pszemraj commited on
Commit
5f8f910
1 Parent(s): b57852a

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +29 -0
  2. eval_results.json +14 -0
  3. train_results.json +18 -0
  4. trainer_state.json +827 -0
all_results.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "before_init_mem_cpu": 699817984,
3
+ "before_init_mem_gpu": 0,
4
+ "epoch": 5.98,
5
+ "eval_accuracy": 0.544073054441718,
6
+ "eval_loss": 1.9982993602752686,
7
+ "eval_mem_cpu_alloc_delta": 9977856,
8
+ "eval_mem_cpu_peaked_delta": 60305408,
9
+ "eval_mem_gpu_alloc_delta": 0,
10
+ "eval_mem_gpu_peaked_delta": 2357757952,
11
+ "eval_runtime": 481.5484,
12
+ "eval_samples": 588,
13
+ "eval_samples_per_second": 1.221,
14
+ "eval_steps_per_second": 0.305,
15
+ "init_mem_cpu_alloc_delta": -283938816,
16
+ "init_mem_cpu_peaked_delta": 332918784,
17
+ "init_mem_gpu_alloc_delta": 334744576,
18
+ "init_mem_gpu_peaked_delta": 0,
19
+ "perplexity": 7.376500655769286,
20
+ "train_loss": 2.0183091745143984,
21
+ "train_mem_cpu_alloc_delta": 1155149824,
22
+ "train_mem_cpu_peaked_delta": 308719616,
23
+ "train_mem_gpu_alloc_delta": 988995584,
24
+ "train_mem_gpu_peaked_delta": 4500811776,
25
+ "train_runtime": 35067.5673,
26
+ "train_samples": 5325,
27
+ "train_samples_per_second": 0.911,
28
+ "train_steps_per_second": 0.007
29
+ }
eval_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.98,
3
+ "eval_accuracy": 0.544073054441718,
4
+ "eval_loss": 1.9982993602752686,
5
+ "eval_mem_cpu_alloc_delta": 9977856,
6
+ "eval_mem_cpu_peaked_delta": 60305408,
7
+ "eval_mem_gpu_alloc_delta": 0,
8
+ "eval_mem_gpu_peaked_delta": 2357757952,
9
+ "eval_runtime": 481.5484,
10
+ "eval_samples": 588,
11
+ "eval_samples_per_second": 1.221,
12
+ "eval_steps_per_second": 0.305,
13
+ "perplexity": 7.376500655769286
14
+ }
train_results.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "before_init_mem_cpu": 699817984,
3
+ "before_init_mem_gpu": 0,
4
+ "epoch": 5.98,
5
+ "init_mem_cpu_alloc_delta": -283938816,
6
+ "init_mem_cpu_peaked_delta": 332918784,
7
+ "init_mem_gpu_alloc_delta": 334744576,
8
+ "init_mem_gpu_peaked_delta": 0,
9
+ "train_loss": 2.0183091745143984,
10
+ "train_mem_cpu_alloc_delta": 1155149824,
11
+ "train_mem_cpu_peaked_delta": 308719616,
12
+ "train_mem_gpu_alloc_delta": 988995584,
13
+ "train_mem_gpu_peaked_delta": 4500811776,
14
+ "train_runtime": 35067.5673,
15
+ "train_samples": 5325,
16
+ "train_samples_per_second": 0.911,
17
+ "train_steps_per_second": 0.007
18
+ }
trainer_state.json ADDED
@@ -0,0 +1,827 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.984984984984985,
5
+ "global_step": 246,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.05,
12
+ "learning_rate": 0.00015384615384615385,
13
+ "loss": 2.9074,
14
+ "step": 2
15
+ },
16
+ {
17
+ "epoch": 0.1,
18
+ "learning_rate": 0.0003076923076923077,
19
+ "loss": 2.7922,
20
+ "step": 4
21
+ },
22
+ {
23
+ "epoch": 0.14,
24
+ "learning_rate": 0.0004615384615384616,
25
+ "loss": 2.6824,
26
+ "step": 6
27
+ },
28
+ {
29
+ "epoch": 0.19,
30
+ "learning_rate": 0.0006153846153846154,
31
+ "loss": 2.6036,
32
+ "step": 8
33
+ },
34
+ {
35
+ "epoch": 0.24,
36
+ "learning_rate": 0.0007692307692307693,
37
+ "loss": 2.5524,
38
+ "step": 10
39
+ },
40
+ {
41
+ "epoch": 0.29,
42
+ "learning_rate": 0.0009230769230769232,
43
+ "loss": 2.5342,
44
+ "step": 12
45
+ },
46
+ {
47
+ "epoch": 0.34,
48
+ "learning_rate": 0.0009999545513138964,
49
+ "loss": 2.5052,
50
+ "step": 14
51
+ },
52
+ {
53
+ "epoch": 0.38,
54
+ "learning_rate": 0.000999591011397558,
55
+ "loss": 2.4689,
56
+ "step": 16
57
+ },
58
+ {
59
+ "epoch": 0.43,
60
+ "learning_rate": 0.000998864195911451,
61
+ "loss": 2.4487,
62
+ "step": 18
63
+ },
64
+ {
65
+ "epoch": 0.48,
66
+ "learning_rate": 0.000997774633356497,
67
+ "loss": 2.4068,
68
+ "step": 20
69
+ },
70
+ {
71
+ "epoch": 0.53,
72
+ "learning_rate": 0.0009963231160036714,
73
+ "loss": 2.3813,
74
+ "step": 22
75
+ },
76
+ {
77
+ "epoch": 0.58,
78
+ "learning_rate": 0.0009945106993179073,
79
+ "loss": 2.3558,
80
+ "step": 24
81
+ },
82
+ {
83
+ "epoch": 0.62,
84
+ "learning_rate": 0.000992338701190618,
85
+ "loss": 2.3425,
86
+ "step": 26
87
+ },
88
+ {
89
+ "epoch": 0.67,
90
+ "learning_rate": 0.0009898087009813986,
91
+ "loss": 2.3254,
92
+ "step": 28
93
+ },
94
+ {
95
+ "epoch": 0.72,
96
+ "learning_rate": 0.000986922538369599,
97
+ "loss": 2.3128,
98
+ "step": 30
99
+ },
100
+ {
101
+ "epoch": 0.77,
102
+ "learning_rate": 0.0009836823120166116,
103
+ "loss": 2.3026,
104
+ "step": 32
105
+ },
106
+ {
107
+ "epoch": 0.82,
108
+ "learning_rate": 0.0009800903780398356,
109
+ "loss": 2.3005,
110
+ "step": 34
111
+ },
112
+ {
113
+ "epoch": 0.86,
114
+ "learning_rate": 0.0009761493482994372,
115
+ "loss": 2.2828,
116
+ "step": 36
117
+ },
118
+ {
119
+ "epoch": 0.91,
120
+ "learning_rate": 0.0009718620884991454,
121
+ "loss": 2.2779,
122
+ "step": 38
123
+ },
124
+ {
125
+ "epoch": 0.96,
126
+ "learning_rate": 0.0009672317161024677,
127
+ "loss": 2.2485,
128
+ "step": 40
129
+ },
130
+ {
131
+ "epoch": 0.98,
132
+ "eval_accuracy": 0.5157865687819605,
133
+ "eval_loss": 2.1457464694976807,
134
+ "eval_runtime": 496.7153,
135
+ "eval_samples_per_second": 1.184,
136
+ "eval_steps_per_second": 0.296,
137
+ "step": 41
138
+ },
139
+ {
140
+ "epoch": 1.02,
141
+ "learning_rate": 0.000962261598065839,
142
+ "loss": 2.9045,
143
+ "step": 42
144
+ },
145
+ {
146
+ "epoch": 1.07,
147
+ "learning_rate": 0.000956955348390353,
148
+ "loss": 2.16,
149
+ "step": 44
150
+ },
151
+ {
152
+ "epoch": 1.12,
153
+ "learning_rate": 0.000951316825493856,
154
+ "loss": 2.1395,
155
+ "step": 46
156
+ },
157
+ {
158
+ "epoch": 1.17,
159
+ "learning_rate": 0.0009453501294053137,
160
+ "loss": 2.1327,
161
+ "step": 48
162
+ },
163
+ {
164
+ "epoch": 1.22,
165
+ "learning_rate": 0.0009390595987834929,
166
+ "loss": 2.128,
167
+ "step": 50
168
+ },
169
+ {
170
+ "epoch": 1.26,
171
+ "learning_rate": 0.000932449807762122,
172
+ "loss": 2.1251,
173
+ "step": 52
174
+ },
175
+ {
176
+ "epoch": 1.31,
177
+ "learning_rate": 0.0009255255626238293,
178
+ "loss": 2.1138,
179
+ "step": 54
180
+ },
181
+ {
182
+ "epoch": 1.36,
183
+ "learning_rate": 0.0009182918983052741,
184
+ "loss": 2.118,
185
+ "step": 56
186
+ },
187
+ {
188
+ "epoch": 1.41,
189
+ "learning_rate": 0.0009107540747360124,
190
+ "loss": 2.1034,
191
+ "step": 58
192
+ },
193
+ {
194
+ "epoch": 1.46,
195
+ "learning_rate": 0.0009029175730137609,
196
+ "loss": 2.1016,
197
+ "step": 60
198
+ },
199
+ {
200
+ "epoch": 1.5,
201
+ "learning_rate": 0.0008947880914188397,
202
+ "loss": 2.1027,
203
+ "step": 62
204
+ },
205
+ {
206
+ "epoch": 1.55,
207
+ "learning_rate": 0.0008863715412706897,
208
+ "loss": 2.0969,
209
+ "step": 64
210
+ },
211
+ {
212
+ "epoch": 1.6,
213
+ "learning_rate": 0.0008776740426294817,
214
+ "loss": 2.0926,
215
+ "step": 66
216
+ },
217
+ {
218
+ "epoch": 1.65,
219
+ "learning_rate": 0.0008687019198459394,
220
+ "loss": 2.0946,
221
+ "step": 68
222
+ },
223
+ {
224
+ "epoch": 1.7,
225
+ "learning_rate": 0.0008594616969626134,
226
+ "loss": 2.0968,
227
+ "step": 70
228
+ },
229
+ {
230
+ "epoch": 1.74,
231
+ "learning_rate": 0.0008499600929699501,
232
+ "loss": 2.0932,
233
+ "step": 72
234
+ },
235
+ {
236
+ "epoch": 1.79,
237
+ "learning_rate": 0.0008402040169206053,
238
+ "loss": 2.0813,
239
+ "step": 74
240
+ },
241
+ {
242
+ "epoch": 1.84,
243
+ "learning_rate": 0.0008302005629055549,
244
+ "loss": 2.0815,
245
+ "step": 76
246
+ },
247
+ {
248
+ "epoch": 1.89,
249
+ "learning_rate": 0.0008199570048956552,
250
+ "loss": 2.0809,
251
+ "step": 78
252
+ },
253
+ {
254
+ "epoch": 1.94,
255
+ "learning_rate": 0.0008094807914524047,
256
+ "loss": 2.0704,
257
+ "step": 80
258
+ },
259
+ {
260
+ "epoch": 1.98,
261
+ "learning_rate": 0.0007987795403117528,
262
+ "loss": 2.0757,
263
+ "step": 82
264
+ },
265
+ {
266
+ "epoch": 1.98,
267
+ "eval_accuracy": 0.530352903624793,
268
+ "eval_loss": 2.058377742767334,
269
+ "eval_runtime": 493.7321,
270
+ "eval_samples_per_second": 1.191,
271
+ "eval_steps_per_second": 0.298,
272
+ "step": 82
273
+ },
274
+ {
275
+ "epoch": 2.05,
276
+ "learning_rate": 0.0007878610328448947,
277
+ "loss": 2.615,
278
+ "step": 84
279
+ },
280
+ {
281
+ "epoch": 2.1,
282
+ "learning_rate": 0.0007767332084000785,
283
+ "loss": 1.9638,
284
+ "step": 86
285
+ },
286
+ {
287
+ "epoch": 2.14,
288
+ "learning_rate": 0.0007654041585295399,
289
+ "loss": 1.979,
290
+ "step": 88
291
+ },
292
+ {
293
+ "epoch": 2.19,
294
+ "learning_rate": 0.0007538821211057647,
295
+ "loss": 1.9628,
296
+ "step": 90
297
+ },
298
+ {
299
+ "epoch": 2.24,
300
+ "learning_rate": 0.0007421754743313514,
301
+ "loss": 1.9717,
302
+ "step": 92
303
+ },
304
+ {
305
+ "epoch": 2.29,
306
+ "learning_rate": 0.0007302927306468364,
307
+ "loss": 1.9625,
308
+ "step": 94
309
+ },
310
+ {
311
+ "epoch": 2.34,
312
+ "learning_rate": 0.0007182425305409072,
313
+ "loss": 1.9683,
314
+ "step": 96
315
+ },
316
+ {
317
+ "epoch": 2.38,
318
+ "learning_rate": 0.0007060336362675068,
319
+ "loss": 1.9556,
320
+ "step": 98
321
+ },
322
+ {
323
+ "epoch": 2.43,
324
+ "learning_rate": 0.0006936749254743951,
325
+ "loss": 1.9565,
326
+ "step": 100
327
+ },
328
+ {
329
+ "epoch": 2.48,
330
+ "learning_rate": 0.000681175384747805,
331
+ "loss": 1.9601,
332
+ "step": 102
333
+ },
334
+ {
335
+ "epoch": 2.53,
336
+ "learning_rate": 0.0006685441030778817,
337
+ "loss": 1.9612,
338
+ "step": 104
339
+ },
340
+ {
341
+ "epoch": 2.58,
342
+ "learning_rate": 0.0006557902652496611,
343
+ "loss": 1.9539,
344
+ "step": 106
345
+ },
346
+ {
347
+ "epoch": 2.62,
348
+ "learning_rate": 0.0006429231451643907,
349
+ "loss": 1.9563,
350
+ "step": 108
351
+ },
352
+ {
353
+ "epoch": 2.67,
354
+ "learning_rate": 0.0006299520990960496,
355
+ "loss": 1.9597,
356
+ "step": 110
357
+ },
358
+ {
359
+ "epoch": 2.72,
360
+ "learning_rate": 0.000616886558887973,
361
+ "loss": 1.9625,
362
+ "step": 112
363
+ },
364
+ {
365
+ "epoch": 2.77,
366
+ "learning_rate": 0.0006037360250945242,
367
+ "loss": 1.965,
368
+ "step": 114
369
+ },
370
+ {
371
+ "epoch": 2.82,
372
+ "learning_rate": 0.0005905100600728067,
373
+ "loss": 1.9588,
374
+ "step": 116
375
+ },
376
+ {
377
+ "epoch": 2.86,
378
+ "learning_rate": 0.0005772182810294344,
379
+ "loss": 1.965,
380
+ "step": 118
381
+ },
382
+ {
383
+ "epoch": 2.91,
384
+ "learning_rate": 0.0005638703530274187,
385
+ "loss": 1.9595,
386
+ "step": 120
387
+ },
388
+ {
389
+ "epoch": 2.96,
390
+ "learning_rate": 0.0005504759819582581,
391
+ "loss": 1.966,
392
+ "step": 122
393
+ },
394
+ {
395
+ "epoch": 2.98,
396
+ "eval_accuracy": 0.5375845352803879,
397
+ "eval_loss": 2.0210342407226562,
398
+ "eval_runtime": 475.4248,
399
+ "eval_samples_per_second": 1.237,
400
+ "eval_steps_per_second": 0.309,
401
+ "step": 123
402
+ },
403
+ {
404
+ "epoch": 3.02,
405
+ "learning_rate": 0.000537044907484338,
406
+ "loss": 2.5101,
407
+ "step": 124
408
+ },
409
+ {
410
+ "epoch": 3.07,
411
+ "learning_rate": 0.0005235868959567755,
412
+ "loss": 1.8782,
413
+ "step": 126
414
+ },
415
+ {
416
+ "epoch": 3.12,
417
+ "learning_rate": 0.0005101117333138558,
418
+ "loss": 1.8556,
419
+ "step": 128
420
+ },
421
+ {
422
+ "epoch": 3.17,
423
+ "learning_rate": 0.0004966292179652285,
424
+ "loss": 1.8622,
425
+ "step": 130
426
+ },
427
+ {
428
+ "epoch": 3.22,
429
+ "learning_rate": 0.00048314915366703335,
430
+ "loss": 1.8668,
431
+ "step": 132
432
+ },
433
+ {
434
+ "epoch": 3.26,
435
+ "learning_rate": 0.0004696813423931381,
436
+ "loss": 1.873,
437
+ "step": 134
438
+ },
439
+ {
440
+ "epoch": 3.31,
441
+ "learning_rate": 0.000456235577207674,
442
+ "loss": 1.8694,
443
+ "step": 136
444
+ },
445
+ {
446
+ "epoch": 3.36,
447
+ "learning_rate": 0.00044282163514404915,
448
+ "loss": 1.8742,
449
+ "step": 138
450
+ },
451
+ {
452
+ "epoch": 3.41,
453
+ "learning_rate": 0.00042944927009561784,
454
+ "loss": 1.8694,
455
+ "step": 140
456
+ },
457
+ {
458
+ "epoch": 3.46,
459
+ "learning_rate": 0.00041612820572317757,
460
+ "loss": 1.8569,
461
+ "step": 142
462
+ },
463
+ {
464
+ "epoch": 3.5,
465
+ "learning_rate": 0.0004028681283844471,
466
+ "loss": 1.8619,
467
+ "step": 144
468
+ },
469
+ {
470
+ "epoch": 3.55,
471
+ "learning_rate": 0.0003896786800906733,
472
+ "loss": 1.8734,
473
+ "step": 146
474
+ },
475
+ {
476
+ "epoch": 3.6,
477
+ "learning_rate": 0.00037656945149547956,
478
+ "loss": 1.8677,
479
+ "step": 148
480
+ },
481
+ {
482
+ "epoch": 3.65,
483
+ "learning_rate": 0.00036354997492106257,
484
+ "loss": 1.872,
485
+ "step": 150
486
+ },
487
+ {
488
+ "epoch": 3.7,
489
+ "learning_rate": 0.00035062971742680246,
490
+ "loss": 1.847,
491
+ "step": 152
492
+ },
493
+ {
494
+ "epoch": 3.74,
495
+ "learning_rate": 0.00033781807392532895,
496
+ "loss": 1.8733,
497
+ "step": 154
498
+ },
499
+ {
500
+ "epoch": 3.79,
501
+ "learning_rate": 0.00032512436035104966,
502
+ "loss": 1.8657,
503
+ "step": 156
504
+ },
505
+ {
506
+ "epoch": 3.84,
507
+ "learning_rate": 0.00031255780688610506,
508
+ "loss": 1.8581,
509
+ "step": 158
510
+ },
511
+ {
512
+ "epoch": 3.89,
513
+ "learning_rate": 0.00030012755124868095,
514
+ "loss": 1.8672,
515
+ "step": 160
516
+ },
517
+ {
518
+ "epoch": 3.94,
519
+ "learning_rate": 0.00028784263204855175,
520
+ "loss": 1.8585,
521
+ "step": 162
522
+ },
523
+ {
524
+ "epoch": 3.98,
525
+ "learning_rate": 0.00027571198221469395,
526
+ "loss": 1.8602,
527
+ "step": 164
528
+ },
529
+ {
530
+ "epoch": 3.98,
531
+ "eval_accuracy": 0.5422011424315571,
532
+ "eval_loss": 2.0011844635009766,
533
+ "eval_runtime": 493.1263,
534
+ "eval_samples_per_second": 1.192,
535
+ "eval_steps_per_second": 0.298,
536
+ "step": 164
537
+ },
538
+ {
539
+ "epoch": 4.05,
540
+ "learning_rate": 0.0002637444224997421,
541
+ "loss": 2.4006,
542
+ "step": 166
543
+ },
544
+ {
545
+ "epoch": 4.1,
546
+ "learning_rate": 0.00025194865506601504,
547
+ "loss": 1.8118,
548
+ "step": 168
549
+ },
550
+ {
551
+ "epoch": 4.14,
552
+ "learning_rate": 0.00024033325715777377,
553
+ "loss": 1.7888,
554
+ "step": 170
555
+ },
556
+ {
557
+ "epoch": 4.19,
558
+ "learning_rate": 0.00022890667486431295,
559
+ "loss": 1.798,
560
+ "step": 172
561
+ },
562
+ {
563
+ "epoch": 4.24,
564
+ "learning_rate": 0.00021767721697842242,
565
+ "loss": 1.8052,
566
+ "step": 174
567
+ },
568
+ {
569
+ "epoch": 4.29,
570
+ "learning_rate": 0.00020665304895468113,
571
+ "loss": 1.8054,
572
+ "step": 176
573
+ },
574
+ {
575
+ "epoch": 4.34,
576
+ "learning_rate": 0.0001958421869719807,
577
+ "loss": 1.7966,
578
+ "step": 178
579
+ },
580
+ {
581
+ "epoch": 4.38,
582
+ "learning_rate": 0.00018525249210459344,
583
+ "loss": 1.8049,
584
+ "step": 180
585
+ },
586
+ {
587
+ "epoch": 4.43,
588
+ "learning_rate": 0.00017489166460602496,
589
+ "loss": 1.7955,
590
+ "step": 182
591
+ },
592
+ {
593
+ "epoch": 4.48,
594
+ "learning_rate": 0.0001647672383098045,
595
+ "loss": 1.8062,
596
+ "step": 184
597
+ },
598
+ {
599
+ "epoch": 4.53,
600
+ "learning_rate": 0.00015488657515129006,
601
+ "loss": 1.8132,
602
+ "step": 186
603
+ },
604
+ {
605
+ "epoch": 4.58,
606
+ "learning_rate": 0.0001452568598144668,
607
+ "loss": 1.8027,
608
+ "step": 188
609
+ },
610
+ {
611
+ "epoch": 4.62,
612
+ "learning_rate": 0.00013588509450763282,
613
+ "loss": 1.8066,
614
+ "step": 190
615
+ },
616
+ {
617
+ "epoch": 4.67,
618
+ "learning_rate": 0.0001267780938717722,
619
+ "loss": 1.7989,
620
+ "step": 192
621
+ },
622
+ {
623
+ "epoch": 4.72,
624
+ "learning_rate": 0.00011794248002531643,
625
+ "loss": 1.7966,
626
+ "step": 194
627
+ },
628
+ {
629
+ "epoch": 4.77,
630
+ "learning_rate": 0.00010938467774889881,
631
+ "loss": 1.8074,
632
+ "step": 196
633
+ },
634
+ {
635
+ "epoch": 4.82,
636
+ "learning_rate": 0.00010111090981359961,
637
+ "loss": 1.8073,
638
+ "step": 198
639
+ },
640
+ {
641
+ "epoch": 4.86,
642
+ "learning_rate": 9.312719245608486e-05,
643
+ "loss": 1.8022,
644
+ "step": 200
645
+ },
646
+ {
647
+ "epoch": 4.91,
648
+ "learning_rate": 8.54393310039246e-05,
649
+ "loss": 1.8014,
650
+ "step": 202
651
+ },
652
+ {
653
+ "epoch": 4.96,
654
+ "learning_rate": 7.805291565427064e-05,
655
+ "loss": 1.8089,
656
+ "step": 204
657
+ },
658
+ {
659
+ "epoch": 4.98,
660
+ "eval_accuracy": 0.5435859583325021,
661
+ "eval_loss": 1.9977257251739502,
662
+ "eval_runtime": 497.2866,
663
+ "eval_samples_per_second": 1.182,
664
+ "eval_steps_per_second": 0.296,
665
+ "step": 205
666
+ },
667
+ {
668
+ "epoch": 5.02,
669
+ "learning_rate": 7.097331740896996e-05,
670
+ "loss": 2.3557,
671
+ "step": 206
672
+ },
673
+ {
674
+ "epoch": 5.07,
675
+ "learning_rate": 6.420568416906064e-05,
676
+ "loss": 1.7835,
677
+ "step": 208
678
+ },
679
+ {
680
+ "epoch": 5.12,
681
+ "learning_rate": 5.775493699149753e-05,
682
+ "loss": 1.7641,
683
+ "step": 210
684
+ },
685
+ {
686
+ "epoch": 5.17,
687
+ "learning_rate": 5.16257665108254e-05,
688
+ "loss": 1.7754,
689
+ "step": 212
690
+ },
691
+ {
692
+ "epoch": 5.22,
693
+ "learning_rate": 4.582262952840355e-05,
694
+ "loss": 1.7765,
695
+ "step": 214
696
+ },
697
+ {
698
+ "epoch": 5.26,
699
+ "learning_rate": 4.034974577166023e-05,
700
+ "loss": 1.7732,
701
+ "step": 216
702
+ },
703
+ {
704
+ "epoch": 5.31,
705
+ "learning_rate": 3.521109482573515e-05,
706
+ "loss": 1.7663,
707
+ "step": 218
708
+ },
709
+ {
710
+ "epoch": 5.36,
711
+ "learning_rate": 3.0410413239740975e-05,
712
+ "loss": 1.7779,
713
+ "step": 220
714
+ },
715
+ {
716
+ "epoch": 5.41,
717
+ "learning_rate": 2.5951191809746145e-05,
718
+ "loss": 1.7739,
719
+ "step": 222
720
+ },
721
+ {
722
+ "epoch": 5.46,
723
+ "learning_rate": 2.1836673040456943e-05,
724
+ "loss": 1.7602,
725
+ "step": 224
726
+ },
727
+ {
728
+ "epoch": 5.5,
729
+ "learning_rate": 1.8069848787443555e-05,
730
+ "loss": 1.7773,
731
+ "step": 226
732
+ },
733
+ {
734
+ "epoch": 5.55,
735
+ "learning_rate": 1.4653458081624272e-05,
736
+ "loss": 1.7841,
737
+ "step": 228
738
+ },
739
+ {
740
+ "epoch": 5.6,
741
+ "learning_rate": 1.1589985137590519e-05,
742
+ "loss": 1.7746,
743
+ "step": 230
744
+ },
745
+ {
746
+ "epoch": 5.65,
747
+ "learning_rate": 8.881657547219868e-06,
748
+ "loss": 1.7801,
749
+ "step": 232
750
+ },
751
+ {
752
+ "epoch": 5.7,
753
+ "learning_rate": 6.530444659892443e-06,
754
+ "loss": 1.7639,
755
+ "step": 234
756
+ },
757
+ {
758
+ "epoch": 5.74,
759
+ "learning_rate": 4.538056150485858e-06,
760
+ "loss": 1.7771,
761
+ "step": 236
762
+ },
763
+ {
764
+ "epoch": 5.79,
765
+ "learning_rate": 2.905940776192384e-06,
766
+ "loss": 1.7724,
767
+ "step": 238
768
+ },
769
+ {
770
+ "epoch": 5.84,
771
+ "learning_rate": 1.6352853230609532e-06,
772
+ "loss": 1.7876,
773
+ "step": 240
774
+ },
775
+ {
776
+ "epoch": 5.89,
777
+ "learning_rate": 7.270137430306356e-07,
778
+ "loss": 1.7718,
779
+ "step": 242
780
+ },
781
+ {
782
+ "epoch": 5.94,
783
+ "learning_rate": 1.817864820827242e-07,
784
+ "loss": 1.779,
785
+ "step": 244
786
+ },
787
+ {
788
+ "epoch": 5.98,
789
+ "learning_rate": 0.0,
790
+ "loss": 1.7698,
791
+ "step": 246
792
+ },
793
+ {
794
+ "epoch": 5.98,
795
+ "eval_accuracy": 0.544073054441718,
796
+ "eval_loss": 1.9982993602752686,
797
+ "eval_runtime": 495.216,
798
+ "eval_samples_per_second": 1.187,
799
+ "eval_steps_per_second": 0.297,
800
+ "step": 246
801
+ },
802
+ {
803
+ "before_init_mem_cpu": 699817984,
804
+ "before_init_mem_gpu": 0,
805
+ "epoch": 5.98,
806
+ "init_mem_cpu_alloc_delta": -283938816,
807
+ "init_mem_cpu_peaked_delta": 332918784,
808
+ "init_mem_gpu_alloc_delta": 334744576,
809
+ "init_mem_gpu_peaked_delta": 0,
810
+ "step": 246,
811
+ "total_flos": 8328311331618816.0,
812
+ "train_loss": 2.0183091745143984,
813
+ "train_mem_cpu_alloc_delta": 1155149824,
814
+ "train_mem_cpu_peaked_delta": 308719616,
815
+ "train_mem_gpu_alloc_delta": 988995584,
816
+ "train_mem_gpu_peaked_delta": 4500811776,
817
+ "train_runtime": 35067.5673,
818
+ "train_samples_per_second": 0.911,
819
+ "train_steps_per_second": 0.007
820
+ }
821
+ ],
822
+ "max_steps": 246,
823
+ "num_train_epochs": 6,
824
+ "total_flos": 8328311331618816.0,
825
+ "trial_name": null,
826
+ "trial_params": null
827
+ }