LegrandFrederic commited on
Commit
683faf9
·
verified ·
1 Parent(s): 038cbf6

Upload trainer_state.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. trainer_state.json +455 -0
trainer_state.json ADDED
@@ -0,0 +1,455 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 6.662921348314606,
5
+ "eval_steps": 500,
6
+ "global_step": 593,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.11235955056179775,
13
+ "grad_norm": 3.5541882514953613,
14
+ "learning_rate": 6.666666666666667e-05,
15
+ "loss": 0.9261,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.2247191011235955,
20
+ "grad_norm": 1.3109055757522583,
21
+ "learning_rate": 0.00013333333333333334,
22
+ "loss": 0.3754,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.33707865168539325,
27
+ "grad_norm": 1.4723390340805054,
28
+ "learning_rate": 0.0002,
29
+ "loss": 0.2649,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.449438202247191,
34
+ "grad_norm": 0.7010077238082886,
35
+ "learning_rate": 0.000199844353174683,
36
+ "loss": 0.2166,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.5617977528089888,
41
+ "grad_norm": 1.1943029165267944,
42
+ "learning_rate": 0.00019937789721741653,
43
+ "loss": 0.1767,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.6741573033707865,
48
+ "grad_norm": 0.5831332206726074,
49
+ "learning_rate": 0.00019860208417597864,
50
+ "loss": 0.1481,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.7865168539325843,
55
+ "grad_norm": 1.139088749885559,
56
+ "learning_rate": 0.00019751932910710805,
57
+ "loss": 0.1318,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.898876404494382,
62
+ "grad_norm": 1.130523681640625,
63
+ "learning_rate": 0.00019613300255858616,
64
+ "loss": 0.1277,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 1.0112359550561798,
69
+ "grad_norm": 0.8881850242614746,
70
+ "learning_rate": 0.0001944474200769355,
71
+ "loss": 0.1202,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 1.1235955056179776,
76
+ "grad_norm": 0.6445964574813843,
77
+ "learning_rate": 0.00019246782877339766,
78
+ "loss": 0.101,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 1.2359550561797752,
83
+ "grad_norm": 0.45679864287376404,
84
+ "learning_rate": 0.00019020039099000907,
85
+ "loss": 0.1078,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 1.348314606741573,
90
+ "grad_norm": 0.5455164909362793,
91
+ "learning_rate": 0.0001876521651166215,
92
+ "loss": 0.0931,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 1.4606741573033708,
97
+ "grad_norm": 0.5435442924499512,
98
+ "learning_rate": 0.00018483108361858262,
99
+ "loss": 0.0866,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 1.5730337078651684,
104
+ "grad_norm": 0.4356015622615814,
105
+ "learning_rate": 0.00018174592834347504,
106
+ "loss": 0.0899,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 1.6853932584269664,
111
+ "grad_norm": 0.6584963798522949,
112
+ "learning_rate": 0.00017840630318378232,
113
+ "loss": 0.0905,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 1.797752808988764,
118
+ "grad_norm": 0.500027060508728,
119
+ "learning_rate": 0.00017482260418058164,
120
+ "loss": 0.091,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 1.9101123595505618,
125
+ "grad_norm": 0.720208466053009,
126
+ "learning_rate": 0.00017100598716132773,
127
+ "loss": 0.0805,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 2.0224719101123596,
132
+ "grad_norm": 0.7830264568328857,
133
+ "learning_rate": 0.0001669683330124706,
134
+ "loss": 0.0779,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 2.134831460674157,
139
+ "grad_norm": 0.6720086932182312,
140
+ "learning_rate": 0.0001627222106950102,
141
+ "loss": 0.0731,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 2.247191011235955,
146
+ "grad_norm": 0.4705169200897217,
147
+ "learning_rate": 0.0001582808381181189,
148
+ "loss": 0.0705,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 2.359550561797753,
153
+ "grad_norm": 0.40476611256599426,
154
+ "learning_rate": 0.0001536580409926296,
155
+ "loss": 0.0715,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 2.4719101123595504,
160
+ "grad_norm": 0.5198217034339905,
161
+ "learning_rate": 0.0001488682097924756,
162
+ "loss": 0.0658,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 2.5842696629213484,
167
+ "grad_norm": 0.3450973331928253,
168
+ "learning_rate": 0.00014392625495805912,
169
+ "loss": 0.0613,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 2.696629213483146,
174
+ "grad_norm": 0.4224022626876831,
175
+ "learning_rate": 0.00013884756048099687,
176
+ "loss": 0.0663,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 2.808988764044944,
181
+ "grad_norm": 0.3117457926273346,
182
+ "learning_rate": 0.00013364793601473106,
183
+ "loss": 0.0577,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 2.9213483146067416,
188
+ "grad_norm": 0.2872294783592224,
189
+ "learning_rate": 0.00012834356766008197,
190
+ "loss": 0.0626,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 3.033707865168539,
195
+ "grad_norm": 0.4661742150783539,
196
+ "learning_rate": 0.0001229509675789439,
197
+ "loss": 0.0619,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 3.146067415730337,
202
+ "grad_norm": 0.3543541729450226,
203
+ "learning_rate": 0.00011748692259297347,
204
+ "loss": 0.0494,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 3.258426966292135,
209
+ "grad_norm": 0.3537474572658539,
210
+ "learning_rate": 0.00011196844192727984,
211
+ "loss": 0.052,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 3.370786516853933,
216
+ "grad_norm": 0.3915044367313385,
217
+ "learning_rate": 0.00010641270426178676,
218
+ "loss": 0.0547,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 3.4831460674157304,
223
+ "grad_norm": 0.3162959814071655,
224
+ "learning_rate": 0.00010083700425509279,
225
+ "loss": 0.0524,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 3.595505617977528,
230
+ "grad_norm": 0.38653671741485596,
231
+ "learning_rate": 9.52586987072972e-05,
232
+ "loss": 0.0465,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 3.7078651685393256,
237
+ "grad_norm": 0.3417101800441742,
238
+ "learning_rate": 8.969515252938322e-05,
239
+ "loss": 0.0482,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 3.8202247191011236,
244
+ "grad_norm": 0.28841620683670044,
245
+ "learning_rate": 8.41636846873528e-05,
246
+ "loss": 0.0564,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 3.932584269662921,
251
+ "grad_norm": 0.3317897915840149,
252
+ "learning_rate": 7.868151428938502e-05,
253
+ "loss": 0.0458,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 4.044943820224719,
258
+ "grad_norm": 0.1963573545217514,
259
+ "learning_rate": 7.326570698384568e-05,
260
+ "loss": 0.0498,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 4.157303370786517,
265
+ "grad_norm": 0.2925940454006195,
266
+ "learning_rate": 6.793312183500759e-05,
267
+ "loss": 0.0411,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 4.269662921348314,
272
+ "grad_norm": 0.2857739329338074,
273
+ "learning_rate": 6.270035884185367e-05,
274
+ "loss": 0.0391,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 4.382022471910112,
279
+ "grad_norm": 0.3383257985115051,
280
+ "learning_rate": 5.758370726333434e-05,
281
+ "loss": 0.0448,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 4.49438202247191,
286
+ "grad_norm": 0.2550479471683502,
287
+ "learning_rate": 5.2599094910938594e-05,
288
+ "loss": 0.0382,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 4.606741573033708,
293
+ "grad_norm": 0.21575377881526947,
294
+ "learning_rate": 4.7762038566428155e-05,
295
+ "loss": 0.043,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 4.719101123595506,
300
+ "grad_norm": 0.24605822563171387,
301
+ "learning_rate": 4.3087595679081096e-05,
302
+ "loss": 0.0417,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 4.831460674157303,
307
+ "grad_norm": 0.2827017903327942,
308
+ "learning_rate": 3.8590317492808236e-05,
309
+ "loss": 0.0421,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 4.943820224719101,
314
+ "grad_norm": 0.2163945734500885,
315
+ "learning_rate": 3.428420374905483e-05,
316
+ "loss": 0.0412,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 5.056179775280899,
321
+ "grad_norm": 0.22983698546886444,
322
+ "learning_rate": 3.0182659106494192e-05,
323
+ "loss": 0.0393,
324
+ "step": 450
325
+ },
326
+ {
327
+ "epoch": 5.168539325842697,
328
+ "grad_norm": 0.25033506751060486,
329
+ "learning_rate": 2.629845141317656e-05,
330
+ "loss": 0.0358,
331
+ "step": 460
332
+ },
333
+ {
334
+ "epoch": 5.280898876404494,
335
+ "grad_norm": 0.21685783565044403,
336
+ "learning_rate": 2.264367196102869e-05,
337
+ "loss": 0.0332,
338
+ "step": 470
339
+ },
340
+ {
341
+ "epoch": 5.393258426966292,
342
+ "grad_norm": 0.2170724868774414,
343
+ "learning_rate": 1.9229697846429773e-05,
344
+ "loss": 0.0379,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 5.50561797752809,
349
+ "grad_norm": 0.17588071525096893,
350
+ "learning_rate": 1.606715655403289e-05,
351
+ "loss": 0.033,
352
+ "step": 490
353
+ },
354
+ {
355
+ "epoch": 5.617977528089888,
356
+ "grad_norm": 0.3224557936191559,
357
+ "learning_rate": 1.3165892874079899e-05,
358
+ "loss": 0.0379,
359
+ "step": 500
360
+ },
361
+ {
362
+ "epoch": 5.730337078651686,
363
+ "grad_norm": 0.443591445684433,
364
+ "learning_rate": 1.0534938256194671e-05,
365
+ "loss": 0.0352,
366
+ "step": 510
367
+ },
368
+ {
369
+ "epoch": 5.842696629213483,
370
+ "grad_norm": 0.23971615731716156,
371
+ "learning_rate": 8.182482695053728e-06,
372
+ "loss": 0.0343,
373
+ "step": 520
374
+ },
375
+ {
376
+ "epoch": 5.955056179775281,
377
+ "grad_norm": 0.33371302485466003,
378
+ "learning_rate": 6.1158492354529195e-06,
379
+ "loss": 0.0346,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 6.067415730337078,
384
+ "grad_norm": 0.2301403433084488,
385
+ "learning_rate": 4.3414711761338375e-06,
386
+ "loss": 0.0343,
387
+ "step": 540
388
+ },
389
+ {
390
+ "epoch": 6.179775280898877,
391
+ "grad_norm": 0.16674034297466278,
392
+ "learning_rate": 2.8648720433333996e-06,
393
+ "loss": 0.03,
394
+ "step": 550
395
+ },
396
+ {
397
+ "epoch": 6.292134831460674,
398
+ "grad_norm": 0.2350272834300995,
399
+ "learning_rate": 1.6906483963973207e-06,
400
+ "loss": 0.0344,
401
+ "step": 560
402
+ },
403
+ {
404
+ "epoch": 6.404494382022472,
405
+ "grad_norm": 0.1408814638853073,
406
+ "learning_rate": 8.224555189827565e-07,
407
+ "loss": 0.0387,
408
+ "step": 570
409
+ },
410
+ {
411
+ "epoch": 6.51685393258427,
412
+ "grad_norm": 0.2586834132671356,
413
+ "learning_rate": 2.629960403923715e-07,
414
+ "loss": 0.0329,
415
+ "step": 580
416
+ },
417
+ {
418
+ "epoch": 6.629213483146067,
419
+ "grad_norm": 0.21313035488128662,
420
+ "learning_rate": 1.4011522460866122e-08,
421
+ "loss": 0.0324,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 6.662921348314606,
426
+ "step": 593,
427
+ "total_flos": 8.40154581763943e+16,
428
+ "train_loss": 0.0879759074865666,
429
+ "train_runtime": 704.4365,
430
+ "train_samples_per_second": 53.876,
431
+ "train_steps_per_second": 0.842
432
+ }
433
+ ],
434
+ "logging_steps": 10,
435
+ "max_steps": 593,
436
+ "num_input_tokens_seen": 0,
437
+ "num_train_epochs": 7,
438
+ "save_steps": 10000,
439
+ "stateful_callbacks": {
440
+ "TrainerControl": {
441
+ "args": {
442
+ "should_epoch_stop": false,
443
+ "should_evaluate": false,
444
+ "should_log": false,
445
+ "should_save": true,
446
+ "should_training_stop": true
447
+ },
448
+ "attributes": {}
449
+ }
450
+ },
451
+ "total_flos": 8.40154581763943e+16,
452
+ "train_batch_size": 64,
453
+ "trial_name": null,
454
+ "trial_params": null
455
+ }