forked from ppwwyyxx/GroupNorm-reproduce
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathlog.log
1935 lines (1935 loc) · 140 KB
/
log.log
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
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
[1219 18:00:20 @imagenet-resnet-gn.py:54] Running on 8 towers. Batch size per tower: 32
[1219 18:00:21 @ilsvrc.py:128] [ILSVRC12] Assuming directory /XXX/imagenet/val has 'train' structure.
[1219 18:00:21 @imagenet-resnet-gn.py:61] BASELR: 0.1
[1219 18:00:21 @training.py:52] [DataParallel] Training a model of 8 towers.
[1219 18:00:21 @interface.py:34] Automatically applying QueueInput on the DataFlow.
[1219 18:00:21 @interface.py:46] Automatically applying StagingInput on the DataFlow.
[1219 18:00:21 @input_source.py:219] Setting up the queue 'QueueInput/input_queue' for CPU prefetching ...
[1219 18:00:21 @training.py:112] Building graph for training tower 0 on device /gpu:0 ...
[1219 18:00:21 @registry.py:121] conv0 input: [None, 3, 230, 230]
[1219 18:00:21 @registry.py:121] conv0/gn input: [None, 64, 112, 112]
[1219 18:00:21 @registry.py:129] conv0/gn output: [None, 64, 112, 112]
[1219 18:00:21 @registry.py:129] conv0 output: [None, 64, 112, 112]
[1219 18:00:21 @registry.py:121] pool0 input: [None, 64, 114, 114]
[1219 18:00:21 @registry.py:129] pool0 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/conv1 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/conv1/gn input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/conv1/gn output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/conv1 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/conv2 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/conv2/gn input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/conv2/gn output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/conv2 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/conv3 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/conv3/gn input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/conv3/gn output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/conv3 output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/convshortcut input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block0/convshortcut/gn input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/convshortcut/gn output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block0/convshortcut output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block1/conv1 input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block1/conv1/gn input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block1/conv1/gn output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block1/conv1 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block1/conv2 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block1/conv2/gn input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block1/conv2/gn output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block1/conv2 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block1/conv3 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block1/conv3/gn input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block1/conv3/gn output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block1/conv3 output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block2/conv1 input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block2/conv1/gn input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block2/conv1/gn output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block2/conv1 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block2/conv2 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block2/conv2/gn input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block2/conv2/gn output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block2/conv2 output: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block2/conv3 input: [None, 64, 56, 56]
[1219 18:00:21 @registry.py:121] group0/block2/conv3/gn input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block2/conv3/gn output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:129] group0/block2/conv3 output: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group1/block0/conv1 input: [None, 256, 56, 56]
[1219 18:00:21 @registry.py:121] group1/block0/conv1/gn input: [None, 128, 56, 56]
[1219 18:00:21 @registry.py:129] group1/block0/conv1/gn output: [None, 128, 56, 56]
[1219 18:00:21 @registry.py:129] group1/block0/conv1 output: [None, 128, 56, 56]
[1219 18:00:21 @registry.py:121] group1/block0/conv2 input: [None, 128, 58, 58]
[1219 18:00:21 @registry.py:121] group1/block0/conv2/gn input: [None, 128, 28, 28]
[1219 18:00:21 @registry.py:129] group1/block0/conv2/gn output: [None, 128, 28, 28]
[1219 18:00:21 @registry.py:129] group1/block0/conv2 output: [None, 128, 28, 28]
[1219 18:00:21 @registry.py:121] group1/block0/conv3 input: [None, 128, 28, 28]
[1219 18:00:21 @registry.py:121] group1/block0/conv3/gn input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block0/conv3/gn output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block0/conv3 output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block0/convshortcut input: [None, 256, 56, 56]
[1219 18:00:22 @registry.py:121] group1/block0/convshortcut/gn input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block0/convshortcut/gn output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block0/convshortcut output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block1/conv1 input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block1/conv1/gn input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block1/conv1/gn output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block1/conv1 output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block1/conv2 input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block1/conv2/gn input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block1/conv2/gn output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block1/conv2 output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block1/conv3 input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block1/conv3/gn input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block1/conv3/gn output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block1/conv3 output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block2/conv1 input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block2/conv1/gn input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block2/conv1/gn output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block2/conv1 output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block2/conv2 input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block2/conv2/gn input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block2/conv2/gn output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block2/conv2 output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block2/conv3 input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block2/conv3/gn input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block2/conv3/gn output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block2/conv3 output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block3/conv1 input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block3/conv1/gn input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block3/conv1/gn output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block3/conv1 output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block3/conv2 input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block3/conv2/gn input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block3/conv2/gn output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block3/conv2 output: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block3/conv3 input: [None, 128, 28, 28]
[1219 18:00:22 @registry.py:121] group1/block3/conv3/gn input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block3/conv3/gn output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:129] group1/block3/conv3 output: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group2/block0/conv1 input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group2/block0/conv1/gn input: [None, 256, 28, 28]
[1219 18:00:22 @registry.py:129] group2/block0/conv1/gn output: [None, 256, 28, 28]
[1219 18:00:22 @registry.py:129] group2/block0/conv1 output: [None, 256, 28, 28]
[1219 18:00:22 @registry.py:121] group2/block0/conv2 input: [None, 256, 30, 30]
[1219 18:00:22 @registry.py:121] group2/block0/conv2/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block0/conv2/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block0/conv2 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block0/conv3 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block0/conv3/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block0/conv3/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block0/conv3 output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block0/convshortcut input: [None, 512, 28, 28]
[1219 18:00:22 @registry.py:121] group2/block0/convshortcut/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block0/convshortcut/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block0/convshortcut output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block1/conv1 input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block1/conv1/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block1/conv1/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block1/conv1 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block1/conv2 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block1/conv2/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block1/conv2/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block1/conv2 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block1/conv3 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block1/conv3/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block1/conv3/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block1/conv3 output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block2/conv1 input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block2/conv1/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block2/conv1/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block2/conv1 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block2/conv2 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block2/conv2/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block2/conv2/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block2/conv2 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block2/conv3 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block2/conv3/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block2/conv3/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block2/conv3 output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block3/conv1 input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block3/conv1/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block3/conv1/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block3/conv1 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block3/conv2 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block3/conv2/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block3/conv2/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block3/conv2 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block3/conv3 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block3/conv3/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block3/conv3/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block3/conv3 output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block4/conv1 input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block4/conv1/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block4/conv1/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block4/conv1 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block4/conv2 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block4/conv2/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block4/conv2/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block4/conv2 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block4/conv3 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block4/conv3/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block4/conv3/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block4/conv3 output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block5/conv1 input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block5/conv1/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block5/conv1/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block5/conv1 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block5/conv2 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block5/conv2/gn input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block5/conv2/gn output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block5/conv2 output: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block5/conv3 input: [None, 256, 14, 14]
[1219 18:00:22 @registry.py:121] group2/block5/conv3/gn input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block5/conv3/gn output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:129] group2/block5/conv3 output: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group3/block0/conv1 input: [None, 1024, 14, 14]
[1219 18:00:22 @registry.py:121] group3/block0/conv1/gn input: [None, 512, 14, 14]
[1219 18:00:22 @registry.py:129] group3/block0/conv1/gn output: [None, 512, 14, 14]
[1219 18:00:22 @registry.py:129] group3/block0/conv1 output: [None, 512, 14, 14]
[1219 18:00:22 @registry.py:121] group3/block0/conv2 input: [None, 512, 16, 16]
[1219 18:00:22 @registry.py:121] group3/block0/conv2/gn input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block0/conv2/gn output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block0/conv2 output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block0/conv3 input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block0/conv3/gn input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block0/conv3/gn output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block0/conv3 output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block0/convshortcut input: [None, 1024, 14, 14]
[1219 18:00:23 @registry.py:121] group3/block0/convshortcut/gn input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block0/convshortcut/gn output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block0/convshortcut output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block1/conv1 input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block1/conv1/gn input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block1/conv1/gn output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block1/conv1 output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block1/conv2 input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block1/conv2/gn input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block1/conv2/gn output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block1/conv2 output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block1/conv3 input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block1/conv3/gn input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block1/conv3/gn output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block1/conv3 output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block2/conv1 input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block2/conv1/gn input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block2/conv1/gn output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block2/conv1 output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block2/conv2 input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block2/conv2/gn input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block2/conv2/gn output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block2/conv2 output: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block2/conv3 input: [None, 512, 7, 7]
[1219 18:00:23 @registry.py:121] group3/block2/conv3/gn input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block2/conv3/gn output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] group3/block2/conv3 output: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:121] gap input: [None, 2048, 7, 7]
[1219 18:00:23 @registry.py:129] gap output: [None, 2048]
[1219 18:00:23 @registry.py:121] linear input: [None, 2048]
[1219 18:00:23 @registry.py:129] linear output: [None, 1000]
[1219 18:00:23 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:00:23 @regularize.py:20] The following tensors will be regularized: conv0/W:0, conv0/gn/beta:0, conv0/gn/gamma:0, group0/block0/conv1/W:0, group0/block0/conv1/gn/beta:0, group0/block0/conv1/gn/gamma:0, group0/block0/conv2/W:0, group0/block0/conv2/gn/beta:0, group0/block0/conv2/gn/gamma:0, group0/block0/conv3/W:0, group0/block0/conv3/gn/beta:0, group0/block0/conv3/gn/gamma:0, group0/block0/convshortcut/W:0, group0/block0/convshortcut/gn/beta:0, group0/block0/convshortcut/gn/gamma:0, group0/block1/conv1/W:0, group0/block1/conv1/gn/beta:0, group0/block1/conv1/gn/gamma:0, group0/block1/conv2/W:0, group0/block1/conv2/gn/beta:0, group0/block1/conv2/gn/gamma:0, group0/block1/conv3/W:0, group0/block1/conv3/gn/beta:0, group0/block1/conv3/gn/gamma:0, group0/block2/conv1/W:0, group0/block2/conv1/gn/beta:0, group0/block2/conv1/gn/gamma:0, group0/block2/conv2/W:0, group0/block2/conv2/gn/beta:0, group0/block2/conv2/gn/gamma:0, group0/block2/conv3/W:0, group0/block2/conv3/gn/beta:0, group0/block2/conv3/gn/gamma:0, group1/block0/conv1/W:0, group1/block0/conv1/gn/beta:0, group1/block0/conv1/gn/gamma:0, group1/block0/conv2/W:0, group1/block0/conv2/gn/beta:0, group1/block0/conv2/gn/gamma:0, group1/block0/conv3/W:0, group1/block0/conv3/gn/beta:0, group1/block0/conv3/gn/gamma:0, group1/block0/convshortcut/W:0, group1/block0/convshortcut/gn/beta:0, group1/block0/convshortcut/gn/gamma:0, group1/block1/conv1/W:0, group1/block1/conv1/gn/beta:0, group1/block1/conv1/gn/gamma:0, group1/block1/conv2/W:0, group1/block1/conv2/gn/beta:0, group1/block1/conv2/gn/gamma:0, group1/block1/conv3/W:0, group1/block1/conv3/gn/beta:0, group1/block1/conv3/gn/gamma:0, group1/block2/conv1/W:0, group1/block2/conv1/gn/beta:0, group1/block2/conv1/gn/gamma:0, group1/block2/conv2/W:0, group1/block2/conv2/gn/beta:0, group1/block2/conv2/gn/gamma:0, group1/block2/conv3/W:0, group1/block2/conv3/gn/beta:0, group1/block2/conv3/gn/gamma:0, group1/block3/conv1/W:0, group1/block3/conv1/gn/beta:0, group1/block3/conv1/gn/gamma:0, group1/block3/conv2/W:0, group1/block3/conv2/gn/beta:0, group1/block3/conv2/gn/gamma:0, group1/block3/conv3/W:0, group1/block3/conv3/gn/beta:0, group1/block3/conv3/gn/gamma:0, group2/block0/conv1/W:0, group2/block0/conv1/gn/beta:0, group2/block0/conv1/gn/gamma:0, group2/block0/conv2/W:0, group2/block0/conv2/gn/beta:0, group2/block0/conv2/gn/gamma:0, group2/block0/conv3/W:0, group2/block0/conv3/gn/beta:0, group2/block0/conv3/gn/gamma:0, group2/block0/convshortcut/W:0, group2/block0/convshortcut/gn/beta:0, group2/block0/convshortcut/gn/gamma:0, group2/block1/conv1/W:0, group2/block1/conv1/gn/beta:0, group2/block1/conv1/gn/gamma:0, group2/block1/conv2/W:0, group2/block1/conv2/gn/beta:0, group2/block1/conv2/gn/gamma:0, group2/block1/conv3/W:0, group2/block1/conv3/gn/beta:0, group2/block1/conv3/gn/gamma:0, group2/block2/conv1/W:0, group2/block2/conv1/gn/beta:0, group2/block2/conv1/gn/gamma:0, group2/block2/conv2/W:0, group2/block2/conv2/gn/beta:0, group2/block2/conv2/gn/gamma:0, group2/block2/conv3/W:0, group2/block2/conv3/gn/beta:0, group2/block2/conv3/gn/gamma:0, group2/block3/conv1/W:0, group2/block3/conv1/gn/beta:0, group2/block3/conv1/gn/gamma:0, group2/block3/conv2/W:0, group2/block3/conv2/gn/beta:0, group2/block3/conv2/gn/gamma:0, group2/block3/conv3/W:0, group2/block3/conv3/gn/beta:0, group2/block3/conv3/gn/gamma:0, group2/block4/conv1/W:0, group2/block4/conv1/gn/beta:0, group2/block4/conv1/gn/gamma:0, group2/block4/conv2/W:0, group2/block4/conv2/gn/beta:0, group2/block4/conv2/gn/gamma:0, group2/block4/conv3/W:0, group2/block4/conv3/gn/beta:0, group2/block4/conv3/gn/gamma:0, group2/block5/conv1/W:0, group2/block5/conv1/gn/beta:0, group2/block5/conv1/gn/gamma:0, group2/block5/conv2/W:0, group2/block5/conv2/gn/beta:0, group2/block5/conv2/gn/gamma:0, group2/block5/conv3/W:0, group2/block5/conv3/gn/beta:0, group2/block5/conv3/gn/gamma:0, group3/block0/conv1/W:0, group3/block0/conv1/gn/beta:0, group3/block0/conv1/gn/gamma:0, group3/block0/conv2/W:0, group3/block0/conv2/gn/beta:0, group3/block0/conv2/gn/gamma:0, group3/block0/conv3/W:0, group3/block0/conv3/gn/beta:0, group3/block0/conv3/gn/gamma:0, group3/block0/convshortcut/W:0, group3/block0/convshortcut/gn/beta:0, group3/block0/convshortcut/gn/gamma:0, group3/block1/conv1/W:0, group3/block1/conv1/gn/beta:0, group3/block1/conv1/gn/gamma:0, group3/block1/conv2/W:0, group3/block1/conv2/gn/beta:0, group3/block1/conv2/gn/gamma:0, group3/block1/conv3/W:0, group3/block1/conv3/gn/beta:0, group3/block1/conv3/gn/gamma:0, group3/block2/conv1/W:0, group3/block2/conv1/gn/beta:0, group3/block2/conv1/gn/gamma:0, group3/block2/conv2/W:0, group3/block2/conv2/gn/beta:0, group3/block2/conv2/gn/gamma:0, group3/block2/conv3/W:0, group3/block2/conv3/gn/beta:0, group3/block2/conv3/gn/gamma:0, linear/W:0
[1219 18:00:28 @training.py:112] Building graph for training tower 1 on device /gpu:1 ...
[1219 18:00:30 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:00:34 @training.py:112] Building graph for training tower 2 on device /gpu:2 ...
[1219 18:00:36 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:00:41 @training.py:112] Building graph for training tower 3 on device /gpu:3 ...
[1219 18:00:43 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:00:47 @training.py:112] Building graph for training tower 4 on device /gpu:4 ...
[1219 18:00:49 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:00:54 @training.py:112] Building graph for training tower 5 on device /gpu:5 ...
[1219 18:00:56 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:01:00 @training.py:112] Building graph for training tower 6 on device /gpu:6 ...
[1219 18:01:02 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:01:07 @training.py:112] Building graph for training tower 7 on device /gpu:7 ...
[1219 18:01:08 @regularize.py:95] regularize_cost() found 160 variables to regularize.
[1219 18:01:13 @utils.py:364] Will pack 161 gradients of total dimension=25557032 into 8 splits.
[1219 18:01:23 @training.py:345] 'sync_variables_from_main_tower' includes 2254 operations.
[1219 18:01:23 @model_utils.py:64] Trainable Variables:
name shape dim
------------------------------------- ------------------ -------
conv0/W:0 [7, 7, 3, 64] 9408
conv0/gn/beta:0 [64] 64
conv0/gn/gamma:0 [64] 64
group0/block0/conv1/W:0 [1, 1, 64, 64] 4096
group0/block0/conv1/gn/beta:0 [64] 64
group0/block0/conv1/gn/gamma:0 [64] 64
group0/block0/conv2/W:0 [3, 3, 64, 64] 36864
group0/block0/conv2/gn/beta:0 [64] 64
group0/block0/conv2/gn/gamma:0 [64] 64
group0/block0/conv3/W:0 [1, 1, 64, 256] 16384
group0/block0/conv3/gn/beta:0 [256] 256
group0/block0/conv3/gn/gamma:0 [256] 256
group0/block0/convshortcut/W:0 [1, 1, 64, 256] 16384
group0/block0/convshortcut/gn/beta:0 [256] 256
group0/block0/convshortcut/gn/gamma:0 [256] 256
group0/block1/conv1/W:0 [1, 1, 256, 64] 16384
group0/block1/conv1/gn/beta:0 [64] 64
group0/block1/conv1/gn/gamma:0 [64] 64
group0/block1/conv2/W:0 [3, 3, 64, 64] 36864
group0/block1/conv2/gn/beta:0 [64] 64
group0/block1/conv2/gn/gamma:0 [64] 64
group0/block1/conv3/W:0 [1, 1, 64, 256] 16384
group0/block1/conv3/gn/beta:0 [256] 256
group0/block1/conv3/gn/gamma:0 [256] 256
group0/block2/conv1/W:0 [1, 1, 256, 64] 16384
group0/block2/conv1/gn/beta:0 [64] 64
group0/block2/conv1/gn/gamma:0 [64] 64
group0/block2/conv2/W:0 [3, 3, 64, 64] 36864
group0/block2/conv2/gn/beta:0 [64] 64
group0/block2/conv2/gn/gamma:0 [64] 64
group0/block2/conv3/W:0 [1, 1, 64, 256] 16384
group0/block2/conv3/gn/beta:0 [256] 256
group0/block2/conv3/gn/gamma:0 [256] 256
group1/block0/conv1/W:0 [1, 1, 256, 128] 32768
group1/block0/conv1/gn/beta:0 [128] 128
group1/block0/conv1/gn/gamma:0 [128] 128
group1/block0/conv2/W:0 [3, 3, 128, 128] 147456
group1/block0/conv2/gn/beta:0 [128] 128
group1/block0/conv2/gn/gamma:0 [128] 128
group1/block0/conv3/W:0 [1, 1, 128, 512] 65536
group1/block0/conv3/gn/beta:0 [512] 512
group1/block0/conv3/gn/gamma:0 [512] 512
group1/block0/convshortcut/W:0 [1, 1, 256, 512] 131072
group1/block0/convshortcut/gn/beta:0 [512] 512
group1/block0/convshortcut/gn/gamma:0 [512] 512
group1/block1/conv1/W:0 [1, 1, 512, 128] 65536
group1/block1/conv1/gn/beta:0 [128] 128
group1/block1/conv1/gn/gamma:0 [128] 128
group1/block1/conv2/W:0 [3, 3, 128, 128] 147456
group1/block1/conv2/gn/beta:0 [128] 128
group1/block1/conv2/gn/gamma:0 [128] 128
group1/block1/conv3/W:0 [1, 1, 128, 512] 65536
group1/block1/conv3/gn/beta:0 [512] 512
group1/block1/conv3/gn/gamma:0 [512] 512
group1/block2/conv1/W:0 [1, 1, 512, 128] 65536
group1/block2/conv1/gn/beta:0 [128] 128
group1/block2/conv1/gn/gamma:0 [128] 128
group1/block2/conv2/W:0 [3, 3, 128, 128] 147456
group1/block2/conv2/gn/beta:0 [128] 128
group1/block2/conv2/gn/gamma:0 [128] 128
group1/block2/conv3/W:0 [1, 1, 128, 512] 65536
group1/block2/conv3/gn/beta:0 [512] 512
group1/block2/conv3/gn/gamma:0 [512] 512
group1/block3/conv1/W:0 [1, 1, 512, 128] 65536
group1/block3/conv1/gn/beta:0 [128] 128
group1/block3/conv1/gn/gamma:0 [128] 128
group1/block3/conv2/W:0 [3, 3, 128, 128] 147456
group1/block3/conv2/gn/beta:0 [128] 128
group1/block3/conv2/gn/gamma:0 [128] 128
group1/block3/conv3/W:0 [1, 1, 128, 512] 65536
group1/block3/conv3/gn/beta:0 [512] 512
group1/block3/conv3/gn/gamma:0 [512] 512
group2/block0/conv1/W:0 [1, 1, 512, 256] 131072
group2/block0/conv1/gn/beta:0 [256] 256
group2/block0/conv1/gn/gamma:0 [256] 256
group2/block0/conv2/W:0 [3, 3, 256, 256] 589824
group2/block0/conv2/gn/beta:0 [256] 256
group2/block0/conv2/gn/gamma:0 [256] 256
group2/block0/conv3/W:0 [1, 1, 256, 1024] 262144
group2/block0/conv3/gn/beta:0 [1024] 1024
group2/block0/conv3/gn/gamma:0 [1024] 1024
group2/block0/convshortcut/W:0 [1, 1, 512, 1024] 524288
group2/block0/convshortcut/gn/beta:0 [1024] 1024
group2/block0/convshortcut/gn/gamma:0 [1024] 1024
group2/block1/conv1/W:0 [1, 1, 1024, 256] 262144
group2/block1/conv1/gn/beta:0 [256] 256
group2/block1/conv1/gn/gamma:0 [256] 256
group2/block1/conv2/W:0 [3, 3, 256, 256] 589824
group2/block1/conv2/gn/beta:0 [256] 256
group2/block1/conv2/gn/gamma:0 [256] 256
group2/block1/conv3/W:0 [1, 1, 256, 1024] 262144
group2/block1/conv3/gn/beta:0 [1024] 1024
group2/block1/conv3/gn/gamma:0 [1024] 1024
group2/block2/conv1/W:0 [1, 1, 1024, 256] 262144
group2/block2/conv1/gn/beta:0 [256] 256
group2/block2/conv1/gn/gamma:0 [256] 256
group2/block2/conv2/W:0 [3, 3, 256, 256] 589824
group2/block2/conv2/gn/beta:0 [256] 256
group2/block2/conv2/gn/gamma:0 [256] 256
group2/block2/conv3/W:0 [1, 1, 256, 1024] 262144
group2/block2/conv3/gn/beta:0 [1024] 1024
group2/block2/conv3/gn/gamma:0 [1024] 1024
group2/block3/conv1/W:0 [1, 1, 1024, 256] 262144
group2/block3/conv1/gn/beta:0 [256] 256
group2/block3/conv1/gn/gamma:0 [256] 256
group2/block3/conv2/W:0 [3, 3, 256, 256] 589824
group2/block3/conv2/gn/beta:0 [256] 256
group2/block3/conv2/gn/gamma:0 [256] 256
group2/block3/conv3/W:0 [1, 1, 256, 1024] 262144
group2/block3/conv3/gn/beta:0 [1024] 1024
group2/block3/conv3/gn/gamma:0 [1024] 1024
group2/block4/conv1/W:0 [1, 1, 1024, 256] 262144
group2/block4/conv1/gn/beta:0 [256] 256
group2/block4/conv1/gn/gamma:0 [256] 256
group2/block4/conv2/W:0 [3, 3, 256, 256] 589824
group2/block4/conv2/gn/beta:0 [256] 256
group2/block4/conv2/gn/gamma:0 [256] 256
group2/block4/conv3/W:0 [1, 1, 256, 1024] 262144
group2/block4/conv3/gn/beta:0 [1024] 1024
group2/block4/conv3/gn/gamma:0 [1024] 1024
group2/block5/conv1/W:0 [1, 1, 1024, 256] 262144
group2/block5/conv1/gn/beta:0 [256] 256
group2/block5/conv1/gn/gamma:0 [256] 256
group2/block5/conv2/W:0 [3, 3, 256, 256] 589824
group2/block5/conv2/gn/beta:0 [256] 256
group2/block5/conv2/gn/gamma:0 [256] 256
group2/block5/conv3/W:0 [1, 1, 256, 1024] 262144
group2/block5/conv3/gn/beta:0 [1024] 1024
group2/block5/conv3/gn/gamma:0 [1024] 1024
group3/block0/conv1/W:0 [1, 1, 1024, 512] 524288
group3/block0/conv1/gn/beta:0 [512] 512
group3/block0/conv1/gn/gamma:0 [512] 512
group3/block0/conv2/W:0 [3, 3, 512, 512] 2359296
group3/block0/conv2/gn/beta:0 [512] 512
group3/block0/conv2/gn/gamma:0 [512] 512
group3/block0/conv3/W:0 [1, 1, 512, 2048] 1048576
group3/block0/conv3/gn/beta:0 [2048] 2048
group3/block0/conv3/gn/gamma:0 [2048] 2048
group3/block0/convshortcut/W:0 [1, 1, 1024, 2048] 2097152
group3/block0/convshortcut/gn/beta:0 [2048] 2048
group3/block0/convshortcut/gn/gamma:0 [2048] 2048
group3/block1/conv1/W:0 [1, 1, 2048, 512] 1048576
group3/block1/conv1/gn/beta:0 [512] 512
group3/block1/conv1/gn/gamma:0 [512] 512
group3/block1/conv2/W:0 [3, 3, 512, 512] 2359296
group3/block1/conv2/gn/beta:0 [512] 512
group3/block1/conv2/gn/gamma:0 [512] 512
group3/block1/conv3/W:0 [1, 1, 512, 2048] 1048576
group3/block1/conv3/gn/beta:0 [2048] 2048
group3/block1/conv3/gn/gamma:0 [2048] 2048
group3/block2/conv1/W:0 [1, 1, 2048, 512] 1048576
group3/block2/conv1/gn/beta:0 [512] 512
group3/block2/conv1/gn/gamma:0 [512] 512
group3/block2/conv2/W:0 [3, 3, 512, 512] 2359296
group3/block2/conv2/gn/beta:0 [512] 512
group3/block2/conv2/gn/gamma:0 [512] 512
group3/block2/conv3/W:0 [1, 1, 512, 2048] 1048576
group3/block2/conv3/gn/beta:0 [2048] 2048
group3/block2/conv3/gn/gamma:0 [2048] 2048
linear/W:0 [2048, 1000] 2048000
linear/b:0 [1000] 1000
Total #vars=161, #params=25557032, size=97.49MB
[1219 18:01:23 @base.py:209] Setup callbacks graph ...
[1219 18:01:25 @input_source.py:219] Setting up the queue 'DataParallelInferenceRunner/QueueInput/input_queue' for CPU prefetching ...
[1219 18:01:25 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower0' on device /gpu:0 ...
[1219 18:01:26 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower1' on device /gpu:1 with variable scope 'tower1'...
[1219 18:01:28 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower2' on device /gpu:2 with variable scope 'tower2'...
[1219 18:01:29 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower3' on device /gpu:3 with variable scope 'tower3'...
[1219 18:01:30 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower4' on device /gpu:4 with variable scope 'tower4'...
[1219 18:01:31 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower5' on device /gpu:5 with variable scope 'tower5'...
[1219 18:01:32 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower6' on device /gpu:6 with variable scope 'tower6'...
[1219 18:01:33 @inference_runner.py:239] [InferenceRunner] Building tower 'InferenceTower7' on device /gpu:7 with variable scope 'tower7'...
[1219 18:01:34 @summary.py:46] [MovingAverageSummary] 4 operations in collection 'MOVING_SUMMARY_OPS' will be run with session hooks.
[1219 18:01:34 @summary.py:93] Summarizing collection 'summaries' of size 7.
[1219 18:01:49 @base.py:230] Creating the session ...
[1219 18:02:24 @base.py:236] Initializing the session ...
[1219 18:02:24 @base.py:243] Graph Finalized.
[1219 18:02:25 @concurrency.py:37] Starting EnqueueThread QueueInput/input_queue ...
[1219 18:02:25 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 18:02:37 @param.py:158] [HyperParamSetter] At global_step=0, learning_rate is set to 0.100000
[1219 18:02:38 @concurrency.py:37] Starting EnqueueThread DataParallelInferenceRunner/QueueInput/input_queue ...
[1219 18:02:38 @inference_runner.py:101] [InferenceRunner] Will eval 1563 iterations
[1219 18:02:39 @base.py:275] Start Epoch 1 ...
[1219 18:02:39 @input_source.py:550] Pre-filling StagingArea ...
[1219 18:02:40 @input_source.py:554] 1 element was put into StagingArea on each tower.
[1219 18:22:11 @base.py:285] Epoch 1 (global_step 5004) finished, time:19 minutes 31 seconds.
[1219 18:22:11 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 18:22:12 @saver.py:77] Model saved to train_log/ResNet50-GN/model-5004.
[1219 18:22:12 @misc.py:109] Estimated Time Left: 1 day 9 hours 57 minutes 15 seconds
[1219 18:22:59 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 50
[1219 18:22:59 @monitor.py:459] QueueInput/queue_size: 49.461
[1219 18:22:59 @monitor.py:459] l2_regularize_loss: 1.2797
[1219 18:22:59 @monitor.py:459] learning_rate: 0.1
[1219 18:22:59 @monitor.py:459] train-error-top1: 0.86814
[1219 18:22:59 @monitor.py:459] train-error-top5: 0.71173
[1219 18:22:59 @monitor.py:459] val-error-top1: 0.8546
[1219 18:22:59 @monitor.py:459] val-error-top5: 0.67004
[1219 18:22:59 @monitor.py:459] xentropy-loss: 4.7791
[1219 18:22:59 @group.py:48] Callbacks took 48.057 sec in total. DataParallelInferenceRunner: 46.1 seconds
[1219 18:22:59 @base.py:275] Start Epoch 2 ...
[1219 18:38:29 @base.py:285] Epoch 2 (global_step 10008) finished, time:15 minutes 30 seconds.
[1219 18:38:29 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 18:38:30 @saver.py:77] Model saved to train_log/ResNet50-GN/model-10008.
[1219 18:38:30 @misc.py:109] Estimated Time Left: 1 day 6 hours 47 minutes 45 seconds
[1219 18:39:07 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 29.469
[1219 18:39:07 @monitor.py:459] QueueInput/queue_size: 48.011
[1219 18:39:07 @monitor.py:459] l2_regularize_loss: 0.88098
[1219 18:39:07 @monitor.py:459] learning_rate: 0.1
[1219 18:39:07 @monitor.py:459] train-error-top1: 0.77193
[1219 18:39:07 @monitor.py:459] train-error-top5: 0.54807
[1219 18:39:07 @monitor.py:459] val-error-top1: 0.72072
[1219 18:39:07 @monitor.py:459] val-error-top5: 0.4778
[1219 18:39:07 @monitor.py:459] xentropy-loss: 3.9111
[1219 18:39:07 @group.py:48] Callbacks took 37.733 sec in total. DataParallelInferenceRunner: 36.9 seconds
[1219 18:39:07 @base.py:275] Start Epoch 3 ...
[1219 18:54:38 @base.py:285] Epoch 3 (global_step 15012) finished, time:15 minutes 30 seconds.
[1219 18:54:38 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 18:54:38 @saver.py:77] Model saved to train_log/ResNet50-GN/model-15012.
[1219 18:54:38 @misc.py:109] Estimated Time Left: 1 day 5 hours 28 minutes 44 seconds
[1219 18:55:14 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.935
[1219 18:55:14 @monitor.py:459] QueueInput/queue_size: 49.552
[1219 18:55:14 @monitor.py:459] l2_regularize_loss: 0.83173
[1219 18:55:14 @monitor.py:459] learning_rate: 0.1
[1219 18:55:14 @monitor.py:459] train-error-top1: 0.7008
[1219 18:55:14 @monitor.py:459] train-error-top5: 0.45408
[1219 18:55:14 @monitor.py:459] val-error-top1: 0.63336
[1219 18:55:14 @monitor.py:459] val-error-top5: 0.3745
[1219 18:55:14 @monitor.py:459] xentropy-loss: 3.3647
[1219 18:55:14 @group.py:48] Callbacks took 36.523 sec in total. DataParallelInferenceRunner: 35.6 seconds
[1219 18:55:14 @base.py:275] Start Epoch 4 ...
[1219 19:10:51 @base.py:285] Epoch 4 (global_step 20016) finished, time:15 minutes 37 seconds.
[1219 19:10:51 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 19:10:52 @saver.py:77] Model saved to train_log/ResNet50-GN/model-20016.
[1219 19:10:52 @misc.py:109] Estimated Time Left: 1 day 4 hours 43 minutes 20 seconds
[1219 19:11:28 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.442
[1219 19:11:28 @monitor.py:459] QueueInput/queue_size: 49.912
[1219 19:11:28 @monitor.py:459] l2_regularize_loss: 0.87067
[1219 19:11:28 @monitor.py:459] learning_rate: 0.1
[1219 19:11:28 @monitor.py:459] train-error-top1: 0.62967
[1219 19:11:28 @monitor.py:459] train-error-top5: 0.37793
[1219 19:11:28 @monitor.py:459] val-error-top1: 0.58458
[1219 19:11:28 @monitor.py:459] val-error-top5: 0.31886
[1219 19:11:28 @monitor.py:459] xentropy-loss: 2.8733
[1219 19:11:28 @group.py:48] Callbacks took 36.817 sec in total. DataParallelInferenceRunner: 36 seconds
[1219 19:11:28 @base.py:275] Start Epoch 5 ...
[1219 19:27:04 @base.py:285] Epoch 5 (global_step 25020) finished, time:15 minutes 36 seconds.
[1219 19:27:04 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 19:27:06 @saver.py:77] Model saved to train_log/ResNet50-GN/model-25020.
[1219 19:27:06 @misc.py:109] Estimated Time Left: 1 day 4 hours 9 minutes 29 seconds
[1219 19:27:41 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.526
[1219 19:27:41 @monitor.py:459] QueueInput/queue_size: 49.532
[1219 19:27:41 @monitor.py:459] l2_regularize_loss: 0.91692
[1219 19:27:41 @monitor.py:459] learning_rate: 0.1
[1219 19:27:41 @monitor.py:459] train-error-top1: 0.58465
[1219 19:27:41 @monitor.py:459] train-error-top5: 0.32825
[1219 19:27:41 @monitor.py:459] val-error-top1: 0.55642
[1219 19:27:41 @monitor.py:459] val-error-top5: 0.29786
[1219 19:27:41 @monitor.py:459] xentropy-loss: 2.6878
[1219 19:27:41 @group.py:48] Callbacks took 37.119 sec in total. DataParallelInferenceRunner: 35.9 seconds
[1219 19:27:41 @base.py:275] Start Epoch 6 ...
[1219 19:43:12 @base.py:285] Epoch 6 (global_step 30024) finished, time:15 minutes 30 seconds.
[1219 19:43:12 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 19:43:13 @saver.py:77] Model saved to train_log/ResNet50-GN/model-30024.
[1219 19:43:13 @misc.py:109] Estimated Time Left: 1 day 2 hours 44 minutes
[1219 19:43:49 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.46
[1219 19:43:49 @monitor.py:459] QueueInput/queue_size: 48.997
[1219 19:43:49 @monitor.py:459] l2_regularize_loss: 0.95708
[1219 19:43:49 @monitor.py:459] learning_rate: 0.1
[1219 19:43:49 @monitor.py:459] train-error-top1: 0.57648
[1219 19:43:49 @monitor.py:459] train-error-top5: 0.33639
[1219 19:43:49 @monitor.py:459] val-error-top1: 0.54158
[1219 19:43:49 @monitor.py:459] val-error-top5: 0.27638
[1219 19:43:49 @monitor.py:459] xentropy-loss: 2.7237
[1219 19:43:49 @group.py:48] Callbacks took 36.890 sec in total. DataParallelInferenceRunner: 36 seconds
[1219 19:43:49 @base.py:275] Start Epoch 7 ...
[1219 19:59:21 @base.py:285] Epoch 7 (global_step 35028) finished, time:15 minutes 31 seconds.
[1219 19:59:21 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 19:59:21 @saver.py:77] Model saved to train_log/ResNet50-GN/model-35028.
[1219 19:59:21 @misc.py:109] Estimated Time Left: 1 day 2 hours 24 minutes 50 seconds
[1219 19:59:56 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.562
[1219 19:59:56 @monitor.py:459] QueueInput/queue_size: 48.651
[1219 19:59:56 @monitor.py:459] l2_regularize_loss: 0.98861
[1219 19:59:56 @monitor.py:459] learning_rate: 0.1
[1219 19:59:56 @monitor.py:459] train-error-top1: 0.56901
[1219 19:59:56 @monitor.py:459] train-error-top5: 0.32027
[1219 19:59:56 @monitor.py:459] val-error-top1: 0.52586
[1219 19:59:56 @monitor.py:459] val-error-top5: 0.26476
[1219 19:59:56 @monitor.py:459] xentropy-loss: 2.564
[1219 19:59:56 @group.py:48] Callbacks took 35.628 sec in total. DataParallelInferenceRunner: 34.9 seconds
[1219 19:59:56 @base.py:275] Start Epoch 8 ...
[1219 20:15:28 @base.py:285] Epoch 8 (global_step 40032) finished, time:15 minutes 32 seconds.
[1219 20:15:28 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 20:15:29 @saver.py:77] Model saved to train_log/ResNet50-GN/model-40032.
[1219 20:15:29 @misc.py:109] Estimated Time Left: 1 day 2 hours 8 minutes 25 seconds
[1219 20:16:03 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.157
[1219 20:16:03 @monitor.py:459] QueueInput/queue_size: 48.999
[1219 20:16:03 @monitor.py:459] l2_regularize_loss: 1.0137
[1219 20:16:03 @monitor.py:459] learning_rate: 0.1
[1219 20:16:03 @monitor.py:459] train-error-top1: 0.59191
[1219 20:16:03 @monitor.py:459] train-error-top5: 0.32787
[1219 20:16:03 @monitor.py:459] val-error-top1: 0.50692
[1219 20:16:03 @monitor.py:459] val-error-top5: 0.24922
[1219 20:16:03 @monitor.py:459] xentropy-loss: 2.6461
[1219 20:16:03 @group.py:48] Callbacks took 34.187 sec in total. DataParallelInferenceRunner: 33.3 seconds
[1219 20:16:03 @base.py:275] Start Epoch 9 ...
[1219 20:31:36 @base.py:285] Epoch 9 (global_step 45036) finished, time:15 minutes 33 seconds.
[1219 20:31:36 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 20:31:37 @saver.py:77] Model saved to train_log/ResNet50-GN/model-45036.
[1219 20:31:37 @misc.py:109] Estimated Time Left: 1 day 1 hour 50 minutes 21 seconds
[1219 20:32:11 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.715
[1219 20:32:11 @monitor.py:459] QueueInput/queue_size: 49.825
[1219 20:32:11 @monitor.py:459] l2_regularize_loss: 1.0338
[1219 20:32:11 @monitor.py:459] learning_rate: 0.1
[1219 20:32:11 @monitor.py:459] train-error-top1: 0.56983
[1219 20:32:11 @monitor.py:459] train-error-top5: 0.30268
[1219 20:32:11 @monitor.py:459] val-error-top1: 0.50174
[1219 20:32:11 @monitor.py:459] val-error-top5: 0.24244
[1219 20:32:11 @monitor.py:459] xentropy-loss: 2.5496
[1219 20:32:11 @group.py:48] Callbacks took 35.239 sec in total. DataParallelInferenceRunner: 34.3 seconds
[1219 20:32:11 @base.py:275] Start Epoch 10 ...
[1219 20:47:40 @base.py:285] Epoch 10 (global_step 50040) finished, time:15 minutes 28 seconds.
[1219 20:47:40 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 20:47:41 @saver.py:77] Model saved to train_log/ResNet50-GN/model-50040.
[1219 20:47:41 @misc.py:109] Estimated Time Left: 1 day 1 hour 31 minutes 3 seconds
[1219 20:48:16 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.166
[1219 20:48:16 @monitor.py:459] QueueInput/queue_size: 49.828
[1219 20:48:16 @monitor.py:459] l2_regularize_loss: 1.0499
[1219 20:48:16 @monitor.py:459] learning_rate: 0.1
[1219 20:48:16 @monitor.py:459] train-error-top1: 0.5498
[1219 20:48:16 @monitor.py:459] train-error-top5: 0.30908
[1219 20:48:16 @monitor.py:459] val-error-top1: 0.4873
[1219 20:48:16 @monitor.py:459] val-error-top5: 0.2309
[1219 20:48:16 @monitor.py:459] xentropy-loss: 2.5269
[1219 20:48:16 @group.py:48] Callbacks took 36.483 sec in total. DataParallelInferenceRunner: 35.5 seconds
[1219 20:48:16 @base.py:275] Start Epoch 11 ...
[1219 21:03:47 @base.py:285] Epoch 11 (global_step 55044) finished, time:15 minutes 31 seconds.
[1219 21:03:47 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 21:03:48 @saver.py:77] Model saved to train_log/ResNet50-GN/model-55044.
[1219 21:03:48 @misc.py:109] Estimated Time Left: 1 day 1 hour 14 minutes 59 seconds
[1219 21:04:24 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.633
[1219 21:04:24 @monitor.py:459] QueueInput/queue_size: 49.935
[1219 21:04:24 @monitor.py:459] l2_regularize_loss: 1.0631
[1219 21:04:24 @monitor.py:459] learning_rate: 0.1
[1219 21:04:24 @monitor.py:459] train-error-top1: 0.53359
[1219 21:04:24 @monitor.py:459] train-error-top5: 0.28452
[1219 21:04:24 @monitor.py:459] val-error-top1: 0.48854
[1219 21:04:24 @monitor.py:459] val-error-top5: 0.23106
[1219 21:04:24 @monitor.py:459] xentropy-loss: 2.4048
[1219 21:04:24 @group.py:48] Callbacks took 37.206 sec in total. DataParallelInferenceRunner: 36 seconds
[1219 21:04:24 @base.py:275] Start Epoch 12 ...
[1219 21:19:57 @base.py:285] Epoch 12 (global_step 60048) finished, time:15 minutes 32 seconds.
[1219 21:19:57 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 21:19:58 @saver.py:77] Model saved to train_log/ResNet50-GN/model-60048.
[1219 21:19:58 @misc.py:109] Estimated Time Left: 1 day 59 minutes 12 seconds
[1219 21:20:34 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.712
[1219 21:20:34 @monitor.py:459] QueueInput/queue_size: 50
[1219 21:20:34 @monitor.py:459] l2_regularize_loss: 1.0736
[1219 21:20:34 @monitor.py:459] learning_rate: 0.1
[1219 21:20:34 @monitor.py:459] train-error-top1: 0.53609
[1219 21:20:34 @monitor.py:459] train-error-top5: 0.28247
[1219 21:20:34 @monitor.py:459] val-error-top1: 0.47426
[1219 21:20:34 @monitor.py:459] val-error-top5: 0.2221
[1219 21:20:34 @monitor.py:459] xentropy-loss: 2.375
[1219 21:20:34 @group.py:48] Callbacks took 36.893 sec in total. DataParallelInferenceRunner: 36 seconds
[1219 21:20:34 @base.py:275] Start Epoch 13 ...
[1219 21:36:03 @base.py:285] Epoch 13 (global_step 65052) finished, time:15 minutes 29 seconds.
[1219 21:36:03 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 21:36:04 @saver.py:77] Model saved to train_log/ResNet50-GN/model-65052.
[1219 21:36:04 @misc.py:109] Estimated Time Left: 1 day 42 minutes 47 seconds
[1219 21:36:41 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.932
[1219 21:36:41 @monitor.py:459] QueueInput/queue_size: 48.953
[1219 21:36:41 @monitor.py:459] l2_regularize_loss: 1.0819
[1219 21:36:41 @monitor.py:459] learning_rate: 0.1
[1219 21:36:41 @monitor.py:459] train-error-top1: 0.54885
[1219 21:36:41 @monitor.py:459] train-error-top5: 0.30349
[1219 21:36:41 @monitor.py:459] val-error-top1: 0.47376
[1219 21:36:41 @monitor.py:459] val-error-top5: 0.2207
[1219 21:36:41 @monitor.py:459] xentropy-loss: 2.4507
[1219 21:36:41 @group.py:48] Callbacks took 38.089 sec in total. DataParallelInferenceRunner: 36.8 seconds
[1219 21:36:41 @base.py:275] Start Epoch 14 ...
[1219 21:52:13 @base.py:285] Epoch 14 (global_step 70056) finished, time:15 minutes 31 seconds.
[1219 21:52:13 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 21:52:14 @saver.py:77] Model saved to train_log/ResNet50-GN/model-70056.
[1219 21:52:14 @misc.py:109] Estimated Time Left: 1 day 27 minutes 10 seconds
[1219 21:52:49 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.796
[1219 21:52:49 @monitor.py:459] QueueInput/queue_size: 47.88
[1219 21:52:49 @monitor.py:459] l2_regularize_loss: 1.091
[1219 21:52:49 @monitor.py:459] learning_rate: 0.1
[1219 21:52:49 @monitor.py:459] train-error-top1: 0.53949
[1219 21:52:49 @monitor.py:459] train-error-top5: 0.31891
[1219 21:52:49 @monitor.py:459] val-error-top1: 0.45954
[1219 21:52:49 @monitor.py:459] val-error-top5: 0.20906
[1219 21:52:49 @monitor.py:459] xentropy-loss: 2.4856
[1219 21:52:49 @group.py:48] Callbacks took 36.288 sec in total. DataParallelInferenceRunner: 35.5 seconds
[1219 21:52:49 @base.py:275] Start Epoch 15 ...
[1219 22:08:18 @base.py:285] Epoch 15 (global_step 75060) finished, time:15 minutes 28 seconds.
[1219 22:08:18 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 22:08:19 @saver.py:77] Model saved to train_log/ResNet50-GN/model-75060.
[1219 22:08:19 @misc.py:109] Estimated Time Left: 1 day 11 minutes 37 seconds
[1219 22:08:55 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.915
[1219 22:08:55 @monitor.py:459] QueueInput/queue_size: 47.703
[1219 22:08:55 @monitor.py:459] l2_regularize_loss: 1.0995
[1219 22:08:55 @monitor.py:459] learning_rate: 0.1
[1219 22:08:55 @monitor.py:459] train-error-top1: 0.501
[1219 22:08:55 @monitor.py:459] train-error-top5: 0.275
[1219 22:08:55 @monitor.py:459] val-error-top1: 0.46338
[1219 22:08:55 @monitor.py:459] val-error-top5: 0.21188
[1219 22:08:55 @monitor.py:459] xentropy-loss: 2.2878
[1219 22:08:55 @group.py:48] Callbacks took 36.888 sec in total. DataParallelInferenceRunner: 35.8 seconds
[1219 22:08:55 @base.py:275] Start Epoch 16 ...
[1219 22:24:28 @base.py:285] Epoch 16 (global_step 80064) finished, time:15 minutes 33 seconds.
[1219 22:24:28 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 22:24:29 @saver.py:77] Model saved to train_log/ResNet50-GN/model-80064.
[1219 22:24:29 @misc.py:109] Estimated Time Left: 23 hours 56 minutes 9 seconds
[1219 22:25:05 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.124
[1219 22:25:05 @monitor.py:459] QueueInput/queue_size: 49.965
[1219 22:25:05 @monitor.py:459] l2_regularize_loss: 1.1048
[1219 22:25:05 @monitor.py:459] learning_rate: 0.1
[1219 22:25:05 @monitor.py:459] train-error-top1: 0.557
[1219 22:25:05 @monitor.py:459] train-error-top5: 0.29174
[1219 22:25:05 @monitor.py:459] val-error-top1: 0.4623
[1219 22:25:05 @monitor.py:459] val-error-top5: 0.21038
[1219 22:25:05 @monitor.py:459] xentropy-loss: 2.4528
[1219 22:25:05 @group.py:48] Callbacks took 36.312 sec in total. DataParallelInferenceRunner: 35.6 seconds
[1219 22:25:05 @base.py:275] Start Epoch 17 ...
[1219 22:40:38 @base.py:285] Epoch 17 (global_step 85068) finished, time:15 minutes 33 seconds.
[1219 22:40:38 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 22:40:39 @saver.py:77] Model saved to train_log/ResNet50-GN/model-85068.
[1219 22:40:39 @misc.py:109] Estimated Time Left: 23 hours 40 minutes 17 seconds
[1219 22:41:16 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.803
[1219 22:41:16 @monitor.py:459] QueueInput/queue_size: 47.944
[1219 22:41:16 @monitor.py:459] l2_regularize_loss: 1.1098
[1219 22:41:16 @monitor.py:459] learning_rate: 0.1
[1219 22:41:16 @monitor.py:459] train-error-top1: 0.5087
[1219 22:41:16 @monitor.py:459] train-error-top5: 0.28771
[1219 22:41:16 @monitor.py:459] val-error-top1: 0.45336
[1219 22:41:16 @monitor.py:459] val-error-top5: 0.2029
[1219 22:41:16 @monitor.py:459] xentropy-loss: 2.3087
[1219 22:41:16 @group.py:48] Callbacks took 37.270 sec in total. DataParallelInferenceRunner: 36.1 seconds
[1219 22:41:16 @base.py:275] Start Epoch 18 ...
[1219 22:56:48 @base.py:285] Epoch 18 (global_step 90072) finished, time:15 minutes 32 seconds.
[1219 22:56:48 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 22:56:49 @saver.py:77] Model saved to train_log/ResNet50-GN/model-90072.
[1219 22:56:49 @misc.py:109] Estimated Time Left: 23 hours 24 minutes 48 seconds
[1219 22:57:25 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.492
[1219 22:57:25 @monitor.py:459] QueueInput/queue_size: 49.494
[1219 22:57:25 @monitor.py:459] l2_regularize_loss: 1.1133
[1219 22:57:25 @monitor.py:459] learning_rate: 0.1
[1219 22:57:25 @monitor.py:459] train-error-top1: 0.5159
[1219 22:57:25 @monitor.py:459] train-error-top5: 0.26345
[1219 22:57:25 @monitor.py:459] val-error-top1: 0.45738
[1219 22:57:25 @monitor.py:459] val-error-top5: 0.20724
[1219 22:57:25 @monitor.py:459] xentropy-loss: 2.3015
[1219 22:57:25 @group.py:48] Callbacks took 37.318 sec in total. DataParallelInferenceRunner: 36.5 seconds
[1219 22:57:25 @base.py:275] Start Epoch 19 ...
[1219 23:12:55 @base.py:285] Epoch 19 (global_step 95076) finished, time:15 minutes 30 seconds.
[1219 23:12:55 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 23:12:56 @saver.py:77] Model saved to train_log/ResNet50-GN/model-95076.
[1219 23:12:56 @misc.py:109] Estimated Time Left: 23 hours 8 minutes 10 seconds
[1219 23:13:34 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.909
[1219 23:13:34 @monitor.py:459] QueueInput/queue_size: 49.923
[1219 23:13:34 @monitor.py:459] l2_regularize_loss: 1.1188
[1219 23:13:34 @monitor.py:459] learning_rate: 0.1
[1219 23:13:34 @monitor.py:459] train-error-top1: 0.52166
[1219 23:13:34 @monitor.py:459] train-error-top5: 0.28931
[1219 23:13:34 @monitor.py:459] val-error-top1: 0.45408
[1219 23:13:34 @monitor.py:459] val-error-top5: 0.20412
[1219 23:13:34 @monitor.py:459] xentropy-loss: 2.4203
[1219 23:13:34 @group.py:48] Callbacks took 38.218 sec in total. DataParallelInferenceRunner: 37.1 seconds
[1219 23:13:34 @base.py:275] Start Epoch 20 ...
[1219 23:29:09 @base.py:285] Epoch 20 (global_step 100080) finished, time:15 minutes 35 seconds.
[1219 23:29:09 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 23:29:11 @saver.py:77] Model saved to train_log/ResNet50-GN/model-100080.
[1219 23:29:11 @misc.py:109] Estimated Time Left: 22 hours 54 minutes 37 seconds
[1219 23:29:47 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.083
[1219 23:29:47 @monitor.py:459] QueueInput/queue_size: 49.781
[1219 23:29:47 @monitor.py:459] l2_regularize_loss: 1.1206
[1219 23:29:47 @monitor.py:459] learning_rate: 0.1
[1219 23:29:47 @monitor.py:459] train-error-top1: 0.52061
[1219 23:29:47 @monitor.py:459] train-error-top5: 0.27471
[1219 23:29:47 @monitor.py:459] val-error-top1: 0.45924
[1219 23:29:47 @monitor.py:459] val-error-top5: 0.20794
[1219 23:29:47 @monitor.py:459] xentropy-loss: 2.3392
[1219 23:29:47 @group.py:48] Callbacks took 37.090 sec in total. DataParallelInferenceRunner: 35.6 seconds
[1219 23:29:47 @base.py:275] Start Epoch 21 ...
[1219 23:45:21 @base.py:285] Epoch 21 (global_step 105084) finished, time:15 minutes 34 seconds.
[1219 23:45:21 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1219 23:45:22 @saver.py:77] Model saved to train_log/ResNet50-GN/model-105084.
[1219 23:45:22 @misc.py:109] Estimated Time Left: 22 hours 38 minutes 42 seconds
[1219 23:45:59 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.62
[1219 23:45:59 @monitor.py:459] QueueInput/queue_size: 49.845
[1219 23:45:59 @monitor.py:459] l2_regularize_loss: 1.1243
[1219 23:45:59 @monitor.py:459] learning_rate: 0.1
[1219 23:45:59 @monitor.py:459] train-error-top1: 0.51423
[1219 23:45:59 @monitor.py:459] train-error-top5: 0.28954
[1219 23:45:59 @monitor.py:459] val-error-top1: 0.44494
[1219 23:45:59 @monitor.py:459] val-error-top5: 0.1969
[1219 23:45:59 @monitor.py:459] xentropy-loss: 2.3954
[1219 23:45:59 @group.py:48] Callbacks took 37.889 sec in total. DataParallelInferenceRunner: 36.9 seconds
[1219 23:45:59 @base.py:275] Start Epoch 22 ...
[1220 00:01:32 @base.py:285] Epoch 22 (global_step 110088) finished, time:15 minutes 33 seconds.
[1220 00:01:32 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 00:01:33 @saver.py:77] Model saved to train_log/ResNet50-GN/model-110088.
[1220 00:01:33 @misc.py:109] Estimated Time Left: 22 hours 22 minutes 54 seconds
[1220 00:02:10 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.798
[1220 00:02:10 @monitor.py:459] QueueInput/queue_size: 49.574
[1220 00:02:10 @monitor.py:459] l2_regularize_loss: 1.1261
[1220 00:02:10 @monitor.py:459] learning_rate: 0.1
[1220 00:02:10 @monitor.py:459] train-error-top1: 0.53043
[1220 00:02:10 @monitor.py:459] train-error-top5: 0.27477
[1220 00:02:10 @monitor.py:459] val-error-top1: 0.44198
[1220 00:02:10 @monitor.py:459] val-error-top5: 0.19586
[1220 00:02:10 @monitor.py:459] xentropy-loss: 2.3135
[1220 00:02:10 @group.py:48] Callbacks took 38.307 sec in total. DataParallelInferenceRunner: 37.1 seconds
[1220 00:02:10 @base.py:275] Start Epoch 23 ...
[1220 00:17:47 @base.py:285] Epoch 23 (global_step 115092) finished, time:15 minutes 36 seconds.
[1220 00:17:47 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 00:17:48 @saver.py:77] Model saved to train_log/ResNet50-GN/model-115092.
[1220 00:17:48 @misc.py:109] Estimated Time Left: 22 hours 8 minutes 22 seconds
[1220 00:18:25 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.479
[1220 00:18:25 @monitor.py:459] QueueInput/queue_size: 48.993
[1220 00:18:25 @monitor.py:459] l2_regularize_loss: 1.1295
[1220 00:18:25 @monitor.py:459] learning_rate: 0.1
[1220 00:18:25 @monitor.py:459] train-error-top1: 0.49912
[1220 00:18:25 @monitor.py:459] train-error-top5: 0.24787
[1220 00:18:25 @monitor.py:459] val-error-top1: 0.44954
[1220 00:18:25 @monitor.py:459] val-error-top5: 0.20046
[1220 00:18:25 @monitor.py:459] xentropy-loss: 2.2321
[1220 00:18:25 @group.py:48] Callbacks took 37.587 sec in total. DataParallelInferenceRunner: 36.4 seconds
[1220 00:18:25 @base.py:275] Start Epoch 24 ...
[1220 00:33:55 @base.py:285] Epoch 24 (global_step 120096) finished, time:15 minutes 29 seconds.
[1220 00:33:55 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 00:33:56 @saver.py:77] Model saved to train_log/ResNet50-GN/model-120096.
[1220 00:33:56 @misc.py:109] Estimated Time Left: 21 hours 52 minutes 4 seconds
[1220 00:34:31 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.04
[1220 00:34:31 @monitor.py:459] QueueInput/queue_size: 49.9
[1220 00:34:31 @monitor.py:459] l2_regularize_loss: 1.1331
[1220 00:34:31 @monitor.py:459] learning_rate: 0.1
[1220 00:34:31 @monitor.py:459] train-error-top1: 0.5335
[1220 00:34:31 @monitor.py:459] train-error-top5: 0.2757
[1220 00:34:31 @monitor.py:459] val-error-top1: 0.44488
[1220 00:34:31 @monitor.py:459] val-error-top5: 0.19456
[1220 00:34:31 @monitor.py:459] xentropy-loss: 2.3378
[1220 00:34:31 @group.py:48] Callbacks took 36.842 sec in total. DataParallelInferenceRunner: 35.5 seconds
[1220 00:34:31 @base.py:275] Start Epoch 25 ...
[1220 00:50:00 @base.py:285] Epoch 25 (global_step 125100) finished, time:15 minutes 28 seconds.
[1220 00:50:00 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 00:50:01 @saver.py:77] Model saved to train_log/ResNet50-GN/model-125100.
[1220 00:50:01 @misc.py:109] Estimated Time Left: 21 hours 33 minutes 18 seconds
[1220 00:50:37 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.926
[1220 00:50:37 @monitor.py:459] QueueInput/queue_size: 49.998
[1220 00:50:37 @monitor.py:459] l2_regularize_loss: 1.134
[1220 00:50:37 @monitor.py:459] learning_rate: 0.1
[1220 00:50:37 @monitor.py:459] train-error-top1: 0.48561
[1220 00:50:37 @monitor.py:459] train-error-top5: 0.24956
[1220 00:50:37 @monitor.py:459] val-error-top1: 0.4447
[1220 00:50:37 @monitor.py:459] val-error-top5: 0.19728
[1220 00:50:37 @monitor.py:459] xentropy-loss: 2.1617
[1220 00:50:37 @group.py:48] Callbacks took 36.991 sec in total. DataParallelInferenceRunner: 36.1 seconds
[1220 00:50:37 @base.py:275] Start Epoch 26 ...
[1220 01:06:08 @base.py:285] Epoch 26 (global_step 130104) finished, time:15 minutes 30 seconds.
[1220 01:06:08 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 01:06:09 @saver.py:77] Model saved to train_log/ResNet50-GN/model-130104.
[1220 01:06:09 @misc.py:109] Estimated Time Left: 21 hours 16 minutes 25 seconds
[1220 01:06:43 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.621
[1220 01:06:43 @monitor.py:459] QueueInput/queue_size: 48.992
[1220 01:06:43 @monitor.py:459] l2_regularize_loss: 1.1352
[1220 01:06:43 @monitor.py:459] learning_rate: 0.1
[1220 01:06:43 @monitor.py:459] train-error-top1: 0.5127
[1220 01:06:43 @monitor.py:459] train-error-top5: 0.24991
[1220 01:06:43 @monitor.py:459] val-error-top1: 0.44478
[1220 01:06:43 @monitor.py:459] val-error-top5: 0.20092
[1220 01:06:43 @monitor.py:459] xentropy-loss: 2.2358
[1220 01:06:43 @group.py:48] Callbacks took 35.102 sec in total. DataParallelInferenceRunner: 33.8 seconds
[1220 01:06:43 @base.py:275] Start Epoch 27 ...
[1220 01:22:15 @base.py:285] Epoch 27 (global_step 135108) finished, time:15 minutes 32 seconds.
[1220 01:22:15 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 01:22:16 @saver.py:77] Model saved to train_log/ResNet50-GN/model-135108.
[1220 01:22:16 @misc.py:109] Estimated Time Left: 20 hours 59 minutes 6 seconds
[1220 01:22:52 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.094
[1220 01:22:52 @monitor.py:459] QueueInput/queue_size: 49.572
[1220 01:22:52 @monitor.py:459] l2_regularize_loss: 1.1381
[1220 01:22:52 @monitor.py:459] learning_rate: 0.1
[1220 01:22:52 @monitor.py:459] train-error-top1: 0.5043
[1220 01:22:52 @monitor.py:459] train-error-top5: 0.25436
[1220 01:22:52 @monitor.py:459] val-error-top1: 0.44062
[1220 01:22:52 @monitor.py:459] val-error-top5: 0.19466
[1220 01:22:52 @monitor.py:459] xentropy-loss: 2.2019
[1220 01:22:52 @group.py:48] Callbacks took 37.332 sec in total. DataParallelInferenceRunner: 36.4 seconds
[1220 01:22:52 @base.py:275] Start Epoch 28 ...
[1220 01:38:24 @base.py:285] Epoch 28 (global_step 140112) finished, time:15 minutes 31 seconds.
[1220 01:38:24 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 01:38:25 @saver.py:77] Model saved to train_log/ResNet50-GN/model-140112.
[1220 01:38:25 @misc.py:109] Estimated Time Left: 20 hours 41 minutes 19 seconds
[1220 01:39:00 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.769
[1220 01:39:00 @monitor.py:459] QueueInput/queue_size: 49.486
[1220 01:39:00 @monitor.py:459] l2_regularize_loss: 1.1409
[1220 01:39:00 @monitor.py:459] learning_rate: 0.1
[1220 01:39:00 @monitor.py:459] train-error-top1: 0.50842
[1220 01:39:00 @monitor.py:459] train-error-top5: 0.24898
[1220 01:39:00 @monitor.py:459] val-error-top1: 0.44114
[1220 01:39:00 @monitor.py:459] val-error-top5: 0.19278
[1220 01:39:00 @monitor.py:459] xentropy-loss: 2.2024
[1220 01:39:00 @group.py:48] Callbacks took 35.809 sec in total. DataParallelInferenceRunner: 34.8 seconds
[1220 01:39:00 @base.py:275] Start Epoch 29 ...
[1220 01:54:30 @base.py:285] Epoch 29 (global_step 145116) finished, time:15 minutes 30 seconds.
[1220 01:54:30 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 01:54:31 @saver.py:77] Model saved to train_log/ResNet50-GN/model-145116.
[1220 01:54:31 @misc.py:109] Estimated Time Left: 20 hours 24 minutes 57 seconds
[1220 01:55:07 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.611
[1220 01:55:07 @monitor.py:459] QueueInput/queue_size: 49.681
[1220 01:55:07 @monitor.py:459] l2_regularize_loss: 1.1417
[1220 01:55:07 @monitor.py:459] learning_rate: 0.1
[1220 01:55:07 @monitor.py:459] train-error-top1: 0.47591
[1220 01:55:07 @monitor.py:459] train-error-top5: 0.23456
[1220 01:55:07 @monitor.py:459] val-error-top1: 0.4435
[1220 01:55:07 @monitor.py:459] val-error-top5: 0.19094
[1220 01:55:07 @monitor.py:459] xentropy-loss: 2.1657
[1220 01:55:07 @group.py:48] Callbacks took 36.137 sec in total. DataParallelInferenceRunner: 35.3 seconds
[1220 01:55:07 @base.py:275] Start Epoch 30 ...
[1220 02:10:38 @base.py:285] Epoch 30 (global_step 150120) finished, time:15 minutes 31 seconds.
[1220 02:10:38 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 02:10:39 @saver.py:77] Model saved to train_log/ResNet50-GN/model-150120.
[1220 02:10:39 @misc.py:109] Estimated Time Left: 20 hours 9 minutes 35 seconds
[1220 02:10:39 @param.py:161] [HyperParamSetter] At global_step=150120, learning_rate changes from 0.100000 to 0.010000
[1220 02:11:16 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.155
[1220 02:11:16 @monitor.py:459] QueueInput/queue_size: 49.062
[1220 02:11:16 @monitor.py:459] l2_regularize_loss: 1.1435
[1220 02:11:16 @monitor.py:459] learning_rate: 0.1
[1220 02:11:16 @monitor.py:459] train-error-top1: 0.48194
[1220 02:11:16 @monitor.py:459] train-error-top5: 0.2432
[1220 02:11:16 @monitor.py:459] val-error-top1: 0.44082
[1220 02:11:16 @monitor.py:459] val-error-top5: 0.19136
[1220 02:11:16 @monitor.py:459] xentropy-loss: 2.1438
[1220 02:11:16 @group.py:48] Callbacks took 37.686 sec in total. DataParallelInferenceRunner: 36.6 seconds
[1220 02:11:16 @base.py:275] Start Epoch 31 ...
[1220 02:26:48 @base.py:285] Epoch 31 (global_step 155124) finished, time:15 minutes 32 seconds.
[1220 02:26:48 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 02:26:49 @saver.py:77] Model saved to train_log/ResNet50-GN/model-155124.
[1220 02:26:49 @misc.py:109] Estimated Time Left: 19 hours 53 minutes 56 seconds
[1220 02:27:24 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.64
[1220 02:27:24 @monitor.py:459] QueueInput/queue_size: 47.092
[1220 02:27:24 @monitor.py:459] l2_regularize_loss: 1.0546
[1220 02:27:24 @monitor.py:459] learning_rate: 0.01
[1220 02:27:24 @monitor.py:459] train-error-top1: 0.35191
[1220 02:27:24 @monitor.py:459] train-error-top5: 0.15394
[1220 02:27:24 @monitor.py:459] val-error-top1: 0.3235
[1220 02:27:24 @monitor.py:459] val-error-top5: 0.1176
[1220 02:27:24 @monitor.py:459] xentropy-loss: 1.5135
[1220 02:27:24 @group.py:48] Callbacks took 35.643 sec in total. DataParallelInferenceRunner: 34.7 seconds
[1220 02:27:24 @base.py:275] Start Epoch 32 ...
[1220 02:42:55 @base.py:285] Epoch 32 (global_step 160128) finished, time:15 minutes 31 seconds.
[1220 02:42:55 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 02:42:56 @saver.py:77] Model saved to train_log/ResNet50-GN/model-160128.
[1220 02:42:56 @misc.py:109] Estimated Time Left: 19 hours 37 minutes 44 seconds
[1220 02:43:31 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.869
[1220 02:43:31 @monitor.py:459] QueueInput/queue_size: 47.987
[1220 02:43:31 @monitor.py:459] l2_regularize_loss: 0.97688
[1220 02:43:31 @monitor.py:459] learning_rate: 0.01
[1220 02:43:31 @monitor.py:459] train-error-top1: 0.35762
[1220 02:43:31 @monitor.py:459] train-error-top5: 0.14639
[1220 02:43:31 @monitor.py:459] val-error-top1: 0.3135
[1220 02:43:31 @monitor.py:459] val-error-top5: 0.11126
[1220 02:43:31 @monitor.py:459] xentropy-loss: 1.4937
[1220 02:43:31 @group.py:48] Callbacks took 36.246 sec in total. DataParallelInferenceRunner: 35.3 seconds
[1220 02:43:31 @base.py:275] Start Epoch 33 ...
[1220 02:59:01 @base.py:285] Epoch 33 (global_step 165132) finished, time:15 minutes 29 seconds.
[1220 02:59:01 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 02:59:02 @saver.py:77] Model saved to train_log/ResNet50-GN/model-165132.
[1220 02:59:02 @misc.py:109] Estimated Time Left: 19 hours 20 minutes 51 seconds
[1220 02:59:37 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.269
[1220 02:59:37 @monitor.py:459] QueueInput/queue_size: 49.925
[1220 02:59:37 @monitor.py:459] l2_regularize_loss: 0.90961
[1220 02:59:37 @monitor.py:459] learning_rate: 0.01
[1220 02:59:37 @monitor.py:459] train-error-top1: 0.35251
[1220 02:59:37 @monitor.py:459] train-error-top5: 0.15808
[1220 02:59:37 @monitor.py:459] val-error-top1: 0.31148
[1220 02:59:37 @monitor.py:459] val-error-top5: 0.10948
[1220 02:59:37 @monitor.py:459] xentropy-loss: 1.496
[1220 02:59:37 @group.py:48] Callbacks took 36.017 sec in total. DataParallelInferenceRunner: 34.8 seconds
[1220 02:59:37 @base.py:275] Start Epoch 34 ...
[1220 03:15:06 @base.py:285] Epoch 34 (global_step 170136) finished, time:15 minutes 29 seconds.
[1220 03:15:06 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 03:15:08 @saver.py:77] Model saved to train_log/ResNet50-GN/model-170136.
[1220 03:15:08 @misc.py:109] Estimated Time Left: 19 hours 4 minutes 35 seconds
[1220 03:15:43 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.599
[1220 03:15:43 @monitor.py:459] QueueInput/queue_size: 48.927
[1220 03:15:43 @monitor.py:459] l2_regularize_loss: 0.8516
[1220 03:15:43 @monitor.py:459] learning_rate: 0.01
[1220 03:15:43 @monitor.py:459] train-error-top1: 0.33564
[1220 03:15:43 @monitor.py:459] train-error-top5: 0.14487
[1220 03:15:43 @monitor.py:459] val-error-top1: 0.30738
[1220 03:15:43 @monitor.py:459] val-error-top5: 0.10738
[1220 03:15:43 @monitor.py:459] xentropy-loss: 1.4514
[1220 03:15:43 @group.py:48] Callbacks took 36.312 sec in total. DataParallelInferenceRunner: 35.2 seconds
[1220 03:15:43 @base.py:275] Start Epoch 35 ...
[1220 03:31:13 @base.py:285] Epoch 35 (global_step 175140) finished, time:15 minutes 30 seconds.
[1220 03:31:13 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 03:31:15 @saver.py:77] Model saved to train_log/ResNet50-GN/model-175140.
[1220 03:31:15 @misc.py:109] Estimated Time Left: 18 hours 48 minutes 19 seconds
[1220 03:31:52 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.888
[1220 03:31:52 @monitor.py:459] QueueInput/queue_size: 48.892
[1220 03:31:52 @monitor.py:459] l2_regularize_loss: 0.8017
[1220 03:31:52 @monitor.py:459] learning_rate: 0.01
[1220 03:31:52 @monitor.py:459] train-error-top1: 0.32469
[1220 03:31:52 @monitor.py:459] train-error-top5: 0.13863
[1220 03:31:52 @monitor.py:459] val-error-top1: 0.30706
[1220 03:31:52 @monitor.py:459] val-error-top5: 0.10536
[1220 03:31:52 @monitor.py:459] xentropy-loss: 1.3783
[1220 03:31:52 @group.py:48] Callbacks took 38.211 sec in total. DataParallelInferenceRunner: 36.8 seconds
[1220 03:31:52 @base.py:275] Start Epoch 36 ...
[1220 03:47:21 @base.py:285] Epoch 36 (global_step 180144) finished, time:15 minutes 29 seconds.
[1220 03:47:21 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 03:47:22 @saver.py:77] Model saved to train_log/ResNet50-GN/model-180144.
[1220 03:47:22 @misc.py:109] Estimated Time Left: 18 hours 31 minutes 34 seconds
[1220 03:47:58 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.501
[1220 03:47:58 @monitor.py:459] QueueInput/queue_size: 49.811
[1220 03:47:58 @monitor.py:459] l2_regularize_loss: 0.75919
[1220 03:47:58 @monitor.py:459] learning_rate: 0.01
[1220 03:47:58 @monitor.py:459] train-error-top1: 0.32261
[1220 03:47:58 @monitor.py:459] train-error-top5: 0.12497
[1220 03:47:58 @monitor.py:459] val-error-top1: 0.3048
[1220 03:47:58 @monitor.py:459] val-error-top5: 0.10556
[1220 03:47:58 @monitor.py:459] xentropy-loss: 1.3603
[1220 03:47:58 @group.py:48] Callbacks took 37.048 sec in total. DataParallelInferenceRunner: 36.2 seconds
[1220 03:47:58 @base.py:275] Start Epoch 37 ...
[1220 04:03:29 @base.py:285] Epoch 37 (global_step 185148) finished, time:15 minutes 30 seconds.
[1220 04:03:29 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 04:03:31 @saver.py:77] Model saved to train_log/ResNet50-GN/model-185148.
[1220 04:03:31 @misc.py:109] Estimated Time Left: 18 hours 15 minutes 48 seconds
[1220 04:04:08 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 25.773
[1220 04:04:08 @monitor.py:459] QueueInput/queue_size: 48.998
[1220 04:04:08 @monitor.py:459] l2_regularize_loss: 0.72315
[1220 04:04:08 @monitor.py:459] learning_rate: 0.01
[1220 04:04:08 @monitor.py:459] train-error-top1: 0.34545
[1220 04:04:08 @monitor.py:459] train-error-top5: 0.13876
[1220 04:04:08 @monitor.py:459] val-error-top1: 0.30378
[1220 04:04:08 @monitor.py:459] val-error-top5: 0.10418
[1220 04:04:08 @monitor.py:459] xentropy-loss: 1.4463
[1220 04:04:08 @group.py:48] Callbacks took 38.396 sec in total. DataParallelInferenceRunner: 37.2 seconds
[1220 04:04:08 @base.py:275] Start Epoch 38 ...
[1220 04:19:39 @base.py:285] Epoch 38 (global_step 190152) finished, time:15 minutes 31 seconds.
[1220 04:19:39 @graph.py:73] Running Op sync_variables/sync_variables_from_main_tower ...
[1220 04:19:40 @saver.py:77] Model saved to train_log/ResNet50-GN/model-190152.
[1220 04:19:40 @misc.py:109] Estimated Time Left: 18 hours 37 seconds
[1220 04:20:18 @monitor.py:459] DataParallelInferenceRunner/QueueInput/queue_size: 26.056
[1220 04:20:18 @monitor.py:459] QueueInput/queue_size: 49.971
[1220 04:20:18 @monitor.py:459] l2_regularize_loss: 0.69282
[1220 04:20:18 @monitor.py:459] learning_rate: 0.01
[1220 04:20:18 @monitor.py:459] train-error-top1: 0.34373
[1220 04:20:18 @monitor.py:459] train-error-top5: 0.14335
[1220 04:20:18 @monitor.py:459] val-error-top1: 0.30328
[1220 04:20:18 @monitor.py:459] val-error-top5: 0.10402