filippogreppi commited on
Commit
0988c8c
·
verified ·
1 Parent(s): 6a9f941

Add trainer_state.json of the last checkpoint

Browse files
Files changed (1) hide show
  1. trainer_state.json +744 -0
trainer_state.json ADDED
@@ -0,0 +1,744 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 908,
3
+ "best_metric": 0.5586269497871399,
4
+ "best_model_checkpoint": "/outputs/work/smollm3-3b_structured/checkpoints/checkpoint-908",
5
+ "epoch": 2.0,
6
+ "eval_steps": 100,
7
+ "global_step": 908,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.02203856749311295,
14
+ "grad_norm": 0.9534511566162109,
15
+ "learning_rate": 3.3333333333333335e-05,
16
+ "loss": 1.2359769821166993,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0440771349862259,
21
+ "grad_norm": 0.6362833976745605,
22
+ "learning_rate": 7.037037037037038e-05,
23
+ "loss": 1.0208109855651855,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.06611570247933884,
28
+ "grad_norm": 0.6407135128974915,
29
+ "learning_rate": 9.999872841297521e-05,
30
+ "loss": 0.8830883979797364,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0881542699724518,
35
+ "grad_norm": 0.8838831186294556,
36
+ "learning_rate": 9.995422965786025e-05,
37
+ "loss": 0.8082539558410644,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.11019283746556474,
42
+ "grad_norm": 0.5342389345169067,
43
+ "learning_rate": 9.984621621406273e-05,
44
+ "loss": 0.7579926490783692,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.1322314049586777,
49
+ "grad_norm": 0.8452595472335815,
50
+ "learning_rate": 9.96748254161048e-05,
51
+ "loss": 0.769981575012207,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.15426997245179064,
56
+ "grad_norm": 0.5976874232292175,
57
+ "learning_rate": 9.944027518013187e-05,
58
+ "loss": 0.6899113655090332,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.1763085399449036,
63
+ "grad_norm": 0.6876652836799622,
64
+ "learning_rate": 9.914286372684153e-05,
65
+ "loss": 0.6597393989562989,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.19834710743801653,
70
+ "grad_norm": 0.6256213784217834,
71
+ "learning_rate": 9.878296920230869e-05,
72
+ "loss": 0.6918736934661865,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.22038567493112948,
77
+ "grad_norm": 0.6878316402435303,
78
+ "learning_rate": 9.83610491971886e-05,
79
+ "loss": 0.7168015956878662,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.22038567493112948,
84
+ "eval_loss": 0.6784517168998718,
85
+ "eval_runtime": 11.9227,
86
+ "eval_samples_per_second": 16.104,
87
+ "eval_steps_per_second": 16.104,
88
+ "step": 100
89
+ },
90
+ {
91
+ "epoch": 0.24242424242424243,
92
+ "grad_norm": 0.6689047813415527,
93
+ "learning_rate": 9.787764016490992e-05,
94
+ "loss": 0.6953184127807617,
95
+ "step": 110
96
+ },
97
+ {
98
+ "epoch": 0.2644628099173554,
99
+ "grad_norm": 0.47161027789115906,
100
+ "learning_rate": 9.733335673959671e-05,
101
+ "loss": 0.6750634193420411,
102
+ "step": 120
103
+ },
104
+ {
105
+ "epoch": 0.2865013774104683,
106
+ "grad_norm": 0.7510103583335876,
107
+ "learning_rate": 9.672889095458734e-05,
108
+ "loss": 0.6649476528167725,
109
+ "step": 130
110
+ },
111
+ {
112
+ "epoch": 0.3085399449035813,
113
+ "grad_norm": 0.6093885898590088,
114
+ "learning_rate": 9.606501136254327e-05,
115
+ "loss": 0.6267831802368165,
116
+ "step": 140
117
+ },
118
+ {
119
+ "epoch": 0.3305785123966942,
120
+ "grad_norm": 0.6549955606460571,
121
+ "learning_rate": 9.534256205826684e-05,
122
+ "loss": 0.6659124374389649,
123
+ "step": 150
124
+ },
125
+ {
126
+ "epoch": 0.3526170798898072,
127
+ "grad_norm": 0.5473728775978088,
128
+ "learning_rate": 9.456246160547057e-05,
129
+ "loss": 0.6131270408630372,
130
+ "step": 160
131
+ },
132
+ {
133
+ "epoch": 0.3746556473829201,
134
+ "grad_norm": 0.6127820014953613,
135
+ "learning_rate": 9.372570186886225e-05,
136
+ "loss": 0.609403657913208,
137
+ "step": 170
138
+ },
139
+ {
140
+ "epoch": 0.39669421487603307,
141
+ "grad_norm": 0.6445918083190918,
142
+ "learning_rate": 9.283334675303084e-05,
143
+ "loss": 0.6382901668548584,
144
+ "step": 180
145
+ },
146
+ {
147
+ "epoch": 0.418732782369146,
148
+ "grad_norm": 0.7743238210678101,
149
+ "learning_rate": 9.188653084973692e-05,
150
+ "loss": 0.6500725746154785,
151
+ "step": 190
152
+ },
153
+ {
154
+ "epoch": 0.44077134986225897,
155
+ "grad_norm": 0.7458897829055786,
156
+ "learning_rate": 9.088645799532729e-05,
157
+ "loss": 0.6170906543731689,
158
+ "step": 200
159
+ },
160
+ {
161
+ "epoch": 0.44077134986225897,
162
+ "eval_loss": 0.6293469071388245,
163
+ "eval_runtime": 11.9203,
164
+ "eval_samples_per_second": 16.107,
165
+ "eval_steps_per_second": 16.107,
166
+ "step": 200
167
+ },
168
+ {
169
+ "epoch": 0.4628099173553719,
170
+ "grad_norm": 0.5193160772323608,
171
+ "learning_rate": 8.983439974010794e-05,
172
+ "loss": 0.6225500583648682,
173
+ "step": 210
174
+ },
175
+ {
176
+ "epoch": 0.48484848484848486,
177
+ "grad_norm": 0.5633059740066528,
178
+ "learning_rate": 8.873169373162174e-05,
179
+ "loss": 0.6362756729125977,
180
+ "step": 220
181
+ },
182
+ {
183
+ "epoch": 0.5068870523415978,
184
+ "grad_norm": 0.5231749415397644,
185
+ "learning_rate": 8.757974201388605e-05,
186
+ "loss": 0.6010201454162598,
187
+ "step": 230
188
+ },
189
+ {
190
+ "epoch": 0.5289256198347108,
191
+ "grad_norm": 0.6744521260261536,
192
+ "learning_rate": 8.638000924475336e-05,
193
+ "loss": 0.6185548305511475,
194
+ "step": 240
195
+ },
196
+ {
197
+ "epoch": 0.5509641873278237,
198
+ "grad_norm": 0.6033552885055542,
199
+ "learning_rate": 8.513402083366054e-05,
200
+ "loss": 0.5994134426116944,
201
+ "step": 250
202
+ },
203
+ {
204
+ "epoch": 0.5730027548209367,
205
+ "grad_norm": 0.5614070296287537,
206
+ "learning_rate": 8.384336100213546e-05,
207
+ "loss": 0.5692955493927002,
208
+ "step": 260
209
+ },
210
+ {
211
+ "epoch": 0.5950413223140496,
212
+ "grad_norm": 0.6417648196220398,
213
+ "learning_rate": 8.250967076952635e-05,
214
+ "loss": 0.6048506736755371,
215
+ "step": 270
216
+ },
217
+ {
218
+ "epoch": 0.6170798898071626,
219
+ "grad_norm": 0.6965922713279724,
220
+ "learning_rate": 8.113464586651516e-05,
221
+ "loss": 0.6014030456542969,
222
+ "step": 280
223
+ },
224
+ {
225
+ "epoch": 0.6391184573002755,
226
+ "grad_norm": 0.6192167401313782,
227
+ "learning_rate": 7.972003457906773e-05,
228
+ "loss": 0.634767723083496,
229
+ "step": 290
230
+ },
231
+ {
232
+ "epoch": 0.6611570247933884,
233
+ "grad_norm": 0.720318615436554,
234
+ "learning_rate": 7.826763552556222e-05,
235
+ "loss": 0.5867681503295898,
236
+ "step": 300
237
+ },
238
+ {
239
+ "epoch": 0.6611570247933884,
240
+ "eval_loss": 0.6011102795600891,
241
+ "eval_runtime": 12.0266,
242
+ "eval_samples_per_second": 15.965,
243
+ "eval_steps_per_second": 15.965,
244
+ "step": 300
245
+ },
246
+ {
247
+ "epoch": 0.6831955922865014,
248
+ "grad_norm": 0.7235103249549866,
249
+ "learning_rate": 7.677929536992194e-05,
250
+ "loss": 0.5783907413482666,
251
+ "step": 310
252
+ },
253
+ {
254
+ "epoch": 0.7052341597796143,
255
+ "grad_norm": 0.7653256058692932,
256
+ "learning_rate": 7.525690647366032e-05,
257
+ "loss": 0.6062646865844726,
258
+ "step": 320
259
+ },
260
+ {
261
+ "epoch": 0.7272727272727273,
262
+ "grad_norm": 0.5779842734336853,
263
+ "learning_rate": 7.370240448982344e-05,
264
+ "loss": 0.6198268413543702,
265
+ "step": 330
266
+ },
267
+ {
268
+ "epoch": 0.7493112947658402,
269
+ "grad_norm": 0.709947943687439,
270
+ "learning_rate": 7.211776590188898e-05,
271
+ "loss": 0.6091408729553223,
272
+ "step": 340
273
+ },
274
+ {
275
+ "epoch": 0.7713498622589532,
276
+ "grad_norm": 0.595930278301239,
277
+ "learning_rate": 7.050500551075121e-05,
278
+ "loss": 0.6241033554077149,
279
+ "step": 350
280
+ },
281
+ {
282
+ "epoch": 0.7933884297520661,
283
+ "grad_norm": 0.5700712203979492,
284
+ "learning_rate": 6.886617387298689e-05,
285
+ "loss": 0.6038753509521484,
286
+ "step": 360
287
+ },
288
+ {
289
+ "epoch": 0.8154269972451791,
290
+ "grad_norm": 0.6845744252204895,
291
+ "learning_rate": 6.720335469365943e-05,
292
+ "loss": 0.5801003932952881,
293
+ "step": 370
294
+ },
295
+ {
296
+ "epoch": 0.837465564738292,
297
+ "grad_norm": 0.5995373725891113,
298
+ "learning_rate": 6.551866217697602e-05,
299
+ "loss": 0.5841936588287353,
300
+ "step": 380
301
+ },
302
+ {
303
+ "epoch": 0.859504132231405,
304
+ "grad_norm": 0.702873170375824,
305
+ "learning_rate": 6.381423833816653e-05,
306
+ "loss": 0.5742908477783203,
307
+ "step": 390
308
+ },
309
+ {
310
+ "epoch": 0.8815426997245179,
311
+ "grad_norm": 0.5957848429679871,
312
+ "learning_rate": 6.209225028000173e-05,
313
+ "loss": 0.6035511493682861,
314
+ "step": 400
315
+ },
316
+ {
317
+ "epoch": 0.8815426997245179,
318
+ "eval_loss": 0.5868632197380066,
319
+ "eval_runtime": 11.9519,
320
+ "eval_samples_per_second": 16.064,
321
+ "eval_steps_per_second": 16.064,
322
+ "step": 400
323
+ },
324
+ {
325
+ "epoch": 0.9035812672176309,
326
+ "grad_norm": 0.6939573884010315,
327
+ "learning_rate": 6.0354887437413965e-05,
328
+ "loss": 0.5780395030975342,
329
+ "step": 410
330
+ },
331
+ {
332
+ "epoch": 0.9256198347107438,
333
+ "grad_norm": 0.6161796450614929,
334
+ "learning_rate": 5.860435879372319e-05,
335
+ "loss": 0.5949501037597656,
336
+ "step": 420
337
+ },
338
+ {
339
+ "epoch": 0.9476584022038568,
340
+ "grad_norm": 0.6215364336967468,
341
+ "learning_rate": 5.684289007200819e-05,
342
+ "loss": 0.5814109802246094,
343
+ "step": 430
344
+ },
345
+ {
346
+ "epoch": 0.9696969696969697,
347
+ "grad_norm": 0.523289680480957,
348
+ "learning_rate": 5.5072720905193984e-05,
349
+ "loss": 0.5734921932220459,
350
+ "step": 440
351
+ },
352
+ {
353
+ "epoch": 0.9917355371900827,
354
+ "grad_norm": 0.6269870400428772,
355
+ "learning_rate": 5.329610198845322e-05,
356
+ "loss": 0.5888432502746582,
357
+ "step": 450
358
+ },
359
+ {
360
+ "epoch": 1.0132231404958678,
361
+ "grad_norm": 0.6019392013549805,
362
+ "learning_rate": 5.151529221754262e-05,
363
+ "loss": 0.5435478687286377,
364
+ "step": 460
365
+ },
366
+ {
367
+ "epoch": 1.0352617079889808,
368
+ "grad_norm": 0.6337491869926453,
369
+ "learning_rate": 4.97325558167126e-05,
370
+ "loss": 0.5148682594299316,
371
+ "step": 470
372
+ },
373
+ {
374
+ "epoch": 1.0573002754820937,
375
+ "grad_norm": 0.6463139653205872,
376
+ "learning_rate": 4.795015945984198e-05,
377
+ "loss": 0.4946620941162109,
378
+ "step": 480
379
+ },
380
+ {
381
+ "epoch": 1.0793388429752067,
382
+ "grad_norm": 0.667147159576416,
383
+ "learning_rate": 4.617036938845806e-05,
384
+ "loss": 0.5063100337982178,
385
+ "step": 490
386
+ },
387
+ {
388
+ "epoch": 1.1013774104683196,
389
+ "grad_norm": 0.5854341983795166,
390
+ "learning_rate": 4.43954485303065e-05,
391
+ "loss": 0.5141865730285644,
392
+ "step": 500
393
+ },
394
+ {
395
+ "epoch": 1.1013774104683196,
396
+ "eval_loss": 0.5751394629478455,
397
+ "eval_runtime": 11.933,
398
+ "eval_samples_per_second": 16.09,
399
+ "eval_steps_per_second": 16.09,
400
+ "step": 500
401
+ },
402
+ {
403
+ "epoch": 1.1234159779614326,
404
+ "grad_norm": 0.7505147457122803,
405
+ "learning_rate": 4.2627653622134346e-05,
406
+ "loss": 0.5038840770721436,
407
+ "step": 510
408
+ },
409
+ {
410
+ "epoch": 1.1454545454545455,
411
+ "grad_norm": 0.6712458729743958,
412
+ "learning_rate": 4.086923234034493e-05,
413
+ "loss": 0.49640936851501466,
414
+ "step": 520
415
+ },
416
+ {
417
+ "epoch": 1.1674931129476585,
418
+ "grad_norm": 0.6768338084220886,
419
+ "learning_rate": 3.912242044317233e-05,
420
+ "loss": 0.5040760993957519,
421
+ "step": 530
422
+ },
423
+ {
424
+ "epoch": 1.1895316804407714,
425
+ "grad_norm": 0.6140052080154419,
426
+ "learning_rate": 3.738943892800951e-05,
427
+ "loss": 0.4928180694580078,
428
+ "step": 540
429
+ },
430
+ {
431
+ "epoch": 1.2115702479338843,
432
+ "grad_norm": 0.7098944187164307,
433
+ "learning_rate": 3.5672491207504e-05,
434
+ "loss": 0.49433393478393556,
435
+ "step": 550
436
+ },
437
+ {
438
+ "epoch": 1.2336088154269973,
439
+ "grad_norm": 0.603570282459259,
440
+ "learning_rate": 3.397376030801215e-05,
441
+ "loss": 0.5312057495117187,
442
+ "step": 560
443
+ },
444
+ {
445
+ "epoch": 1.2556473829201102,
446
+ "grad_norm": 0.8194121718406677,
447
+ "learning_rate": 3.229540609397334e-05,
448
+ "loss": 0.5198305130004883,
449
+ "step": 570
450
+ },
451
+ {
452
+ "epoch": 1.2776859504132232,
453
+ "grad_norm": 0.7442294955253601,
454
+ "learning_rate": 3.0639562521733935e-05,
455
+ "loss": 0.4971327304840088,
456
+ "step": 580
457
+ },
458
+ {
459
+ "epoch": 1.2997245179063361,
460
+ "grad_norm": 0.664591372013092,
461
+ "learning_rate": 2.900833492631191e-05,
462
+ "loss": 0.4868955612182617,
463
+ "step": 590
464
+ },
465
+ {
466
+ "epoch": 1.321763085399449,
467
+ "grad_norm": 0.7322269678115845,
468
+ "learning_rate": 2.7403797344552696e-05,
469
+ "loss": 0.5075324058532715,
470
+ "step": 600
471
+ },
472
+ {
473
+ "epoch": 1.321763085399449,
474
+ "eval_loss": 0.5670042634010315,
475
+ "eval_runtime": 11.8846,
476
+ "eval_samples_per_second": 16.155,
477
+ "eval_steps_per_second": 16.155,
478
+ "step": 600
479
+ },
480
+ {
481
+ "epoch": 1.343801652892562,
482
+ "grad_norm": 0.7925372123718262,
483
+ "learning_rate": 2.5827989878078972e-05,
484
+ "loss": 0.5191412925720215,
485
+ "step": 610
486
+ },
487
+ {
488
+ "epoch": 1.365840220385675,
489
+ "grad_norm": 0.7049777507781982,
490
+ "learning_rate": 2.428291609938786e-05,
491
+ "loss": 0.4850438117980957,
492
+ "step": 620
493
+ },
494
+ {
495
+ "epoch": 1.387878787878788,
496
+ "grad_norm": 0.5821607112884521,
497
+ "learning_rate": 2.2770540504393224e-05,
498
+ "loss": 0.4845143795013428,
499
+ "step": 630
500
+ },
501
+ {
502
+ "epoch": 1.4099173553719009,
503
+ "grad_norm": 0.8042203187942505,
504
+ "learning_rate": 2.1292786014652398e-05,
505
+ "loss": 0.5041770935058594,
506
+ "step": 640
507
+ },
508
+ {
509
+ "epoch": 1.4319559228650138,
510
+ "grad_norm": 0.6269323229789734,
511
+ "learning_rate": 1.9851531532452665e-05,
512
+ "loss": 0.4969661712646484,
513
+ "step": 650
514
+ },
515
+ {
516
+ "epoch": 1.4539944903581268,
517
+ "grad_norm": 0.6438784599304199,
518
+ "learning_rate": 1.8448609551866647e-05,
519
+ "loss": 0.49288215637207033,
520
+ "step": 660
521
+ },
522
+ {
523
+ "epoch": 1.4760330578512397,
524
+ "grad_norm": 0.6186577081680298,
525
+ "learning_rate": 1.7085803828813546e-05,
526
+ "loss": 0.5123368263244629,
527
+ "step": 670
528
+ },
529
+ {
530
+ "epoch": 1.4980716253443527,
531
+ "grad_norm": 0.8740533590316772,
532
+ "learning_rate": 1.5764847113089094e-05,
533
+ "loss": 0.47509055137634276,
534
+ "step": 680
535
+ },
536
+ {
537
+ "epoch": 1.5201101928374654,
538
+ "grad_norm": 0.7086982727050781,
539
+ "learning_rate": 1.4487418945247438e-05,
540
+ "loss": 0.49436202049255373,
541
+ "step": 690
542
+ },
543
+ {
544
+ "epoch": 1.5421487603305786,
545
+ "grad_norm": 0.6316142082214355,
546
+ "learning_rate": 1.3255143521136498e-05,
547
+ "loss": 0.5204075336456299,
548
+ "step": 700
549
+ },
550
+ {
551
+ "epoch": 1.5421487603305786,
552
+ "eval_loss": 0.5617139339447021,
553
+ "eval_runtime": 12.2107,
554
+ "eval_samples_per_second": 15.724,
555
+ "eval_steps_per_second": 15.724,
556
+ "step": 700
557
+ },
558
+ {
559
+ "epoch": 1.5641873278236913,
560
+ "grad_norm": 0.6398480534553528,
561
+ "learning_rate": 1.206958762680157e-05,
562
+ "loss": 0.49640421867370604,
563
+ "step": 710
564
+ },
565
+ {
566
+ "epoch": 1.5862258953168045,
567
+ "grad_norm": 0.7017738819122314,
568
+ "learning_rate": 1.0932258646383336e-05,
569
+ "loss": 0.5167774677276611,
570
+ "step": 720
571
+ },
572
+ {
573
+ "epoch": 1.6082644628099172,
574
+ "grad_norm": 0.8112418055534363,
575
+ "learning_rate": 9.844602645542584e-06,
576
+ "loss": 0.4898106575012207,
577
+ "step": 730
578
+ },
579
+ {
580
+ "epoch": 1.6303030303030304,
581
+ "grad_norm": 0.6042963862419128,
582
+ "learning_rate": 8.808002532849047e-06,
583
+ "loss": 0.5056878089904785,
584
+ "step": 740
585
+ },
586
+ {
587
+ "epoch": 1.652341597796143,
588
+ "grad_norm": 0.6272949576377869,
589
+ "learning_rate": 7.823776301471591e-06,
590
+ "loss": 0.49520001411437986,
591
+ "step": 750
592
+ },
593
+ {
594
+ "epoch": 1.6743801652892563,
595
+ "grad_norm": 0.6539549231529236,
596
+ "learning_rate": 6.893175353405756e-06,
597
+ "loss": 0.4806521892547607,
598
+ "step": 760
599
+ },
600
+ {
601
+ "epoch": 1.696418732782369,
602
+ "grad_norm": 0.6603189706802368,
603
+ "learning_rate": 6.017382908369051e-06,
604
+ "loss": 0.4939537048339844,
605
+ "step": 770
606
+ },
607
+ {
608
+ "epoch": 1.7184573002754822,
609
+ "grad_norm": 0.9547717571258545,
610
+ "learning_rate": 5.197512499387175e-06,
611
+ "loss": 0.5203057765960694,
612
+ "step": 780
613
+ },
614
+ {
615
+ "epoch": 1.7404958677685949,
616
+ "grad_norm": 0.7032685279846191,
617
+ "learning_rate": 4.434606556983878e-06,
618
+ "loss": 0.5253971576690674,
619
+ "step": 790
620
+ },
621
+ {
622
+ "epoch": 1.762534435261708,
623
+ "grad_norm": 0.6500975489616394,
624
+ "learning_rate": 3.729635083774724e-06,
625
+ "loss": 0.5071902275085449,
626
+ "step": 800
627
+ },
628
+ {
629
+ "epoch": 1.762534435261708,
630
+ "eval_loss": 0.5592079162597656,
631
+ "eval_runtime": 11.8791,
632
+ "eval_samples_per_second": 16.163,
633
+ "eval_steps_per_second": 16.163,
634
+ "step": 800
635
+ },
636
+ {
637
+ "epoch": 1.7845730027548208,
638
+ "grad_norm": 0.5473243594169617,
639
+ "learning_rate": 3.083494421149752e-06,
640
+ "loss": 0.5045226097106934,
641
+ "step": 810
642
+ },
643
+ {
644
+ "epoch": 1.806611570247934,
645
+ "grad_norm": 0.6320291757583618,
646
+ "learning_rate": 2.49700610961342e-06,
647
+ "loss": 0.5111465454101562,
648
+ "step": 820
649
+ },
650
+ {
651
+ "epoch": 1.8286501377410467,
652
+ "grad_norm": 0.6422545909881592,
653
+ "learning_rate": 1.9709158442305365e-06,
654
+ "loss": 0.48850622177124026,
655
+ "step": 830
656
+ },
657
+ {
658
+ "epoch": 1.8506887052341598,
659
+ "grad_norm": 0.7762777209281921,
660
+ "learning_rate": 1.5058925265066194e-06,
661
+ "loss": 0.4583238124847412,
662
+ "step": 840
663
+ },
664
+ {
665
+ "epoch": 1.8727272727272726,
666
+ "grad_norm": 0.6142582297325134,
667
+ "learning_rate": 1.1025274139079057e-06,
668
+ "loss": 0.4907660484313965,
669
+ "step": 850
670
+ },
671
+ {
672
+ "epoch": 1.8947658402203857,
673
+ "grad_norm": 0.8006341457366943,
674
+ "learning_rate": 7.613333681025236e-07,
675
+ "loss": 0.4667381286621094,
676
+ "step": 860
677
+ },
678
+ {
679
+ "epoch": 1.9168044077134985,
680
+ "grad_norm": 0.8411538004875183,
681
+ "learning_rate": 4.827442028785655e-07,
682
+ "loss": 0.5014886379241943,
683
+ "step": 870
684
+ },
685
+ {
686
+ "epoch": 1.9388429752066116,
687
+ "grad_norm": 0.6041914820671082,
688
+ "learning_rate": 2.67114132568258e-07,
689
+ "loss": 0.5158967018127442,
690
+ "step": 880
691
+ },
692
+ {
693
+ "epoch": 1.9608815426997244,
694
+ "grad_norm": 0.6050441861152649,
695
+ "learning_rate": 1.1471732167938065e-07,
696
+ "loss": 0.49015021324157715,
697
+ "step": 890
698
+ },
699
+ {
700
+ "epoch": 1.9829201101928375,
701
+ "grad_norm": 0.6653909087181091,
702
+ "learning_rate": 2.57475363067472e-08,
703
+ "loss": 0.5256523609161377,
704
+ "step": 900
705
+ },
706
+ {
707
+ "epoch": 1.9829201101928375,
708
+ "eval_loss": 0.5588088035583496,
709
+ "eval_runtime": 11.8638,
710
+ "eval_samples_per_second": 16.184,
711
+ "eval_steps_per_second": 16.184,
712
+ "step": 900
713
+ },
714
+ {
715
+ "epoch": 2.0,
716
+ "eval_loss": 0.5586269497871399,
717
+ "eval_runtime": 11.8357,
718
+ "eval_samples_per_second": 16.222,
719
+ "eval_steps_per_second": 16.222,
720
+ "step": 908
721
+ }
722
+ ],
723
+ "logging_steps": 10,
724
+ "max_steps": 908,
725
+ "num_input_tokens_seen": 0,
726
+ "num_train_epochs": 2,
727
+ "save_steps": 100,
728
+ "stateful_callbacks": {
729
+ "TrainerControl": {
730
+ "args": {
731
+ "should_epoch_stop": false,
732
+ "should_evaluate": false,
733
+ "should_log": false,
734
+ "should_save": true,
735
+ "should_training_stop": true
736
+ },
737
+ "attributes": {}
738
+ }
739
+ },
740
+ "total_flos": 1.1401079002855834e+17,
741
+ "train_batch_size": 1,
742
+ "trial_name": null,
743
+ "trial_params": null
744
+ }