File size: 29,702 Bytes
7df5122
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "eval_steps": 500,
  "global_step": 150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 95.85,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 195.6,
      "completions/mean_length": 95.85,
      "completions/mean_terminated_length": 72.8880989074707,
      "completions/min_length": 36.6,
      "completions/min_terminated_length": 36.6,
      "epoch": 0.5,
      "frac_reward_zero_std": 0.2,
      "grad_norm": 0.3041034936904907,
      "kl": 0.00021898818758927518,
      "learning_rate": 4e-05,
      "loss": 0.0329527884721756,
      "num_tokens": 21444.0,
      "reward": 0.7767857074737549,
      "reward_std": 0.7150911569595337,
      "rewards/reward_fn/mean": 0.7767857074737549,
      "rewards/reward_fn/std": 0.9959433197975158,
      "step": 5
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 95.1,
      "completions/clipped_ratio": 0.0875,
      "completions/max_length": 225.2,
      "completions/max_terminated_length": 159.2,
      "completions/mean_length": 95.1,
      "completions/mean_terminated_length": 80.34466400146485,
      "completions/min_length": 34.6,
      "completions/min_terminated_length": 34.6,
      "epoch": 1.0,
      "frac_reward_zero_std": 0.4,
      "grad_norm": 0.2708108425140381,
      "kl": 0.006061523332027719,
      "learning_rate": 9e-05,
      "loss": -0.016053691506385803,
      "num_tokens": 42220.0,
      "reward": 0.6875000119209289,
      "reward_std": 0.5175672590732574,
      "rewards/reward_fn/mean": 0.6875,
      "rewards/reward_fn/std": 0.8034753799438477,
      "step": 10
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 146.275,
      "completions/clipped_ratio": 0.325,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 225.4,
      "completions/mean_length": 146.275,
      "completions/mean_terminated_length": 92.8663101196289,
      "completions/min_length": 32.6,
      "completions/min_terminated_length": 32.6,
      "epoch": 1.5,
      "frac_reward_zero_std": 0.2,
      "grad_norm": 0.27529817819595337,
      "kl": 0.01560373855754733,
      "learning_rate": 9.989068136093873e-05,
      "loss": 0.010054035484790802,
      "num_tokens": 69198.0,
      "reward": 1.1558928728103637,
      "reward_std": 0.8465379416942597,
      "rewards/reward_fn/mean": 1.1558928728103637,
      "rewards/reward_fn/std": 1.1404805421829223,
      "step": 15
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 75.85,
      "completions/clipped_ratio": 0.05,
      "completions/max_length": 220.6,
      "completions/max_terminated_length": 151.0,
      "completions/mean_length": 75.85,
      "completions/mean_terminated_length": 66.14857330322266,
      "completions/min_length": 32.8,
      "completions/min_terminated_length": 32.8,
      "epoch": 2.0,
      "frac_reward_zero_std": 0.35,
      "grad_norm": 0.353487491607666,
      "kl": 0.027986684162169696,
      "learning_rate": 9.944739353007344e-05,
      "loss": -0.030885905027389526,
      "num_tokens": 87334.0,
      "reward": 0.7932142853736878,
      "reward_std": 0.4118983089923859,
      "rewards/reward_fn/mean": 0.7932142853736878,
      "rewards/reward_fn/std": 0.7167340397834778,
      "step": 20
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 90.85,
      "completions/clipped_ratio": 0.0625,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 216.4,
      "completions/mean_length": 90.85,
      "completions/mean_terminated_length": 79.819287109375,
      "completions/min_length": 28.0,
      "completions/min_terminated_length": 28.0,
      "epoch": 2.5,
      "frac_reward_zero_std": 0.15,
      "grad_norm": 0.3318491578102112,
      "kl": 0.04692814685404301,
      "learning_rate": 9.86663298624003e-05,
      "loss": 0.05783940553665161,
      "num_tokens": 108330.0,
      "reward": 1.1987500190734863,
      "reward_std": 0.5907092690467834,
      "rewards/reward_fn/mean": 1.1987500190734863,
      "rewards/reward_fn/std": 0.9482298731803894,
      "step": 25
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 91.775,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 248.2,
      "completions/max_terminated_length": 212.6,
      "completions/mean_length": 91.775,
      "completions/mean_terminated_length": 68.68722763061524,
      "completions/min_length": 25.2,
      "completions/min_terminated_length": 25.2,
      "epoch": 3.0,
      "frac_reward_zero_std": 0.4,
      "grad_norm": 0.24605490267276764,
      "kl": 0.07254143096506596,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.03934891819953919,
      "num_tokens": 129012.0,
      "reward": 1.0242857217788697,
      "reward_std": 0.5802445828914642,
      "rewards/reward_fn/mean": 1.0242857217788697,
      "rewards/reward_fn/std": 0.8706011414527893,
      "step": 30
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 59.7875,
      "completions/clipped_ratio": 0.0375,
      "completions/max_length": 175.6,
      "completions/max_terminated_length": 118.4,
      "completions/mean_length": 59.7875,
      "completions/mean_terminated_length": 52.19559555053711,
      "completions/min_length": 25.4,
      "completions/min_terminated_length": 25.4,
      "epoch": 3.5,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 0.2970605790615082,
      "kl": 0.053099883161485194,
      "learning_rate": 9.611448774886924e-05,
      "loss": 0.06977529525756836,
      "num_tokens": 147103.0,
      "reward": 0.9473214387893677,
      "reward_std": 0.3923970699310303,
      "rewards/reward_fn/mean": 0.9473214268684387,
      "rewards/reward_fn/std": 0.649278998374939,
      "step": 35
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 47.85,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 94.8,
      "completions/max_terminated_length": 94.8,
      "completions/mean_length": 47.85,
      "completions/mean_terminated_length": 47.85,
      "completions/min_length": 31.2,
      "completions/min_terminated_length": 31.2,
      "epoch": 4.0,
      "frac_reward_zero_std": 0.85,
      "grad_norm": 0.16646049916744232,
      "kl": 0.056573960930109024,
      "learning_rate": 9.43611409721806e-05,
      "loss": -0.0031644195318222047,
      "num_tokens": 164643.0,
      "reward": 0.7899999976158142,
      "reward_std": 0.12681145668029786,
      "rewards/reward_fn/mean": 0.7899999976158142,
      "rewards/reward_fn/std": 0.5133758962154389,
      "step": 40
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 49.2625,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 93.6,
      "completions/max_terminated_length": 93.6,
      "completions/mean_length": 49.2625,
      "completions/mean_terminated_length": 49.2625,
      "completions/min_length": 33.8,
      "completions/min_terminated_length": 33.8,
      "epoch": 4.5,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 0.3159082531929016,
      "kl": 0.0353464649990201,
      "learning_rate": 9.230476262104677e-05,
      "loss": -0.03417131304740906,
      "num_tokens": 182568.0,
      "reward": 0.7924999833106995,
      "reward_std": 0.3821385264396667,
      "rewards/reward_fn/mean": 0.7924999833106995,
      "rewards/reward_fn/std": 0.62625293135643,
      "step": 45
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 58.25,
      "completions/clipped_ratio": 0.0125,
      "completions/max_length": 180.6,
      "completions/max_terminated_length": 153.2,
      "completions/mean_length": 58.25,
      "completions/mean_terminated_length": 55.67250061035156,
      "completions/min_length": 33.2,
      "completions/min_terminated_length": 33.2,
      "epoch": 5.0,
      "frac_reward_zero_std": 0.55,
      "grad_norm": 0.27360936999320984,
      "kl": 0.04483541352674365,
      "learning_rate": 8.995939984474624e-05,
      "loss": 0.026852786540985107,
      "num_tokens": 200660.0,
      "reward": 0.774821424484253,
      "reward_std": 0.32772802114486693,
      "rewards/reward_fn/mean": 0.774821424484253,
      "rewards/reward_fn/std": 0.5432307064533234,
      "step": 50
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 51.225,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 117.4,
      "completions/max_terminated_length": 117.4,
      "completions/mean_length": 51.225,
      "completions/mean_terminated_length": 51.225,
      "completions/min_length": 36.6,
      "completions/min_terminated_length": 36.6,
      "epoch": 5.5,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 0.21470840275287628,
      "kl": 0.04195364024490118,
      "learning_rate": 8.73410738492077e-05,
      "loss": 0.023307889699935913,
      "num_tokens": 218442.0,
      "reward": 1.0364285826683044,
      "reward_std": 0.37462076991796495,
      "rewards/reward_fn/mean": 1.0364285826683044,
      "rewards/reward_fn/std": 0.6156843721866607,
      "step": 55
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 53.225,
      "completions/clipped_ratio": 0.0125,
      "completions/max_length": 159.8,
      "completions/max_terminated_length": 123.8,
      "completions/mean_length": 53.225,
      "completions/mean_terminated_length": 50.68583374023437,
      "completions/min_length": 33.8,
      "completions/min_terminated_length": 33.8,
      "epoch": 6.0,
      "frac_reward_zero_std": 0.45,
      "grad_norm": 0.29306381940841675,
      "kl": 0.05784560404717922,
      "learning_rate": 8.44676704559283e-05,
      "loss": 0.015561078488826752,
      "num_tokens": 236112.0,
      "reward": 1.1410714387893677,
      "reward_std": 0.3540264666080475,
      "rewards/reward_fn/mean": 1.1410714387893677,
      "rewards/reward_fn/std": 0.607547128200531,
      "step": 60
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 62.2625,
      "completions/clipped_ratio": 0.05,
      "completions/max_length": 153.2,
      "completions/max_terminated_length": 117.4,
      "completions/mean_length": 62.2625,
      "completions/mean_terminated_length": 52.46428680419922,
      "completions/min_length": 30.6,
      "completions/min_terminated_length": 30.6,
      "epoch": 6.5,
      "frac_reward_zero_std": 0.65,
      "grad_norm": 0.2615426480770111,
      "kl": 0.09820330440998078,
      "learning_rate": 8.135881792367686e-05,
      "loss": 0.04314278066158295,
      "num_tokens": 254033.0,
      "reward": 1.2110714197158814,
      "reward_std": 0.2716316431760788,
      "rewards/reward_fn/mean": 1.2110714197158814,
      "rewards/reward_fn/std": 0.6701113641262054,
      "step": 65
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 82.5625,
      "completions/clipped_ratio": 0.1125,
      "completions/max_length": 234.4,
      "completions/max_terminated_length": 147.4,
      "completions/mean_length": 82.5625,
      "completions/mean_terminated_length": 60.485835266113284,
      "completions/min_length": 35.4,
      "completions/min_terminated_length": 35.4,
      "epoch": 7.0,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 0.1326226145029068,
      "kl": 0.07194480895996094,
      "learning_rate": 7.803575286758364e-05,
      "loss": 0.03152734041213989,
      "num_tokens": 275222.0,
      "reward": 1.320892834663391,
      "reward_std": 0.35234126150608064,
      "rewards/reward_fn/mean": 1.320892858505249,
      "rewards/reward_fn/std": 0.676600456237793,
      "step": 70
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 72.7625,
      "completions/clipped_ratio": 0.0625,
      "completions/max_length": 220.4,
      "completions/max_terminated_length": 166.8,
      "completions/mean_length": 72.7625,
      "completions/mean_terminated_length": 59.95151138305664,
      "completions/min_length": 30.2,
      "completions/min_terminated_length": 30.2,
      "epoch": 7.5,
      "frac_reward_zero_std": 0.7,
      "grad_norm": 0.15138906240463257,
      "kl": 0.09165919721126556,
      "learning_rate": 7.452117519152542e-05,
      "loss": 0.0013809207826852798,
      "num_tokens": 294407.0,
      "reward": 1.3473214387893677,
      "reward_std": 0.17192934565246104,
      "rewards/reward_fn/mean": 1.3473214387893677,
      "rewards/reward_fn/std": 0.6935437321662903,
      "step": 75
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 80.9375,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 222.4,
      "completions/max_terminated_length": 156.0,
      "completions/mean_length": 80.9375,
      "completions/mean_terminated_length": 55.99818344116211,
      "completions/min_length": 31.6,
      "completions/min_terminated_length": 31.6,
      "epoch": 8.0,
      "frac_reward_zero_std": 0.65,
      "grad_norm": 0.19589942693710327,
      "kl": 0.0683181069791317,
      "learning_rate": 7.083909302476453e-05,
      "loss": 0.033023273944854735,
      "num_tokens": 314502.0,
      "reward": 1.1685714244842529,
      "reward_std": 0.27699449211359023,
      "rewards/reward_fn/mean": 1.1685714244842529,
      "rewards/reward_fn/std": 0.5244716942310333,
      "step": 80
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 60.6625,
      "completions/clipped_ratio": 0.05,
      "completions/max_length": 207.2,
      "completions/max_terminated_length": 100.0,
      "completions/mean_length": 60.6625,
      "completions/mean_terminated_length": 50.27571563720703,
      "completions/min_length": 35.2,
      "completions/min_terminated_length": 35.2,
      "epoch": 8.5,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 0.20066438615322113,
      "kl": 0.0837025498971343,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.022454433143138885,
      "num_tokens": 334007.0,
      "reward": 1.3910714387893677,
      "reward_std": 0.21785714849829674,
      "rewards/reward_fn/mean": 1.3910714387893677,
      "rewards/reward_fn/std": 0.6235540926456451,
      "step": 85
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 54.025,
      "completions/clipped_ratio": 0.025,
      "completions/max_length": 143.4,
      "completions/max_terminated_length": 117.2,
      "completions/mean_length": 54.025,
      "completions/mean_terminated_length": 49.025001525878906,
      "completions/min_length": 30.2,
      "completions/min_terminated_length": 30.2,
      "epoch": 9.0,
      "frac_reward_zero_std": 0.8,
      "grad_norm": 0.18161360919475555,
      "kl": 0.05805329587310552,
      "learning_rate": 6.307399704769099e-05,
      "loss": -0.04096371531486511,
      "num_tokens": 351041.0,
      "reward": 1.24375,
      "reward_std": 0.2125,
      "rewards/reward_fn/mean": 1.24375,
      "rewards/reward_fn/std": 0.584508627653122,
      "step": 90
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 54.65,
      "completions/clipped_ratio": 0.0125,
      "completions/max_length": 145.0,
      "completions/max_terminated_length": 142.2,
      "completions/mean_length": 54.65,
      "completions/mean_terminated_length": 52.23083419799805,
      "completions/min_length": 35.4,
      "completions/min_terminated_length": 35.4,
      "epoch": 9.5,
      "frac_reward_zero_std": 0.9,
      "grad_norm": 0.10424701869487762,
      "kl": 0.05452606473118067,
      "learning_rate": 5.90440267166055e-05,
      "loss": 0.0021311525255441667,
      "num_tokens": 368361.0,
      "reward": 1.1996428489685058,
      "reward_std": 0.058449314022436735,
      "rewards/reward_fn/mean": 1.1996428489685058,
      "rewards/reward_fn/std": 0.4604207634925842,
      "step": 95
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 68.3875,
      "completions/clipped_ratio": 0.0625,
      "completions/max_length": 254.4,
      "completions/max_terminated_length": 140.8,
      "completions/mean_length": 68.3875,
      "completions/mean_terminated_length": 55.740359497070315,
      "completions/min_length": 36.2,
      "completions/min_terminated_length": 36.2,
      "epoch": 10.0,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 0.13536690175533295,
      "kl": 0.0854075826704502,
      "learning_rate": 5.495227651252315e-05,
      "loss": -0.025432443618774413,
      "num_tokens": 388400.0,
      "reward": 1.3283928632736206,
      "reward_std": 0.2009493112564087,
      "rewards/reward_fn/mean": 1.3283928632736206,
      "rewards/reward_fn/std": 0.5586225628852844,
      "step": 100
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 77.6,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 115.6,
      "completions/mean_length": 77.6,
      "completions/mean_terminated_length": 52.14154052734375,
      "completions/min_length": 30.4,
      "completions/min_terminated_length": 30.4,
      "epoch": 10.5,
      "frac_reward_zero_std": 0.9,
      "grad_norm": 0.0006729933083988726,
      "kl": 0.07917941156774759,
      "learning_rate": 5.0826697238317935e-05,
      "loss": -0.0014228260144591332,
      "num_tokens": 408808.0,
      "reward": 1.4,
      "reward_std": 0.1,
      "rewards/reward_fn/mean": 1.4,
      "rewards/reward_fn/std": 0.5159838497638702,
      "step": 105
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 71.6375,
      "completions/clipped_ratio": 0.1,
      "completions/max_length": 223.8,
      "completions/max_terminated_length": 125.4,
      "completions/mean_length": 71.6375,
      "completions/mean_terminated_length": 50.991825103759766,
      "completions/min_length": 34.6,
      "completions/min_terminated_length": 34.6,
      "epoch": 11.0,
      "frac_reward_zero_std": 0.85,
      "grad_norm": 0.1310124397277832,
      "kl": 0.0553528118878603,
      "learning_rate": 4.669547078371504e-05,
      "loss": 0.027294650673866272,
      "num_tokens": 427787.0,
      "reward": 1.15,
      "reward_std": 0.15773502588272095,
      "rewards/reward_fn/mean": 1.15,
      "rewards/reward_fn/std": 0.482165002822876,
      "step": 110
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 75.975,
      "completions/clipped_ratio": 0.1375,
      "completions/max_length": 214.8,
      "completions/max_terminated_length": 76.2,
      "completions/mean_length": 75.975,
      "completions/mean_terminated_length": 47.73154296875,
      "completions/min_length": 36.2,
      "completions/min_terminated_length": 36.2,
      "epoch": 11.5,
      "frac_reward_zero_std": 0.9,
      "grad_norm": 0.001892779255285859,
      "kl": 0.056033550202846526,
      "learning_rate": 4.2586817614407895e-05,
      "loss": -0.03689469993114471,
      "num_tokens": 447185.0,
      "reward": 1.2867857217788696,
      "reward_std": 0.026428571343421935,
      "rewards/reward_fn/mean": 1.2867857217788696,
      "rewards/reward_fn/std": 0.3760030448436737,
      "step": 115
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 76.725,
      "completions/clipped_ratio": 0.1375,
      "completions/max_length": 218.2,
      "completions/max_terminated_length": 82.8,
      "completions/mean_length": 76.725,
      "completions/mean_terminated_length": 48.34801483154297,
      "completions/min_length": 30.4,
      "completions/min_terminated_length": 30.4,
      "epoch": 12.0,
      "frac_reward_zero_std": 0.95,
      "grad_norm": 0.00067809724714607,
      "kl": 0.06624810677021742,
      "learning_rate": 3.852880399766243e-05,
      "loss": 0.005734995752573013,
      "num_tokens": 466979.0,
      "reward": 1.325,
      "reward_std": 0.05,
      "rewards/reward_fn/mean": 1.325,
      "rewards/reward_fn/std": 0.4854445576667786,
      "step": 120
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 70.3625,
      "completions/clipped_ratio": 0.0875,
      "completions/max_length": 183.8,
      "completions/max_terminated_length": 99.4,
      "completions/mean_length": 70.3625,
      "completions/mean_terminated_length": 52.45879211425781,
      "completions/min_length": 38.0,
      "completions/min_terminated_length": 38.0,
      "epoch": 12.5,
      "frac_reward_zero_std": 0.9,
      "grad_norm": 0.0024091172963380814,
      "kl": 0.0920485682785511,
      "learning_rate": 3.4549150281252636e-05,
      "loss": -0.03069324791431427,
      "num_tokens": 486616.0,
      "reward": 1.3125,
      "reward_std": 0.075,
      "rewards/reward_fn/mean": 1.3125,
      "rewards/reward_fn/std": 0.46982967257499697,
      "step": 125
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 71.8625,
      "completions/clipped_ratio": 0.1,
      "completions/max_length": 214.8,
      "completions/max_terminated_length": 120.2,
      "completions/mean_length": 71.8625,
      "completions/mean_terminated_length": 51.76010284423828,
      "completions/min_length": 33.4,
      "completions/min_terminated_length": 33.4,
      "epoch": 13.0,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0004440450284164399,
      "kl": 0.053088791808113454,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.229191156104207e-05,
      "num_tokens": 505905.0,
      "reward": 1.35,
      "reward_std": 0.0,
      "rewards/reward_fn/mean": 1.35,
      "rewards/reward_fn/std": 0.39928138852119444,
      "step": 130
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 84.5,
      "completions/clipped_ratio": 0.175,
      "completions/max_length": 182.8,
      "completions/max_terminated_length": 76.6,
      "completions/mean_length": 84.5,
      "completions/mean_terminated_length": 48.71715393066406,
      "completions/min_length": 38.0,
      "completions/min_terminated_length": 38.0,
      "epoch": 13.5,
      "frac_reward_zero_std": 0.95,
      "grad_norm": 0.0007964075193740427,
      "kl": 0.07870542965829372,
      "learning_rate": 2.693294185106562e-05,
      "loss": 0.0014319854788482189,
      "num_tokens": 526773.0,
      "reward": 1.3875,
      "reward_std": 0.025,
      "rewards/reward_fn/mean": 1.3875,
      "rewards/reward_fn/std": 0.38874412178993223,
      "step": 135
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 70.6375,
      "completions/clipped_ratio": 0.1,
      "completions/max_length": 219.2,
      "completions/max_terminated_length": 104.2,
      "completions/mean_length": 70.6375,
      "completions/mean_terminated_length": 50.06154937744141,
      "completions/min_length": 37.0,
      "completions/min_terminated_length": 37.0,
      "epoch": 14.0,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.00017480542010162026,
      "kl": 0.05223398320376873,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.521520506590605e-05,
      "num_tokens": 545864.0,
      "reward": 1.3,
      "reward_std": 0.0,
      "rewards/reward_fn/mean": 1.3,
      "rewards/reward_fn/std": 0.38544455766677854,
      "step": 140
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 75.7,
      "completions/clipped_ratio": 0.1125,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 134.4,
      "completions/mean_length": 75.7,
      "completions/mean_terminated_length": 52.85439682006836,
      "completions/min_length": 36.6,
      "completions/min_terminated_length": 36.6,
      "epoch": 14.5,
      "frac_reward_zero_std": 0.95,
      "grad_norm": 0.0014995499514043331,
      "kl": 0.061327320709824565,
      "learning_rate": 1.9945942635848748e-05,
      "loss": -0.039644479751586914,
      "num_tokens": 565456.0,
      "reward": 1.275,
      "reward_std": 0.05,
      "rewards/reward_fn/mean": 1.275,
      "rewards/reward_fn/std": 0.40599284172058103,
      "step": 145
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completion_length": 88.6875,
      "completions/clipped_ratio": 0.1875,
      "completions/max_length": 214.6,
      "completions/max_terminated_length": 98.8,
      "completions/mean_length": 88.6875,
      "completions/mean_terminated_length": 51.92777633666992,
      "completions/min_length": 35.0,
      "completions/min_terminated_length": 35.0,
      "epoch": 15.0,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.001117849606089294,
      "kl": 0.05267696371302009,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.003726157359779e-05,
      "num_tokens": 586207.0,
      "reward": 1.35,
      "reward_std": 0.0,
      "rewards/reward_fn/mean": 1.35,
      "rewards/reward_fn/std": 0.37160772681236265,
      "step": 150
    }
  ],
  "logging_steps": 5,
  "max_steps": 200,
  "num_input_tokens_seen": 586207,
  "num_train_epochs": 20,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}