-
Notifications
You must be signed in to change notification settings - Fork 21
/
Copy pathtwi_logs
1000 lines (959 loc) · 53.3 KB
/
twi_logs
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
[epoch: 1 step: 5] train loss: 1.13864 time: 0:00:02
[epoch: 1 step: 10] train loss: 1.15238 time: 0:00:03
[epoch: 1 step: 15] train loss: 1.12349 time: 0:00:04
[epoch: 1 step: 20] train loss: 1.13857 time: 0:00:05
[epoch: 1 step: 25] train loss: 1.11944 time: 0:00:06
[epoch: 1 step: 30] train loss: 1.12575 time: 0:00:07
[epoch: 1 step: 35] train loss: 1.11639 time: 0:00:08
[epoch: 1 step: 40] train loss: 1.10686 time: 0:00:09
[epoch: 1 step: 45] train loss: 1.1045 time: 0:00:10
[epoch: 1 step: 50] train loss: 1.09135 time: 0:00:11
[epoch: 1 step: 55] train loss: 1.06109 time: 0:00:12
[epoch: 1 step: 60] train loss: 1.07924 time: 0:00:13
[epoch: 1 step: 65] train loss: 1.06343 time: 0:00:14
[epoch: 1 step: 70] train loss: 1.043 time: 0:00:15
[epoch: 1 step: 75] train loss: 1.04289 time: 0:00:16
[epoch: 1 step: 80] train loss: 1.04709 time: 0:00:18
[epoch: 1 step: 85] train loss: 1.04227 time: 0:00:19
[epoch: 1 step: 90] train loss: 1.00301 time: 0:00:20
[epoch: 1 step: 95] train loss: 1.05562 time: 0:00:21
[epoch: 1 step: 100] train loss: 1.02691 time: 0:00:22
Evaluate data in 0.98 seconds!
Evaluation on dev at Epoch 1/40. Step:100/4000:
AccuracyMetric: acc=0.5
ClassifyFPreRecMetric: f=0.222222, pre=0.166667, rec=0.333333
[epoch: 2 step: 105] train loss: 1.00201 time: 0:00:24
[epoch: 2 step: 110] train loss: 1.07761 time: 0:00:25
[epoch: 2 step: 115] train loss: 1.00473 time: 0:00:26
[epoch: 2 step: 120] train loss: 0.975775 time: 0:00:27
[epoch: 2 step: 125] train loss: 1.00496 time: 0:00:28
[epoch: 2 step: 130] train loss: 1.00688 time: 0:00:30
[epoch: 2 step: 135] train loss: 0.945054 time: 0:00:31
[epoch: 2 step: 140] train loss: 0.938269 time: 0:00:32
[epoch: 2 step: 145] train loss: 1.03178 time: 0:00:33
[epoch: 2 step: 150] train loss: 0.982394 time: 0:00:34
[epoch: 2 step: 155] train loss: 0.989195 time: 0:00:35
[epoch: 2 step: 160] train loss: 0.995123 time: 0:00:36
[epoch: 2 step: 165] train loss: 0.962398 time: 0:00:37
[epoch: 2 step: 170] train loss: 0.976378 time: 0:00:38
[epoch: 2 step: 175] train loss: 0.918939 time: 0:00:39
[epoch: 2 step: 180] train loss: 0.861031 time: 0:00:40
[epoch: 2 step: 185] train loss: 0.865126 time: 0:00:41
[epoch: 2 step: 190] train loss: 0.790967 time: 0:00:42
[epoch: 2 step: 195] train loss: 0.844504 time: 0:00:44
[epoch: 2 step: 200] train loss: 0.804385 time: 0:00:45
Evaluate data in 0.98 seconds!
Evaluation on dev at Epoch 2/40. Step:200/4000:
AccuracyMetric: acc=0.657514
ClassifyFPreRecMetric: f=0.612687, pre=0.716308, rec=0.589595
[epoch: 3 step: 205] train loss: 0.767893 time: 0:00:47
[epoch: 3 step: 210] train loss: 0.758124 time: 0:00:49
[epoch: 3 step: 215] train loss: 0.733636 time: 0:00:50
[epoch: 3 step: 220] train loss: 0.72308 time: 0:00:51
[epoch: 3 step: 225] train loss: 0.699978 time: 0:00:52
[epoch: 3 step: 230] train loss: 0.735783 time: 0:00:53
[epoch: 3 step: 235] train loss: 0.699985 time: 0:00:54
[epoch: 3 step: 240] train loss: 0.691995 time: 0:00:55
[epoch: 3 step: 245] train loss: 0.724457 time: 0:00:57
[epoch: 3 step: 250] train loss: 0.69061 time: 0:00:58
[epoch: 3 step: 255] train loss: 0.740627 time: 0:00:59
[epoch: 3 step: 260] train loss: 0.692188 time: 0:01:00
[epoch: 3 step: 265] train loss: 0.631286 time: 0:01:01
[epoch: 3 step: 270] train loss: 0.717986 time: 0:01:02
[epoch: 3 step: 275] train loss: 0.665581 time: 0:01:03
[epoch: 3 step: 280] train loss: 0.657309 time: 0:01:04
[epoch: 3 step: 285] train loss: 0.660009 time: 0:01:06
[epoch: 3 step: 290] train loss: 0.564226 time: 0:01:07
[epoch: 3 step: 295] train loss: 0.636999 time: 0:01:08
Evaluate data in 1.05 seconds!
Evaluation on dev at Epoch 3/40. Step:299/4000:
AccuracyMetric: acc=0.728324
ClassifyFPreRecMetric: f=0.724915, pre=0.722091, rec=0.732177
[epoch: 4 step: 300] train loss: 0.658039 time: 0:01:11
[epoch: 4 step: 305] train loss: 0.58347 time: 0:01:12
[epoch: 4 step: 310] train loss: 0.47611 time: 0:01:13
[epoch: 4 step: 315] train loss: 0.547356 time: 0:01:14
[epoch: 4 step: 320] train loss: 0.550639 time: 0:01:15
[epoch: 4 step: 325] train loss: 0.629732 time: 0:01:16
[epoch: 4 step: 330] train loss: 0.601335 time: 0:01:17
[epoch: 4 step: 335] train loss: 0.639741 time: 0:01:18
[epoch: 4 step: 340] train loss: 0.521111 time: 0:01:19
[epoch: 4 step: 345] train loss: 0.61345 time: 0:01:20
[epoch: 4 step: 350] train loss: 0.502645 time: 0:01:22
[epoch: 4 step: 355] train loss: 0.567266 time: 0:01:23
[epoch: 4 step: 360] train loss: 0.601676 time: 0:01:24
[epoch: 4 step: 365] train loss: 0.587274 time: 0:01:25
[epoch: 4 step: 370] train loss: 0.543396 time: 0:01:26
[epoch: 4 step: 375] train loss: 0.595919 time: 0:01:27
[epoch: 4 step: 380] train loss: 0.583232 time: 0:01:28
[epoch: 4 step: 385] train loss: 0.589684 time: 0:01:29
[epoch: 4 step: 390] train loss: 0.465112 time: 0:01:30
[epoch: 4 step: 395] train loss: 0.600252 time: 0:01:31
Evaluate data in 1.0 seconds!
Evaluation on dev at Epoch 4/40. Step:398/4000:
AccuracyMetric: acc=0.75
ClassifyFPreRecMetric: f=0.7368, pre=0.743676, rec=0.731214
[epoch: 5 step: 400] train loss: 0.512496 time: 0:01:34
[epoch: 5 step: 405] train loss: 0.440739 time: 0:01:35
[epoch: 5 step: 410] train loss: 0.501291 time: 0:01:36
[epoch: 5 step: 415] train loss: 0.456617 time: 0:01:38
[epoch: 5 step: 420] train loss: 0.396239 time: 0:01:39
[epoch: 5 step: 425] train loss: 0.423322 time: 0:01:40
[epoch: 5 step: 430] train loss: 0.434894 time: 0:01:41
[epoch: 5 step: 435] train loss: 0.590604 time: 0:01:42
[epoch: 5 step: 440] train loss: 0.535231 time: 0:01:43
[epoch: 5 step: 445] train loss: 0.493281 time: 0:01:44
[epoch: 5 step: 450] train loss: 0.50266 time: 0:01:45
[epoch: 5 step: 455] train loss: 0.3946 time: 0:01:46
[epoch: 5 step: 460] train loss: 0.417611 time: 0:01:47
[epoch: 5 step: 465] train loss: 0.503198 time: 0:01:48
[epoch: 5 step: 470] train loss: 0.525553 time: 0:01:49
[epoch: 5 step: 475] train loss: 0.551626 time: 0:01:51
[epoch: 5 step: 480] train loss: 0.498308 time: 0:01:52
[epoch: 5 step: 485] train loss: 0.461482 time: 0:01:53
[epoch: 5 step: 490] train loss: 0.428833 time: 0:01:54
[epoch: 5 step: 495] train loss: 0.346066 time: 0:01:55
Evaluate data in 1.05 seconds!
Evaluation on dev at Epoch 5/40. Step:497/4000:
AccuracyMetric: acc=0.754335
ClassifyFPreRecMetric: f=0.743216, pre=0.742073, rec=0.746628
[epoch: 6 step: 500] train loss: 0.33558 time: 0:01:58
[epoch: 6 step: 505] train loss: 0.438859 time: 0:01:59
[epoch: 6 step: 510] train loss: 0.354131 time: 0:02:00
[epoch: 6 step: 515] train loss: 0.312225 time: 0:02:01
[epoch: 6 step: 520] train loss: 0.29336 time: 0:02:02
[epoch: 6 step: 525] train loss: 0.332068 time: 0:02:03
[epoch: 6 step: 530] train loss: 0.379834 time: 0:02:04
[epoch: 6 step: 535] train loss: 0.35035 time: 0:02:05
[epoch: 6 step: 540] train loss: 0.261282 time: 0:02:06
[epoch: 6 step: 545] train loss: 0.332007 time: 0:02:08
[epoch: 6 step: 550] train loss: 0.234517 time: 0:02:09
[epoch: 6 step: 555] train loss: 0.426128 time: 0:02:10
[epoch: 6 step: 560] train loss: 0.391886 time: 0:02:11
[epoch: 6 step: 565] train loss: 0.35235 time: 0:02:12
[epoch: 6 step: 570] train loss: 0.318142 time: 0:02:13
[epoch: 6 step: 575] train loss: 0.381427 time: 0:02:14
[epoch: 6 step: 580] train loss: 0.378412 time: 0:02:15
[epoch: 6 step: 585] train loss: 0.44842 time: 0:02:16
[epoch: 6 step: 590] train loss: 0.352888 time: 0:02:17
[epoch: 6 step: 595] train loss: 0.4177 time: 0:02:18
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 6/40. Step:597/4000:
AccuracyMetric: acc=0.736994
ClassifyFPreRecMetric: f=0.725998, pre=0.733863, rec=0.72158
[epoch: 7 step: 600] train loss: 0.333546 time: 0:02:21
[epoch: 7 step: 605] train loss: 0.227657 time: 0:02:22
[epoch: 7 step: 610] train loss: 0.230945 time: 0:02:23
[epoch: 7 step: 615] train loss: 0.286516 time: 0:02:24
[epoch: 7 step: 620] train loss: 0.206006 time: 0:02:25
[epoch: 7 step: 625] train loss: 0.288706 time: 0:02:26
[epoch: 7 step: 630] train loss: 0.286619 time: 0:02:27
[epoch: 7 step: 635] train loss: 0.285225 time: 0:02:28
[epoch: 7 step: 640] train loss: 0.287965 time: 0:02:29
[epoch: 7 step: 645] train loss: 0.357559 time: 0:02:30
[epoch: 7 step: 650] train loss: 0.346628 time: 0:02:32
[epoch: 7 step: 655] train loss: 0.303289 time: 0:02:33
[epoch: 7 step: 660] train loss: 0.313846 time: 0:02:34
[epoch: 7 step: 665] train loss: 0.30264 time: 0:02:35
[epoch: 7 step: 670] train loss: 0.242804 time: 0:02:36
[epoch: 7 step: 675] train loss: 0.268508 time: 0:02:37
[epoch: 7 step: 680] train loss: 0.249239 time: 0:02:38
[epoch: 7 step: 685] train loss: 0.310368 time: 0:02:39
[epoch: 7 step: 690] train loss: 0.392958 time: 0:02:40
[epoch: 7 step: 695] train loss: 0.269223 time: 0:02:41
Evaluate data in 1.0 seconds!
Evaluation on dev at Epoch 7/40. Step:696/4000:
AccuracyMetric: acc=0.751445
ClassifyFPreRecMetric: f=0.732612, pre=0.760121, rec=0.7158
[epoch: 8 step: 700] train loss: 0.164022 time: 0:02:44
[epoch: 8 step: 705] train loss: 0.193238 time: 0:02:45
[epoch: 8 step: 710] train loss: 0.167269 time: 0:02:46
[epoch: 8 step: 715] train loss: 0.223865 time: 0:02:47
[epoch: 8 step: 720] train loss: 0.173995 time: 0:02:48
[epoch: 8 step: 725] train loss: 0.224364 time: 0:02:49
[epoch: 8 step: 730] train loss: 0.15605 time: 0:02:50
[epoch: 8 step: 735] train loss: 0.119168 time: 0:02:52
[epoch: 8 step: 740] train loss: 0.22077 time: 0:02:53
[epoch: 8 step: 745] train loss: 0.109266 time: 0:02:54
[epoch: 8 step: 750] train loss: 0.199142 time: 0:02:55
[epoch: 8 step: 755] train loss: 0.232091 time: 0:02:56
[epoch: 8 step: 760] train loss: 0.243842 time: 0:02:57
[epoch: 8 step: 765] train loss: 0.242281 time: 0:02:58
[epoch: 8 step: 770] train loss: 0.196771 time: 0:02:59
[epoch: 8 step: 775] train loss: 0.279491 time: 0:03:00
[epoch: 8 step: 780] train loss: 0.208339 time: 0:03:01
[epoch: 8 step: 785] train loss: 0.174755 time: 0:03:03
[epoch: 8 step: 790] train loss: 0.246353 time: 0:03:04
[epoch: 8 step: 795] train loss: 0.238257 time: 0:03:05
Evaluate data in 1.04 seconds!
Evaluation on dev at Epoch 8/40. Step:796/4000:
AccuracyMetric: acc=0.734104
ClassifyFPreRecMetric: f=0.716099, pre=0.739178, rec=0.701349
[epoch: 9 step: 800] train loss: 0.149851 time: 0:03:07
[epoch: 9 step: 805] train loss: 0.159235 time: 0:03:08
[epoch: 9 step: 810] train loss: 0.0678576 time: 0:03:09
[epoch: 9 step: 815] train loss: 0.156339 time: 0:03:10
[epoch: 9 step: 820] train loss: 0.161922 time: 0:03:12
[epoch: 9 step: 825] train loss: 0.12394 time: 0:03:13
[epoch: 9 step: 830] train loss: 0.1091 time: 0:03:14
[epoch: 9 step: 835] train loss: 0.138638 time: 0:03:15
[epoch: 9 step: 840] train loss: 0.160729 time: 0:03:16
[epoch: 9 step: 845] train loss: 0.261103 time: 0:03:17
[epoch: 9 step: 850] train loss: 0.177083 time: 0:03:18
[epoch: 9 step: 855] train loss: 0.182482 time: 0:03:19
[epoch: 9 step: 860] train loss: 0.16409 time: 0:03:20
[epoch: 9 step: 865] train loss: 0.0985997 time: 0:03:21
[epoch: 9 step: 870] train loss: 0.157505 time: 0:03:22
[epoch: 9 step: 875] train loss: 0.153448 time: 0:03:24
[epoch: 9 step: 880] train loss: 0.13917 time: 0:03:25
[epoch: 9 step: 885] train loss: 0.185474 time: 0:03:26
[epoch: 9 step: 890] train loss: 0.203157 time: 0:03:27
[epoch: 9 step: 895] train loss: 0.108233 time: 0:03:28
Evaluate data in 1.0 seconds!
Evaluation on dev at Epoch 9/40. Step:895/4000:
AccuracyMetric: acc=0.74422
ClassifyFPreRecMetric: f=0.735337, pre=0.731344, rec=0.740848
[epoch: 10 step: 900] train loss: 0.0632544 time: 0:03:30
[epoch: 10 step: 905] train loss: 0.0999093 time: 0:03:31
[epoch: 10 step: 910] train loss: 0.105881 time: 0:03:32
[epoch: 10 step: 915] train loss: 0.114817 time: 0:03:33
[epoch: 10 step: 920] train loss: 0.161772 time: 0:03:34
[epoch: 10 step: 925] train loss: 0.114378 time: 0:03:36
[epoch: 10 step: 930] train loss: 0.173446 time: 0:03:37
[epoch: 10 step: 935] train loss: 0.0641023 time: 0:03:38
[epoch: 10 step: 940] train loss: 0.0694527 time: 0:03:39
[epoch: 10 step: 945] train loss: 0.102927 time: 0:03:40
[epoch: 10 step: 950] train loss: 0.0901521 time: 0:03:41
[epoch: 10 step: 955] train loss: 0.177204 time: 0:03:42
[epoch: 10 step: 960] train loss: 0.132496 time: 0:03:43
[epoch: 10 step: 965] train loss: 0.155208 time: 0:03:44
[epoch: 10 step: 970] train loss: 0.12102 time: 0:03:45
[epoch: 10 step: 975] train loss: 0.163838 time: 0:03:47
[epoch: 10 step: 980] train loss: 0.150052 time: 0:03:48
[epoch: 10 step: 985] train loss: 0.131243 time: 0:03:49
[epoch: 10 step: 990] train loss: 0.0953849 time: 0:03:50
[epoch: 10 step: 995] train loss: 0.106316 time: 0:03:51
Evaluate data in 0.98 seconds!
Evaluation on dev at Epoch 10/40. Step:995/4000:
AccuracyMetric: acc=0.758671
ClassifyFPreRecMetric: f=0.743407, pre=0.759272, rec=0.732177
[epoch: 11 step: 1000] train loss: 0.0945375 time: 0:03:54
[epoch: 11 step: 1005] train loss: 0.0946657 time: 0:03:55
[epoch: 11 step: 1010] train loss: 0.0703632 time: 0:03:56
[epoch: 11 step: 1015] train loss: 0.0530382 time: 0:03:57
[epoch: 11 step: 1020] train loss: 0.0842621 time: 0:03:58
[epoch: 11 step: 1025] train loss: 0.0958701 time: 0:03:59
[epoch: 11 step: 1030] train loss: 0.0749797 time: 0:04:00
[epoch: 11 step: 1035] train loss: 0.0801487 time: 0:04:02
[epoch: 11 step: 1040] train loss: 0.0785323 time: 0:04:03
[epoch: 11 step: 1045] train loss: 0.0904252 time: 0:04:04
[epoch: 11 step: 1050] train loss: 0.115877 time: 0:04:05
[epoch: 11 step: 1055] train loss: 0.0884827 time: 0:04:06
[epoch: 11 step: 1060] train loss: 0.0897125 time: 0:04:07
[epoch: 11 step: 1065] train loss: 0.0611966 time: 0:04:08
[epoch: 11 step: 1070] train loss: 0.176501 time: 0:04:09
[epoch: 11 step: 1075] train loss: 0.10571 time: 0:04:10
[epoch: 11 step: 1080] train loss: 0.0845378 time: 0:04:12
[epoch: 11 step: 1085] train loss: 0.12642 time: 0:04:13
[epoch: 11 step: 1090] train loss: 0.051064 time: 0:04:14
Evaluate data in 1.06 seconds!
Evaluation on dev at Epoch 11/40. Step:1094/4000:
AccuracyMetric: acc=0.765896
ClassifyFPreRecMetric: f=0.756273, pre=0.75236, rec=0.761079
[epoch: 12 step: 1095] train loss: 0.113942 time: 0:04:17
[epoch: 12 step: 1100] train loss: 0.062657 time: 0:04:18
[epoch: 12 step: 1105] train loss: 0.0711043 time: 0:04:19
[epoch: 12 step: 1110] train loss: 0.0378352 time: 0:04:20
[epoch: 12 step: 1115] train loss: 0.0709116 time: 0:04:21
[epoch: 12 step: 1120] train loss: 0.086386 time: 0:04:22
[epoch: 12 step: 1125] train loss: 0.0641496 time: 0:04:23
[epoch: 12 step: 1130] train loss: 0.0783225 time: 0:04:24
[epoch: 12 step: 1135] train loss: 0.0443819 time: 0:04:26
[epoch: 12 step: 1140] train loss: 0.0454565 time: 0:04:27
[epoch: 12 step: 1145] train loss: 0.107261 time: 0:04:28
[epoch: 12 step: 1150] train loss: 0.0486325 time: 0:04:29
[epoch: 12 step: 1155] train loss: 0.0911526 time: 0:04:30
[epoch: 12 step: 1160] train loss: 0.052619 time: 0:04:31
[epoch: 12 step: 1165] train loss: 0.0375513 time: 0:04:32
[epoch: 12 step: 1170] train loss: 0.0638773 time: 0:04:33
[epoch: 12 step: 1175] train loss: 0.0573005 time: 0:04:34
[epoch: 12 step: 1180] train loss: 0.156544 time: 0:04:36
[epoch: 12 step: 1185] train loss: 0.0720308 time: 0:04:37
[epoch: 12 step: 1190] train loss: 0.0324415 time: 0:04:38
Evaluate data in 1.03 seconds!
Evaluation on dev at Epoch 12/40. Step:1193/4000:
AccuracyMetric: acc=0.741329
ClassifyFPreRecMetric: f=0.734705, pre=0.730566, rec=0.740848
[epoch: 13 step: 1195] train loss: 0.0637136 time: 0:04:40
[epoch: 13 step: 1200] train loss: 0.0468718 time: 0:04:41
[epoch: 13 step: 1205] train loss: 0.142961 time: 0:04:42
[epoch: 13 step: 1210] train loss: 0.0737235 time: 0:04:43
[epoch: 13 step: 1215] train loss: 0.129606 time: 0:04:45
[epoch: 13 step: 1220] train loss: 0.0831341 time: 0:04:46
[epoch: 13 step: 1225] train loss: 0.0674375 time: 0:04:47
[epoch: 13 step: 1230] train loss: 0.052384 time: 0:04:48
[epoch: 13 step: 1235] train loss: 0.0586535 time: 0:04:49
[epoch: 13 step: 1240] train loss: 0.0782512 time: 0:04:50
[epoch: 13 step: 1245] train loss: 0.0695851 time: 0:04:51
[epoch: 13 step: 1250] train loss: 0.0776924 time: 0:04:52
[epoch: 13 step: 1255] train loss: 0.103773 time: 0:04:53
[epoch: 13 step: 1260] train loss: 0.0307629 time: 0:04:55
[epoch: 13 step: 1265] train loss: 0.0612557 time: 0:04:56
[epoch: 13 step: 1270] train loss: 0.0921784 time: 0:04:57
[epoch: 13 step: 1275] train loss: 0.0290776 time: 0:04:58
[epoch: 13 step: 1280] train loss: 0.0591548 time: 0:04:59
[epoch: 13 step: 1285] train loss: 0.112332 time: 0:05:00
[epoch: 13 step: 1290] train loss: 0.0503268 time: 0:05:01
Evaluate data in 0.99 seconds!
Evaluation on dev at Epoch 13/40. Step:1293/4000:
AccuracyMetric: acc=0.75
ClassifyFPreRecMetric: f=0.738832, pre=0.743477, rec=0.735067
[epoch: 14 step: 1295] train loss: 0.048085 time: 0:05:04
[epoch: 14 step: 1300] train loss: 0.0601558 time: 0:05:05
[epoch: 14 step: 1305] train loss: 0.030848 time: 0:05:06
[epoch: 14 step: 1310] train loss: 0.0740223 time: 0:05:07
[epoch: 14 step: 1315] train loss: 0.0716093 time: 0:05:08
[epoch: 14 step: 1320] train loss: 0.06616 time: 0:05:09
[epoch: 14 step: 1325] train loss: 0.0507408 time: 0:05:10
[epoch: 14 step: 1330] train loss: 0.0568515 time: 0:05:11
[epoch: 14 step: 1335] train loss: 0.0230883 time: 0:05:12
[epoch: 14 step: 1340] train loss: 0.0409247 time: 0:05:13
[epoch: 14 step: 1345] train loss: 0.0240557 time: 0:05:15
[epoch: 14 step: 1350] train loss: 0.0651398 time: 0:05:16
[epoch: 14 step: 1355] train loss: 0.13528 time: 0:05:17
[epoch: 14 step: 1360] train loss: 0.0693462 time: 0:05:18
[epoch: 14 step: 1365] train loss: 0.0482034 time: 0:05:19
[epoch: 14 step: 1370] train loss: 0.0704539 time: 0:05:20
[epoch: 14 step: 1375] train loss: 0.0403976 time: 0:05:21
[epoch: 14 step: 1380] train loss: 0.0616848 time: 0:05:23
[epoch: 14 step: 1385] train loss: 0.0308961 time: 0:05:24
[epoch: 14 step: 1390] train loss: 0.0790122 time: 0:05:25
Evaluate data in 1.02 seconds!
Evaluation on dev at Epoch 14/40. Step:1393/4000:
AccuracyMetric: acc=0.764451
ClassifyFPreRecMetric: f=0.756292, pre=0.759579, rec=0.755299
[epoch: 15 step: 1395] train loss: 0.0938962 time: 0:05:28
[epoch: 15 step: 1400] train loss: 0.0654544 time: 0:05:29
[epoch: 15 step: 1405] train loss: 0.0528128 time: 0:05:30
[epoch: 15 step: 1410] train loss: 0.0573165 time: 0:05:31
[epoch: 15 step: 1415] train loss: 0.0436426 time: 0:05:33
[epoch: 15 step: 1420] train loss: 0.0493191 time: 0:05:34
[epoch: 15 step: 1425] train loss: 0.0938521 time: 0:05:35
[epoch: 15 step: 1430] train loss: 0.0466928 time: 0:05:36
[epoch: 15 step: 1435] train loss: 0.0605802 time: 0:05:37
[epoch: 15 step: 1440] train loss: 0.0497639 time: 0:05:38
[epoch: 15 step: 1445] train loss: 0.0685136 time: 0:05:40
[epoch: 15 step: 1450] train loss: 0.0669511 time: 0:05:41
[epoch: 15 step: 1455] train loss: 0.0835088 time: 0:05:42
[epoch: 15 step: 1460] train loss: 0.0556267 time: 0:05:43
[epoch: 15 step: 1465] train loss: 0.05363 time: 0:05:44
[epoch: 15 step: 1470] train loss: 0.0349031 time: 0:05:45
[epoch: 15 step: 1475] train loss: 0.0324451 time: 0:05:46
[epoch: 15 step: 1480] train loss: 0.068616 time: 0:05:47
[epoch: 15 step: 1485] train loss: 0.0775718 time: 0:05:49
[epoch: 15 step: 1490] train loss: 0.0606611 time: 0:05:50
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 15/40. Step:1492/4000:
AccuracyMetric: acc=0.774566
ClassifyFPreRecMetric: f=0.768735, pre=0.765939, rec=0.77553
[epoch: 16 step: 1495] train loss: 0.0106068 time: 0:05:53
[epoch: 16 step: 1500] train loss: 0.0361037 time: 0:05:54
[epoch: 16 step: 1505] train loss: 0.0344819 time: 0:05:55
[epoch: 16 step: 1510] train loss: 0.034899 time: 0:05:56
[epoch: 16 step: 1515] train loss: 0.056474 time: 0:05:57
[epoch: 16 step: 1520] train loss: 0.0241321 time: 0:05:58
[epoch: 16 step: 1525] train loss: 0.0480917 time: 0:05:59
[epoch: 16 step: 1530] train loss: 0.0296029 time: 0:06:00
[epoch: 16 step: 1535] train loss: 0.0593281 time: 0:06:02
[epoch: 16 step: 1540] train loss: 0.0787024 time: 0:06:03
[epoch: 16 step: 1545] train loss: 0.0309458 time: 0:06:04
[epoch: 16 step: 1550] train loss: 0.0335981 time: 0:06:05
[epoch: 16 step: 1555] train loss: 0.0326619 time: 0:06:06
[epoch: 16 step: 1560] train loss: 0.0324079 time: 0:06:07
[epoch: 16 step: 1565] train loss: 0.0398079 time: 0:06:08
[epoch: 16 step: 1570] train loss: 0.0425405 time: 0:06:09
[epoch: 16 step: 1575] train loss: 0.0184658 time: 0:06:11
[epoch: 16 step: 1580] train loss: 0.0162482 time: 0:06:12
[epoch: 16 step: 1585] train loss: 0.0195364 time: 0:06:13
[epoch: 16 step: 1590] train loss: 0.0611992 time: 0:06:14
Evaluate data in 1.01 seconds!
Evaluation on dev at Epoch 16/40. Step:1591/4000:
AccuracyMetric: acc=0.768786
ClassifyFPreRecMetric: f=0.754859, pre=0.765839, rec=0.746628
[epoch: 17 step: 1595] train loss: 0.0352147 time: 0:06:16
[epoch: 17 step: 1600] train loss: 0.00769891 time: 0:06:17
[epoch: 17 step: 1605] train loss: 0.0271968 time: 0:06:19
[epoch: 17 step: 1610] train loss: 0.0380527 time: 0:06:20
[epoch: 17 step: 1615] train loss: 0.0324014 time: 0:06:21
[epoch: 17 step: 1620] train loss: 0.0159971 time: 0:06:22
[epoch: 17 step: 1625] train loss: 0.0370239 time: 0:06:23
[epoch: 17 step: 1630] train loss: 0.0575973 time: 0:06:24
[epoch: 17 step: 1635] train loss: 0.0246751 time: 0:06:25
[epoch: 17 step: 1640] train loss: 0.0352842 time: 0:06:26
[epoch: 17 step: 1645] train loss: 0.0189498 time: 0:06:28
[epoch: 17 step: 1650] train loss: 0.0611896 time: 0:06:29
[epoch: 17 step: 1655] train loss: 0.0299268 time: 0:06:30
[epoch: 17 step: 1660] train loss: 0.0524472 time: 0:06:31
[epoch: 17 step: 1665] train loss: 0.0734842 time: 0:06:32
[epoch: 17 step: 1670] train loss: 0.0852396 time: 0:06:33
[epoch: 17 step: 1675] train loss: 0.0444644 time: 0:06:34
[epoch: 17 step: 1680] train loss: 0.0583292 time: 0:06:35
[epoch: 17 step: 1685] train loss: 0.0253034 time: 0:06:36
[epoch: 17 step: 1690] train loss: 0.0210962 time: 0:06:38
Evaluate data in 1.02 seconds!
Evaluation on dev at Epoch 17/40. Step:1690/4000:
AccuracyMetric: acc=0.760116
ClassifyFPreRecMetric: f=0.749788, pre=0.755484, rec=0.745665
[epoch: 18 step: 1695] train loss: 0.0202321 time: 0:06:40
[epoch: 18 step: 1700] train loss: 0.0481104 time: 0:06:41
[epoch: 18 step: 1705] train loss: 0.0182659 time: 0:06:42
[epoch: 18 step: 1710] train loss: 0.0662594 time: 0:06:43
[epoch: 18 step: 1715] train loss: 0.0315656 time: 0:06:44
[epoch: 18 step: 1720] train loss: 0.0437782 time: 0:06:45
[epoch: 18 step: 1725] train loss: 0.0402139 time: 0:06:46
[epoch: 18 step: 1730] train loss: 0.0396788 time: 0:06:48
[epoch: 18 step: 1735] train loss: 0.0169144 time: 0:06:49
[epoch: 18 step: 1740] train loss: 0.0802155 time: 0:06:50
[epoch: 18 step: 1745] train loss: 0.0538738 time: 0:06:51
[epoch: 18 step: 1750] train loss: 0.030575 time: 0:06:52
[epoch: 18 step: 1755] train loss: 0.0597589 time: 0:06:53
[epoch: 18 step: 1760] train loss: 0.0508305 time: 0:06:54
[epoch: 18 step: 1765] train loss: 0.0440908 time: 0:06:55
[epoch: 18 step: 1770] train loss: 0.0373899 time: 0:06:56
[epoch: 18 step: 1775] train loss: 0.055136 time: 0:06:58
[epoch: 18 step: 1780] train loss: 0.0593063 time: 0:06:59
[epoch: 18 step: 1785] train loss: 0.0372199 time: 0:07:00
Evaluate data in 1.03 seconds!
Evaluation on dev at Epoch 18/40. Step:1789/4000:
AccuracyMetric: acc=0.754335
ClassifyFPreRecMetric: f=0.746169, pre=0.750339, rec=0.742775
[epoch: 19 step: 1790] train loss: 0.0136426 time: 0:07:02
[epoch: 19 step: 1795] train loss: 0.0481108 time: 0:07:03
[epoch: 19 step: 1800] train loss: 0.043816 time: 0:07:04
[epoch: 19 step: 1805] train loss: 0.0609251 time: 0:07:06
[epoch: 19 step: 1810] train loss: 0.0593326 time: 0:07:07
[epoch: 19 step: 1815] train loss: 0.0427309 time: 0:07:08
[epoch: 19 step: 1820] train loss: 0.0482327 time: 0:07:09
[epoch: 19 step: 1825] train loss: 0.090029 time: 0:07:10
[epoch: 19 step: 1830] train loss: 0.0251999 time: 0:07:11
[epoch: 19 step: 1835] train loss: 0.053522 time: 0:07:12
[epoch: 19 step: 1840] train loss: 0.0272572 time: 0:07:13
[epoch: 19 step: 1845] train loss: 0.0509301 time: 0:07:14
[epoch: 19 step: 1850] train loss: 0.038865 time: 0:07:15
[epoch: 19 step: 1855] train loss: 0.0769282 time: 0:07:16
[epoch: 19 step: 1860] train loss: 0.054126 time: 0:07:17
[epoch: 19 step: 1865] train loss: 0.0375402 time: 0:07:19
[epoch: 19 step: 1870] train loss: 0.0436563 time: 0:07:20
[epoch: 19 step: 1875] train loss: 0.0453058 time: 0:07:21
[epoch: 19 step: 1880] train loss: 0.0352668 time: 0:07:22
[epoch: 19 step: 1885] train loss: 0.0468595 time: 0:07:23
Evaluate data in 1.04 seconds!
Evaluation on dev at Epoch 19/40. Step:1888/4000:
AccuracyMetric: acc=0.764451
ClassifyFPreRecMetric: f=0.75649, pre=0.752517, rec=0.761079
[epoch: 20 step: 1890] train loss: 0.0179826 time: 0:07:25
[epoch: 20 step: 1895] train loss: 0.0555838 time: 0:07:26
[epoch: 20 step: 1900] train loss: 0.00891653 time: 0:07:27
[epoch: 20 step: 1905] train loss: 0.0311756 time: 0:07:29
[epoch: 20 step: 1910] train loss: 0.0181034 time: 0:07:30
[epoch: 20 step: 1915] train loss: 0.0578097 time: 0:07:31
[epoch: 20 step: 1920] train loss: 0.0157227 time: 0:07:32
[epoch: 20 step: 1925] train loss: 0.0392162 time: 0:07:33
[epoch: 20 step: 1930] train loss: 0.0513587 time: 0:07:34
[epoch: 20 step: 1935] train loss: 0.038797 time: 0:07:35
[epoch: 20 step: 1940] train loss: 0.0537462 time: 0:07:36
[epoch: 20 step: 1945] train loss: 0.0660937 time: 0:07:37
[epoch: 20 step: 1950] train loss: 0.0679525 time: 0:07:38
[epoch: 20 step: 1955] train loss: 0.0156045 time: 0:07:40
[epoch: 20 step: 1960] train loss: 0.0264108 time: 0:07:41
[epoch: 20 step: 1965] train loss: 0.034072 time: 0:07:42
[epoch: 20 step: 1970] train loss: 0.0524176 time: 0:07:43
[epoch: 20 step: 1975] train loss: 0.0164103 time: 0:07:44
[epoch: 20 step: 1980] train loss: 0.028831 time: 0:07:45
[epoch: 20 step: 1985] train loss: 0.0115804 time: 0:07:46
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 20/40. Step:1987/4000:
AccuracyMetric: acc=0.757225
ClassifyFPreRecMetric: f=0.748988, pre=0.759065, rec=0.743738
[epoch: 21 step: 1990] train loss: 0.0742115 time: 0:07:49
[epoch: 21 step: 1995] train loss: 0.0357113 time: 0:07:50
[epoch: 21 step: 2000] train loss: 0.0779999 time: 0:07:51
[epoch: 21 step: 2005] train loss: 0.0511898 time: 0:07:52
[epoch: 21 step: 2010] train loss: 0.0709834 time: 0:07:53
[epoch: 21 step: 2015] train loss: 0.0723567 time: 0:07:54
[epoch: 21 step: 2020] train loss: 0.0472189 time: 0:07:55
[epoch: 21 step: 2025] train loss: 0.0251427 time: 0:07:56
[epoch: 21 step: 2030] train loss: 0.0365861 time: 0:07:57
[epoch: 21 step: 2035] train loss: 0.0496484 time: 0:07:58
[epoch: 21 step: 2040] train loss: 0.0170396 time: 0:08:00
[epoch: 21 step: 2045] train loss: 0.0396211 time: 0:08:01
[epoch: 21 step: 2050] train loss: 0.0382273 time: 0:08:02
[epoch: 21 step: 2055] train loss: 0.0251051 time: 0:08:03
[epoch: 21 step: 2060] train loss: 0.0266952 time: 0:08:04
[epoch: 21 step: 2065] train loss: 0.0458569 time: 0:08:05
[epoch: 21 step: 2070] train loss: 0.0526841 time: 0:08:06
[epoch: 21 step: 2075] train loss: 0.0566048 time: 0:08:07
[epoch: 21 step: 2080] train loss: 0.0500018 time: 0:08:08
[epoch: 21 step: 2085] train loss: 0.0172405 time: 0:08:09
Evaluate data in 1.03 seconds!
Evaluation on dev at Epoch 21/40. Step:2086/4000:
AccuracyMetric: acc=0.74711
ClassifyFPreRecMetric: f=0.742678, pre=0.735304, rec=0.753372
[epoch: 22 step: 2090] train loss: 0.061325 time: 0:08:12
[epoch: 22 step: 2095] train loss: 0.0336349 time: 0:08:13
[epoch: 22 step: 2100] train loss: 0.0625857 time: 0:08:14
[epoch: 22 step: 2105] train loss: 0.0527022 time: 0:08:15
[epoch: 22 step: 2110] train loss: 0.0350976 time: 0:08:16
[epoch: 22 step: 2115] train loss: 0.0296317 time: 0:08:17
[epoch: 22 step: 2120] train loss: 0.00859019 time: 0:08:18
[epoch: 22 step: 2125] train loss: 0.026802 time: 0:08:19
[epoch: 22 step: 2130] train loss: 0.0581335 time: 0:08:20
[epoch: 22 step: 2135] train loss: 0.0342154 time: 0:08:21
[epoch: 22 step: 2140] train loss: 0.0105781 time: 0:08:22
[epoch: 22 step: 2145] train loss: 0.00922428 time: 0:08:24
[epoch: 22 step: 2150] train loss: 0.0421383 time: 0:08:25
[epoch: 22 step: 2155] train loss: 0.0173676 time: 0:08:26
[epoch: 22 step: 2160] train loss: 0.0289083 time: 0:08:27
[epoch: 22 step: 2165] train loss: 0.0452846 time: 0:08:28
[epoch: 22 step: 2170] train loss: 0.0236736 time: 0:08:29
[epoch: 22 step: 2175] train loss: 0.0794543 time: 0:08:30
[epoch: 22 step: 2180] train loss: 0.0304108 time: 0:08:31
[epoch: 22 step: 2185] train loss: 0.0704948 time: 0:08:32
Evaluate data in 1.11 seconds!
Evaluation on dev at Epoch 22/40. Step:2185/4000:
AccuracyMetric: acc=0.739884
ClassifyFPreRecMetric: f=0.731847, pre=0.724101, rec=0.744701
[epoch: 23 step: 2190] train loss: 0.0487476 time: 0:08:35
[epoch: 23 step: 2195] train loss: 0.0211978 time: 0:08:36
[epoch: 23 step: 2200] train loss: 0.0450151 time: 0:08:37
[epoch: 23 step: 2205] train loss: 0.0363668 time: 0:08:38
[epoch: 23 step: 2210] train loss: 0.0235301 time: 0:08:39
[epoch: 23 step: 2215] train loss: 0.037947 time: 0:08:40
[epoch: 23 step: 2220] train loss: 0.0160405 time: 0:08:42
[epoch: 23 step: 2225] train loss: 0.0230165 time: 0:08:43
[epoch: 23 step: 2230] train loss: 0.0236386 time: 0:08:44
[epoch: 23 step: 2235] train loss: 0.0160591 time: 0:08:45
[epoch: 23 step: 2240] train loss: 0.015652 time: 0:08:46
[epoch: 23 step: 2245] train loss: 0.0153331 time: 0:08:47
[epoch: 23 step: 2250] train loss: 0.0132059 time: 0:08:48
[epoch: 23 step: 2255] train loss: 0.013806 time: 0:08:49
[epoch: 23 step: 2260] train loss: 0.0393783 time: 0:08:50
[epoch: 23 step: 2265] train loss: 0.0254626 time: 0:08:51
[epoch: 23 step: 2270] train loss: 0.0373299 time: 0:08:53
[epoch: 23 step: 2275] train loss: 0.0413595 time: 0:08:54
[epoch: 23 step: 2280] train loss: 0.0631317 time: 0:08:55
[epoch: 23 step: 2285] train loss: 0.0116424 time: 0:08:56
Evaluate data in 1.02 seconds!
Evaluation on dev at Epoch 23/40. Step:2285/4000:
AccuracyMetric: acc=0.713873
ClassifyFPreRecMetric: f=0.710952, pre=0.708266, rec=0.742775
[epoch: 24 step: 2290] train loss: 0.0581359 time: 0:08:58
[epoch: 24 step: 2295] train loss: 0.0558862 time: 0:08:59
[epoch: 24 step: 2300] train loss: 0.0473512 time: 0:09:01
[epoch: 24 step: 2305] train loss: 0.0223213 time: 0:09:02
[epoch: 24 step: 2310] train loss: 0.0350982 time: 0:09:03
[epoch: 24 step: 2315] train loss: 0.0201827 time: 0:09:04
[epoch: 24 step: 2320] train loss: 0.0168095 time: 0:09:05
[epoch: 24 step: 2325] train loss: 0.0210689 time: 0:09:06
[epoch: 24 step: 2330] train loss: 0.0194665 time: 0:09:07
[epoch: 24 step: 2335] train loss: 0.020499 time: 0:09:08
[epoch: 24 step: 2340] train loss: 0.0325294 time: 0:09:09
[epoch: 24 step: 2345] train loss: 0.0197196 time: 0:09:10
[epoch: 24 step: 2350] train loss: 0.0101744 time: 0:09:11
[epoch: 24 step: 2355] train loss: 0.0100923 time: 0:09:12
[epoch: 24 step: 2360] train loss: 0.029394 time: 0:09:13
[epoch: 24 step: 2365] train loss: 0.0426516 time: 0:09:15
[epoch: 24 step: 2370] train loss: 0.00997483 time: 0:09:16
[epoch: 24 step: 2375] train loss: 0.0444601 time: 0:09:17
[epoch: 24 step: 2380] train loss: 0.0246397 time: 0:09:18
[epoch: 24 step: 2385] train loss: 0.0433294 time: 0:09:19
Evaluate data in 1.07 seconds!
Evaluation on dev at Epoch 24/40. Step:2385/4000:
AccuracyMetric: acc=0.739884
ClassifyFPreRecMetric: f=0.734312, pre=0.726359, rec=0.748555
[epoch: 25 step: 2390] train loss: 0.0308945 time: 0:09:21
[epoch: 25 step: 2395] train loss: 0.0193565 time: 0:09:22
[epoch: 25 step: 2400] train loss: 0.0131431 time: 0:09:23
[epoch: 25 step: 2405] train loss: 0.00615937 time: 0:09:25
[epoch: 25 step: 2410] train loss: 0.033512 time: 0:09:26
[epoch: 25 step: 2415] train loss: 0.0241989 time: 0:09:27
[epoch: 25 step: 2420] train loss: 0.0524742 time: 0:09:28
[epoch: 25 step: 2425] train loss: 0.00476483 time: 0:09:29
[epoch: 25 step: 2430] train loss: 0.021179 time: 0:09:30
[epoch: 25 step: 2435] train loss: 0.0416606 time: 0:09:31
[epoch: 25 step: 2440] train loss: 0.0551788 time: 0:09:32
[epoch: 25 step: 2445] train loss: 0.0484138 time: 0:09:33
[epoch: 25 step: 2450] train loss: 0.0574146 time: 0:09:34
[epoch: 25 step: 2455] train loss: 0.0346037 time: 0:09:35
[epoch: 25 step: 2460] train loss: 0.0108237 time: 0:09:37
[epoch: 25 step: 2465] train loss: 0.0294231 time: 0:09:38
[epoch: 25 step: 2470] train loss: 0.0397738 time: 0:09:39
[epoch: 25 step: 2475] train loss: 0.0276438 time: 0:09:40
[epoch: 25 step: 2480] train loss: 0.0281831 time: 0:09:41
[epoch: 25 step: 2485] train loss: 0.0365072 time: 0:09:42
Evaluate data in 1.02 seconds!
Evaluation on dev at Epoch 25/40. Step:2485/4000:
AccuracyMetric: acc=0.734104
ClassifyFPreRecMetric: f=0.725915, pre=0.720841, rec=0.732177
[epoch: 26 step: 2490] train loss: 0.0202263 time: 0:09:45
[epoch: 26 step: 2495] train loss: 0.037064 time: 0:09:46
[epoch: 26 step: 2500] train loss: 0.0463464 time: 0:09:47
[epoch: 26 step: 2505] train loss: 0.0291489 time: 0:09:48
[epoch: 26 step: 2510] train loss: 0.0347035 time: 0:09:49
[epoch: 26 step: 2515] train loss: 0.0235692 time: 0:09:50
[epoch: 26 step: 2520] train loss: 0.0279975 time: 0:09:51
[epoch: 26 step: 2525] train loss: 0.0317 time: 0:09:52
[epoch: 26 step: 2530] train loss: 0.0438738 time: 0:09:54
[epoch: 26 step: 2535] train loss: 0.0186653 time: 0:09:55
[epoch: 26 step: 2540] train loss: 0.0466459 time: 0:09:56
[epoch: 26 step: 2545] train loss: 0.0310462 time: 0:09:57
[epoch: 26 step: 2550] train loss: 0.0370185 time: 0:09:58
[epoch: 26 step: 2555] train loss: 0.0182208 time: 0:09:59
[epoch: 26 step: 2560] train loss: 0.0110572 time: 0:10:00
[epoch: 26 step: 2565] train loss: 0.0654995 time: 0:10:01
[epoch: 26 step: 2570] train loss: 0.00942638 time: 0:10:02
[epoch: 26 step: 2575] train loss: 0.0491904 time: 0:10:03
[epoch: 26 step: 2580] train loss: 0.0248037 time: 0:10:05
Evaluate data in 1.05 seconds!
Evaluation on dev at Epoch 26/40. Step:2584/4000:
AccuracyMetric: acc=0.764451
ClassifyFPreRecMetric: f=0.7548, pre=0.751762, rec=0.758189
[epoch: 27 step: 2585] train loss: 0.0237237 time: 0:10:07
[epoch: 27 step: 2590] train loss: 0.0165207 time: 0:10:08
[epoch: 27 step: 2595] train loss: 0.0128053 time: 0:10:09
[epoch: 27 step: 2600] train loss: 0.00591129 time: 0:10:10
[epoch: 27 step: 2605] train loss: 0.0109911 time: 0:10:11
[epoch: 27 step: 2610] train loss: 0.0162573 time: 0:10:13
[epoch: 27 step: 2615] train loss: 0.00380667 time: 0:10:14
[epoch: 27 step: 2620] train loss: 0.0182462 time: 0:10:15
[epoch: 27 step: 2625] train loss: 0.0132512 time: 0:10:16
[epoch: 27 step: 2630] train loss: 0.00973649 time: 0:10:17
[epoch: 27 step: 2635] train loss: 0.0292162 time: 0:10:18
[epoch: 27 step: 2640] train loss: 0.002876 time: 0:10:19
[epoch: 27 step: 2645] train loss: 0.00523205 time: 0:10:20
[epoch: 27 step: 2650] train loss: 0.00260634 time: 0:10:21
[epoch: 27 step: 2655] train loss: 0.00446276 time: 0:10:22
[epoch: 27 step: 2660] train loss: 0.003339 time: 0:10:24
[epoch: 27 step: 2665] train loss: 0.0136002 time: 0:10:25
[epoch: 27 step: 2670] train loss: 0.0121612 time: 0:10:26
[epoch: 27 step: 2675] train loss: 0.0188614 time: 0:10:27
[epoch: 27 step: 2680] train loss: 0.00106157 time: 0:10:28
Evaluate data in 1.06 seconds!
Evaluation on dev at Epoch 27/40. Step:2683/4000:
AccuracyMetric: acc=0.760116
ClassifyFPreRecMetric: f=0.753636, pre=0.747169, rec=0.762042
[epoch: 28 step: 2685] train loss: 0.0145727 time: 0:10:30
[epoch: 28 step: 2690] train loss: 0.00362587 time: 0:10:32
[epoch: 28 step: 2695] train loss: 0.00415807 time: 0:10:33
[epoch: 28 step: 2700] train loss: 0.00722865 time: 0:10:34
[epoch: 28 step: 2705] train loss: 0.00384618 time: 0:10:35
[epoch: 28 step: 2710] train loss: 0.00312952 time: 0:10:36
[epoch: 28 step: 2715] train loss: 0.0062763 time: 0:10:37
[epoch: 28 step: 2720] train loss: 0.0113585 time: 0:10:38
[epoch: 28 step: 2725] train loss: 0.0142391 time: 0:10:39
[epoch: 28 step: 2730] train loss: 0.00774773 time: 0:10:40
[epoch: 28 step: 2735] train loss: 0.0168115 time: 0:10:41
[epoch: 28 step: 2740] train loss: 0.00285072 time: 0:10:43
[epoch: 28 step: 2745] train loss: 0.0104214 time: 0:10:44
[epoch: 28 step: 2750] train loss: 0.0489625 time: 0:10:45
[epoch: 28 step: 2755] train loss: 0.0127681 time: 0:10:46
[epoch: 28 step: 2760] train loss: 0.0310872 time: 0:10:47
[epoch: 28 step: 2765] train loss: 0.0284757 time: 0:10:48
[epoch: 28 step: 2770] train loss: 0.0406551 time: 0:10:49
[epoch: 28 step: 2775] train loss: 0.103437 time: 0:10:50
[epoch: 28 step: 2780] train loss: 0.0376491 time: 0:10:52
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 28/40. Step:2782/4000:
AccuracyMetric: acc=0.735549
ClassifyFPreRecMetric: f=0.716532, pre=0.737291, rec=0.703276
[epoch: 29 step: 2785] train loss: 0.0805964 time: 0:10:54
[epoch: 29 step: 2790] train loss: 0.120751 time: 0:10:55
[epoch: 29 step: 2795] train loss: 0.0103458 time: 0:10:56
[epoch: 29 step: 2800] train loss: 0.129276 time: 0:10:57
[epoch: 29 step: 2805] train loss: 0.0395072 time: 0:10:59
[epoch: 29 step: 2810] train loss: 0.137211 time: 0:11:00
[epoch: 29 step: 2815] train loss: 0.0294531 time: 0:11:01
[epoch: 29 step: 2820] train loss: 0.118781 time: 0:11:02
[epoch: 29 step: 2825] train loss: 0.097555 time: 0:11:03
[epoch: 29 step: 2830] train loss: 0.0571168 time: 0:11:04
[epoch: 29 step: 2835] train loss: 0.0362304 time: 0:11:05
[epoch: 29 step: 2840] train loss: 0.0686007 time: 0:11:06
[epoch: 29 step: 2845] train loss: 0.0169458 time: 0:11:07
[epoch: 29 step: 2850] train loss: 0.0218714 time: 0:11:09
[epoch: 29 step: 2855] train loss: 0.0424088 time: 0:11:10
[epoch: 29 step: 2860] train loss: 0.0352423 time: 0:11:11
[epoch: 29 step: 2865] train loss: 0.015848 time: 0:11:12
[epoch: 29 step: 2870] train loss: 0.037257 time: 0:11:13
[epoch: 29 step: 2875] train loss: 0.0263848 time: 0:11:14
[epoch: 29 step: 2880] train loss: 0.0136789 time: 0:11:15
Evaluate data in 0.99 seconds!
Evaluation on dev at Epoch 29/40. Step:2882/4000:
AccuracyMetric: acc=0.74422
ClassifyFPreRecMetric: f=0.737205, pre=0.736928, rec=0.740848
[epoch: 30 step: 2885] train loss: 0.0119094 time: 0:11:17
[epoch: 30 step: 2890] train loss: 0.00750754 time: 0:11:19
[epoch: 30 step: 2895] train loss: 0.0410767 time: 0:11:20
[epoch: 30 step: 2900] train loss: 0.0548536 time: 0:11:21
[epoch: 30 step: 2905] train loss: 0.110413 time: 0:11:22
[epoch: 30 step: 2910] train loss: 0.0470082 time: 0:11:23
[epoch: 30 step: 2915] train loss: 0.0553099 time: 0:11:24
[epoch: 30 step: 2920] train loss: 0.0250072 time: 0:11:25
[epoch: 30 step: 2925] train loss: 0.0405053 time: 0:11:26
[epoch: 30 step: 2930] train loss: 0.0504158 time: 0:11:28
[epoch: 30 step: 2935] train loss: 0.034178 time: 0:11:29
[epoch: 30 step: 2940] train loss: 0.00408491 time: 0:11:30
[epoch: 30 step: 2945] train loss: 0.0503086 time: 0:11:31
[epoch: 30 step: 2950] train loss: 0.0818106 time: 0:11:32
[epoch: 30 step: 2955] train loss: 0.0414844 time: 0:11:33
[epoch: 30 step: 2960] train loss: 0.0184179 time: 0:11:34
[epoch: 30 step: 2965] train loss: 0.0369965 time: 0:11:35
[epoch: 30 step: 2970] train loss: 0.0081557 time: 0:11:37
[epoch: 30 step: 2975] train loss: 0.0253334 time: 0:11:38
[epoch: 30 step: 2980] train loss: 0.0275307 time: 0:11:39
Evaluate data in 1.05 seconds!
Evaluation on dev at Epoch 30/40. Step:2981/4000:
AccuracyMetric: acc=0.739884
ClassifyFPreRecMetric: f=0.736413, pre=0.727333, rec=0.753372
[epoch: 31 step: 2985] train loss: 0.0160446 time: 0:11:41
[epoch: 31 step: 2990] train loss: 0.0142177 time: 0:11:42
[epoch: 31 step: 2995] train loss: 0.0127497 time: 0:11:44
[epoch: 31 step: 3000] train loss: 0.0110957 time: 0:11:45
[epoch: 31 step: 3005] train loss: 0.0141594 time: 0:11:46
[epoch: 31 step: 3010] train loss: 0.0397089 time: 0:11:47
[epoch: 31 step: 3015] train loss: 0.00845332 time: 0:11:48
[epoch: 31 step: 3020] train loss: 0.0283848 time: 0:11:49
[epoch: 31 step: 3025] train loss: 0.00876734 time: 0:11:50
[epoch: 31 step: 3030] train loss: 0.00761751 time: 0:11:52
[epoch: 31 step: 3035] train loss: 0.00245171 time: 0:11:53
[epoch: 31 step: 3040] train loss: 0.0279476 time: 0:11:54
[epoch: 31 step: 3045] train loss: 0.024666 time: 0:11:55
[epoch: 31 step: 3050] train loss: 0.00358941 time: 0:11:56
[epoch: 31 step: 3055] train loss: 0.0113073 time: 0:11:57
[epoch: 31 step: 3060] train loss: 0.0139717 time: 0:11:58
[epoch: 31 step: 3065] train loss: 0.0188745 time: 0:11:59
[epoch: 31 step: 3070] train loss: 0.0187576 time: 0:12:00
[epoch: 31 step: 3075] train loss: 0.00409914 time: 0:12:02
[epoch: 31 step: 3080] train loss: 0.0196871 time: 0:12:03
Evaluate data in 1.05 seconds!
Evaluation on dev at Epoch 31/40. Step:3081/4000:
AccuracyMetric: acc=0.74711
ClassifyFPreRecMetric: f=0.735913, pre=0.735648, rec=0.742775
[epoch: 32 step: 3085] train loss: 0.00660522 time: 0:12:05
[epoch: 32 step: 3090] train loss: 0.042598 time: 0:12:06
[epoch: 32 step: 3095] train loss: 0.0364246 time: 0:12:07
[epoch: 32 step: 3100] train loss: 0.0132036 time: 0:12:08
[epoch: 32 step: 3105] train loss: 0.0120029 time: 0:12:09
[epoch: 32 step: 3110] train loss: 0.00574468 time: 0:12:11
[epoch: 32 step: 3115] train loss: 0.0162101 time: 0:12:12
[epoch: 32 step: 3120] train loss: 0.0434415 time: 0:12:13
[epoch: 32 step: 3125] train loss: 0.0341415 time: 0:12:14
[epoch: 32 step: 3130] train loss: 0.0355587 time: 0:12:15
[epoch: 32 step: 3135] train loss: 0.0197093 time: 0:12:16
[epoch: 32 step: 3140] train loss: 0.0290261 time: 0:12:17
[epoch: 32 step: 3145] train loss: 0.0583945 time: 0:12:19
[epoch: 32 step: 3150] train loss: 0.0264685 time: 0:12:20
[epoch: 32 step: 3155] train loss: 0.0320532 time: 0:12:21
[epoch: 32 step: 3160] train loss: 0.0241357 time: 0:12:22
[epoch: 32 step: 3165] train loss: 0.0192306 time: 0:12:23
[epoch: 32 step: 3170] train loss: 0.00652523 time: 0:12:24
[epoch: 32 step: 3175] train loss: 0.0323015 time: 0:12:25
[epoch: 32 step: 3180] train loss: 0.0173433 time: 0:12:26
Evaluate data in 1.01 seconds!
Evaluation on dev at Epoch 32/40. Step:3180/4000:
AccuracyMetric: acc=0.739884
ClassifyFPreRecMetric: f=0.728269, pre=0.727497, rec=0.731214
[epoch: 33 step: 3185] train loss: 0.00637909 time: 0:12:29
[epoch: 33 step: 3190] train loss: 0.0100387 time: 0:12:30
[epoch: 33 step: 3195] train loss: 0.00410429 time: 0:12:31
[epoch: 33 step: 3200] train loss: 0.00748123 time: 0:12:32
[epoch: 33 step: 3205] train loss: 0.0135016 time: 0:12:33
[epoch: 33 step: 3210] train loss: 0.0532061 time: 0:12:34
[epoch: 33 step: 3215] train loss: 0.026976 time: 0:12:35
[epoch: 33 step: 3220] train loss: 0.0619597 time: 0:12:37
[epoch: 33 step: 3225] train loss: 0.0159952 time: 0:12:38
[epoch: 33 step: 3230] train loss: 0.0812861 time: 0:12:39
[epoch: 33 step: 3235] train loss: 0.122266 time: 0:12:40
[epoch: 33 step: 3240] train loss: 0.113538 time: 0:12:41
[epoch: 33 step: 3245] train loss: 0.0266255 time: 0:12:42
[epoch: 33 step: 3250] train loss: 0.0522224 time: 0:12:43
[epoch: 33 step: 3255] train loss: 0.0297226 time: 0:12:45
[epoch: 33 step: 3260] train loss: 0.0347748 time: 0:12:46
[epoch: 33 step: 3265] train loss: 0.0258506 time: 0:12:47
[epoch: 33 step: 3270] train loss: 0.0128289 time: 0:12:48
[epoch: 33 step: 3275] train loss: 0.0218095 time: 0:12:49
Evaluate data in 1.03 seconds!
Evaluation on dev at Epoch 33/40. Step:3279/4000:
AccuracyMetric: acc=0.729769
ClassifyFPreRecMetric: f=0.718252, pre=0.71653, rec=0.72447
[epoch: 34 step: 3280] train loss: 0.00326808 time: 0:12:51
[epoch: 34 step: 3285] train loss: 0.00635173 time: 0:12:52
[epoch: 34 step: 3290] train loss: 0.00788111 time: 0:12:53
[epoch: 34 step: 3295] train loss: 0.00536279 time: 0:12:54
[epoch: 34 step: 3300] train loss: 0.0245352 time: 0:12:56
[epoch: 34 step: 3305] train loss: 0.0258414 time: 0:12:57
[epoch: 34 step: 3310] train loss: 0.00124637 time: 0:12:58
[epoch: 34 step: 3315] train loss: 0.0260081 time: 0:12:59
[epoch: 34 step: 3320] train loss: 0.00551642 time: 0:13:00
[epoch: 34 step: 3325] train loss: 0.00830744 time: 0:13:01
[epoch: 34 step: 3330] train loss: 0.0193638 time: 0:13:02
[epoch: 34 step: 3335] train loss: 0.0060313 time: 0:13:03
[epoch: 34 step: 3340] train loss: 0.0507447 time: 0:13:04
[epoch: 34 step: 3345] train loss: 0.0446804 time: 0:13:05
[epoch: 34 step: 3350] train loss: 0.0584427 time: 0:13:06
[epoch: 34 step: 3355] train loss: 0.0348987 time: 0:13:07
[epoch: 34 step: 3360] train loss: 0.0141275 time: 0:13:09
[epoch: 34 step: 3365] train loss: 0.00956455 time: 0:13:10
[epoch: 34 step: 3370] train loss: 0.0135289 time: 0:13:11
[epoch: 34 step: 3375] train loss: 0.0165924 time: 0:13:12
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 34/40. Step:3379/4000:
AccuracyMetric: acc=0.742775
ClassifyFPreRecMetric: f=0.735476, pre=0.730517, rec=0.743738
[epoch: 35 step: 3380] train loss: 0.0387775 time: 0:13:14
[epoch: 35 step: 3385] train loss: 0.021852 time: 0:13:15
[epoch: 35 step: 3390] train loss: 0.0163006 time: 0:13:16
[epoch: 35 step: 3395] train loss: 0.0285978 time: 0:13:18
[epoch: 35 step: 3400] train loss: 0.0158669 time: 0:13:19
[epoch: 35 step: 3405] train loss: 0.0154483 time: 0:13:20
[epoch: 35 step: 3410] train loss: 0.0293378 time: 0:13:21
[epoch: 35 step: 3415] train loss: 0.0141502 time: 0:13:22
[epoch: 35 step: 3420] train loss: 0.019966 time: 0:13:23
[epoch: 35 step: 3425] train loss: 0.00263632 time: 0:13:24
[epoch: 35 step: 3430] train loss: 0.00717409 time: 0:13:25
[epoch: 35 step: 3435] train loss: 0.0112033 time: 0:13:26
[epoch: 35 step: 3440] train loss: 0.0260018 time: 0:13:27
[epoch: 35 step: 3445] train loss: 0.00289766 time: 0:13:29
[epoch: 35 step: 3450] train loss: 0.00802878 time: 0:13:30
[epoch: 35 step: 3455] train loss: 0.0112625 time: 0:13:31
[epoch: 35 step: 3460] train loss: 0.00948343 time: 0:13:32
[epoch: 35 step: 3465] train loss: 0.0129518 time: 0:13:33
[epoch: 35 step: 3470] train loss: 0.00740895 time: 0:13:34
[epoch: 35 step: 3475] train loss: 0.0204329 time: 0:13:35
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 35/40. Step:3478/4000:
AccuracyMetric: acc=0.75289
ClassifyFPreRecMetric: f=0.744025, pre=0.740834, rec=0.747592
[epoch: 36 step: 3480] train loss: 0.00382822 time: 0:13:37
[epoch: 36 step: 3485] train loss: 0.0352342 time: 0:13:38
[epoch: 36 step: 3490] train loss: 0.052569 time: 0:13:40
[epoch: 36 step: 3495] train loss: 0.0096562 time: 0:13:41
[epoch: 36 step: 3500] train loss: 0.0137752 time: 0:13:42
[epoch: 36 step: 3505] train loss: 0.0237168 time: 0:13:43
[epoch: 36 step: 3510] train loss: 0.00352897 time: 0:13:44
[epoch: 36 step: 3515] train loss: 0.0282929 time: 0:13:45
[epoch: 36 step: 3520] train loss: 0.0311635 time: 0:13:46
[epoch: 36 step: 3525] train loss: 0.0122684 time: 0:13:47
[epoch: 36 step: 3530] train loss: 0.0174832 time: 0:13:48
[epoch: 36 step: 3535] train loss: 0.0122037 time: 0:13:49
[epoch: 36 step: 3540] train loss: 0.00911354 time: 0:13:50
[epoch: 36 step: 3545] train loss: 0.0437233 time: 0:13:52
[epoch: 36 step: 3550] train loss: 0.00543887 time: 0:13:53
[epoch: 36 step: 3555] train loss: 0.0212632 time: 0:13:54
[epoch: 36 step: 3560] train loss: 0.00353659 time: 0:13:55
[epoch: 36 step: 3565] train loss: 0.0292816 time: 0:13:56
[epoch: 36 step: 3570] train loss: 0.0267267 time: 0:13:57
[epoch: 36 step: 3575] train loss: 0.00590311 time: 0:13:58
Evaluate data in 1.09 seconds!
Evaluation on dev at Epoch 36/40. Step:3578/4000:
AccuracyMetric: acc=0.739884
ClassifyFPreRecMetric: f=0.732396, pre=0.727731, rec=0.737958
[epoch: 37 step: 3580] train loss: 0.0474948 time: 0:14:01
[epoch: 37 step: 3585] train loss: 0.0149129 time: 0:14:02
[epoch: 37 step: 3590] train loss: 0.0327283 time: 0:14:03
[epoch: 37 step: 3595] train loss: 0.00772098 time: 0:14:04
[epoch: 37 step: 3600] train loss: 0.00885975 time: 0:14:05
[epoch: 37 step: 3605] train loss: 0.0202815 time: 0:14:06
[epoch: 37 step: 3610] train loss: 0.00725806 time: 0:14:07
[epoch: 37 step: 3615] train loss: 0.00595141 time: 0:14:08
[epoch: 37 step: 3620] train loss: 0.0416214 time: 0:14:10
[epoch: 37 step: 3625] train loss: 0.0180997 time: 0:14:11
[epoch: 37 step: 3630] train loss: 0.0133148 time: 0:14:12
[epoch: 37 step: 3635] train loss: 0.00268015 time: 0:14:13
[epoch: 37 step: 3640] train loss: 0.0127212 time: 0:14:14
[epoch: 37 step: 3645] train loss: 0.00657735 time: 0:14:15
[epoch: 37 step: 3650] train loss: 0.00348991 time: 0:14:16
[epoch: 37 step: 3655] train loss: 0.0126788 time: 0:14:17
[epoch: 37 step: 3660] train loss: 0.021383 time: 0:14:18
[epoch: 37 step: 3665] train loss: 0.00422911 time: 0:14:19
[epoch: 37 step: 3670] train loss: 0.00666931 time: 0:14:21
[epoch: 37 step: 3675] train loss: 0.0035347 time: 0:14:22
Evaluate data in 1.07 seconds!
Evaluation on dev at Epoch 37/40. Step:3678/4000:
AccuracyMetric: acc=0.75
ClassifyFPreRecMetric: f=0.742118, pre=0.73446, rec=0.756262
[epoch: 38 step: 3680] train loss: 0.0159097 time: 0:14:24
[epoch: 38 step: 3685] train loss: 0.0080168 time: 0:14:25
[epoch: 38 step: 3690] train loss: 0.00205557 time: 0:14:26
[epoch: 38 step: 3695] train loss: 0.00515276 time: 0:14:27
[epoch: 38 step: 3700] train loss: 0.0044554 time: 0:14:28
[epoch: 38 step: 3705] train loss: 0.0132944 time: 0:14:30
[epoch: 38 step: 3710] train loss: 0.00412378 time: 0:14:31
[epoch: 38 step: 3715] train loss: 0.0141759 time: 0:14:32
[epoch: 38 step: 3720] train loss: 0.0285306 time: 0:14:33
[epoch: 38 step: 3725] train loss: 0.0183273 time: 0:14:34
[epoch: 38 step: 3730] train loss: 0.00369076 time: 0:14:35
[epoch: 38 step: 3735] train loss: 0.00682234 time: 0:14:36
[epoch: 38 step: 3740] train loss: 0.0194611 time: 0:14:37
[epoch: 38 step: 3745] train loss: 0.0153373 time: 0:14:38
[epoch: 38 step: 3750] train loss: 0.0173112 time: 0:14:40
[epoch: 38 step: 3755] train loss: 0.0132194 time: 0:14:41
[epoch: 38 step: 3760] train loss: 0.0149764 time: 0:14:42
[epoch: 38 step: 3765] train loss: 0.0296497 time: 0:14:43
[epoch: 38 step: 3770] train loss: 0.0444196 time: 0:14:44
[epoch: 38 step: 3775] train loss: 0.0369212 time: 0:14:45
Evaluate data in 0.99 seconds!
Evaluation on dev at Epoch 38/40. Step:3777/4000:
AccuracyMetric: acc=0.734104
ClassifyFPreRecMetric: f=0.723387, pre=0.723265, rec=0.738921
[epoch: 39 step: 3780] train loss: 0.00448706 time: 0:14:47
[epoch: 39 step: 3785] train loss: 0.00182417 time: 0:14:48
[epoch: 39 step: 3790] train loss: 0.0562886 time: 0:14:50
[epoch: 39 step: 3795] train loss: 0.0317504 time: 0:14:51
[epoch: 39 step: 3800] train loss: 0.0139837 time: 0:14:52
[epoch: 39 step: 3805] train loss: 0.0216679 time: 0:14:53
[epoch: 39 step: 3810] train loss: 0.033562 time: 0:14:54
[epoch: 39 step: 3815] train loss: 0.0186327 time: 0:14:55
[epoch: 39 step: 3820] train loss: 0.0283727 time: 0:14:56
[epoch: 39 step: 3825] train loss: 0.0192097 time: 0:14:57
[epoch: 39 step: 3830] train loss: 0.00824567 time: 0:14:58
[epoch: 39 step: 3835] train loss: 0.0170784 time: 0:15:00
[epoch: 39 step: 3840] train loss: 0.00728181 time: 0:15:01
[epoch: 39 step: 3845] train loss: 0.00785138 time: 0:15:02
[epoch: 39 step: 3850] train loss: 0.00440814 time: 0:15:03
[epoch: 39 step: 3855] train loss: 0.0189377 time: 0:15:04
[epoch: 39 step: 3860] train loss: 0.0336996 time: 0:15:05
[epoch: 39 step: 3865] train loss: 0.0322053 time: 0:15:06
[epoch: 39 step: 3870] train loss: 0.0325616 time: 0:15:07
[epoch: 39 step: 3875] train loss: 0.0140382 time: 0:15:08
Evaluate data in 1.02 seconds!
Evaluation on dev at Epoch 39/40. Step:3876/4000:
AccuracyMetric: acc=0.745665
ClassifyFPreRecMetric: f=0.733856, pre=0.733946, rec=0.734104
[epoch: 40 step: 3880] train loss: 0.0120895 time: 0:15:11
[epoch: 40 step: 3885] train loss: 0.00877676 time: 0:15:12
[epoch: 40 step: 3890] train loss: 0.00399007 time: 0:15:13
[epoch: 40 step: 3895] train loss: 0.0146379 time: 0:15:14
[epoch: 40 step: 3900] train loss: 0.00443851 time: 0:15:15
[epoch: 40 step: 3905] train loss: 0.0133478 time: 0:15:16
[epoch: 40 step: 3910] train loss: 0.00979301 time: 0:15:17
[epoch: 40 step: 3915] train loss: 0.00818301 time: 0:15:18
[epoch: 40 step: 3920] train loss: 0.0149235 time: 0:15:20
[epoch: 40 step: 3925] train loss: 0.0103201 time: 0:15:21
[epoch: 40 step: 3930] train loss: 0.00629345 time: 0:15:22
[epoch: 40 step: 3935] train loss: 0.0191421 time: 0:15:23
[epoch: 40 step: 3940] train loss: 0.0154709 time: 0:15:24
[epoch: 40 step: 3945] train loss: 0.0063882 time: 0:15:25
[epoch: 40 step: 3950] train loss: 0.00733271 time: 0:15:26
[epoch: 40 step: 3955] train loss: 0.0107157 time: 0:15:27
[epoch: 40 step: 3960] train loss: 0.00975401 time: 0:15:29
[epoch: 40 step: 3965] train loss: 0.0331033 time: 0:15:30
[epoch: 40 step: 3970] train loss: 0.0132963 time: 0:15:31
[epoch: 40 step: 3975] train loss: 0.0036747 time: 0:15:32
Evaluate data in 1.05 seconds!
Evaluation on dev at Epoch 40/40. Step:3975/4000:
AccuracyMetric: acc=0.74711
ClassifyFPreRecMetric: f=0.738392, pre=0.734412, rec=0.743738
Reloaded the best model.
In Epoch:15/Step:1492, got best dev performance:
AccuracyMetric: acc=0.774566
ClassifyFPreRecMetric: f=0.768735, pre=0.765939, rec=0.77553