-
Notifications
You must be signed in to change notification settings - Fork 4
/
DCMH_MSRVTT_1kA.txt
2470 lines (2470 loc) · 226 KB
/
DCMH_MSRVTT_1kA.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
Experiment directory: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA
Preparing the dataloaders ...
Loading dataset MSRVTT_jsfusion_trainval in ram ...
Finish loading dataset MSRVTT_jsfusion_trainval in ram, taking 546.9244015216827 s.
Loading dataset MSRVTT_jsfusion_test in ram ...
Finish loading dataset MSRVTT_jsfusion_test in ram, taking 69.83318901062012 s.
Loading dataset MSRVTT_jsfusion_test in ram ...
Finish loading dataset MSRVTT_jsfusion_test in ram, taking 60.15059280395508 s.
Training ...
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch0.pth ...
Done in 6.558s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch0.pth ...
Done in 8.222s
epoch : 0
loss : 0
learning_rate : 5e-05
n_samples : 0
n_steps : 0
MSRVTT_jsfusion_test/t2v_metrics/R1: 0.1
MSRVTT_jsfusion_test/t2v_metrics/R5: 0.5
MSRVTT_jsfusion_test/t2v_metrics/R10: 1.0
MSRVTT_jsfusion_test/t2v_metrics/R50: 4.5
MSRVTT_jsfusion_test/t2v_metrics/MedR: 502.5
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 499.848
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 0.3684031498640387
MSRVTT_jsfusion_test/v2t_metrics/R1: 0.1
MSRVTT_jsfusion_test/v2t_metrics/R5: 0.4
MSRVTT_jsfusion_test/v2t_metrics/R10: 1.0
MSRVTT_jsfusion_test/v2t_metrics/R50: 4.6
MSRVTT_jsfusion_test/v2t_metrics/MedR: 503.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 500.8045
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 0.3419951893353394
mnt_best : 0.3684031498640387
not_improved_count: 0
Train Epoch: 1 [1/250 128/32000 (0%)] Loss: 2.68994 (semantic_loss: 0.74034, quant_loss: 1.94922, bit_balance_loss: 0.00038) batch_time=22.97347
Train Epoch: 1 [12/250 1536/32000 (5%)] Loss: 1.99995 (semantic_loss: 0.04833, quant_loss: 1.95117, bit_balance_loss: 0.00045) batch_time=0.34628
Train Epoch: 1 [23/250 2944/32000 (9%)] Loss: 1.99769 (semantic_loss: 0.04604, quant_loss: 1.95117, bit_balance_loss: 0.00047) batch_time=0.33036
Train Epoch: 1 [34/250 4352/32000 (14%)] Loss: 1.99667 (semantic_loss: 0.04599, quant_loss: 1.95020, bit_balance_loss: 0.00048) batch_time=0.32395
Train Epoch: 1 [45/250 5760/32000 (18%)] Loss: 1.99752 (semantic_loss: 0.04586, quant_loss: 1.95117, bit_balance_loss: 0.00048) batch_time=0.32902
Train Epoch: 1 [56/250 7168/32000 (22%)] Loss: 1.99643 (semantic_loss: 0.04575, quant_loss: 1.95020, bit_balance_loss: 0.00048) batch_time=0.35001
Train Epoch: 1 [67/250 8576/32000 (27%)] Loss: 1.99643 (semantic_loss: 0.04575, quant_loss: 1.95020, bit_balance_loss: 0.00048) batch_time=0.32516
Train Epoch: 1 [78/250 9984/32000 (31%)] Loss: 1.99738 (semantic_loss: 0.04573, quant_loss: 1.95117, bit_balance_loss: 0.00048) batch_time=0.33290
Train Epoch: 1 [89/250 11392/32000 (36%)] Loss: 1.99738 (semantic_loss: 0.04573, quant_loss: 1.95117, bit_balance_loss: 0.00048) batch_time=0.32463
Train Epoch: 1 [100/250 12800/32000 (40%)] Loss: 1.99735 (semantic_loss: 0.04570, quant_loss: 1.95117, bit_balance_loss: 0.00048) batch_time=0.34818
Train Epoch: 1 [111/250 14208/32000 (44%)] Loss: 1.99637 (semantic_loss: 0.04570, quant_loss: 1.95020, bit_balance_loss: 0.00047) batch_time=0.36100
Train Epoch: 1 [122/250 15616/32000 (49%)] Loss: 1.99637 (semantic_loss: 0.04571, quant_loss: 1.95020, bit_balance_loss: 0.00047) batch_time=0.32936
Train Epoch: 1 [133/250 17024/32000 (53%)] Loss: 1.99736 (semantic_loss: 0.04573, quant_loss: 1.95117, bit_balance_loss: 0.00046) batch_time=0.35055
Train Epoch: 1 [144/250 18432/32000 (58%)] Loss: 1.99733 (semantic_loss: 0.04570, quant_loss: 1.95117, bit_balance_loss: 0.00046) batch_time=0.32414
Train Epoch: 1 [155/250 19840/32000 (62%)] Loss: 1.99736 (semantic_loss: 0.04573, quant_loss: 1.95117, bit_balance_loss: 0.00046) batch_time=0.34215
Train Epoch: 1 [166/250 21248/32000 (66%)] Loss: 1.99732 (semantic_loss: 0.04569, quant_loss: 1.95117, bit_balance_loss: 0.00045) batch_time=0.35292
Train Epoch: 1 [177/250 22656/32000 (71%)] Loss: 1.99730 (semantic_loss: 0.04568, quant_loss: 1.95117, bit_balance_loss: 0.00045) batch_time=0.33110
Train Epoch: 1 [188/250 24064/32000 (75%)] Loss: 1.99732 (semantic_loss: 0.04571, quant_loss: 1.95117, bit_balance_loss: 0.00045) batch_time=0.39705
Train Epoch: 1 [199/250 25472/32000 (80%)] Loss: 1.99629 (semantic_loss: 0.04565, quant_loss: 1.95020, bit_balance_loss: 0.00044) batch_time=0.32698
Train Epoch: 1 [210/250 26880/32000 (84%)] Loss: 1.99723 (semantic_loss: 0.04562, quant_loss: 1.95117, bit_balance_loss: 0.00044) batch_time=5.77071
Train Epoch: 1 [221/250 28288/32000 (88%)] Loss: 1.99619 (semantic_loss: 0.04557, quant_loss: 1.95020, bit_balance_loss: 0.00043) batch_time=0.34086
Train Epoch: 1 [232/250 29696/32000 (93%)] Loss: 1.99623 (semantic_loss: 0.04561, quant_loss: 1.95020, bit_balance_loss: 0.00043) batch_time=0.34422
Train Epoch: 1 [243/250 31104/32000 (97%)] Loss: 1.99623 (semantic_loss: 0.04561, quant_loss: 1.95020, bit_balance_loss: 0.00042) batch_time=0.32681
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch1.pth ...
Done in 3.870s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch1.pth ...
Done in 7.779s
epoch : 1
loss : 2.002893678188324
learning_rate : 5e-05
n_samples : 32000
n_steps : 250
MSRVTT_jsfusion_test/t2v_metrics/R1: 0.1
MSRVTT_jsfusion_test/t2v_metrics/R5: 0.8
MSRVTT_jsfusion_test/t2v_metrics/R10: 1.6
MSRVTT_jsfusion_test/t2v_metrics/R50: 7.3
MSRVTT_jsfusion_test/t2v_metrics/MedR: 454.25
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 470.219
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 0.5039684199579494
MSRVTT_jsfusion_test/v2t_metrics/R1: 0.0
MSRVTT_jsfusion_test/v2t_metrics/R5: 0.6
MSRVTT_jsfusion_test/v2t_metrics/R10: 1.6
MSRVTT_jsfusion_test/v2t_metrics/R50: 5.6
MSRVTT_jsfusion_test/v2t_metrics/MedR: 454.25
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 476.6665
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 0.0
mnt_best : 0.5039684199579494
not_improved_count: 0
Train Epoch: 2 [1/250 128/32000 (0%)] Loss: 1.99702 (semantic_loss: 0.04543, quant_loss: 1.95117, bit_balance_loss: 0.00042) batch_time=27.67402
Train Epoch: 2 [12/250 1536/32000 (5%)] Loss: 1.99699 (semantic_loss: 0.04541, quant_loss: 1.95117, bit_balance_loss: 0.00041) batch_time=0.33488
Train Epoch: 2 [23/250 2944/32000 (9%)] Loss: 1.99622 (semantic_loss: 0.04561, quant_loss: 1.95020, bit_balance_loss: 0.00042) batch_time=0.33755
Train Epoch: 2 [34/250 4352/32000 (14%)] Loss: 1.99605 (semantic_loss: 0.04545, quant_loss: 1.95020, bit_balance_loss: 0.00041) batch_time=0.35204
Train Epoch: 2 [45/250 5760/32000 (18%)] Loss: 1.99589 (semantic_loss: 0.04529, quant_loss: 1.95020, bit_balance_loss: 0.00040) batch_time=0.33885
Train Epoch: 2 [56/250 7168/32000 (22%)] Loss: 1.99682 (semantic_loss: 0.04526, quant_loss: 1.95117, bit_balance_loss: 0.00039) batch_time=0.34402
Train Epoch: 2 [67/250 8576/32000 (27%)] Loss: 1.99669 (semantic_loss: 0.04512, quant_loss: 1.95117, bit_balance_loss: 0.00039) batch_time=0.34548
Train Epoch: 2 [78/250 9984/32000 (31%)] Loss: 1.99559 (semantic_loss: 0.04501, quant_loss: 1.95020, bit_balance_loss: 0.00039) batch_time=4.30400
Train Epoch: 2 [89/250 11392/32000 (36%)] Loss: 1.99546 (semantic_loss: 0.04488, quant_loss: 1.95020, bit_balance_loss: 0.00038) batch_time=0.33674
Train Epoch: 2 [100/250 12800/32000 (40%)] Loss: 1.99608 (semantic_loss: 0.04551, quant_loss: 1.95020, bit_balance_loss: 0.00038) batch_time=0.33340
Train Epoch: 2 [111/250 14208/32000 (44%)] Loss: 1.99591 (semantic_loss: 0.04534, quant_loss: 1.95020, bit_balance_loss: 0.00038) batch_time=0.32623
Train Epoch: 2 [122/250 15616/32000 (49%)] Loss: 1.99476 (semantic_loss: 0.04419, quant_loss: 1.95020, bit_balance_loss: 0.00037) batch_time=0.32464
Train Epoch: 2 [133/250 17024/32000 (53%)] Loss: 1.99588 (semantic_loss: 0.04434, quant_loss: 1.95117, bit_balance_loss: 0.00036) batch_time=0.33094
Train Epoch: 2 [144/250 18432/32000 (58%)] Loss: 1.99450 (semantic_loss: 0.04395, quant_loss: 1.95020, bit_balance_loss: 0.00036) batch_time=1.05294
Train Epoch: 2 [155/250 19840/32000 (62%)] Loss: 1.99451 (semantic_loss: 0.04395, quant_loss: 1.95020, bit_balance_loss: 0.00036) batch_time=0.32559
Train Epoch: 2 [166/250 21248/32000 (66%)] Loss: 1.99548 (semantic_loss: 0.04395, quant_loss: 1.95117, bit_balance_loss: 0.00035) batch_time=0.32195
Train Epoch: 2 [177/250 22656/32000 (71%)] Loss: 1.99576 (semantic_loss: 0.04424, quant_loss: 1.95117, bit_balance_loss: 0.00035) batch_time=0.34834
Train Epoch: 2 [188/250 24064/32000 (75%)] Loss: 1.99493 (semantic_loss: 0.04439, quant_loss: 1.95020, bit_balance_loss: 0.00035) batch_time=0.35347
Train Epoch: 2 [199/250 25472/32000 (80%)] Loss: 1.99388 (semantic_loss: 0.04334, quant_loss: 1.95020, bit_balance_loss: 0.00034) batch_time=0.33256
Train Epoch: 2 [210/250 26880/32000 (84%)] Loss: 1.99487 (semantic_loss: 0.04336, quant_loss: 1.95117, bit_balance_loss: 0.00034) batch_time=0.34094
Train Epoch: 2 [221/250 28288/32000 (88%)] Loss: 1.99385 (semantic_loss: 0.04332, quant_loss: 1.95020, bit_balance_loss: 0.00034) batch_time=0.32201
Train Epoch: 2 [232/250 29696/32000 (93%)] Loss: 1.99467 (semantic_loss: 0.04316, quant_loss: 1.95117, bit_balance_loss: 0.00033) batch_time=0.33833
Train Epoch: 2 [243/250 31104/32000 (97%)] Loss: 1.99425 (semantic_loss: 0.04275, quant_loss: 1.95117, bit_balance_loss: 0.00033) batch_time=0.32278
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch2.pth ...
Done in 4.159s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch2.pth ...
Done in 7.883s
removing stale ckpt [epoch 1] [took 0.00s]
removing stale ckpt [epoch 0] [took 0.00s]
epoch : 2
loss : 1.9955712633132934
learning_rate : 4.75e-05
n_samples : 64000
n_steps : 500
MSRVTT_jsfusion_test/t2v_metrics/R1: 0.2
MSRVTT_jsfusion_test/t2v_metrics/R5: 1.7
MSRVTT_jsfusion_test/t2v_metrics/R10: 2.8
MSRVTT_jsfusion_test/t2v_metrics/R50: 13.8
MSRVTT_jsfusion_test/t2v_metrics/MedR: 256.5
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 290.412
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 0.9837369468917456
MSRVTT_jsfusion_test/v2t_metrics/R1: 0.1
MSRVTT_jsfusion_test/v2t_metrics/R5: 1.0
MSRVTT_jsfusion_test/v2t_metrics/R10: 2.5
MSRVTT_jsfusion_test/v2t_metrics/R50: 11.8
MSRVTT_jsfusion_test/v2t_metrics/MedR: 238.5
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 283.8205
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 0.6299605249474366
mnt_best : 0.9837369468917456
not_improved_count: 0
Train Epoch: 3 [1/250 128/32000 (0%)] Loss: 1.99304 (semantic_loss: 0.04252, quant_loss: 1.95020, bit_balance_loss: 0.00033) batch_time=24.52264
Train Epoch: 3 [12/250 1536/32000 (5%)] Loss: 1.99399 (semantic_loss: 0.04250, quant_loss: 1.95117, bit_balance_loss: 0.00032) batch_time=2.49927
Train Epoch: 3 [23/250 2944/32000 (9%)] Loss: 1.99287 (semantic_loss: 0.04235, quant_loss: 1.95020, bit_balance_loss: 0.00032) batch_time=0.34087
Train Epoch: 3 [34/250 4352/32000 (14%)] Loss: 1.99219 (semantic_loss: 0.04169, quant_loss: 1.95020, bit_balance_loss: 0.00031) batch_time=0.34155
Train Epoch: 3 [45/250 5760/32000 (18%)] Loss: 1.99159 (semantic_loss: 0.04206, quant_loss: 1.94922, bit_balance_loss: 0.00031) batch_time=0.34564
Train Epoch: 3 [56/250 7168/32000 (22%)] Loss: 1.99173 (semantic_loss: 0.04123, quant_loss: 1.95020, bit_balance_loss: 0.00031) batch_time=0.34210
Train Epoch: 3 [67/250 8576/32000 (27%)] Loss: 1.99086 (semantic_loss: 0.04133, quant_loss: 1.94922, bit_balance_loss: 0.00031) batch_time=0.34120
Train Epoch: 3 [78/250 9984/32000 (31%)] Loss: 1.99252 (semantic_loss: 0.04105, quant_loss: 1.95117, bit_balance_loss: 0.00030) batch_time=0.36533
Train Epoch: 3 [89/250 11392/32000 (36%)] Loss: 1.99131 (semantic_loss: 0.04081, quant_loss: 1.95020, bit_balance_loss: 0.00030) batch_time=0.34230
Train Epoch: 3 [100/250 12800/32000 (40%)] Loss: 1.99120 (semantic_loss: 0.04071, quant_loss: 1.95020, bit_balance_loss: 0.00030) batch_time=0.33869
Train Epoch: 3 [111/250 14208/32000 (44%)] Loss: 1.99121 (semantic_loss: 0.03975, quant_loss: 1.95117, bit_balance_loss: 0.00029) batch_time=0.33668
Train Epoch: 3 [122/250 15616/32000 (49%)] Loss: 1.98890 (semantic_loss: 0.03939, quant_loss: 1.94922, bit_balance_loss: 0.00029) batch_time=0.32599
Train Epoch: 3 [133/250 17024/32000 (53%)] Loss: 1.98969 (semantic_loss: 0.03921, quant_loss: 1.95020, bit_balance_loss: 0.00029) batch_time=0.34284
Train Epoch: 3 [144/250 18432/32000 (58%)] Loss: 1.99016 (semantic_loss: 0.03968, quant_loss: 1.95020, bit_balance_loss: 0.00028) batch_time=0.33150
Train Epoch: 3 [155/250 19840/32000 (62%)] Loss: 1.98819 (semantic_loss: 0.03869, quant_loss: 1.94922, bit_balance_loss: 0.00028) batch_time=0.34062
Train Epoch: 3 [166/250 21248/32000 (66%)] Loss: 1.98764 (semantic_loss: 0.03717, quant_loss: 1.95020, bit_balance_loss: 0.00027) batch_time=0.32692
Train Epoch: 3 [177/250 22656/32000 (71%)] Loss: 1.98730 (semantic_loss: 0.03781, quant_loss: 1.94922, bit_balance_loss: 0.00027) batch_time=0.33965
Train Epoch: 3 [188/250 24064/32000 (75%)] Loss: 1.98535 (semantic_loss: 0.03683, quant_loss: 1.94824, bit_balance_loss: 0.00027) batch_time=0.35133
Train Epoch: 3 [199/250 25472/32000 (80%)] Loss: 1.98684 (semantic_loss: 0.03638, quant_loss: 1.95020, bit_balance_loss: 0.00027) batch_time=0.32899
Train Epoch: 3 [210/250 26880/32000 (84%)] Loss: 1.98708 (semantic_loss: 0.03662, quant_loss: 1.95020, bit_balance_loss: 0.00026) batch_time=0.32451
Train Epoch: 3 [221/250 28288/32000 (88%)] Loss: 1.98647 (semantic_loss: 0.03698, quant_loss: 1.94922, bit_balance_loss: 0.00026) batch_time=2.12151
Train Epoch: 3 [232/250 29696/32000 (93%)] Loss: 1.98596 (semantic_loss: 0.03648, quant_loss: 1.94922, bit_balance_loss: 0.00026) batch_time=0.33921
Train Epoch: 3 [243/250 31104/32000 (97%)] Loss: 1.98399 (semantic_loss: 0.03451, quant_loss: 1.94922, bit_balance_loss: 0.00026) batch_time=0.33004
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch3.pth ...
Done in 4.098s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch3.pth ...
Done in 8.047s
removing stale ckpt [epoch 2] [took 0.00s]
epoch : 3
loss : 1.9893670625686646
learning_rate : 4.5125e-05
n_samples : 96000
n_steps : 750
MSRVTT_jsfusion_test/t2v_metrics/R1: 2.0
MSRVTT_jsfusion_test/t2v_metrics/R5: 8.7
MSRVTT_jsfusion_test/t2v_metrics/R10: 15.0
MSRVTT_jsfusion_test/t2v_metrics/R50: 48.8
MSRVTT_jsfusion_test/t2v_metrics/MedR: 52.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 112.439
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 6.3906765283993066
MSRVTT_jsfusion_test/v2t_metrics/R1: 2.8
MSRVTT_jsfusion_test/v2t_metrics/R5: 10.1
MSRVTT_jsfusion_test/v2t_metrics/R10: 17.0
MSRVTT_jsfusion_test/v2t_metrics/R50: 51.4
MSRVTT_jsfusion_test/v2t_metrics/MedR: 48.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 105.4345
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 7.833865463574119
mnt_best : 6.3906765283993066
not_improved_count: 0
Train Epoch: 4 [1/250 128/32000 (0%)] Loss: 1.98367 (semantic_loss: 0.03420, quant_loss: 1.94922, bit_balance_loss: 0.00025) batch_time=25.84425
Train Epoch: 4 [12/250 1536/32000 (5%)] Loss: 1.98639 (semantic_loss: 0.03594, quant_loss: 1.95020, bit_balance_loss: 0.00026) batch_time=0.67224
Train Epoch: 4 [23/250 2944/32000 (9%)] Loss: 1.98547 (semantic_loss: 0.03502, quant_loss: 1.95020, bit_balance_loss: 0.00025) batch_time=1.18812
Train Epoch: 4 [34/250 4352/32000 (14%)] Loss: 1.98436 (semantic_loss: 0.03490, quant_loss: 1.94922, bit_balance_loss: 0.00025) batch_time=0.33996
Train Epoch: 4 [45/250 5760/32000 (18%)] Loss: 1.98309 (semantic_loss: 0.03363, quant_loss: 1.94922, bit_balance_loss: 0.00024) batch_time=0.33626
Train Epoch: 4 [56/250 7168/32000 (22%)] Loss: 1.98308 (semantic_loss: 0.03362, quant_loss: 1.94922, bit_balance_loss: 0.00024) batch_time=0.34410
Train Epoch: 4 [67/250 8576/32000 (27%)] Loss: 1.98359 (semantic_loss: 0.03413, quant_loss: 1.94922, bit_balance_loss: 0.00024) batch_time=2.96249
Train Epoch: 4 [78/250 9984/32000 (31%)] Loss: 1.98372 (semantic_loss: 0.03328, quant_loss: 1.95020, bit_balance_loss: 0.00024) batch_time=0.34048
Train Epoch: 4 [89/250 11392/32000 (36%)] Loss: 1.98223 (semantic_loss: 0.03278, quant_loss: 1.94922, bit_balance_loss: 0.00024) batch_time=0.32326
Train Epoch: 4 [100/250 12800/32000 (40%)] Loss: 1.98431 (semantic_loss: 0.03388, quant_loss: 1.95020, bit_balance_loss: 0.00024) batch_time=0.33253
Train Epoch: 4 [111/250 14208/32000 (44%)] Loss: 1.98296 (semantic_loss: 0.03253, quant_loss: 1.95020, bit_balance_loss: 0.00023) batch_time=1.49998
Train Epoch: 4 [122/250 15616/32000 (49%)] Loss: 1.98356 (semantic_loss: 0.03411, quant_loss: 1.94922, bit_balance_loss: 0.00023) batch_time=0.57107
Train Epoch: 4 [133/250 17024/32000 (53%)] Loss: 1.98433 (semantic_loss: 0.03391, quant_loss: 1.95020, bit_balance_loss: 0.00023) batch_time=0.35517
Train Epoch: 4 [144/250 18432/32000 (58%)] Loss: 1.98237 (semantic_loss: 0.03292, quant_loss: 1.94922, bit_balance_loss: 0.00023) batch_time=1.37063
Train Epoch: 4 [155/250 19840/32000 (62%)] Loss: 1.98449 (semantic_loss: 0.03407, quant_loss: 1.95020, bit_balance_loss: 0.00023) batch_time=0.32443
Train Epoch: 4 [166/250 21248/32000 (66%)] Loss: 1.98375 (semantic_loss: 0.03333, quant_loss: 1.95020, bit_balance_loss: 0.00023) batch_time=0.33849
Train Epoch: 4 [177/250 22656/32000 (71%)] Loss: 1.98120 (semantic_loss: 0.03176, quant_loss: 1.94922, bit_balance_loss: 0.00022) batch_time=0.33056
Train Epoch: 4 [188/250 24064/32000 (75%)] Loss: 1.98336 (semantic_loss: 0.03294, quant_loss: 1.95020, bit_balance_loss: 0.00022) batch_time=0.35123
Train Epoch: 4 [199/250 25472/32000 (80%)] Loss: 1.98178 (semantic_loss: 0.03234, quant_loss: 1.94922, bit_balance_loss: 0.00022) batch_time=0.38238
Train Epoch: 4 [210/250 26880/32000 (84%)] Loss: 1.98297 (semantic_loss: 0.03256, quant_loss: 1.95020, bit_balance_loss: 0.00022) batch_time=0.34366
Train Epoch: 4 [221/250 28288/32000 (88%)] Loss: 1.98205 (semantic_loss: 0.03164, quant_loss: 1.95020, bit_balance_loss: 0.00022) batch_time=0.34690
Train Epoch: 4 [232/250 29696/32000 (93%)] Loss: 1.98047 (semantic_loss: 0.03007, quant_loss: 1.95020, bit_balance_loss: 0.00021) batch_time=0.35810
Train Epoch: 4 [243/250 31104/32000 (97%)] Loss: 1.98109 (semantic_loss: 0.03166, quant_loss: 1.94922, bit_balance_loss: 0.00021) batch_time=0.34256
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch4.pth ...
Done in 4.208s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch4.pth ...
Done in 21.501s
removing stale ckpt [epoch 3] [took 0.01s]
epoch : 4
loss : 1.9828839597702026
learning_rate : 4.2868749999999995e-05
n_samples : 128000
n_steps : 1000
MSRVTT_jsfusion_test/t2v_metrics/R1: 4.1
MSRVTT_jsfusion_test/t2v_metrics/R5: 15.6
MSRVTT_jsfusion_test/t2v_metrics/R10: 25.9
MSRVTT_jsfusion_test/t2v_metrics/R50: 63.7
MSRVTT_jsfusion_test/t2v_metrics/MedR: 31.25
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 74.5245
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 11.832306375644585
MSRVTT_jsfusion_test/v2t_metrics/R1: 3.1
MSRVTT_jsfusion_test/v2t_metrics/R5: 14.1
MSRVTT_jsfusion_test/v2t_metrics/R10: 26.0
MSRVTT_jsfusion_test/v2t_metrics/R50: 64.8
MSRVTT_jsfusion_test/v2t_metrics/MedR: 31.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 71.724
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 10.435615075493853
mnt_best : 11.832306375644585
not_improved_count: 0
Train Epoch: 5 [1/250 128/32000 (0%)] Loss: 1.98151 (semantic_loss: 0.03208, quant_loss: 1.94922, bit_balance_loss: 0.00021) batch_time=30.94472
Train Epoch: 5 [12/250 1536/32000 (5%)] Loss: 1.97998 (semantic_loss: 0.03056, quant_loss: 1.94922, bit_balance_loss: 0.00020) batch_time=0.35294
Train Epoch: 5 [23/250 2944/32000 (9%)] Loss: 1.98053 (semantic_loss: 0.03110, quant_loss: 1.94922, bit_balance_loss: 0.00021) batch_time=0.34209
Train Epoch: 5 [34/250 4352/32000 (14%)] Loss: 1.97996 (semantic_loss: 0.03054, quant_loss: 1.94922, bit_balance_loss: 0.00020) batch_time=0.33087
Train Epoch: 5 [45/250 5760/32000 (18%)] Loss: 1.97876 (semantic_loss: 0.02934, quant_loss: 1.94922, bit_balance_loss: 0.00020) batch_time=0.37877
Train Epoch: 5 [56/250 7168/32000 (22%)] Loss: 1.98018 (semantic_loss: 0.03076, quant_loss: 1.94922, bit_balance_loss: 0.00020) batch_time=0.32354
Train Epoch: 5 [67/250 8576/32000 (27%)] Loss: 1.98065 (semantic_loss: 0.03123, quant_loss: 1.94922, bit_balance_loss: 0.00020) batch_time=0.32341
Train Epoch: 5 [78/250 9984/32000 (31%)] Loss: 1.98090 (semantic_loss: 0.03051, quant_loss: 1.95020, bit_balance_loss: 0.00020) batch_time=0.34839
Train Epoch: 5 [89/250 11392/32000 (36%)] Loss: 1.97900 (semantic_loss: 0.02861, quant_loss: 1.95020, bit_balance_loss: 0.00019) batch_time=0.33749
Train Epoch: 5 [100/250 12800/32000 (40%)] Loss: 1.98030 (semantic_loss: 0.03089, quant_loss: 1.94922, bit_balance_loss: 0.00019) batch_time=0.33259
Train Epoch: 5 [111/250 14208/32000 (44%)] Loss: 1.97942 (semantic_loss: 0.02904, quant_loss: 1.95020, bit_balance_loss: 0.00019) batch_time=0.34503
Train Epoch: 5 [122/250 15616/32000 (49%)] Loss: 1.97979 (semantic_loss: 0.02941, quant_loss: 1.95020, bit_balance_loss: 0.00019) batch_time=0.33206
Train Epoch: 5 [133/250 17024/32000 (53%)] Loss: 1.98059 (semantic_loss: 0.03020, quant_loss: 1.95020, bit_balance_loss: 0.00019) batch_time=0.34651
Train Epoch: 5 [144/250 18432/32000 (58%)] Loss: 1.97912 (semantic_loss: 0.02874, quant_loss: 1.95020, bit_balance_loss: 0.00019) batch_time=0.35917
Train Epoch: 5 [155/250 19840/32000 (62%)] Loss: 1.97847 (semantic_loss: 0.02809, quant_loss: 1.95020, bit_balance_loss: 0.00019) batch_time=0.33773
Train Epoch: 5 [166/250 21248/32000 (66%)] Loss: 1.97972 (semantic_loss: 0.03031, quant_loss: 1.94922, bit_balance_loss: 0.00019) batch_time=0.35197
Train Epoch: 5 [177/250 22656/32000 (71%)] Loss: 1.97900 (semantic_loss: 0.02863, quant_loss: 1.95020, bit_balance_loss: 0.00018) batch_time=0.34012
Train Epoch: 5 [188/250 24064/32000 (75%)] Loss: 1.97848 (semantic_loss: 0.02908, quant_loss: 1.94922, bit_balance_loss: 0.00018) batch_time=0.32895
Train Epoch: 5 [199/250 25472/32000 (80%)] Loss: 1.97903 (semantic_loss: 0.02963, quant_loss: 1.94922, bit_balance_loss: 0.00018) batch_time=0.33839
Train Epoch: 5 [210/250 26880/32000 (84%)] Loss: 1.97972 (semantic_loss: 0.02935, quant_loss: 1.95020, bit_balance_loss: 0.00018) batch_time=0.36511
Train Epoch: 5 [221/250 28288/32000 (88%)] Loss: 1.97850 (semantic_loss: 0.02911, quant_loss: 1.94922, bit_balance_loss: 0.00018) batch_time=0.33600
Train Epoch: 5 [232/250 29696/32000 (93%)] Loss: 1.97782 (semantic_loss: 0.02843, quant_loss: 1.94922, bit_balance_loss: 0.00018) batch_time=0.34379
Train Epoch: 5 [243/250 31104/32000 (97%)] Loss: 1.97941 (semantic_loss: 0.03001, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.38014
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch5.pth ...
Done in 12.206s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch5.pth ...
Done in 16.068s
removing stale ckpt [epoch 4] [took 0.00s]
epoch : 5
loss : 1.9797719712257384
learning_rate : 4.072531249999999e-05
n_samples : 160000
n_steps : 1250
MSRVTT_jsfusion_test/t2v_metrics/R1: 6.8
MSRVTT_jsfusion_test/t2v_metrics/R5: 22.1
MSRVTT_jsfusion_test/t2v_metrics/R10: 32.9
MSRVTT_jsfusion_test/t2v_metrics/R50: 69.3
MSRVTT_jsfusion_test/t2v_metrics/MedR: 21.5
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 61.515
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 17.035924032637332
MSRVTT_jsfusion_test/v2t_metrics/R1: 4.9
MSRVTT_jsfusion_test/v2t_metrics/R5: 20.3
MSRVTT_jsfusion_test/v2t_metrics/R10: 32.5
MSRVTT_jsfusion_test/v2t_metrics/R50: 70.5
MSRVTT_jsfusion_test/v2t_metrics/MedR: 20.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 60.411
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 14.78626525876031
mnt_best : 17.035924032637332
not_improved_count: 0
Train Epoch: 6 [1/250 128/32000 (0%)] Loss: 1.97860 (semantic_loss: 0.02823, quant_loss: 1.95020, bit_balance_loss: 0.00018) batch_time=25.66846
Train Epoch: 6 [12/250 1536/32000 (5%)] Loss: 1.97635 (semantic_loss: 0.02696, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.32844
Train Epoch: 6 [23/250 2944/32000 (9%)] Loss: 1.97772 (semantic_loss: 0.02832, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.33309
Train Epoch: 6 [34/250 4352/32000 (14%)] Loss: 1.98121 (semantic_loss: 0.02986, quant_loss: 1.95117, bit_balance_loss: 0.00017) batch_time=0.34432
Train Epoch: 6 [45/250 5760/32000 (18%)] Loss: 1.97765 (semantic_loss: 0.02825, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.33056
Train Epoch: 6 [56/250 7168/32000 (22%)] Loss: 1.97847 (semantic_loss: 0.02908, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.35378
Train Epoch: 6 [67/250 8576/32000 (27%)] Loss: 1.97796 (semantic_loss: 0.02760, quant_loss: 1.95020, bit_balance_loss: 0.00017) batch_time=0.35475
Train Epoch: 6 [78/250 9984/32000 (31%)] Loss: 1.97894 (semantic_loss: 0.02955, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.32255
Train Epoch: 6 [89/250 11392/32000 (36%)] Loss: 1.97942 (semantic_loss: 0.03003, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.33031
Train Epoch: 6 [100/250 12800/32000 (40%)] Loss: 1.97844 (semantic_loss: 0.02905, quant_loss: 1.94922, bit_balance_loss: 0.00017) batch_time=0.32299
Train Epoch: 6 [111/250 14208/32000 (44%)] Loss: 1.97894 (semantic_loss: 0.02858, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.32768
Train Epoch: 6 [122/250 15616/32000 (49%)] Loss: 1.97868 (semantic_loss: 0.02832, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.34093
Train Epoch: 6 [133/250 17024/32000 (53%)] Loss: 1.97759 (semantic_loss: 0.02723, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=2.03154
Train Epoch: 6 [144/250 18432/32000 (58%)] Loss: 1.97680 (semantic_loss: 0.02742, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=1.97644
Train Epoch: 6 [155/250 19840/32000 (62%)] Loss: 1.97766 (semantic_loss: 0.02731, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.33963
Train Epoch: 6 [166/250 21248/32000 (66%)] Loss: 1.97568 (semantic_loss: 0.02630, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=0.33865
Train Epoch: 6 [177/250 22656/32000 (71%)] Loss: 1.97842 (semantic_loss: 0.02807, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.34252
Train Epoch: 6 [188/250 24064/32000 (75%)] Loss: 1.97919 (semantic_loss: 0.02884, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.32597
Train Epoch: 6 [199/250 25472/32000 (80%)] Loss: 1.97680 (semantic_loss: 0.02743, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=2.47898
Train Epoch: 6 [210/250 26880/32000 (84%)] Loss: 1.97966 (semantic_loss: 0.02931, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.34378
Train Epoch: 6 [221/250 28288/32000 (88%)] Loss: 1.97708 (semantic_loss: 0.02770, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=1.62352
Train Epoch: 6 [232/250 29696/32000 (93%)] Loss: 1.97571 (semantic_loss: 0.02634, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.34669
Train Epoch: 6 [243/250 31104/32000 (97%)] Loss: 1.97633 (semantic_loss: 0.02696, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=0.32764
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch6.pth ...
Done in 3.913s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch6.pth ...
Done in 7.607s
removing stale ckpt [epoch 5] [took 0.00s]
epoch : 6
loss : 1.9776311101913453
learning_rate : 3.868904687499999e-05
n_samples : 192000
n_steps : 1500
MSRVTT_jsfusion_test/t2v_metrics/R1: 7.4
MSRVTT_jsfusion_test/t2v_metrics/R5: 25.8
MSRVTT_jsfusion_test/t2v_metrics/R10: 39.9
MSRVTT_jsfusion_test/t2v_metrics/R50: 76.1
MSRVTT_jsfusion_test/t2v_metrics/MedR: 17.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 50.0435
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 19.676209611080754
MSRVTT_jsfusion_test/v2t_metrics/R1: 7.1
MSRVTT_jsfusion_test/v2t_metrics/R5: 26.0
MSRVTT_jsfusion_test/v2t_metrics/R10: 39.7
MSRVTT_jsfusion_test/v2t_metrics/R50: 75.1
MSRVTT_jsfusion_test/v2t_metrics/MedR: 16.5
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 50.818
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 19.4240923621854
mnt_best : 19.676209611080754
not_improved_count: 0
Train Epoch: 7 [1/250 128/32000 (0%)] Loss: 1.97488 (semantic_loss: 0.02551, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=29.61169
Train Epoch: 7 [12/250 1536/32000 (5%)] Loss: 1.97713 (semantic_loss: 0.02775, quant_loss: 1.94922, bit_balance_loss: 0.00016) batch_time=0.45703
Train Epoch: 7 [23/250 2944/32000 (9%)] Loss: 1.97719 (semantic_loss: 0.02684, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.34296
Train Epoch: 7 [34/250 4352/32000 (14%)] Loss: 1.97592 (semantic_loss: 0.02655, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=1.57383
Train Epoch: 7 [45/250 5760/32000 (18%)] Loss: 1.97617 (semantic_loss: 0.02583, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.34423
Train Epoch: 7 [56/250 7168/32000 (22%)] Loss: 1.97494 (semantic_loss: 0.02557, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.33076
Train Epoch: 7 [67/250 8576/32000 (27%)] Loss: 1.97640 (semantic_loss: 0.02605, quant_loss: 1.95020, bit_balance_loss: 0.00016) batch_time=0.88075
Train Epoch: 7 [78/250 9984/32000 (31%)] Loss: 1.97757 (semantic_loss: 0.02723, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.34393
Train Epoch: 7 [89/250 11392/32000 (36%)] Loss: 1.97702 (semantic_loss: 0.02765, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.33866
Train Epoch: 7 [100/250 12800/32000 (40%)] Loss: 1.97844 (semantic_loss: 0.02810, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.35191
Train Epoch: 7 [111/250 14208/32000 (44%)] Loss: 1.97608 (semantic_loss: 0.02573, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.35797
Train Epoch: 7 [122/250 15616/32000 (49%)] Loss: 1.97728 (semantic_loss: 0.02694, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.34227
Train Epoch: 7 [133/250 17024/32000 (53%)] Loss: 1.97593 (semantic_loss: 0.02558, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33701
Train Epoch: 7 [144/250 18432/32000 (58%)] Loss: 1.97578 (semantic_loss: 0.02641, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.32540
Train Epoch: 7 [155/250 19840/32000 (62%)] Loss: 1.97670 (semantic_loss: 0.02733, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.32619
Train Epoch: 7 [166/250 21248/32000 (66%)] Loss: 1.97527 (semantic_loss: 0.02492, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.35803
Train Epoch: 7 [177/250 22656/32000 (71%)] Loss: 1.97558 (semantic_loss: 0.02622, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.37538
Train Epoch: 7 [188/250 24064/32000 (75%)] Loss: 1.97342 (semantic_loss: 0.02503, quant_loss: 1.94824, bit_balance_loss: 0.00015) batch_time=0.32589
Train Epoch: 7 [199/250 25472/32000 (80%)] Loss: 1.97566 (semantic_loss: 0.02629, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.36427
Train Epoch: 7 [210/250 26880/32000 (84%)] Loss: 1.97432 (semantic_loss: 0.02397, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.55100
Train Epoch: 7 [221/250 28288/32000 (88%)] Loss: 1.97547 (semantic_loss: 0.02513, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33333
Train Epoch: 7 [232/250 29696/32000 (93%)] Loss: 1.97485 (semantic_loss: 0.02548, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.35949
Train Epoch: 7 [243/250 31104/32000 (97%)] Loss: 1.97692 (semantic_loss: 0.02658, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.35162
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch7.pth ...
Done in 3.959s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch7.pth ...
Done in 7.767s
removing stale ckpt [epoch 6] [took 0.00s]
epoch : 7
loss : 1.9759862656593323
learning_rate : 3.675459453124999e-05
n_samples : 224000
n_steps : 1750
MSRVTT_jsfusion_test/t2v_metrics/R1: 9.0
MSRVTT_jsfusion_test/t2v_metrics/R5: 28.8
MSRVTT_jsfusion_test/t2v_metrics/R10: 42.0
MSRVTT_jsfusion_test/t2v_metrics/R50: 78.3
MSRVTT_jsfusion_test/t2v_metrics/MedR: 15.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 45.1885
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 22.162977005374003
MSRVTT_jsfusion_test/v2t_metrics/R1: 8.7
MSRVTT_jsfusion_test/v2t_metrics/R5: 29.3
MSRVTT_jsfusion_test/v2t_metrics/R10: 42.5
MSRVTT_jsfusion_test/v2t_metrics/R50: 77.5
MSRVTT_jsfusion_test/v2t_metrics/MedR: 14.5
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 44.519
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 22.127139184781417
mnt_best : 22.162977005374003
not_improved_count: 0
Train Epoch: 8 [1/250 128/32000 (0%)] Loss: 1.97545 (semantic_loss: 0.02609, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=29.08181
Train Epoch: 8 [12/250 1536/32000 (5%)] Loss: 1.97689 (semantic_loss: 0.02655, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.32351
Train Epoch: 8 [23/250 2944/32000 (9%)] Loss: 1.97449 (semantic_loss: 0.02415, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33379
Train Epoch: 8 [34/250 4352/32000 (14%)] Loss: 1.97573 (semantic_loss: 0.02538, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33053
Train Epoch: 8 [45/250 5760/32000 (18%)] Loss: 1.97501 (semantic_loss: 0.02467, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.35077
Train Epoch: 8 [56/250 7168/32000 (22%)] Loss: 1.97569 (semantic_loss: 0.02535, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.32053
Train Epoch: 8 [67/250 8576/32000 (27%)] Loss: 1.97547 (semantic_loss: 0.02513, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=2.98766
Train Epoch: 8 [78/250 9984/32000 (31%)] Loss: 1.97411 (semantic_loss: 0.02474, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.33794
Train Epoch: 8 [89/250 11392/32000 (36%)] Loss: 1.97630 (semantic_loss: 0.02596, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.36877
Train Epoch: 8 [100/250 12800/32000 (40%)] Loss: 1.97501 (semantic_loss: 0.02467, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.32742
Train Epoch: 8 [111/250 14208/32000 (44%)] Loss: 1.97376 (semantic_loss: 0.02342, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32047
Train Epoch: 8 [122/250 15616/32000 (49%)] Loss: 1.97538 (semantic_loss: 0.02601, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.43797
Train Epoch: 8 [133/250 17024/32000 (53%)] Loss: 1.97304 (semantic_loss: 0.02367, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.96069
Train Epoch: 8 [144/250 18432/32000 (58%)] Loss: 1.97483 (semantic_loss: 0.02546, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.33601
Train Epoch: 8 [155/250 19840/32000 (62%)] Loss: 1.97570 (semantic_loss: 0.02634, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.38556
Train Epoch: 8 [166/250 21248/32000 (66%)] Loss: 1.97404 (semantic_loss: 0.02467, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.33069
Train Epoch: 8 [177/250 22656/32000 (71%)] Loss: 1.97567 (semantic_loss: 0.02631, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.32924
Train Epoch: 8 [188/250 24064/32000 (75%)] Loss: 1.97653 (semantic_loss: 0.02619, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33155
Train Epoch: 8 [199/250 25472/32000 (80%)] Loss: 1.97433 (semantic_loss: 0.02497, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33886
Train Epoch: 8 [210/250 26880/32000 (84%)] Loss: 1.97475 (semantic_loss: 0.02539, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.37317
Train Epoch: 8 [221/250 28288/32000 (88%)] Loss: 1.97252 (semantic_loss: 0.02218, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34071
Train Epoch: 8 [232/250 29696/32000 (93%)] Loss: 1.97458 (semantic_loss: 0.02522, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.31998
Train Epoch: 8 [243/250 31104/32000 (97%)] Loss: 1.97380 (semantic_loss: 0.02444, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.32104
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch8.pth ...
Done in 18.063s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch8.pth ...
Done in 21.734s
removing stale ckpt [epoch 7] [took 0.00s]
epoch : 8
loss : 1.974881987094879
learning_rate : 3.4916864804687486e-05
n_samples : 256000
n_steps : 2000
MSRVTT_jsfusion_test/t2v_metrics/R1: 10.5
MSRVTT_jsfusion_test/t2v_metrics/R5: 30.3
MSRVTT_jsfusion_test/t2v_metrics/R10: 43.6
MSRVTT_jsfusion_test/t2v_metrics/R50: 77.7
MSRVTT_jsfusion_test/t2v_metrics/MedR: 14.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 44.226
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 24.02736462053925
MSRVTT_jsfusion_test/v2t_metrics/R1: 9.9
MSRVTT_jsfusion_test/v2t_metrics/R5: 30.7
MSRVTT_jsfusion_test/v2t_metrics/R10: 44.4
MSRVTT_jsfusion_test/v2t_metrics/R50: 79.9
MSRVTT_jsfusion_test/v2t_metrics/MedR: 13.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 43.1955
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 23.807777040868576
mnt_best : 24.02736462053925
not_improved_count: 0
Train Epoch: 9 [1/250 128/32000 (0%)] Loss: 1.97520 (semantic_loss: 0.02486, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=25.96196
Train Epoch: 9 [12/250 1536/32000 (5%)] Loss: 1.97404 (semantic_loss: 0.02468, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=2.22275
Train Epoch: 9 [23/250 2944/32000 (9%)] Loss: 1.97166 (semantic_loss: 0.02229, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.71324
Train Epoch: 9 [34/250 4352/32000 (14%)] Loss: 1.97284 (semantic_loss: 0.02347, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.56360
Train Epoch: 9 [45/250 5760/32000 (18%)] Loss: 1.97120 (semantic_loss: 0.02184, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33420
Train Epoch: 9 [56/250 7168/32000 (22%)] Loss: 1.97331 (semantic_loss: 0.02395, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32978
Train Epoch: 9 [67/250 8576/32000 (27%)] Loss: 1.97237 (semantic_loss: 0.02301, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32839
Train Epoch: 9 [78/250 9984/32000 (31%)] Loss: 1.97200 (semantic_loss: 0.02166, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33374
Train Epoch: 9 [89/250 11392/32000 (36%)] Loss: 1.97264 (semantic_loss: 0.02327, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.35652
Train Epoch: 9 [100/250 12800/32000 (40%)] Loss: 1.97335 (semantic_loss: 0.02301, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33607
Train Epoch: 9 [111/250 14208/32000 (44%)] Loss: 1.97382 (semantic_loss: 0.02446, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35435
Train Epoch: 9 [122/250 15616/32000 (49%)] Loss: 1.97420 (semantic_loss: 0.02484, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.37089
Train Epoch: 9 [133/250 17024/32000 (53%)] Loss: 1.97386 (semantic_loss: 0.02352, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33946
Train Epoch: 9 [144/250 18432/32000 (58%)] Loss: 1.97422 (semantic_loss: 0.02388, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.37548
Train Epoch: 9 [155/250 19840/32000 (62%)] Loss: 1.97527 (semantic_loss: 0.02493, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33815
Train Epoch: 9 [166/250 21248/32000 (66%)] Loss: 1.97331 (semantic_loss: 0.02395, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35505
Train Epoch: 9 [177/250 22656/32000 (71%)] Loss: 1.97337 (semantic_loss: 0.02304, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33342
Train Epoch: 9 [188/250 24064/32000 (75%)] Loss: 1.97193 (semantic_loss: 0.02257, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.33307
Train Epoch: 9 [199/250 25472/32000 (80%)] Loss: 1.97539 (semantic_loss: 0.02505, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.32648
Train Epoch: 9 [210/250 26880/32000 (84%)] Loss: 1.97300 (semantic_loss: 0.02363, quant_loss: 1.94922, bit_balance_loss: 0.00015) batch_time=0.40197
Train Epoch: 9 [221/250 28288/32000 (88%)] Loss: 1.97241 (semantic_loss: 0.02305, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34184
Train Epoch: 9 [232/250 29696/32000 (93%)] Loss: 1.97507 (semantic_loss: 0.02474, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34223
Train Epoch: 9 [243/250 31104/32000 (97%)] Loss: 1.97392 (semantic_loss: 0.02358, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.34206
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch9.pth ...
Done in 3.798s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch9.pth ...
Done in 7.457s
removing stale ckpt [epoch 8] [took 0.00s]
epoch : 9
loss : 1.9737976188659667
learning_rate : 3.317102156445311e-05
n_samples : 288000
n_steps : 2250
MSRVTT_jsfusion_test/t2v_metrics/R1: 10.4
MSRVTT_jsfusion_test/t2v_metrics/R5: 34.5
MSRVTT_jsfusion_test/t2v_metrics/R10: 47.9
MSRVTT_jsfusion_test/t2v_metrics/R50: 79.6
MSRVTT_jsfusion_test/t2v_metrics/MedR: 12.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 42.0315
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 25.806512387625688
MSRVTT_jsfusion_test/v2t_metrics/R1: 12.1
MSRVTT_jsfusion_test/v2t_metrics/R5: 35.4
MSRVTT_jsfusion_test/v2t_metrics/R10: 49.1
MSRVTT_jsfusion_test/v2t_metrics/R50: 80.9
MSRVTT_jsfusion_test/v2t_metrics/MedR: 11.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 40.673
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 27.603026868201866
mnt_best : 25.806512387625688
not_improved_count: 0
Train Epoch: 10 [1/250 128/32000 (0%)] Loss: 1.97439 (semantic_loss: 0.02405, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=34.64514
Train Epoch: 10 [12/250 1536/32000 (5%)] Loss: 1.97377 (semantic_loss: 0.02440, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32886
Train Epoch: 10 [23/250 2944/32000 (9%)] Loss: 1.97356 (semantic_loss: 0.02322, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34312
Train Epoch: 10 [34/250 4352/32000 (14%)] Loss: 1.97361 (semantic_loss: 0.02327, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36421
Train Epoch: 10 [45/250 5760/32000 (18%)] Loss: 1.97392 (semantic_loss: 0.02358, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34345
Train Epoch: 10 [56/250 7168/32000 (22%)] Loss: 1.97434 (semantic_loss: 0.02400, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33982
Train Epoch: 10 [67/250 8576/32000 (27%)] Loss: 1.97192 (semantic_loss: 0.02256, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32800
Train Epoch: 10 [78/250 9984/32000 (31%)] Loss: 1.97259 (semantic_loss: 0.02226, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=2.71621
Train Epoch: 10 [89/250 11392/32000 (36%)] Loss: 1.97659 (semantic_loss: 0.02625, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=1.39139
Train Epoch: 10 [100/250 12800/32000 (40%)] Loss: 1.97476 (semantic_loss: 0.02442, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34165
Train Epoch: 10 [111/250 14208/32000 (44%)] Loss: 1.97129 (semantic_loss: 0.02193, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34638
Train Epoch: 10 [122/250 15616/32000 (49%)] Loss: 1.97349 (semantic_loss: 0.02412, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33273
Train Epoch: 10 [133/250 17024/32000 (53%)] Loss: 1.97347 (semantic_loss: 0.02313, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32621
Train Epoch: 10 [144/250 18432/32000 (58%)] Loss: 1.97256 (semantic_loss: 0.02222, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32883
Train Epoch: 10 [155/250 19840/32000 (62%)] Loss: 1.97290 (semantic_loss: 0.02256, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32626
Train Epoch: 10 [166/250 21248/32000 (66%)] Loss: 1.97613 (semantic_loss: 0.02580, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33112
Train Epoch: 10 [177/250 22656/32000 (71%)] Loss: 1.97145 (semantic_loss: 0.02209, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33426
Train Epoch: 10 [188/250 24064/32000 (75%)] Loss: 1.97271 (semantic_loss: 0.02335, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34032
Train Epoch: 10 [199/250 25472/32000 (80%)] Loss: 1.97232 (semantic_loss: 0.02198, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.33800
Train Epoch: 10 [210/250 26880/32000 (84%)] Loss: 1.97365 (semantic_loss: 0.02331, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32796
Train Epoch: 10 [221/250 28288/32000 (88%)] Loss: 1.97211 (semantic_loss: 0.02275, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34693
Train Epoch: 10 [232/250 29696/32000 (93%)] Loss: 1.97313 (semantic_loss: 0.02280, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33457
Train Epoch: 10 [243/250 31104/32000 (97%)] Loss: 1.97353 (semantic_loss: 0.02319, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.44650
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch10.pth ...
Done in 3.786s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch10.pth ...
Done in 7.617s
removing stale ckpt [epoch 9] [took 0.00s]
epoch : 10
loss : 1.9729892263412476
learning_rate : 3.151247048623045e-05
n_samples : 320000
n_steps : 2500
MSRVTT_jsfusion_test/t2v_metrics/R1: 12.1
MSRVTT_jsfusion_test/t2v_metrics/R5: 33.2
MSRVTT_jsfusion_test/t2v_metrics/R10: 48.2
MSRVTT_jsfusion_test/t2v_metrics/R50: 80.6
MSRVTT_jsfusion_test/t2v_metrics/MedR: 11.5
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 38.947
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 26.852836285912055
MSRVTT_jsfusion_test/v2t_metrics/R1: 11.4
MSRVTT_jsfusion_test/v2t_metrics/R5: 33.4
MSRVTT_jsfusion_test/v2t_metrics/R10: 48.2
MSRVTT_jsfusion_test/v2t_metrics/R50: 81.8
MSRVTT_jsfusion_test/v2t_metrics/MedR: 11.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 37.8105
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 26.377448602061047
mnt_best : 26.852836285912055
not_improved_count: 0
Train Epoch: 11 [1/250 128/32000 (0%)] Loss: 1.97116 (semantic_loss: 0.02180, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=31.32738
Train Epoch: 11 [12/250 1536/32000 (5%)] Loss: 1.97347 (semantic_loss: 0.02313, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34292
Train Epoch: 11 [23/250 2944/32000 (9%)] Loss: 1.97150 (semantic_loss: 0.02116, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=2.44624
Train Epoch: 11 [34/250 4352/32000 (14%)] Loss: 1.97387 (semantic_loss: 0.02353, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36694
Train Epoch: 11 [45/250 5760/32000 (18%)] Loss: 1.97233 (semantic_loss: 0.02297, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33960
Train Epoch: 11 [56/250 7168/32000 (22%)] Loss: 1.97325 (semantic_loss: 0.02291, quant_loss: 1.95020, bit_balance_loss: 0.00015) batch_time=0.37048
Train Epoch: 11 [67/250 8576/32000 (27%)] Loss: 1.97339 (semantic_loss: 0.02305, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34649
Train Epoch: 11 [78/250 9984/32000 (31%)] Loss: 1.97259 (semantic_loss: 0.02323, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.36936
Train Epoch: 11 [89/250 11392/32000 (36%)] Loss: 1.97218 (semantic_loss: 0.02184, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33934
Train Epoch: 11 [100/250 12800/32000 (40%)] Loss: 1.97443 (semantic_loss: 0.02409, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34164
Train Epoch: 11 [111/250 14208/32000 (44%)] Loss: 1.97378 (semantic_loss: 0.02345, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.45050
Train Epoch: 11 [122/250 15616/32000 (49%)] Loss: 1.97223 (semantic_loss: 0.02288, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34065
Train Epoch: 11 [133/250 17024/32000 (53%)] Loss: 1.97203 (semantic_loss: 0.02268, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33708
Train Epoch: 11 [144/250 18432/32000 (58%)] Loss: 1.97099 (semantic_loss: 0.02163, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34635
Train Epoch: 11 [155/250 19840/32000 (62%)] Loss: 1.97121 (semantic_loss: 0.02088, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.38816
Train Epoch: 11 [166/250 21248/32000 (66%)] Loss: 1.97042 (semantic_loss: 0.02105, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34196
Train Epoch: 11 [177/250 22656/32000 (71%)] Loss: 1.97241 (semantic_loss: 0.02305, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.80074
Train Epoch: 11 [188/250 24064/32000 (75%)] Loss: 1.97240 (semantic_loss: 0.02303, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33488
Train Epoch: 11 [199/250 25472/32000 (80%)] Loss: 1.97072 (semantic_loss: 0.02136, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34533
Train Epoch: 11 [210/250 26880/32000 (84%)] Loss: 1.97380 (semantic_loss: 0.02444, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32803
Train Epoch: 11 [221/250 28288/32000 (88%)] Loss: 1.97190 (semantic_loss: 0.02254, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33899
Train Epoch: 11 [232/250 29696/32000 (93%)] Loss: 1.97052 (semantic_loss: 0.02117, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33738
Train Epoch: 11 [243/250 31104/32000 (97%)] Loss: 1.97281 (semantic_loss: 0.02247, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.58384
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch11.pth ...
Done in 3.895s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch11.pth ...
Done in 7.611s
removing stale ckpt [epoch 10] [took 0.00s]
epoch : 11
loss : 1.9723797092437745
learning_rate : 2.993684696191893e-05
n_samples : 352000
n_steps : 2750
MSRVTT_jsfusion_test/t2v_metrics/R1: 12.3
MSRVTT_jsfusion_test/t2v_metrics/R5: 35.4
MSRVTT_jsfusion_test/t2v_metrics/R10: 49.6
MSRVTT_jsfusion_test/t2v_metrics/R50: 80.6
MSRVTT_jsfusion_test/t2v_metrics/MedR: 11.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 39.242
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 27.848171402382903
MSRVTT_jsfusion_test/v2t_metrics/R1: 14.1
MSRVTT_jsfusion_test/v2t_metrics/R5: 36.6
MSRVTT_jsfusion_test/v2t_metrics/R10: 50.1
MSRVTT_jsfusion_test/v2t_metrics/R50: 82.1
MSRVTT_jsfusion_test/v2t_metrics/MedR: 10.25
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 38.26
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 29.569635742119843
mnt_best : 27.848171402382903
not_improved_count: 0
Train Epoch: 12 [1/250 128/32000 (0%)] Loss: 1.97174 (semantic_loss: 0.02141, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=32.12269
Train Epoch: 12 [12/250 1536/32000 (5%)] Loss: 1.97271 (semantic_loss: 0.02237, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35051
Train Epoch: 12 [23/250 2944/32000 (9%)] Loss: 1.96907 (semantic_loss: 0.01972, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33783
Train Epoch: 12 [34/250 4352/32000 (14%)] Loss: 1.97050 (semantic_loss: 0.02114, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35880
Train Epoch: 12 [45/250 5760/32000 (18%)] Loss: 1.97198 (semantic_loss: 0.02165, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35126
Train Epoch: 12 [56/250 7168/32000 (22%)] Loss: 1.97487 (semantic_loss: 0.02453, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33884
Train Epoch: 12 [67/250 8576/32000 (27%)] Loss: 1.97247 (semantic_loss: 0.02214, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32716
Train Epoch: 12 [78/250 9984/32000 (31%)] Loss: 1.97147 (semantic_loss: 0.02113, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35572
Train Epoch: 12 [89/250 11392/32000 (36%)] Loss: 1.97206 (semantic_loss: 0.02172, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34481
Train Epoch: 12 [100/250 12800/32000 (40%)] Loss: 1.97225 (semantic_loss: 0.02290, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34446
Train Epoch: 12 [111/250 14208/32000 (44%)] Loss: 1.97123 (semantic_loss: 0.02285, quant_loss: 1.94824, bit_balance_loss: 0.00014) batch_time=0.33387
Train Epoch: 12 [122/250 15616/32000 (49%)] Loss: 1.97147 (semantic_loss: 0.02113, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33942
Train Epoch: 12 [133/250 17024/32000 (53%)] Loss: 1.97107 (semantic_loss: 0.02171, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33627
Train Epoch: 12 [144/250 18432/32000 (58%)] Loss: 1.97092 (semantic_loss: 0.02156, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33145
Train Epoch: 12 [155/250 19840/32000 (62%)] Loss: 1.97314 (semantic_loss: 0.02281, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33930
Train Epoch: 12 [166/250 21248/32000 (66%)] Loss: 1.97207 (semantic_loss: 0.02173, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33331
Train Epoch: 12 [177/250 22656/32000 (71%)] Loss: 1.97145 (semantic_loss: 0.02209, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35273
Train Epoch: 12 [188/250 24064/32000 (75%)] Loss: 1.97302 (semantic_loss: 0.02268, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32924
Train Epoch: 12 [199/250 25472/32000 (80%)] Loss: 1.97107 (semantic_loss: 0.02171, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35578
Train Epoch: 12 [210/250 26880/32000 (84%)] Loss: 1.97239 (semantic_loss: 0.02206, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35270
Train Epoch: 12 [221/250 28288/32000 (88%)] Loss: 1.97221 (semantic_loss: 0.02187, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36533
Train Epoch: 12 [232/250 29696/32000 (93%)] Loss: 1.97024 (semantic_loss: 0.01991, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36974
Train Epoch: 12 [243/250 31104/32000 (97%)] Loss: 1.97141 (semantic_loss: 0.02108, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33004
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch12.pth ...
Done in 3.964s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch12.pth ...
Done in 8.041s
removing stale ckpt [epoch 11] [took 0.00s]
epoch : 12
loss : 1.9716937117576598
learning_rate : 2.844000461382298e-05
n_samples : 384000
n_steps : 3000
MSRVTT_jsfusion_test/t2v_metrics/R1: 12.5
MSRVTT_jsfusion_test/t2v_metrics/R5: 36.2
MSRVTT_jsfusion_test/t2v_metrics/R10: 50.7
MSRVTT_jsfusion_test/t2v_metrics/R50: 82.1
MSRVTT_jsfusion_test/t2v_metrics/MedR: 10.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 36.4805
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 28.414641494846443
MSRVTT_jsfusion_test/v2t_metrics/R1: 12.4
MSRVTT_jsfusion_test/v2t_metrics/R5: 38.6
MSRVTT_jsfusion_test/v2t_metrics/R10: 51.9
MSRVTT_jsfusion_test/v2t_metrics/R50: 81.6
MSRVTT_jsfusion_test/v2t_metrics/MedR: 10.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 36.6785
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 29.178219198647007
mnt_best : 28.414641494846443
not_improved_count: 0
Train Epoch: 13 [1/250 128/32000 (0%)] Loss: 1.97230 (semantic_loss: 0.02196, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=27.74543
Train Epoch: 13 [12/250 1536/32000 (5%)] Loss: 1.97160 (semantic_loss: 0.02127, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36037
Train Epoch: 13 [23/250 2944/32000 (9%)] Loss: 1.97128 (semantic_loss: 0.02192, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.36287
Train Epoch: 13 [34/250 4352/32000 (14%)] Loss: 1.97107 (semantic_loss: 0.02073, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.71264
Train Epoch: 13 [45/250 5760/32000 (18%)] Loss: 1.97082 (semantic_loss: 0.02049, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34811
Train Epoch: 13 [56/250 7168/32000 (22%)] Loss: 1.97025 (semantic_loss: 0.02089, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34717
Train Epoch: 13 [67/250 8576/32000 (27%)] Loss: 1.97294 (semantic_loss: 0.02261, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36064
Train Epoch: 13 [78/250 9984/32000 (31%)] Loss: 1.97024 (semantic_loss: 0.02089, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35638
Train Epoch: 13 [89/250 11392/32000 (36%)] Loss: 1.96851 (semantic_loss: 0.01915, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.36998
Train Epoch: 13 [100/250 12800/32000 (40%)] Loss: 1.97192 (semantic_loss: 0.02256, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34756
Train Epoch: 13 [111/250 14208/32000 (44%)] Loss: 1.97054 (semantic_loss: 0.02021, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36122
Train Epoch: 13 [122/250 15616/32000 (49%)] Loss: 1.97308 (semantic_loss: 0.02275, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34918
Train Epoch: 13 [133/250 17024/32000 (53%)] Loss: 1.96986 (semantic_loss: 0.02050, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34114
Train Epoch: 13 [144/250 18432/32000 (58%)] Loss: 1.97280 (semantic_loss: 0.02246, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33744
Train Epoch: 13 [155/250 19840/32000 (62%)] Loss: 1.96934 (semantic_loss: 0.01999, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34297
Train Epoch: 13 [166/250 21248/32000 (66%)] Loss: 1.97305 (semantic_loss: 0.02272, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34328
Train Epoch: 13 [177/250 22656/32000 (71%)] Loss: 1.97272 (semantic_loss: 0.02238, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33536
Train Epoch: 13 [188/250 24064/32000 (75%)] Loss: 1.97186 (semantic_loss: 0.02152, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33970
Train Epoch: 13 [199/250 25472/32000 (80%)] Loss: 1.96999 (semantic_loss: 0.01966, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34932
Train Epoch: 13 [210/250 26880/32000 (84%)] Loss: 1.97004 (semantic_loss: 0.02068, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33585
Train Epoch: 13 [221/250 28288/32000 (88%)] Loss: 1.97090 (semantic_loss: 0.02154, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33629
Train Epoch: 13 [232/250 29696/32000 (93%)] Loss: 1.97209 (semantic_loss: 0.02175, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36215
Train Epoch: 13 [243/250 31104/32000 (97%)] Loss: 1.97119 (semantic_loss: 0.02086, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33919
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch13.pth ...
Done in 3.879s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch13.pth ...
Done in 7.465s
removing stale ckpt [epoch 12] [took 0.00s]
epoch : 13
loss : 1.9710919966697693
learning_rate : 2.7018004383131832e-05
n_samples : 416000
n_steps : 3250
MSRVTT_jsfusion_test/t2v_metrics/R1: 13.0
MSRVTT_jsfusion_test/t2v_metrics/R5: 37.6
MSRVTT_jsfusion_test/t2v_metrics/R10: 50.9
MSRVTT_jsfusion_test/t2v_metrics/R50: 82.5
MSRVTT_jsfusion_test/t2v_metrics/MedR: 10.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 35.135
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 29.19328675344571
MSRVTT_jsfusion_test/v2t_metrics/R1: 14.3
MSRVTT_jsfusion_test/v2t_metrics/R5: 38.6
MSRVTT_jsfusion_test/v2t_metrics/R10: 53.7
MSRVTT_jsfusion_test/v2t_metrics/R50: 83.1
MSRVTT_jsfusion_test/v2t_metrics/MedR: 9.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 34.7395
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 30.947996752418568
mnt_best : 29.19328675344571
not_improved_count: 0
Train Epoch: 14 [1/250 128/32000 (0%)] Loss: 1.97062 (semantic_loss: 0.02029, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=29.41704
Train Epoch: 14 [12/250 1536/32000 (5%)] Loss: 1.97272 (semantic_loss: 0.02238, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32949
Train Epoch: 14 [23/250 2944/32000 (9%)] Loss: 1.97093 (semantic_loss: 0.02157, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33227
Train Epoch: 14 [34/250 4352/32000 (14%)] Loss: 1.97055 (semantic_loss: 0.02021, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33660
Train Epoch: 14 [45/250 5760/32000 (18%)] Loss: 1.97062 (semantic_loss: 0.02126, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32996
Train Epoch: 14 [56/250 7168/32000 (22%)] Loss: 1.96947 (semantic_loss: 0.02011, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32662
Train Epoch: 14 [67/250 8576/32000 (27%)] Loss: 1.96835 (semantic_loss: 0.01900, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33676
Train Epoch: 14 [78/250 9984/32000 (31%)] Loss: 1.96988 (semantic_loss: 0.01955, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33285
Train Epoch: 14 [89/250 11392/32000 (36%)] Loss: 1.97026 (semantic_loss: 0.01992, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33596
Train Epoch: 14 [100/250 12800/32000 (40%)] Loss: 1.97033 (semantic_loss: 0.02097, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35301
Train Epoch: 14 [111/250 14208/32000 (44%)] Loss: 1.97046 (semantic_loss: 0.02111, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33916
Train Epoch: 14 [122/250 15616/32000 (49%)] Loss: 1.97088 (semantic_loss: 0.02152, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34296
Train Epoch: 14 [133/250 17024/32000 (53%)] Loss: 1.97013 (semantic_loss: 0.01980, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32406
Train Epoch: 14 [144/250 18432/32000 (58%)] Loss: 1.96925 (semantic_loss: 0.01989, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34644
Train Epoch: 14 [155/250 19840/32000 (62%)] Loss: 1.96806 (semantic_loss: 0.01870, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32481
Train Epoch: 14 [166/250 21248/32000 (66%)] Loss: 1.97100 (semantic_loss: 0.02164, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32781
Train Epoch: 14 [177/250 22656/32000 (71%)] Loss: 1.97178 (semantic_loss: 0.02145, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32821
Train Epoch: 14 [188/250 24064/32000 (75%)] Loss: 1.96992 (semantic_loss: 0.01959, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34976
Train Epoch: 14 [199/250 25472/32000 (80%)] Loss: 1.97018 (semantic_loss: 0.01984, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=6.53753
Train Epoch: 14 [210/250 26880/32000 (84%)] Loss: 1.97073 (semantic_loss: 0.02138, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33645
Train Epoch: 14 [221/250 28288/32000 (88%)] Loss: 1.97068 (semantic_loss: 0.02035, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33672
Train Epoch: 14 [232/250 29696/32000 (93%)] Loss: 1.97038 (semantic_loss: 0.02004, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33759
Train Epoch: 14 [243/250 31104/32000 (97%)] Loss: 1.97118 (semantic_loss: 0.02085, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32813
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch14.pth ...
Done in 3.931s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch14.pth ...
Done in 8.721s
removing stale ckpt [epoch 13] [took 0.00s]
epoch : 14
loss : 1.9705258221626283
learning_rate : 2.566710416397524e-05
n_samples : 448000
n_steps : 3500
MSRVTT_jsfusion_test/t2v_metrics/R1: 12.3
MSRVTT_jsfusion_test/t2v_metrics/R5: 39.4
MSRVTT_jsfusion_test/t2v_metrics/R10: 52.5
MSRVTT_jsfusion_test/t2v_metrics/R50: 83.5
MSRVTT_jsfusion_test/t2v_metrics/MedR: 10.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 36.1765
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 29.411705737136675
MSRVTT_jsfusion_test/v2t_metrics/R1: 13.2
MSRVTT_jsfusion_test/v2t_metrics/R5: 41.0
MSRVTT_jsfusion_test/v2t_metrics/R10: 54.7
MSRVTT_jsfusion_test/v2t_metrics/R50: 84.3
MSRVTT_jsfusion_test/v2t_metrics/MedR: 8.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 35.5925
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 30.934875422693057
mnt_best : 29.411705737136675
not_improved_count: 0
Train Epoch: 15 [1/250 128/32000 (0%)] Loss: 1.96814 (semantic_loss: 0.01878, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=31.90274
Train Epoch: 15 [12/250 1536/32000 (5%)] Loss: 1.97028 (semantic_loss: 0.01994, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32878
Train Epoch: 15 [23/250 2944/32000 (9%)] Loss: 1.96929 (semantic_loss: 0.01994, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33210
Train Epoch: 15 [34/250 4352/32000 (14%)] Loss: 1.97132 (semantic_loss: 0.02099, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=1.09976
Train Epoch: 15 [45/250 5760/32000 (18%)] Loss: 1.96903 (semantic_loss: 0.01967, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32354
Train Epoch: 15 [56/250 7168/32000 (22%)] Loss: 1.96883 (semantic_loss: 0.01947, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.80725
Train Epoch: 15 [67/250 8576/32000 (27%)] Loss: 1.96965 (semantic_loss: 0.02030, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.34606
Train Epoch: 15 [78/250 9984/32000 (31%)] Loss: 1.96800 (semantic_loss: 0.01865, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.36205
Train Epoch: 15 [89/250 11392/32000 (36%)] Loss: 1.97049 (semantic_loss: 0.02113, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34163
Train Epoch: 15 [100/250 12800/32000 (40%)] Loss: 1.96915 (semantic_loss: 0.01980, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34203
Train Epoch: 15 [111/250 14208/32000 (44%)] Loss: 1.96888 (semantic_loss: 0.01952, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33980
Train Epoch: 15 [122/250 15616/32000 (49%)] Loss: 1.97036 (semantic_loss: 0.02002, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33136
Train Epoch: 15 [133/250 17024/32000 (53%)] Loss: 1.96904 (semantic_loss: 0.01870, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.37776
Train Epoch: 15 [144/250 18432/32000 (58%)] Loss: 1.97047 (semantic_loss: 0.02014, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32953
Train Epoch: 15 [155/250 19840/32000 (62%)] Loss: 1.96989 (semantic_loss: 0.02053, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33918
Train Epoch: 15 [166/250 21248/32000 (66%)] Loss: 1.96951 (semantic_loss: 0.02015, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34371
Train Epoch: 15 [177/250 22656/32000 (71%)] Loss: 1.97002 (semantic_loss: 0.01969, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32295
Train Epoch: 15 [188/250 24064/32000 (75%)] Loss: 1.96855 (semantic_loss: 0.01822, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32450
Train Epoch: 15 [199/250 25472/32000 (80%)] Loss: 1.97021 (semantic_loss: 0.01988, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.44555
Train Epoch: 15 [210/250 26880/32000 (84%)] Loss: 1.96850 (semantic_loss: 0.01915, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32939
Train Epoch: 15 [221/250 28288/32000 (88%)] Loss: 1.97001 (semantic_loss: 0.02066, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34026
Train Epoch: 15 [232/250 29696/32000 (93%)] Loss: 1.97278 (semantic_loss: 0.02245, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33184
Train Epoch: 15 [243/250 31104/32000 (97%)] Loss: 1.96894 (semantic_loss: 0.01958, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33569
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch15.pth ...
Done in 4.336s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch15.pth ...
Done in 9.230s
removing stale ckpt [epoch 14] [took 0.00s]
epoch : 15
loss : 1.9699746775627136
learning_rate : 2.4383748955776477e-05
n_samples : 480000
n_steps : 3750
MSRVTT_jsfusion_test/t2v_metrics/R1: 13.6
MSRVTT_jsfusion_test/t2v_metrics/R5: 39.5
MSRVTT_jsfusion_test/t2v_metrics/R10: 53.4
MSRVTT_jsfusion_test/t2v_metrics/R50: 83.0
MSRVTT_jsfusion_test/t2v_metrics/MedR: 9.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 34.8075
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 30.612050446293168
MSRVTT_jsfusion_test/v2t_metrics/R1: 14.4
MSRVTT_jsfusion_test/v2t_metrics/R5: 41.5
MSRVTT_jsfusion_test/v2t_metrics/R10: 55.7
MSRVTT_jsfusion_test/v2t_metrics/R50: 83.4
MSRVTT_jsfusion_test/v2t_metrics/MedR: 8.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 34.188
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 32.167842076598355
mnt_best : 30.612050446293168
not_improved_count: 0
Train Epoch: 16 [1/250 128/32000 (0%)] Loss: 1.96898 (semantic_loss: 0.01865, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=32.11603
Train Epoch: 16 [12/250 1536/32000 (5%)] Loss: 1.96929 (semantic_loss: 0.01993, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33362
Train Epoch: 16 [23/250 2944/32000 (9%)] Loss: 1.97000 (semantic_loss: 0.01967, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33392
Train Epoch: 16 [34/250 4352/32000 (14%)] Loss: 1.96904 (semantic_loss: 0.01871, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.37830
Train Epoch: 16 [45/250 5760/32000 (18%)] Loss: 1.97027 (semantic_loss: 0.01994, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33387
Train Epoch: 16 [56/250 7168/32000 (22%)] Loss: 1.96855 (semantic_loss: 0.01919, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33720
Train Epoch: 16 [67/250 8576/32000 (27%)] Loss: 1.96907 (semantic_loss: 0.01971, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32981
Train Epoch: 16 [78/250 9984/32000 (31%)] Loss: 1.96820 (semantic_loss: 0.01884, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33099
Train Epoch: 16 [89/250 11392/32000 (36%)] Loss: 1.96865 (semantic_loss: 0.01929, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32228
Train Epoch: 16 [100/250 12800/32000 (40%)] Loss: 1.97266 (semantic_loss: 0.02232, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.85404
Train Epoch: 16 [111/250 14208/32000 (44%)] Loss: 1.96912 (semantic_loss: 0.01878, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32658
Train Epoch: 16 [122/250 15616/32000 (49%)] Loss: 1.97097 (semantic_loss: 0.02064, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32598
Train Epoch: 16 [133/250 17024/32000 (53%)] Loss: 1.96929 (semantic_loss: 0.01896, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33394
Train Epoch: 16 [144/250 18432/32000 (58%)] Loss: 1.96821 (semantic_loss: 0.01885, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33204
Train Epoch: 16 [155/250 19840/32000 (62%)] Loss: 1.97107 (semantic_loss: 0.02074, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32437
Train Epoch: 16 [166/250 21248/32000 (66%)] Loss: 1.96992 (semantic_loss: 0.01958, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33767
Train Epoch: 16 [177/250 22656/32000 (71%)] Loss: 1.96841 (semantic_loss: 0.01808, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35637
Train Epoch: 16 [188/250 24064/32000 (75%)] Loss: 1.97061 (semantic_loss: 0.02027, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33985
Train Epoch: 16 [199/250 25472/32000 (80%)] Loss: 1.97116 (semantic_loss: 0.02083, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.35571
Train Epoch: 16 [210/250 26880/32000 (84%)] Loss: 1.96851 (semantic_loss: 0.01915, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32364
Train Epoch: 16 [221/250 28288/32000 (88%)] Loss: 1.97108 (semantic_loss: 0.02075, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32627
Train Epoch: 16 [232/250 29696/32000 (93%)] Loss: 1.96913 (semantic_loss: 0.01978, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33241
Train Epoch: 16 [243/250 31104/32000 (97%)] Loss: 1.96917 (semantic_loss: 0.01884, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32566
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch16.pth ...
Done in 3.810s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch16.pth ...
Done in 7.896s
removing stale ckpt [epoch 15] [took 0.23s]
epoch : 16
loss : 1.9694899773597718
learning_rate : 2.3164561507987653e-05
n_samples : 512000
n_steps : 4000
MSRVTT_jsfusion_test/t2v_metrics/R1: 15.5
MSRVTT_jsfusion_test/t2v_metrics/R5: 39.7
MSRVTT_jsfusion_test/t2v_metrics/R10: 55.7
MSRVTT_jsfusion_test/t2v_metrics/R50: 83.6
MSRVTT_jsfusion_test/t2v_metrics/MedR: 8.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 34.097
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 32.48322447856545
MSRVTT_jsfusion_test/v2t_metrics/R1: 14.8
MSRVTT_jsfusion_test/v2t_metrics/R5: 41.4
MSRVTT_jsfusion_test/v2t_metrics/R10: 56.7
MSRVTT_jsfusion_test/v2t_metrics/R50: 82.6
MSRVTT_jsfusion_test/v2t_metrics/MedR: 8.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 33.8205
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 32.629847080073866
mnt_best : 32.48322447856545
not_improved_count: 0
Train Epoch: 17 [1/250 128/32000 (0%)] Loss: 1.96893 (semantic_loss: 0.01957, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=25.30211
Train Epoch: 17 [12/250 1536/32000 (5%)] Loss: 1.96928 (semantic_loss: 0.01895, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32611
Train Epoch: 17 [23/250 2944/32000 (9%)] Loss: 1.97006 (semantic_loss: 0.01973, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34513
Train Epoch: 17 [34/250 4352/32000 (14%)] Loss: 1.96963 (semantic_loss: 0.01930, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33445
Train Epoch: 17 [45/250 5760/32000 (18%)] Loss: 1.97130 (semantic_loss: 0.02096, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33560
Train Epoch: 17 [56/250 7168/32000 (22%)] Loss: 1.96934 (semantic_loss: 0.01999, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35344
Train Epoch: 17 [67/250 8576/32000 (27%)] Loss: 1.97035 (semantic_loss: 0.02100, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.34470
Train Epoch: 17 [78/250 9984/32000 (31%)] Loss: 1.96901 (semantic_loss: 0.01965, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34946
Train Epoch: 17 [89/250 11392/32000 (36%)] Loss: 1.96816 (semantic_loss: 0.01978, quant_loss: 1.94824, bit_balance_loss: 0.00014) batch_time=0.34522
Train Epoch: 17 [100/250 12800/32000 (40%)] Loss: 1.96844 (semantic_loss: 0.01909, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33476
Train Epoch: 17 [111/250 14208/32000 (44%)] Loss: 1.96782 (semantic_loss: 0.01846, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33555
Train Epoch: 17 [122/250 15616/32000 (49%)] Loss: 1.96979 (semantic_loss: 0.01946, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33006
Train Epoch: 17 [133/250 17024/32000 (53%)] Loss: 1.97031 (semantic_loss: 0.02096, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.66092
Train Epoch: 17 [144/250 18432/32000 (58%)] Loss: 1.96892 (semantic_loss: 0.01956, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35362
Train Epoch: 17 [155/250 19840/32000 (62%)] Loss: 1.96964 (semantic_loss: 0.01931, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34774
Train Epoch: 17 [166/250 21248/32000 (66%)] Loss: 1.96792 (semantic_loss: 0.01856, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33908
Train Epoch: 17 [177/250 22656/32000 (71%)] Loss: 1.96861 (semantic_loss: 0.01926, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.35446
Train Epoch: 17 [188/250 24064/32000 (75%)] Loss: 1.96853 (semantic_loss: 0.01918, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32811
Train Epoch: 17 [199/250 25472/32000 (80%)] Loss: 1.96806 (semantic_loss: 0.01871, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33046
Train Epoch: 17 [210/250 26880/32000 (84%)] Loss: 1.96875 (semantic_loss: 0.01841, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32868
Train Epoch: 17 [221/250 28288/32000 (88%)] Loss: 1.97039 (semantic_loss: 0.02006, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.36715
Train Epoch: 17 [232/250 29696/32000 (93%)] Loss: 1.97000 (semantic_loss: 0.01967, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.67887
Train Epoch: 17 [243/250 31104/32000 (97%)] Loss: 1.96993 (semantic_loss: 0.01960, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32400
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch17.pth ...
Done in 4.034s
removing stale ckpt [epoch 16] [took 0.00s]
epoch : 17
loss : 1.9692578253746034
learning_rate : 2.2006333432588268e-05
n_samples : 544000
n_steps : 4250
MSRVTT_jsfusion_test/t2v_metrics/R1: 14.8
MSRVTT_jsfusion_test/t2v_metrics/R5: 40.5
MSRVTT_jsfusion_test/t2v_metrics/R10: 55.5
MSRVTT_jsfusion_test/t2v_metrics/R50: 83.5
MSRVTT_jsfusion_test/t2v_metrics/MedR: 8.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 33.7415
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 32.16152058971961
MSRVTT_jsfusion_test/v2t_metrics/R1: 15.8
MSRVTT_jsfusion_test/v2t_metrics/R5: 43.5
MSRVTT_jsfusion_test/v2t_metrics/R10: 56.6
MSRVTT_jsfusion_test/v2t_metrics/R50: 84.8
MSRVTT_jsfusion_test/v2t_metrics/MedR: 7.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 32.473
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 33.88344751039688
mnt_best : 32.48322447856545
not_improved_count: 1
Train Epoch: 18 [1/250 128/32000 (0%)] Loss: 1.97001 (semantic_loss: 0.02065, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=32.97279
Train Epoch: 18 [12/250 1536/32000 (5%)] Loss: 1.96957 (semantic_loss: 0.02022, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=2.18891
Train Epoch: 18 [23/250 2944/32000 (9%)] Loss: 1.96915 (semantic_loss: 0.01882, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33359
Train Epoch: 18 [34/250 4352/32000 (14%)] Loss: 1.96805 (semantic_loss: 0.01869, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32448
Train Epoch: 18 [45/250 5760/32000 (18%)] Loss: 1.96853 (semantic_loss: 0.01917, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.36690
Train Epoch: 18 [56/250 7168/32000 (22%)] Loss: 1.96774 (semantic_loss: 0.01838, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.35332
Train Epoch: 18 [67/250 8576/32000 (27%)] Loss: 1.96864 (semantic_loss: 0.01831, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.37670
Train Epoch: 18 [78/250 9984/32000 (31%)] Loss: 1.97007 (semantic_loss: 0.01974, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32137
Train Epoch: 18 [89/250 11392/32000 (36%)] Loss: 1.97123 (semantic_loss: 0.02090, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33211
Train Epoch: 18 [100/250 12800/32000 (40%)] Loss: 1.96697 (semantic_loss: 0.01762, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.35046
Train Epoch: 18 [111/250 14208/32000 (44%)] Loss: 1.96842 (semantic_loss: 0.01809, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32676
Train Epoch: 18 [122/250 15616/32000 (49%)] Loss: 1.97040 (semantic_loss: 0.02007, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32740
Train Epoch: 18 [133/250 17024/32000 (53%)] Loss: 1.96928 (semantic_loss: 0.01895, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35731
Train Epoch: 18 [144/250 18432/32000 (58%)] Loss: 1.96853 (semantic_loss: 0.01820, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.37198
Train Epoch: 18 [155/250 19840/32000 (62%)] Loss: 1.97164 (semantic_loss: 0.02131, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32633
Train Epoch: 18 [166/250 21248/32000 (66%)] Loss: 1.96857 (semantic_loss: 0.01824, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34087
Train Epoch: 18 [177/250 22656/32000 (71%)] Loss: 1.96972 (semantic_loss: 0.01939, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.57843
Train Epoch: 18 [188/250 24064/32000 (75%)] Loss: 1.96808 (semantic_loss: 0.01775, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.35509
Train Epoch: 18 [199/250 25472/32000 (80%)] Loss: 1.96980 (semantic_loss: 0.01946, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=3.19903
Train Epoch: 18 [210/250 26880/32000 (84%)] Loss: 1.96857 (semantic_loss: 0.01823, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=1.69988
Train Epoch: 18 [221/250 28288/32000 (88%)] Loss: 1.96717 (semantic_loss: 0.01782, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.40585
Train Epoch: 18 [232/250 29696/32000 (93%)] Loss: 1.96718 (semantic_loss: 0.01783, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33986
Train Epoch: 18 [243/250 31104/32000 (97%)] Loss: 1.96765 (semantic_loss: 0.01830, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34444
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch18.pth ...
Done in 4.060s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch18.pth ...
Done in 8.019s
removing stale ckpt [epoch 17] [took 0.00s]
epoch : 18
loss : 1.9689448275566102
learning_rate : 2.0906016760958855e-05
n_samples : 576000
n_steps : 4500
MSRVTT_jsfusion_test/t2v_metrics/R1: 14.9
MSRVTT_jsfusion_test/t2v_metrics/R5: 41.9
MSRVTT_jsfusion_test/t2v_metrics/R10: 56.6
MSRVTT_jsfusion_test/t2v_metrics/R50: 83.6
MSRVTT_jsfusion_test/t2v_metrics/MedR: 8.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 33.386
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 32.814987325882335
MSRVTT_jsfusion_test/v2t_metrics/R1: 16.0
MSRVTT_jsfusion_test/v2t_metrics/R5: 43.7
MSRVTT_jsfusion_test/v2t_metrics/R10: 58.7
MSRVTT_jsfusion_test/v2t_metrics/R50: 84.3
MSRVTT_jsfusion_test/v2t_metrics/MedR: 7.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 31.91
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 34.49423414109665
mnt_best : 32.814987325882335
not_improved_count: 0
Train Epoch: 19 [1/250 128/32000 (0%)] Loss: 1.96740 (semantic_loss: 0.01805, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=28.37734
Train Epoch: 19 [12/250 1536/32000 (5%)] Loss: 1.96892 (semantic_loss: 0.01956, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33215
Train Epoch: 19 [23/250 2944/32000 (9%)] Loss: 1.96863 (semantic_loss: 0.01830, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33674
Train Epoch: 19 [34/250 4352/32000 (14%)] Loss: 1.96831 (semantic_loss: 0.01896, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33743
Train Epoch: 19 [45/250 5760/32000 (18%)] Loss: 1.96874 (semantic_loss: 0.01841, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.35802
Train Epoch: 19 [56/250 7168/32000 (22%)] Loss: 1.96641 (semantic_loss: 0.01705, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33643
Train Epoch: 19 [67/250 8576/32000 (27%)] Loss: 1.96848 (semantic_loss: 0.01815, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=1.83501
Train Epoch: 19 [78/250 9984/32000 (31%)] Loss: 1.96737 (semantic_loss: 0.01704, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34312
Train Epoch: 19 [89/250 11392/32000 (36%)] Loss: 1.96828 (semantic_loss: 0.01795, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33231
Train Epoch: 19 [100/250 12800/32000 (40%)] Loss: 1.96774 (semantic_loss: 0.01838, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32719
Train Epoch: 19 [111/250 14208/32000 (44%)] Loss: 1.96901 (semantic_loss: 0.01868, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33152
Train Epoch: 19 [122/250 15616/32000 (49%)] Loss: 1.96735 (semantic_loss: 0.01702, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32258
Train Epoch: 19 [133/250 17024/32000 (53%)] Loss: 1.96932 (semantic_loss: 0.01899, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33970
Train Epoch: 19 [144/250 18432/32000 (58%)] Loss: 1.96865 (semantic_loss: 0.01930, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32520
Train Epoch: 19 [155/250 19840/32000 (62%)] Loss: 1.96822 (semantic_loss: 0.01887, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.35675
Train Epoch: 19 [166/250 21248/32000 (66%)] Loss: 1.96680 (semantic_loss: 0.01745, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32395
Train Epoch: 19 [177/250 22656/32000 (71%)] Loss: 1.96665 (semantic_loss: 0.01729, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34097
Train Epoch: 19 [188/250 24064/32000 (75%)] Loss: 1.96844 (semantic_loss: 0.01908, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34773
Train Epoch: 19 [199/250 25472/32000 (80%)] Loss: 1.96720 (semantic_loss: 0.01785, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32999
Train Epoch: 19 [210/250 26880/32000 (84%)] Loss: 1.96819 (semantic_loss: 0.01884, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32605
Train Epoch: 19 [221/250 28288/32000 (88%)] Loss: 1.96856 (semantic_loss: 0.01823, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33109
Train Epoch: 19 [232/250 29696/32000 (93%)] Loss: 1.96907 (semantic_loss: 0.01874, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34252
Train Epoch: 19 [243/250 31104/32000 (97%)] Loss: 1.96894 (semantic_loss: 0.01861, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32684
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch19.pth ...
Done in 3.881s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch19.pth ...
Done in 7.878s
removing stale ckpt [epoch 18] [took 0.00s]
epoch : 19
loss : 1.968436700820923
learning_rate : 1.986071592291091e-05
n_samples : 608000
n_steps : 4750
MSRVTT_jsfusion_test/t2v_metrics/R1: 16.3
MSRVTT_jsfusion_test/t2v_metrics/R5: 41.7
MSRVTT_jsfusion_test/t2v_metrics/R10: 55.3
MSRVTT_jsfusion_test/t2v_metrics/R50: 84.4
MSRVTT_jsfusion_test/t2v_metrics/MedR: 8.0
MSRVTT_jsfusion_test/t2v_metrics/MeanR: 32.7385
MSRVTT_jsfusion_test/t2v_metrics/geometric_mean_R1-R5-R10: 33.497798325640446
MSRVTT_jsfusion_test/v2t_metrics/R1: 14.9
MSRVTT_jsfusion_test/v2t_metrics/R5: 43.2
MSRVTT_jsfusion_test/v2t_metrics/R10: 58.4
MSRVTT_jsfusion_test/v2t_metrics/R50: 85.5
MSRVTT_jsfusion_test/v2t_metrics/MedR: 7.0
MSRVTT_jsfusion_test/v2t_metrics/MeanR: 31.44
MSRVTT_jsfusion_test/v2t_metrics/geometric_mean_R1-R5-R10: 33.498674336789286
mnt_best : 33.497798325640446
not_improved_count: 0
Train Epoch: 20 [1/250 128/32000 (0%)] Loss: 1.96757 (semantic_loss: 0.01724, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=26.86768
Train Epoch: 20 [12/250 1536/32000 (5%)] Loss: 1.96709 (semantic_loss: 0.01774, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.33212
Train Epoch: 20 [23/250 2944/32000 (9%)] Loss: 1.97026 (semantic_loss: 0.01994, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33806
Train Epoch: 20 [34/250 4352/32000 (14%)] Loss: 1.96619 (semantic_loss: 0.01684, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32712
Train Epoch: 20 [45/250 5760/32000 (18%)] Loss: 1.96671 (semantic_loss: 0.01638, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.34164
Train Epoch: 20 [56/250 7168/32000 (22%)] Loss: 1.96842 (semantic_loss: 0.01907, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.32400
Train Epoch: 20 [67/250 8576/32000 (27%)] Loss: 1.96867 (semantic_loss: 0.01834, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.88469
Train Epoch: 20 [78/250 9984/32000 (31%)] Loss: 1.96723 (semantic_loss: 0.01788, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=1.77187
Train Epoch: 20 [89/250 11392/32000 (36%)] Loss: 1.96837 (semantic_loss: 0.01804, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.33537
Train Epoch: 20 [100/250 12800/32000 (40%)] Loss: 1.96875 (semantic_loss: 0.01940, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34271
Train Epoch: 20 [111/250 14208/32000 (44%)] Loss: 1.96802 (semantic_loss: 0.01769, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33601
Train Epoch: 20 [122/250 15616/32000 (49%)] Loss: 1.96955 (semantic_loss: 0.01922, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32761
Train Epoch: 20 [133/250 17024/32000 (53%)] Loss: 1.96743 (semantic_loss: 0.01808, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=5.77753
Train Epoch: 20 [144/250 18432/32000 (58%)] Loss: 1.96934 (semantic_loss: 0.01803, quant_loss: 1.95117, bit_balance_loss: 0.00014) batch_time=0.33187
Train Epoch: 20 [155/250 19840/32000 (62%)] Loss: 1.96909 (semantic_loss: 0.01876, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=1.55490
Train Epoch: 20 [166/250 21248/32000 (66%)] Loss: 1.96999 (semantic_loss: 0.01966, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.35033
Train Epoch: 20 [177/250 22656/32000 (71%)] Loss: 1.96648 (semantic_loss: 0.01712, quant_loss: 1.94922, bit_balance_loss: 0.00014) batch_time=0.34092
Train Epoch: 20 [188/250 24064/32000 (75%)] Loss: 1.96820 (semantic_loss: 0.01885, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.32580
Train Epoch: 20 [199/250 25472/32000 (80%)] Loss: 1.96950 (semantic_loss: 0.01918, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.33346
Train Epoch: 20 [210/250 26880/32000 (84%)] Loss: 1.96824 (semantic_loss: 0.01791, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32928
Train Epoch: 20 [221/250 28288/32000 (88%)] Loss: 1.96857 (semantic_loss: 0.01824, quant_loss: 1.95020, bit_balance_loss: 0.00014) batch_time=0.32350
Train Epoch: 20 [232/250 29696/32000 (93%)] Loss: 1.96766 (semantic_loss: 0.01733, quant_loss: 1.95020, bit_balance_loss: 0.00013) batch_time=0.32976
Train Epoch: 20 [243/250 31104/32000 (97%)] Loss: 1.96834 (semantic_loss: 0.01898, quant_loss: 1.94922, bit_balance_loss: 0.00013) batch_time=0.33435
Saving checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch20.pth ...
Done in 4.114s
Updating 'best' checkpoint: /apdcephfs/share_47076/gimwang/HCQ/exps/DCMH_MSRVTT_1kA/checkpoint-epoch20.pth ...
Done in 8.918s
removing stale ckpt [epoch 19] [took 0.00s]
epoch : 20
loss : 1.9683022408485413
learning_rate : 1.8867680126765363e-05
n_samples : 640000
n_steps : 5000
MSRVTT_jsfusion_test/t2v_metrics/R1: 16.2
MSRVTT_jsfusion_test/t2v_metrics/R5: 42.3
MSRVTT_jsfusion_test/t2v_metrics/R10: 56.7
MSRVTT_jsfusion_test/t2v_metrics/R50: 84.5