ucmp137538 commited on
Commit
2b8951d
·
verified ·
1 Parent(s): 96539dd

Model save

Browse files
Files changed (4) hide show
  1. README.md +1 -1
  2. all_results.json +6 -6
  3. train_results.json +6 -6
  4. trainer_state.json +766 -1252
README.md CHANGED
@@ -27,7 +27,7 @@ print(output["generated_text"])
27
 
28
  ## Training procedure
29
 
30
- [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/mingzeli/PreThink_MemAgent/runs/umrk1n8k)
31
 
32
 
33
  This model was trained with SFT.
 
27
 
28
  ## Training procedure
29
 
30
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/mingzeli/PreThink_MemAgent/runs/88u8xb90)
31
 
32
 
33
  This model was trained with SFT.
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 2.7972111648056934e+17,
3
- "train_loss": 1.0216455256655104,
4
- "train_runtime": 2780.2712,
5
- "train_samples": 5375,
6
- "train_samples_per_second": 9.666,
7
- "train_steps_per_second": 0.076
8
  }
 
1
  {
2
+ "total_flos": 5.614755170526167e+17,
3
+ "train_loss": 1.1423310605036159,
4
+ "train_runtime": 3470.2952,
5
+ "train_samples": 12431,
6
+ "train_samples_per_second": 10.746,
7
+ "train_steps_per_second": 0.042
8
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 2.7972111648056934e+17,
3
- "train_loss": 1.0216455256655104,
4
- "train_runtime": 2780.2712,
5
- "train_samples": 5375,
6
- "train_samples_per_second": 9.666,
7
- "train_steps_per_second": 0.076
8
  }
 
1
  {
2
+ "total_flos": 5.614755170526167e+17,
3
+ "train_loss": 1.1423310605036159,
4
+ "train_runtime": 3470.2952,
5
+ "train_samples": 12431,
6
+ "train_samples_per_second": 10.746,
7
+ "train_steps_per_second": 0.042
8
  }
trainer_state.json CHANGED
@@ -2,1707 +2,1221 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 5.0,
6
- "eval_steps": 500,
7
- "global_step": 210,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
- "epoch": 0.023809523809523808,
14
- "grad_norm": 6.538216213930027,
15
  "learning_rate": 0.0,
16
- "loss": 1.9338,
17
- "num_tokens": 301499.0,
18
  "step": 1
19
  },
20
  {
21
- "epoch": 0.047619047619047616,
22
- "grad_norm": 6.773412169837184,
23
- "learning_rate": 4.2857142857142855e-06,
24
- "loss": 1.9236,
25
- "num_tokens": 603383.0,
26
  "step": 2
27
  },
28
  {
29
- "epoch": 0.07142857142857142,
30
- "grad_norm": 4.749957858339624,
31
- "learning_rate": 8.571428571428571e-06,
32
- "loss": 1.8766,
33
- "num_tokens": 863622.0,
34
  "step": 3
35
  },
36
  {
37
- "epoch": 0.09523809523809523,
38
- "grad_norm": 2.2238600419377277,
39
- "learning_rate": 1.2857142857142857e-05,
40
- "loss": 1.7506,
41
- "num_tokens": 1179346.0,
42
  "step": 4
43
  },
44
  {
45
- "epoch": 0.11904761904761904,
46
- "grad_norm": 3.7552483574413404,
47
- "learning_rate": 1.7142857142857142e-05,
48
- "loss": 1.7277,
49
- "num_tokens": 1490946.0,
50
  "step": 5
51
  },
52
  {
53
- "epoch": 0.14285714285714285,
54
- "grad_norm": 4.15755928790991,
55
- "learning_rate": 2.1428571428571428e-05,
56
- "loss": 1.658,
57
- "num_tokens": 1719799.0,
58
  "step": 6
59
  },
60
  {
61
- "epoch": 0.16666666666666666,
62
- "grad_norm": 2.245417974821929,
63
- "learning_rate": 2.5714285714285714e-05,
64
- "loss": 1.6608,
65
- "num_tokens": 2009958.0,
66
  "step": 7
67
  },
68
  {
69
- "epoch": 0.19047619047619047,
70
- "grad_norm": 3.1872785546255233,
71
- "learning_rate": 3e-05,
72
- "loss": 1.6032,
73
- "num_tokens": 2334131.0,
74
  "step": 8
75
  },
76
  {
77
- "epoch": 0.21428571428571427,
78
- "grad_norm": 2.062217429493514,
79
- "learning_rate": 2.999838339925525e-05,
80
- "loss": 1.6093,
81
- "num_tokens": 2659726.0,
82
  "step": 9
83
  },
84
  {
85
- "epoch": 0.23809523809523808,
86
- "grad_norm": 30.952811910237916,
87
- "learning_rate": 2.9993533984191048e-05,
88
- "loss": 1.6153,
89
- "num_tokens": 2997515.0,
90
  "step": 10
91
  },
92
  {
93
- "epoch": 0.2619047619047619,
94
- "grad_norm": 5.437630168475735,
95
- "learning_rate": 2.9985452916224895e-05,
96
- "loss": 1.6272,
97
- "num_tokens": 3298571.0,
98
  "step": 11
99
  },
100
  {
101
- "epoch": 0.2857142857142857,
102
- "grad_norm": 9.067784859415095,
103
- "learning_rate": 2.9974142130743523e-05,
104
- "loss": 1.6427,
105
- "num_tokens": 3621523.0,
106
  "step": 12
107
  },
108
  {
109
- "epoch": 0.30952380952380953,
110
- "grad_norm": 8.297445832624902,
111
- "learning_rate": 2.9959604336639406e-05,
112
- "loss": 1.6557,
113
- "num_tokens": 3918652.0,
114
  "step": 13
115
  },
116
  {
117
- "epoch": 0.3333333333333333,
118
- "grad_norm": 18.611016423707234,
119
- "learning_rate": 2.9941843015662005e-05,
120
- "loss": 1.6543,
121
- "num_tokens": 4189760.0,
122
  "step": 14
123
  },
124
  {
125
- "epoch": 0.35714285714285715,
126
- "grad_norm": 37.588129897999515,
127
- "learning_rate": 2.9920862421583856e-05,
128
- "loss": 1.6303,
129
- "num_tokens": 4543297.0,
130
  "step": 15
131
  },
132
  {
133
- "epoch": 0.38095238095238093,
134
- "grad_norm": 4.123366988000153,
135
- "learning_rate": 2.9896667579181855e-05,
136
- "loss": 1.6285,
137
- "num_tokens": 4829061.0,
138
  "step": 16
139
  },
140
  {
141
- "epoch": 0.40476190476190477,
142
- "grad_norm": 11.611181067813334,
143
- "learning_rate": 2.9869264283033827e-05,
144
- "loss": 1.5812,
145
- "num_tokens": 5089606.0,
146
  "step": 17
147
  },
148
  {
149
- "epoch": 0.42857142857142855,
150
- "grad_norm": 2.2261338342365398,
151
- "learning_rate": 2.9838659096130715e-05,
152
- "loss": 1.5597,
153
- "num_tokens": 5390312.0,
154
  "step": 18
155
  },
156
  {
157
- "epoch": 0.4523809523809524,
158
- "grad_norm": 2.1851134690595377,
159
- "learning_rate": 2.9804859348304814e-05,
160
- "loss": 1.5541,
161
- "num_tokens": 5637391.0,
162
  "step": 19
163
  },
164
  {
165
- "epoch": 0.47619047619047616,
166
- "grad_norm": 19.42274392919098,
167
- "learning_rate": 2.976787313447427e-05,
168
- "loss": 1.5582,
169
- "num_tokens": 5900436.0,
170
  "step": 20
171
  },
172
  {
173
- "epoch": 0.5,
174
- "grad_norm": 2.743718856278709,
175
- "learning_rate": 2.9727709312704392e-05,
176
- "loss": 1.5783,
177
- "num_tokens": 6208557.0,
178
  "step": 21
179
  },
180
  {
181
- "epoch": 0.5238095238095238,
182
- "grad_norm": 1.270567239695732,
183
- "learning_rate": 2.968437750208617e-05,
184
- "loss": 1.571,
185
- "num_tokens": 6508211.0,
186
  "step": 22
187
  },
188
  {
189
- "epoch": 0.5476190476190477,
190
- "grad_norm": 1.3295343234029675,
191
- "learning_rate": 2.963788808043254e-05,
192
- "loss": 1.5558,
193
- "num_tokens": 6815212.0,
194
  "step": 23
195
  },
196
  {
197
- "epoch": 0.5714285714285714,
198
- "grad_norm": 1.0645487974082066,
199
- "learning_rate": 2.9588252181792933e-05,
200
- "loss": 1.5579,
201
- "num_tokens": 7165341.0,
202
  "step": 24
203
  },
204
  {
205
- "epoch": 0.5952380952380952,
206
- "grad_norm": 0.8618665557232967,
207
- "learning_rate": 2.953548169378672e-05,
208
- "loss": 1.5311,
209
- "num_tokens": 7446406.0,
210
  "step": 25
211
  },
212
  {
213
- "epoch": 0.6190476190476191,
214
- "grad_norm": 0.8833172294647831,
215
- "learning_rate": 2.9479589254756145e-05,
216
- "loss": 1.5395,
217
- "num_tokens": 7716401.0,
218
  "step": 26
219
  },
220
  {
221
- "epoch": 0.6428571428571429,
222
- "grad_norm": 0.8730876262810777,
223
- "learning_rate": 2.9420588250739514e-05,
224
- "loss": 1.5083,
225
- "num_tokens": 7991093.0,
226
  "step": 27
227
  },
228
  {
229
- "epoch": 0.6666666666666666,
230
- "grad_norm": 9.830926234059927,
231
- "learning_rate": 2.9358492812265286e-05,
232
- "loss": 1.562,
233
- "num_tokens": 8326807.0,
234
  "step": 28
235
  },
236
  {
237
- "epoch": 0.6904761904761905,
238
- "grad_norm": 3.439648574877625,
239
- "learning_rate": 2.9293317810967836e-05,
240
- "loss": 1.5434,
241
- "num_tokens": 8608746.0,
242
  "step": 29
243
  },
244
  {
245
- "epoch": 0.7142857142857143,
246
- "grad_norm": 1.4403398373980578,
247
- "learning_rate": 2.9225078856025776e-05,
248
- "loss": 1.55,
249
- "num_tokens": 8887444.0,
250
  "step": 30
251
  },
252
  {
253
- "epoch": 0.7380952380952381,
254
- "grad_norm": 0.9256924750760002,
255
- "learning_rate": 2.915379229042361e-05,
256
- "loss": 1.5387,
257
- "num_tokens": 9158119.0,
258
  "step": 31
259
  },
260
  {
261
- "epoch": 0.7619047619047619,
262
- "grad_norm": 0.9693333317012788,
263
- "learning_rate": 2.9079475187037645e-05,
264
- "loss": 1.5525,
265
- "num_tokens": 9455380.0,
266
  "step": 32
267
  },
268
  {
269
- "epoch": 0.7857142857142857,
270
- "grad_norm": 4.418422539715986,
271
- "learning_rate": 2.900214534454708e-05,
272
- "loss": 1.562,
273
- "num_tokens": 9819963.0,
274
  "step": 33
275
  },
276
  {
277
- "epoch": 0.8095238095238095,
278
- "grad_norm": 1.4976653598911862,
279
- "learning_rate": 2.8921821283171333e-05,
280
- "loss": 1.5134,
281
- "num_tokens": 10120795.0,
282
  "step": 34
283
  },
284
  {
285
- "epoch": 0.8333333333333334,
286
- "grad_norm": 0.9155113033240181,
287
- "learning_rate": 2.883852224023446e-05,
288
- "loss": 1.5485,
289
- "num_tokens": 10448963.0,
290
  "step": 35
291
  },
292
  {
293
- "epoch": 0.8571428571428571,
294
- "grad_norm": 1.078789553362349,
295
- "learning_rate": 2.875226816555792e-05,
296
- "loss": 1.5239,
297
- "num_tokens": 10726463.0,
298
  "step": 36
299
  },
300
  {
301
- "epoch": 0.8809523809523809,
302
- "grad_norm": 0.8912072449466584,
303
- "learning_rate": 2.8663079716682657e-05,
304
- "loss": 1.5328,
305
- "num_tokens": 11034561.0,
306
  "step": 37
307
  },
308
  {
309
- "epoch": 0.9047619047619048,
310
- "grad_norm": 0.8137040759182133,
311
- "learning_rate": 2.8570978253921695e-05,
312
- "loss": 1.5173,
313
- "num_tokens": 11357514.0,
314
  "step": 38
315
  },
316
  {
317
- "epoch": 0.9285714285714286,
318
- "grad_norm": 0.9066560545161888,
319
- "learning_rate": 2.8475985835244395e-05,
320
- "loss": 1.5032,
321
- "num_tokens": 11618431.0,
322
  "step": 39
323
  },
324
  {
325
- "epoch": 0.9523809523809523,
326
- "grad_norm": 0.789402718151679,
327
- "learning_rate": 2.8378125210993705e-05,
328
- "loss": 1.5219,
329
- "num_tokens": 11938221.0,
330
  "step": 40
331
  },
332
  {
333
- "epoch": 0.9761904761904762,
334
- "grad_norm": 0.984207528765468,
335
- "learning_rate": 2.8277419818437477e-05,
336
- "loss": 1.536,
337
- "num_tokens": 12260374.0,
338
  "step": 41
339
  },
340
  {
341
- "epoch": 1.0,
342
- "grad_norm": 0.7141475421535004,
343
- "learning_rate": 2.817389377615535e-05,
344
- "loss": 1.5138,
345
- "num_tokens": 12557020.0,
346
  "step": 42
347
  },
348
  {
349
- "epoch": 1.0238095238095237,
350
- "grad_norm": 0.9363198060842514,
351
- "learning_rate": 2.8067571878262455e-05,
352
- "loss": 1.3474,
353
- "num_tokens": 12816238.0,
354
  "step": 43
355
  },
356
  {
357
- "epoch": 1.0476190476190477,
358
- "grad_norm": 0.6926582015960785,
359
- "learning_rate": 2.7958479588471274e-05,
360
- "loss": 1.3267,
361
- "num_tokens": 13122015.0,
362
  "step": 44
363
  },
364
  {
365
- "epoch": 1.0714285714285714,
366
- "grad_norm": 0.9454215939995514,
367
- "learning_rate": 2.784664303399321e-05,
368
- "loss": 1.339,
369
- "num_tokens": 13397622.0,
370
  "step": 45
371
  },
372
  {
373
- "epoch": 1.0952380952380953,
374
- "grad_norm": 0.7721777219877292,
375
- "learning_rate": 2.7732088999281182e-05,
376
- "loss": 1.3203,
377
- "num_tokens": 13699593.0,
378
  "step": 46
379
  },
380
  {
381
- "epoch": 1.119047619047619,
382
- "grad_norm": 0.8317013732600135,
383
- "learning_rate": 2.7614844919614878e-05,
384
- "loss": 1.3393,
385
- "num_tokens": 14012389.0,
386
  "step": 47
387
  },
388
  {
389
- "epoch": 1.1428571428571428,
390
- "grad_norm": 0.8420079333202011,
391
- "learning_rate": 2.749493887453007e-05,
392
- "loss": 1.321,
393
- "num_tokens": 14310791.0,
394
  "step": 48
395
  },
396
  {
397
- "epoch": 1.1666666666666667,
398
- "grad_norm": 0.808825236941867,
399
- "learning_rate": 2.7372399581093692e-05,
400
- "loss": 1.3444,
401
- "num_tokens": 14658237.0,
402
  "step": 49
403
  },
404
  {
405
- "epoch": 1.1904761904761905,
406
- "grad_norm": 0.9423947902880564,
407
- "learning_rate": 2.724725638702619e-05,
408
- "loss": 1.3208,
409
- "num_tokens": 14932524.0,
410
  "step": 50
411
  },
412
  {
413
- "epoch": 1.2142857142857142,
414
- "grad_norm": 0.8920368672000416,
415
- "learning_rate": 2.7119539263672863e-05,
416
- "loss": 1.3663,
417
- "num_tokens": 15221313.0,
 
 
 
 
 
 
 
 
 
418
  "step": 51
419
  },
420
  {
421
- "epoch": 1.2380952380952381,
422
- "grad_norm": 1.037715092791166,
423
- "learning_rate": 2.698927879882581e-05,
424
- "loss": 1.3133,
425
- "num_tokens": 15503438.0,
426
  "step": 52
427
  },
428
  {
429
- "epoch": 1.2619047619047619,
430
- "grad_norm": 0.9566394653382752,
431
- "learning_rate": 2.6856506189398304e-05,
432
- "loss": 1.3418,
433
- "num_tokens": 15812209.0,
434
  "step": 53
435
  },
436
  {
437
- "epoch": 1.2857142857142856,
438
- "grad_norm": 0.7773982206501678,
439
- "learning_rate": 2.672125323395319e-05,
440
- "loss": 1.3214,
441
- "num_tokens": 16133281.0,
442
  "step": 54
443
  },
444
  {
445
- "epoch": 1.3095238095238095,
446
- "grad_norm": 0.7309052587570527,
447
- "learning_rate": 2.6583552325087277e-05,
448
- "loss": 1.3106,
449
- "num_tokens": 16469546.0,
450
  "step": 55
451
  },
452
  {
453
- "epoch": 1.3333333333333333,
454
- "grad_norm": 0.731975052614095,
455
- "learning_rate": 2.644343644167344e-05,
456
- "loss": 1.315,
457
- "num_tokens": 16740412.0,
458
  "step": 56
459
  },
460
  {
461
- "epoch": 1.3571428571428572,
462
- "grad_norm": 0.888533248470318,
463
- "learning_rate": 2.6300939140962265e-05,
464
- "loss": 1.3347,
465
- "num_tokens": 17069209.0,
466
  "step": 57
467
  },
468
  {
469
- "epoch": 1.380952380952381,
470
- "grad_norm": 0.8174152731954014,
471
- "learning_rate": 2.615609455054523e-05,
472
- "loss": 1.3252,
473
- "num_tokens": 17390230.0,
474
  "step": 58
475
  },
476
  {
477
- "epoch": 1.4047619047619047,
478
- "grad_norm": 0.7703407162710431,
479
- "learning_rate": 2.6008937360181247e-05,
480
- "loss": 1.333,
481
- "num_tokens": 17751056.0,
482
  "step": 59
483
  },
484
  {
485
- "epoch": 1.4285714285714286,
486
- "grad_norm": 0.8093422656169592,
487
- "learning_rate": 2.5859502813488634e-05,
488
- "loss": 1.2688,
489
- "num_tokens": 18029513.0,
490
  "step": 60
491
  },
492
  {
493
- "epoch": 1.4523809523809523,
494
- "grad_norm": 0.7448128748019907,
495
- "learning_rate": 2.570782669950435e-05,
496
- "loss": 1.2866,
497
- "num_tokens": 18342199.0,
498
  "step": 61
499
  },
500
  {
501
- "epoch": 1.4761904761904763,
502
- "grad_norm": 0.797247555706446,
503
- "learning_rate": 2.5553945344112646e-05,
504
- "loss": 1.3474,
505
- "num_tokens": 18654129.0,
506
  "step": 62
507
  },
508
  {
509
- "epoch": 1.5,
510
- "grad_norm": 0.8270748581084111,
511
- "learning_rate": 2.539789560134521e-05,
512
- "loss": 1.3099,
513
- "num_tokens": 18959464.0,
514
  "step": 63
515
  },
516
  {
517
- "epoch": 1.5238095238095237,
518
- "grad_norm": 0.813301315394994,
519
- "learning_rate": 2.5239714844554674e-05,
520
- "loss": 1.3041,
521
- "num_tokens": 19214541.0,
522
  "step": 64
523
  },
524
  {
525
- "epoch": 1.5476190476190477,
526
- "grad_norm": 0.951995926855497,
527
- "learning_rate": 2.5079440957463894e-05,
528
- "loss": 1.2476,
529
- "num_tokens": 19491950.0,
530
  "step": 65
531
  },
532
  {
533
- "epoch": 1.5714285714285714,
534
- "grad_norm": 0.7172356466200617,
535
- "learning_rate": 2.4917112325092904e-05,
536
- "loss": 1.2979,
537
- "num_tokens": 19789279.0,
538
  "step": 66
539
  },
540
  {
541
- "epoch": 1.5952380952380953,
542
- "grad_norm": 0.842364452878783,
543
- "learning_rate": 2.475276782456585e-05,
544
- "loss": 1.2964,
545
- "num_tokens": 20050385.0,
546
  "step": 67
547
  },
548
  {
549
- "epoch": 1.619047619047619,
550
- "grad_norm": 0.8775202665218085,
551
- "learning_rate": 2.4586446815800056e-05,
552
- "loss": 1.2306,
553
- "num_tokens": 20293468.0,
554
  "step": 68
555
  },
556
  {
557
- "epoch": 1.6428571428571428,
558
- "grad_norm": 0.8077565121484018,
559
- "learning_rate": 2.441818913207947e-05,
560
- "loss": 1.2124,
561
- "num_tokens": 20548563.0,
562
  "step": 69
563
  },
564
  {
565
- "epoch": 1.6666666666666665,
566
- "grad_norm": 0.7891520078797498,
567
- "learning_rate": 2.4248035070514732e-05,
568
- "loss": 1.31,
569
- "num_tokens": 20899914.0,
570
  "step": 70
571
  },
572
  {
573
- "epoch": 1.6904761904761905,
574
- "grad_norm": 0.8935642914804406,
575
- "learning_rate": 2.4076025382392166e-05,
576
- "loss": 1.2904,
577
- "num_tokens": 21207831.0,
578
  "step": 71
579
  },
580
  {
581
- "epoch": 1.7142857142857144,
582
- "grad_norm": 0.7791700089762551,
583
- "learning_rate": 2.3902201263413968e-05,
584
- "loss": 1.3009,
585
- "num_tokens": 21535523.0,
586
  "step": 72
587
  },
588
  {
589
- "epoch": 1.7380952380952381,
590
- "grad_norm": 0.6803499303828192,
591
- "learning_rate": 2.372660434383203e-05,
592
- "loss": 1.3084,
593
- "num_tokens": 21866131.0,
594
  "step": 73
595
  },
596
  {
597
- "epoch": 1.7619047619047619,
598
- "grad_norm": 0.8712043390870515,
599
- "learning_rate": 2.3549276678477605e-05,
600
- "loss": 1.2974,
601
- "num_tokens": 22201523.0,
602
  "step": 74
603
  },
604
  {
605
- "epoch": 1.7857142857142856,
606
- "grad_norm": 0.8755800517621928,
607
- "learning_rate": 2.337026073668934e-05,
608
- "loss": 1.2914,
609
- "num_tokens": 22488112.0,
610
  "step": 75
611
  },
612
  {
613
- "epoch": 1.8095238095238095,
614
- "grad_norm": 0.6269216466100511,
615
- "learning_rate": 2.3189599392142027e-05,
616
- "loss": 1.3065,
617
- "num_tokens": 22813071.0,
618
  "step": 76
619
  },
620
  {
621
- "epoch": 1.8333333333333335,
622
- "grad_norm": 0.8732441594396502,
623
- "learning_rate": 2.3007335912578496e-05,
624
- "loss": 1.2387,
625
- "num_tokens": 23107173.0,
626
  "step": 77
627
  },
628
  {
629
- "epoch": 1.8571428571428572,
630
- "grad_norm": 0.8321205173092046,
631
- "learning_rate": 2.282351394944717e-05,
632
- "loss": 1.3084,
633
- "num_tokens": 23422105.0,
634
  "step": 78
635
  },
636
  {
637
- "epoch": 1.880952380952381,
638
- "grad_norm": 0.6201826285098905,
639
- "learning_rate": 2.263817752744767e-05,
640
- "loss": 1.2937,
641
- "num_tokens": 23788463.0,
642
  "step": 79
643
  },
644
  {
645
- "epoch": 1.9047619047619047,
646
- "grad_norm": 0.864663931903862,
647
- "learning_rate": 2.2451371033987086e-05,
648
- "loss": 1.2315,
649
- "num_tokens": 24016057.0,
650
  "step": 80
651
  },
652
  {
653
- "epoch": 1.9285714285714286,
654
- "grad_norm": 0.8151031466504542,
655
- "learning_rate": 2.226313920854934e-05,
656
- "loss": 1.2065,
657
- "num_tokens": 24286443.0,
658
  "step": 81
659
  },
660
  {
661
- "epoch": 1.9523809523809523,
662
- "grad_norm": 0.69377209773933,
663
- "learning_rate": 2.207352713198024e-05,
664
- "loss": 1.2522,
665
- "num_tokens": 24564564.0,
666
  "step": 82
667
  },
668
  {
669
- "epoch": 1.9761904761904763,
670
- "grad_norm": 0.7749074972232748,
671
- "learning_rate": 2.1882580215690765e-05,
672
- "loss": 1.3075,
673
- "num_tokens": 24856761.0,
674
  "step": 83
675
  },
676
  {
677
- "epoch": 2.0,
678
- "grad_norm": 0.836947163785287,
679
- "learning_rate": 2.169034419078124e-05,
680
- "loss": 1.2472,
681
- "num_tokens": 25119155.0,
682
  "step": 84
683
  },
684
  {
685
- "epoch": 2.0238095238095237,
686
- "grad_norm": 1.302298749181937,
687
- "learning_rate": 2.1496865097088846e-05,
688
- "loss": 1.0524,
689
- "num_tokens": 25461378.0,
690
  "step": 85
691
  },
692
  {
693
- "epoch": 2.0476190476190474,
694
- "grad_norm": 0.9938997123428022,
695
- "learning_rate": 2.130218927216128e-05,
696
- "loss": 0.9867,
697
- "num_tokens": 25757738.0,
698
  "step": 86
699
  },
700
  {
701
- "epoch": 2.0714285714285716,
702
- "grad_norm": 2.239246853283492,
703
- "learning_rate": 2.110636334015907e-05,
704
- "loss": 0.9849,
705
- "num_tokens": 26050090.0,
706
  "step": 87
707
  },
708
  {
709
- "epoch": 2.0952380952380953,
710
- "grad_norm": 2.255240802193634,
711
- "learning_rate": 2.0909434200689264e-05,
712
- "loss": 1.0166,
713
- "num_tokens": 26323746.0,
714
  "step": 88
715
  },
716
  {
717
- "epoch": 2.119047619047619,
718
- "grad_norm": 1.0279577082424196,
719
- "learning_rate": 2.0711449017573122e-05,
720
- "loss": 1.016,
721
- "num_tokens": 26607004.0,
722
  "step": 89
723
  },
724
  {
725
- "epoch": 2.142857142857143,
726
- "grad_norm": 0.8902435973371925,
727
- "learning_rate": 2.0512455207550557e-05,
728
- "loss": 0.9812,
729
- "num_tokens": 26898427.0,
730
  "step": 90
731
  },
732
  {
733
- "epoch": 2.1666666666666665,
734
- "grad_norm": 0.9183593054714583,
735
- "learning_rate": 2.0312500428924023e-05,
736
- "loss": 1.0121,
737
- "num_tokens": 27205357.0,
738
  "step": 91
739
  },
740
  {
741
- "epoch": 2.1904761904761907,
742
- "grad_norm": 0.9376600266290521,
743
- "learning_rate": 2.0111632570144487e-05,
744
- "loss": 0.988,
745
- "num_tokens": 27496694.0,
746
  "step": 92
747
  },
748
  {
749
- "epoch": 2.2142857142857144,
750
- "grad_norm": 0.7220165463532561,
751
- "learning_rate": 1.9909899738342324e-05,
752
- "loss": 1.0119,
753
- "num_tokens": 27794697.0,
754
  "step": 93
755
  },
756
  {
757
- "epoch": 2.238095238095238,
758
- "grad_norm": 0.9053661036106623,
759
- "learning_rate": 1.9707350247805862e-05,
760
- "loss": 1.0194,
761
- "num_tokens": 28115378.0,
762
  "step": 94
763
  },
764
  {
765
- "epoch": 2.261904761904762,
766
- "grad_norm": 0.8729413169116081,
767
- "learning_rate": 1.9504032608410243e-05,
768
- "loss": 0.9866,
769
- "num_tokens": 28404844.0,
770
  "step": 95
771
  },
772
  {
773
- "epoch": 2.2857142857142856,
774
- "grad_norm": 0.7627072905875467,
775
- "learning_rate": 1.9299995513999512e-05,
776
- "loss": 1.0105,
777
- "num_tokens": 28717491.0,
778
  "step": 96
779
  },
780
  {
781
- "epoch": 2.3095238095238093,
782
- "grad_norm": 0.7347382833732581,
783
- "learning_rate": 1.90952878307246e-05,
784
- "loss": 0.9667,
785
- "num_tokens": 28969175.0,
786
  "step": 97
787
  },
788
  {
789
- "epoch": 2.3333333333333335,
790
- "grad_norm": 0.8054230888254635,
791
- "learning_rate": 1.888995858534005e-05,
792
- "loss": 1.0089,
793
- "num_tokens": 29278525.0,
794
  "step": 98
795
  },
796
  {
797
- "epoch": 2.357142857142857,
798
- "grad_norm": 0.7107149711009891,
799
- "learning_rate": 1.8684056953462328e-05,
800
- "loss": 0.9881,
801
- "num_tokens": 29589990.0,
802
  "step": 99
803
  },
804
  {
805
- "epoch": 2.380952380952381,
806
- "grad_norm": 0.709536600984241,
807
- "learning_rate": 1.8477632247792365e-05,
808
- "loss": 0.9676,
809
- "num_tokens": 29851176.0,
810
  "step": 100
811
  },
812
  {
813
- "epoch": 2.4047619047619047,
814
- "grad_norm": 0.7059346761233823,
815
- "learning_rate": 1.827073390630542e-05,
816
- "loss": 0.9866,
817
- "num_tokens": 30150417.0,
818
  "step": 101
819
  },
820
  {
821
- "epoch": 2.4285714285714284,
822
- "grad_norm": 0.6475677168167608,
823
- "learning_rate": 1.806341148041082e-05,
824
- "loss": 0.9679,
825
- "num_tokens": 30476868.0,
826
  "step": 102
827
  },
828
  {
829
- "epoch": 2.4523809523809526,
830
- "grad_norm": 0.6574153556079124,
831
- "learning_rate": 1.785571462308458e-05,
832
- "loss": 0.961,
833
- "num_tokens": 30720675.0,
 
 
 
 
 
 
 
 
 
834
  "step": 103
835
  },
836
  {
837
- "epoch": 2.4761904761904763,
838
- "grad_norm": 0.782684876535666,
839
- "learning_rate": 1.764769307697769e-05,
840
- "loss": 0.9449,
841
- "num_tokens": 30943687.0,
842
  "step": 104
843
  },
844
  {
845
- "epoch": 2.5,
846
- "grad_norm": 0.6959657002281912,
847
- "learning_rate": 1.7439396662502947e-05,
848
- "loss": 0.9588,
849
- "num_tokens": 31226585.0,
850
  "step": 105
851
  },
852
  {
853
- "epoch": 2.5238095238095237,
854
- "grad_norm": 0.6638836519525753,
855
- "learning_rate": 1.723087526590314e-05,
856
- "loss": 0.9657,
857
- "num_tokens": 31536165.0,
858
  "step": 106
859
  },
860
  {
861
- "epoch": 2.5476190476190474,
862
- "grad_norm": 0.5543755171176452,
863
- "learning_rate": 1.702217882730345e-05,
864
- "loss": 0.9605,
865
- "num_tokens": 31859352.0,
866
  "step": 107
867
  },
868
  {
869
- "epoch": 2.571428571428571,
870
- "grad_norm": 0.7519329941062585,
871
- "learning_rate": 1.6813357328751003e-05,
872
- "loss": 0.951,
873
- "num_tokens": 32141589.0,
874
  "step": 108
875
  },
876
  {
877
- "epoch": 2.5952380952380953,
878
- "grad_norm": 0.6384580732849452,
879
- "learning_rate": 1.660446078224433e-05,
880
- "loss": 0.9641,
881
- "num_tokens": 32415823.0,
882
  "step": 109
883
  },
884
  {
885
- "epoch": 2.619047619047619,
886
- "grad_norm": 0.6375498198212793,
887
- "learning_rate": 1.6395539217755673e-05,
888
- "loss": 0.9767,
889
- "num_tokens": 32738959.0,
890
  "step": 110
891
  },
892
  {
893
- "epoch": 2.642857142857143,
894
- "grad_norm": 0.5909380929927559,
895
- "learning_rate": 1.6186642671249e-05,
896
- "loss": 0.9553,
897
- "num_tokens": 33046758.0,
898
  "step": 111
899
  },
900
  {
901
- "epoch": 2.6666666666666665,
902
- "grad_norm": 0.6627084816829975,
903
- "learning_rate": 1.597782117269655e-05,
904
- "loss": 0.9918,
905
- "num_tokens": 33356268.0,
906
  "step": 112
907
  },
908
  {
909
- "epoch": 2.6904761904761907,
910
- "grad_norm": 0.6209028590481098,
911
- "learning_rate": 1.5769124734096862e-05,
912
- "loss": 0.9946,
913
- "num_tokens": 33669521.0,
914
  "step": 113
915
  },
916
  {
917
- "epoch": 2.7142857142857144,
918
- "grad_norm": 0.5682420553998002,
919
- "learning_rate": 1.5560603337497055e-05,
920
- "loss": 0.9809,
921
- "num_tokens": 34006046.0,
922
  "step": 114
923
  },
924
  {
925
- "epoch": 2.738095238095238,
926
- "grad_norm": 0.6102593452284443,
927
- "learning_rate": 1.5352306923022314e-05,
928
- "loss": 0.9949,
929
- "num_tokens": 34329314.0,
930
  "step": 115
931
  },
932
  {
933
- "epoch": 2.761904761904762,
934
- "grad_norm": 0.5963721089952566,
935
- "learning_rate": 1.5144285376915424e-05,
936
- "loss": 0.9496,
937
- "num_tokens": 34649593.0,
938
  "step": 116
939
  },
940
  {
941
- "epoch": 2.7857142857142856,
942
- "grad_norm": 0.6334479812477452,
943
- "learning_rate": 1.4936588519589182e-05,
944
- "loss": 0.9865,
945
- "num_tokens": 34968486.0,
946
  "step": 117
947
  },
948
  {
949
- "epoch": 2.8095238095238093,
950
- "grad_norm": 0.579179313504508,
951
- "learning_rate": 1.4729266093694578e-05,
952
- "loss": 1.0112,
953
- "num_tokens": 35335674.0,
954
  "step": 118
955
  },
956
  {
957
- "epoch": 2.8333333333333335,
958
- "grad_norm": 0.6547466538591331,
959
- "learning_rate": 1.4522367752207636e-05,
960
- "loss": 0.9476,
961
- "num_tokens": 35620162.0,
962
  "step": 119
963
  },
964
  {
965
- "epoch": 2.857142857142857,
966
- "grad_norm": 0.6519356419157716,
967
- "learning_rate": 1.4315943046537676e-05,
968
- "loss": 0.9574,
969
- "num_tokens": 35888853.0,
970
  "step": 120
971
  },
972
  {
973
- "epoch": 2.880952380952381,
974
- "grad_norm": 0.6056113333547287,
975
- "learning_rate": 1.411004141465995e-05,
976
- "loss": 0.9453,
977
- "num_tokens": 36190378.0,
978
  "step": 121
979
  },
980
  {
981
- "epoch": 2.9047619047619047,
982
- "grad_norm": 0.5999985950614662,
983
- "learning_rate": 1.3904712169275403e-05,
984
- "loss": 0.9878,
985
- "num_tokens": 36499565.0,
986
  "step": 122
987
  },
988
  {
989
- "epoch": 2.928571428571429,
990
- "grad_norm": 0.6170461063881166,
991
- "learning_rate": 1.3700004486000488e-05,
992
- "loss": 1.003,
993
- "num_tokens": 36819385.0,
994
  "step": 123
995
  },
996
  {
997
- "epoch": 2.9523809523809526,
998
- "grad_norm": 0.5609993983165859,
999
- "learning_rate": 1.3495967391589758e-05,
1000
- "loss": 0.9709,
1001
- "num_tokens": 37126003.0,
1002
  "step": 124
1003
  },
1004
  {
1005
- "epoch": 2.9761904761904763,
1006
- "grad_norm": 0.686886397394039,
1007
- "learning_rate": 1.3292649752194144e-05,
1008
- "loss": 0.9514,
1009
- "num_tokens": 37385037.0,
1010
  "step": 125
1011
  },
1012
  {
1013
- "epoch": 3.0,
1014
- "grad_norm": 0.6489847174148443,
1015
- "learning_rate": 1.309010026165768e-05,
1016
- "loss": 0.9919,
1017
- "num_tokens": 37681802.0,
1018
  "step": 126
1019
  },
1020
  {
1021
- "epoch": 3.0238095238095237,
1022
- "grad_norm": 1.3407025275092483,
1023
- "learning_rate": 1.288836742985552e-05,
1024
- "loss": 0.7691,
1025
- "num_tokens": 37984864.0,
1026
  "step": 127
1027
  },
1028
  {
1029
- "epoch": 3.0476190476190474,
1030
- "grad_norm": 1.1240882365427978,
1031
- "learning_rate": 1.2687499571075978e-05,
1032
- "loss": 0.7481,
1033
- "num_tokens": 38255245.0,
1034
  "step": 128
1035
  },
1036
  {
1037
- "epoch": 3.0714285714285716,
1038
- "grad_norm": 0.7591608642049189,
1039
- "learning_rate": 1.2487544792449443e-05,
1040
- "loss": 0.7418,
1041
- "num_tokens": 38525517.0,
1042
  "step": 129
1043
  },
1044
  {
1045
- "epoch": 3.0952380952380953,
1046
- "grad_norm": 1.4687837390641727,
1047
- "learning_rate": 1.2288550982426879e-05,
1048
- "loss": 0.7639,
1049
- "num_tokens": 38849309.0,
1050
  "step": 130
1051
  },
1052
  {
1053
- "epoch": 3.119047619047619,
1054
- "grad_norm": 1.6813633030275168,
1055
- "learning_rate": 1.2090565799310738e-05,
1056
- "loss": 0.7419,
1057
- "num_tokens": 39220621.0,
1058
  "step": 131
1059
  },
1060
  {
1061
- "epoch": 3.142857142857143,
1062
- "grad_norm": 1.5003572460299635,
1063
- "learning_rate": 1.1893636659840927e-05,
1064
- "loss": 0.7561,
1065
- "num_tokens": 39564336.0,
1066
  "step": 132
1067
  },
1068
  {
1069
- "epoch": 3.1666666666666665,
1070
- "grad_norm": 1.0050005240430444,
1071
- "learning_rate": 1.169781072783872e-05,
1072
- "loss": 0.7044,
1073
- "num_tokens": 39847434.0,
1074
  "step": 133
1075
  },
1076
  {
1077
- "epoch": 3.1904761904761907,
1078
- "grad_norm": 0.8969397973329072,
1079
- "learning_rate": 1.1503134902911155e-05,
1080
- "loss": 0.722,
1081
- "num_tokens": 40157495.0,
1082
  "step": 134
1083
  },
1084
  {
1085
- "epoch": 3.2142857142857144,
1086
- "grad_norm": 0.8424389883668345,
1087
- "learning_rate": 1.1309655809218764e-05,
1088
- "loss": 0.7033,
1089
- "num_tokens": 40469966.0,
1090
  "step": 135
1091
  },
1092
  {
1093
- "epoch": 3.238095238095238,
1094
- "grad_norm": 0.8450270760954434,
1095
- "learning_rate": 1.1117419784309236e-05,
1096
- "loss": 0.7047,
1097
- "num_tokens": 40771634.0,
1098
  "step": 136
1099
  },
1100
  {
1101
- "epoch": 3.261904761904762,
1102
- "grad_norm": 0.7538567142823028,
1103
- "learning_rate": 1.0926472868019764e-05,
1104
- "loss": 0.715,
1105
- "num_tokens": 41087014.0,
1106
  "step": 137
1107
  },
1108
  {
1109
- "epoch": 3.2857142857142856,
1110
- "grad_norm": 0.799941449070327,
1111
- "learning_rate": 1.0736860791450658e-05,
1112
- "loss": 0.7477,
1113
- "num_tokens": 41435947.0,
1114
  "step": 138
1115
  },
1116
  {
1117
- "epoch": 3.3095238095238093,
1118
- "grad_norm": 0.821321998423862,
1119
- "learning_rate": 1.0548628966012917e-05,
1120
- "loss": 0.7057,
1121
- "num_tokens": 41733553.0,
1122
  "step": 139
1123
  },
1124
  {
1125
- "epoch": 3.3333333333333335,
1126
- "grad_norm": 0.8039593601908853,
1127
- "learning_rate": 1.036182247255233e-05,
1128
- "loss": 0.6771,
1129
- "num_tokens": 41998508.0,
1130
  "step": 140
1131
  },
1132
  {
1133
- "epoch": 3.357142857142857,
1134
- "grad_norm": 0.7486362398650765,
1135
- "learning_rate": 1.0176486050552834e-05,
1136
- "loss": 0.7025,
1137
- "num_tokens": 42328033.0,
1138
  "step": 141
1139
  },
1140
  {
1141
- "epoch": 3.380952380952381,
1142
- "grad_norm": 0.7435262398434506,
1143
- "learning_rate": 9.99266408742151e-06,
1144
- "loss": 0.6721,
1145
- "num_tokens": 42576679.0,
1146
  "step": 142
1147
  },
1148
  {
1149
- "epoch": 3.4047619047619047,
1150
- "grad_norm": 0.640772240723941,
1151
- "learning_rate": 9.810400607857975e-06,
1152
- "loss": 0.7008,
1153
- "num_tokens": 42881848.0,
1154
  "step": 143
1155
  },
1156
  {
1157
- "epoch": 3.4285714285714284,
1158
- "grad_norm": 0.6750307967979159,
1159
- "learning_rate": 9.629739263310663e-06,
1160
- "loss": 0.6576,
1161
- "num_tokens": 43180370.0,
1162
  "step": 144
1163
  },
1164
  {
1165
- "epoch": 3.4523809523809526,
1166
- "grad_norm": 0.6349593868773542,
1167
- "learning_rate": 9.4507233215224e-06,
1168
- "loss": 0.6895,
1169
- "num_tokens": 43485194.0,
1170
  "step": 145
1171
  },
1172
  {
1173
- "epoch": 3.4761904761904763,
1174
- "grad_norm": 0.7007765296778146,
1175
- "learning_rate": 9.273395656167974e-06,
1176
- "loss": 0.6616,
1177
- "num_tokens": 43735211.0,
1178
  "step": 146
1179
  },
1180
  {
1181
- "epoch": 3.5,
1182
- "grad_norm": 0.6838183617031265,
1183
- "learning_rate": 9.097798736586033e-06,
1184
- "loss": 0.6899,
1185
- "num_tokens": 44018672.0,
1186
  "step": 147
1187
  },
1188
  {
1189
- "epoch": 3.5238095238095237,
1190
- "grad_norm": 0.6632040105335824,
1191
- "learning_rate": 8.92397461760784e-06,
1192
- "loss": 0.6846,
1193
- "num_tokens": 44328901.0,
1194
- "step": 148
1195
- },
1196
- {
1197
- "epoch": 3.5476190476190474,
1198
- "grad_norm": 0.6283627701359418,
1199
- "learning_rate": 8.751964929485264e-06,
1200
- "loss": 0.7087,
1201
- "num_tokens": 44639895.0,
1202
- "step": 149
1203
- },
1204
- {
1205
- "epoch": 3.571428571428571,
1206
- "grad_norm": 0.6949482066022805,
1207
- "learning_rate": 8.58181086792053e-06,
1208
- "loss": 0.7229,
1209
- "num_tokens": 44912674.0,
1210
- "step": 150
1211
- },
1212
- {
1213
- "epoch": 3.5952380952380953,
1214
- "grad_norm": 0.6161930464924805,
1215
- "learning_rate": 8.413553184199947e-06,
1216
- "loss": 0.7318,
1217
- "num_tokens": 45239369.0,
1218
- "step": 151
1219
- },
1220
- {
1221
- "epoch": 3.619047619047619,
1222
- "grad_norm": 0.6775157046640012,
1223
- "learning_rate": 8.247232175434151e-06,
1224
- "loss": 0.69,
1225
- "num_tokens": 45487078.0,
1226
- "step": 152
1227
- },
1228
- {
1229
- "epoch": 3.642857142857143,
1230
- "grad_norm": 0.6598743917563118,
1231
- "learning_rate": 8.082887674907099e-06,
1232
- "loss": 0.6823,
1233
- "num_tokens": 45764060.0,
1234
- "step": 153
1235
- },
1236
- {
1237
- "epoch": 3.6666666666666665,
1238
- "grad_norm": 0.6389058360603095,
1239
- "learning_rate": 7.92055904253611e-06,
1240
- "loss": 0.7304,
1241
- "num_tokens": 46055608.0,
1242
- "step": 154
1243
- },
1244
- {
1245
- "epoch": 3.6904761904761907,
1246
- "grad_norm": 0.6077489003813343,
1247
- "learning_rate": 7.760285155445328e-06,
1248
- "loss": 0.7093,
1249
- "num_tokens": 46359953.0,
1250
- "step": 155
1251
- },
1252
- {
1253
- "epoch": 3.7142857142857144,
1254
- "grad_norm": 0.5967397524054626,
1255
- "learning_rate": 7.602104398654793e-06,
1256
- "loss": 0.7066,
1257
- "num_tokens": 46668754.0,
1258
- "step": 156
1259
- },
1260
- {
1261
- "epoch": 3.738095238095238,
1262
- "grad_norm": 0.6350745338711014,
1263
- "learning_rate": 7.446054655887351e-06,
1264
- "loss": 0.7024,
1265
- "num_tokens": 46984092.0,
1266
- "step": 157
1267
- },
1268
- {
1269
- "epoch": 3.761904761904762,
1270
- "grad_norm": 0.654479549343766,
1271
- "learning_rate": 7.292173300495655e-06,
1272
- "loss": 0.7024,
1273
- "num_tokens": 47279370.0,
1274
- "step": 158
1275
- },
1276
- {
1277
- "epoch": 3.7857142857142856,
1278
- "grad_norm": 0.6241512933913473,
1279
- "learning_rate": 7.140497186511365e-06,
1280
- "loss": 0.6782,
1281
- "num_tokens": 47607126.0,
1282
- "step": 159
1283
- },
1284
- {
1285
- "epoch": 3.8095238095238093,
1286
- "grad_norm": 0.6892381674511956,
1287
- "learning_rate": 6.99106263981875e-06,
1288
- "loss": 0.6708,
1289
- "num_tokens": 47854285.0,
1290
- "step": 160
1291
- },
1292
- {
1293
- "epoch": 3.8333333333333335,
1294
- "grad_norm": 0.6054186909041117,
1295
- "learning_rate": 6.843905449454775e-06,
1296
- "loss": 0.7098,
1297
- "num_tokens": 48172146.0,
1298
- "step": 161
1299
- },
1300
- {
1301
- "epoch": 3.857142857142857,
1302
- "grad_norm": 0.5956432849446107,
1303
- "learning_rate": 6.699060859037737e-06,
1304
- "loss": 0.7313,
1305
- "num_tokens": 48512170.0,
1306
- "step": 162
1307
- },
1308
- {
1309
- "epoch": 3.880952380952381,
1310
- "grad_norm": 0.6733062415864223,
1311
- "learning_rate": 6.556563558326561e-06,
1312
- "loss": 0.7001,
1313
- "num_tokens": 48798674.0,
1314
- "step": 163
1315
- },
1316
- {
1317
- "epoch": 3.9047619047619047,
1318
- "grad_norm": 0.6484125556278886,
1319
- "learning_rate": 6.416447674912726e-06,
1320
- "loss": 0.7064,
1321
- "num_tokens": 49078950.0,
1322
- "step": 164
1323
- },
1324
- {
1325
- "epoch": 3.928571428571429,
1326
- "grad_norm": 0.6651238762004111,
1327
- "learning_rate": 6.278746766046815e-06,
1328
- "loss": 0.6613,
1329
- "num_tokens": 49321513.0,
1330
- "step": 165
1331
- },
1332
- {
1333
- "epoch": 3.9523809523809526,
1334
- "grad_norm": 0.6123780278745324,
1335
- "learning_rate": 6.143493810601696e-06,
1336
- "loss": 0.6737,
1337
- "num_tokens": 49622007.0,
1338
- "step": 166
1339
- },
1340
- {
1341
- "epoch": 3.9761904761904763,
1342
- "grad_norm": 0.6147575021892541,
1343
- "learning_rate": 6.010721201174187e-06,
1344
- "loss": 0.7213,
1345
- "num_tokens": 49949396.0,
1346
- "step": 167
1347
- },
1348
- {
1349
- "epoch": 4.0,
1350
- "grad_norm": 0.6525996500721447,
1351
- "learning_rate": 5.880460736327139e-06,
1352
- "loss": 0.6751,
1353
- "num_tokens": 50240879.0,
1354
- "step": 168
1355
- },
1356
- {
1357
- "epoch": 4.023809523809524,
1358
- "grad_norm": 1.1945930661064657,
1359
- "learning_rate": 5.7527436129738094e-06,
1360
- "loss": 0.5781,
1361
- "num_tokens": 50522550.0,
1362
- "step": 169
1363
- },
1364
- {
1365
- "epoch": 4.0476190476190474,
1366
- "grad_norm": 1.0060804692820782,
1367
- "learning_rate": 5.62760041890631e-06,
1368
- "loss": 0.5498,
1369
- "num_tokens": 50813206.0,
1370
- "step": 170
1371
- },
1372
- {
1373
- "epoch": 4.071428571428571,
1374
- "grad_norm": 0.751291910016424,
1375
- "learning_rate": 5.505061125469931e-06,
1376
- "loss": 0.5308,
1377
- "num_tokens": 51062794.0,
1378
- "step": 171
1379
- },
1380
- {
1381
- "epoch": 4.095238095238095,
1382
- "grad_norm": 0.7527801055673746,
1383
- "learning_rate": 5.385155080385125e-06,
1384
- "loss": 0.5223,
1385
- "num_tokens": 51335547.0,
1386
- "step": 172
1387
- },
1388
- {
1389
- "epoch": 4.119047619047619,
1390
- "grad_norm": 0.9363987669688705,
1391
- "learning_rate": 5.267911000718819e-06,
1392
- "loss": 0.5258,
1393
- "num_tokens": 51637656.0,
1394
- "step": 173
1395
- },
1396
- {
1397
- "epoch": 4.142857142857143,
1398
- "grad_norm": 1.1015381943904987,
1399
- "learning_rate": 5.153356966006791e-06,
1400
- "loss": 0.5196,
1401
- "num_tokens": 51953140.0,
1402
- "step": 174
1403
- },
1404
- {
1405
- "epoch": 4.166666666666667,
1406
- "grad_norm": 1.0393498678201005,
1407
- "learning_rate": 5.041520411528727e-06,
1408
- "loss": 0.55,
1409
- "num_tokens": 52262499.0,
1410
- "step": 175
1411
- },
1412
- {
1413
- "epoch": 4.190476190476191,
1414
- "grad_norm": 0.8250332666446762,
1415
- "learning_rate": 4.932428121737547e-06,
1416
- "loss": 0.5226,
1417
- "num_tokens": 52568631.0,
1418
- "step": 176
1419
- },
1420
- {
1421
- "epoch": 4.214285714285714,
1422
- "grad_norm": 0.7655426183040523,
1423
- "learning_rate": 4.826106223844648e-06,
1424
- "loss": 0.5127,
1425
- "num_tokens": 52834140.0,
1426
- "step": 177
1427
- },
1428
- {
1429
- "epoch": 4.238095238095238,
1430
- "grad_norm": 0.6995684804290214,
1431
- "learning_rate": 4.722580181562527e-06,
1432
- "loss": 0.494,
1433
- "num_tokens": 53156720.0,
1434
- "step": 178
1435
- },
1436
- {
1437
- "epoch": 4.261904761904762,
1438
- "grad_norm": 0.7748083731595724,
1439
- "learning_rate": 4.621874789006295e-06,
1440
- "loss": 0.5381,
1441
- "num_tokens": 53491051.0,
1442
- "step": 179
1443
- },
1444
- {
1445
- "epoch": 4.285714285714286,
1446
- "grad_norm": 0.8131952113654631,
1447
- "learning_rate": 4.524014164755603e-06,
1448
- "loss": 0.4968,
1449
- "num_tokens": 53787155.0,
1450
- "step": 180
1451
- },
1452
- {
1453
- "epoch": 4.309523809523809,
1454
- "grad_norm": 0.7367042370774411,
1455
- "learning_rate": 4.429021746078311e-06,
1456
- "loss": 0.4861,
1457
- "num_tokens": 54049978.0,
1458
- "step": 181
1459
- },
1460
- {
1461
- "epoch": 4.333333333333333,
1462
- "grad_norm": 0.7194704571250619,
1463
- "learning_rate": 4.336920283317344e-06,
1464
- "loss": 0.5212,
1465
- "num_tokens": 54305304.0,
1466
- "step": 182
1467
- },
1468
- {
1469
- "epoch": 4.357142857142857,
1470
- "grad_norm": 0.6527660461115374,
1471
- "learning_rate": 4.247731834442083e-06,
1472
- "loss": 0.5263,
1473
- "num_tokens": 54632438.0,
1474
- "step": 183
1475
- },
1476
- {
1477
- "epoch": 4.380952380952381,
1478
- "grad_norm": 0.7266514800309486,
1479
- "learning_rate": 4.161477759765542e-06,
1480
- "loss": 0.5202,
1481
- "num_tokens": 54954887.0,
1482
- "step": 184
1483
- },
1484
- {
1485
- "epoch": 4.404761904761905,
1486
- "grad_norm": 0.761296256236184,
1487
- "learning_rate": 4.078178716828666e-06,
1488
- "loss": 0.5379,
1489
- "num_tokens": 55245795.0,
1490
- "step": 185
1491
- },
1492
- {
1493
- "epoch": 4.428571428571429,
1494
- "grad_norm": 0.7945856951237223,
1495
- "learning_rate": 3.9978546554529185e-06,
1496
- "loss": 0.5176,
1497
- "num_tokens": 55546573.0,
1498
- "step": 186
1499
- },
1500
- {
1501
- "epoch": 4.4523809523809526,
1502
- "grad_norm": 0.7468660256983087,
1503
- "learning_rate": 3.920524812962359e-06,
1504
- "loss": 0.5096,
1505
- "num_tokens": 55843166.0,
1506
- "step": 187
1507
- },
1508
- {
1509
- "epoch": 4.476190476190476,
1510
- "grad_norm": 0.7193745095120948,
1511
- "learning_rate": 3.846207709576389e-06,
1512
- "loss": 0.5205,
1513
- "num_tokens": 56138218.0,
1514
- "step": 188
1515
- },
1516
- {
1517
- "epoch": 4.5,
1518
- "grad_norm": 0.6655820277052442,
1519
- "learning_rate": 3.7749211439742257e-06,
1520
- "loss": 0.5041,
1521
- "num_tokens": 56417291.0,
1522
- "step": 189
1523
- },
1524
- {
1525
- "epoch": 4.523809523809524,
1526
- "grad_norm": 0.7239518179183401,
1527
- "learning_rate": 3.7066821890321687e-06,
1528
- "loss": 0.4846,
1529
- "num_tokens": 56692919.0,
1530
- "step": 190
1531
- },
1532
- {
1533
- "epoch": 4.5476190476190474,
1534
- "grad_norm": 0.6477616515263318,
1535
- "learning_rate": 3.6415071877347165e-06,
1536
- "loss": 0.539,
1537
- "num_tokens": 57031682.0,
1538
- "step": 191
1539
- },
1540
- {
1541
- "epoch": 4.571428571428571,
1542
- "grad_norm": 0.6640615854897481,
1543
- "learning_rate": 3.579411749260486e-06,
1544
- "loss": 0.4944,
1545
- "num_tokens": 57344445.0,
1546
- "step": 192
1547
- },
1548
- {
1549
- "epoch": 4.595238095238095,
1550
- "grad_norm": 0.6848846128939778,
1551
- "learning_rate": 3.520410745243858e-06,
1552
- "loss": 0.5121,
1553
- "num_tokens": 57613210.0,
1554
- "step": 193
1555
- },
1556
- {
1557
- "epoch": 4.619047619047619,
1558
- "grad_norm": 0.6422019750084208,
1559
- "learning_rate": 3.4645183062132826e-06,
1560
- "loss": 0.5193,
1561
- "num_tokens": 57942367.0,
1562
- "step": 194
1563
- },
1564
- {
1565
- "epoch": 4.642857142857143,
1566
- "grad_norm": 0.7657789249156284,
1567
- "learning_rate": 3.411747818207066e-06,
1568
- "loss": 0.4979,
1569
- "num_tokens": 58201641.0,
1570
- "step": 195
1571
- },
1572
- {
1573
- "epoch": 4.666666666666667,
1574
- "grad_norm": 0.6739307543671282,
1575
- "learning_rate": 3.3621119195674597e-06,
1576
- "loss": 0.4888,
1577
- "num_tokens": 58517521.0,
1578
- "step": 196
1579
- },
1580
- {
1581
- "epoch": 4.690476190476191,
1582
- "grad_norm": 0.6865319310088059,
1583
- "learning_rate": 3.3156224979138306e-06,
1584
- "loss": 0.494,
1585
- "num_tokens": 58816395.0,
1586
- "step": 197
1587
- },
1588
- {
1589
- "epoch": 4.714285714285714,
1590
- "grad_norm": 0.6584126661804732,
1591
- "learning_rate": 3.2722906872956088e-06,
1592
- "loss": 0.5181,
1593
- "num_tokens": 59120541.0,
1594
- "step": 198
1595
- },
1596
- {
1597
- "epoch": 4.738095238095238,
1598
- "grad_norm": 0.6099916205741043,
1599
- "learning_rate": 3.232126865525731e-06,
1600
- "loss": 0.5293,
1601
- "num_tokens": 59461204.0,
1602
- "step": 199
1603
- },
1604
- {
1605
- "epoch": 4.761904761904762,
1606
- "grad_norm": 0.667910578763116,
1607
- "learning_rate": 3.1951406516951874e-06,
1608
- "loss": 0.5116,
1609
- "num_tokens": 59751334.0,
1610
- "step": 200
1611
- },
1612
- {
1613
- "epoch": 4.785714285714286,
1614
- "grad_norm": 0.6141509992681458,
1615
- "learning_rate": 3.1613409038692853e-06,
1616
- "loss": 0.5124,
1617
- "num_tokens": 60077027.0,
1618
- "step": 201
1619
- },
1620
- {
1621
- "epoch": 4.809523809523809,
1622
- "grad_norm": 0.6302007739092258,
1623
- "learning_rate": 3.130735716966174e-06,
1624
- "loss": 0.5405,
1625
- "num_tokens": 60408543.0,
1626
- "step": 202
1627
- },
1628
- {
1629
- "epoch": 4.833333333333333,
1630
- "grad_norm": 0.6664208365053094,
1631
- "learning_rate": 3.1033324208181422e-06,
1632
- "loss": 0.5142,
1633
- "num_tokens": 60702718.0,
1634
- "step": 203
1635
- },
1636
- {
1637
- "epoch": 4.857142857142857,
1638
- "grad_norm": 0.6536331401194854,
1639
- "learning_rate": 3.0791375784161456e-06,
1640
- "loss": 0.5193,
1641
- "num_tokens": 61018938.0,
1642
- "step": 204
1643
- },
1644
- {
1645
- "epoch": 4.880952380952381,
1646
- "grad_norm": 0.6969216972121804,
1647
- "learning_rate": 3.058156984337999e-06,
1648
- "loss": 0.5061,
1649
- "num_tokens": 61279305.0,
1650
- "step": 205
1651
- },
1652
- {
1653
- "epoch": 4.904761904761905,
1654
- "grad_norm": 0.6535453587519624,
1655
- "learning_rate": 3.0403956633605944e-06,
1656
- "loss": 0.4729,
1657
- "num_tokens": 61565633.0,
1658
- "step": 206
1659
- },
1660
- {
1661
- "epoch": 4.928571428571429,
1662
- "grad_norm": 0.6753649990944919,
1663
- "learning_rate": 3.0258578692564804e-06,
1664
- "loss": 0.4772,
1665
- "num_tokens": 61841335.0,
1666
- "step": 207
1667
- },
1668
- {
1669
- "epoch": 4.9523809523809526,
1670
- "grad_norm": 0.6358980505171887,
1671
- "learning_rate": 3.014547083775106e-06,
1672
- "loss": 0.5398,
1673
- "num_tokens": 62169402.0,
1674
- "step": 208
1675
- },
1676
- {
1677
- "epoch": 4.976190476190476,
1678
- "grad_norm": 0.5805504932681311,
1679
- "learning_rate": 3.0064660158089516e-06,
1680
- "loss": 0.524,
1681
- "num_tokens": 62538856.0,
1682
- "step": 209
1683
- },
1684
- {
1685
- "epoch": 5.0,
1686
- "grad_norm": 0.6817532772499482,
1687
- "learning_rate": 3.001616600744755e-06,
1688
- "loss": 0.5121,
1689
- "num_tokens": 62797894.0,
1690
- "step": 210
1691
- },
1692
- {
1693
- "epoch": 5.0,
1694
- "step": 210,
1695
- "total_flos": 2.7972111648056934e+17,
1696
- "train_loss": 1.0216455256655104,
1697
- "train_runtime": 2780.2712,
1698
- "train_samples_per_second": 9.666,
1699
- "train_steps_per_second": 0.076
1700
  }
1701
  ],
1702
  "logging_steps": 1,
1703
- "max_steps": 210,
1704
  "num_input_tokens_seen": 0,
1705
- "num_train_epochs": 5,
1706
  "save_steps": 500,
1707
  "stateful_callbacks": {
1708
  "TrainerControl": {
@@ -1716,8 +1230,8 @@
1716
  "attributes": {}
1717
  }
1718
  },
1719
- "total_flos": 2.7972111648056934e+17,
1720
- "train_batch_size": 2,
1721
  "trial_name": null,
1722
  "trial_params": null
1723
  }
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 3.0,
6
+ "eval_steps": 51,
7
+ "global_step": 147,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
+ "epoch": 0.02056555269922879,
14
+ "grad_norm": 7.168717487255555,
15
  "learning_rate": 0.0,
16
+ "loss": 1.6868,
17
+ "num_tokens": 809024.0,
18
  "step": 1
19
  },
20
  {
21
+ "epoch": 0.04113110539845758,
22
+ "grad_norm": 7.291622078201161,
23
+ "learning_rate": 2.0000000000000003e-06,
24
+ "loss": 1.6858,
25
+ "num_tokens": 1592788.0,
26
  "step": 2
27
  },
28
  {
29
+ "epoch": 0.061696658097686374,
30
+ "grad_norm": 6.633388372265764,
31
+ "learning_rate": 4.000000000000001e-06,
32
+ "loss": 1.6593,
33
+ "num_tokens": 2354703.0,
34
  "step": 3
35
  },
36
  {
37
+ "epoch": 0.08226221079691516,
38
+ "grad_norm": 3.5110553987567465,
39
+ "learning_rate": 6e-06,
40
+ "loss": 1.5722,
41
+ "num_tokens": 3143138.0,
42
  "step": 4
43
  },
44
  {
45
+ "epoch": 0.10282776349614396,
46
+ "grad_norm": 12.715495674379813,
47
+ "learning_rate": 8.000000000000001e-06,
48
+ "loss": 1.5493,
49
+ "num_tokens": 3888034.0,
50
  "step": 5
51
  },
52
  {
53
+ "epoch": 0.12339331619537275,
54
+ "grad_norm": 24.03794663222114,
55
+ "learning_rate": 1e-05,
56
+ "loss": 1.5373,
57
+ "num_tokens": 4682804.0,
58
  "step": 6
59
  },
60
  {
61
+ "epoch": 0.14395886889460155,
62
+ "grad_norm": 19.871112801805985,
63
+ "learning_rate": 9.998898745083595e-06,
64
+ "loss": 1.4966,
65
+ "num_tokens": 5508918.0,
66
  "step": 7
67
  },
68
  {
69
+ "epoch": 0.16452442159383032,
70
+ "grad_norm": 12.796838534757285,
71
+ "learning_rate": 9.995595519339882e-06,
72
+ "loss": 1.418,
73
+ "num_tokens": 6259535.0,
74
  "step": 8
75
  },
76
  {
77
+ "epoch": 0.18508997429305912,
78
+ "grad_norm": 5.341991620474575,
79
+ "learning_rate": 9.99009193952157e-06,
80
+ "loss": 1.4099,
81
+ "num_tokens": 6992386.0,
82
  "step": 9
83
  },
84
  {
85
+ "epoch": 0.20565552699228792,
86
+ "grad_norm": 4.762493826344396,
87
+ "learning_rate": 9.982390699337253e-06,
88
+ "loss": 1.3722,
89
+ "num_tokens": 7728349.0,
90
  "step": 10
91
  },
92
  {
93
+ "epoch": 0.2262210796915167,
94
+ "grad_norm": 4.232985415108636,
95
+ "learning_rate": 9.97249556813298e-06,
96
+ "loss": 1.3674,
97
+ "num_tokens": 8421223.0,
98
  "step": 11
99
  },
100
  {
101
+ "epoch": 0.2467866323907455,
102
+ "grad_norm": 4.678412892018652,
103
+ "learning_rate": 9.960411389047366e-06,
104
+ "loss": 1.3579,
105
+ "num_tokens": 9252435.0,
106
  "step": 12
107
  },
108
  {
109
+ "epoch": 0.26735218508997427,
110
+ "grad_norm": 4.981692703295692,
111
+ "learning_rate": 9.946144076641138e-06,
112
+ "loss": 1.3257,
113
+ "num_tokens": 10072798.0,
114
  "step": 13
115
  },
116
  {
117
+ "epoch": 0.2879177377892031,
118
+ "grad_norm": 2.8431033488668582,
119
+ "learning_rate": 9.929700614002265e-06,
120
+ "loss": 1.3241,
121
+ "num_tokens": 10915700.0,
122
  "step": 14
123
  },
124
  {
125
+ "epoch": 0.30848329048843187,
126
+ "grad_norm": 6.237403081050188,
127
+ "learning_rate": 9.911089049328113e-06,
128
+ "loss": 1.2926,
129
+ "num_tokens": 11690475.0,
130
  "step": 15
131
  },
132
  {
133
+ "epoch": 0.32904884318766064,
134
+ "grad_norm": 3.3556394601847654,
135
+ "learning_rate": 9.890318491986282e-06,
136
+ "loss": 1.3084,
137
+ "num_tokens": 12454353.0,
138
  "step": 16
139
  },
140
  {
141
+ "epoch": 0.3496143958868895,
142
+ "grad_norm": 3.4957166279626715,
143
+ "learning_rate": 9.867399108056054e-06,
144
+ "loss": 1.2421,
145
+ "num_tokens": 13239119.0,
146
  "step": 17
147
  },
148
  {
149
+ "epoch": 0.37017994858611825,
150
+ "grad_norm": 2.531289170109121,
151
+ "learning_rate": 9.842342115352647e-06,
152
+ "loss": 1.2823,
153
+ "num_tokens": 14021595.0,
154
  "step": 18
155
  },
156
  {
157
+ "epoch": 0.390745501285347,
158
+ "grad_norm": 1.6632412232515026,
159
+ "learning_rate": 9.815159777936682e-06,
160
+ "loss": 1.2517,
161
+ "num_tokens": 14783289.0,
162
  "step": 19
163
  },
164
  {
165
+ "epoch": 0.41131105398457585,
166
+ "grad_norm": 5.91844570481689,
167
+ "learning_rate": 9.785865400111593e-06,
168
+ "loss": 1.2599,
169
+ "num_tokens": 15628684.0,
170
  "step": 20
171
  },
172
  {
173
+ "epoch": 0.4318766066838046,
174
+ "grad_norm": 7.193576968428877,
175
+ "learning_rate": 9.754473319911866e-06,
176
+ "loss": 1.252,
177
+ "num_tokens": 16396270.0,
178
  "step": 21
179
  },
180
  {
181
+ "epoch": 0.4524421593830334,
182
+ "grad_norm": 5.504834853944556,
183
+ "learning_rate": 9.720998902085354e-06,
184
+ "loss": 1.2151,
185
+ "num_tokens": 17186501.0,
186
  "step": 22
187
  },
188
  {
189
+ "epoch": 0.4730077120822622,
190
+ "grad_norm": 1.264259592512111,
191
+ "learning_rate": 9.685458530573035e-06,
192
+ "loss": 1.2377,
193
+ "num_tokens": 17997203.0,
194
  "step": 23
195
  },
196
  {
197
+ "epoch": 0.493573264781491,
198
+ "grad_norm": 3.713191417559263,
199
+ "learning_rate": 9.647869600489954e-06,
200
+ "loss": 1.2181,
201
+ "num_tokens": 18817970.0,
202
  "step": 24
203
  },
204
  {
205
+ "epoch": 0.5141388174807198,
206
+ "grad_norm": 3.1765726456521626,
207
+ "learning_rate": 9.60825050961125e-06,
208
+ "loss": 1.227,
209
+ "num_tokens": 19633513.0,
210
  "step": 25
211
  },
212
  {
213
+ "epoch": 0.5347043701799485,
214
+ "grad_norm": 1.1060798252537263,
215
+ "learning_rate": 9.566620649367418e-06,
216
+ "loss": 1.2185,
217
+ "num_tokens": 20390710.0,
218
  "step": 26
219
  },
220
  {
221
+ "epoch": 0.5552699228791774,
222
+ "grad_norm": 2.8920027077334307,
223
+ "learning_rate": 9.523000395353242e-06,
224
+ "loss": 1.2012,
225
+ "num_tokens": 21154825.0,
226
  "step": 27
227
  },
228
  {
229
+ "epoch": 0.5758354755784062,
230
+ "grad_norm": 3.0008962834932205,
231
+ "learning_rate": 9.477411097355025e-06,
232
+ "loss": 1.1862,
233
+ "num_tokens": 21973044.0,
234
  "step": 28
235
  },
236
  {
237
+ "epoch": 0.596401028277635,
238
+ "grad_norm": 0.8612249291394044,
239
+ "learning_rate": 9.429875068901018e-06,
240
+ "loss": 1.2349,
241
+ "num_tokens": 22763384.0,
242
  "step": 29
243
  },
244
  {
245
+ "epoch": 0.6169665809768637,
246
+ "grad_norm": 1.5747643898508015,
247
+ "learning_rate": 9.380415576340127e-06,
248
+ "loss": 1.2552,
249
+ "num_tokens": 23624743.0,
250
  "step": 30
251
  },
252
  {
253
+ "epoch": 0.6375321336760925,
254
+ "grad_norm": 1.0439033686337735,
255
+ "learning_rate": 9.329056827454279e-06,
256
+ "loss": 1.2262,
257
+ "num_tokens": 24445122.0,
258
  "step": 31
259
  },
260
  {
261
+ "epoch": 0.6580976863753213,
262
+ "grad_norm": 1.1431587522720466,
263
+ "learning_rate": 9.275823959610019e-06,
264
+ "loss": 1.1802,
265
+ "num_tokens": 25269608.0,
266
  "step": 32
267
  },
268
  {
269
+ "epoch": 0.6786632390745502,
270
+ "grad_norm": 0.9682411879222097,
271
+ "learning_rate": 9.22074302745509e-06,
272
+ "loss": 1.2086,
273
+ "num_tokens": 26077592.0,
274
  "step": 33
275
  },
276
  {
277
+ "epoch": 0.699228791773779,
278
+ "grad_norm": 2.0072489147760595,
279
+ "learning_rate": 9.163840990166085e-06,
280
+ "loss": 1.2066,
281
+ "num_tokens": 26890903.0,
282
  "step": 34
283
  },
284
  {
285
+ "epoch": 0.7197943444730077,
286
+ "grad_norm": 1.4975825224817174,
287
+ "learning_rate": 9.105145698253367e-06,
288
+ "loss": 1.1348,
289
+ "num_tokens": 27707185.0,
290
  "step": 35
291
  },
292
  {
293
+ "epoch": 0.7403598971722365,
294
+ "grad_norm": 0.9880199998455155,
295
+ "learning_rate": 9.044685879929734e-06,
296
+ "loss": 1.2621,
297
+ "num_tokens": 28536664.0,
298
  "step": 36
299
  },
300
  {
301
+ "epoch": 0.7609254498714653,
302
+ "grad_norm": 2.668219422837114,
303
+ "learning_rate": 8.982491127049482e-06,
304
+ "loss": 1.1934,
305
+ "num_tokens": 29362087.0,
306
  "step": 37
307
  },
308
  {
309
+ "epoch": 0.781491002570694,
310
+ "grad_norm": 1.1933259433753431,
311
+ "learning_rate": 8.918591880624783e-06,
312
+ "loss": 1.1959,
313
+ "num_tokens": 30117101.0,
314
  "step": 38
315
  },
316
  {
317
+ "epoch": 0.8020565552699229,
318
+ "grad_norm": 3.820068913066678,
319
+ "learning_rate": 8.85301941592643e-06,
320
+ "loss": 1.1953,
321
+ "num_tokens": 30947725.0,
322
  "step": 39
323
  },
324
  {
325
+ "epoch": 0.8226221079691517,
326
+ "grad_norm": 3.2235005439130355,
327
+ "learning_rate": 8.785805827176256e-06,
328
+ "loss": 1.1483,
329
+ "num_tokens": 31745760.0,
330
  "step": 40
331
  },
332
  {
333
+ "epoch": 0.8431876606683805,
334
+ "grad_norm": 1.3876292035353346,
335
+ "learning_rate": 8.716984011838731e-06,
336
+ "loss": 1.1777,
337
+ "num_tokens": 32569405.0,
338
  "step": 41
339
  },
340
  {
341
+ "epoch": 0.8637532133676092,
342
+ "grad_norm": 1.8637438280292022,
343
+ "learning_rate": 8.646587654519413e-06,
344
+ "loss": 1.1803,
345
+ "num_tokens": 33393264.0,
346
  "step": 42
347
  },
348
  {
349
+ "epoch": 0.884318766066838,
350
+ "grad_norm": 1.195519553673873,
351
+ "learning_rate": 8.574651210478121e-06,
352
+ "loss": 1.211,
353
+ "num_tokens": 34224768.0,
354
  "step": 43
355
  },
356
  {
357
+ "epoch": 0.9048843187660668,
358
+ "grad_norm": 0.8382740981864941,
359
+ "learning_rate": 8.501209888764928e-06,
360
+ "loss": 1.1545,
361
+ "num_tokens": 35047872.0,
362
  "step": 44
363
  },
364
  {
365
+ "epoch": 0.9254498714652957,
366
+ "grad_norm": 1.760524895981456,
367
+ "learning_rate": 8.426299634987217e-06,
368
+ "loss": 1.1532,
369
+ "num_tokens": 35875180.0,
370
  "step": 45
371
  },
372
  {
373
+ "epoch": 0.9460154241645244,
374
+ "grad_norm": 3.878430468083381,
375
+ "learning_rate": 8.349957113716213e-06,
376
+ "loss": 1.1812,
377
+ "num_tokens": 36683234.0,
378
  "step": 46
379
  },
380
  {
381
+ "epoch": 0.9665809768637532,
382
+ "grad_norm": 3.037524297964966,
383
+ "learning_rate": 8.272219690541639e-06,
384
+ "loss": 1.2384,
385
+ "num_tokens": 37448006.0,
386
  "step": 47
387
  },
388
  {
389
+ "epoch": 0.987146529562982,
390
+ "grad_norm": 1.4759778040121787,
391
+ "learning_rate": 8.19312541378326e-06,
392
+ "loss": 1.1638,
393
+ "num_tokens": 38273059.0,
394
  "step": 48
395
  },
396
  {
397
+ "epoch": 1.0,
398
+ "grad_norm": 1.4759778040121787,
399
+ "learning_rate": 8.112712995868244e-06,
400
+ "loss": 1.181,
401
+ "num_tokens": 38793091.0,
402
  "step": 49
403
  },
404
  {
405
+ "epoch": 1.0205655526992288,
406
+ "grad_norm": 3.8386221295253287,
407
+ "learning_rate": 8.031021794383513e-06,
408
+ "loss": 1.1309,
409
+ "num_tokens": 39621652.0,
410
  "step": 50
411
  },
412
  {
413
+ "epoch": 1.0411311053984575,
414
+ "grad_norm": 3.3524575596264032,
415
+ "learning_rate": 7.948091792812296e-06,
416
+ "loss": 1.0887,
417
+ "num_tokens": 40386138.0,
418
+ "step": 51
419
+ },
420
+ {
421
+ "epoch": 1.0411311053984575,
422
+ "eval_loss": 0.9421821236610413,
423
+ "eval_num_tokens": 40386138.0,
424
+ "eval_runtime": 31.2484,
425
+ "eval_samples_per_second": 44.226,
426
+ "eval_steps_per_second": 5.536,
427
  "step": 51
428
  },
429
  {
430
+ "epoch": 1.0616966580976863,
431
+ "grad_norm": 2.5943580288080614,
432
+ "learning_rate": 7.863963580964344e-06,
433
+ "loss": 1.2202,
434
+ "num_tokens": 41267887.0,
435
  "step": 52
436
  },
437
  {
438
+ "epoch": 1.082262210796915,
439
+ "grad_norm": 1.0418412079638837,
440
+ "learning_rate": 7.778678335109399e-06,
441
+ "loss": 1.1243,
442
+ "num_tokens": 42061937.0,
443
  "step": 53
444
  },
445
  {
446
+ "epoch": 1.1028277634961439,
447
+ "grad_norm": 1.7618278256873732,
448
+ "learning_rate": 7.692277797823585e-06,
449
+ "loss": 1.0871,
450
+ "num_tokens": 42839063.0,
451
  "step": 54
452
  },
453
  {
454
+ "epoch": 1.1233933161953726,
455
+ "grad_norm": 1.6933944438474042,
456
+ "learning_rate": 7.6048042575586675e-06,
457
+ "loss": 1.1047,
458
+ "num_tokens": 43584262.0,
459
  "step": 55
460
  },
461
  {
462
+ "epoch": 1.1439588688946016,
463
+ "grad_norm": 1.0074442814131914,
464
+ "learning_rate": 7.516300527944104e-06,
465
+ "loss": 1.0769,
466
+ "num_tokens": 44448960.0,
467
  "step": 56
468
  },
469
  {
470
+ "epoch": 1.1645244215938304,
471
+ "grad_norm": 1.614591216728204,
472
+ "learning_rate": 7.426809926832049e-06,
473
+ "loss": 1.1515,
474
+ "num_tokens": 45212858.0,
475
  "step": 57
476
  },
477
  {
478
+ "epoch": 1.1850899742930592,
479
+ "grad_norm": 1.2719018890548137,
480
+ "learning_rate": 7.336376255095592e-06,
481
+ "loss": 1.1,
482
+ "num_tokens": 45985576.0,
483
  "step": 58
484
  },
485
  {
486
+ "epoch": 1.205655526992288,
487
+ "grad_norm": 0.7493358958267934,
488
+ "learning_rate": 7.24504377519056e-06,
489
+ "loss": 1.1753,
490
+ "num_tokens": 46750204.0,
491
  "step": 59
492
  },
493
  {
494
+ "epoch": 1.2262210796915167,
495
+ "grad_norm": 2.1100937351617874,
496
+ "learning_rate": 7.152857189491406e-06,
497
+ "loss": 1.1719,
498
+ "num_tokens": 47534251.0,
499
  "step": 60
500
  },
501
  {
502
+ "epoch": 1.2467866323907455,
503
+ "grad_norm": 2.292323991944126,
504
+ "learning_rate": 7.059861618411789e-06,
505
+ "loss": 1.0608,
506
+ "num_tokens": 48344753.0,
507
  "step": 61
508
  },
509
  {
510
+ "epoch": 1.2673521850899743,
511
+ "grad_norm": 0.8232509355203044,
512
+ "learning_rate": 6.966102578320531e-06,
513
+ "loss": 1.123,
514
+ "num_tokens": 49162942.0,
515
  "step": 62
516
  },
517
  {
518
+ "epoch": 1.287917737789203,
519
+ "grad_norm": 3.585688770272051,
520
+ "learning_rate": 6.87162595926379e-06,
521
+ "loss": 1.1799,
522
+ "num_tokens": 49945314.0,
523
  "step": 63
524
  },
525
  {
526
+ "epoch": 1.3084832904884318,
527
+ "grad_norm": 4.4862385716926605,
528
+ "learning_rate": 6.776478002504335e-06,
529
+ "loss": 1.1507,
530
+ "num_tokens": 50765754.0,
531
  "step": 64
532
  },
533
  {
534
+ "epoch": 1.3290488431876606,
535
+ "grad_norm": 2.963644902210829,
536
+ "learning_rate": 6.680705277888907e-06,
537
+ "loss": 1.1385,
538
+ "num_tokens": 51580961.0,
539
  "step": 65
540
  },
541
  {
542
+ "epoch": 1.3496143958868894,
543
+ "grad_norm": 0.9625449299027389,
544
+ "learning_rate": 6.584354661054765e-06,
545
+ "loss": 1.0606,
546
+ "num_tokens": 52327855.0,
547
  "step": 66
548
  },
549
  {
550
+ "epoch": 1.3701799485861184,
551
+ "grad_norm": 3.402724405424802,
552
+ "learning_rate": 6.487473310486574e-06,
553
+ "loss": 1.1162,
554
+ "num_tokens": 53102088.0,
555
  "step": 67
556
  },
557
  {
558
+ "epoch": 1.390745501285347,
559
+ "grad_norm": 1.9902964062133557,
560
+ "learning_rate": 6.390108644434828e-06,
561
+ "loss": 1.0108,
562
+ "num_tokens": 53877061.0,
563
  "step": 68
564
  },
565
  {
566
+ "epoch": 1.411311053984576,
567
+ "grad_norm": 0.8883185890120039,
568
+ "learning_rate": 6.292308317707163e-06,
569
+ "loss": 1.1061,
570
+ "num_tokens": 54680526.0,
571
  "step": 69
572
  },
573
  {
574
+ "epoch": 1.4318766066838047,
575
+ "grad_norm": 3.1753697126676976,
576
+ "learning_rate": 6.19412019834386e-06,
577
+ "loss": 1.1154,
578
+ "num_tokens": 55519612.0,
579
  "step": 70
580
  },
581
  {
582
+ "epoch": 1.4524421593830334,
583
+ "grad_norm": 1.8774651070058619,
584
+ "learning_rate": 6.095592344189002e-06,
585
+ "loss": 1.0765,
586
+ "num_tokens": 56336361.0,
587
  "step": 71
588
  },
589
  {
590
+ "epoch": 1.4730077120822622,
591
+ "grad_norm": 0.7156088232150171,
592
+ "learning_rate": 5.996772979368715e-06,
593
+ "loss": 1.1009,
594
+ "num_tokens": 57166304.0,
595
  "step": 72
596
  },
597
  {
598
+ "epoch": 1.493573264781491,
599
+ "grad_norm": 1.594991892592125,
600
+ "learning_rate": 5.897710470688042e-06,
601
+ "loss": 1.0415,
602
+ "num_tokens": 57940003.0,
603
  "step": 73
604
  },
605
  {
606
+ "epoch": 1.5141388174807198,
607
+ "grad_norm": 1.2034191805379335,
608
+ "learning_rate": 5.798453303957968e-06,
609
+ "loss": 1.0978,
610
+ "num_tokens": 58721507.0,
611
  "step": 74
612
  },
613
  {
614
+ "epoch": 1.5347043701799485,
615
+ "grad_norm": 2.43542768552457,
616
+ "learning_rate": 5.6990500602642056e-06,
617
+ "loss": 1.1483,
618
+ "num_tokens": 59499744.0,
619
  "step": 75
620
  },
621
  {
622
+ "epoch": 1.5552699228791775,
623
+ "grad_norm": 1.0836491203509007,
624
+ "learning_rate": 5.5995493921893415e-06,
625
+ "loss": 1.1097,
626
+ "num_tokens": 60318482.0,
627
  "step": 76
628
  },
629
  {
630
+ "epoch": 1.575835475578406,
631
+ "grad_norm": 0.7098795195940786,
632
+ "learning_rate": 5.500000000000001e-06,
633
+ "loss": 1.0371,
634
+ "num_tokens": 61070625.0,
635
  "step": 77
636
  },
637
  {
638
+ "epoch": 1.596401028277635,
639
+ "grad_norm": 0.7002726718932227,
640
+ "learning_rate": 5.40045060781066e-06,
641
+ "loss": 1.1046,
642
+ "num_tokens": 61874321.0,
643
  "step": 78
644
  },
645
  {
646
+ "epoch": 1.6169665809768636,
647
+ "grad_norm": 0.5013188744398371,
648
+ "learning_rate": 5.300949939735797e-06,
649
+ "loss": 1.0814,
650
+ "num_tokens": 62747678.0,
651
  "step": 79
652
  },
653
  {
654
+ "epoch": 1.6375321336760926,
655
+ "grad_norm": 0.930487154531265,
656
+ "learning_rate": 5.201546696042033e-06,
657
+ "loss": 1.1056,
658
+ "num_tokens": 63600515.0,
659
  "step": 80
660
  },
661
  {
662
+ "epoch": 1.6580976863753212,
663
+ "grad_norm": 0.5762398846789,
664
+ "learning_rate": 5.10228952931196e-06,
665
+ "loss": 1.1254,
666
+ "num_tokens": 64367564.0,
667
  "step": 81
668
  },
669
  {
670
+ "epoch": 1.6786632390745502,
671
+ "grad_norm": 0.4494307449174887,
672
+ "learning_rate": 5.003227020631287e-06,
673
+ "loss": 1.067,
674
+ "num_tokens": 65129369.0,
675
  "step": 82
676
  },
677
  {
678
+ "epoch": 1.699228791773779,
679
+ "grad_norm": 0.5249083910569988,
680
+ "learning_rate": 4.904407655811e-06,
681
+ "loss": 1.0899,
682
+ "num_tokens": 65908690.0,
683
  "step": 83
684
  },
685
  {
686
+ "epoch": 1.7197943444730077,
687
+ "grad_norm": 1.20880116726979,
688
+ "learning_rate": 4.80587980165614e-06,
689
+ "loss": 1.1388,
690
+ "num_tokens": 66729195.0,
691
  "step": 84
692
  },
693
  {
694
+ "epoch": 1.7403598971722365,
695
+ "grad_norm": 1.057560348088788,
696
+ "learning_rate": 4.707691682292838e-06,
697
+ "loss": 1.1105,
698
+ "num_tokens": 67475591.0,
699
  "step": 85
700
  },
701
  {
702
+ "epoch": 1.7609254498714653,
703
+ "grad_norm": 0.6863002803853495,
704
+ "learning_rate": 4.609891355565172e-06,
705
+ "loss": 1.0767,
706
+ "num_tokens": 68256417.0,
707
  "step": 86
708
  },
709
  {
710
+ "epoch": 1.781491002570694,
711
+ "grad_norm": 0.6819115141106705,
712
+ "learning_rate": 4.5125266895134275e-06,
713
+ "loss": 1.0941,
714
+ "num_tokens": 69028462.0,
715
  "step": 87
716
  },
717
  {
718
+ "epoch": 1.8020565552699228,
719
+ "grad_norm": 1.4347921037396176,
720
+ "learning_rate": 4.415645338945236e-06,
721
+ "loss": 1.091,
722
+ "num_tokens": 69808820.0,
723
  "step": 88
724
  },
725
  {
726
+ "epoch": 1.8226221079691518,
727
+ "grad_norm": 1.4693715473624118,
728
+ "learning_rate": 4.3192947221110956e-06,
729
+ "loss": 1.0714,
730
+ "num_tokens": 70645343.0,
731
  "step": 89
732
  },
733
  {
734
+ "epoch": 1.8431876606683804,
735
+ "grad_norm": 0.6582970773448227,
736
+ "learning_rate": 4.223521997495665e-06,
737
+ "loss": 1.0957,
738
+ "num_tokens": 71458144.0,
739
  "step": 90
740
  },
741
  {
742
+ "epoch": 1.8637532133676094,
743
+ "grad_norm": 1.1825217028469224,
744
+ "learning_rate": 4.128374040736211e-06,
745
+ "loss": 1.1132,
746
+ "num_tokens": 72282385.0,
747
  "step": 91
748
  },
749
  {
750
+ "epoch": 1.884318766066838,
751
+ "grad_norm": 1.1298665815485143,
752
+ "learning_rate": 4.033897421679472e-06,
753
+ "loss": 1.0389,
754
+ "num_tokens": 73088829.0,
755
  "step": 92
756
  },
757
  {
758
+ "epoch": 1.904884318766067,
759
+ "grad_norm": 1.3570211954476283,
760
+ "learning_rate": 3.940138381588214e-06,
761
+ "loss": 1.1499,
762
+ "num_tokens": 73938027.0,
763
  "step": 93
764
  },
765
  {
766
+ "epoch": 1.9254498714652957,
767
+ "grad_norm": 1.280064760591942,
768
+ "learning_rate": 3.847142810508596e-06,
769
+ "loss": 1.0316,
770
+ "num_tokens": 74712404.0,
771
  "step": 94
772
  },
773
  {
774
+ "epoch": 1.9460154241645244,
775
+ "grad_norm": 1.446677445007892,
776
+ "learning_rate": 3.7549562248094396e-06,
777
+ "loss": 1.0729,
778
+ "num_tokens": 75476326.0,
779
  "step": 95
780
  },
781
  {
782
+ "epoch": 1.9665809768637532,
783
+ "grad_norm": 0.7546075726383373,
784
+ "learning_rate": 3.6636237449044077e-06,
785
+ "loss": 1.0677,
786
+ "num_tokens": 76280080.0,
787
  "step": 96
788
  },
789
  {
790
+ "epoch": 1.987146529562982,
791
+ "grad_norm": 0.6418441333219697,
792
+ "learning_rate": 3.5731900731679525e-06,
793
+ "loss": 1.0859,
794
+ "num_tokens": 77079865.0,
795
  "step": 97
796
  },
797
  {
798
+ "epoch": 2.0,
799
+ "grad_norm": 1.1588829155309626,
800
+ "learning_rate": 3.483699472055897e-06,
801
+ "loss": 1.0656,
802
+ "num_tokens": 77568391.0,
803
  "step": 98
804
  },
805
  {
806
+ "epoch": 2.020565552699229,
807
+ "grad_norm": 0.6238842093615174,
808
+ "learning_rate": 3.3951957424413327e-06,
809
+ "loss": 1.0429,
810
+ "num_tokens": 78366813.0,
811
  "step": 99
812
  },
813
  {
814
+ "epoch": 2.0411311053984575,
815
+ "grad_norm": 0.5553403632072313,
816
+ "learning_rate": 3.307722202176417e-06,
817
+ "loss": 1.0947,
818
+ "num_tokens": 79224595.0,
819
  "step": 100
820
  },
821
  {
822
+ "epoch": 2.0616966580976865,
823
+ "grad_norm": 0.5248026371177484,
824
+ "learning_rate": 3.2213216648906037e-06,
825
+ "loss": 1.0545,
826
+ "num_tokens": 80031127.0,
827
  "step": 101
828
  },
829
  {
830
+ "epoch": 2.082262210796915,
831
+ "grad_norm": 0.5022337615857637,
832
+ "learning_rate": 3.136036419035656e-06,
833
+ "loss": 1.0642,
834
+ "num_tokens": 80790581.0,
835
  "step": 102
836
  },
837
  {
838
+ "epoch": 2.082262210796915,
839
+ "eval_loss": 0.8722440600395203,
840
+ "eval_num_tokens": 80790581.0,
841
+ "eval_runtime": 30.28,
842
+ "eval_samples_per_second": 45.641,
843
+ "eval_steps_per_second": 5.713,
844
+ "step": 102
845
+ },
846
+ {
847
+ "epoch": 2.102827763496144,
848
+ "grad_norm": 0.9132791992130541,
849
+ "learning_rate": 3.051908207187705e-06,
850
+ "loss": 1.0463,
851
+ "num_tokens": 81572710.0,
852
  "step": 103
853
  },
854
  {
855
+ "epoch": 2.1233933161953726,
856
+ "grad_norm": 0.49676184487454156,
857
+ "learning_rate": 2.9689782056164874e-06,
858
+ "loss": 1.0363,
859
+ "num_tokens": 82392421.0,
860
  "step": 104
861
  },
862
  {
863
+ "epoch": 2.1439588688946016,
864
+ "grad_norm": 0.41641122076638964,
865
+ "learning_rate": 2.887287004131757e-06,
866
+ "loss": 1.0129,
867
+ "num_tokens": 83197018.0,
868
  "step": 105
869
  },
870
  {
871
+ "epoch": 2.16452442159383,
872
+ "grad_norm": 0.6305455673369688,
873
+ "learning_rate": 2.8068745862167423e-06,
874
+ "loss": 1.0604,
875
+ "num_tokens": 84017921.0,
876
  "step": 106
877
  },
878
  {
879
+ "epoch": 2.185089974293059,
880
+ "grad_norm": 0.45877001918075977,
881
+ "learning_rate": 2.727780309458361e-06,
882
+ "loss": 1.0573,
883
+ "num_tokens": 84789176.0,
884
  "step": 107
885
  },
886
  {
887
+ "epoch": 2.2056555269922877,
888
+ "grad_norm": 0.6286596726421219,
889
+ "learning_rate": 2.6500428862837878e-06,
890
+ "loss": 0.9936,
891
+ "num_tokens": 85561595.0,
892
  "step": 108
893
  },
894
  {
895
+ "epoch": 2.2262210796915167,
896
+ "grad_norm": 0.7376081520239532,
897
+ "learning_rate": 2.5737003650127834e-06,
898
+ "loss": 0.9638,
899
+ "num_tokens": 86334568.0,
900
  "step": 109
901
  },
902
  {
903
+ "epoch": 2.2467866323907453,
904
+ "grad_norm": 1.0507755635770515,
905
+ "learning_rate": 2.498790111235072e-06,
906
+ "loss": 1.0676,
907
+ "num_tokens": 87089590.0,
908
  "step": 110
909
  },
910
  {
911
+ "epoch": 2.2673521850899743,
912
+ "grad_norm": 0.751812284582568,
913
+ "learning_rate": 2.4253487895218796e-06,
914
+ "loss": 1.003,
915
+ "num_tokens": 87870458.0,
916
  "step": 111
917
  },
918
  {
919
+ "epoch": 2.2879177377892033,
920
+ "grad_norm": 0.7258055709708501,
921
+ "learning_rate": 2.353412345480587e-06,
922
+ "loss": 1.0957,
923
+ "num_tokens": 88697761.0,
924
  "step": 112
925
  },
926
  {
927
+ "epoch": 2.308483290488432,
928
+ "grad_norm": 0.9702567291939189,
929
+ "learning_rate": 2.2830159881612695e-06,
930
+ "loss": 0.9815,
931
+ "num_tokens": 89498858.0,
932
  "step": 113
933
  },
934
  {
935
+ "epoch": 2.329048843187661,
936
+ "grad_norm": 1.706294287936,
937
+ "learning_rate": 2.2141941728237467e-06,
938
+ "loss": 1.02,
939
+ "num_tokens": 90207780.0,
940
  "step": 114
941
  },
942
  {
943
+ "epoch": 2.3496143958868894,
944
+ "grad_norm": 0.7383798216635169,
945
+ "learning_rate": 2.146980584073573e-06,
946
+ "loss": 1.0215,
947
+ "num_tokens": 90929907.0,
948
  "step": 115
949
  },
950
  {
951
+ "epoch": 2.3701799485861184,
952
+ "grad_norm": 0.8092138470162481,
953
+ "learning_rate": 2.081408119375219e-06,
954
+ "loss": 1.0648,
955
+ "num_tokens": 91776397.0,
956
  "step": 116
957
  },
958
  {
959
+ "epoch": 2.390745501285347,
960
+ "grad_norm": 0.8621943427931472,
961
+ "learning_rate": 2.0175088729505206e-06,
962
+ "loss": 1.0926,
963
+ "num_tokens": 92513206.0,
964
  "step": 117
965
  },
966
  {
967
+ "epoch": 2.411311053984576,
968
+ "grad_norm": 1.787970295069879,
969
+ "learning_rate": 1.955314120070269e-06,
970
+ "loss": 0.9965,
971
+ "num_tokens": 93306104.0,
972
  "step": 118
973
  },
974
  {
975
+ "epoch": 2.4318766066838045,
976
+ "grad_norm": 1.479554722978421,
977
+ "learning_rate": 1.8948543017466337e-06,
978
+ "loss": 1.0851,
979
+ "num_tokens": 94155482.0,
980
  "step": 119
981
  },
982
  {
983
+ "epoch": 2.4524421593830334,
984
+ "grad_norm": 0.49335005217781264,
985
+ "learning_rate": 1.8361590098339168e-06,
986
+ "loss": 1.037,
987
+ "num_tokens": 94936066.0,
988
  "step": 120
989
  },
990
  {
991
+ "epoch": 2.4730077120822624,
992
+ "grad_norm": 0.6721016458518935,
993
+ "learning_rate": 1.7792569725449114e-06,
994
+ "loss": 1.0131,
995
+ "num_tokens": 95818245.0,
996
  "step": 121
997
  },
998
  {
999
+ "epoch": 2.493573264781491,
1000
+ "grad_norm": 1.0956037260235203,
1001
+ "learning_rate": 1.724176040389982e-06,
1002
+ "loss": 1.0814,
1003
+ "num_tokens": 96635430.0,
1004
  "step": 122
1005
  },
1006
  {
1007
+ "epoch": 2.5141388174807195,
1008
+ "grad_norm": 1.107512998866383,
1009
+ "learning_rate": 1.6709431725457203e-06,
1010
+ "loss": 1.0615,
1011
+ "num_tokens": 97438156.0,
1012
  "step": 123
1013
  },
1014
  {
1015
+ "epoch": 2.5347043701799485,
1016
+ "grad_norm": 0.5872510671862957,
1017
+ "learning_rate": 1.619584423659875e-06,
1018
+ "loss": 1.042,
1019
+ "num_tokens": 98247321.0,
1020
  "step": 124
1021
  },
1022
  {
1023
+ "epoch": 2.5552699228791775,
1024
+ "grad_norm": 0.9339232970249086,
1025
+ "learning_rate": 1.5701249310989827e-06,
1026
+ "loss": 1.0533,
1027
+ "num_tokens": 99018604.0,
1028
  "step": 125
1029
  },
1030
  {
1031
+ "epoch": 2.575835475578406,
1032
+ "grad_norm": 0.9907046236471151,
1033
+ "learning_rate": 1.5225889026449754e-06,
1034
+ "loss": 1.0472,
1035
+ "num_tokens": 99814782.0,
1036
  "step": 126
1037
  },
1038
  {
1039
+ "epoch": 2.596401028277635,
1040
+ "grad_norm": 0.42612961753245276,
1041
+ "learning_rate": 1.4769996046467599e-06,
1042
+ "loss": 1.1264,
1043
+ "num_tokens": 100614217.0,
1044
  "step": 127
1045
  },
1046
  {
1047
+ "epoch": 2.6169665809768636,
1048
+ "grad_norm": 0.9091136187984498,
1049
+ "learning_rate": 1.4333793506325832e-06,
1050
+ "loss": 1.0193,
1051
+ "num_tokens": 101396655.0,
1052
  "step": 128
1053
  },
1054
  {
1055
+ "epoch": 2.6375321336760926,
1056
+ "grad_norm": 0.7892032837871268,
1057
+ "learning_rate": 1.3917494903887516e-06,
1058
+ "loss": 0.9758,
1059
+ "num_tokens": 102215959.0,
1060
  "step": 129
1061
  },
1062
  {
1063
+ "epoch": 2.658097686375321,
1064
+ "grad_norm": 0.5128890053331989,
1065
+ "learning_rate": 1.3521303995100479e-06,
1066
+ "loss": 1.057,
1067
+ "num_tokens": 103010668.0,
1068
  "step": 130
1069
  },
1070
  {
1071
+ "epoch": 2.67866323907455,
1072
+ "grad_norm": 1.0590105248836392,
1073
+ "learning_rate": 1.3145414694269671e-06,
1074
+ "loss": 1.0391,
1075
+ "num_tokens": 103845595.0,
1076
  "step": 131
1077
  },
1078
  {
1079
+ "epoch": 2.6992287917737787,
1080
+ "grad_norm": 0.8126843666598889,
1081
+ "learning_rate": 1.2790010979146467e-06,
1082
+ "loss": 0.9984,
1083
+ "num_tokens": 104715999.0,
1084
  "step": 132
1085
  },
1086
  {
1087
+ "epoch": 2.7197943444730077,
1088
+ "grad_norm": 0.47820659260495135,
1089
+ "learning_rate": 1.2455266800881342e-06,
1090
+ "loss": 1.0,
1091
+ "num_tokens": 105488928.0,
1092
  "step": 133
1093
  },
1094
  {
1095
+ "epoch": 2.7403598971722367,
1096
+ "grad_norm": 0.5317897918745331,
1097
+ "learning_rate": 1.2141345998884092e-06,
1098
+ "loss": 1.0498,
1099
+ "num_tokens": 106186987.0,
1100
  "step": 134
1101
  },
1102
  {
1103
+ "epoch": 2.7609254498714653,
1104
+ "grad_norm": 0.9341792757623867,
1105
+ "learning_rate": 1.1848402220633198e-06,
1106
+ "loss": 0.9566,
1107
+ "num_tokens": 106967242.0,
1108
  "step": 135
1109
  },
1110
  {
1111
+ "epoch": 2.781491002570694,
1112
+ "grad_norm": 1.4588060292984133,
1113
+ "learning_rate": 1.1576578846473558e-06,
1114
+ "loss": 1.0252,
1115
+ "num_tokens": 107817871.0,
1116
  "step": 136
1117
  },
1118
  {
1119
+ "epoch": 2.802056555269923,
1120
+ "grad_norm": 0.8215750745057664,
1121
+ "learning_rate": 1.1326008919439476e-06,
1122
+ "loss": 1.0712,
1123
+ "num_tokens": 108629125.0,
1124
  "step": 137
1125
  },
1126
  {
1127
+ "epoch": 2.822622107969152,
1128
+ "grad_norm": 0.38163574145212953,
1129
+ "learning_rate": 1.1096815080137196e-06,
1130
+ "loss": 1.0421,
1131
+ "num_tokens": 109428401.0,
1132
  "step": 138
1133
  },
1134
  {
1135
+ "epoch": 2.8431876606683804,
1136
+ "grad_norm": 1.0563512897683773,
1137
+ "learning_rate": 1.0889109506718873e-06,
1138
+ "loss": 0.961,
1139
+ "num_tokens": 110292149.0,
1140
  "step": 139
1141
  },
1142
  {
1143
+ "epoch": 2.8637532133676094,
1144
+ "grad_norm": 0.6548608506755417,
1145
+ "learning_rate": 1.070299385997735e-06,
1146
+ "loss": 0.9862,
1147
+ "num_tokens": 111130565.0,
1148
  "step": 140
1149
  },
1150
  {
1151
+ "epoch": 2.884318766066838,
1152
+ "grad_norm": 0.7840147082599769,
1153
+ "learning_rate": 1.053855923358862e-06,
1154
+ "loss": 1.101,
1155
+ "num_tokens": 111859971.0,
1156
  "step": 141
1157
  },
1158
  {
1159
+ "epoch": 2.904884318766067,
1160
+ "grad_norm": 0.8638836969151383,
1161
+ "learning_rate": 1.0395886109526346e-06,
1162
+ "loss": 1.0626,
1163
+ "num_tokens": 112658866.0,
1164
  "step": 142
1165
  },
1166
  {
1167
+ "epoch": 2.925449871465296,
1168
+ "grad_norm": 0.5112313921441834,
1169
+ "learning_rate": 1.0275044318670219e-06,
1170
+ "loss": 1.0679,
1171
+ "num_tokens": 113427101.0,
1172
  "step": 143
1173
  },
1174
  {
1175
+ "epoch": 2.9460154241645244,
1176
+ "grad_norm": 1.1240389638051913,
1177
+ "learning_rate": 1.0176093006627485e-06,
1178
+ "loss": 0.9497,
1179
+ "num_tokens": 114240330.0,
1180
  "step": 144
1181
  },
1182
  {
1183
+ "epoch": 2.966580976863753,
1184
+ "grad_norm": 1.2175079530173951,
1185
+ "learning_rate": 1.0099080604784303e-06,
1186
+ "loss": 0.9541,
1187
+ "num_tokens": 115051111.0,
1188
  "step": 145
1189
  },
1190
  {
1191
+ "epoch": 2.987146529562982,
1192
+ "grad_norm": 0.6533936705036852,
1193
+ "learning_rate": 1.0044044806601188e-06,
1194
+ "loss": 1.0145,
1195
+ "num_tokens": 115853478.0,
1196
  "step": 146
1197
  },
1198
  {
1199
+ "epoch": 3.0,
1200
+ "grad_norm": 0.6533936705036852,
1201
+ "learning_rate": 1.0011012549164066e-06,
1202
+ "loss": 1.0527,
1203
+ "num_tokens": 116344566.0,
1204
  "step": 147
1205
  },
1206
  {
1207
+ "epoch": 3.0,
1208
+ "step": 147,
1209
+ "total_flos": 5.614755170526167e+17,
1210
+ "train_loss": 1.1423310605036159,
1211
+ "train_runtime": 3470.2952,
1212
+ "train_samples_per_second": 10.746,
1213
+ "train_steps_per_second": 0.042
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1214
  }
1215
  ],
1216
  "logging_steps": 1,
1217
+ "max_steps": 147,
1218
  "num_input_tokens_seen": 0,
1219
+ "num_train_epochs": 3,
1220
  "save_steps": 500,
1221
  "stateful_callbacks": {
1222
  "TrainerControl": {
 
1230
  "attributes": {}
1231
  }
1232
  },
1233
+ "total_flos": 5.614755170526167e+17,
1234
+ "train_batch_size": 4,
1235
  "trial_name": null,
1236
  "trial_params": null
1237
  }