,
684
+ "learning_rate": 1.81352845453767e-05,
685
+ "loss": 0.4897,
686
+ "step": 565
687
+ },
688
+ {
689
+ "epoch": 2.97,
690
+ "learning_rate": 1.7742883068638447e-05,
691
+ "loss": 0.5732,
692
+ "step": 570
693
+ },
694
+ {
695
+ "epoch": 2.99,
696
+ "learning_rate": 1.7352424497586163e-05,
697
+ "loss": 0.5019,
698
+ "step": 575
699
+ },
700
+ {
701
+ "epoch": 3.02,
702
+ "learning_rate": 1.6964013367420966e-05,
703
+ "loss": 0.3509,
704
+ "step": 580
705
+ },
706
+ {
707
+ "epoch": 3.05,
708
+ "learning_rate": 1.65777536651945e-05,
709
+ "loss": 0.2495,
710
+ "step": 585
711
+ },
712
+ {
713
+ "epoch": 3.07,
714
+ "learning_rate": 1.6193748801969163e-05,
715
+ "loss": 0.265,
716
+ "step": 590
717
+ },
718
+ {
719
+ "epoch": 3.1,
720
+ "learning_rate": 1.5812101585132417e-05,
721
+ "loss": 0.2412,
722
+ "step": 595
723
+ },
724
+ {
725
+ "epoch": 3.12,
726
+ "learning_rate": 1.5432914190872757e-05,
727
+ "loss": 0.2662,
728
+ "step": 600
729
+ },
730
+ {
731
+ "epoch": 3.15,
732
+ "learning_rate": 1.5056288136824726e-05,
733
+ "loss": 0.241,
734
+ "step": 605
735
+ },
736
+ {
737
+ "epoch": 3.18,
738
+ "learning_rate": 1.4682324254890134e-05,
739
+ "loss": 0.2568,
740
+ "step": 610
741
+ },
742
+ {
743
+ "epoch": 3.2,
744
+ "learning_rate": 1.4311122664242954e-05,
745
+ "loss": 0.2462,
746
+ "step": 615
747
+ },
748
+ {
749
+ "epoch": 3.23,
750
+ "learning_rate": 1.3942782744524973e-05,
751
+ "loss": 0.2399,
752
+ "step": 620
753
+ },
754
+ {
755
+ "epoch": 3.26,
756
+ "learning_rate": 1.3577403109239484e-05,
757
+ "loss": 0.2486,
758
+ "step": 625
759
+ },
760
+ {
761
+ "epoch": 3.28,
762
+ "learning_rate": 1.3215081579350058e-05,
763
+ "loss": 0.2343,
764
+ "step": 630
765
+ },
766
+ {
767
+ "epoch": 3.31,
768
+ "learning_rate": 1.2855915157091497e-05,
769
+ "loss": 0.2531,
770
+ "step": 635
771
+ },
772
+ {
773
+ "epoch": 3.33,
774
+ "learning_rate": 1.2500000000000006e-05,
775
+ "loss": 0.2529,
776
+ "step": 640
777
+ },
778
+ {
779
+ "epoch": 3.36,
780
+ "learning_rate": 1.2147431395169459e-05,
781
+ "loss": 0.2451,
782
+ "step": 645
783
+ },
784
+ {
785
+ "epoch": 3.39,
786
+ "learning_rate": 1.1798303733740802e-05,
787
+ "loss": 0.2397,
788
+ "step": 650
789
+ },
790
+ {
791
+ "epoch": 3.41,
792
+ "learning_rate": 1.1452710485631216e-05,
793
+ "loss": 0.2397,
794
+ "step": 655
795
+ },
796
+ {
797
+ "epoch": 3.44,
798
+ "learning_rate": 1.1110744174509952e-05,
799
+ "loss": 0.2515,
800
+ "step": 660
801
+ },
802
+ {
803
+ "epoch": 3.46,
804
+ "learning_rate": 1.0772496353027538e-05,
805
+ "loss": 0.2551,
806
+ "step": 665
807
+ },
808
+ {
809
+ "epoch": 3.49,
810
+ "learning_rate": 1.043805757830495e-05,
811
+ "loss": 0.293,
812
+ "step": 670
813
+ },
814
+ {
815
+ "epoch": 3.52,
816
+ "learning_rate": 1.0107517387689166e-05,
817
+ "loss": 0.2861,
818
+ "step": 675
819
+ },
820
+ {
821
+ "epoch": 3.54,
822
+ "learning_rate": 9.780964274781984e-06,
823
+ "loss": 0.2599,
824
+ "step": 680
825
+ },
826
+ {
827
+ "epoch": 3.57,
828
+ "learning_rate": 9.45848566574807e-06,
829
+ "loss": 0.2521,
830
+ "step": 685
831
+ },
832
+ {
833
+ "epoch": 3.59,
834
+ "learning_rate": 9.140167895908867e-06,
835
+ "loss": 0.264,
836
+ "step": 690
837
+ },
838
+ {
839
+ "epoch": 3.62,
840
+ "learning_rate": 8.82609618662857e-06,
841
+ "loss": 0.257,
842
+ "step": 695
843
+ },
844
+ {
845
+ "epoch": 3.65,
846
+ "learning_rate": 8.51635462249828e-06,
847
+ "loss": 0.2621,
848
+ "step": 700
849
+ },
850
+ {
851
+ "epoch": 3.67,
852
+ "learning_rate": 8.211026128824539e-06,
853
+ "loss": 0.2529,
854
+ "step": 705
855
+ },
856
+ {
857
+ "epoch": 3.7,
858
+ "learning_rate": 7.910192449428217e-06,
859
+ "loss": 0.25,
860
+ "step": 710
861
+ },
862
+ {
863
+ "epoch": 3.72,
864
+ "learning_rate": 7.613934124759711e-06,
865
+ "loss": 0.2652,
866
+ "step": 715
867
+ },
868
+ {
869
+ "epoch": 3.75,
870
+ "learning_rate": 7.3223304703363135e-06,
871
+ "loss": 0.2448,
872
+ "step": 720
873
+ },
874
+ {
875
+ "epoch": 3.78,
876
+ "learning_rate": 7.035459555507548e-06,
877
+ "loss": 0.238,
878
+ "step": 725
879
+ },
880
+ {
881
+ "epoch": 3.8,
882
+ "learning_rate": 6.753398182554116e-06,
883
+ "loss": 0.2335,
884
+ "step": 730
885
+ },
886
+ {
887
+ "epoch": 3.83,
888
+ "learning_rate": 6.476221866126029e-06,
889
+ "loss": 0.2498,
890
+ "step": 735
891
+ },
892
+ {
893
+ "epoch": 3.85,
894
+ "learning_rate": 6.204004813025568e-06,
895
+ "loss": 0.2657,
896
+ "step": 740
897
+ },
898
+ {
899
+ "epoch": 3.88,
900
+ "learning_rate": 5.936819902340299e-06,
901
+ "loss": 0.2439,
902
+ "step": 745
903
+ },
904
+ {
905
+ "epoch": 3.91,
906
+ "learning_rate": 5.674738665931575e-06,
907
+ "loss": 0.2455,
908
+ "step": 750
909
+ },
910
+ {
911
+ "epoch": 3.93,
912
+ "learning_rate": 5.417831269283741e-06,
913
+ "loss": 0.2569,
914
+ "step": 755
915
+ },
916
+ {
917
+ "epoch": 3.96,
918
+ "learning_rate": 5.166166492719124e-06,
919
+ "loss": 0.2566,
920
+ "step": 760
921
+ },
922
+ {
923
+ "epoch": 3.98,
924
+ "learning_rate": 4.91981171298388e-06,
925
+ "loss": 0.2502,
926
+ "step": 765
927
+ },
928
+ {
929
+ "epoch": 4.01,
930
+ "learning_rate": 4.678832885209622e-06,
931
+ "loss": 0.2081,
932
+ "step": 770
933
+ },
934
+ {
935
+ "epoch": 4.04,
936
+ "learning_rate": 4.4432945252556284e-06,
937
+ "loss": 0.1486,
938
+ "step": 775
939
+ },
940
+ {
941
+ "epoch": 4.06,
942
+ "learning_rate": 4.213259692436367e-06,
943
+ "loss": 0.1647,
944
+ "step": 780
945
+ },
946
+ {
947
+ "epoch": 4.09,
948
+ "learning_rate": 3.988789972639045e-06,
949
+ "loss": 0.152,
950
+ "step": 785
951
+ },
952
+ {
953
+ "epoch": 4.11,
954
+ "learning_rate": 3.769945461835531e-06,
955
+ "loss": 0.1594,
956
+ "step": 790
957
+ },
958
+ {
959
+ "epoch": 4.14,
960
+ "learning_rate": 3.5567847499932e-06,
961
+ "loss": 0.1576,
962
+ "step": 795
963
+ },
964
+ {
965
+ "epoch": 4.17,
966
+ "learning_rate": 3.3493649053890326e-06,
967
+ "loss": 0.1622,
968
+ "step": 800
969
+ },
970
+ {
971
+ "epoch": 4.19,
972
+ "learning_rate": 3.147741459331033e-06,
973
+ "loss": 0.1564,
974
+ "step": 805
975
+ },
976
+ {
977
+ "epoch": 4.22,
978
+ "learning_rate": 2.9519683912911266e-06,
979
+ "loss": 0.1668,
980
+ "step": 810
981
+ },
982
+ {
983
+ "epoch": 4.24,
984
+ "learning_rate": 2.7620981144536013e-06,
985
+ "loss": 0.1479,
986
+ "step": 815
987
+ },
988
+ {
989
+ "epoch": 4.27,
990
+ "learning_rate": 2.578181461682794e-06,
991
+ "loss": 0.1558,
992
+ "step": 820
993
+ },
994
+ {
995
+ "epoch": 4.3,
996
+ "learning_rate": 2.4002676719139166e-06,
997
+ "loss": 0.1483,
998
+ "step": 825
999
+ },
1000
+ {
1001
+ "epoch": 4.32,
1002
+ "learning_rate": 2.2284043769706027e-06,
1003
+ "loss": 0.1556,
1004
+ "step": 830
1005
+ },
1006
+ {
1007
+ "epoch": 4.35,
1008
+ "learning_rate": 2.0626375888127187e-06,
1009
+ "loss": 0.1634,
1010
+ "step": 835
1011
+ },
1012
+ {
1013
+ "epoch": 4.38,
1014
+ "learning_rate": 1.9030116872178316e-06,
1015
+ "loss": 0.1428,
1016
+ "step": 840
1017
+ },
1018
+ {
1019
+ "epoch": 4.4,
1020
+ "learning_rate": 1.7495694078996983e-06,
1021
+ "loss": 0.1491,
1022
+ "step": 845
1023
+ },
1024
+ {
1025
+ "epoch": 4.43,
1026
+ "learning_rate": 1.6023518310668618e-06,
1027
+ "loss": 0.1633,
1028
+ "step": 850
1029
+ },
1030
+ {
1031
+ "epoch": 4.45,
1032
+ "learning_rate": 1.4613983704244826e-06,
1033
+ "loss": 0.1471,
1034
+ "step": 855
1035
+ },
1036
+ {
1037
+ "epoch": 4.48,
1038
+ "learning_rate": 1.3267467626223606e-06,
1039
+ "loss": 0.1481,
1040
+ "step": 860
1041
+ },
1042
+ {
1043
+ "epoch": 4.51,
1044
+ "learning_rate": 1.1984330571518931e-06,
1045
+ "loss": 0.1464,
1046
+ "step": 865
1047
+ },
1048
+ {
1049
+ "epoch": 4.53,
1050
+ "learning_rate": 1.0764916066947794e-06,
1051
+ "loss": 0.1658,
1052
+ "step": 870
1053
+ },
1054
+ {
1055
+ "epoch": 4.56,
1056
+ "learning_rate": 9.609550579259496e-07,
1057
+ "loss": 0.1535,
1058
+ "step": 875
1059
+ },
1060
+ {
1061
+ "epoch": 4.58,
1062
+ "learning_rate": 8.51854342773295e-07,
1063
+ "loss": 0.1548,
1064
+ "step": 880
1065
+ },
1066
+ {
1067
+ "epoch": 4.61,
1068
+ "learning_rate": 7.492186701364007e-07,
1069
+ "loss": 0.1452,
1070
+ "step": 885
1071
+ },
1072
+ {
1073
+ "epoch": 4.64,
1074
+ "learning_rate": 6.530755180666592e-07,
1075
+ "loss": 0.1518,
1076
+ "step": 890
1077
+ },
1078
+ {
1079
+ "epoch": 4.66,
1080
+ "learning_rate": 5.634506264107053e-07,
1081
+ "loss": 0.153,
1082
+ "step": 895
1083
+ },
1084
+ {
1085
+ "epoch": 4.69,
1086
+ "learning_rate": 4.803679899192392e-07,
1087
+ "loss": 0.1568,
1088
+ "step": 900
1089
+ },
1090
+ {
1091
+ "epoch": 4.71,
1092
+ "learning_rate": 4.0384985182306265e-07,
1093
+ "loss": 0.1527,
1094
+ "step": 905
1095
+ },
1096
+ {
1097
+ "epoch": 4.74,
1098
+ "learning_rate": 3.339166978780256e-07,
1099
+ "loss": 0.144,
1100
+ "step": 910
1101
+ },
1102
+ {
1103
+ "epoch": 4.77,
1104
+ "learning_rate": 2.705872508804747e-07,
1105
+ "loss": 0.1509,
1106
+ "step": 915
1107
+ },
1108
+ {
1109
+ "epoch": 4.79,
1110
+ "learning_rate": 2.1387846565474045e-07,
1111
+ "loss": 0.1418,
1112
+ "step": 920
1113
+ },
1114
+ {
1115
+ "epoch": 4.82,
1116
+ "learning_rate": 1.6380552451389086e-07,
1117
+ "loss": 0.168,
1118
+ "step": 925
1119
+ },
1120
+ {
1121
+ "epoch": 4.84,
1122
+ "learning_rate": 1.2038183319507955e-07,
1123
+ "loss": 0.156,
1124
+ "step": 930
1125
+ },
1126
+ {
1127
+ "epoch": 4.87,
1128
+ "learning_rate": 8.361901727049204e-08,
1129
+ "loss": 0.1364,
1130
+ "step": 935
1131
+ },
1132
+ {
1133
+ "epoch": 4.9,
1134
+ "learning_rate": 5.352691903491303e-08,
1135
+ "loss": 0.1482,
1136
+ "step": 940
1137
+ },
1138
+ {
1139
+ "epoch": 4.92,
1140
+ "learning_rate": 3.011359487068987e-08,
1141
+ "loss": 0.1574,
1142
+ "step": 945
1143
+ },
1144
+ {
1145
+ "epoch": 4.95,
1146
+ "learning_rate": 1.3385313090857887e-08,
1147
+ "loss": 0.1496,
1148
+ "step": 950
1149
+ },
1150
+ {
1151
+ "epoch": 4.97,
1152
+ "learning_rate": 3.346552260954927e-09,
1153
+ "loss": 0.1475,
1154
+ "step": 955
1155
+ },
1156
+ {
1157
+ "epoch": 5.0,
1158
+ "learning_rate": 0.0,
1159
+ "loss": 0.1461,
1160
+ "step": 960
1161
+ },
1162
+ {
1163
+ "epoch": 5.0,
1164
+ "step": 960,
1165
+ "total_flos": 1778473878159360.0,
1166
+ "train_loss": 0.8035268472507596,
1167
+ "train_runtime": 399.0968,
1168
+ "train_samples_per_second": 4.798,
1169
+ "train_steps_per_second": 2.405
1170
+ }
1171
+ ],
1172
+ "max_steps": 960,
1173
+ "num_train_epochs": 5,
1174
+ "total_flos": 1778473878159360.0,
1175
+ "trial_name": null,
1176
+ "trial_params": null
1177
+ }