Created by: JepsonWong
add dataloader for seq2seq.
1.7.0:
-- Epoch:[0]; Batch:[100]; Time: 0.13797 s; ppl: 485.27556
-- Epoch:[0]; Batch:[200]; Time: 0.12886 s; ppl: 240.92520
-- Epoch:[0]; Batch:[300]; Time: 0.14664 s; ppl: 146.76782
-- Epoch:[0]; Batch:[400]; Time: 0.11538 s; ppl: 98.72650
-- Epoch:[0]; Batch:[500]; Time: 0.14769 s; ppl: 72.46521
-- Epoch:[0]; Batch:[600]; Time: 0.10891 s; ppl: 50.58025
-- Epoch:[0]; Batch:[700]; Time: 0.11890 s; ppl: 43.79331
-- Epoch:[0]; Batch:[800]; Time: 0.12919 s; ppl: 36.08316
-- Epoch:[0]; Batch:[900]; Time: 0.13054 s; ppl: 33.17884
-- Epoch:[0]; Batch:[1000]; Time: 0.10040 s; ppl: 33.15334
Train epoch:[0]; Epoch Time: 327.73248; avg_time: 0.31075 s/step
-- Epoch:[1]; Batch:[100]; Time: 0.10352 s; ppl: 25.16106
-- Epoch:[1]; Batch:[200]; Time: 0.12170 s; ppl: 22.02668
-- Epoch:[1]; Batch:[300]; Time: 0.11957 s; ppl: 19.21704
-- Epoch:[1]; Batch:[400]; Time: 0.10764 s; ppl: 19.65568
-- Epoch:[1]; Batch:[500]; Time: 0.14002 s; ppl: 18.72767
-- Epoch:[1]; Batch:[600]; Time: 0.10404 s; ppl: 16.13224
-- Epoch:[1]; Batch:[700]; Time: 0.11563 s; ppl: 16.32774
-- Epoch:[1]; Batch:[800]; Time: 0.12431 s; ppl: 15.09768
-- Epoch:[1]; Batch:[900]; Time: 0.13960 s; ppl: 15.16519
-- Epoch:[1]; Batch:[1000]; Time: 0.10603 s; ppl: 16.14104
Train epoch:[1]; Epoch Time: 317.63054; avg_time: 0.30145 s/step
-- Epoch:[2]; Batch:[100]; Time: 0.11607 s; ppl: 13.42104
-- Epoch:[2]; Batch:[200]; Time: 0.12969 s; ppl: 12.45739
-- Epoch:[2]; Batch:[300]; Time: 0.13076 s; ppl: 11.25073
-- Epoch:[2]; Batch:[400]; Time: 0.11045 s; ppl: 12.08184
-- Epoch:[2]; Batch:[500]; Time: 0.13646 s; ppl: 11.88435
-- Epoch:[2]; Batch:[600]; Time: 0.10717 s; ppl: 10.73674
-- Epoch:[2]; Batch:[700]; Time: 0.11687 s; ppl: 11.02143
-- Epoch:[2]; Batch:[800]; Time: 0.13232 s; ppl: 10.60439
-- Epoch:[2]; Batch:[900]; Time: 0.14060 s; ppl: 10.46024
-- Epoch:[2]; Batch:[1000]; Time: 0.09651 s; ppl: 11.37829
Train epoch:[2]; Epoch Time: 321.35660; avg_time: 0.30495 s/step
-- Epoch:[3]; Batch:[100]; Time: 0.12971 s; ppl: 9.84421
-- Epoch:[3]; Batch:[200]; Time: 0.17458 s; ppl: 9.23038
-- Epoch:[3]; Batch:[300]; Time: 0.12176 s; ppl: 8.49308
-- Epoch:[3]; Batch:[400]; Time: 0.11894 s; ppl: 9.14306
-- Epoch:[3]; Batch:[500]; Time: 0.13966 s; ppl: 9.24141
-- Epoch:[3]; Batch:[600]; Time: 0.10183 s; ppl: 8.41742
-- Epoch:[3]; Batch:[700]; Time: 0.11645 s; ppl: 8.62797
-- Epoch:[3]; Batch:[800]; Time: 0.12236 s; ppl: 8.60114
-- Epoch:[3]; Batch:[900]; Time: 0.13539 s; ppl: 8.35749
-- Epoch:[3]; Batch:[1000]; Time: 0.09380 s; ppl: 9.18386
Train epoch:[3]; Epoch Time: 321.00874; avg_time: 0.30430 s/step
-- Epoch:[4]; Batch:[100]; Time: 0.10334 s; ppl: 8.12115
-- Epoch:[4]; Batch:[200]; Time: 0.11693 s; ppl: 7.55782
-- Epoch:[4]; Batch:[300]; Time: 0.10830 s; ppl: 7.11914
-- Epoch:[4]; Batch:[400]; Time: 0.11142 s; ppl: 7.55348
-- Epoch:[4]; Batch:[500]; Time: 0.12647 s; ppl: 7.67579
-- Epoch:[4]; Batch:[600]; Time: 0.09058 s; ppl: 7.06502
-- Epoch:[4]; Batch:[700]; Time: 0.13929 s; ppl: 7.21603
-- Epoch:[4]; Batch:[800]; Time: 0.11377 s; ppl: 7.19111
-- Epoch:[4]; Batch:[900]; Time: 0.12153 s; ppl: 7.02148
-- Epoch:[4]; Batch:[1000]; Time: 0.08657 s; ppl: 7.84391
Train epoch:[4]; Epoch Time: 294.01391; avg_time: 0.27894 s/step
-- Epoch:[5]; Batch:[100]; Time: 0.10274 s; ppl: 6.90528
-- Epoch:[5]; Batch:[200]; Time: 0.11319 s; ppl: 6.40949
-- Epoch:[5]; Batch:[300]; Time: 0.11147 s; ppl: 6.10432
-- Epoch:[5]; Batch:[400]; Time: 0.09536 s; ppl: 6.62869
-- Epoch:[5]; Batch:[500]; Time: 0.12387 s; ppl: 6.63936
-- Epoch:[5]; Batch:[600]; Time: 0.09200 s; ppl: 6.09945
-- Epoch:[5]; Batch:[700]; Time: 0.09753 s; ppl: 6.30576
-- Epoch:[5]; Batch:[800]; Time: 0.12054 s; ppl: 6.20876
-- Epoch:[5]; Batch:[900]; Time: 0.12753 s; ppl: 6.14133
-- Epoch:[5]; Batch:[1000]; Time: 0.09157 s; ppl: 6.72946
Train epoch:[5]; Epoch Time: 290.87147; avg_time: 0.27603 s/step
develop:
-- Epoch:[0]; Batch:[100]; Time: 0.09927 s; ppl: 485.69727
-- Epoch:[0]; Batch:[200]; Time: 0.11030 s; ppl: 240.97714
-- Epoch:[0]; Batch:[300]; Time: 0.10571 s; ppl: 147.61790
-- Epoch:[0]; Batch:[400]; Time: 0.09628 s; ppl: 99.06590
-- Epoch:[0]; Batch:[500]; Time: 0.11892 s; ppl: 72.37745
-- Epoch:[0]; Batch:[600]; Time: 0.09523 s; ppl: 50.20113
-- Epoch:[0]; Batch:[700]; Time: 0.10502 s; ppl: 43.76393
-- Epoch:[0]; Batch:[800]; Time: 0.11429 s; ppl: 35.95067
-- Epoch:[0]; Batch:[900]; Time: 0.12320 s; ppl: 32.83636
-- Epoch:[0]; Batch:[1000]; Time: 0.08886 s; ppl: 33.21380
Train epoch:[0]; Epoch Time: 284.08296; avg_time: 0.26946 s/step
-- Epoch:[1]; Batch:[100]; Time: 0.11073 s; ppl: 25.03152
-- Epoch:[1]; Batch:[200]; Time: 0.12003 s; ppl: 21.86988
-- Epoch:[1]; Batch:[300]; Time: 0.10475 s; ppl: 19.31184
-- Epoch:[1]; Batch:[400]; Time: 0.09941 s; ppl: 19.33428
-- Epoch:[1]; Batch:[500]; Time: 0.15979 s; ppl: 19.01874
-- Epoch:[1]; Batch:[600]; Time: 0.09984 s; ppl: 16.07280
-- Epoch:[1]; Batch:[700]; Time: 0.09836 s; ppl: 16.51026
-- Epoch:[1]; Batch:[800]; Time: 0.10797 s; ppl: 15.06073
-- Epoch:[1]; Batch:[900]; Time: 0.11527 s; ppl: 15.03655
-- Epoch:[1]; Batch:[1000]; Time: 0.08916 s; ppl: 16.28758
Train epoch:[1]; Epoch Time: 292.74497; avg_time: 0.27830 s/step
-- Epoch:[2]; Batch:[100]; Time: 0.09836 s; ppl: 13.60783
-- Epoch:[2]; Batch:[200]; Time: 0.10850 s; ppl: 12.42435
-- Epoch:[2]; Batch:[300]; Time: 0.10372 s; ppl: 11.19597
-- Epoch:[2]; Batch:[400]; Time: 0.09497 s; ppl: 11.83070
-- Epoch:[2]; Batch:[500]; Time: 0.11871 s; ppl: 11.94742
-- Epoch:[2]; Batch:[600]; Time: 0.08622 s; ppl: 10.64706
-- Epoch:[2]; Batch:[700]; Time: 0.09909 s; ppl: 11.17663
-- Epoch:[2]; Batch:[800]; Time: 0.11402 s; ppl: 10.53449
-- Epoch:[2]; Batch:[900]; Time: 0.11905 s; ppl: 10.55764
-- Epoch:[2]; Batch:[1000]; Time: 0.08283 s; ppl: 11.45152
Train epoch:[2]; Epoch Time: 275.63878; avg_time: 0.26152 s/step
-- Epoch:[3]; Batch:[100]; Time: 0.10964 s; ppl: 9.97863
-- Epoch:[3]; Batch:[200]; Time: 0.11491 s; ppl: 9.36300
-- Epoch:[3]; Batch:[300]; Time: 0.11184 s; ppl: 8.47306
-- Epoch:[3]; Batch:[400]; Time: 0.10569 s; ppl: 9.06469
-- Epoch:[3]; Batch:[500]; Time: 0.12949 s; ppl: 9.30301
-- Epoch:[3]; Batch:[600]; Time: 0.09394 s; ppl: 8.35555
-- Epoch:[3]; Batch:[700]; Time: 0.11191 s; ppl: 8.70492
-- Epoch:[3]; Batch:[800]; Time: 0.14475 s; ppl: 8.35808
-- Epoch:[3]; Batch:[900]; Time: 0.16802 s; ppl: 8.35319
-- Epoch:[3]; Batch:[1000]; Time: 0.10665 s; ppl: 9.02744
Train epoch:[3]; Epoch Time: 317.41175; avg_time: 0.30108 s/step
-- Epoch:[4]; Batch:[100]; Time: 0.12824 s; ppl: 8.10008
-- Epoch:[4]; Batch:[200]; Time: 0.11959 s; ppl: 7.57185
-- Epoch:[4]; Batch:[300]; Time: 0.11344 s; ppl: 7.03630
-- Epoch:[4]; Batch:[400]; Time: 0.10568 s; ppl: 7.56011
-- Epoch:[4]; Batch:[500]; Time: 0.13163 s; ppl: 7.85504
-- Epoch:[4]; Batch:[600]; Time: 0.09790 s; ppl: 6.98017
-- Epoch:[4]; Batch:[700]; Time: 0.11025 s; ppl: 7.26515
-- Epoch:[4]; Batch:[800]; Time: 0.11490 s; ppl: 7.15001
-- Epoch:[4]; Batch:[900]; Time: 0.13308 s; ppl: 7.06806
-- Epoch:[4]; Batch:[1000]; Time: 0.08176 s; ppl: 7.71039
Train epoch:[4]; Epoch Time: 312.62300; avg_time: 0.29619 s/step
-- Epoch:[5]; Batch:[100]; Time: 0.11303 s; ppl: 6.87338
-- Epoch:[5]; Batch:[200]; Time: 0.17841 s; ppl: 6.46120
-- Epoch:[5]; Batch:[300]; Time: 0.11422 s; ppl: 6.07377
-- Epoch:[5]; Batch:[400]; Time: 0.10600 s; ppl: 6.62446
-- Epoch:[5]; Batch:[500]; Time: 0.14465 s; ppl: 6.72500
-- Epoch:[5]; Batch:[600]; Time: 0.11218 s; ppl: 6.02742
-- Epoch:[5]; Batch:[700]; Time: 0.10870 s; ppl: 6.40271
-- Epoch:[5]; Batch:[800]; Time: 0.11464 s; ppl: 6.28339
-- Epoch:[5]; Batch:[900]; Time: 0.13164 s; ppl: 6.25134
-- Epoch:[5]; Batch:[1000]; Time: 0.09276 s; ppl: 6.79961
Train epoch:[5]; Epoch Time: 316.02392; avg_time: 0.29946 s/step