RASMUS commited on
Commit
7caa0a3
1 Parent(s): e48e398

Upload 6 files

Browse files
checkpoint-34077/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:806f7f705f6f9c165e29f1d0191aa96d497c54970e6bf03ccf25d46b8b4857a5
3
+ size 2362860
checkpoint-34077/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a8521d1f88f84bb03f9a1a9a799150b8d537364ddca0031e22b02b919826a23
3
+ size 4730636
checkpoint-34077/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b457a94b1434789b520439b2d787ff907bf4b2dc2d93cf901827a24448f0cf7
3
+ size 14244
checkpoint-34077/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cefb1ff66bead85b347dd9fff851894b1fe220800280b0202b2d9e26f8ae6261
3
+ size 1064
checkpoint-34077/trainer_state.json ADDED
@@ -0,0 +1,527 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 34077,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.044017959327405584,
13
+ "grad_norm": 0.6358331441879272,
14
+ "learning_rate": 1.9706546937817296e-05,
15
+ "loss": 2.539,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.08803591865481117,
20
+ "grad_norm": 0.4060266613960266,
21
+ "learning_rate": 1.9413093875634593e-05,
22
+ "loss": 1.2396,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.13205387798221674,
27
+ "grad_norm": 4.214008808135986,
28
+ "learning_rate": 1.911964081345189e-05,
29
+ "loss": 0.9035,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.17607183730962234,
34
+ "grad_norm": 1.6060826778411865,
35
+ "learning_rate": 1.882618775126919e-05,
36
+ "loss": 0.7033,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.2200897966370279,
41
+ "grad_norm": 4.546453475952148,
42
+ "learning_rate": 1.8532734689086482e-05,
43
+ "loss": 0.6154,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.2641077559644335,
48
+ "grad_norm": 1.295145034790039,
49
+ "learning_rate": 1.8239281626903777e-05,
50
+ "loss": 0.5954,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.3081257152918391,
55
+ "grad_norm": 5.553115367889404,
56
+ "learning_rate": 1.7945828564721074e-05,
57
+ "loss": 0.5736,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.3521436746192447,
62
+ "grad_norm": 6.943908214569092,
63
+ "learning_rate": 1.765237550253837e-05,
64
+ "loss": 0.541,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.3961616339466502,
69
+ "grad_norm": 1.2528984546661377,
70
+ "learning_rate": 1.7358922440355666e-05,
71
+ "loss": 0.5497,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.4401795932740558,
76
+ "grad_norm": 5.552521705627441,
77
+ "learning_rate": 1.7065469378172963e-05,
78
+ "loss": 0.5508,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.4841975526014614,
83
+ "grad_norm": 2.075917959213257,
84
+ "learning_rate": 1.6772016315990257e-05,
85
+ "loss": 0.5487,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.528215511928867,
90
+ "grad_norm": 2.551165819168091,
91
+ "learning_rate": 1.6478563253807555e-05,
92
+ "loss": 0.5185,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.5722334712562726,
97
+ "grad_norm": 1.3819420337677002,
98
+ "learning_rate": 1.6185110191624852e-05,
99
+ "loss": 0.5338,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.6162514305836782,
104
+ "grad_norm": 1.1496446132659912,
105
+ "learning_rate": 1.5891657129442147e-05,
106
+ "loss": 0.5291,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.6602693899110837,
111
+ "grad_norm": 2.217291831970215,
112
+ "learning_rate": 1.5598204067259444e-05,
113
+ "loss": 0.5396,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.7042873492384893,
118
+ "grad_norm": 0.9094293117523193,
119
+ "learning_rate": 1.5304751005076738e-05,
120
+ "loss": 0.5232,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.7483053085658948,
125
+ "grad_norm": 1.1623644828796387,
126
+ "learning_rate": 1.5011297942894036e-05,
127
+ "loss": 0.5318,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.7923232678933004,
132
+ "grad_norm": 4.72590970993042,
133
+ "learning_rate": 1.4717844880711332e-05,
134
+ "loss": 0.5251,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.836341227220706,
139
+ "grad_norm": 3.225889205932617,
140
+ "learning_rate": 1.4424391818528627e-05,
141
+ "loss": 0.5254,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.8803591865481116,
146
+ "grad_norm": 1.6131037473678589,
147
+ "learning_rate": 1.4130938756345925e-05,
148
+ "loss": 0.5248,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 0.9243771458755172,
153
+ "grad_norm": 1.3447926044464111,
154
+ "learning_rate": 1.3837485694163219e-05,
155
+ "loss": 0.5351,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 0.9683951052029228,
160
+ "grad_norm": 2.7240824699401855,
161
+ "learning_rate": 1.3544032631980515e-05,
162
+ "loss": 0.5108,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 1.0,
167
+ "eval_loss": 0.48541975021362305,
168
+ "eval_mse": 0.48541975021362305,
169
+ "eval_runtime": 4.5323,
170
+ "eval_samples_per_second": 2227.781,
171
+ "eval_steps_per_second": 139.443,
172
+ "step": 11359
173
+ },
174
+ {
175
+ "epoch": 1.0124130645303284,
176
+ "grad_norm": 1.8170539140701294,
177
+ "learning_rate": 1.3250579569797812e-05,
178
+ "loss": 0.5247,
179
+ "step": 11500
180
+ },
181
+ {
182
+ "epoch": 1.056431023857734,
183
+ "grad_norm": 3.0960614681243896,
184
+ "learning_rate": 1.2957126507615108e-05,
185
+ "loss": 0.5071,
186
+ "step": 12000
187
+ },
188
+ {
189
+ "epoch": 1.1004489831851396,
190
+ "grad_norm": 1.9027760028839111,
191
+ "learning_rate": 1.2663673445432406e-05,
192
+ "loss": 0.4881,
193
+ "step": 12500
194
+ },
195
+ {
196
+ "epoch": 1.144466942512545,
197
+ "grad_norm": 2.4669666290283203,
198
+ "learning_rate": 1.23702203832497e-05,
199
+ "loss": 0.497,
200
+ "step": 13000
201
+ },
202
+ {
203
+ "epoch": 1.1884849018399506,
204
+ "grad_norm": 4.2728095054626465,
205
+ "learning_rate": 1.2076767321066996e-05,
206
+ "loss": 0.5037,
207
+ "step": 13500
208
+ },
209
+ {
210
+ "epoch": 1.2325028611673563,
211
+ "grad_norm": 2.4737207889556885,
212
+ "learning_rate": 1.1783314258884293e-05,
213
+ "loss": 0.5081,
214
+ "step": 14000
215
+ },
216
+ {
217
+ "epoch": 1.2765208204947618,
218
+ "grad_norm": 3.5890121459960938,
219
+ "learning_rate": 1.1489861196701589e-05,
220
+ "loss": 0.5156,
221
+ "step": 14500
222
+ },
223
+ {
224
+ "epoch": 1.3205387798221675,
225
+ "grad_norm": 0.9171675443649292,
226
+ "learning_rate": 1.1196408134518883e-05,
227
+ "loss": 0.5041,
228
+ "step": 15000
229
+ },
230
+ {
231
+ "epoch": 1.364556739149573,
232
+ "grad_norm": 0.7652862071990967,
233
+ "learning_rate": 1.090295507233618e-05,
234
+ "loss": 0.4898,
235
+ "step": 15500
236
+ },
237
+ {
238
+ "epoch": 1.4085746984769787,
239
+ "grad_norm": 1.9014813899993896,
240
+ "learning_rate": 1.0609502010153477e-05,
241
+ "loss": 0.5096,
242
+ "step": 16000
243
+ },
244
+ {
245
+ "epoch": 1.4525926578043842,
246
+ "grad_norm": 3.0211782455444336,
247
+ "learning_rate": 1.0316048947970774e-05,
248
+ "loss": 0.5192,
249
+ "step": 16500
250
+ },
251
+ {
252
+ "epoch": 1.4966106171317897,
253
+ "grad_norm": 1.619178295135498,
254
+ "learning_rate": 1.0022595885788068e-05,
255
+ "loss": 0.5106,
256
+ "step": 17000
257
+ },
258
+ {
259
+ "epoch": 1.5406285764591954,
260
+ "grad_norm": 2.8226287364959717,
261
+ "learning_rate": 9.729142823605366e-06,
262
+ "loss": 0.5127,
263
+ "step": 17500
264
+ },
265
+ {
266
+ "epoch": 1.584646535786601,
267
+ "grad_norm": 1.5159013271331787,
268
+ "learning_rate": 9.435689761422662e-06,
269
+ "loss": 0.521,
270
+ "step": 18000
271
+ },
272
+ {
273
+ "epoch": 1.6286644951140063,
274
+ "grad_norm": 1.251665472984314,
275
+ "learning_rate": 9.142236699239957e-06,
276
+ "loss": 0.5104,
277
+ "step": 18500
278
+ },
279
+ {
280
+ "epoch": 1.672682454441412,
281
+ "grad_norm": 1.2266521453857422,
282
+ "learning_rate": 8.848783637057253e-06,
283
+ "loss": 0.4991,
284
+ "step": 19000
285
+ },
286
+ {
287
+ "epoch": 1.7167004137688178,
288
+ "grad_norm": 3.0873475074768066,
289
+ "learning_rate": 8.555330574874549e-06,
290
+ "loss": 0.495,
291
+ "step": 19500
292
+ },
293
+ {
294
+ "epoch": 1.7607183730962233,
295
+ "grad_norm": 1.2593939304351807,
296
+ "learning_rate": 8.261877512691847e-06,
297
+ "loss": 0.5173,
298
+ "step": 20000
299
+ },
300
+ {
301
+ "epoch": 1.8047363324236287,
302
+ "grad_norm": 1.4156285524368286,
303
+ "learning_rate": 7.96842445050914e-06,
304
+ "loss": 0.5047,
305
+ "step": 20500
306
+ },
307
+ {
308
+ "epoch": 1.8487542917510345,
309
+ "grad_norm": 1.6748478412628174,
310
+ "learning_rate": 7.674971388326438e-06,
311
+ "loss": 0.5194,
312
+ "step": 21000
313
+ },
314
+ {
315
+ "epoch": 1.8927722510784402,
316
+ "grad_norm": 2.2740602493286133,
317
+ "learning_rate": 7.381518326143734e-06,
318
+ "loss": 0.5165,
319
+ "step": 21500
320
+ },
321
+ {
322
+ "epoch": 1.9367902104058454,
323
+ "grad_norm": 3.551628589630127,
324
+ "learning_rate": 7.088065263961031e-06,
325
+ "loss": 0.5218,
326
+ "step": 22000
327
+ },
328
+ {
329
+ "epoch": 1.9808081697332511,
330
+ "grad_norm": 3.0715420246124268,
331
+ "learning_rate": 6.794612201778326e-06,
332
+ "loss": 0.5048,
333
+ "step": 22500
334
+ },
335
+ {
336
+ "epoch": 2.0,
337
+ "eval_loss": 0.46880388259887695,
338
+ "eval_mse": 0.46880388259887695,
339
+ "eval_runtime": 4.5138,
340
+ "eval_samples_per_second": 2236.922,
341
+ "eval_steps_per_second": 140.015,
342
+ "step": 22718
343
+ },
344
+ {
345
+ "epoch": 2.024826129060657,
346
+ "grad_norm": 1.7051458358764648,
347
+ "learning_rate": 6.501159139595622e-06,
348
+ "loss": 0.5349,
349
+ "step": 23000
350
+ },
351
+ {
352
+ "epoch": 2.0688440883880626,
353
+ "grad_norm": 2.6633028984069824,
354
+ "learning_rate": 6.207706077412918e-06,
355
+ "loss": 0.4895,
356
+ "step": 23500
357
+ },
358
+ {
359
+ "epoch": 2.112862047715468,
360
+ "grad_norm": 1.9820642471313477,
361
+ "learning_rate": 5.914253015230215e-06,
362
+ "loss": 0.5008,
363
+ "step": 24000
364
+ },
365
+ {
366
+ "epoch": 2.1568800070428735,
367
+ "grad_norm": 1.2028346061706543,
368
+ "learning_rate": 5.620799953047511e-06,
369
+ "loss": 0.4816,
370
+ "step": 24500
371
+ },
372
+ {
373
+ "epoch": 2.2008979663702792,
374
+ "grad_norm": 1.2734020948410034,
375
+ "learning_rate": 5.3273468908648065e-06,
376
+ "loss": 0.4952,
377
+ "step": 25000
378
+ },
379
+ {
380
+ "epoch": 2.2449159256976845,
381
+ "grad_norm": 1.3878185749053955,
382
+ "learning_rate": 5.033893828682102e-06,
383
+ "loss": 0.5007,
384
+ "step": 25500
385
+ },
386
+ {
387
+ "epoch": 2.28893388502509,
388
+ "grad_norm": 2.8702523708343506,
389
+ "learning_rate": 4.740440766499399e-06,
390
+ "loss": 0.5138,
391
+ "step": 26000
392
+ },
393
+ {
394
+ "epoch": 2.332951844352496,
395
+ "grad_norm": 2.3792970180511475,
396
+ "learning_rate": 4.446987704316695e-06,
397
+ "loss": 0.4914,
398
+ "step": 26500
399
+ },
400
+ {
401
+ "epoch": 2.376969803679901,
402
+ "grad_norm": 1.627300500869751,
403
+ "learning_rate": 4.1535346421339915e-06,
404
+ "loss": 0.5116,
405
+ "step": 27000
406
+ },
407
+ {
408
+ "epoch": 2.420987763007307,
409
+ "grad_norm": 3.074572801589966,
410
+ "learning_rate": 3.860081579951287e-06,
411
+ "loss": 0.5001,
412
+ "step": 27500
413
+ },
414
+ {
415
+ "epoch": 2.4650057223347126,
416
+ "grad_norm": 3.1290581226348877,
417
+ "learning_rate": 3.566628517768583e-06,
418
+ "loss": 0.5009,
419
+ "step": 28000
420
+ },
421
+ {
422
+ "epoch": 2.5090236816621183,
423
+ "grad_norm": 1.4590922594070435,
424
+ "learning_rate": 3.273175455585879e-06,
425
+ "loss": 0.4948,
426
+ "step": 28500
427
+ },
428
+ {
429
+ "epoch": 2.5530416409895236,
430
+ "grad_norm": 1.742438554763794,
431
+ "learning_rate": 2.9797223934031757e-06,
432
+ "loss": 0.4972,
433
+ "step": 29000
434
+ },
435
+ {
436
+ "epoch": 2.5970596003169293,
437
+ "grad_norm": 1.680262565612793,
438
+ "learning_rate": 2.6862693312204715e-06,
439
+ "loss": 0.5051,
440
+ "step": 29500
441
+ },
442
+ {
443
+ "epoch": 2.641077559644335,
444
+ "grad_norm": 2.896779775619507,
445
+ "learning_rate": 2.3928162690377677e-06,
446
+ "loss": 0.4851,
447
+ "step": 30000
448
+ },
449
+ {
450
+ "epoch": 2.6850955189717407,
451
+ "grad_norm": 2.4820384979248047,
452
+ "learning_rate": 2.0993632068550636e-06,
453
+ "loss": 0.5053,
454
+ "step": 30500
455
+ },
456
+ {
457
+ "epoch": 2.729113478299146,
458
+ "grad_norm": 4.39924430847168,
459
+ "learning_rate": 1.8059101446723598e-06,
460
+ "loss": 0.4989,
461
+ "step": 31000
462
+ },
463
+ {
464
+ "epoch": 2.7731314376265517,
465
+ "grad_norm": 1.578857660293579,
466
+ "learning_rate": 1.5124570824896559e-06,
467
+ "loss": 0.5038,
468
+ "step": 31500
469
+ },
470
+ {
471
+ "epoch": 2.8171493969539574,
472
+ "grad_norm": 4.331024646759033,
473
+ "learning_rate": 1.219004020306952e-06,
474
+ "loss": 0.505,
475
+ "step": 32000
476
+ },
477
+ {
478
+ "epoch": 2.8611673562813627,
479
+ "grad_norm": 1.4179637432098389,
480
+ "learning_rate": 9.25550958124248e-07,
481
+ "loss": 0.5041,
482
+ "step": 32500
483
+ },
484
+ {
485
+ "epoch": 2.9051853156087684,
486
+ "grad_norm": 4.137328147888184,
487
+ "learning_rate": 6.320978959415442e-07,
488
+ "loss": 0.4894,
489
+ "step": 33000
490
+ },
491
+ {
492
+ "epoch": 2.949203274936174,
493
+ "grad_norm": 1.429250717163086,
494
+ "learning_rate": 3.386448337588403e-07,
495
+ "loss": 0.4982,
496
+ "step": 33500
497
+ },
498
+ {
499
+ "epoch": 2.9932212342635793,
500
+ "grad_norm": 3.0007667541503906,
501
+ "learning_rate": 4.519177157613641e-08,
502
+ "loss": 0.5104,
503
+ "step": 34000
504
+ }
505
+ ],
506
+ "logging_steps": 500,
507
+ "max_steps": 34077,
508
+ "num_input_tokens_seen": 0,
509
+ "num_train_epochs": 3,
510
+ "save_steps": 500,
511
+ "stateful_callbacks": {
512
+ "TrainerControl": {
513
+ "args": {
514
+ "should_epoch_stop": false,
515
+ "should_evaluate": false,
516
+ "should_log": false,
517
+ "should_save": true,
518
+ "should_training_stop": true
519
+ },
520
+ "attributes": {}
521
+ }
522
+ },
523
+ "total_flos": 0.0,
524
+ "train_batch_size": 8,
525
+ "trial_name": null,
526
+ "trial_params": null
527
+ }
checkpoint-34077/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2b4c01c40ec2e08be31e8df6930ce5b43d017b587f69cdd80247de2f235be9a
3
+ size 5112