-
Notifications
You must be signed in to change notification settings - Fork 1
/
log_1-76.txt
3638 lines (3638 loc) · 364 KB
/
log_1-76.txt
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
2021-12-28 01:16:41,354
AMP: False
AUG:
AUTO_AUGMENT: None
COLOR_JITTER: 0.4
CUTMIX: 1.0
CUTMIX_MINMAX: None
MIXUP: 0.8
MIXUP_MODE: batch
MIXUP_PROB: 1.0
MIXUP_SWITCH_PROB: 0.5
RE_COUNT: 1
RE_MODE: pixel
RE_PROB: 0.25
BASE: ['']
DATA:
BATCH_SIZE: 256
BATCH_SIZE_EVAL: 8
CROP_PCT: 0.9
DATASET: imagenet2012
DATA_PATH: ./Light_ILSVRC2012
IMAGENET_MEAN: [0.485, 0.456, 0.406]
IMAGENET_STD: [0.229, 0.224, 0.225]
IMAGE_SIZE: 224
NUM_WORKERS: 8
EVAL: False
LOCAL_RANK: 0
MODEL:
ATTENTION_DROPOUT: 0.0
DISTILL: False
DROPOUT: 0.0
DROP_PATH: 0.1
NAME: pit_ti
NUM_CLASSES: 1000
PRETRAINED: None
RESUME: None
TRANS:
BASE_DIMS: [32, 32, 32]
DEPTH: [2, 6, 4]
HEADS: [2, 4, 8]
PATCH_SIZE: 16
STRIDE: 8
TYPE: PiT
NGPUS: 4
REPORT_FREQ: 50
SAVE: /root/paddlejob/workspace/output/train-20211228-01-15-41
SAVE_FREQ: 10
SEED: 0
TAG: default
TRAIN:
ACCUM_ITER: 1
AUTO_AUGMENT: True
BASE_LR: 0.0005
COLOR_JITTER: 0.4
COOLDOWN_EPOCHS: 10
CUTMIX_ALPHA: 1.0
CUTMIX_MINMAX: None
DISTILLATION_ALPHA: 0.5
DISTILLATION_TAU: 1.0
DISTILLATION_TYPE: none
END_LR: 5e-06
GRAD_CLIP: 5.0
LAST_EPOCH: 0
LINEAR_SCALED_LR: None
LR_SCHEDULER:
DECAY_EPOCHS: 30
DECAY_RATE: 0.1
MILESTONES: 30, 60, 90
NAME: warmupcosine
MIXUP_ALPHA: 0.8
MIXUP_MODE: batch
MIXUP_PROB: 1.0
MIXUP_SWITCH_PROB: 0.5
MODEL_EMA: True
MODEL_EMA_DECAY: 0.99996
NUM_EPOCHS: 300
OPTIMIZER:
BETAS: (0.9, 0.999)
EPS: 1e-08
MOMENTUM: 0.9
NAME: AdamW
RANDOM_ERASE_COUNT: 1
RANDOM_ERASE_MODE: pixel
RANDOM_ERASE_PROB: 0.25
RANDOM_ERASE_SPLIT: False
SMOOTHING: 0.1
TEACHER_MODEL: ./regnety_160
WARMUP_EPOCHS: 20
WARMUP_START_LR: 5e-07
WEIGHT_DECAY: 0.05
VALIDATE_FREQ: 2
2021-12-28 01:16:41,354 ----- world_size = 4, local_rank = 0
2021-12-28 01:16:41,489 ----- Total # of train batch (single gpu): 1251
2021-12-28 01:16:41,489 ----- Total # of val batch (single gpu): 1563
2021-12-28 01:16:41,491 Start training from epoch 1.
2021-12-28 01:16:41,491 Now training epoch 1. LR=0.000050
2021-12-28 01:18:04,598 Epoch[001/310], Step[0000/1251], Loss: 7.0940(7.0940), Acc: 0.0010(0.0010)
2021-12-28 01:19:08,363 Epoch[001/310], Step[0050/1251], Loss: 6.9415(7.0088), Acc: 0.0020(0.0011)
2021-12-28 01:20:10,503 Epoch[001/310], Step[0100/1251], Loss: 6.9237(6.9746), Acc: 0.0000(0.0012)
2021-12-28 01:21:11,457 Epoch[001/310], Step[0150/1251], Loss: 6.9087(6.9556), Acc: 0.0020(0.0012)
2021-12-28 01:22:14,286 Epoch[001/310], Step[0200/1251], Loss: 6.9103(6.9440), Acc: 0.0020(0.0012)
2021-12-28 01:23:18,092 Epoch[001/310], Step[0250/1251], Loss: 6.9053(6.9359), Acc: 0.0029(0.0013)
2021-12-28 01:24:20,135 Epoch[001/310], Step[0300/1251], Loss: 6.8939(6.9298), Acc: 0.0039(0.0015)
2021-12-28 01:25:24,095 Epoch[001/310], Step[0350/1251], Loss: 6.9114(6.9249), Acc: 0.0020(0.0015)
2021-12-28 01:26:26,810 Epoch[001/310], Step[0400/1251], Loss: 6.8918(6.9205), Acc: 0.0029(0.0016)
2021-12-28 01:27:30,353 Epoch[001/310], Step[0450/1251], Loss: 6.8826(6.9166), Acc: 0.0020(0.0017)
2021-12-28 01:28:33,786 Epoch[001/310], Step[0500/1251], Loss: 6.8696(6.9127), Acc: 0.0049(0.0018)
2021-12-28 01:29:35,682 Epoch[001/310], Step[0550/1251], Loss: 6.8517(6.9087), Acc: 0.0029(0.0019)
2021-12-28 01:30:40,043 Epoch[001/310], Step[0600/1251], Loss: 6.8667(6.9045), Acc: 0.0000(0.0020)
2021-12-28 01:31:43,991 Epoch[001/310], Step[0650/1251], Loss: 6.8660(6.9000), Acc: 0.0039(0.0021)
2021-12-28 01:32:46,993 Epoch[001/310], Step[0700/1251], Loss: 6.8412(6.8954), Acc: 0.0098(0.0022)
2021-12-28 01:33:50,664 Epoch[001/310], Step[0750/1251], Loss: 6.8220(6.8906), Acc: 0.0010(0.0023)
2021-12-28 01:34:53,366 Epoch[001/310], Step[0800/1251], Loss: 6.8184(6.8856), Acc: 0.0059(0.0024)
2021-12-28 01:35:56,907 Epoch[001/310], Step[0850/1251], Loss: 6.8315(6.8799), Acc: 0.0010(0.0025)
2021-12-28 01:36:59,922 Epoch[001/310], Step[0900/1251], Loss: 6.8223(6.8743), Acc: 0.0049(0.0026)
2021-12-28 01:38:03,417 Epoch[001/310], Step[0950/1251], Loss: 6.7653(6.8687), Acc: 0.0088(0.0028)
2021-12-28 01:39:06,382 Epoch[001/310], Step[1000/1251], Loss: 6.7393(6.8631), Acc: 0.0020(0.0029)
2021-12-28 01:40:09,642 Epoch[001/310], Step[1050/1251], Loss: 6.7027(6.8576), Acc: 0.0098(0.0030)
2021-12-28 01:41:13,291 Epoch[001/310], Step[1100/1251], Loss: 6.7285(6.8519), Acc: 0.0059(0.0031)
2021-12-28 01:42:15,692 Epoch[001/310], Step[1150/1251], Loss: 6.7719(6.8464), Acc: 0.0088(0.0033)
2021-12-28 01:43:19,812 Epoch[001/310], Step[1200/1251], Loss: 6.7301(6.8411), Acc: 0.0088(0.0034)
2021-12-28 01:44:22,935 Epoch[001/310], Step[1250/1251], Loss: 6.6901(6.8359), Acc: 0.0068(0.0035)
2021-12-28 01:44:24,848 ----- Epoch[001/310], Train Loss: 6.8359, Train Acc: 0.0035, time: 1663.35
2021-12-28 01:44:24,848 Now training epoch 2. LR=0.000100
2021-12-28 01:45:41,090 Epoch[002/310], Step[0000/1251], Loss: 6.6706(6.6706), Acc: 0.0039(0.0039)
2021-12-28 01:46:42,212 Epoch[002/310], Step[0050/1251], Loss: 6.6461(6.7129), Acc: 0.0039(0.0070)
2021-12-28 01:47:42,999 Epoch[002/310], Step[0100/1251], Loss: 6.7165(6.7035), Acc: 0.0020(0.0067)
2021-12-28 01:48:44,362 Epoch[002/310], Step[0150/1251], Loss: 6.6357(6.6989), Acc: 0.0068(0.0068)
2021-12-28 01:49:46,698 Epoch[002/310], Step[0200/1251], Loss: 6.7086(6.6934), Acc: 0.0039(0.0070)
2021-12-28 01:50:50,094 Epoch[002/310], Step[0250/1251], Loss: 6.6614(6.6867), Acc: 0.0078(0.0073)
2021-12-28 01:51:50,687 Epoch[002/310], Step[0300/1251], Loss: 6.5899(6.6815), Acc: 0.0078(0.0076)
2021-12-28 01:52:53,891 Epoch[002/310], Step[0350/1251], Loss: 6.6899(6.6762), Acc: 0.0088(0.0079)
2021-12-28 01:53:56,844 Epoch[002/310], Step[0400/1251], Loss: 6.5864(6.6709), Acc: 0.0068(0.0081)
2021-12-28 01:55:00,218 Epoch[002/310], Step[0450/1251], Loss: 6.5619(6.6667), Acc: 0.0117(0.0082)
2021-12-28 01:56:03,437 Epoch[002/310], Step[0500/1251], Loss: 6.6267(6.6621), Acc: 0.0029(0.0083)
2021-12-28 01:57:05,172 Epoch[002/310], Step[0550/1251], Loss: 6.5709(6.6581), Acc: 0.0049(0.0084)
2021-12-28 01:58:07,259 Epoch[002/310], Step[0600/1251], Loss: 6.6521(6.6537), Acc: 0.0020(0.0086)
2021-12-28 01:59:10,790 Epoch[002/310], Step[0650/1251], Loss: 6.6175(6.6481), Acc: 0.0059(0.0087)
2021-12-28 02:00:11,775 Epoch[002/310], Step[0700/1251], Loss: 6.5885(6.6430), Acc: 0.0156(0.0089)
2021-12-28 02:01:15,222 Epoch[002/310], Step[0750/1251], Loss: 6.5930(6.6392), Acc: 0.0088(0.0090)
2021-12-28 02:02:17,228 Epoch[002/310], Step[0800/1251], Loss: 6.5865(6.6363), Acc: 0.0117(0.0092)
2021-12-28 02:03:20,822 Epoch[002/310], Step[0850/1251], Loss: 6.5710(6.6329), Acc: 0.0146(0.0093)
2021-12-28 02:04:23,056 Epoch[002/310], Step[0900/1251], Loss: 6.5508(6.6303), Acc: 0.0137(0.0094)
2021-12-28 02:05:24,450 Epoch[002/310], Step[0950/1251], Loss: 6.5499(6.6278), Acc: 0.0107(0.0095)
2021-12-28 02:06:27,162 Epoch[002/310], Step[1000/1251], Loss: 6.5868(6.6243), Acc: 0.0078(0.0097)
2021-12-28 02:07:31,261 Epoch[002/310], Step[1050/1251], Loss: 6.5730(6.6208), Acc: 0.0078(0.0097)
2021-12-28 02:08:35,254 Epoch[002/310], Step[1100/1251], Loss: 6.5274(6.6183), Acc: 0.0068(0.0099)
2021-12-28 02:09:38,729 Epoch[002/310], Step[1150/1251], Loss: 6.5782(6.6155), Acc: 0.0146(0.0100)
2021-12-28 02:10:42,967 Epoch[002/310], Step[1200/1251], Loss: 6.5596(6.6126), Acc: 0.0117(0.0101)
2021-12-28 02:11:44,045 Epoch[002/310], Step[1250/1251], Loss: 6.4667(6.6101), Acc: 0.0137(0.0103)
2021-12-28 02:11:45,984 ----- Validation after Epoch: 2
2021-12-28 02:12:47,828 Val Step[0000/1563], Loss: 6.1526 (6.1526), Acc@1: 0.0938 (0.0938), Acc@5: 0.1250 (0.1250)
2021-12-28 02:12:49,402 Val Step[0050/1563], Loss: 6.1569 (5.7906), Acc@1: 0.0000 (0.0386), Acc@5: 0.0000 (0.1544)
2021-12-28 02:12:50,843 Val Step[0100/1563], Loss: 5.5235 (5.8577), Acc@1: 0.0312 (0.0347), Acc@5: 0.1875 (0.1256)
2021-12-28 02:12:52,380 Val Step[0150/1563], Loss: 6.3391 (5.9068), Acc@1: 0.0000 (0.0294), Acc@5: 0.0625 (0.1136)
2021-12-28 02:12:53,880 Val Step[0200/1563], Loss: 6.4078 (5.8920), Acc@1: 0.0000 (0.0320), Acc@5: 0.0000 (0.1185)
2021-12-28 02:12:55,349 Val Step[0250/1563], Loss: 6.2010 (5.8724), Acc@1: 0.0000 (0.0335), Acc@5: 0.0312 (0.1282)
2021-12-28 02:12:56,809 Val Step[0300/1563], Loss: 6.0884 (5.9241), Acc@1: 0.0000 (0.0284), Acc@5: 0.0000 (0.1114)
2021-12-28 02:12:58,324 Val Step[0350/1563], Loss: 5.5859 (5.9384), Acc@1: 0.0625 (0.0259), Acc@5: 0.4375 (0.1048)
2021-12-28 02:12:59,779 Val Step[0400/1563], Loss: 5.4140 (5.9457), Acc@1: 0.0000 (0.0258), Acc@5: 0.0625 (0.1011)
2021-12-28 02:13:01,427 Val Step[0450/1563], Loss: 5.4875 (5.9498), Acc@1: 0.0000 (0.0245), Acc@5: 0.0938 (0.0974)
2021-12-28 02:13:02,988 Val Step[0500/1563], Loss: 5.0737 (5.9432), Acc@1: 0.0312 (0.0257), Acc@5: 0.5000 (0.1009)
2021-12-28 02:13:04,591 Val Step[0550/1563], Loss: 6.0668 (5.9316), Acc@1: 0.0312 (0.0283), Acc@5: 0.0625 (0.1070)
2021-12-28 02:13:06,188 Val Step[0600/1563], Loss: 5.2369 (5.9256), Acc@1: 0.1562 (0.0284), Acc@5: 0.4688 (0.1088)
2021-12-28 02:13:07,640 Val Step[0650/1563], Loss: 6.6271 (5.9238), Acc@1: 0.0000 (0.0311), Acc@5: 0.0312 (0.1136)
2021-12-28 02:13:09,123 Val Step[0700/1563], Loss: 6.6836 (5.9466), Acc@1: 0.0000 (0.0298), Acc@5: 0.0000 (0.1086)
2021-12-28 02:13:10,705 Val Step[0750/1563], Loss: 5.5981 (5.9514), Acc@1: 0.0312 (0.0301), Acc@5: 0.1875 (0.1094)
2021-12-28 02:13:12,082 Val Step[0800/1563], Loss: 5.6226 (5.9540), Acc@1: 0.1562 (0.0301), Acc@5: 0.3750 (0.1104)
2021-12-28 02:13:13,596 Val Step[0850/1563], Loss: 6.6143 (5.9702), Acc@1: 0.0000 (0.0294), Acc@5: 0.0000 (0.1077)
2021-12-28 02:13:15,144 Val Step[0900/1563], Loss: 5.9602 (5.9649), Acc@1: 0.0000 (0.0309), Acc@5: 0.0000 (0.1101)
2021-12-28 02:13:16,780 Val Step[0950/1563], Loss: 6.1074 (5.9744), Acc@1: 0.0312 (0.0313), Acc@5: 0.0938 (0.1091)
2021-12-28 02:13:18,296 Val Step[1000/1563], Loss: 4.8796 (5.9771), Acc@1: 0.3438 (0.0320), Acc@5: 0.6562 (0.1100)
2021-12-28 02:13:19,751 Val Step[1050/1563], Loss: 6.0258 (5.9772), Acc@1: 0.0000 (0.0319), Acc@5: 0.0000 (0.1104)
2021-12-28 02:13:21,302 Val Step[1100/1563], Loss: 6.3236 (5.9832), Acc@1: 0.0000 (0.0325), Acc@5: 0.0938 (0.1099)
2021-12-28 02:13:22,833 Val Step[1150/1563], Loss: 6.2608 (5.9910), Acc@1: 0.0312 (0.0315), Acc@5: 0.0625 (0.1080)
2021-12-28 02:13:24,343 Val Step[1200/1563], Loss: 6.0101 (5.9969), Acc@1: 0.1250 (0.0317), Acc@5: 0.2812 (0.1073)
2021-12-28 02:13:25,949 Val Step[1250/1563], Loss: 5.3095 (6.0005), Acc@1: 0.0625 (0.0317), Acc@5: 0.3125 (0.1071)
2021-12-28 02:13:27,417 Val Step[1300/1563], Loss: 5.9796 (6.0067), Acc@1: 0.0000 (0.0316), Acc@5: 0.0312 (0.1057)
2021-12-28 02:13:28,951 Val Step[1350/1563], Loss: 5.5858 (6.0123), Acc@1: 0.0000 (0.0310), Acc@5: 0.0312 (0.1039)
2021-12-28 02:13:30,465 Val Step[1400/1563], Loss: 5.8377 (6.0112), Acc@1: 0.0625 (0.0315), Acc@5: 0.1875 (0.1043)
2021-12-28 02:13:32,013 Val Step[1450/1563], Loss: 5.8420 (6.0050), Acc@1: 0.0000 (0.0322), Acc@5: 0.0000 (0.1056)
2021-12-28 02:13:33,460 Val Step[1500/1563], Loss: 6.0359 (5.9799), Acc@1: 0.0938 (0.0352), Acc@5: 0.1250 (0.1124)
2021-12-28 02:13:34,981 Val Step[1550/1563], Loss: 4.9260 (5.9561), Acc@1: 0.2812 (0.0390), Acc@5: 0.5625 (0.1196)
2021-12-28 02:13:35,838 ----- Epoch[002/310], Validation Loss: 5.9529, Validation Acc@1: 0.0398, Validation Acc@5: 0.1208, time: 109.85
2021-12-28 02:13:35,838 ----- Epoch[002/310], Train Loss: 6.6101, Train Acc: 0.0103, time: 1641.13, Best Val(epoch2) Acc@1: 0.0398
2021-12-28 02:13:36,019 Max accuracy so far: 0.0398 at epoch_2
2021-12-28 02:13:36,019 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 02:13:36,019 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 02:13:36,091 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 02:13:36,091 Now training epoch 3. LR=0.000150
2021-12-28 02:14:53,186 Epoch[003/310], Step[0000/1251], Loss: 6.5208(6.5208), Acc: 0.0078(0.0078)
2021-12-28 02:15:55,994 Epoch[003/310], Step[0050/1251], Loss: 6.5465(6.5358), Acc: 0.0088(0.0126)
2021-12-28 02:16:58,346 Epoch[003/310], Step[0100/1251], Loss: 6.4940(6.5424), Acc: 0.0166(0.0129)
2021-12-28 02:17:59,748 Epoch[003/310], Step[0150/1251], Loss: 6.5350(6.5424), Acc: 0.0195(0.0128)
2021-12-28 02:19:00,756 Epoch[003/310], Step[0200/1251], Loss: 6.4596(6.5453), Acc: 0.0205(0.0129)
2021-12-28 02:20:03,623 Epoch[003/310], Step[0250/1251], Loss: 6.4331(6.5404), Acc: 0.0342(0.0133)
2021-12-28 02:21:05,933 Epoch[003/310], Step[0300/1251], Loss: 6.5202(6.5360), Acc: 0.0156(0.0133)
2021-12-28 02:22:07,631 Epoch[003/310], Step[0350/1251], Loss: 6.5801(6.5345), Acc: 0.0176(0.0136)
2021-12-28 02:23:09,008 Epoch[003/310], Step[0400/1251], Loss: 6.5020(6.5320), Acc: 0.0205(0.0136)
2021-12-28 02:24:12,965 Epoch[003/310], Step[0450/1251], Loss: 6.4971(6.5285), Acc: 0.0195(0.0136)
2021-12-28 02:25:15,701 Epoch[003/310], Step[0500/1251], Loss: 6.5883(6.5258), Acc: 0.0039(0.0137)
2021-12-28 02:26:18,041 Epoch[003/310], Step[0550/1251], Loss: 6.3944(6.5245), Acc: 0.0234(0.0138)
2021-12-28 02:27:20,578 Epoch[003/310], Step[0600/1251], Loss: 6.4589(6.5219), Acc: 0.0195(0.0140)
2021-12-28 02:28:22,936 Epoch[003/310], Step[0650/1251], Loss: 6.6498(6.5198), Acc: 0.0146(0.0140)
2021-12-28 02:29:24,666 Epoch[003/310], Step[0700/1251], Loss: 6.4827(6.5182), Acc: 0.0010(0.0141)
2021-12-28 02:30:27,740 Epoch[003/310], Step[0750/1251], Loss: 6.4262(6.5171), Acc: 0.0107(0.0141)
2021-12-28 02:31:30,578 Epoch[003/310], Step[0800/1251], Loss: 6.5057(6.5151), Acc: 0.0166(0.0142)
2021-12-28 02:32:33,378 Epoch[003/310], Step[0850/1251], Loss: 6.5124(6.5135), Acc: 0.0127(0.0143)
2021-12-28 02:33:36,483 Epoch[003/310], Step[0900/1251], Loss: 6.5201(6.5115), Acc: 0.0127(0.0142)
2021-12-28 02:34:38,106 Epoch[003/310], Step[0950/1251], Loss: 6.4418(6.5091), Acc: 0.0059(0.0143)
2021-12-28 02:35:39,644 Epoch[003/310], Step[1000/1251], Loss: 6.5024(6.5066), Acc: 0.0176(0.0145)
2021-12-28 02:36:41,864 Epoch[003/310], Step[1050/1251], Loss: 6.4407(6.5039), Acc: 0.0146(0.0146)
2021-12-28 02:37:44,734 Epoch[003/310], Step[1100/1251], Loss: 6.4088(6.5017), Acc: 0.0137(0.0147)
2021-12-28 02:38:45,944 Epoch[003/310], Step[1150/1251], Loss: 6.5124(6.5012), Acc: 0.0205(0.0148)
2021-12-28 02:39:48,435 Epoch[003/310], Step[1200/1251], Loss: 6.3755(6.4985), Acc: 0.0146(0.0148)
2021-12-28 02:40:49,419 Epoch[003/310], Step[1250/1251], Loss: 6.4717(6.4974), Acc: 0.0137(0.0149)
2021-12-28 02:40:51,623 ----- Epoch[003/310], Train Loss: 6.4974, Train Acc: 0.0149, time: 1635.53, Best Val(epoch2) Acc@1: 0.0398
2021-12-28 02:40:51,623 Now training epoch 4. LR=0.000200
2021-12-28 02:42:12,677 Epoch[004/310], Step[0000/1251], Loss: 6.4337(6.4337), Acc: 0.0039(0.0039)
2021-12-28 02:43:14,505 Epoch[004/310], Step[0050/1251], Loss: 6.4698(6.4608), Acc: 0.0205(0.0175)
2021-12-28 02:44:14,824 Epoch[004/310], Step[0100/1251], Loss: 6.4222(6.4627), Acc: 0.0195(0.0181)
2021-12-28 02:45:16,376 Epoch[004/310], Step[0150/1251], Loss: 6.4593(6.4582), Acc: 0.0254(0.0185)
2021-12-28 02:46:18,569 Epoch[004/310], Step[0200/1251], Loss: 6.5867(6.4538), Acc: 0.0117(0.0182)
2021-12-28 02:47:19,847 Epoch[004/310], Step[0250/1251], Loss: 6.4943(6.4537), Acc: 0.0059(0.0177)
2021-12-28 02:48:22,048 Epoch[004/310], Step[0300/1251], Loss: 6.3752(6.4499), Acc: 0.0059(0.0181)
2021-12-28 02:49:23,861 Epoch[004/310], Step[0350/1251], Loss: 6.3365(6.4445), Acc: 0.0137(0.0183)
2021-12-28 02:50:25,157 Epoch[004/310], Step[0400/1251], Loss: 6.4265(6.4406), Acc: 0.0127(0.0185)
2021-12-28 02:51:27,093 Epoch[004/310], Step[0450/1251], Loss: 6.2875(6.4395), Acc: 0.0322(0.0185)
2021-12-28 02:52:28,893 Epoch[004/310], Step[0500/1251], Loss: 6.4583(6.4390), Acc: 0.0088(0.0186)
2021-12-28 02:53:31,676 Epoch[004/310], Step[0550/1251], Loss: 6.4258(6.4368), Acc: 0.0166(0.0188)
2021-12-28 02:54:35,388 Epoch[004/310], Step[0600/1251], Loss: 6.4289(6.4348), Acc: 0.0156(0.0189)
2021-12-28 02:55:37,051 Epoch[004/310], Step[0650/1251], Loss: 6.3990(6.4345), Acc: 0.0176(0.0189)
2021-12-28 02:56:41,069 Epoch[004/310], Step[0700/1251], Loss: 6.4094(6.4337), Acc: 0.0215(0.0189)
2021-12-28 02:57:44,303 Epoch[004/310], Step[0750/1251], Loss: 6.3095(6.4324), Acc: 0.0273(0.0189)
2021-12-28 02:58:47,879 Epoch[004/310], Step[0800/1251], Loss: 6.3775(6.4319), Acc: 0.0205(0.0189)
2021-12-28 02:59:50,598 Epoch[004/310], Step[0850/1251], Loss: 6.4454(6.4319), Acc: 0.0312(0.0189)
2021-12-28 03:00:54,205 Epoch[004/310], Step[0900/1251], Loss: 6.4107(6.4296), Acc: 0.0146(0.0191)
2021-12-28 03:01:57,258 Epoch[004/310], Step[0950/1251], Loss: 6.3756(6.4282), Acc: 0.0127(0.0192)
2021-12-28 03:02:59,066 Epoch[004/310], Step[1000/1251], Loss: 6.4438(6.4256), Acc: 0.0234(0.0193)
2021-12-28 03:04:01,330 Epoch[004/310], Step[1050/1251], Loss: 6.5477(6.4250), Acc: 0.0098(0.0194)
2021-12-28 03:05:04,406 Epoch[004/310], Step[1100/1251], Loss: 6.5235(6.4228), Acc: 0.0186(0.0195)
2021-12-28 03:06:05,574 Epoch[004/310], Step[1150/1251], Loss: 6.0741(6.4198), Acc: 0.0459(0.0197)
2021-12-28 03:07:08,039 Epoch[004/310], Step[1200/1251], Loss: 6.3781(6.4182), Acc: 0.0283(0.0199)
2021-12-28 03:08:10,408 Epoch[004/310], Step[1250/1251], Loss: 6.4743(6.4167), Acc: 0.0215(0.0199)
2021-12-28 03:08:12,337 ----- Validation after Epoch: 4
2021-12-28 03:09:11,773 Val Step[0000/1563], Loss: 5.5857 (5.5857), Acc@1: 0.0938 (0.0938), Acc@5: 0.2500 (0.2500)
2021-12-28 03:09:13,337 Val Step[0050/1563], Loss: 5.6522 (5.1074), Acc@1: 0.0000 (0.0974), Acc@5: 0.0938 (0.2966)
2021-12-28 03:09:14,807 Val Step[0100/1563], Loss: 4.8835 (5.3068), Acc@1: 0.0000 (0.0733), Acc@5: 0.2812 (0.2259)
2021-12-28 03:09:16,298 Val Step[0150/1563], Loss: 4.9888 (5.3062), Acc@1: 0.1562 (0.0791), Acc@5: 0.3438 (0.2281)
2021-12-28 03:09:17,768 Val Step[0200/1563], Loss: 5.9477 (5.3002), Acc@1: 0.0000 (0.0813), Acc@5: 0.0000 (0.2261)
2021-12-28 03:09:19,261 Val Step[0250/1563], Loss: 6.2631 (5.2610), Acc@1: 0.0000 (0.0864), Acc@5: 0.0000 (0.2446)
2021-12-28 03:09:20,797 Val Step[0300/1563], Loss: 6.0147 (5.3521), Acc@1: 0.0000 (0.0751), Acc@5: 0.0312 (0.2183)
2021-12-28 03:09:22,266 Val Step[0350/1563], Loss: 4.8924 (5.3735), Acc@1: 0.1875 (0.0715), Acc@5: 0.4062 (0.2123)
2021-12-28 03:09:23,703 Val Step[0400/1563], Loss: 4.8037 (5.3806), Acc@1: 0.0000 (0.0702), Acc@5: 0.1250 (0.2106)
2021-12-28 03:09:25,325 Val Step[0450/1563], Loss: 4.6478 (5.4105), Acc@1: 0.0625 (0.0655), Acc@5: 0.5312 (0.2022)
2021-12-28 03:09:26,861 Val Step[0500/1563], Loss: 4.1501 (5.3955), Acc@1: 0.2188 (0.0672), Acc@5: 0.5312 (0.2048)
2021-12-28 03:09:28,424 Val Step[0550/1563], Loss: 5.9484 (5.3908), Acc@1: 0.0312 (0.0707), Acc@5: 0.0625 (0.2072)
2021-12-28 03:09:30,122 Val Step[0600/1563], Loss: 4.1452 (5.3886), Acc@1: 0.3125 (0.0713), Acc@5: 0.6562 (0.2086)
2021-12-28 03:09:31,669 Val Step[0650/1563], Loss: 5.9684 (5.3911), Acc@1: 0.0312 (0.0747), Acc@5: 0.0625 (0.2114)
2021-12-28 03:09:33,328 Val Step[0700/1563], Loss: 6.6606 (5.4258), Acc@1: 0.0000 (0.0718), Acc@5: 0.0000 (0.2046)
2021-12-28 03:09:35,060 Val Step[0750/1563], Loss: 5.5674 (5.4416), Acc@1: 0.0000 (0.0707), Acc@5: 0.1562 (0.2031)
2021-12-28 03:09:36,582 Val Step[0800/1563], Loss: 5.1198 (5.4501), Acc@1: 0.2188 (0.0707), Acc@5: 0.5000 (0.2033)
2021-12-28 03:09:38,137 Val Step[0850/1563], Loss: 6.0277 (5.4708), Acc@1: 0.0312 (0.0693), Acc@5: 0.0625 (0.1998)
2021-12-28 03:09:39,700 Val Step[0900/1563], Loss: 5.5742 (5.4694), Acc@1: 0.0312 (0.0705), Acc@5: 0.1250 (0.2014)
2021-12-28 03:09:41,324 Val Step[0950/1563], Loss: 5.6599 (5.4824), Acc@1: 0.0625 (0.0701), Acc@5: 0.2812 (0.1989)
2021-12-28 03:09:42,840 Val Step[1000/1563], Loss: 4.0546 (5.4876), Acc@1: 0.4688 (0.0704), Acc@5: 0.7500 (0.1990)
2021-12-28 03:09:44,380 Val Step[1050/1563], Loss: 5.6944 (5.4926), Acc@1: 0.0000 (0.0701), Acc@5: 0.0000 (0.1986)
2021-12-28 03:09:45,883 Val Step[1100/1563], Loss: 6.5358 (5.5044), Acc@1: 0.0000 (0.0695), Acc@5: 0.0000 (0.1964)
2021-12-28 03:09:47,354 Val Step[1150/1563], Loss: 5.6395 (5.5186), Acc@1: 0.0625 (0.0681), Acc@5: 0.1875 (0.1925)
2021-12-28 03:09:48,793 Val Step[1200/1563], Loss: 4.7829 (5.5274), Acc@1: 0.4062 (0.0678), Acc@5: 0.5938 (0.1914)
2021-12-28 03:09:50,369 Val Step[1250/1563], Loss: 5.0767 (5.5373), Acc@1: 0.1250 (0.0674), Acc@5: 0.2812 (0.1897)
2021-12-28 03:09:51,778 Val Step[1300/1563], Loss: 5.4526 (5.5446), Acc@1: 0.0312 (0.0663), Acc@5: 0.1562 (0.1874)
2021-12-28 03:09:53,231 Val Step[1350/1563], Loss: 5.3696 (5.5563), Acc@1: 0.0000 (0.0653), Acc@5: 0.0000 (0.1848)
2021-12-28 03:09:54,680 Val Step[1400/1563], Loss: 5.6455 (5.5588), Acc@1: 0.0000 (0.0651), Acc@5: 0.1250 (0.1845)
2021-12-28 03:09:56,118 Val Step[1450/1563], Loss: 5.3385 (5.5540), Acc@1: 0.0000 (0.0660), Acc@5: 0.1875 (0.1854)
2021-12-28 03:09:57,637 Val Step[1500/1563], Loss: 5.4705 (5.5207), Acc@1: 0.0312 (0.0693), Acc@5: 0.1875 (0.1934)
2021-12-28 03:09:59,109 Val Step[1550/1563], Loss: 3.3275 (5.4918), Acc@1: 0.6250 (0.0746), Acc@5: 0.7812 (0.2006)
2021-12-28 03:09:59,950 ----- Epoch[004/310], Validation Loss: 5.4884, Validation Acc@1: 0.0754, Validation Acc@5: 0.2016, time: 107.61
2021-12-28 03:09:59,951 ----- Epoch[004/310], Train Loss: 6.4167, Train Acc: 0.0199, time: 1640.71, Best Val(epoch4) Acc@1: 0.0754
2021-12-28 03:10:00,172 Max accuracy so far: 0.0754 at epoch_4
2021-12-28 03:10:00,173 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 03:10:00,173 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 03:10:00,254 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 03:10:00,255 Now training epoch 5. LR=0.000250
2021-12-28 03:11:19,125 Epoch[005/310], Step[0000/1251], Loss: 6.2711(6.2711), Acc: 0.0342(0.0342)
2021-12-28 03:12:18,692 Epoch[005/310], Step[0050/1251], Loss: 6.5073(6.3789), Acc: 0.0146(0.0222)
2021-12-28 03:13:17,763 Epoch[005/310], Step[0100/1251], Loss: 6.4768(6.3824), Acc: 0.0215(0.0236)
2021-12-28 03:14:20,346 Epoch[005/310], Step[0150/1251], Loss: 6.3189(6.3667), Acc: 0.0332(0.0244)
2021-12-28 03:15:22,150 Epoch[005/310], Step[0200/1251], Loss: 6.4432(6.3697), Acc: 0.0195(0.0235)
2021-12-28 03:16:24,484 Epoch[005/310], Step[0250/1251], Loss: 6.4518(6.3725), Acc: 0.0312(0.0236)
2021-12-28 03:17:27,989 Epoch[005/310], Step[0300/1251], Loss: 6.4499(6.3703), Acc: 0.0127(0.0233)
2021-12-28 03:18:31,095 Epoch[005/310], Step[0350/1251], Loss: 6.3843(6.3693), Acc: 0.0264(0.0234)
2021-12-28 03:19:33,706 Epoch[005/310], Step[0400/1251], Loss: 6.4956(6.3681), Acc: 0.0146(0.0236)
2021-12-28 03:20:36,633 Epoch[005/310], Step[0450/1251], Loss: 6.1938(6.3684), Acc: 0.0400(0.0236)
2021-12-28 03:21:38,147 Epoch[005/310], Step[0500/1251], Loss: 6.4161(6.3650), Acc: 0.0273(0.0238)
2021-12-28 03:22:39,815 Epoch[005/310], Step[0550/1251], Loss: 6.3705(6.3635), Acc: 0.0068(0.0238)
2021-12-28 03:23:43,921 Epoch[005/310], Step[0600/1251], Loss: 6.2083(6.3632), Acc: 0.0186(0.0238)
2021-12-28 03:24:45,677 Epoch[005/310], Step[0650/1251], Loss: 6.4322(6.3622), Acc: 0.0303(0.0237)
2021-12-28 03:25:49,008 Epoch[005/310], Step[0700/1251], Loss: 6.2162(6.3600), Acc: 0.0312(0.0237)
2021-12-28 03:26:51,870 Epoch[005/310], Step[0750/1251], Loss: 6.4254(6.3571), Acc: 0.0234(0.0239)
2021-12-28 03:27:54,426 Epoch[005/310], Step[0800/1251], Loss: 6.3007(6.3534), Acc: 0.0449(0.0241)
2021-12-28 03:28:58,564 Epoch[005/310], Step[0850/1251], Loss: 6.5177(6.3515), Acc: 0.0146(0.0242)
2021-12-28 03:30:00,525 Epoch[005/310], Step[0900/1251], Loss: 6.2390(6.3502), Acc: 0.0195(0.0243)
2021-12-28 03:31:03,099 Epoch[005/310], Step[0950/1251], Loss: 5.9928(6.3486), Acc: 0.0469(0.0245)
2021-12-28 03:32:05,374 Epoch[005/310], Step[1000/1251], Loss: 6.5459(6.3474), Acc: 0.0127(0.0245)
2021-12-28 03:33:08,579 Epoch[005/310], Step[1050/1251], Loss: 6.3215(6.3449), Acc: 0.0244(0.0247)
2021-12-28 03:34:12,671 Epoch[005/310], Step[1100/1251], Loss: 6.2211(6.3434), Acc: 0.0146(0.0249)
2021-12-28 03:35:15,678 Epoch[005/310], Step[1150/1251], Loss: 6.4775(6.3409), Acc: 0.0156(0.0251)
2021-12-28 03:36:19,216 Epoch[005/310], Step[1200/1251], Loss: 6.2204(6.3393), Acc: 0.0332(0.0252)
2021-12-28 03:37:22,665 Epoch[005/310], Step[1250/1251], Loss: 6.4387(6.3376), Acc: 0.0244(0.0252)
2021-12-28 03:37:24,643 ----- Epoch[005/310], Train Loss: 6.3376, Train Acc: 0.0252, time: 1644.38, Best Val(epoch4) Acc@1: 0.0754
2021-12-28 03:37:24,643 Now training epoch 6. LR=0.000300
2021-12-28 03:38:43,521 Epoch[006/310], Step[0000/1251], Loss: 6.3832(6.3832), Acc: 0.0166(0.0166)
2021-12-28 03:39:46,185 Epoch[006/310], Step[0050/1251], Loss: 6.3771(6.3250), Acc: 0.0391(0.0277)
2021-12-28 03:40:48,606 Epoch[006/310], Step[0100/1251], Loss: 6.0324(6.3163), Acc: 0.0479(0.0279)
2021-12-28 03:41:51,480 Epoch[006/310], Step[0150/1251], Loss: 6.3904(6.3213), Acc: 0.0234(0.0280)
2021-12-28 03:42:53,903 Epoch[006/310], Step[0200/1251], Loss: 6.2318(6.3172), Acc: 0.0361(0.0273)
2021-12-28 03:43:57,297 Epoch[006/310], Step[0250/1251], Loss: 6.3587(6.3079), Acc: 0.0127(0.0275)
2021-12-28 03:45:00,054 Epoch[006/310], Step[0300/1251], Loss: 6.2288(6.3094), Acc: 0.0322(0.0276)
2021-12-28 03:46:02,448 Epoch[006/310], Step[0350/1251], Loss: 6.1735(6.3069), Acc: 0.0117(0.0279)
2021-12-28 03:47:02,187 Epoch[006/310], Step[0400/1251], Loss: 6.4323(6.3060), Acc: 0.0322(0.0280)
2021-12-28 03:48:01,871 Epoch[006/310], Step[0450/1251], Loss: 6.1138(6.3005), Acc: 0.0264(0.0285)
2021-12-28 03:49:04,047 Epoch[006/310], Step[0500/1251], Loss: 6.4421(6.2980), Acc: 0.0176(0.0285)
2021-12-28 03:50:06,262 Epoch[006/310], Step[0550/1251], Loss: 6.2870(6.2979), Acc: 0.0459(0.0285)
2021-12-28 03:51:08,180 Epoch[006/310], Step[0600/1251], Loss: 6.2377(6.2931), Acc: 0.0205(0.0286)
2021-12-28 03:52:10,823 Epoch[006/310], Step[0650/1251], Loss: 6.1608(6.2910), Acc: 0.0283(0.0287)
2021-12-28 03:53:14,364 Epoch[006/310], Step[0700/1251], Loss: 6.1621(6.2884), Acc: 0.0293(0.0288)
2021-12-28 03:54:17,734 Epoch[006/310], Step[0750/1251], Loss: 6.2115(6.2851), Acc: 0.0449(0.0291)
2021-12-28 03:55:17,120 Epoch[006/310], Step[0800/1251], Loss: 6.4529(6.2834), Acc: 0.0264(0.0294)
2021-12-28 03:56:19,067 Epoch[006/310], Step[0850/1251], Loss: 6.2526(6.2824), Acc: 0.0264(0.0295)
2021-12-28 03:57:22,218 Epoch[006/310], Step[0900/1251], Loss: 6.0972(6.2802), Acc: 0.0381(0.0296)
2021-12-28 03:58:25,537 Epoch[006/310], Step[0950/1251], Loss: 6.2266(6.2776), Acc: 0.0508(0.0298)
2021-12-28 03:59:28,743 Epoch[006/310], Step[1000/1251], Loss: 6.2157(6.2748), Acc: 0.0410(0.0298)
2021-12-28 04:00:32,699 Epoch[006/310], Step[1050/1251], Loss: 6.1232(6.2727), Acc: 0.0557(0.0298)
2021-12-28 04:01:35,307 Epoch[006/310], Step[1100/1251], Loss: 6.3269(6.2704), Acc: 0.0225(0.0299)
2021-12-28 04:02:36,923 Epoch[006/310], Step[1150/1251], Loss: 6.1745(6.2677), Acc: 0.0635(0.0301)
2021-12-28 04:03:39,400 Epoch[006/310], Step[1200/1251], Loss: 6.5095(6.2665), Acc: 0.0244(0.0303)
2021-12-28 04:04:42,437 Epoch[006/310], Step[1250/1251], Loss: 6.2276(6.2656), Acc: 0.0234(0.0305)
2021-12-28 04:04:44,585 ----- Validation after Epoch: 6
2021-12-28 04:06:01,772 Val Step[0000/1563], Loss: 4.5633 (4.5633), Acc@1: 0.4375 (0.4375), Acc@5: 0.6250 (0.6250)
2021-12-28 04:06:03,477 Val Step[0050/1563], Loss: 5.5109 (4.6850), Acc@1: 0.0000 (0.1789), Acc@5: 0.0312 (0.3732)
2021-12-28 04:06:05,020 Val Step[0100/1563], Loss: 4.4117 (4.8921), Acc@1: 0.0625 (0.1334), Acc@5: 0.5938 (0.3082)
2021-12-28 04:06:06,585 Val Step[0150/1563], Loss: 4.7017 (4.8700), Acc@1: 0.2188 (0.1364), Acc@5: 0.4688 (0.3181)
2021-12-28 04:06:08,130 Val Step[0200/1563], Loss: 5.3797 (4.8908), Acc@1: 0.0312 (0.1357), Acc@5: 0.0938 (0.3148)
2021-12-28 04:06:09,676 Val Step[0250/1563], Loss: 5.8486 (4.8120), Acc@1: 0.0000 (0.1460), Acc@5: 0.0625 (0.3379)
2021-12-28 04:06:11,192 Val Step[0300/1563], Loss: 5.5279 (4.9053), Acc@1: 0.0000 (0.1266), Acc@5: 0.0625 (0.3060)
2021-12-28 04:06:12,706 Val Step[0350/1563], Loss: 5.1189 (4.9267), Acc@1: 0.1250 (0.1188), Acc@5: 0.3438 (0.2953)
2021-12-28 04:06:14,252 Val Step[0400/1563], Loss: 4.8271 (4.9363), Acc@1: 0.0625 (0.1169), Acc@5: 0.2812 (0.2923)
2021-12-28 04:06:15,762 Val Step[0450/1563], Loss: 3.3774 (4.9552), Acc@1: 0.0625 (0.1111), Acc@5: 0.7812 (0.2864)
2021-12-28 04:06:17,317 Val Step[0500/1563], Loss: 3.9731 (4.9440), Acc@1: 0.2812 (0.1122), Acc@5: 0.5938 (0.2886)
2021-12-28 04:06:19,010 Val Step[0550/1563], Loss: 4.7111 (4.9255), Acc@1: 0.1250 (0.1178), Acc@5: 0.4062 (0.2949)
2021-12-28 04:06:20,609 Val Step[0600/1563], Loss: 3.6423 (4.9401), Acc@1: 0.3750 (0.1163), Acc@5: 0.7188 (0.2916)
2021-12-28 04:06:22,095 Val Step[0650/1563], Loss: 5.1332 (4.9398), Acc@1: 0.0625 (0.1186), Acc@5: 0.2812 (0.2934)
2021-12-28 04:06:23,555 Val Step[0700/1563], Loss: 6.3114 (4.9704), Acc@1: 0.0000 (0.1155), Acc@5: 0.0000 (0.2883)
2021-12-28 04:06:25,156 Val Step[0750/1563], Loss: 4.6611 (4.9849), Acc@1: 0.1562 (0.1143), Acc@5: 0.4375 (0.2853)
2021-12-28 04:06:26,760 Val Step[0800/1563], Loss: 4.8228 (4.9978), Acc@1: 0.3125 (0.1126), Acc@5: 0.4375 (0.2838)
2021-12-28 04:06:28,212 Val Step[0850/1563], Loss: 6.0093 (5.0180), Acc@1: 0.0000 (0.1100), Acc@5: 0.0625 (0.2785)
2021-12-28 04:06:29,836 Val Step[0900/1563], Loss: 4.6084 (5.0112), Acc@1: 0.1875 (0.1112), Acc@5: 0.4062 (0.2817)
2021-12-28 04:06:31,427 Val Step[0950/1563], Loss: 5.4470 (5.0249), Acc@1: 0.0938 (0.1106), Acc@5: 0.2188 (0.2789)
2021-12-28 04:06:33,013 Val Step[1000/1563], Loss: 3.0228 (5.0330), Acc@1: 0.6250 (0.1107), Acc@5: 0.8125 (0.2774)
2021-12-28 04:06:34,570 Val Step[1050/1563], Loss: 5.3903 (5.0369), Acc@1: 0.0000 (0.1093), Acc@5: 0.0312 (0.2764)
2021-12-28 04:06:36,146 Val Step[1100/1563], Loss: 6.0367 (5.0460), Acc@1: 0.0000 (0.1098), Acc@5: 0.0312 (0.2745)
2021-12-28 04:06:37,633 Val Step[1150/1563], Loss: 5.4603 (5.0609), Acc@1: 0.0312 (0.1077), Acc@5: 0.1875 (0.2706)
2021-12-28 04:06:39,095 Val Step[1200/1563], Loss: 4.5543 (5.0723), Acc@1: 0.3750 (0.1068), Acc@5: 0.5000 (0.2682)
2021-12-28 04:06:40,602 Val Step[1250/1563], Loss: 4.0460 (5.0803), Acc@1: 0.3750 (0.1067), Acc@5: 0.5938 (0.2669)
2021-12-28 04:06:41,997 Val Step[1300/1563], Loss: 5.3720 (5.0869), Acc@1: 0.0000 (0.1055), Acc@5: 0.0938 (0.2646)
2021-12-28 04:06:43,465 Val Step[1350/1563], Loss: 4.7858 (5.1007), Acc@1: 0.0000 (0.1035), Acc@5: 0.0938 (0.2610)
2021-12-28 04:06:44,953 Val Step[1400/1563], Loss: 5.0671 (5.1027), Acc@1: 0.0625 (0.1032), Acc@5: 0.2812 (0.2608)
2021-12-28 04:06:46,393 Val Step[1450/1563], Loss: 5.1176 (5.1004), Acc@1: 0.0312 (0.1038), Acc@5: 0.1562 (0.2613)
2021-12-28 04:06:48,037 Val Step[1500/1563], Loss: 5.3115 (5.0704), Acc@1: 0.0938 (0.1081), Acc@5: 0.2812 (0.2689)
2021-12-28 04:06:49,546 Val Step[1550/1563], Loss: 2.5734 (5.0449), Acc@1: 0.7500 (0.1126), Acc@5: 0.8438 (0.2755)
2021-12-28 04:06:50,395 ----- Epoch[006/310], Validation Loss: 5.0382, Validation Acc@1: 0.1141, Validation Acc@5: 0.2775, time: 125.81
2021-12-28 04:06:50,395 ----- Epoch[006/310], Train Loss: 6.2656, Train Acc: 0.0305, time: 1639.94, Best Val(epoch6) Acc@1: 0.1141
2021-12-28 04:06:50,607 Max accuracy so far: 0.1141 at epoch_6
2021-12-28 04:06:50,608 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 04:06:50,608 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 04:06:50,690 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 04:06:50,691 Now training epoch 7. LR=0.000350
2021-12-28 04:08:10,652 Epoch[007/310], Step[0000/1251], Loss: 6.1249(6.1249), Acc: 0.0371(0.0371)
2021-12-28 04:09:11,614 Epoch[007/310], Step[0050/1251], Loss: 6.2588(6.2178), Acc: 0.0166(0.0355)
2021-12-28 04:10:12,470 Epoch[007/310], Step[0100/1251], Loss: 6.2578(6.2173), Acc: 0.0400(0.0337)
2021-12-28 04:11:14,628 Epoch[007/310], Step[0150/1251], Loss: 6.1997(6.2183), Acc: 0.0479(0.0328)
2021-12-28 04:12:17,507 Epoch[007/310], Step[0200/1251], Loss: 6.1353(6.2176), Acc: 0.0254(0.0337)
2021-12-28 04:13:19,782 Epoch[007/310], Step[0250/1251], Loss: 6.3358(6.2147), Acc: 0.0303(0.0344)
2021-12-28 04:14:22,669 Epoch[007/310], Step[0300/1251], Loss: 6.1561(6.2097), Acc: 0.0293(0.0342)
2021-12-28 04:15:24,996 Epoch[007/310], Step[0350/1251], Loss: 6.3189(6.2076), Acc: 0.0205(0.0346)
2021-12-28 04:16:26,856 Epoch[007/310], Step[0400/1251], Loss: 6.0281(6.2121), Acc: 0.0322(0.0344)
2021-12-28 04:17:30,418 Epoch[007/310], Step[0450/1251], Loss: 6.0641(6.2103), Acc: 0.0293(0.0347)
2021-12-28 04:18:33,407 Epoch[007/310], Step[0500/1251], Loss: 6.3147(6.2107), Acc: 0.0322(0.0347)
2021-12-28 04:19:35,661 Epoch[007/310], Step[0550/1251], Loss: 6.1012(6.2088), Acc: 0.0391(0.0348)
2021-12-28 04:20:37,729 Epoch[007/310], Step[0600/1251], Loss: 6.4069(6.2091), Acc: 0.0215(0.0350)
2021-12-28 04:21:39,673 Epoch[007/310], Step[0650/1251], Loss: 6.3839(6.2051), Acc: 0.0195(0.0353)
2021-12-28 04:22:41,466 Epoch[007/310], Step[0700/1251], Loss: 6.2860(6.2024), Acc: 0.0205(0.0356)
2021-12-28 04:23:44,693 Epoch[007/310], Step[0750/1251], Loss: 6.3975(6.2006), Acc: 0.0410(0.0357)
2021-12-28 04:24:46,326 Epoch[007/310], Step[0800/1251], Loss: 6.0580(6.1996), Acc: 0.0273(0.0358)
2021-12-28 04:25:49,436 Epoch[007/310], Step[0850/1251], Loss: 6.3677(6.1981), Acc: 0.0293(0.0361)
2021-12-28 04:26:50,950 Epoch[007/310], Step[0900/1251], Loss: 6.0862(6.1959), Acc: 0.0244(0.0362)
2021-12-28 04:27:53,429 Epoch[007/310], Step[0950/1251], Loss: 6.1484(6.1939), Acc: 0.0547(0.0362)
2021-12-28 04:28:54,093 Epoch[007/310], Step[1000/1251], Loss: 5.9857(6.1934), Acc: 0.0391(0.0364)
2021-12-28 04:29:57,004 Epoch[007/310], Step[1050/1251], Loss: 6.1963(6.1923), Acc: 0.0430(0.0366)
2021-12-28 04:30:58,615 Epoch[007/310], Step[1100/1251], Loss: 6.3372(6.1920), Acc: 0.0186(0.0367)
2021-12-28 04:32:00,906 Epoch[007/310], Step[1150/1251], Loss: 6.3451(6.1915), Acc: 0.0420(0.0368)
2021-12-28 04:33:02,650 Epoch[007/310], Step[1200/1251], Loss: 6.0608(6.1892), Acc: 0.0420(0.0370)
2021-12-28 04:34:05,629 Epoch[007/310], Step[1250/1251], Loss: 6.3188(6.1875), Acc: 0.0303(0.0372)
2021-12-28 04:34:07,943 ----- Epoch[007/310], Train Loss: 6.1875, Train Acc: 0.0372, time: 1637.25, Best Val(epoch6) Acc@1: 0.1141
2021-12-28 04:34:07,943 Now training epoch 8. LR=0.000400
2021-12-28 04:35:29,412 Epoch[008/310], Step[0000/1251], Loss: 6.1598(6.1598), Acc: 0.0400(0.0400)
2021-12-28 04:36:29,589 Epoch[008/310], Step[0050/1251], Loss: 6.0677(6.1395), Acc: 0.0664(0.0439)
2021-12-28 04:37:31,980 Epoch[008/310], Step[0100/1251], Loss: 5.9164(6.1500), Acc: 0.0322(0.0408)
2021-12-28 04:38:35,089 Epoch[008/310], Step[0150/1251], Loss: 6.2306(6.1479), Acc: 0.0537(0.0402)
2021-12-28 04:39:37,239 Epoch[008/310], Step[0200/1251], Loss: 6.3538(6.1446), Acc: 0.0186(0.0398)
2021-12-28 04:40:41,226 Epoch[008/310], Step[0250/1251], Loss: 5.9127(6.1374), Acc: 0.0332(0.0398)
2021-12-28 04:41:44,920 Epoch[008/310], Step[0300/1251], Loss: 6.2269(6.1285), Acc: 0.0615(0.0408)
2021-12-28 04:42:48,458 Epoch[008/310], Step[0350/1251], Loss: 6.1453(6.1303), Acc: 0.0479(0.0408)
2021-12-28 04:43:51,587 Epoch[008/310], Step[0400/1251], Loss: 6.2626(6.1299), Acc: 0.0283(0.0410)
2021-12-28 04:44:54,894 Epoch[008/310], Step[0450/1251], Loss: 6.1011(6.1269), Acc: 0.0645(0.0418)
2021-12-28 04:45:57,388 Epoch[008/310], Step[0500/1251], Loss: 6.0233(6.1257), Acc: 0.0703(0.0418)
2021-12-28 04:47:00,433 Epoch[008/310], Step[0550/1251], Loss: 6.1314(6.1222), Acc: 0.0332(0.0417)
2021-12-28 04:48:02,484 Epoch[008/310], Step[0600/1251], Loss: 6.2690(6.1198), Acc: 0.0518(0.0423)
2021-12-28 04:49:06,106 Epoch[008/310], Step[0650/1251], Loss: 5.8832(6.1203), Acc: 0.0459(0.0426)
2021-12-28 04:50:10,161 Epoch[008/310], Step[0700/1251], Loss: 6.2126(6.1209), Acc: 0.0557(0.0429)
2021-12-28 04:51:13,393 Epoch[008/310], Step[0750/1251], Loss: 5.8652(6.1181), Acc: 0.0576(0.0430)
2021-12-28 04:52:16,707 Epoch[008/310], Step[0800/1251], Loss: 6.1133(6.1137), Acc: 0.0469(0.0433)
2021-12-28 04:53:19,773 Epoch[008/310], Step[0850/1251], Loss: 6.0166(6.1093), Acc: 0.0557(0.0437)
2021-12-28 04:54:23,694 Epoch[008/310], Step[0900/1251], Loss: 6.2069(6.1086), Acc: 0.0332(0.0437)
2021-12-28 04:55:27,836 Epoch[008/310], Step[0950/1251], Loss: 6.2659(6.1079), Acc: 0.0557(0.0437)
2021-12-28 04:56:29,508 Epoch[008/310], Step[1000/1251], Loss: 6.3163(6.1074), Acc: 0.0283(0.0437)
2021-12-28 04:57:33,031 Epoch[008/310], Step[1050/1251], Loss: 5.7493(6.1072), Acc: 0.0566(0.0436)
2021-12-28 04:58:35,480 Epoch[008/310], Step[1100/1251], Loss: 6.1479(6.1046), Acc: 0.0479(0.0439)
2021-12-28 04:59:38,360 Epoch[008/310], Step[1150/1251], Loss: 5.9699(6.1021), Acc: 0.0625(0.0440)
2021-12-28 05:00:41,408 Epoch[008/310], Step[1200/1251], Loss: 6.1207(6.1003), Acc: 0.0127(0.0439)
2021-12-28 05:01:43,995 Epoch[008/310], Step[1250/1251], Loss: 6.0955(6.0959), Acc: 0.0762(0.0444)
2021-12-28 05:01:46,154 ----- Validation after Epoch: 8
2021-12-28 05:02:53,459 Val Step[0000/1563], Loss: 4.2232 (4.2232), Acc@1: 0.3750 (0.3750), Acc@5: 0.5625 (0.5625)
2021-12-28 05:02:54,988 Val Step[0050/1563], Loss: 5.3471 (4.1724), Acc@1: 0.0625 (0.2279), Acc@5: 0.0625 (0.4761)
2021-12-28 05:02:56,508 Val Step[0100/1563], Loss: 4.2527 (4.5074), Acc@1: 0.2188 (0.1764), Acc@5: 0.5625 (0.3815)
2021-12-28 05:02:58,057 Val Step[0150/1563], Loss: 3.9361 (4.4241), Acc@1: 0.3750 (0.1898), Acc@5: 0.5625 (0.4000)
2021-12-28 05:02:59,505 Val Step[0200/1563], Loss: 5.2120 (4.4405), Acc@1: 0.0312 (0.1920), Acc@5: 0.1562 (0.3960)
2021-12-28 05:03:01,058 Val Step[0250/1563], Loss: 4.7525 (4.3457), Acc@1: 0.0625 (0.2044), Acc@5: 0.2500 (0.4137)
2021-12-28 05:03:02,625 Val Step[0300/1563], Loss: 4.8609 (4.3945), Acc@1: 0.0000 (0.1842), Acc@5: 0.1562 (0.3905)
2021-12-28 05:03:04,094 Val Step[0350/1563], Loss: 4.1755 (4.3781), Acc@1: 0.1250 (0.1805), Acc@5: 0.4062 (0.3918)
2021-12-28 05:03:05,644 Val Step[0400/1563], Loss: 4.3990 (4.3451), Acc@1: 0.0000 (0.1792), Acc@5: 0.0938 (0.3967)
2021-12-28 05:03:07,220 Val Step[0450/1563], Loss: 3.0670 (4.3572), Acc@1: 0.0312 (0.1728), Acc@5: 0.8125 (0.3907)
2021-12-28 05:03:08,766 Val Step[0500/1563], Loss: 3.4612 (4.3369), Acc@1: 0.3750 (0.1752), Acc@5: 0.6562 (0.3951)
2021-12-28 05:03:10,262 Val Step[0550/1563], Loss: 3.6726 (4.3301), Acc@1: 0.3750 (0.1790), Acc@5: 0.6250 (0.3983)
2021-12-28 05:03:11,877 Val Step[0600/1563], Loss: 2.9074 (4.3485), Acc@1: 0.5312 (0.1764), Acc@5: 0.7500 (0.3935)
2021-12-28 05:03:13,359 Val Step[0650/1563], Loss: 4.5197 (4.3566), Acc@1: 0.1562 (0.1783), Acc@5: 0.5000 (0.3939)
2021-12-28 05:03:14,815 Val Step[0700/1563], Loss: 5.8308 (4.4032), Acc@1: 0.0312 (0.1735), Acc@5: 0.1562 (0.3857)
2021-12-28 05:03:16,318 Val Step[0750/1563], Loss: 4.0523 (4.4388), Acc@1: 0.3438 (0.1699), Acc@5: 0.5625 (0.3787)
2021-12-28 05:03:17,860 Val Step[0800/1563], Loss: 4.4569 (4.4670), Acc@1: 0.3438 (0.1669), Acc@5: 0.5938 (0.3745)
2021-12-28 05:03:19,308 Val Step[0850/1563], Loss: 5.0408 (4.4992), Acc@1: 0.0625 (0.1642), Acc@5: 0.2188 (0.3681)
2021-12-28 05:03:20,829 Val Step[0900/1563], Loss: 4.3009 (4.5020), Acc@1: 0.2500 (0.1655), Acc@5: 0.5000 (0.3696)
2021-12-28 05:03:22,433 Val Step[0950/1563], Loss: 5.1388 (4.5251), Acc@1: 0.1562 (0.1642), Acc@5: 0.3438 (0.3654)
2021-12-28 05:03:23,908 Val Step[1000/1563], Loss: 2.8231 (4.5383), Acc@1: 0.5625 (0.1638), Acc@5: 0.8125 (0.3639)
2021-12-28 05:03:25,375 Val Step[1050/1563], Loss: 4.4427 (4.5463), Acc@1: 0.0938 (0.1622), Acc@5: 0.4062 (0.3630)
2021-12-28 05:03:26,948 Val Step[1100/1563], Loss: 5.9044 (4.5624), Acc@1: 0.0000 (0.1618), Acc@5: 0.0312 (0.3596)
2021-12-28 05:03:28,459 Val Step[1150/1563], Loss: 4.7928 (4.5845), Acc@1: 0.1250 (0.1595), Acc@5: 0.3750 (0.3555)
2021-12-28 05:03:29,995 Val Step[1200/1563], Loss: 3.0901 (4.6010), Acc@1: 0.5938 (0.1578), Acc@5: 0.6562 (0.3527)
2021-12-28 05:03:31,538 Val Step[1250/1563], Loss: 3.7739 (4.6168), Acc@1: 0.3438 (0.1571), Acc@5: 0.6250 (0.3494)
2021-12-28 05:03:33,038 Val Step[1300/1563], Loss: 4.5142 (4.6266), Acc@1: 0.0938 (0.1555), Acc@5: 0.4688 (0.3474)
2021-12-28 05:03:34,641 Val Step[1350/1563], Loss: 4.1959 (4.6451), Acc@1: 0.0625 (0.1530), Acc@5: 0.2188 (0.3425)
2021-12-28 05:03:36,200 Val Step[1400/1563], Loss: 4.9120 (4.6528), Acc@1: 0.0312 (0.1517), Acc@5: 0.2188 (0.3412)
2021-12-28 05:03:37,668 Val Step[1450/1563], Loss: 4.9033 (4.6534), Acc@1: 0.0000 (0.1524), Acc@5: 0.1562 (0.3411)
2021-12-28 05:03:39,117 Val Step[1500/1563], Loss: 4.6982 (4.6249), Acc@1: 0.1875 (0.1564), Acc@5: 0.4062 (0.3479)
2021-12-28 05:03:40,583 Val Step[1550/1563], Loss: 1.8617 (4.5963), Acc@1: 0.7500 (0.1622), Acc@5: 0.8750 (0.3541)
2021-12-28 05:03:41,442 ----- Epoch[008/310], Validation Loss: 4.5880, Validation Acc@1: 0.1640, Validation Acc@5: 0.3561, time: 115.29
2021-12-28 05:03:41,442 ----- Epoch[008/310], Train Loss: 6.0959, Train Acc: 0.0444, time: 1658.21, Best Val(epoch8) Acc@1: 0.1640
2021-12-28 05:03:41,647 Max accuracy so far: 0.1640 at epoch_8
2021-12-28 05:03:41,648 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 05:03:41,648 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 05:03:41,743 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 05:03:41,744 Now training epoch 9. LR=0.000450
2021-12-28 05:05:11,722 Epoch[009/310], Step[0000/1251], Loss: 6.1375(6.1375), Acc: 0.0537(0.0537)
2021-12-28 05:06:14,082 Epoch[009/310], Step[0050/1251], Loss: 6.4220(6.0726), Acc: 0.0264(0.0480)
2021-12-28 05:07:15,942 Epoch[009/310], Step[0100/1251], Loss: 5.9458(6.0661), Acc: 0.0430(0.0489)
2021-12-28 05:08:19,244 Epoch[009/310], Step[0150/1251], Loss: 6.0973(6.0586), Acc: 0.0615(0.0492)
2021-12-28 05:09:20,625 Epoch[009/310], Step[0200/1251], Loss: 5.9580(6.0590), Acc: 0.0791(0.0492)
2021-12-28 05:10:22,951 Epoch[009/310], Step[0250/1251], Loss: 5.9317(6.0513), Acc: 0.0605(0.0488)
2021-12-28 05:11:24,442 Epoch[009/310], Step[0300/1251], Loss: 5.8982(6.0513), Acc: 0.0566(0.0492)
2021-12-28 05:12:28,234 Epoch[009/310], Step[0350/1251], Loss: 6.0040(6.0509), Acc: 0.0273(0.0494)
2021-12-28 05:13:31,232 Epoch[009/310], Step[0400/1251], Loss: 5.9850(6.0504), Acc: 0.0693(0.0490)
2021-12-28 05:14:34,215 Epoch[009/310], Step[0450/1251], Loss: 5.7783(6.0447), Acc: 0.0850(0.0495)
2021-12-28 05:15:36,559 Epoch[009/310], Step[0500/1251], Loss: 5.9475(6.0422), Acc: 0.0684(0.0495)
2021-12-28 05:16:38,939 Epoch[009/310], Step[0550/1251], Loss: 5.7596(6.0385), Acc: 0.0684(0.0499)
2021-12-28 05:17:41,469 Epoch[009/310], Step[0600/1251], Loss: 6.0748(6.0374), Acc: 0.0566(0.0505)
2021-12-28 05:18:45,370 Epoch[009/310], Step[0650/1251], Loss: 6.2354(6.0360), Acc: 0.0293(0.0503)
2021-12-28 05:19:48,612 Epoch[009/310], Step[0700/1251], Loss: 5.7653(6.0344), Acc: 0.0928(0.0502)
2021-12-28 05:20:51,897 Epoch[009/310], Step[0750/1251], Loss: 5.9508(6.0323), Acc: 0.0615(0.0506)
2021-12-28 05:21:54,084 Epoch[009/310], Step[0800/1251], Loss: 5.9214(6.0305), Acc: 0.0762(0.0512)
2021-12-28 05:22:57,213 Epoch[009/310], Step[0850/1251], Loss: 5.8969(6.0296), Acc: 0.0264(0.0515)
2021-12-28 05:24:00,534 Epoch[009/310], Step[0900/1251], Loss: 6.0071(6.0262), Acc: 0.0684(0.0520)
2021-12-28 05:25:04,377 Epoch[009/310], Step[0950/1251], Loss: 5.9693(6.0216), Acc: 0.0352(0.0524)
2021-12-28 05:26:08,354 Epoch[009/310], Step[1000/1251], Loss: 5.9004(6.0211), Acc: 0.0420(0.0523)
2021-12-28 05:27:11,898 Epoch[009/310], Step[1050/1251], Loss: 6.0209(6.0180), Acc: 0.0342(0.0527)
2021-12-28 05:28:15,226 Epoch[009/310], Step[1100/1251], Loss: 6.0232(6.0151), Acc: 0.0576(0.0528)
2021-12-28 05:29:18,970 Epoch[009/310], Step[1150/1251], Loss: 5.7622(6.0125), Acc: 0.0244(0.0530)
2021-12-28 05:30:21,213 Epoch[009/310], Step[1200/1251], Loss: 6.0653(6.0093), Acc: 0.0322(0.0532)
2021-12-28 05:31:24,372 Epoch[009/310], Step[1250/1251], Loss: 6.1896(6.0062), Acc: 0.0166(0.0535)
2021-12-28 05:31:26,336 ----- Epoch[009/310], Train Loss: 6.0062, Train Acc: 0.0535, time: 1664.59, Best Val(epoch8) Acc@1: 0.1640
2021-12-28 05:31:26,336 Now training epoch 10. LR=0.000500
2021-12-28 05:32:45,897 Epoch[010/310], Step[0000/1251], Loss: 5.9057(5.9057), Acc: 0.0771(0.0771)
2021-12-28 05:33:44,684 Epoch[010/310], Step[0050/1251], Loss: 5.8278(6.0196), Acc: 0.0420(0.0511)
2021-12-28 05:34:46,429 Epoch[010/310], Step[0100/1251], Loss: 6.1851(6.0031), Acc: 0.0625(0.0541)
2021-12-28 05:35:49,353 Epoch[010/310], Step[0150/1251], Loss: 5.5877(5.9804), Acc: 0.0781(0.0562)
2021-12-28 05:36:52,636 Epoch[010/310], Step[0200/1251], Loss: 5.8691(5.9794), Acc: 0.0508(0.0551)
2021-12-28 05:37:55,469 Epoch[010/310], Step[0250/1251], Loss: 5.5436(5.9822), Acc: 0.0840(0.0550)
2021-12-28 05:38:57,420 Epoch[010/310], Step[0300/1251], Loss: 5.6745(5.9744), Acc: 0.0684(0.0554)
2021-12-28 05:40:00,564 Epoch[010/310], Step[0350/1251], Loss: 6.2661(5.9726), Acc: 0.0264(0.0562)
2021-12-28 05:41:02,936 Epoch[010/310], Step[0400/1251], Loss: 5.8947(5.9635), Acc: 0.1016(0.0570)
2021-12-28 05:42:05,370 Epoch[010/310], Step[0450/1251], Loss: 5.7852(5.9632), Acc: 0.1113(0.0574)
2021-12-28 05:43:07,686 Epoch[010/310], Step[0500/1251], Loss: 6.1793(5.9571), Acc: 0.0615(0.0582)
2021-12-28 05:44:09,071 Epoch[010/310], Step[0550/1251], Loss: 5.8962(5.9543), Acc: 0.0361(0.0588)
2021-12-28 05:45:11,855 Epoch[010/310], Step[0600/1251], Loss: 5.6894(5.9519), Acc: 0.0684(0.0592)
2021-12-28 05:46:15,616 Epoch[010/310], Step[0650/1251], Loss: 5.9433(5.9501), Acc: 0.0947(0.0591)
2021-12-28 05:47:17,403 Epoch[010/310], Step[0700/1251], Loss: 6.1404(5.9512), Acc: 0.0762(0.0590)
2021-12-28 05:48:20,781 Epoch[010/310], Step[0750/1251], Loss: 5.7930(5.9446), Acc: 0.0732(0.0591)
2021-12-28 05:49:22,759 Epoch[010/310], Step[0800/1251], Loss: 5.6043(5.9418), Acc: 0.0586(0.0595)
2021-12-28 05:50:25,374 Epoch[010/310], Step[0850/1251], Loss: 6.1905(5.9380), Acc: 0.0479(0.0599)
2021-12-28 05:51:28,053 Epoch[010/310], Step[0900/1251], Loss: 5.8473(5.9343), Acc: 0.0859(0.0603)
2021-12-28 05:52:32,409 Epoch[010/310], Step[0950/1251], Loss: 6.1441(5.9327), Acc: 0.0459(0.0604)
2021-12-28 05:53:35,948 Epoch[010/310], Step[1000/1251], Loss: 6.3020(5.9314), Acc: 0.0449(0.0603)
2021-12-28 05:54:39,337 Epoch[010/310], Step[1050/1251], Loss: 5.7227(5.9293), Acc: 0.0674(0.0603)
2021-12-28 05:55:43,189 Epoch[010/310], Step[1100/1251], Loss: 6.0977(5.9276), Acc: 0.0605(0.0603)
2021-12-28 05:56:46,178 Epoch[010/310], Step[1150/1251], Loss: 5.9378(5.9244), Acc: 0.0400(0.0605)
2021-12-28 05:57:48,697 Epoch[010/310], Step[1200/1251], Loss: 5.5351(5.9224), Acc: 0.1016(0.0611)
2021-12-28 05:58:51,712 Epoch[010/310], Step[1250/1251], Loss: 5.9723(5.9210), Acc: 0.0645(0.0613)
2021-12-28 05:58:53,745 ----- Validation after Epoch: 10
2021-12-28 05:59:56,848 Val Step[0000/1563], Loss: 3.3614 (3.3614), Acc@1: 0.5938 (0.5938), Acc@5: 0.6562 (0.6562)
2021-12-28 05:59:58,403 Val Step[0050/1563], Loss: 5.0017 (3.6205), Acc@1: 0.0312 (0.2947), Acc@5: 0.2500 (0.5729)
2021-12-28 05:59:59,976 Val Step[0100/1563], Loss: 3.7050 (3.9848), Acc@1: 0.3438 (0.2376), Acc@5: 0.6875 (0.4833)
2021-12-28 06:00:01,461 Val Step[0150/1563], Loss: 3.4145 (3.9175), Acc@1: 0.3125 (0.2554), Acc@5: 0.6875 (0.4977)
2021-12-28 06:00:03,024 Val Step[0200/1563], Loss: 4.3618 (3.9550), Acc@1: 0.1562 (0.2565), Acc@5: 0.4062 (0.4939)
2021-12-28 06:00:04,511 Val Step[0250/1563], Loss: 4.6667 (3.8546), Acc@1: 0.0938 (0.2743), Acc@5: 0.3750 (0.5110)
2021-12-28 06:00:06,026 Val Step[0300/1563], Loss: 4.5568 (3.9448), Acc@1: 0.0312 (0.2500), Acc@5: 0.2188 (0.4864)
2021-12-28 06:00:07,477 Val Step[0350/1563], Loss: 4.3282 (3.9784), Acc@1: 0.1562 (0.2382), Acc@5: 0.4062 (0.4767)
2021-12-28 06:00:08,922 Val Step[0400/1563], Loss: 4.1717 (3.9758), Acc@1: 0.1250 (0.2345), Acc@5: 0.4062 (0.4776)
2021-12-28 06:00:10,526 Val Step[0450/1563], Loss: 2.4690 (3.9968), Acc@1: 0.3125 (0.2271), Acc@5: 0.8438 (0.4728)
2021-12-28 06:00:12,007 Val Step[0500/1563], Loss: 2.6461 (3.9822), Acc@1: 0.5625 (0.2295), Acc@5: 0.7812 (0.4769)
2021-12-28 06:00:13,582 Val Step[0550/1563], Loss: 3.3437 (3.9610), Acc@1: 0.3750 (0.2349), Acc@5: 0.6875 (0.4815)
2021-12-28 06:00:15,207 Val Step[0600/1563], Loss: 3.0816 (3.9733), Acc@1: 0.4375 (0.2343), Acc@5: 0.7500 (0.4795)
2021-12-28 06:00:16,713 Val Step[0650/1563], Loss: 4.3881 (3.9699), Acc@1: 0.1875 (0.2360), Acc@5: 0.5000 (0.4796)
2021-12-28 06:00:18,187 Val Step[0700/1563], Loss: 5.5253 (4.0084), Acc@1: 0.0312 (0.2314), Acc@5: 0.1562 (0.4706)
2021-12-28 06:00:19,662 Val Step[0750/1563], Loss: 3.6820 (4.0355), Acc@1: 0.2812 (0.2278), Acc@5: 0.6562 (0.4635)
2021-12-28 06:00:21,267 Val Step[0800/1563], Loss: 4.2590 (4.0586), Acc@1: 0.2812 (0.2246), Acc@5: 0.5312 (0.4599)
2021-12-28 06:00:22,875 Val Step[0850/1563], Loss: 4.3357 (4.0815), Acc@1: 0.1875 (0.2214), Acc@5: 0.5312 (0.4561)
2021-12-28 06:00:24,521 Val Step[0900/1563], Loss: 3.0156 (4.0778), Acc@1: 0.4375 (0.2229), Acc@5: 0.6562 (0.4566)
2021-12-28 06:00:26,151 Val Step[0950/1563], Loss: 4.2880 (4.0977), Acc@1: 0.3125 (0.2213), Acc@5: 0.5625 (0.4525)
2021-12-28 06:00:27,737 Val Step[1000/1563], Loss: 2.4319 (4.1093), Acc@1: 0.6250 (0.2202), Acc@5: 0.8750 (0.4506)
2021-12-28 06:00:29,350 Val Step[1050/1563], Loss: 3.5920 (4.1148), Acc@1: 0.2188 (0.2182), Acc@5: 0.5625 (0.4490)
2021-12-28 06:00:30,950 Val Step[1100/1563], Loss: 4.9806 (4.1329), Acc@1: 0.0938 (0.2170), Acc@5: 0.2812 (0.4448)
2021-12-28 06:00:32,494 Val Step[1150/1563], Loss: 4.4535 (4.1544), Acc@1: 0.3125 (0.2138), Acc@5: 0.4375 (0.4397)
2021-12-28 06:00:34,084 Val Step[1200/1563], Loss: 3.5266 (4.1698), Acc@1: 0.5000 (0.2127), Acc@5: 0.5938 (0.4364)
2021-12-28 06:00:35,806 Val Step[1250/1563], Loss: 3.0643 (4.1868), Acc@1: 0.5625 (0.2111), Acc@5: 0.7812 (0.4325)
2021-12-28 06:00:37,399 Val Step[1300/1563], Loss: 3.9191 (4.1959), Acc@1: 0.1875 (0.2092), Acc@5: 0.5625 (0.4301)
2021-12-28 06:00:38,944 Val Step[1350/1563], Loss: 3.6659 (4.2122), Acc@1: 0.0625 (0.2066), Acc@5: 0.4688 (0.4263)
2021-12-28 06:00:40,444 Val Step[1400/1563], Loss: 3.5362 (4.2166), Acc@1: 0.3750 (0.2057), Acc@5: 0.6250 (0.4253)
2021-12-28 06:00:41,897 Val Step[1450/1563], Loss: 3.8765 (4.2127), Acc@1: 0.1562 (0.2068), Acc@5: 0.5312 (0.4257)
2021-12-28 06:00:43,291 Val Step[1500/1563], Loss: 4.0640 (4.1851), Acc@1: 0.2500 (0.2113), Acc@5: 0.5312 (0.4324)
2021-12-28 06:00:44,742 Val Step[1550/1563], Loss: 2.0297 (4.1610), Acc@1: 0.7500 (0.2158), Acc@5: 0.8125 (0.4368)
2021-12-28 06:00:45,656 ----- Epoch[010/310], Validation Loss: 4.1550, Validation Acc@1: 0.2171, Validation Acc@5: 0.4382, time: 111.91
2021-12-28 06:00:45,656 ----- Epoch[010/310], Train Loss: 5.9210, Train Acc: 0.0613, time: 1647.40, Best Val(epoch10) Acc@1: 0.2171
2021-12-28 06:00:45,854 Max accuracy so far: 0.2171 at epoch_10
2021-12-28 06:00:45,854 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 06:00:45,854 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 06:00:45,948 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 06:00:46,114 ----- Save model: /root/paddlejob/workspace/output/train-20211228-01-15-41/PiT-Epoch-10-Loss-5.902993043263753.pdparams
2021-12-28 06:00:46,114 ----- Save optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/PiT-Epoch-10-Loss-5.902993043263753.pdopt
2021-12-28 06:00:46,186 ----- Save ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/PiT-Epoch-10-Loss-5.902993043263753-EMA.pdparams
2021-12-28 06:00:46,187 Now training epoch 11. LR=0.000550
2021-12-28 06:02:09,873 Epoch[011/310], Step[0000/1251], Loss: 5.8402(5.8402), Acc: 0.0859(0.0859)
2021-12-28 06:03:12,813 Epoch[011/310], Step[0050/1251], Loss: 5.7676(5.8775), Acc: 0.0947(0.0660)
2021-12-28 06:04:15,080 Epoch[011/310], Step[0100/1251], Loss: 5.7378(5.8863), Acc: 0.0771(0.0642)
2021-12-28 06:05:18,122 Epoch[011/310], Step[0150/1251], Loss: 6.3123(5.8754), Acc: 0.0156(0.0647)
2021-12-28 06:06:19,941 Epoch[011/310], Step[0200/1251], Loss: 5.9759(5.8723), Acc: 0.0537(0.0660)
2021-12-28 06:07:23,400 Epoch[011/310], Step[0250/1251], Loss: 5.7472(5.8702), Acc: 0.0391(0.0658)
2021-12-28 06:08:25,907 Epoch[011/310], Step[0300/1251], Loss: 6.2743(5.8620), Acc: 0.0488(0.0665)
2021-12-28 06:09:27,581 Epoch[011/310], Step[0350/1251], Loss: 5.3266(5.8579), Acc: 0.0967(0.0672)
2021-12-28 06:10:29,807 Epoch[011/310], Step[0400/1251], Loss: 5.7600(5.8611), Acc: 0.0537(0.0668)
2021-12-28 06:11:32,882 Epoch[011/310], Step[0450/1251], Loss: 6.1723(5.8570), Acc: 0.0420(0.0677)
2021-12-28 06:12:35,011 Epoch[011/310], Step[0500/1251], Loss: 6.1038(5.8551), Acc: 0.0703(0.0682)
2021-12-28 06:13:37,176 Epoch[011/310], Step[0550/1251], Loss: 5.8383(5.8546), Acc: 0.0869(0.0683)
2021-12-28 06:14:41,401 Epoch[011/310], Step[0600/1251], Loss: 5.7110(5.8537), Acc: 0.0557(0.0684)
2021-12-28 06:15:45,093 Epoch[011/310], Step[0650/1251], Loss: 6.0166(5.8511), Acc: 0.0742(0.0684)
2021-12-28 06:16:48,432 Epoch[011/310], Step[0700/1251], Loss: 5.8511(5.8470), Acc: 0.1064(0.0691)
2021-12-28 06:17:52,181 Epoch[011/310], Step[0750/1251], Loss: 5.5441(5.8464), Acc: 0.0811(0.0690)
2021-12-28 06:18:55,389 Epoch[011/310], Step[0800/1251], Loss: 5.3270(5.8432), Acc: 0.1406(0.0696)
2021-12-28 06:19:58,815 Epoch[011/310], Step[0850/1251], Loss: 5.6028(5.8402), Acc: 0.0742(0.0700)
2021-12-28 06:21:02,047 Epoch[011/310], Step[0900/1251], Loss: 5.7378(5.8377), Acc: 0.0605(0.0702)
2021-12-28 06:22:04,979 Epoch[011/310], Step[0950/1251], Loss: 5.9499(5.8371), Acc: 0.0410(0.0702)
2021-12-28 06:23:08,912 Epoch[011/310], Step[1000/1251], Loss: 5.8664(5.8348), Acc: 0.1055(0.0703)
2021-12-28 06:24:10,683 Epoch[011/310], Step[1050/1251], Loss: 5.6577(5.8299), Acc: 0.1104(0.0709)
2021-12-28 06:25:12,401 Epoch[011/310], Step[1100/1251], Loss: 6.1046(5.8230), Acc: 0.0596(0.0717)
2021-12-28 06:26:12,418 Epoch[011/310], Step[1150/1251], Loss: 5.7338(5.8200), Acc: 0.0771(0.0721)
2021-12-28 06:27:15,493 Epoch[011/310], Step[1200/1251], Loss: 5.9164(5.8185), Acc: 0.0752(0.0723)
2021-12-28 06:28:18,271 Epoch[011/310], Step[1250/1251], Loss: 6.1286(5.8177), Acc: 0.0693(0.0727)
2021-12-28 06:28:20,468 ----- Epoch[011/310], Train Loss: 5.8177, Train Acc: 0.0727, time: 1654.28, Best Val(epoch10) Acc@1: 0.2171
2021-12-28 06:28:20,468 Now training epoch 12. LR=0.000600
2021-12-28 06:29:38,665 Epoch[012/310], Step[0000/1251], Loss: 6.0570(6.0570), Acc: 0.0381(0.0381)
2021-12-28 06:30:41,314 Epoch[012/310], Step[0050/1251], Loss: 5.8181(5.7380), Acc: 0.1064(0.0820)
2021-12-28 06:31:43,372 Epoch[012/310], Step[0100/1251], Loss: 5.7663(5.7348), Acc: 0.0967(0.0798)
2021-12-28 06:32:46,346 Epoch[012/310], Step[0150/1251], Loss: 5.7197(5.7267), Acc: 0.0449(0.0813)
2021-12-28 06:33:48,556 Epoch[012/310], Step[0200/1251], Loss: 5.5393(5.7278), Acc: 0.0791(0.0801)
2021-12-28 06:34:51,709 Epoch[012/310], Step[0250/1251], Loss: 5.6361(5.7302), Acc: 0.0869(0.0803)
2021-12-28 06:35:54,997 Epoch[012/310], Step[0300/1251], Loss: 5.7490(5.7251), Acc: 0.1104(0.0806)
2021-12-28 06:36:56,802 Epoch[012/310], Step[0350/1251], Loss: 5.8216(5.7301), Acc: 0.0293(0.0810)
2021-12-28 06:37:59,556 Epoch[012/310], Step[0400/1251], Loss: 6.0363(5.7352), Acc: 0.0234(0.0801)
2021-12-28 06:39:02,422 Epoch[012/310], Step[0450/1251], Loss: 5.6102(5.7361), Acc: 0.0244(0.0798)
2021-12-28 06:40:03,979 Epoch[012/310], Step[0500/1251], Loss: 5.8202(5.7335), Acc: 0.0918(0.0802)
2021-12-28 06:41:06,250 Epoch[012/310], Step[0550/1251], Loss: 5.5210(5.7274), Acc: 0.1133(0.0805)
2021-12-28 06:42:08,431 Epoch[012/310], Step[0600/1251], Loss: 5.6166(5.7269), Acc: 0.0469(0.0811)
2021-12-28 06:43:11,050 Epoch[012/310], Step[0650/1251], Loss: 5.7082(5.7314), Acc: 0.0889(0.0812)
2021-12-28 06:44:14,543 Epoch[012/310], Step[0700/1251], Loss: 5.7577(5.7295), Acc: 0.0645(0.0816)
2021-12-28 06:45:17,407 Epoch[012/310], Step[0750/1251], Loss: 5.8784(5.7307), Acc: 0.0498(0.0810)
2021-12-28 06:46:20,037 Epoch[012/310], Step[0800/1251], Loss: 5.4949(5.7295), Acc: 0.0918(0.0810)
2021-12-28 06:47:21,101 Epoch[012/310], Step[0850/1251], Loss: 5.6947(5.7287), Acc: 0.0557(0.0811)
2021-12-28 06:48:24,420 Epoch[012/310], Step[0900/1251], Loss: 5.7077(5.7290), Acc: 0.1045(0.0813)
2021-12-28 06:49:26,528 Epoch[012/310], Step[0950/1251], Loss: 5.2144(5.7243), Acc: 0.0986(0.0816)
2021-12-28 06:50:29,110 Epoch[012/310], Step[1000/1251], Loss: 5.6991(5.7224), Acc: 0.1113(0.0820)
2021-12-28 06:51:32,578 Epoch[012/310], Step[1050/1251], Loss: 5.5236(5.7197), Acc: 0.1455(0.0823)
2021-12-28 06:52:36,548 Epoch[012/310], Step[1100/1251], Loss: 5.9494(5.7191), Acc: 0.1074(0.0824)
2021-12-28 06:53:40,458 Epoch[012/310], Step[1150/1251], Loss: 6.0036(5.7179), Acc: 0.0684(0.0825)
2021-12-28 06:54:44,039 Epoch[012/310], Step[1200/1251], Loss: 5.7296(5.7139), Acc: 0.1172(0.0829)
2021-12-28 06:55:48,005 Epoch[012/310], Step[1250/1251], Loss: 5.9216(5.7130), Acc: 0.0625(0.0831)
2021-12-28 06:55:50,042 ----- Validation after Epoch: 12
2021-12-28 06:56:49,210 Val Step[0000/1563], Loss: 2.2751 (2.2751), Acc@1: 0.6562 (0.6562), Acc@5: 0.7812 (0.7812)
2021-12-28 06:56:50,761 Val Step[0050/1563], Loss: 4.5798 (2.9692), Acc@1: 0.1250 (0.4062), Acc@5: 0.3438 (0.6906)
2021-12-28 06:56:52,295 Val Step[0100/1563], Loss: 3.5935 (3.4439), Acc@1: 0.2812 (0.3202), Acc@5: 0.5938 (0.5789)
2021-12-28 06:56:53,832 Val Step[0150/1563], Loss: 2.8282 (3.3259), Acc@1: 0.4688 (0.3413), Acc@5: 0.6562 (0.6024)
2021-12-28 06:56:55,284 Val Step[0200/1563], Loss: 4.0620 (3.3768), Acc@1: 0.0625 (0.3388), Acc@5: 0.3125 (0.5927)
2021-12-28 06:56:56,722 Val Step[0250/1563], Loss: 3.9279 (3.2616), Acc@1: 0.2500 (0.3576), Acc@5: 0.5000 (0.6129)
2021-12-28 06:56:58,351 Val Step[0300/1563], Loss: 4.0314 (3.3429), Acc@1: 0.0000 (0.3312), Acc@5: 0.3438 (0.5906)
2021-12-28 06:56:59,932 Val Step[0350/1563], Loss: 3.5568 (3.3600), Acc@1: 0.1875 (0.3226), Acc@5: 0.5625 (0.5882)
2021-12-28 06:57:01,529 Val Step[0400/1563], Loss: 3.4484 (3.3540), Acc@1: 0.0938 (0.3177), Acc@5: 0.6250 (0.5888)
2021-12-28 06:57:03,214 Val Step[0450/1563], Loss: 2.0452 (3.3638), Acc@1: 0.1250 (0.3114), Acc@5: 0.9375 (0.5892)
2021-12-28 06:57:04,769 Val Step[0500/1563], Loss: 2.0934 (3.3606), Acc@1: 0.6875 (0.3123), Acc@5: 0.8438 (0.5891)
2021-12-28 06:57:06,379 Val Step[0550/1563], Loss: 2.6090 (3.3342), Acc@1: 0.4688 (0.3201), Acc@5: 0.7500 (0.5932)
2021-12-28 06:57:08,025 Val Step[0600/1563], Loss: 2.5502 (3.3505), Acc@1: 0.5312 (0.3182), Acc@5: 0.8125 (0.5894)
2021-12-28 06:57:09,554 Val Step[0650/1563], Loss: 3.0434 (3.3669), Acc@1: 0.3750 (0.3175), Acc@5: 0.7188 (0.5866)
2021-12-28 06:57:11,049 Val Step[0700/1563], Loss: 5.0194 (3.4197), Acc@1: 0.0938 (0.3116), Acc@5: 0.2812 (0.5764)
2021-12-28 06:57:12,714 Val Step[0750/1563], Loss: 3.8146 (3.4640), Acc@1: 0.2188 (0.3059), Acc@5: 0.5000 (0.5672)
2021-12-28 06:57:14,246 Val Step[0800/1563], Loss: 3.8243 (3.5115), Acc@1: 0.3125 (0.2994), Acc@5: 0.6562 (0.5586)
2021-12-28 06:57:15,739 Val Step[0850/1563], Loss: 3.6224 (3.5470), Acc@1: 0.2812 (0.2952), Acc@5: 0.5000 (0.5513)
2021-12-28 06:57:17,227 Val Step[0900/1563], Loss: 1.9592 (3.5490), Acc@1: 0.7188 (0.2973), Acc@5: 0.8750 (0.5516)
2021-12-28 06:57:18,784 Val Step[0950/1563], Loss: 3.9040 (3.5790), Acc@1: 0.3125 (0.2946), Acc@5: 0.6250 (0.5461)
2021-12-28 06:57:20,310 Val Step[1000/1563], Loss: 2.2156 (3.6065), Acc@1: 0.6250 (0.2905), Acc@5: 0.7812 (0.5412)
2021-12-28 06:57:21,771 Val Step[1050/1563], Loss: 2.6299 (3.6245), Acc@1: 0.5625 (0.2868), Acc@5: 0.8125 (0.5374)
2021-12-28 06:57:23,276 Val Step[1100/1563], Loss: 4.6088 (3.6465), Acc@1: 0.1875 (0.2842), Acc@5: 0.3438 (0.5322)
2021-12-28 06:57:24,798 Val Step[1150/1563], Loss: 3.7726 (3.6724), Acc@1: 0.3750 (0.2803), Acc@5: 0.4688 (0.5271)
2021-12-28 06:57:26,361 Val Step[1200/1563], Loss: 3.0892 (3.6948), Acc@1: 0.5625 (0.2780), Acc@5: 0.6250 (0.5228)
2021-12-28 06:57:27,856 Val Step[1250/1563], Loss: 3.2957 (3.7205), Acc@1: 0.4062 (0.2746), Acc@5: 0.6250 (0.5173)
2021-12-28 06:57:29,332 Val Step[1300/1563], Loss: 4.7067 (3.7308), Acc@1: 0.1562 (0.2733), Acc@5: 0.2500 (0.5146)
2021-12-28 06:57:30,834 Val Step[1350/1563], Loss: 3.4855 (3.7562), Acc@1: 0.0938 (0.2690), Acc@5: 0.4688 (0.5094)
2021-12-28 06:57:32,327 Val Step[1400/1563], Loss: 3.6936 (3.7663), Acc@1: 0.3438 (0.2673), Acc@5: 0.6250 (0.5075)
2021-12-28 06:57:33,763 Val Step[1450/1563], Loss: 3.9730 (3.7684), Acc@1: 0.1250 (0.2672), Acc@5: 0.5938 (0.5066)
2021-12-28 06:57:35,389 Val Step[1500/1563], Loss: 4.1823 (3.7428), Acc@1: 0.1562 (0.2714), Acc@5: 0.5000 (0.5127)
2021-12-28 06:57:36,931 Val Step[1550/1563], Loss: 1.3357 (3.7194), Acc@1: 0.8438 (0.2759), Acc@5: 0.8750 (0.5167)
2021-12-28 06:57:37,838 ----- Epoch[012/310], Validation Loss: 3.7119, Validation Acc@1: 0.2777, Validation Acc@5: 0.5181, time: 107.79
2021-12-28 06:57:37,838 ----- Epoch[012/310], Train Loss: 5.7130, Train Acc: 0.0831, time: 1649.57, Best Val(epoch12) Acc@1: 0.2777
2021-12-28 06:57:38,049 Max accuracy so far: 0.2777 at epoch_12
2021-12-28 06:57:38,050 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 06:57:38,050 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 06:57:38,124 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 06:57:38,124 Now training epoch 13. LR=0.000650
2021-12-28 06:58:54,229 Epoch[013/310], Step[0000/1251], Loss: 5.5341(5.5341), Acc: 0.0918(0.0918)
2021-12-28 06:59:54,837 Epoch[013/310], Step[0050/1251], Loss: 5.4687(5.6141), Acc: 0.1006(0.0999)
2021-12-28 07:00:57,748 Epoch[013/310], Step[0100/1251], Loss: 5.7414(5.6487), Acc: 0.1260(0.0924)
2021-12-28 07:02:00,705 Epoch[013/310], Step[0150/1251], Loss: 5.4365(5.6701), Acc: 0.1523(0.0884)
2021-12-28 07:03:03,530 Epoch[013/310], Step[0200/1251], Loss: 5.6929(5.6722), Acc: 0.1182(0.0888)
2021-12-28 07:04:05,344 Epoch[013/310], Step[0250/1251], Loss: 5.8056(5.6752), Acc: 0.0781(0.0881)
2021-12-28 07:05:08,415 Epoch[013/310], Step[0300/1251], Loss: 5.6439(5.6797), Acc: 0.0420(0.0872)
2021-12-28 07:06:10,664 Epoch[013/310], Step[0350/1251], Loss: 5.7731(5.6783), Acc: 0.0820(0.0870)
2021-12-28 07:07:13,501 Epoch[013/310], Step[0400/1251], Loss: 5.8148(5.6739), Acc: 0.0762(0.0877)
2021-12-28 07:08:16,755 Epoch[013/310], Step[0450/1251], Loss: 5.1762(5.6707), Acc: 0.0986(0.0886)
2021-12-28 07:09:19,298 Epoch[013/310], Step[0500/1251], Loss: 6.1760(5.6609), Acc: 0.0420(0.0897)
2021-12-28 07:10:22,161 Epoch[013/310], Step[0550/1251], Loss: 6.1408(5.6653), Acc: 0.0674(0.0898)
2021-12-28 07:11:26,006 Epoch[013/310], Step[0600/1251], Loss: 5.6699(5.6625), Acc: 0.1084(0.0888)
2021-12-28 07:12:28,855 Epoch[013/310], Step[0650/1251], Loss: 5.4778(5.6606), Acc: 0.0254(0.0887)
2021-12-28 07:13:31,472 Epoch[013/310], Step[0700/1251], Loss: 5.2617(5.6608), Acc: 0.0996(0.0883)
2021-12-28 07:14:32,632 Epoch[013/310], Step[0750/1251], Loss: 5.5408(5.6597), Acc: 0.0488(0.0890)
2021-12-28 07:15:34,333 Epoch[013/310], Step[0800/1251], Loss: 5.1423(5.6569), Acc: 0.1865(0.0894)
2021-12-28 07:16:36,654 Epoch[013/310], Step[0850/1251], Loss: 5.6327(5.6529), Acc: 0.0879(0.0898)
2021-12-28 07:17:39,660 Epoch[013/310], Step[0900/1251], Loss: 5.3896(5.6519), Acc: 0.0986(0.0898)
2021-12-28 07:18:43,016 Epoch[013/310], Step[0950/1251], Loss: 5.7526(5.6491), Acc: 0.0713(0.0897)
2021-12-28 07:19:43,960 Epoch[013/310], Step[1000/1251], Loss: 5.9515(5.6457), Acc: 0.1133(0.0906)
2021-12-28 07:20:47,378 Epoch[013/310], Step[1050/1251], Loss: 5.6315(5.6461), Acc: 0.0977(0.0906)
2021-12-28 07:21:48,969 Epoch[013/310], Step[1100/1251], Loss: 5.6471(5.6443), Acc: 0.1270(0.0904)
2021-12-28 07:22:51,580 Epoch[013/310], Step[1150/1251], Loss: 5.2454(5.6449), Acc: 0.1504(0.0908)
2021-12-28 07:23:54,996 Epoch[013/310], Step[1200/1251], Loss: 5.3272(5.6404), Acc: 0.1738(0.0912)
2021-12-28 07:24:57,198 Epoch[013/310], Step[1250/1251], Loss: 5.7538(5.6371), Acc: 0.0557(0.0914)
2021-12-28 07:24:59,723 ----- Epoch[013/310], Train Loss: 5.6371, Train Acc: 0.0914, time: 1641.59, Best Val(epoch12) Acc@1: 0.2777
2021-12-28 07:24:59,723 Now training epoch 14. LR=0.000700
2021-12-28 07:26:18,015 Epoch[014/310], Step[0000/1251], Loss: 5.1568(5.1568), Acc: 0.0674(0.0674)
2021-12-28 07:27:20,746 Epoch[014/310], Step[0050/1251], Loss: 5.4195(5.5535), Acc: 0.1396(0.1049)
2021-12-28 07:28:22,525 Epoch[014/310], Step[0100/1251], Loss: 5.4277(5.5332), Acc: 0.0479(0.1008)
2021-12-28 07:29:24,088 Epoch[014/310], Step[0150/1251], Loss: 5.7859(5.5383), Acc: 0.0986(0.0988)
2021-12-28 07:30:25,624 Epoch[014/310], Step[0200/1251], Loss: 5.5962(5.5371), Acc: 0.0928(0.1006)
2021-12-28 07:31:28,045 Epoch[014/310], Step[0250/1251], Loss: 5.7579(5.5505), Acc: 0.0459(0.1002)
2021-12-28 07:32:30,149 Epoch[014/310], Step[0300/1251], Loss: 5.5318(5.5492), Acc: 0.0654(0.0989)
2021-12-28 07:33:32,917 Epoch[014/310], Step[0350/1251], Loss: 5.3217(5.5579), Acc: 0.1611(0.0987)
2021-12-28 07:34:36,805 Epoch[014/310], Step[0400/1251], Loss: 5.3984(5.5593), Acc: 0.0771(0.0987)
2021-12-28 07:35:38,611 Epoch[014/310], Step[0450/1251], Loss: 5.4874(5.5625), Acc: 0.1533(0.0989)
2021-12-28 07:36:40,433 Epoch[014/310], Step[0500/1251], Loss: 5.9691(5.5571), Acc: 0.0527(0.1004)
2021-12-28 07:37:43,129 Epoch[014/310], Step[0550/1251], Loss: 5.9723(5.5585), Acc: 0.0850(0.1005)
2021-12-28 07:38:46,662 Epoch[014/310], Step[0600/1251], Loss: 5.3409(5.5601), Acc: 0.0918(0.1009)
2021-12-28 07:39:50,418 Epoch[014/310], Step[0650/1251], Loss: 5.6691(5.5593), Acc: 0.0664(0.1009)
2021-12-28 07:40:53,981 Epoch[014/310], Step[0700/1251], Loss: 5.4780(5.5566), Acc: 0.1377(0.1009)
2021-12-28 07:41:56,762 Epoch[014/310], Step[0750/1251], Loss: 5.6194(5.5575), Acc: 0.0967(0.1012)
2021-12-28 07:43:00,920 Epoch[014/310], Step[0800/1251], Loss: 5.2731(5.5568), Acc: 0.0742(0.1006)
2021-12-28 07:44:01,699 Epoch[014/310], Step[0850/1251], Loss: 5.4809(5.5515), Acc: 0.1152(0.1006)
2021-12-28 07:45:05,461 Epoch[014/310], Step[0900/1251], Loss: 5.2640(5.5519), Acc: 0.1719(0.1006)
2021-12-28 07:46:08,962 Epoch[014/310], Step[0950/1251], Loss: 5.4471(5.5491), Acc: 0.1221(0.1006)
2021-12-28 07:47:12,786 Epoch[014/310], Step[1000/1251], Loss: 5.7125(5.5472), Acc: 0.0947(0.1008)
2021-12-28 07:48:16,148 Epoch[014/310], Step[1050/1251], Loss: 5.7761(5.5443), Acc: 0.1230(0.1013)
2021-12-28 07:49:18,663 Epoch[014/310], Step[1100/1251], Loss: 5.7787(5.5437), Acc: 0.1055(0.1013)
2021-12-28 07:50:22,249 Epoch[014/310], Step[1150/1251], Loss: 5.3068(5.5390), Acc: 0.1787(0.1018)
2021-12-28 07:51:24,646 Epoch[014/310], Step[1200/1251], Loss: 5.5040(5.5364), Acc: 0.1045(0.1021)
2021-12-28 07:52:28,075 Epoch[014/310], Step[1250/1251], Loss: 5.7638(5.5323), Acc: 0.1553(0.1028)
2021-12-28 07:52:30,051 ----- Validation after Epoch: 14
2021-12-28 07:53:29,809 Val Step[0000/1563], Loss: 2.1811 (2.1811), Acc@1: 0.7188 (0.7188), Acc@5: 0.8438 (0.8438)
2021-12-28 07:53:31,440 Val Step[0050/1563], Loss: 4.4849 (2.6487), Acc@1: 0.1250 (0.4792), Acc@5: 0.2812 (0.7439)
2021-12-28 07:53:32,987 Val Step[0100/1563], Loss: 3.6699 (3.1255), Acc@1: 0.2500 (0.3787), Acc@5: 0.5625 (0.6399)
2021-12-28 07:53:34,660 Val Step[0150/1563], Loss: 2.4119 (2.9714), Acc@1: 0.5938 (0.4048), Acc@5: 0.6250 (0.6612)
2021-12-28 07:53:36,157 Val Step[0200/1563], Loss: 2.8017 (3.0192), Acc@1: 0.3125 (0.3996), Acc@5: 0.7188 (0.6519)
2021-12-28 07:53:37,656 Val Step[0250/1563], Loss: 3.3411 (2.9078), Acc@1: 0.3438 (0.4147), Acc@5: 0.5625 (0.6712)
2021-12-28 07:53:39,139 Val Step[0300/1563], Loss: 3.4233 (2.9918), Acc@1: 0.1875 (0.3895), Acc@5: 0.5312 (0.6527)
2021-12-28 07:53:40,638 Val Step[0350/1563], Loss: 3.0630 (3.0171), Acc@1: 0.3438 (0.3790), Acc@5: 0.7188 (0.6489)
2021-12-28 07:53:42,131 Val Step[0400/1563], Loss: 3.2515 (3.0016), Acc@1: 0.0938 (0.3735), Acc@5: 0.6562 (0.6549)
2021-12-28 07:53:43,760 Val Step[0450/1563], Loss: 1.7834 (3.0129), Acc@1: 0.3125 (0.3678), Acc@5: 0.9062 (0.6529)
2021-12-28 07:53:45,314 Val Step[0500/1563], Loss: 1.4457 (3.0006), Acc@1: 0.7500 (0.3696), Acc@5: 0.9062 (0.6548)
2021-12-28 07:53:46,847 Val Step[0550/1563], Loss: 2.4378 (2.9789), Acc@1: 0.5000 (0.3770), Acc@5: 0.8125 (0.6572)
2021-12-28 07:53:48,456 Val Step[0600/1563], Loss: 1.8327 (2.9943), Acc@1: 0.6875 (0.3757), Acc@5: 0.8125 (0.6537)
2021-12-28 07:53:50,012 Val Step[0650/1563], Loss: 3.3495 (3.0118), Acc@1: 0.1875 (0.3735), Acc@5: 0.7500 (0.6498)
2021-12-28 07:53:51,460 Val Step[0700/1563], Loss: 4.7586 (3.0604), Acc@1: 0.1250 (0.3687), Acc@5: 0.4062 (0.6403)
2021-12-28 07:53:53,019 Val Step[0750/1563], Loss: 3.4697 (3.1058), Acc@1: 0.3125 (0.3627), Acc@5: 0.5625 (0.6311)
2021-12-28 07:53:54,459 Val Step[0800/1563], Loss: 3.4698 (3.1502), Acc@1: 0.3438 (0.3563), Acc@5: 0.6250 (0.6230)
2021-12-28 07:53:56,001 Val Step[0850/1563], Loss: 3.4147 (3.1871), Acc@1: 0.3438 (0.3525), Acc@5: 0.5625 (0.6156)
2021-12-28 07:53:57,615 Val Step[0900/1563], Loss: 1.8324 (3.1897), Acc@1: 0.6875 (0.3548), Acc@5: 0.8438 (0.6148)
2021-12-28 07:53:59,200 Val Step[0950/1563], Loss: 3.5437 (3.2271), Acc@1: 0.5000 (0.3497), Acc@5: 0.5312 (0.6070)
2021-12-28 07:54:00,626 Val Step[1000/1563], Loss: 1.5925 (3.2526), Acc@1: 0.7812 (0.3457), Acc@5: 0.9062 (0.6022)
2021-12-28 07:54:02,129 Val Step[1050/1563], Loss: 2.6973 (3.2688), Acc@1: 0.4375 (0.3424), Acc@5: 0.7812 (0.5997)
2021-12-28 07:54:03,622 Val Step[1100/1563], Loss: 3.8354 (3.2896), Acc@1: 0.2812 (0.3400), Acc@5: 0.5625 (0.5952)
2021-12-28 07:54:05,143 Val Step[1150/1563], Loss: 4.4016 (3.3171), Acc@1: 0.2500 (0.3362), Acc@5: 0.4375 (0.5898)
2021-12-28 07:54:06,732 Val Step[1200/1563], Loss: 2.8305 (3.3401), Acc@1: 0.4375 (0.3336), Acc@5: 0.5938 (0.5856)
2021-12-28 07:54:08,262 Val Step[1250/1563], Loss: 2.1695 (3.3640), Acc@1: 0.6562 (0.3306), Acc@5: 0.8438 (0.5810)
2021-12-28 07:54:09,759 Val Step[1300/1563], Loss: 3.3855 (3.3760), Acc@1: 0.2812 (0.3279), Acc@5: 0.5938 (0.5783)
2021-12-28 07:54:11,254 Val Step[1350/1563], Loss: 3.3267 (3.3990), Acc@1: 0.0625 (0.3235), Acc@5: 0.5000 (0.5736)
2021-12-28 07:54:12,789 Val Step[1400/1563], Loss: 3.1653 (3.4071), Acc@1: 0.3438 (0.3224), Acc@5: 0.7188 (0.5719)
2021-12-28 07:54:14,286 Val Step[1450/1563], Loss: 3.6773 (3.4070), Acc@1: 0.1250 (0.3226), Acc@5: 0.5938 (0.5717)
2021-12-28 07:54:15,833 Val Step[1500/1563], Loss: 3.9023 (3.3786), Acc@1: 0.1562 (0.3274), Acc@5: 0.5312 (0.5776)
2021-12-28 07:54:17,365 Val Step[1550/1563], Loss: 1.3219 (3.3645), Acc@1: 0.8750 (0.3300), Acc@5: 0.8750 (0.5798)
2021-12-28 07:54:18,243 ----- Epoch[014/310], Validation Loss: 3.3590, Validation Acc@1: 0.3315, Validation Acc@5: 0.5807, time: 108.19
2021-12-28 07:54:18,243 ----- Epoch[014/310], Train Loss: 5.5323, Train Acc: 0.1028, time: 1650.33, Best Val(epoch14) Acc@1: 0.3315
2021-12-28 07:54:18,449 Max accuracy so far: 0.3315 at epoch_14
2021-12-28 07:54:18,450 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 07:54:18,450 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 07:54:18,543 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 07:54:18,543 Now training epoch 15. LR=0.000750
2021-12-28 07:55:33,566 Epoch[015/310], Step[0000/1251], Loss: 5.3800(5.3800), Acc: 0.1641(0.1641)
2021-12-28 07:56:35,525 Epoch[015/310], Step[0050/1251], Loss: 5.5559(5.4395), Acc: 0.0889(0.1038)
2021-12-28 07:57:37,782 Epoch[015/310], Step[0100/1251], Loss: 4.8861(5.4635), Acc: 0.0938(0.1039)
2021-12-28 07:58:40,838 Epoch[015/310], Step[0150/1251], Loss: 5.4250(5.4564), Acc: 0.1641(0.1054)
2021-12-28 07:59:43,879 Epoch[015/310], Step[0200/1251], Loss: 5.4568(5.4654), Acc: 0.0732(0.1071)
2021-12-28 08:00:47,928 Epoch[015/310], Step[0250/1251], Loss: 5.6064(5.4700), Acc: 0.0625(0.1069)
2021-12-28 08:01:51,119 Epoch[015/310], Step[0300/1251], Loss: 5.2058(5.4650), Acc: 0.1221(0.1083)
2021-12-28 08:02:53,890 Epoch[015/310], Step[0350/1251], Loss: 5.1110(5.4630), Acc: 0.0713(0.1098)
2021-12-28 08:03:56,768 Epoch[015/310], Step[0400/1251], Loss: 5.3831(5.4696), Acc: 0.0625(0.1095)
2021-12-28 08:04:59,235 Epoch[015/310], Step[0450/1251], Loss: 5.6619(5.4666), Acc: 0.0381(0.1105)
2021-12-28 08:06:01,427 Epoch[015/310], Step[0500/1251], Loss: 5.2747(5.4600), Acc: 0.0303(0.1107)
2021-12-28 08:07:03,265 Epoch[015/310], Step[0550/1251], Loss: 4.9379(5.4610), Acc: 0.2090(0.1107)
2021-12-28 08:08:06,142 Epoch[015/310], Step[0600/1251], Loss: 5.9671(5.4691), Acc: 0.0859(0.1095)
2021-12-28 08:09:08,484 Epoch[015/310], Step[0650/1251], Loss: 5.4935(5.4691), Acc: 0.0869(0.1106)
2021-12-28 08:10:10,939 Epoch[015/310], Step[0700/1251], Loss: 5.4801(5.4697), Acc: 0.1025(0.1109)
2021-12-28 08:11:13,505 Epoch[015/310], Step[0750/1251], Loss: 5.3737(5.4705), Acc: 0.1602(0.1106)
2021-12-28 08:12:16,058 Epoch[015/310], Step[0800/1251], Loss: 5.3650(5.4678), Acc: 0.0664(0.1109)
2021-12-28 08:13:19,499 Epoch[015/310], Step[0850/1251], Loss: 5.4618(5.4674), Acc: 0.1074(0.1110)
2021-12-28 08:14:21,860 Epoch[015/310], Step[0900/1251], Loss: 5.3026(5.4687), Acc: 0.1650(0.1113)
2021-12-28 08:15:25,665 Epoch[015/310], Step[0950/1251], Loss: 5.2481(5.4679), Acc: 0.2080(0.1119)
2021-12-28 08:16:29,794 Epoch[015/310], Step[1000/1251], Loss: 5.1861(5.4648), Acc: 0.0996(0.1124)
2021-12-28 08:17:33,085 Epoch[015/310], Step[1050/1251], Loss: 5.4850(5.4625), Acc: 0.0693(0.1126)
2021-12-28 08:18:36,754 Epoch[015/310], Step[1100/1251], Loss: 5.6788(5.4615), Acc: 0.1182(0.1129)
2021-12-28 08:19:40,778 Epoch[015/310], Step[1150/1251], Loss: 5.5979(5.4572), Acc: 0.1514(0.1136)
2021-12-28 08:20:43,194 Epoch[015/310], Step[1200/1251], Loss: 5.3761(5.4550), Acc: 0.1543(0.1141)
2021-12-28 08:21:43,460 Epoch[015/310], Step[1250/1251], Loss: 5.7135(5.4543), Acc: 0.0791(0.1142)
2021-12-28 08:21:45,435 ----- Epoch[015/310], Train Loss: 5.4543, Train Acc: 0.1142, time: 1646.89, Best Val(epoch14) Acc@1: 0.3315
2021-12-28 08:21:45,435 Now training epoch 16. LR=0.000800
2021-12-28 08:23:05,870 Epoch[016/310], Step[0000/1251], Loss: 5.8383(5.8383), Acc: 0.1006(0.1006)
2021-12-28 08:24:07,610 Epoch[016/310], Step[0050/1251], Loss: 5.0874(5.4113), Acc: 0.1270(0.1180)
2021-12-28 08:25:09,863 Epoch[016/310], Step[0100/1251], Loss: 5.7592(5.4017), Acc: 0.1162(0.1213)
2021-12-28 08:26:13,640 Epoch[016/310], Step[0150/1251], Loss: 4.8757(5.4005), Acc: 0.1152(0.1228)
2021-12-28 08:27:16,440 Epoch[016/310], Step[0200/1251], Loss: 5.5050(5.4063), Acc: 0.1826(0.1214)
2021-12-28 08:28:19,075 Epoch[016/310], Step[0250/1251], Loss: 5.4394(5.4093), Acc: 0.1797(0.1219)
2021-12-28 08:29:22,554 Epoch[016/310], Step[0300/1251], Loss: 5.4271(5.4031), Acc: 0.1045(0.1220)
2021-12-28 08:30:26,531 Epoch[016/310], Step[0350/1251], Loss: 5.4102(5.4028), Acc: 0.0732(0.1217)
2021-12-28 08:31:30,393 Epoch[016/310], Step[0400/1251], Loss: 5.1642(5.3986), Acc: 0.1357(0.1211)
2021-12-28 08:32:33,407 Epoch[016/310], Step[0450/1251], Loss: 5.8684(5.4004), Acc: 0.0840(0.1203)
2021-12-28 08:33:35,735 Epoch[016/310], Step[0500/1251], Loss: 5.5560(5.4010), Acc: 0.0869(0.1214)
2021-12-28 08:34:37,941 Epoch[016/310], Step[0550/1251], Loss: 5.6167(5.3987), Acc: 0.1484(0.1214)
2021-12-28 08:35:39,878 Epoch[016/310], Step[0600/1251], Loss: 5.0835(5.3993), Acc: 0.1230(0.1213)
2021-12-28 08:36:40,508 Epoch[016/310], Step[0650/1251], Loss: 5.7317(5.3966), Acc: 0.1533(0.1220)
2021-12-28 08:37:43,284 Epoch[016/310], Step[0700/1251], Loss: 5.4023(5.3961), Acc: 0.1445(0.1220)
2021-12-28 08:38:45,005 Epoch[016/310], Step[0750/1251], Loss: 5.9730(5.3928), Acc: 0.0791(0.1229)
2021-12-28 08:39:47,753 Epoch[016/310], Step[0800/1251], Loss: 5.3877(5.3925), Acc: 0.1699(0.1232)
2021-12-28 08:40:51,512 Epoch[016/310], Step[0850/1251], Loss: 4.7347(5.3897), Acc: 0.1816(0.1236)
2021-12-28 08:41:55,012 Epoch[016/310], Step[0900/1251], Loss: 5.1141(5.3856), Acc: 0.1953(0.1240)
2021-12-28 08:42:56,953 Epoch[016/310], Step[0950/1251], Loss: 5.3257(5.3831), Acc: 0.1318(0.1242)
2021-12-28 08:44:00,254 Epoch[016/310], Step[1000/1251], Loss: 5.6917(5.3808), Acc: 0.1143(0.1245)
2021-12-28 08:45:03,940 Epoch[016/310], Step[1050/1251], Loss: 5.1012(5.3778), Acc: 0.0996(0.1247)
2021-12-28 08:46:06,148 Epoch[016/310], Step[1100/1251], Loss: 5.4246(5.3772), Acc: 0.1562(0.1247)
2021-12-28 08:47:11,087 Epoch[016/310], Step[1150/1251], Loss: 5.4267(5.3743), Acc: 0.0742(0.1251)
2021-12-28 08:48:14,133 Epoch[016/310], Step[1200/1251], Loss: 5.3510(5.3721), Acc: 0.0967(0.1256)
2021-12-28 08:49:15,612 Epoch[016/310], Step[1250/1251], Loss: 5.5410(5.3715), Acc: 0.1221(0.1257)
2021-12-28 08:49:17,559 ----- Validation after Epoch: 16
2021-12-28 08:50:14,087 Val Step[0000/1563], Loss: 2.0360 (2.0360), Acc@1: 0.7500 (0.7500), Acc@5: 0.8125 (0.8125)
2021-12-28 08:50:15,628 Val Step[0050/1563], Loss: 4.1025 (2.2220), Acc@1: 0.1562 (0.5509), Acc@5: 0.4062 (0.7923)
2021-12-28 08:50:17,122 Val Step[0100/1563], Loss: 3.3811 (2.7731), Acc@1: 0.1875 (0.4350), Acc@5: 0.6875 (0.6844)
2021-12-28 08:50:18,589 Val Step[0150/1563], Loss: 2.2938 (2.6053), Acc@1: 0.6562 (0.4743), Acc@5: 0.7812 (0.7127)
2021-12-28 08:50:20,227 Val Step[0200/1563], Loss: 2.4093 (2.6658), Acc@1: 0.3438 (0.4638), Acc@5: 0.8125 (0.7016)
2021-12-28 08:50:21,790 Val Step[0250/1563], Loss: 2.5843 (2.5700), Acc@1: 0.4062 (0.4777), Acc@5: 0.7812 (0.7214)
2021-12-28 08:50:23,268 Val Step[0300/1563], Loss: 2.5343 (2.6350), Acc@1: 0.4375 (0.4519), Acc@5: 0.9062 (0.7122)
2021-12-28 08:50:24,840 Val Step[0350/1563], Loss: 2.8143 (2.6376), Acc@1: 0.3750 (0.4464), Acc@5: 0.6875 (0.7152)
2021-12-28 08:50:26,342 Val Step[0400/1563], Loss: 3.1348 (2.6236), Acc@1: 0.1562 (0.4428), Acc@5: 0.6875 (0.7201)
2021-12-28 08:50:27,903 Val Step[0450/1563], Loss: 2.1188 (2.6463), Acc@1: 0.2812 (0.4356), Acc@5: 0.8438 (0.7169)
2021-12-28 08:50:29,469 Val Step[0500/1563], Loss: 1.4133 (2.6483), Acc@1: 0.7500 (0.4346), Acc@5: 0.9688 (0.7171)
2021-12-28 08:50:31,043 Val Step[0550/1563], Loss: 2.1658 (2.6270), Acc@1: 0.5312 (0.4422), Acc@5: 0.8125 (0.7198)
2021-12-28 08:50:32,709 Val Step[0600/1563], Loss: 1.8009 (2.6328), Acc@1: 0.7500 (0.4429), Acc@5: 0.8750 (0.7194)
2021-12-28 08:50:34,197 Val Step[0650/1563], Loss: 2.6033 (2.6601), Acc@1: 0.5312 (0.4392), Acc@5: 0.8125 (0.7148)
2021-12-28 08:50:35,806 Val Step[0700/1563], Loss: 4.7333 (2.7218), Acc@1: 0.1250 (0.4301), Acc@5: 0.3750 (0.7053)
2021-12-28 08:50:37,438 Val Step[0750/1563], Loss: 2.9790 (2.7806), Acc@1: 0.5000 (0.4213), Acc@5: 0.6562 (0.6927)
2021-12-28 08:50:39,054 Val Step[0800/1563], Loss: 3.1345 (2.8331), Acc@1: 0.3750 (0.4128), Acc@5: 0.7188 (0.6829)
2021-12-28 08:50:40,590 Val Step[0850/1563], Loss: 3.4384 (2.8732), Acc@1: 0.2812 (0.4072), Acc@5: 0.5312 (0.6756)
2021-12-28 08:50:42,217 Val Step[0900/1563], Loss: 1.5125 (2.8809), Acc@1: 0.7500 (0.4078), Acc@5: 0.8125 (0.6740)
2021-12-28 08:50:43,857 Val Step[0950/1563], Loss: 3.0851 (2.9201), Acc@1: 0.5000 (0.4028), Acc@5: 0.7812 (0.6668)
2021-12-28 08:50:45,419 Val Step[1000/1563], Loss: 1.4290 (2.9474), Acc@1: 0.7812 (0.3986), Acc@5: 0.9062 (0.6619)
2021-12-28 08:50:46,993 Val Step[1050/1563], Loss: 2.3221 (2.9685), Acc@1: 0.5312 (0.3945), Acc@5: 0.7500 (0.6587)
2021-12-28 08:50:48,557 Val Step[1100/1563], Loss: 3.4696 (2.9945), Acc@1: 0.4062 (0.3907), Acc@5: 0.5938 (0.6524)
2021-12-28 08:50:50,160 Val Step[1150/1563], Loss: 3.3387 (3.0268), Acc@1: 0.5000 (0.3863), Acc@5: 0.6562 (0.6460)
2021-12-28 08:50:51,752 Val Step[1200/1563], Loss: 2.1873 (3.0524), Acc@1: 0.5938 (0.3826), Acc@5: 0.7188 (0.6411)
2021-12-28 08:50:53,338 Val Step[1250/1563], Loss: 2.1212 (3.0747), Acc@1: 0.6562 (0.3802), Acc@5: 0.8438 (0.6371)
2021-12-28 08:50:54,843 Val Step[1300/1563], Loss: 3.2478 (3.0911), Acc@1: 0.2812 (0.3768), Acc@5: 0.5625 (0.6336)
2021-12-28 08:50:56,358 Val Step[1350/1563], Loss: 3.9684 (3.1223), Acc@1: 0.0312 (0.3713), Acc@5: 0.3438 (0.6275)
2021-12-28 08:50:57,824 Val Step[1400/1563], Loss: 3.8544 (3.1332), Acc@1: 0.2500 (0.3698), Acc@5: 0.5312 (0.6256)
2021-12-28 08:50:59,336 Val Step[1450/1563], Loss: 3.7140 (3.1357), Acc@1: 0.1562 (0.3692), Acc@5: 0.5625 (0.6252)
2021-12-28 08:51:00,821 Val Step[1500/1563], Loss: 2.8546 (3.1173), Acc@1: 0.3125 (0.3728), Acc@5: 0.8125 (0.6292)
2021-12-28 08:51:02,380 Val Step[1550/1563], Loss: 1.3854 (3.1030), Acc@1: 0.8125 (0.3755), Acc@5: 0.8750 (0.6313)
2021-12-28 08:51:03,462 ----- Epoch[016/310], Validation Loss: 3.1002, Validation Acc@1: 0.3763, Validation Acc@5: 0.6318, time: 105.90
2021-12-28 08:51:03,462 ----- Epoch[016/310], Train Loss: 5.3715, Train Acc: 0.1257, time: 1652.12, Best Val(epoch16) Acc@1: 0.3763
2021-12-28 08:51:03,670 Max accuracy so far: 0.3763 at epoch_16
2021-12-28 08:51:03,671 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 08:51:03,671 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 08:51:03,765 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 08:51:03,765 Now training epoch 17. LR=0.000850
2021-12-28 08:52:20,627 Epoch[017/310], Step[0000/1251], Loss: 5.1037(5.1037), Acc: 0.1641(0.1641)
2021-12-28 08:53:22,361 Epoch[017/310], Step[0050/1251], Loss: 5.0488(5.3625), Acc: 0.1494(0.1372)
2021-12-28 08:54:24,164 Epoch[017/310], Step[0100/1251], Loss: 5.7158(5.3392), Acc: 0.0391(0.1355)
2021-12-28 08:55:23,208 Epoch[017/310], Step[0150/1251], Loss: 5.2770(5.3111), Acc: 0.1445(0.1383)
2021-12-28 08:56:25,850 Epoch[017/310], Step[0200/1251], Loss: 5.4988(5.2967), Acc: 0.1221(0.1372)
2021-12-28 08:57:29,301 Epoch[017/310], Step[0250/1251], Loss: 5.3459(5.2982), Acc: 0.1826(0.1375)
2021-12-28 08:58:31,027 Epoch[017/310], Step[0300/1251], Loss: 5.1956(5.2979), Acc: 0.1309(0.1384)
2021-12-28 08:59:32,820 Epoch[017/310], Step[0350/1251], Loss: 5.1525(5.3041), Acc: 0.1279(0.1382)
2021-12-28 09:00:35,958 Epoch[017/310], Step[0400/1251], Loss: 4.9547(5.3045), Acc: 0.1396(0.1384)
2021-12-28 09:01:38,204 Epoch[017/310], Step[0450/1251], Loss: 5.2554(5.3111), Acc: 0.0938(0.1371)
2021-12-28 09:02:41,948 Epoch[017/310], Step[0500/1251], Loss: 5.2783(5.3150), Acc: 0.1162(0.1368)
2021-12-28 09:03:46,185 Epoch[017/310], Step[0550/1251], Loss: 5.1208(5.3122), Acc: 0.0605(0.1371)
2021-12-28 09:04:48,645 Epoch[017/310], Step[0600/1251], Loss: 4.8664(5.3077), Acc: 0.1660(0.1373)
2021-12-28 09:05:50,848 Epoch[017/310], Step[0650/1251], Loss: 5.1422(5.3115), Acc: 0.0303(0.1366)
2021-12-28 09:06:53,603 Epoch[017/310], Step[0700/1251], Loss: 5.7668(5.3138), Acc: 0.1123(0.1359)
2021-12-28 09:07:56,107 Epoch[017/310], Step[0750/1251], Loss: 5.3071(5.3120), Acc: 0.0898(0.1354)
2021-12-28 09:08:58,472 Epoch[017/310], Step[0800/1251], Loss: 5.2037(5.3082), Acc: 0.1953(0.1354)
2021-12-28 09:10:00,903 Epoch[017/310], Step[0850/1251], Loss: 5.0306(5.3087), Acc: 0.1113(0.1355)
2021-12-28 09:11:04,782 Epoch[017/310], Step[0900/1251], Loss: 5.7665(5.3116), Acc: 0.0752(0.1352)
2021-12-28 09:12:07,766 Epoch[017/310], Step[0950/1251], Loss: 5.4108(5.3086), Acc: 0.0420(0.1353)
2021-12-28 09:13:09,944 Epoch[017/310], Step[1000/1251], Loss: 5.5713(5.3071), Acc: 0.1885(0.1353)
2021-12-28 09:14:12,675 Epoch[017/310], Step[1050/1251], Loss: 5.2835(5.3029), Acc: 0.0928(0.1355)
2021-12-28 09:15:16,378 Epoch[017/310], Step[1100/1251], Loss: 5.5683(5.3015), Acc: 0.1240(0.1355)
2021-12-28 09:16:17,554 Epoch[017/310], Step[1150/1251], Loss: 5.1012(5.3015), Acc: 0.2109(0.1356)
2021-12-28 09:17:19,328 Epoch[017/310], Step[1200/1251], Loss: 5.7105(5.3003), Acc: 0.1367(0.1361)
2021-12-28 09:18:20,065 Epoch[017/310], Step[1250/1251], Loss: 5.5458(5.2991), Acc: 0.0918(0.1361)
2021-12-28 09:18:21,961 ----- Epoch[017/310], Train Loss: 5.2991, Train Acc: 0.1361, time: 1638.19, Best Val(epoch16) Acc@1: 0.3763
2021-12-28 09:18:21,962 Now training epoch 18. LR=0.000900
2021-12-28 09:19:43,879 Epoch[018/310], Step[0000/1251], Loss: 5.1961(5.1961), Acc: 0.2109(0.2109)
2021-12-28 09:20:46,487 Epoch[018/310], Step[0050/1251], Loss: 5.6486(5.2701), Acc: 0.1270(0.1269)
2021-12-28 09:21:47,927 Epoch[018/310], Step[0100/1251], Loss: 5.0782(5.3084), Acc: 0.0908(0.1299)
2021-12-28 09:22:50,987 Epoch[018/310], Step[0150/1251], Loss: 4.5768(5.2954), Acc: 0.1406(0.1317)
2021-12-28 09:23:54,720 Epoch[018/310], Step[0200/1251], Loss: 5.4717(5.2684), Acc: 0.0781(0.1358)
2021-12-28 09:24:57,494 Epoch[018/310], Step[0250/1251], Loss: 5.2406(5.2638), Acc: 0.1816(0.1384)
2021-12-28 09:26:00,744 Epoch[018/310], Step[0300/1251], Loss: 5.4910(5.2582), Acc: 0.1670(0.1402)
2021-12-28 09:27:02,627 Epoch[018/310], Step[0350/1251], Loss: 4.7432(5.2562), Acc: 0.0928(0.1403)
2021-12-28 09:28:04,871 Epoch[018/310], Step[0400/1251], Loss: 5.0113(5.2569), Acc: 0.0088(0.1393)
2021-12-28 09:29:07,069 Epoch[018/310], Step[0450/1251], Loss: 5.0892(5.2594), Acc: 0.1250(0.1398)
2021-12-28 09:30:09,962 Epoch[018/310], Step[0500/1251], Loss: 4.7892(5.2561), Acc: 0.1865(0.1405)
2021-12-28 09:31:12,945 Epoch[018/310], Step[0550/1251], Loss: 5.3265(5.2563), Acc: 0.1777(0.1404)
2021-12-28 09:32:16,871 Epoch[018/310], Step[0600/1251], Loss: 5.3082(5.2528), Acc: 0.1816(0.1404)
2021-12-28 09:33:19,405 Epoch[018/310], Step[0650/1251], Loss: 5.0805(5.2533), Acc: 0.1719(0.1409)
2021-12-28 09:34:21,748 Epoch[018/310], Step[0700/1251], Loss: 4.9975(5.2505), Acc: 0.1348(0.1418)
2021-12-28 09:35:23,773 Epoch[018/310], Step[0750/1251], Loss: 5.0408(5.2520), Acc: 0.1348(0.1416)
2021-12-28 09:36:27,201 Epoch[018/310], Step[0800/1251], Loss: 5.2822(5.2491), Acc: 0.2148(0.1418)
2021-12-28 09:37:28,845 Epoch[018/310], Step[0850/1251], Loss: 5.5053(5.2482), Acc: 0.1182(0.1421)
2021-12-28 09:38:32,897 Epoch[018/310], Step[0900/1251], Loss: 5.2855(5.2458), Acc: 0.0938(0.1414)
2021-12-28 09:39:34,959 Epoch[018/310], Step[0950/1251], Loss: 5.2137(5.2512), Acc: 0.1016(0.1405)
2021-12-28 09:40:37,765 Epoch[018/310], Step[1000/1251], Loss: 5.5302(5.2498), Acc: 0.0547(0.1408)
2021-12-28 09:41:40,714 Epoch[018/310], Step[1050/1251], Loss: 4.7811(5.2490), Acc: 0.2754(0.1406)
2021-12-28 09:42:42,844 Epoch[018/310], Step[1100/1251], Loss: 5.4148(5.2504), Acc: 0.1240(0.1408)
2021-12-28 09:43:45,742 Epoch[018/310], Step[1150/1251], Loss: 5.6055(5.2468), Acc: 0.0840(0.1409)
2021-12-28 09:44:48,335 Epoch[018/310], Step[1200/1251], Loss: 5.2502(5.2466), Acc: 0.1445(0.1412)
2021-12-28 09:45:50,596 Epoch[018/310], Step[1250/1251], Loss: 5.2813(5.2428), Acc: 0.1436(0.1418)
2021-12-28 09:45:52,578 ----- Validation after Epoch: 18
2021-12-28 09:46:48,507 Val Step[0000/1563], Loss: 1.5268 (1.5268), Acc@1: 0.7812 (0.7812), Acc@5: 0.8750 (0.8750)
2021-12-28 09:46:50,091 Val Step[0050/1563], Loss: 4.2408 (2.0612), Acc@1: 0.1875 (0.5790), Acc@5: 0.3750 (0.8107)
2021-12-28 09:46:51,676 Val Step[0100/1563], Loss: 2.8896 (2.5287), Acc@1: 0.4375 (0.4749), Acc@5: 0.7500 (0.7271)
2021-12-28 09:46:53,201 Val Step[0150/1563], Loss: 1.5638 (2.3973), Acc@1: 0.7500 (0.5052), Acc@5: 0.8438 (0.7448)
2021-12-28 09:46:54,809 Val Step[0200/1563], Loss: 2.2995 (2.4465), Acc@1: 0.3438 (0.4961), Acc@5: 0.8438 (0.7387)
2021-12-28 09:46:56,392 Val Step[0250/1563], Loss: 2.4219 (2.3564), Acc@1: 0.5000 (0.5105), Acc@5: 0.7188 (0.7560)
2021-12-28 09:46:57,887 Val Step[0300/1563], Loss: 2.5862 (2.4503), Acc@1: 0.4375 (0.4816), Acc@5: 0.6875 (0.7401)
2021-12-28 09:46:59,396 Val Step[0350/1563], Loss: 3.0197 (2.4731), Acc@1: 0.1562 (0.4718), Acc@5: 0.6562 (0.7370)
2021-12-28 09:47:00,944 Val Step[0400/1563], Loss: 2.2258 (2.4501), Acc@1: 0.5625 (0.4666), Acc@5: 0.8438 (0.7447)
2021-12-28 09:47:02,554 Val Step[0450/1563], Loss: 1.9279 (2.4384), Acc@1: 0.2500 (0.4655), Acc@5: 0.9062 (0.7497)
2021-12-28 09:47:04,140 Val Step[0500/1563], Loss: 1.7482 (2.4373), Acc@1: 0.6250 (0.4674), Acc@5: 0.8750 (0.7508)
2021-12-28 09:47:05,675 Val Step[0550/1563], Loss: 2.3747 (2.4036), Acc@1: 0.5312 (0.4771), Acc@5: 0.7812 (0.7561)
2021-12-28 09:47:07,297 Val Step[0600/1563], Loss: 1.6098 (2.3985), Acc@1: 0.7188 (0.4788), Acc@5: 0.8750 (0.7572)
2021-12-28 09:47:08,823 Val Step[0650/1563], Loss: 2.0081 (2.4212), Acc@1: 0.5938 (0.4764), Acc@5: 0.9375 (0.7530)
2021-12-28 09:47:10,260 Val Step[0700/1563], Loss: 4.1146 (2.4771), Acc@1: 0.2500 (0.4682), Acc@5: 0.4062 (0.7433)
2021-12-28 09:47:11,688 Val Step[0750/1563], Loss: 2.7859 (2.5315), Acc@1: 0.4062 (0.4595), Acc@5: 0.6562 (0.7325)
2021-12-28 09:47:13,111 Val Step[0800/1563], Loss: 3.2664 (2.5859), Acc@1: 0.4375 (0.4516), Acc@5: 0.6875 (0.7234)
2021-12-28 09:47:14,553 Val Step[0850/1563], Loss: 3.1566 (2.6324), Acc@1: 0.3438 (0.4449), Acc@5: 0.5312 (0.7149)
2021-12-28 09:47:16,035 Val Step[0900/1563], Loss: 1.6596 (2.6443), Acc@1: 0.7188 (0.4453), Acc@5: 0.8125 (0.7124)
2021-12-28 09:47:17,584 Val Step[0950/1563], Loss: 2.2636 (2.6785), Acc@1: 0.6562 (0.4404), Acc@5: 0.8125 (0.7059)
2021-12-28 09:47:18,886 Val Step[1000/1563], Loss: 1.2473 (2.7094), Acc@1: 0.7812 (0.4353), Acc@5: 0.9375 (0.7003)
2021-12-28 09:47:20,190 Val Step[1050/1563], Loss: 1.7723 (2.7280), Acc@1: 0.6562 (0.4324), Acc@5: 0.9062 (0.6973)
2021-12-28 09:47:21,630 Val Step[1100/1563], Loss: 3.1174 (2.7546), Acc@1: 0.5000 (0.4286), Acc@5: 0.6562 (0.6910)
2021-12-28 09:47:23,028 Val Step[1150/1563], Loss: 2.8792 (2.7847), Acc@1: 0.4375 (0.4236), Acc@5: 0.7188 (0.6854)
2021-12-28 09:47:24,398 Val Step[1200/1563], Loss: 1.9476 (2.8101), Acc@1: 0.6250 (0.4201), Acc@5: 0.7812 (0.6808)
2021-12-28 09:47:25,745 Val Step[1250/1563], Loss: 1.9234 (2.8368), Acc@1: 0.6875 (0.4170), Acc@5: 0.8438 (0.6759)
2021-12-28 09:47:27,059 Val Step[1300/1563], Loss: 2.4788 (2.8524), Acc@1: 0.5312 (0.4139), Acc@5: 0.7500 (0.6730)
2021-12-28 09:47:28,411 Val Step[1350/1563], Loss: 3.6395 (2.8794), Acc@1: 0.0312 (0.4084), Acc@5: 0.5000 (0.6675)
2021-12-28 09:47:29,737 Val Step[1400/1563], Loss: 2.6695 (2.8901), Acc@1: 0.4375 (0.4060), Acc@5: 0.7812 (0.6657)
2021-12-28 09:47:31,130 Val Step[1450/1563], Loss: 3.0831 (2.8941), Acc@1: 0.4688 (0.4055), Acc@5: 0.6875 (0.6652)
2021-12-28 09:47:32,701 Val Step[1500/1563], Loss: 2.7246 (2.8753), Acc@1: 0.3750 (0.4090), Acc@5: 0.7812 (0.6693)
2021-12-28 09:47:34,467 Val Step[1550/1563], Loss: 1.0971 (2.8621), Acc@1: 0.8438 (0.4116), Acc@5: 0.8750 (0.6713)
2021-12-28 09:47:35,530 ----- Epoch[018/310], Validation Loss: 2.8580, Validation Acc@1: 0.4127, Validation Acc@5: 0.6718, time: 102.95
2021-12-28 09:47:35,530 ----- Epoch[018/310], Train Loss: 5.2428, Train Acc: 0.1418, time: 1650.61, Best Val(epoch18) Acc@1: 0.4127
2021-12-28 09:47:35,745 Max accuracy so far: 0.4127 at epoch_18
2021-12-28 09:47:35,745 ----- Save BEST model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdparams
2021-12-28 09:47:35,745 ----- Save BEST optim: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT.pdopt
2021-12-28 09:47:35,822 ----- Save BEST ema model: /root/paddlejob/workspace/output/train-20211228-01-15-41/Best_PiT-EMA.pdparams
2021-12-28 09:47:35,823 Now training epoch 19. LR=0.000950
2021-12-28 09:48:55,586 Epoch[019/310], Step[0000/1251], Loss: 5.4525(5.4525), Acc: 0.0693(0.0693)
2021-12-28 09:49:57,286 Epoch[019/310], Step[0050/1251], Loss: 4.8555(5.1774), Acc: 0.1240(0.1520)
2021-12-28 09:51:00,054 Epoch[019/310], Step[0100/1251], Loss: 4.3938(5.1803), Acc: 0.2139(0.1543)
2021-12-28 09:52:01,756 Epoch[019/310], Step[0150/1251], Loss: 5.0570(5.1707), Acc: 0.1221(0.1529)
2021-12-28 09:53:03,897 Epoch[019/310], Step[0200/1251], Loss: 5.4110(5.1724), Acc: 0.1582(0.1545)
2021-12-28 09:54:07,447 Epoch[019/310], Step[0250/1251], Loss: 5.0815(5.1820), Acc: 0.1348(0.1523)
2021-12-28 09:55:10,283 Epoch[019/310], Step[0300/1251], Loss: 5.2371(5.1877), Acc: 0.1387(0.1518)
2021-12-28 09:56:12,677 Epoch[019/310], Step[0350/1251], Loss: 4.8365(5.1805), Acc: 0.1084(0.1527)
2021-12-28 09:57:14,710 Epoch[019/310], Step[0400/1251], Loss: 5.3526(5.1839), Acc: 0.1631(0.1516)
2021-12-28 09:58:17,181 Epoch[019/310], Step[0450/1251], Loss: 4.9122(5.1809), Acc: 0.1602(0.1513)
2021-12-28 09:59:19,088 Epoch[019/310], Step[0500/1251], Loss: 5.1359(5.1841), Acc: 0.1982(0.1524)
2021-12-28 10:00:21,400 Epoch[019/310], Step[0550/1251], Loss: 5.7699(5.1838), Acc: 0.1025(0.1526)
2021-12-28 10:01:25,003 Epoch[019/310], Step[0600/1251], Loss: 5.6999(5.1817), Acc: 0.1201(0.1524)
2021-12-28 10:02:27,508 Epoch[019/310], Step[0650/1251], Loss: 4.8749(5.1806), Acc: 0.2305(0.1524)
2021-12-28 10:03:29,971 Epoch[019/310], Step[0700/1251], Loss: 4.9272(5.1794), Acc: 0.1201(0.1525)
2021-12-28 10:04:32,403 Epoch[019/310], Step[0750/1251], Loss: 5.2556(5.1810), Acc: 0.0850(0.1526)
2021-12-28 10:05:36,170 Epoch[019/310], Step[0800/1251], Loss: 4.9095(5.1815), Acc: 0.0117(0.1518)
2021-12-28 10:06:39,800 Epoch[019/310], Step[0850/1251], Loss: 5.6377(5.1778), Acc: 0.1084(0.1522)
2021-12-28 10:07:42,898 Epoch[019/310], Step[0900/1251], Loss: 5.3750(5.1716), Acc: 0.0752(0.1520)
2021-12-28 10:08:46,182 Epoch[019/310], Step[0950/1251], Loss: 5.7957(5.1727), Acc: 0.1162(0.1519)
2021-12-28 10:09:49,807 Epoch[019/310], Step[1000/1251], Loss: 5.1648(5.1706), Acc: 0.1660(0.1517)
2021-12-28 10:10:52,149 Epoch[019/310], Step[1050/1251], Loss: 5.3762(5.1712), Acc: 0.1719(0.1520)
2021-12-28 10:11:55,669 Epoch[019/310], Step[1100/1251], Loss: 5.2752(5.1729), Acc: 0.2012(0.1518)
2021-12-28 10:12:58,161 Epoch[019/310], Step[1150/1251], Loss: 5.3150(5.1707), Acc: 0.1699(0.1520)
2021-12-28 10:13:59,399 Epoch[019/310], Step[1200/1251], Loss: 4.9217(5.1692), Acc: 0.2256(0.1523)
2021-12-28 10:15:01,578 Epoch[019/310], Step[1250/1251], Loss: 5.5460(5.1685), Acc: 0.1377(0.1524)
2021-12-28 10:15:03,488 ----- Epoch[019/310], Train Loss: 5.1685, Train Acc: 0.1524, time: 1647.66, Best Val(epoch18) Acc@1: 0.4127
2021-12-28 10:15:03,488 Now training epoch 20. LR=0.000989
2021-12-28 10:16:28,014 Epoch[020/310], Step[0000/1251], Loss: 4.8232(4.8232), Acc: 0.2012(0.2012)
2021-12-28 10:17:30,327 Epoch[020/310], Step[0050/1251], Loss: 5.0354(5.0938), Acc: 0.1660(0.1621)
2021-12-28 10:18:32,575 Epoch[020/310], Step[0100/1251], Loss: 5.2030(5.1546), Acc: 0.2100(0.1557)
2021-12-28 10:19:34,026 Epoch[020/310], Step[0150/1251], Loss: 4.8789(5.1483), Acc: 0.2383(0.1549)
2021-12-28 10:20:36,029 Epoch[020/310], Step[0200/1251], Loss: 4.9492(5.1470), Acc: 0.2373(0.1572)
2021-12-28 10:21:40,079 Epoch[020/310], Step[0250/1251], Loss: 4.7567(5.1410), Acc: 0.2705(0.1601)
2021-12-28 10:22:43,064 Epoch[020/310], Step[0300/1251], Loss: 4.8628(5.1320), Acc: 0.1816(0.1605)
2021-12-28 10:23:45,330 Epoch[020/310], Step[0350/1251], Loss: 5.2628(5.1292), Acc: 0.1602(0.1593)
2021-12-28 10:24:47,888 Epoch[020/310], Step[0400/1251], Loss: 4.6515(5.1365), Acc: 0.1914(0.1591)
2021-12-28 10:25:51,234 Epoch[020/310], Step[0450/1251], Loss: 5.2496(5.1357), Acc: 0.1240(0.1573)
2021-12-28 10:26:55,277 Epoch[020/310], Step[0500/1251], Loss: 5.3338(5.1404), Acc: 0.1465(0.1577)
2021-12-28 10:27:56,969 Epoch[020/310], Step[0550/1251], Loss: 5.1894(5.1390), Acc: 0.2031(0.1586)
2021-12-28 10:28:59,428 Epoch[020/310], Step[0600/1251], Loss: 5.4165(5.1397), Acc: 0.1455(0.1583)
2021-12-28 10:30:02,959 Epoch[020/310], Step[0650/1251], Loss: 5.0425(5.1327), Acc: 0.0840(0.1584)
2021-12-28 10:31:07,076 Epoch[020/310], Step[0700/1251], Loss: 5.3043(5.1285), Acc: 0.1816(0.1580)
2021-12-28 10:32:09,360 Epoch[020/310], Step[0750/1251], Loss: 4.8467(5.1241), Acc: 0.2334(0.1576)
2021-12-28 10:33:12,873 Epoch[020/310], Step[0800/1251], Loss: 4.7188(5.1243), Acc: 0.0986(0.1576)
2021-12-28 10:34:16,320 Epoch[020/310], Step[0850/1251], Loss: 5.1548(5.1251), Acc: 0.1982(0.1573)
2021-12-28 10:35:19,931 Epoch[020/310], Step[0900/1251], Loss: 5.3781(5.1230), Acc: 0.1641(0.1581)
2021-12-28 10:36:22,164 Epoch[020/310], Step[0950/1251], Loss: 5.7699(5.1221), Acc: 0.1006(0.1576)
2021-12-28 10:37:24,065 Epoch[020/310], Step[1000/1251], Loss: 5.3758(5.1243), Acc: 0.0859(0.1577)
2021-12-28 10:38:27,473 Epoch[020/310], Step[1050/1251], Loss: 5.3547(5.1209), Acc: 0.1182(0.1581)
2021-12-28 10:39:31,061 Epoch[020/310], Step[1100/1251], Loss: 5.3091(5.1192), Acc: 0.1367(0.1584)
2021-12-28 10:40:34,393 Epoch[020/310], Step[1150/1251], Loss: 4.9686(5.1145), Acc: 0.2578(0.1588)
2021-12-28 10:41:38,500 Epoch[020/310], Step[1200/1251], Loss: 5.2136(5.1148), Acc: 0.0820(0.1591)
2021-12-28 10:42:41,263 Epoch[020/310], Step[1250/1251], Loss: 5.2403(5.1144), Acc: 0.1250(0.1590)
2021-12-28 10:42:43,574 ----- Validation after Epoch: 20