Namespace(batch_norm=False, batch_size=256, crop_ratio=0.875, data_dir='~/.mxnet/datasets/imagenet', dtype='float16', hard_weight=0.5, input_size=224, label_smoothing=True, last_gamma=True, log_interval=50, logging_file='mobilenetv3_large.log', lr=2.6, lr_decay=0.1, lr_decay_epoch='40,60', lr_decay_period=0, lr_mode='cosine', mixup=False, mixup_alpha=0.2, mixup_off_epoch=0, mode='hybrid', model='mobilenetv3_large', momentum=0.9, no_wd=True, num_epochs=360, num_gpus=8, num_workers=96, rec_train='/media/ramdisk/rec/train.rec', rec_train_idx='/media/ramdisk/rec/train.idx', rec_val='/media/ramdisk/rec/val.rec', rec_val_idx='/media/ramdisk/rec/val.idx', resume_epoch=0, resume_params='', resume_states='', save_dir='params_mobilenetv3_large', save_frequency=10, teacher=None, temperature=20, use_gn=False, use_pretrained=False, use_rec=True, use_se=False, warmup_epochs=5, warmup_lr=0.0, wd=3e-05)
Epoch[0] Batch [49]	Speed: 3665.393671 samples/sec	accuracy=0.001152	lr=0.041613
Epoch[0] Batch [99]	Speed: 8594.810600 samples/sec	accuracy=0.001338	lr=0.083227
Epoch[0] Batch [149]	Speed: 9071.735556 samples/sec	accuracy=0.002295	lr=0.124840
Epoch[0] Batch [199]	Speed: 8837.865532 samples/sec	accuracy=0.004197	lr=0.166453
Epoch[0] Batch [249]	Speed: 8935.284964 samples/sec	accuracy=0.007172	lr=0.208067
Epoch[0] Batch [299]	Speed: 9003.796814 samples/sec	accuracy=0.011502	lr=0.249680
Epoch[0] Batch [349]	Speed: 9137.496123 samples/sec	accuracy=0.016828	lr=0.291293
Epoch[0] Batch [399]	Speed: 9158.263452 samples/sec	accuracy=0.023375	lr=0.332907
Epoch[0] Batch [449]	Speed: 9235.061843 samples/sec	accuracy=0.030731	lr=0.374520
Epoch[0] Batch [499]	Speed: 9236.162861 samples/sec	accuracy=0.038631	lr=0.416133
Epoch[0] Batch [549]	Speed: 9250.063725 samples/sec	accuracy=0.046935	lr=0.457746
Epoch[0] Batch [599]	Speed: 9217.892454 samples/sec	accuracy=0.055493	lr=0.499360
[Epoch 0] training: accuracy=0.059901
[Epoch 0] speed: 8098 samples/sec	time cost: 163.092042
[Epoch 0] validation: err-top1=0.834141 err-top5=0.625020
Epoch[1] Batch [49]	Speed: 8814.844286 samples/sec	accuracy=0.176748	lr=0.562612
Epoch[1] Batch [99]	Speed: 9195.570694 samples/sec	accuracy=0.184170	lr=0.604225
Epoch[1] Batch [149]	Speed: 9029.040297 samples/sec	accuracy=0.191022	lr=0.645839
Epoch[1] Batch [199]	Speed: 8961.178847 samples/sec	accuracy=0.197852	lr=0.687452
Epoch[1] Batch [249]	Speed: 8601.401981 samples/sec	accuracy=0.202191	lr=0.729065
Epoch[1] Batch [299]	Speed: 8342.156202 samples/sec	accuracy=0.208568	lr=0.770679
Epoch[1] Batch [349]	Speed: 8848.652922 samples/sec	accuracy=0.214856	lr=0.812292
Epoch[1] Batch [399]	Speed: 8227.384369 samples/sec	accuracy=0.220872	lr=0.853905
Epoch[1] Batch [449]	Speed: 8474.723439 samples/sec	accuracy=0.226475	lr=0.895519
Epoch[1] Batch [499]	Speed: 8617.293119 samples/sec	accuracy=0.232195	lr=0.937132
Epoch[1] Batch [549]	Speed: 7749.377864 samples/sec	accuracy=0.237211	lr=0.978745
Epoch[1] Batch [599]	Speed: 8572.018217 samples/sec	accuracy=0.242322	lr=1.020359
[Epoch 1] training: accuracy=0.244900
[Epoch 1] speed: 8614 samples/sec	time cost: 153.257802
[Epoch 1] validation: err-top1=0.668945 err-top5=0.404643
Epoch[2] Batch [49]	Speed: 8565.477413 samples/sec	accuracy=0.317607	lr=1.083611
Epoch[2] Batch [99]	Speed: 8363.702788 samples/sec	accuracy=0.319736	lr=1.125224
Epoch[2] Batch [149]	Speed: 8818.494810 samples/sec	accuracy=0.321667	lr=1.166837
Epoch[2] Batch [199]	Speed: 8992.245121 samples/sec	accuracy=0.324473	lr=1.208451
Epoch[2] Batch [249]	Speed: 8483.077339 samples/sec	accuracy=0.326918	lr=1.250064
Epoch[2] Batch [299]	Speed: 8672.385738 samples/sec	accuracy=0.330003	lr=1.291677
Epoch[2] Batch [349]	Speed: 8312.371700 samples/sec	accuracy=0.332407	lr=1.333291
Epoch[2] Batch [399]	Speed: 8212.561008 samples/sec	accuracy=0.335219	lr=1.374904
Epoch[2] Batch [449]	Speed: 8655.632400 samples/sec	accuracy=0.337595	lr=1.416517
Epoch[2] Batch [499]	Speed: 8094.950546 samples/sec	accuracy=0.340018	lr=1.458131
Epoch[2] Batch [549]	Speed: 8353.094460 samples/sec	accuracy=0.342100	lr=1.499744
Epoch[2] Batch [599]	Speed: 9011.211430 samples/sec	accuracy=0.344740	lr=1.541357
[Epoch 2] training: accuracy=0.345817
[Epoch 2] speed: 8520 samples/sec	time cost: 154.734905
[Epoch 2] validation: err-top1=0.604590 err-top5=0.336426
Epoch[3] Batch [49]	Speed: 8576.591645 samples/sec	accuracy=0.385645	lr=1.603777
Epoch[3] Batch [99]	Speed: 8799.776527 samples/sec	accuracy=0.384937	lr=1.645391
Epoch[3] Batch [149]	Speed: 8546.342264 samples/sec	accuracy=0.384945	lr=1.687004
Epoch[3] Batch [199]	Speed: 8531.566562 samples/sec	accuracy=0.385586	lr=1.728617
Epoch[3] Batch [249]	Speed: 8747.469164 samples/sec	accuracy=0.386672	lr=1.770230
Epoch[3] Batch [299]	Speed: 8723.964346 samples/sec	accuracy=0.387300	lr=1.811844
Epoch[3] Batch [349]	Speed: 8800.613896 samples/sec	accuracy=0.389176	lr=1.853457
Epoch[3] Batch [399]	Speed: 8724.889082 samples/sec	accuracy=0.390408	lr=1.895070
Epoch[3] Batch [449]	Speed: 8841.325481 samples/sec	accuracy=0.391510	lr=1.936684
Epoch[3] Batch [499]	Speed: 8671.140867 samples/sec	accuracy=0.392583	lr=1.978297
Epoch[3] Batch [549]	Speed: 8709.454435 samples/sec	accuracy=0.393463	lr=2.019910
Epoch[3] Batch [599]	Speed: 8801.414811 samples/sec	accuracy=0.394462	lr=2.061524
[Epoch 3] training: accuracy=0.394839
[Epoch 3] speed: 8699 samples/sec	time cost: 151.758688
[Epoch 3] validation: err-top1=0.595703 err-top5=0.335571
Epoch[4] Batch [49]	Speed: 8848.102400 samples/sec	accuracy=0.416904	lr=2.124776
Epoch[4] Batch [99]	Speed: 8835.611418 samples/sec	accuracy=0.417314	lr=2.166389
Epoch[4] Batch [149]	Speed: 8802.658042 samples/sec	accuracy=0.416797	lr=2.208003
Epoch[4] Batch [199]	Speed: 8665.369755 samples/sec	accuracy=0.416147	lr=2.249616
Epoch[4] Batch [249]	Speed: 8905.782810 samples/sec	accuracy=0.415459	lr=2.291229
Epoch[4] Batch [299]	Speed: 8756.738195 samples/sec	accuracy=0.415744	lr=2.332843
Epoch[4] Batch [349]	Speed: 8762.495288 samples/sec	accuracy=0.415862	lr=2.374456
Epoch[4] Batch [399]	Speed: 8486.086281 samples/sec	accuracy=0.416088	lr=2.416069
Epoch[4] Batch [449]	Speed: 8738.068445 samples/sec	accuracy=0.416534	lr=2.457682
Epoch[4] Batch [499]	Speed: 8848.465518 samples/sec	accuracy=0.416799	lr=2.499296
Epoch[4] Batch [549]	Speed: 8818.467288 samples/sec	accuracy=0.417190	lr=2.540909
Epoch[4] Batch [599]	Speed: 8524.670905 samples/sec	accuracy=0.417494	lr=2.582522
[Epoch 4] training: accuracy=0.417655
[Epoch 4] speed: 8754 samples/sec	time cost: 150.821553
[Epoch 4] validation: err-top1=0.562695 err-top5=0.298066
Epoch[5] Batch [49]	Speed: 8694.517289 samples/sec	accuracy=0.431396	lr=2.600000
Epoch[5] Batch [99]	Speed: 8720.173877 samples/sec	accuracy=0.431338	lr=2.599999
Epoch[5] Batch [149]	Speed: 8542.487226 samples/sec	accuracy=0.431820	lr=2.599997
Epoch[5] Batch [199]	Speed: 8748.119310 samples/sec	accuracy=0.431675	lr=2.599995
Epoch[5] Batch [249]	Speed: 9137.709773 samples/sec	accuracy=0.432666	lr=2.599992
Epoch[5] Batch [299]	Speed: 8818.018821 samples/sec	accuracy=0.433291	lr=2.599988
Epoch[5] Batch [349]	Speed: 8665.741633 samples/sec	accuracy=0.433969	lr=2.599984
Epoch[5] Batch [399]	Speed: 8733.357101 samples/sec	accuracy=0.434678	lr=2.599979
Epoch[5] Batch [449]	Speed: 8450.214853 samples/sec	accuracy=0.435237	lr=2.599973
Epoch[5] Batch [499]	Speed: 8741.902781 samples/sec	accuracy=0.435854	lr=2.599967
Epoch[5] Batch [549]	Speed: 8697.167713 samples/sec	accuracy=0.435636	lr=2.599960
Epoch[5] Batch [599]	Speed: 8719.631967 samples/sec	accuracy=0.435606	lr=2.599953
[Epoch 5] training: accuracy=0.435829
[Epoch 5] speed: 8715 samples/sec	time cost: 151.510643
[Epoch 5] validation: err-top1=0.520447 err-top5=0.258708
Epoch[6] Batch [49]	Speed: 8587.087114 samples/sec	accuracy=0.451396	lr=2.599940
Epoch[6] Batch [99]	Speed: 8706.491705 samples/sec	accuracy=0.450166	lr=2.599931
Epoch[6] Batch [149]	Speed: 8760.641294 samples/sec	accuracy=0.450843	lr=2.599921
Epoch[6] Batch [199]	Speed: 8716.581288 samples/sec	accuracy=0.451008	lr=2.599910
Epoch[6] Batch [249]	Speed: 8700.414720 samples/sec	accuracy=0.451594	lr=2.599899
Epoch[6] Batch [299]	Speed: 8964.261162 samples/sec	accuracy=0.451904	lr=2.599888
Epoch[6] Batch [349]	Speed: 8829.263069 samples/sec	accuracy=0.451984	lr=2.599875
Epoch[6] Batch [399]	Speed: 8717.221012 samples/sec	accuracy=0.452704	lr=2.599862
Epoch[6] Batch [449]	Speed: 8717.915510 samples/sec	accuracy=0.453204	lr=2.599848
Epoch[6] Batch [499]	Speed: 8708.334501 samples/sec	accuracy=0.453189	lr=2.599834
Epoch[6] Batch [549]	Speed: 8473.555894 samples/sec	accuracy=0.453546	lr=2.599819
Epoch[6] Batch [599]	Speed: 8803.242258 samples/sec	accuracy=0.454051	lr=2.599803
[Epoch 6] training: accuracy=0.454088
[Epoch 6] speed: 8722 samples/sec	time cost: 151.287979
[Epoch 6] validation: err-top1=0.500122 err-top5=0.242635
Epoch[7] Batch [49]	Speed: 8790.187007 samples/sec	accuracy=0.468418	lr=2.599778
Epoch[7] Batch [99]	Speed: 8757.146347 samples/sec	accuracy=0.468003	lr=2.599761
Epoch[7] Batch [149]	Speed: 8853.425277 samples/sec	accuracy=0.467900	lr=2.599743
Epoch[7] Batch [199]	Speed: 8703.031531 samples/sec	accuracy=0.466870	lr=2.599725
Epoch[7] Batch [249]	Speed: 8754.541687 samples/sec	accuracy=0.466572	lr=2.599705
Epoch[7] Batch [299]	Speed: 8762.350486 samples/sec	accuracy=0.466436	lr=2.599685
Epoch[7] Batch [349]	Speed: 8713.920611 samples/sec	accuracy=0.466410	lr=2.599665
Epoch[7] Batch [399]	Speed: 8740.437941 samples/sec	accuracy=0.466285	lr=2.599644
Epoch[7] Batch [449]	Speed: 8780.326130 samples/sec	accuracy=0.466137	lr=2.599622
Epoch[7] Batch [499]	Speed: 8735.541391 samples/sec	accuracy=0.466029	lr=2.599599
Epoch[7] Batch [549]	Speed: 8790.864570 samples/sec	accuracy=0.466244	lr=2.599576
Epoch[7] Batch [599]	Speed: 8769.245851 samples/sec	accuracy=0.465862	lr=2.599552
[Epoch 7] training: accuracy=0.465981
[Epoch 7] speed: 8767 samples/sec	time cost: 150.988066
[Epoch 7] validation: err-top1=0.507812 err-top5=0.251582
Epoch[8] Batch [49]	Speed: 8956.200190 samples/sec	accuracy=0.481475	lr=2.599515
Epoch[8] Batch [99]	Speed: 8701.084508 samples/sec	accuracy=0.478804	lr=2.599489
Epoch[8] Batch [149]	Speed: 8697.732021 samples/sec	accuracy=0.476706	lr=2.599463
Epoch[8] Batch [199]	Speed: 8708.793071 samples/sec	accuracy=0.476184	lr=2.599436
Epoch[8] Batch [249]	Speed: 8845.195069 samples/sec	accuracy=0.475910	lr=2.599409
Epoch[8] Batch [299]	Speed: 8744.671555 samples/sec	accuracy=0.475758	lr=2.599381
Epoch[8] Batch [349]	Speed: 8811.450583 samples/sec	accuracy=0.475711	lr=2.599352
Epoch[8] Batch [399]	Speed: 8818.317010 samples/sec	accuracy=0.475610	lr=2.599323
Epoch[8] Batch [449]	Speed: 8694.002849 samples/sec	accuracy=0.475295	lr=2.599293
Epoch[8] Batch [499]	Speed: 8793.174397 samples/sec	accuracy=0.475370	lr=2.599262
Epoch[8] Batch [549]	Speed: 8550.352786 samples/sec	accuracy=0.475716	lr=2.599231
Epoch[8] Batch [599]	Speed: 8679.649006 samples/sec	accuracy=0.475641	lr=2.599199
[Epoch 8] training: accuracy=0.475722
[Epoch 8] speed: 8753 samples/sec	time cost: 150.863210
[Epoch 8] validation: err-top1=0.502686 err-top5=0.248657
Epoch[9] Batch [49]	Speed: 8545.783145 samples/sec	accuracy=0.494443	lr=2.599149
Epoch[9] Batch [99]	Speed: 8652.127120 samples/sec	accuracy=0.489961	lr=2.599115
Epoch[9] Batch [149]	Speed: 8753.190167 samples/sec	accuracy=0.487210	lr=2.599081
Epoch[9] Batch [199]	Speed: 8678.219858 samples/sec	accuracy=0.486255	lr=2.599046
Epoch[9] Batch [249]	Speed: 8722.652010 samples/sec	accuracy=0.484996	lr=2.599010
Epoch[9] Batch [299]	Speed: 8712.978932 samples/sec	accuracy=0.484832	lr=2.598974
Epoch[9] Batch [349]	Speed: 8398.036108 samples/sec	accuracy=0.484097	lr=2.598937
Epoch[9] Batch [399]	Speed: 9012.079123 samples/sec	accuracy=0.483412	lr=2.598900
Epoch[9] Batch [449]	Speed: 8780.122763 samples/sec	accuracy=0.482642	lr=2.598861
Epoch[9] Batch [499]	Speed: 8779.096736 samples/sec	accuracy=0.482216	lr=2.598823
Epoch[9] Batch [549]	Speed: 8720.384215 samples/sec	accuracy=0.482011	lr=2.598783
Epoch[9] Batch [599]	Speed: 8983.534125 samples/sec	accuracy=0.481966	lr=2.598743
[Epoch 9] training: accuracy=0.481909
[Epoch 9] speed: 8726 samples/sec	time cost: 151.247604
[Epoch 9] validation: err-top1=0.468418 err-top5=0.218672
Epoch[10] Batch [49]	Speed: 8591.012592 samples/sec	accuracy=0.495059	lr=2.598682
Epoch[10] Batch [99]	Speed: 8736.426822 samples/sec	accuracy=0.491987	lr=2.598640
Epoch[10] Batch [149]	Speed: 8922.487478 samples/sec	accuracy=0.491042	lr=2.598597
Epoch[10] Batch [199]	Speed: 8767.822129 samples/sec	accuracy=0.490198	lr=2.598554
Epoch[10] Batch [249]	Speed: 8961.664246 samples/sec	accuracy=0.489600	lr=2.598511
Epoch[10] Batch [299]	Speed: 8772.189818 samples/sec	accuracy=0.489396	lr=2.598466
Epoch[10] Batch [349]	Speed: 8801.784028 samples/sec	accuracy=0.489121	lr=2.598421
Epoch[10] Batch [399]	Speed: 8855.214503 samples/sec	accuracy=0.489069	lr=2.598376
Epoch[10] Batch [449]	Speed: 8636.715534 samples/sec	accuracy=0.489002	lr=2.598329
Epoch[10] Batch [499]	Speed: 8725.301538 samples/sec	accuracy=0.488840	lr=2.598282
Epoch[10] Batch [549]	Speed: 8644.884765 samples/sec	accuracy=0.488737	lr=2.598235
Epoch[10] Batch [599]	Speed: 8777.406657 samples/sec	accuracy=0.488502	lr=2.598186
[Epoch 10] training: accuracy=0.488668
[Epoch 10] speed: 8772 samples/sec	time cost: 150.535756
[Epoch 10] validation: err-top1=0.493327 err-top5=0.241211
Epoch[11] Batch [49]	Speed: 8833.550306 samples/sec	accuracy=0.502090	lr=2.598112
Epoch[11] Batch [99]	Speed: 9046.387249 samples/sec	accuracy=0.499419	lr=2.598062
Epoch[11] Batch [149]	Speed: 9118.858282 samples/sec	accuracy=0.497871	lr=2.598011
Epoch[11] Batch [199]	Speed: 8652.352316 samples/sec	accuracy=0.496611	lr=2.597960
Epoch[11] Batch [249]	Speed: 8889.162836 samples/sec	accuracy=0.495715	lr=2.597908
Epoch[11] Batch [299]	Speed: 8868.800911 samples/sec	accuracy=0.495467	lr=2.597856
Epoch[11] Batch [349]	Speed: 8800.359639 samples/sec	accuracy=0.494520	lr=2.597803
Epoch[11] Batch [399]	Speed: 8779.341331 samples/sec	accuracy=0.494637	lr=2.597749
Epoch[11] Batch [449]	Speed: 8584.045765 samples/sec	accuracy=0.494602	lr=2.597694
Epoch[11] Batch [499]	Speed: 8959.923524 samples/sec	accuracy=0.494016	lr=2.597639
Epoch[11] Batch [549]	Speed: 8539.886926 samples/sec	accuracy=0.493803	lr=2.597583
Epoch[11] Batch [599]	Speed: 8923.335758 samples/sec	accuracy=0.493485	lr=2.597527
[Epoch 11] training: accuracy=0.493443
[Epoch 11] speed: 8831 samples/sec	time cost: 149.319424
[Epoch 11] validation: err-top1=0.464030 err-top5=0.215373
Epoch[12] Batch [49]	Speed: 8823.653587 samples/sec	accuracy=0.504297	lr=2.597441
Epoch[12] Batch [99]	Speed: 9045.666296 samples/sec	accuracy=0.502227	lr=2.597383
Epoch[12] Batch [149]	Speed: 8881.452827 samples/sec	accuracy=0.500553	lr=2.597324
Epoch[12] Batch [199]	Speed: 8800.443669 samples/sec	accuracy=0.500251	lr=2.597265
Epoch[12] Batch [249]	Speed: 8919.727050 samples/sec	accuracy=0.499887	lr=2.597205
Epoch[12] Batch [299]	Speed: 8712.042051 samples/sec	accuracy=0.499165	lr=2.597144
Epoch[12] Batch [349]	Speed: 8440.736306 samples/sec	accuracy=0.498961	lr=2.597083
Epoch[12] Batch [399]	Speed: 8905.249899 samples/sec	accuracy=0.498639	lr=2.597021
Epoch[12] Batch [449]	Speed: 8430.726560 samples/sec	accuracy=0.498167	lr=2.596959
Epoch[12] Batch [499]	Speed: 8579.868347 samples/sec	accuracy=0.497723	lr=2.596895
Epoch[12] Batch [549]	Speed: 8664.012593 samples/sec	accuracy=0.497329	lr=2.596831
Epoch[12] Batch [599]	Speed: 8739.656265 samples/sec	accuracy=0.497128	lr=2.596767
[Epoch 12] training: accuracy=0.497157
[Epoch 12] speed: 8740 samples/sec	time cost: 151.573766
[Epoch 12] validation: err-top1=0.451934 err-top5=0.205449
Epoch[13] Batch [49]	Speed: 8638.405202 samples/sec	accuracy=0.508447	lr=2.596667
Epoch[13] Batch [99]	Speed: 8727.107973 samples/sec	accuracy=0.505762	lr=2.596601
Epoch[13] Batch [149]	Speed: 8900.865685 samples/sec	accuracy=0.504375	lr=2.596534
Epoch[13] Batch [199]	Speed: 8912.787114 samples/sec	accuracy=0.503076	lr=2.596467
Epoch[13] Batch [249]	Speed: 8736.613953 samples/sec	accuracy=0.502752	lr=2.596399
Epoch[13] Batch [299]	Speed: 8537.437038 samples/sec	accuracy=0.502311	lr=2.596330
Epoch[13] Batch [349]	Speed: 8680.167888 samples/sec	accuracy=0.501925	lr=2.596261
Epoch[13] Batch [399]	Speed: 8697.452852 samples/sec	accuracy=0.501339	lr=2.596191
Epoch[13] Batch [449]	Speed: 8702.312778 samples/sec	accuracy=0.500576	lr=2.596120
Epoch[13] Batch [499]	Speed: 8689.739201 samples/sec	accuracy=0.500407	lr=2.596048
Epoch[13] Batch [549]	Speed: 8715.308842 samples/sec	accuracy=0.500244	lr=2.595976
Epoch[13] Batch [599]	Speed: 8647.343440 samples/sec	accuracy=0.500165	lr=2.595904
[Epoch 13] training: accuracy=0.500268
[Epoch 13] speed: 8716 samples/sec	time cost: 151.783328
[Epoch 13] validation: err-top1=0.449015 err-top5=0.204163
Epoch[14] Batch [49]	Speed: 8588.089527 samples/sec	accuracy=0.511475	lr=2.595793
Epoch[14] Batch [99]	Speed: 8871.676870 samples/sec	accuracy=0.509385	lr=2.595719
Epoch[14] Batch [149]	Speed: 8556.986259 samples/sec	accuracy=0.507835	lr=2.595644
Epoch[14] Batch [199]	Speed: 8734.169799 samples/sec	accuracy=0.506738	lr=2.595569
Epoch[14] Batch [249]	Speed: 8517.528277 samples/sec	accuracy=0.505775	lr=2.595492
Epoch[14] Batch [299]	Speed: 8475.793957 samples/sec	accuracy=0.505386	lr=2.595415
Epoch[14] Batch [349]	Speed: 8704.321034 samples/sec	accuracy=0.504542	lr=2.595338
Epoch[14] Batch [399]	Speed: 8826.428142 samples/sec	accuracy=0.504696	lr=2.595260
Epoch[14] Batch [449]	Speed: 8835.365859 samples/sec	accuracy=0.504460	lr=2.595181
Epoch[14] Batch [499]	Speed: 8934.576964 samples/sec	accuracy=0.504065	lr=2.595101
Epoch[14] Batch [549]	Speed: 8620.735964 samples/sec	accuracy=0.504187	lr=2.595021
Epoch[14] Batch [599]	Speed: 8487.186338 samples/sec	accuracy=0.503969	lr=2.594940
[Epoch 14] training: accuracy=0.503819
[Epoch 14] speed: 8676 samples/sec	time cost: 152.824939
[Epoch 14] validation: err-top1=0.479902 err-top5=0.222617
Epoch[15] Batch [49]	Speed: 8682.084149 samples/sec	accuracy=0.513760	lr=2.594816
Epoch[15] Batch [99]	Speed: 8503.011536 samples/sec	accuracy=0.512109	lr=2.594734
Epoch[15] Batch [149]	Speed: 8559.199363 samples/sec	accuracy=0.510785	lr=2.594651
Epoch[15] Batch [199]	Speed: 8567.479058 samples/sec	accuracy=0.509343	lr=2.594567
Epoch[15] Batch [249]	Speed: 8669.781196 samples/sec	accuracy=0.508645	lr=2.594483
Epoch[15] Batch [299]	Speed: 8500.443790 samples/sec	accuracy=0.508055	lr=2.594398
Epoch[15] Batch [349]	Speed: 8828.422598 samples/sec	accuracy=0.507489	lr=2.594312
Epoch[15] Batch [399]	Speed: 8889.850408 samples/sec	accuracy=0.506995	lr=2.594226
Epoch[15] Batch [449]	Speed: 8575.530103 samples/sec	accuracy=0.506618	lr=2.594139
Epoch[15] Batch [499]	Speed: 8830.447005 samples/sec	accuracy=0.506274	lr=2.594051
Epoch[15] Batch [549]	Speed: 8772.996141 samples/sec	accuracy=0.505790	lr=2.593963
Epoch[15] Batch [599]	Speed: 8589.889925 samples/sec	accuracy=0.505766	lr=2.593874
[Epoch 15] training: accuracy=0.505691
[Epoch 15] speed: 8664 samples/sec	time cost: 152.578741
[Epoch 15] validation: err-top1=0.492778 err-top5=0.236898
Epoch[16] Batch [49]	Speed: 8537.102392 samples/sec	accuracy=0.514775	lr=2.593738
Epoch[16] Batch [99]	Speed: 8777.625505 samples/sec	accuracy=0.513652	lr=2.593647
Epoch[16] Batch [149]	Speed: 8731.123140 samples/sec	accuracy=0.512526	lr=2.593556
Epoch[16] Batch [199]	Speed: 8810.185349 samples/sec	accuracy=0.511621	lr=2.593464
Epoch[16] Batch [249]	Speed: 8740.940101 samples/sec	accuracy=0.510836	lr=2.593371
Epoch[16] Batch [299]	Speed: 8855.523794 samples/sec	accuracy=0.510490	lr=2.593278
Epoch[16] Batch [349]	Speed: 8635.129304 samples/sec	accuracy=0.510091	lr=2.593185
Epoch[16] Batch [399]	Speed: 8675.165140 samples/sec	accuracy=0.509591	lr=2.593090
Epoch[16] Batch [449]	Speed: 8656.825711 samples/sec	accuracy=0.509022	lr=2.592995
Epoch[16] Batch [499]	Speed: 8794.570169 samples/sec	accuracy=0.508538	lr=2.592899
Epoch[16] Batch [549]	Speed: 8767.622384 samples/sec	accuracy=0.508157	lr=2.592803
Epoch[16] Batch [599]	Speed: 8847.614462 samples/sec	accuracy=0.507970	lr=2.592706
[Epoch 16] training: accuracy=0.507811
[Epoch 16] speed: 8731 samples/sec	time cost: 151.506173
[Epoch 16] validation: err-top1=0.472480 err-top5=0.221152
Epoch[17] Batch [49]	Speed: 8693.484776 samples/sec	accuracy=0.518701	lr=2.592559
Epoch[17] Batch [99]	Speed: 8798.971404 samples/sec	accuracy=0.515996	lr=2.592460
Epoch[17] Batch [149]	Speed: 8523.252415 samples/sec	accuracy=0.514388	lr=2.592361
Epoch[17] Batch [199]	Speed: 8522.245126 samples/sec	accuracy=0.513572	lr=2.592261
Epoch[17] Batch [249]	Speed: 8664.357087 samples/sec	accuracy=0.512748	lr=2.592161
Epoch[17] Batch [299]	Speed: 8710.067148 samples/sec	accuracy=0.512949	lr=2.592059
Epoch[17] Batch [349]	Speed: 8600.417985 samples/sec	accuracy=0.511964	lr=2.591957
Epoch[17] Batch [399]	Speed: 8656.370679 samples/sec	accuracy=0.511281	lr=2.591855
Epoch[17] Batch [449]	Speed: 8504.113965 samples/sec	accuracy=0.510753	lr=2.591752
Epoch[17] Batch [499]	Speed: 8310.476857 samples/sec	accuracy=0.510371	lr=2.591648
Epoch[17] Batch [549]	Speed: 8491.602011 samples/sec	accuracy=0.510014	lr=2.591543
Epoch[17] Batch [599]	Speed: 8657.082560 samples/sec	accuracy=0.510014	lr=2.591438
[Epoch 17] training: accuracy=0.509872
[Epoch 17] speed: 8593 samples/sec	time cost: 154.030408
[Epoch 17] validation: err-top1=0.463969 err-top5=0.214315
Epoch[18] Batch [49]	Speed: 8541.293803 samples/sec	accuracy=0.518828	lr=2.591277
Epoch[18] Batch [99]	Speed: 8824.525785 samples/sec	accuracy=0.519805	lr=2.591170
Epoch[18] Batch [149]	Speed: 8830.523802 samples/sec	accuracy=0.517464	lr=2.591063
Epoch[18] Batch [199]	Speed: 8753.839558 samples/sec	accuracy=0.516958	lr=2.590955
Epoch[18] Batch [249]	Speed: 8592.822124 samples/sec	accuracy=0.516359	lr=2.590846
Epoch[18] Batch [299]	Speed: 8736.466984 samples/sec	accuracy=0.515246	lr=2.590737
Epoch[18] Batch [349]	Speed: 8749.032778 samples/sec	accuracy=0.514466	lr=2.590627
Epoch[18] Batch [399]	Speed: 8585.096375 samples/sec	accuracy=0.513944	lr=2.590516
Epoch[18] Batch [449]	Speed: 8488.259336 samples/sec	accuracy=0.513451	lr=2.590405
Epoch[18] Batch [499]	Speed: 8557.168509 samples/sec	accuracy=0.513192	lr=2.590293
Epoch[18] Batch [549]	Speed: 8827.070123 samples/sec	accuracy=0.512804	lr=2.590180
Epoch[18] Batch [599]	Speed: 8740.172387 samples/sec	accuracy=0.512407	lr=2.590067
[Epoch 18] training: accuracy=0.512363
[Epoch 18] speed: 8681 samples/sec	time cost: 152.112186
[Epoch 18] validation: err-top1=0.475444 err-top5=0.222493
Epoch[19] Batch [49]	Speed: 8501.774085 samples/sec	accuracy=0.518623	lr=2.589896
Epoch[19] Batch [99]	Speed: 8841.850586 samples/sec	accuracy=0.518257	lr=2.589781
Epoch[19] Batch [149]	Speed: 8555.589206 samples/sec	accuracy=0.517842	lr=2.589666
Epoch[19] Batch [199]	Speed: 8765.990750 samples/sec	accuracy=0.516809	lr=2.589550
Epoch[19] Batch [249]	Speed: 8527.102128 samples/sec	accuracy=0.516238	lr=2.589433
Epoch[19] Batch [299]	Speed: 8813.166096 samples/sec	accuracy=0.515256	lr=2.589315
Epoch[19] Batch [349]	Speed: 8853.944885 samples/sec	accuracy=0.514741	lr=2.589197
Epoch[19] Batch [399]	Speed: 8525.396319 samples/sec	accuracy=0.514120	lr=2.589079
Epoch[19] Batch [449]	Speed: 8431.750401 samples/sec	accuracy=0.513809	lr=2.588959
Epoch[19] Batch [499]	Speed: 8771.518892 samples/sec	accuracy=0.513733	lr=2.588839
Epoch[19] Batch [549]	Speed: 8887.804011 samples/sec	accuracy=0.513401	lr=2.588719
Epoch[19] Batch [599]	Speed: 8446.568784 samples/sec	accuracy=0.512988	lr=2.588597
[Epoch 19] training: accuracy=0.512946
[Epoch 19] speed: 8658 samples/sec	time cost: 152.998287
[Epoch 19] validation: err-top1=0.475996 err-top5=0.224727
Epoch[20] Batch [49]	Speed: 8740.564765 samples/sec	accuracy=0.520410	lr=2.588412
Epoch[20] Batch [99]	Speed: 8913.737517 samples/sec	accuracy=0.520068	lr=2.588289
Epoch[20] Batch [149]	Speed: 8900.804629 samples/sec	accuracy=0.520020	lr=2.588165
Epoch[20] Batch [199]	Speed: 8752.602765 samples/sec	accuracy=0.519639	lr=2.588041
Epoch[20] Batch [249]	Speed: 8856.955504 samples/sec	accuracy=0.518666	lr=2.587916
Epoch[20] Batch [299]	Speed: 8906.058524 samples/sec	accuracy=0.517671	lr=2.587790
Epoch[20] Batch [349]	Speed: 8780.527173 samples/sec	accuracy=0.517222	lr=2.587664
Epoch[20] Batch [399]	Speed: 8792.525277 samples/sec	accuracy=0.516575	lr=2.587538
Epoch[20] Batch [449]	Speed: 8725.831743 samples/sec	accuracy=0.516359	lr=2.587410
Epoch[20] Batch [499]	Speed: 8692.670659 samples/sec	accuracy=0.516031	lr=2.587282
Epoch[20] Batch [549]	Speed: 8744.806158 samples/sec	accuracy=0.515630	lr=2.587153
Epoch[20] Batch [599]	Speed: 8763.276763 samples/sec	accuracy=0.515480	lr=2.587024
[Epoch 20] training: accuracy=0.515515
[Epoch 20] speed: 8793 samples/sec	time cost: 150.179572
[Epoch 20] validation: err-top1=0.443176 err-top5=0.196452
Epoch[21] Batch [49]	Speed: 8699.517190 samples/sec	accuracy=0.522676	lr=2.586829
Epoch[21] Batch [99]	Speed: 9003.054325 samples/sec	accuracy=0.521997	lr=2.586698
Epoch[21] Batch [149]	Speed: 8702.769656 samples/sec	accuracy=0.520960	lr=2.586566
Epoch[21] Batch [199]	Speed: 8808.189911 samples/sec	accuracy=0.520010	lr=2.586434
Epoch[21] Batch [249]	Speed: 8620.132468 samples/sec	accuracy=0.519596	lr=2.586301
Epoch[21] Batch [299]	Speed: 8907.342948 samples/sec	accuracy=0.519139	lr=2.586167
Epoch[21] Batch [349]	Speed: 8401.072435 samples/sec	accuracy=0.518836	lr=2.586033
Epoch[21] Batch [399]	Speed: 8630.478682 samples/sec	accuracy=0.517828	lr=2.585898
Epoch[21] Batch [449]	Speed: 8664.410573 samples/sec	accuracy=0.517326	lr=2.585762
Epoch[21] Batch [499]	Speed: 8563.940122 samples/sec	accuracy=0.516883	lr=2.585626
Epoch[21] Batch [549]	Speed: 8677.525535 samples/sec	accuracy=0.516191	lr=2.585489
Epoch[21] Batch [599]	Speed: 8692.766543 samples/sec	accuracy=0.515843	lr=2.585352
[Epoch 21] training: accuracy=0.515682
[Epoch 21] speed: 8692 samples/sec	time cost: 152.341206
[Epoch 21] validation: err-top1=0.494121 err-top5=0.238672
Epoch[22] Batch [49]	Speed: 8649.890617 samples/sec	accuracy=0.526865	lr=2.585142
Epoch[22] Batch [99]	Speed: 8615.439221 samples/sec	accuracy=0.524839	lr=2.585003
Epoch[22] Batch [149]	Speed: 8911.824340 samples/sec	accuracy=0.523167	lr=2.584863
Epoch[22] Batch [199]	Speed: 8588.534661 samples/sec	accuracy=0.522432	lr=2.584723
Epoch[22] Batch [249]	Speed: 8818.273919 samples/sec	accuracy=0.521717	lr=2.584582
Epoch[22] Batch [299]	Speed: 8701.815752 samples/sec	accuracy=0.520715	lr=2.584440
Epoch[22] Batch [349]	Speed: 8670.160804 samples/sec	accuracy=0.519810	lr=2.584298
Epoch[22] Batch [399]	Speed: 8846.030902 samples/sec	accuracy=0.519039	lr=2.584155
Epoch[22] Batch [449]	Speed: 8671.546155 samples/sec	accuracy=0.518719	lr=2.584011
Epoch[22] Batch [499]	Speed: 8775.609643 samples/sec	accuracy=0.518228	lr=2.583867
Epoch[22] Batch [549]	Speed: 8990.215298 samples/sec	accuracy=0.517997	lr=2.583722
Epoch[22] Batch [599]	Speed: 8859.709559 samples/sec	accuracy=0.517525	lr=2.583577
[Epoch 22] training: accuracy=0.517389
[Epoch 22] speed: 8761 samples/sec	time cost: 150.907442
[Epoch 22] validation: err-top1=0.429281 err-top5=0.186646
Epoch[23] Batch [49]	Speed: 9004.609087 samples/sec	accuracy=0.528506	lr=2.583354
Epoch[23] Batch [99]	Speed: 8818.913265 samples/sec	accuracy=0.526108	lr=2.583207
Epoch[23] Batch [149]	Speed: 8805.747977 samples/sec	accuracy=0.524284	lr=2.583059
Epoch[23] Batch [199]	Speed: 9013.894843 samples/sec	accuracy=0.523005	lr=2.582911
Epoch[23] Batch [249]	Speed: 8852.276406 samples/sec	accuracy=0.522184	lr=2.582762
Epoch[23] Batch [299]	Speed: 8998.988095 samples/sec	accuracy=0.521370	lr=2.582612
Epoch[23] Batch [349]	Speed: 8856.105918 samples/sec	accuracy=0.520579	lr=2.582462
Epoch[23] Batch [399]	Speed: 8901.896203 samples/sec	accuracy=0.520055	lr=2.582311
Epoch[23] Batch [449]	Speed: 8854.696024 samples/sec	accuracy=0.519488	lr=2.582159
Epoch[23] Batch [499]	Speed: 8738.286225 samples/sec	accuracy=0.518912	lr=2.582007
Epoch[23] Batch [549]	Speed: 8769.087219 samples/sec	accuracy=0.518762	lr=2.581854
Epoch[23] Batch [599]	Speed: 8873.511247 samples/sec	accuracy=0.518341	lr=2.581700
[Epoch 23] training: accuracy=0.518130
[Epoch 23] speed: 8872 samples/sec	time cost: 149.024334
[Epoch 23] validation: err-top1=0.444661 err-top5=0.195475
Epoch[24] Batch [49]	Speed: 8777.783728 samples/sec	accuracy=0.530039	lr=2.581469
Epoch[24] Batch [99]	Speed: 9046.534921 samples/sec	accuracy=0.525586	lr=2.581314
Epoch[24] Batch [149]	Speed: 8921.434396 samples/sec	accuracy=0.524313	lr=2.581158
Epoch[24] Batch [199]	Speed: 9059.374912 samples/sec	accuracy=0.524426	lr=2.581001
Epoch[24] Batch [249]	Speed: 8728.266088 samples/sec	accuracy=0.523186	lr=2.580844
Epoch[24] Batch [299]	Speed: 8812.832994 samples/sec	accuracy=0.522480	lr=2.580687
Epoch[24] Batch [349]	Speed: 8647.123032 samples/sec	accuracy=0.521770	lr=2.580528
Epoch[24] Batch [399]	Speed: 8619.507605 samples/sec	accuracy=0.521682	lr=2.580369
Epoch[24] Batch [449]	Speed: 8679.114402 samples/sec	accuracy=0.521593	lr=2.580210
Epoch[24] Batch [499]	Speed: 9132.377738 samples/sec	accuracy=0.520938	lr=2.580049
Epoch[24] Batch [549]	Speed: 8852.284070 samples/sec	accuracy=0.520605	lr=2.579888
Epoch[24] Batch [599]	Speed: 9141.317860 samples/sec	accuracy=0.520474	lr=2.579727
[Epoch 24] training: accuracy=0.520349
[Epoch 24] speed: 8867 samples/sec	time cost: 149.586171
[Epoch 24] validation: err-top1=0.479805 err-top5=0.222559
Epoch[25] Batch [49]	Speed: 9111.583306 samples/sec	accuracy=0.527393	lr=2.579480
Epoch[25] Batch [99]	Speed: 9085.564491 samples/sec	accuracy=0.527158	lr=2.579317
Epoch[25] Batch [149]	Speed: 8743.102914 samples/sec	accuracy=0.526022	lr=2.579153
Epoch[25] Batch [199]	Speed: 8697.868706 samples/sec	accuracy=0.524138	lr=2.578988
Epoch[25] Batch [249]	Speed: 8804.842483 samples/sec	accuracy=0.523234	lr=2.578823
Epoch[25] Batch [299]	Speed: 8904.085325 samples/sec	accuracy=0.522822	lr=2.578657
Epoch[25] Batch [349]	Speed: 8918.368309 samples/sec	accuracy=0.522674	lr=2.578491
Epoch[25] Batch [399]	Speed: 8707.309825 samples/sec	accuracy=0.522361	lr=2.578324
Epoch[25] Batch [449]	Speed: 8582.111997 samples/sec	accuracy=0.521755	lr=2.578156
Epoch[25] Batch [499]	Speed: 8724.990819 samples/sec	accuracy=0.521535	lr=2.577988
Epoch[25] Batch [549]	Speed: 8618.288070 samples/sec	accuracy=0.521406	lr=2.577819
Epoch[25] Batch [599]	Speed: 8715.904691 samples/sec	accuracy=0.520969	lr=2.577649
[Epoch 25] training: accuracy=0.520895
[Epoch 25] speed: 8797 samples/sec	time cost: 150.228608
[Epoch 25] validation: err-top1=0.448771 err-top5=0.204732
Epoch[26] Batch [49]	Speed: 8811.358751 samples/sec	accuracy=0.530752	lr=2.577394
Epoch[26] Batch [99]	Speed: 8902.922531 samples/sec	accuracy=0.529756	lr=2.577223
Epoch[26] Batch [149]	Speed: 9171.712064 samples/sec	accuracy=0.528249	lr=2.577051
Epoch[26] Batch [199]	Speed: 8881.315086 samples/sec	accuracy=0.526130	lr=2.576878
Epoch[26] Batch [249]	Speed: 8914.216124 samples/sec	accuracy=0.525697	lr=2.576705
Epoch[26] Batch [299]	Speed: 8617.029808 samples/sec	accuracy=0.524650	lr=2.576531
Epoch[26] Batch [349]	Speed: 8751.431407 samples/sec	accuracy=0.523916	lr=2.576357
Epoch[26] Batch [399]	Speed: 8985.261289 samples/sec	accuracy=0.523273	lr=2.576182
Epoch[26] Batch [449]	Speed: 9053.952940 samples/sec	accuracy=0.522998	lr=2.576006
Epoch[26] Batch [499]	Speed: 8990.967528 samples/sec	accuracy=0.522625	lr=2.575830
Epoch[26] Batch [549]	Speed: 8848.989467 samples/sec	accuracy=0.522251	lr=2.575653
Epoch[26] Batch [599]	Speed: 8955.314467 samples/sec	accuracy=0.521543	lr=2.575475
[Epoch 26] training: accuracy=0.521465
[Epoch 26] speed: 8898 samples/sec	time cost: 149.035552
[Epoch 26] validation: err-top1=0.438789 err-top5=0.195449
Epoch[27] Batch [49]	Speed: 8919.039478 samples/sec	accuracy=0.534365	lr=2.575204
Epoch[27] Batch [99]	Speed: 8789.084884 samples/sec	accuracy=0.530269	lr=2.575025
Epoch[27] Batch [149]	Speed: 8446.412642 samples/sec	accuracy=0.528799	lr=2.574845
Epoch[27] Batch [199]	Speed: 8871.884501 samples/sec	accuracy=0.527979	lr=2.574665
Epoch[27] Batch [249]	Speed: 9200.230601 samples/sec	accuracy=0.527256	lr=2.574483
Epoch[27] Batch [299]	Speed: 8606.979682 samples/sec	accuracy=0.526291	lr=2.574302
Epoch[27] Batch [349]	Speed: 8467.823330 samples/sec	accuracy=0.526045	lr=2.574119
Epoch[27] Batch [399]	Speed: 8615.235644 samples/sec	accuracy=0.525194	lr=2.573936
Epoch[27] Batch [449]	Speed: 8947.315795 samples/sec	accuracy=0.524426	lr=2.573752
Epoch[27] Batch [499]	Speed: 8822.345341 samples/sec	accuracy=0.523985	lr=2.573568
Epoch[27] Batch [549]	Speed: 8851.424070 samples/sec	accuracy=0.523735	lr=2.573383
Epoch[27] Batch [599]	Speed: 8591.104185 samples/sec	accuracy=0.523193	lr=2.573198
[Epoch 27] training: accuracy=0.523060
[Epoch 27] speed: 8761 samples/sec	time cost: 150.808162
[Epoch 27] validation: err-top1=0.432576 err-top5=0.192668
Epoch[28] Batch [49]	Speed: 8600.351510 samples/sec	accuracy=0.532480	lr=2.572918
Epoch[28] Batch [99]	Speed: 8886.911720 samples/sec	accuracy=0.530469	lr=2.572731
Epoch[28] Batch [149]	Speed: 8452.063185 samples/sec	accuracy=0.529258	lr=2.572543
Epoch[28] Batch [199]	Speed: 8969.337197 samples/sec	accuracy=0.528254	lr=2.572354
Epoch[28] Batch [249]	Speed: 8644.734428 samples/sec	accuracy=0.526852	lr=2.572165
Epoch[28] Batch [299]	Speed: 8962.603219 samples/sec	accuracy=0.526304	lr=2.571976
Epoch[28] Batch [349]	Speed: 8993.520441 samples/sec	accuracy=0.525837	lr=2.571785
Epoch[28] Batch [399]	Speed: 8718.258464 samples/sec	accuracy=0.525027	lr=2.571594
Epoch[28] Batch [449]	Speed: 8677.339699 samples/sec	accuracy=0.524760	lr=2.571403
Epoch[28] Batch [499]	Speed: 8866.708362 samples/sec	accuracy=0.524445	lr=2.571210
Epoch[28] Batch [549]	Speed: 8866.451187 samples/sec	accuracy=0.524142	lr=2.571017
Epoch[28] Batch [599]	Speed: 8810.499816 samples/sec	accuracy=0.523791	lr=2.570824
[Epoch 28] training: accuracy=0.523666
[Epoch 28] speed: 8782 samples/sec	time cost: 150.811872
[Epoch 28] validation: err-top1=0.454141 err-top5=0.205586
Epoch[29] Batch [49]	Speed: 8838.144694 samples/sec	accuracy=0.532910	lr=2.570528
Epoch[29] Batch [99]	Speed: 9043.774531 samples/sec	accuracy=0.531138	lr=2.570333
Epoch[29] Batch [149]	Speed: 9245.846002 samples/sec	accuracy=0.529424	lr=2.570137
Epoch[29] Batch [199]	Speed: 9005.644886 samples/sec	accuracy=0.529424	lr=2.569941
Epoch[29] Batch [249]	Speed: 8786.764981 samples/sec	accuracy=0.528438	lr=2.569744
Epoch[29] Batch [299]	Speed: 8956.964110 samples/sec	accuracy=0.528088	lr=2.569546
Epoch[29] Batch [349]	Speed: 8743.401574 samples/sec	accuracy=0.527397	lr=2.569348
Epoch[29] Batch [399]	Speed: 8843.621480 samples/sec	accuracy=0.526801	lr=2.569149
Epoch[29] Batch [449]	Speed: 8657.582518 samples/sec	accuracy=0.526426	lr=2.568949
Epoch[29] Batch [499]	Speed: 8844.227172 samples/sec	accuracy=0.525859	lr=2.568749
Epoch[29] Batch [549]	Speed: 8687.500962 samples/sec	accuracy=0.525342	lr=2.568548
Epoch[29] Batch [599]	Speed: 8684.178949 samples/sec	accuracy=0.524921	lr=2.568346
[Epoch 29] training: accuracy=0.524677
[Epoch 29] speed: 8846 samples/sec	time cost: 149.810836
[Epoch 29] validation: err-top1=0.439596 err-top5=0.192647
Epoch[30] Batch [49]	Speed: 8379.533391 samples/sec	accuracy=0.534463	lr=2.568039
Epoch[30] Batch [99]	Speed: 8813.298838 samples/sec	accuracy=0.531592	lr=2.567836
Epoch[30] Batch [149]	Speed: 8856.502748 samples/sec	accuracy=0.530013	lr=2.567632
Epoch[30] Batch [199]	Speed: 8748.673855 samples/sec	accuracy=0.528923	lr=2.567427
Epoch[30] Batch [249]	Speed: 8832.899752 samples/sec	accuracy=0.528404	lr=2.567222
Epoch[30] Batch [299]	Speed: 8878.366621 samples/sec	accuracy=0.527866	lr=2.567017
Epoch[30] Batch [349]	Speed: 8867.669468 samples/sec	accuracy=0.527394	lr=2.566810
Epoch[30] Batch [399]	Speed: 8708.726322 samples/sec	accuracy=0.526621	lr=2.566603
Epoch[30] Batch [449]	Speed: 8743.347821 samples/sec	accuracy=0.526724	lr=2.566396
Epoch[30] Batch [499]	Speed: 8745.474607 samples/sec	accuracy=0.526325	lr=2.566188
Epoch[30] Batch [549]	Speed: 8855.725192 samples/sec	accuracy=0.525817	lr=2.565979
Epoch[30] Batch [599]	Speed: 8631.431754 samples/sec	accuracy=0.525398	lr=2.565769
[Epoch 30] training: accuracy=0.525223
[Epoch 30] speed: 8750 samples/sec	time cost: 151.211364
[Epoch 30] validation: err-top1=0.436259 err-top5=0.188029
Epoch[31] Batch [49]	Speed: 8361.460038 samples/sec	accuracy=0.532578	lr=2.565454
Epoch[31] Batch [99]	Speed: 8810.890401 samples/sec	accuracy=0.529233	lr=2.565243
Epoch[31] Batch [149]	Speed: 8607.881854 samples/sec	accuracy=0.528880	lr=2.565031
Epoch[31] Batch [199]	Speed: 8962.101262 samples/sec	accuracy=0.528608	lr=2.564819
Epoch[31] Batch [249]	Speed: 8839.900275 samples/sec	accuracy=0.528123	lr=2.564606
Epoch[31] Batch [299]	Speed: 8680.816666 samples/sec	accuracy=0.527339	lr=2.564392
Epoch[31] Batch [349]	Speed: 8613.286072 samples/sec	accuracy=0.527028	lr=2.564178
Epoch[31] Batch [399]	Speed: 8611.224639 samples/sec	accuracy=0.526882	lr=2.563963
Epoch[31] Batch [449]	Speed: 8636.593790 samples/sec	accuracy=0.526866	lr=2.563747
Epoch[31] Batch [499]	Speed: 8403.690326 samples/sec	accuracy=0.526611	lr=2.563531
Epoch[31] Batch [549]	Speed: 8682.858720 samples/sec	accuracy=0.526160	lr=2.563314
Epoch[31] Batch [599]	Speed: 8562.023252 samples/sec	accuracy=0.525939	lr=2.563097
[Epoch 31] training: accuracy=0.525873
[Epoch 31] speed: 8650 samples/sec	time cost: 153.123475
[Epoch 31] validation: err-top1=0.453438 err-top5=0.205156
Epoch[32] Batch [49]	Speed: 8564.279095 samples/sec	accuracy=0.533320	lr=2.562765
Epoch[32] Batch [99]	Speed: 8763.044148 samples/sec	accuracy=0.531304	lr=2.562546
Epoch[32] Batch [149]	Speed: 8663.504029 samples/sec	accuracy=0.530921	lr=2.562327
Epoch[32] Batch [199]	Speed: 8944.344586 samples/sec	accuracy=0.529636	lr=2.562106
Epoch[32] Batch [249]	Speed: 8716.058928 samples/sec	accuracy=0.529057	lr=2.561885
Epoch[32] Batch [299]	Speed: 8565.827100 samples/sec	accuracy=0.528494	lr=2.561664
Epoch[32] Batch [349]	Speed: 8647.315758 samples/sec	accuracy=0.528322	lr=2.561442
Epoch[32] Batch [399]	Speed: 8509.718244 samples/sec	accuracy=0.528087	lr=2.561219
Epoch[32] Batch [449]	Speed: 8798.874604 samples/sec	accuracy=0.527598	lr=2.560995
Epoch[32] Batch [499]	Speed: 8656.544102 samples/sec	accuracy=0.527317	lr=2.560771
Epoch[32] Batch [549]	Speed: 8541.336438 samples/sec	accuracy=0.526876	lr=2.560547
Epoch[32] Batch [599]	Speed: 8586.061763 samples/sec	accuracy=0.526685	lr=2.560321
[Epoch 32] training: accuracy=0.526662
[Epoch 32] speed: 8664 samples/sec	time cost: 152.431847
[Epoch 32] validation: err-top1=0.436178 err-top5=0.193319
Epoch[33] Batch [49]	Speed: 8362.038439 samples/sec	accuracy=0.535039	lr=2.559982
Epoch[33] Batch [99]	Speed: 8741.211039 samples/sec	accuracy=0.532446	lr=2.559755
Epoch[33] Batch [149]	Speed: 8586.562821 samples/sec	accuracy=0.531006	lr=2.559528
Epoch[33] Batch [199]	Speed: 8576.743218 samples/sec	accuracy=0.530034	lr=2.559300
Epoch[33] Batch [249]	Speed: 8560.004877 samples/sec	accuracy=0.529277	lr=2.559071
Epoch[33] Batch [299]	Speed: 8498.539597 samples/sec	accuracy=0.528553	lr=2.558841
Epoch[33] Batch [349]	Speed: 8656.794304 samples/sec	accuracy=0.528410	lr=2.558611
Epoch[33] Batch [399]	Speed: 8587.931886 samples/sec	accuracy=0.527721	lr=2.558381
Epoch[33] Batch [449]	Speed: 8685.844556 samples/sec	accuracy=0.527411	lr=2.558149
Epoch[33] Batch [499]	Speed: 8789.632403 samples/sec	accuracy=0.526947	lr=2.557917
Epoch[33] Batch [549]	Speed: 8554.631510 samples/sec	accuracy=0.526669	lr=2.557685
Epoch[33] Batch [599]	Speed: 8674.173831 samples/sec	accuracy=0.526535	lr=2.557451
[Epoch 33] training: accuracy=0.526493
[Epoch 33] speed: 8611 samples/sec	time cost: 153.819971
[Epoch 33] validation: err-top1=0.445371 err-top5=0.196191
Epoch[34] Batch [49]	Speed: 8629.992081 samples/sec	accuracy=0.538203	lr=2.557096
Epoch[34] Batch [99]	Speed: 8900.583284 samples/sec	accuracy=0.534766	lr=2.556861
Epoch[34] Batch [149]	Speed: 8877.207599 samples/sec	accuracy=0.533008	lr=2.556625
Epoch[34] Batch [199]	Speed: 8403.463419 samples/sec	accuracy=0.531401	lr=2.556389
Epoch[34] Batch [249]	Speed: 8495.644660 samples/sec	accuracy=0.530480	lr=2.556153
Epoch[34] Batch [299]	Speed: 9078.687744 samples/sec	accuracy=0.529943	lr=2.555915
Epoch[34] Batch [349]	Speed: 8649.306894 samples/sec	accuracy=0.529415	lr=2.555677
Epoch[34] Batch [399]	Speed: 8646.847798 samples/sec	accuracy=0.529346	lr=2.555439
Epoch[34] Batch [449]	Speed: 8750.429189 samples/sec	accuracy=0.528844	lr=2.555200
Epoch[34] Batch [499]	Speed: 8591.781825 samples/sec	accuracy=0.528566	lr=2.554960
Epoch[34] Batch [549]	Speed: 8702.991499 samples/sec	accuracy=0.528336	lr=2.554719
Epoch[34] Batch [599]	Speed: 8566.323748 samples/sec	accuracy=0.528272	lr=2.554478
[Epoch 34] training: accuracy=0.528046
[Epoch 34] speed: 8686 samples/sec	time cost: 152.020038
[Epoch 34] validation: err-top1=0.490377 err-top5=0.231547
Epoch[35] Batch [49]	Speed: 8720.064817 samples/sec	accuracy=0.539453	lr=2.554115
Epoch[35] Batch [99]	Speed: 8649.877377 samples/sec	accuracy=0.534727	lr=2.553873
Epoch[35] Batch [149]	Speed: 8481.194484 samples/sec	accuracy=0.533184	lr=2.553629
Epoch[35] Batch [199]	Speed: 8770.328139 samples/sec	accuracy=0.532478	lr=2.553385
Epoch[35] Batch [249]	Speed: 8691.981585 samples/sec	accuracy=0.531992	lr=2.553141
Epoch[35] Batch [299]	Speed: 8511.236128 samples/sec	accuracy=0.531281	lr=2.552896
Epoch[35] Batch [349]	Speed: 8425.934892 samples/sec	accuracy=0.530646	lr=2.552650
Epoch[35] Batch [399]	Speed: 8546.181051 samples/sec	accuracy=0.530156	lr=2.552403
Epoch[35] Batch [449]	Speed: 8826.851887 samples/sec	accuracy=0.529595	lr=2.552156
Epoch[35] Batch [499]	Speed: 8808.806839 samples/sec	accuracy=0.528804	lr=2.551909
Epoch[35] Batch [549]	Speed: 8692.191796 samples/sec	accuracy=0.528469	lr=2.551660
Epoch[35] Batch [599]	Speed: 8738.955277 samples/sec	accuracy=0.528095	lr=2.551411
[Epoch 35] training: accuracy=0.528119
[Epoch 35] speed: 8654 samples/sec	time cost: 152.844763
[Epoch 35] validation: err-top1=0.447489 err-top5=0.201192
Epoch[36] Batch [49]	Speed: 8439.599666 samples/sec	accuracy=0.537324	lr=2.551032
Epoch[36] Batch [99]	Speed: 8400.022186 samples/sec	accuracy=0.536006	lr=2.550781
Epoch[36] Batch [149]	Speed: 8718.432075 samples/sec	accuracy=0.533607	lr=2.550530
Epoch[36] Batch [199]	Speed: 8669.624217 samples/sec	accuracy=0.533284	lr=2.550278
Epoch[36] Batch [249]	Speed: 8874.490149 samples/sec	accuracy=0.532359	lr=2.550026
Epoch[36] Batch [299]	Speed: 8596.287592 samples/sec	accuracy=0.531309	lr=2.549773
Epoch[36] Batch [349]	Speed: 8507.452290 samples/sec	accuracy=0.531137	lr=2.549519
Epoch[36] Batch [399]	Speed: 8485.374582 samples/sec	accuracy=0.530889	lr=2.549265
Epoch[36] Batch [449]	Speed: 8756.460403 samples/sec	accuracy=0.530124	lr=2.549010
Epoch[36] Batch [499]	Speed: 9071.967795 samples/sec	accuracy=0.529988	lr=2.548754
Epoch[36] Batch [549]	Speed: 8576.485976 samples/sec	accuracy=0.529627	lr=2.548498
Epoch[36] Batch [599]	Speed: 8655.689441 samples/sec	accuracy=0.529534	lr=2.548241
[Epoch 36] training: accuracy=0.529413
[Epoch 36] speed: 8649 samples/sec	time cost: 153.201579
[Epoch 36] validation: err-top1=0.429199 err-top5=0.185332
Epoch[37] Batch [49]	Speed: 8800.747702 samples/sec	accuracy=0.539268	lr=2.547850
Epoch[37] Batch [99]	Speed: 8377.763868 samples/sec	accuracy=0.537051	lr=2.547591
Epoch[37] Batch [149]	Speed: 8863.659830 samples/sec	accuracy=0.534691	lr=2.547332
Epoch[37] Batch [199]	Speed: 8920.266326 samples/sec	accuracy=0.534146	lr=2.547073
Epoch[37] Batch [249]	Speed: 8620.372437 samples/sec	accuracy=0.533160	lr=2.546812
Epoch[37] Batch [299]	Speed: 8730.611813 samples/sec	accuracy=0.532736	lr=2.546552
Epoch[37] Batch [349]	Speed: 8606.379145 samples/sec	accuracy=0.532215	lr=2.546290
Epoch[37] Batch [399]	Speed: 8250.537194 samples/sec	accuracy=0.531110	lr=2.546028
Epoch[37] Batch [449]	Speed: 8616.760117 samples/sec	accuracy=0.530633	lr=2.545765
Epoch[37] Batch [499]	Speed: 8666.279838 samples/sec	accuracy=0.530110	lr=2.545502
Epoch[37] Batch [549]	Speed: 8730.962334 samples/sec	accuracy=0.529770	lr=2.545238
Epoch[37] Batch [599]	Speed: 8850.964219 samples/sec	accuracy=0.529504	lr=2.544973
[Epoch 37] training: accuracy=0.529403
[Epoch 37] speed: 8670 samples/sec	time cost: 152.364520
[Epoch 37] validation: err-top1=0.471985 err-top5=0.221639
Epoch[38] Batch [49]	Speed: 8934.925467 samples/sec	accuracy=0.538506	lr=2.544575
Epoch[38] Batch [99]	Speed: 9155.378846 samples/sec	accuracy=0.537251	lr=2.544309
Epoch[38] Batch [149]	Speed: 8960.312328 samples/sec	accuracy=0.534697	lr=2.544042
Epoch[38] Batch [199]	Speed: 8731.681390 samples/sec	accuracy=0.533928	lr=2.543775
Epoch[38] Batch [249]	Speed: 8705.838908 samples/sec	accuracy=0.533240	lr=2.543507
Epoch[38] Batch [299]	Speed: 8683.251060 samples/sec	accuracy=0.532363	lr=2.543238
Epoch[38] Batch [349]	Speed: 8707.713205 samples/sec	accuracy=0.531710	lr=2.542969
Epoch[38] Batch [399]	Speed: 8874.115541 samples/sec	accuracy=0.531373	lr=2.542699
Epoch[38] Batch [449]	Speed: 8646.651263 samples/sec	accuracy=0.530934	lr=2.542428
Epoch[38] Batch [499]	Speed: 8749.362322 samples/sec	accuracy=0.530579	lr=2.542157
Epoch[38] Batch [549]	Speed: 8894.279584 samples/sec	accuracy=0.530407	lr=2.541885
Epoch[38] Batch [599]	Speed: 8653.748720 samples/sec	accuracy=0.530138	lr=2.541613
[Epoch 38] training: accuracy=0.530071
[Epoch 38] speed: 8800 samples/sec	time cost: 150.426151
[Epoch 38] validation: err-top1=0.462930 err-top5=0.210000
Epoch[39] Batch [49]	Speed: 8660.625912 samples/sec	accuracy=0.541084	lr=2.541198
Epoch[39] Batch [99]	Speed: 8683.979485 samples/sec	accuracy=0.538228	lr=2.540924
Epoch[39] Batch [149]	Speed: 8810.228542 samples/sec	accuracy=0.536120	lr=2.540649
Epoch[39] Batch [199]	Speed: 8831.603836 samples/sec	accuracy=0.535371	lr=2.540374
Epoch[39] Batch [249]	Speed: 8985.602666 samples/sec	accuracy=0.533908	lr=2.540098
Epoch[39] Batch [299]	Speed: 8533.486094 samples/sec	accuracy=0.533491	lr=2.539821
Epoch[39] Batch [349]	Speed: 8902.310433 samples/sec	accuracy=0.533241	lr=2.539544
Epoch[39] Batch [399]	Speed: 8671.699702 samples/sec	accuracy=0.532531	lr=2.539267
Epoch[39] Batch [449]	Speed: 8713.883662 samples/sec	accuracy=0.532470	lr=2.538988
Epoch[39] Batch [499]	Speed: 8930.835849 samples/sec	accuracy=0.531844	lr=2.538709
Epoch[39] Batch [549]	Speed: 8958.585585 samples/sec	accuracy=0.531596	lr=2.538430
Epoch[39] Batch [599]	Speed: 8625.520037 samples/sec	accuracy=0.531376	lr=2.538150
[Epoch 39] training: accuracy=0.531245
[Epoch 39] speed: 8778 samples/sec	time cost: 150.533763
[Epoch 39] validation: err-top1=0.434224 err-top5=0.189901
Epoch[40] Batch [49]	Speed: 8843.449584 samples/sec	accuracy=0.538828	lr=2.537728
Epoch[40] Batch [99]	Speed: 9162.033592 samples/sec	accuracy=0.538828	lr=2.537446
Epoch[40] Batch [149]	Speed: 8695.664657 samples/sec	accuracy=0.537064	lr=2.537164
Epoch[40] Batch [199]	Speed: 8721.059206 samples/sec	accuracy=0.536025	lr=2.536881
Epoch[40] Batch [249]	Speed: 8664.946689 samples/sec	accuracy=0.535584	lr=2.536597
Epoch[40] Batch [299]	Speed: 8619.169609 samples/sec	accuracy=0.534118	lr=2.536313
Epoch[40] Batch [349]	Speed: 8782.605638 samples/sec	accuracy=0.533975	lr=2.536028
Epoch[40] Batch [399]	Speed: 8693.852383 samples/sec	accuracy=0.533199	lr=2.535743
Epoch[40] Batch [449]	Speed: 8711.538611 samples/sec	accuracy=0.532637	lr=2.535457
Epoch[40] Batch [499]	Speed: 8674.918255 samples/sec	accuracy=0.532074	lr=2.535170
Epoch[40] Batch [549]	Speed: 8801.812888 samples/sec	accuracy=0.531893	lr=2.534883
Epoch[40] Batch [599]	Speed: 8566.244130 samples/sec	accuracy=0.531213	lr=2.534595
[Epoch 40] training: accuracy=0.531014
[Epoch 40] speed: 8740 samples/sec	time cost: 151.310881
[Epoch 40] validation: err-top1=0.457438 err-top5=0.209615
Epoch[41] Batch [49]	Speed: 8628.769751 samples/sec	accuracy=0.539863	lr=2.534156
Epoch[41] Batch [99]	Speed: 8724.929848 samples/sec	accuracy=0.540117	lr=2.533866
Epoch[41] Batch [149]	Speed: 8600.344277 samples/sec	accuracy=0.538415	lr=2.533576
Epoch[41] Batch [199]	Speed: 8721.839860 samples/sec	accuracy=0.537454	lr=2.533286
Epoch[41] Batch [249]	Speed: 8627.718304 samples/sec	accuracy=0.535799	lr=2.532994
Epoch[41] Batch [299]	Speed: 8699.730938 samples/sec	accuracy=0.535050	lr=2.532702
Epoch[41] Batch [349]	Speed: 8620.326761 samples/sec	accuracy=0.534453	lr=2.532410
Epoch[41] Batch [399]	Speed: 8778.074178 samples/sec	accuracy=0.533757	lr=2.532116
Epoch[41] Batch [449]	Speed: 8653.622310 samples/sec	accuracy=0.533101	lr=2.531823
Epoch[41] Batch [499]	Speed: 8543.356894 samples/sec	accuracy=0.532897	lr=2.531528
Epoch[41] Batch [549]	Speed: 8474.928457 samples/sec	accuracy=0.532813	lr=2.531233
Epoch[41] Batch [599]	Speed: 8632.871733 samples/sec	accuracy=0.532312	lr=2.530937
[Epoch 41] training: accuracy=0.532156
[Epoch 41] speed: 8649 samples/sec	time cost: 153.003189
[Epoch 41] validation: err-top1=0.423574 err-top5=0.181797
Epoch[42] Batch [49]	Speed: 8489.697582 samples/sec	accuracy=0.538760	lr=2.530493
Epoch[42] Batch [99]	Speed: 8844.645706 samples/sec	accuracy=0.538599	lr=2.530196
Epoch[42] Batch [149]	Speed: 8887.719224 samples/sec	accuracy=0.538604	lr=2.529898
Epoch[42] Batch [199]	Speed: 8593.424382 samples/sec	accuracy=0.537344	lr=2.529599
Epoch[42] Batch [249]	Speed: 8653.935290 samples/sec	accuracy=0.536469	lr=2.529300
Epoch[42] Batch [299]	Speed: 8689.300567 samples/sec	accuracy=0.535529	lr=2.529000
Epoch[42] Batch [349]	Speed: 8693.805749 samples/sec	accuracy=0.535152	lr=2.528700
Epoch[42] Batch [399]	Speed: 8794.361639 samples/sec	accuracy=0.534138	lr=2.528399
Epoch[42] Batch [449]	Speed: 8775.083949 samples/sec	accuracy=0.533713	lr=2.528098
Epoch[42] Batch [499]	Speed: 8527.914481 samples/sec	accuracy=0.533253	lr=2.527796
Epoch[42] Batch [549]	Speed: 8840.295655 samples/sec	accuracy=0.532787	lr=2.527493
Epoch[42] Batch [599]	Speed: 8737.807479 samples/sec	accuracy=0.532666	lr=2.527189
[Epoch 42] training: accuracy=0.532533
[Epoch 42] speed: 8705 samples/sec	time cost: 151.924844
[Epoch 42] validation: err-top1=0.469238 err-top5=0.216248
Epoch[43] Batch [49]	Speed: 8841.949062 samples/sec	accuracy=0.541484	lr=2.526727
Epoch[43] Batch [99]	Speed: 9085.992724 samples/sec	accuracy=0.539429	lr=2.526422
Epoch[43] Batch [149]	Speed: 8560.522861 samples/sec	accuracy=0.537699	lr=2.526117
Epoch[43] Batch [199]	Speed: 8716.859564 samples/sec	accuracy=0.536787	lr=2.525810
Epoch[43] Batch [249]	Speed: 8864.605090 samples/sec	accuracy=0.536010	lr=2.525504
Epoch[43] Batch [299]	Speed: 8901.019528 samples/sec	accuracy=0.535200	lr=2.525196
Epoch[43] Batch [349]	Speed: 8784.564705 samples/sec	accuracy=0.534802	lr=2.524888
Epoch[43] Batch [399]	Speed: 8913.579164 samples/sec	accuracy=0.534620	lr=2.524580
Epoch[43] Batch [449]	Speed: 8735.486313 samples/sec	accuracy=0.534151	lr=2.524270
Epoch[43] Batch [499]	Speed: 8595.342266 samples/sec	accuracy=0.533641	lr=2.523961
Epoch[43] Batch [549]	Speed: 8969.258714 samples/sec	accuracy=0.533224	lr=2.523650
Epoch[43] Batch [599]	Speed: 8667.297153 samples/sec	accuracy=0.532926	lr=2.523339
[Epoch 43] training: accuracy=0.532820
[Epoch 43] speed: 8801 samples/sec	time cost: 150.564962
[Epoch 43] validation: err-top1=0.460078 err-top5=0.210352
Epoch[44] Batch [49]	Speed: 8553.334020 samples/sec	accuracy=0.541152	lr=2.522865
Epoch[44] Batch [99]	Speed: 8986.696525 samples/sec	accuracy=0.539033	lr=2.522552
Epoch[44] Batch [149]	Speed: 9083.912292 samples/sec	accuracy=0.539040	lr=2.522239
Epoch[44] Batch [199]	Speed: 8759.569614 samples/sec	accuracy=0.537083	lr=2.521925
Epoch[44] Batch [249]	Speed: 8940.532012 samples/sec	accuracy=0.536326	lr=2.521611
Epoch[44] Batch [299]	Speed: 8607.582374 samples/sec	accuracy=0.535964	lr=2.521296
Epoch[44] Batch [349]	Speed: 8810.317097 samples/sec	accuracy=0.535364	lr=2.520980
Epoch[44] Batch [399]	Speed: 8571.130560 samples/sec	accuracy=0.534813	lr=2.520664
Epoch[44] Batch [449]	Speed: 8896.125533 samples/sec	accuracy=0.534344	lr=2.520347
Epoch[44] Batch [499]	Speed: 8702.038957 samples/sec	accuracy=0.533969	lr=2.520029
Epoch[44] Batch [549]	Speed: 8779.044876 samples/sec	accuracy=0.533752	lr=2.519711
Epoch[44] Batch [599]	Speed: 8637.214532 samples/sec	accuracy=0.533372	lr=2.519393
[Epoch 44] training: accuracy=0.533101
[Epoch 44] speed: 8778 samples/sec	time cost: 150.458842
[Epoch 44] validation: err-top1=0.443746 err-top5=0.198547
Epoch[45] Batch [49]	Speed: 8965.073989 samples/sec	accuracy=0.545625	lr=2.518913
Epoch[45] Batch [99]	Speed: 8603.600379 samples/sec	accuracy=0.541216	lr=2.518593
Epoch[45] Batch [149]	Speed: 8643.834780 samples/sec	accuracy=0.539785	lr=2.518272
Epoch[45] Batch [199]	Speed: 8974.262780 samples/sec	accuracy=0.538425	lr=2.517951
Epoch[45] Batch [249]	Speed: 8820.092253 samples/sec	accuracy=0.537504	lr=2.517629
Epoch[45] Batch [299]	Speed: 8570.859288 samples/sec	accuracy=0.536789	lr=2.517306
Epoch[45] Batch [349]	Speed: 8702.368849 samples/sec	accuracy=0.536182	lr=2.516983
Epoch[45] Batch [399]	Speed: 8982.831234 samples/sec	accuracy=0.535844	lr=2.516659
Epoch[45] Batch [449]	Speed: 8977.358607 samples/sec	accuracy=0.535505	lr=2.516334
Epoch[45] Batch [499]	Speed: 8755.300684 samples/sec	accuracy=0.535183	lr=2.516009
Epoch[45] Batch [549]	Speed: 8595.024738 samples/sec	accuracy=0.534574	lr=2.515683
Epoch[45] Batch [599]	Speed: 8632.997017 samples/sec	accuracy=0.534139	lr=2.515357
[Epoch 45] training: accuracy=0.533958
[Epoch 45] speed: 8767 samples/sec	time cost: 151.218700
[Epoch 45] validation: err-top1=0.456016 err-top5=0.210410
Epoch[46] Batch [49]	Speed: 8752.319884 samples/sec	accuracy=0.542344	lr=2.514860
Epoch[46] Batch [99]	Speed: 8562.661829 samples/sec	accuracy=0.541167	lr=2.514532
Epoch[46] Batch [149]	Speed: 8833.299047 samples/sec	accuracy=0.539941	lr=2.514203
Epoch[46] Batch [199]	Speed: 8953.027487 samples/sec	accuracy=0.539360	lr=2.513874
Epoch[46] Batch [249]	Speed: 8936.219719 samples/sec	accuracy=0.538545	lr=2.513544
Epoch[46] Batch [299]	Speed: 8844.204954 samples/sec	accuracy=0.538019	lr=2.513214
Epoch[46] Batch [349]	Speed: 8934.688482 samples/sec	accuracy=0.537414	lr=2.512883
Epoch[46] Batch [399]	Speed: 8890.638571 samples/sec	accuracy=0.536191	lr=2.512552
Epoch[46] Batch [449]	Speed: 8612.366880 samples/sec	accuracy=0.535763	lr=2.512219
Epoch[46] Batch [499]	Speed: 8846.223487 samples/sec	accuracy=0.535128	lr=2.511887
Epoch[46] Batch [549]	Speed: 8970.279277 samples/sec	accuracy=0.534750	lr=2.511553
Epoch[46] Batch [599]	Speed: 8358.423480 samples/sec	accuracy=0.534143	lr=2.511219
[Epoch 46] training: accuracy=0.533951
[Epoch 46] speed: 8780 samples/sec	time cost: 150.661207
[Epoch 46] validation: err-top1=0.453654 err-top5=0.200989
Epoch[47] Batch [49]	Speed: 8947.617014 samples/sec	accuracy=0.543535	lr=2.510717
Epoch[47] Batch [99]	Speed: 8745.811363 samples/sec	accuracy=0.540762	lr=2.510382
Epoch[47] Batch [149]	Speed: 8769.887957 samples/sec	accuracy=0.540257	lr=2.510046
Epoch[47] Batch [199]	Speed: 9000.162729 samples/sec	accuracy=0.539453	lr=2.509709
Epoch[47] Batch [249]	Speed: 8651.672407 samples/sec	accuracy=0.539092	lr=2.509372
Epoch[47] Batch [299]	Speed: 8656.839844 samples/sec	accuracy=0.538162	lr=2.509034
Epoch[47] Batch [349]	Speed: 8772.197343 samples/sec	accuracy=0.537307	lr=2.508695
Epoch[47] Batch [399]	Speed: 8867.539843 samples/sec	accuracy=0.536688	lr=2.508356
Epoch[47] Batch [449]	Speed: 8390.927305 samples/sec	accuracy=0.535725	lr=2.508016
Epoch[47] Batch [499]	Speed: 8911.979672 samples/sec	accuracy=0.535360	lr=2.507676
Epoch[47] Batch [549]	Speed: 8522.441627 samples/sec	accuracy=0.535083	lr=2.507335
Epoch[47] Batch [599]	Speed: 8888.014052 samples/sec	accuracy=0.534667	lr=2.506993
[Epoch 47] training: accuracy=0.534424
[Epoch 47] speed: 8754 samples/sec	time cost: 151.284538
[Epoch 47] validation: err-top1=0.448669 err-top5=0.200399
Epoch[48] Batch [49]	Speed: 8796.023667 samples/sec	accuracy=0.544834	lr=2.506473
Epoch[48] Batch [99]	Speed: 8697.857609 samples/sec	accuracy=0.541025	lr=2.506130
Epoch[48] Batch [149]	Speed: 8816.580668 samples/sec	accuracy=0.539561	lr=2.505786
Epoch[48] Batch [199]	Speed: 8471.525037 samples/sec	accuracy=0.538357	lr=2.505442
Epoch[48] Batch [249]	Speed: 8798.369550 samples/sec	accuracy=0.537887	lr=2.505097
Epoch[48] Batch [299]	Speed: 8683.567767 samples/sec	accuracy=0.537030	lr=2.504752
Epoch[48] Batch [349]	Speed: 8710.534556 samples/sec	accuracy=0.536409	lr=2.504405
Epoch[48] Batch [399]	Speed: 8830.958108 samples/sec	accuracy=0.536155	lr=2.504059
Epoch[48] Batch [449]	Speed: 8777.333650 samples/sec	accuracy=0.535770	lr=2.503711
Epoch[48] Batch [499]	Speed: 9085.841838 samples/sec	accuracy=0.535310	lr=2.503364
Epoch[48] Batch [549]	Speed: 9209.633517 samples/sec	accuracy=0.535064	lr=2.503015
Epoch[48] Batch [599]	Speed: 8807.959782 samples/sec	accuracy=0.534767	lr=2.502666
[Epoch 48] training: accuracy=0.534716
[Epoch 48] speed: 8797 samples/sec	time cost: 150.508950
[Epoch 48] validation: err-top1=0.437930 err-top5=0.193828
Epoch[49] Batch [49]	Speed: 8835.053976 samples/sec	accuracy=0.543516	lr=2.502141
Epoch[49] Batch [99]	Speed: 9133.029264 samples/sec	accuracy=0.540391	lr=2.501790
Epoch[49] Batch [149]	Speed: 9044.169123 samples/sec	accuracy=0.539499	lr=2.501439
Epoch[49] Batch [199]	Speed: 8411.268977 samples/sec	accuracy=0.538342	lr=2.501087
Epoch[49] Batch [249]	Speed: 8750.369466 samples/sec	accuracy=0.537521	lr=2.500735
Epoch[49] Batch [299]	Speed: 8813.966404 samples/sec	accuracy=0.537415	lr=2.500382
Epoch[49] Batch [349]	Speed: 8884.227837 samples/sec	accuracy=0.537153	lr=2.500028
Epoch[49] Batch [399]	Speed: 8993.000139 samples/sec	accuracy=0.537240	lr=2.499674
Epoch[49] Batch [449]	Speed: 8901.501567 samples/sec	accuracy=0.536683	lr=2.499319
Epoch[49] Batch [499]	Speed: 8627.178986 samples/sec	accuracy=0.536220	lr=2.498964
Epoch[49] Batch [549]	Speed: 8785.002408 samples/sec	accuracy=0.535935	lr=2.498608
Epoch[49] Batch [599]	Speed: 8990.969222 samples/sec	accuracy=0.535694	lr=2.498251
[Epoch 49] training: accuracy=0.535375
[Epoch 49] speed: 8835 samples/sec	time cost: 149.966454
[Epoch 49] validation: err-top1=0.440165 err-top5=0.193990
Epoch[50] Batch [49]	Speed: 9090.926843 samples/sec	accuracy=0.542754	lr=2.497708
Epoch[50] Batch [99]	Speed: 8869.410058 samples/sec	accuracy=0.540283	lr=2.497350
Epoch[50] Batch [149]	Speed: 8658.679311 samples/sec	accuracy=0.540303	lr=2.496991
Epoch[50] Batch [199]	Speed: 8879.251139 samples/sec	accuracy=0.539246	lr=2.496632
Epoch[50] Batch [249]	Speed: 8741.078148 samples/sec	accuracy=0.539072	lr=2.496272
Epoch[50] Batch [299]	Speed: 8779.129934 samples/sec	accuracy=0.538141	lr=2.495911
Epoch[50] Batch [349]	Speed: 8797.007348 samples/sec	accuracy=0.537563	lr=2.495550
Epoch[50] Batch [399]	Speed: 8820.891464 samples/sec	accuracy=0.537391	lr=2.495188
Epoch[50] Batch [449]	Speed: 8844.248481 samples/sec	accuracy=0.537222	lr=2.494826
Epoch[50] Batch [499]	Speed: 8739.827884 samples/sec	accuracy=0.536990	lr=2.494463
Epoch[50] Batch [549]	Speed: 8819.354579 samples/sec	accuracy=0.536606	lr=2.494099
Epoch[50] Batch [599]	Speed: 8562.162021 samples/sec	accuracy=0.536348	lr=2.493735
[Epoch 50] training: accuracy=0.536341
[Epoch 50] speed: 8799 samples/sec	time cost: 150.792487
[Epoch 50] validation: err-top1=0.455078 err-top5=0.202539
Epoch[51] Batch [49]	Speed: 8817.335979 samples/sec	accuracy=0.542070	lr=2.493181
Epoch[51] Batch [99]	Speed: 8565.440174 samples/sec	accuracy=0.541050	lr=2.492815
Epoch[51] Batch [149]	Speed: 8790.315279 samples/sec	accuracy=0.539652	lr=2.492449
Epoch[51] Batch [199]	Speed: 8517.825239 samples/sec	accuracy=0.539548	lr=2.492082
Epoch[51] Batch [249]	Speed: 8502.935279 samples/sec	accuracy=0.539105	lr=2.491714
Epoch[51] Batch [299]	Speed: 8903.782600 samples/sec	accuracy=0.538586	lr=2.491346
Epoch[51] Batch [349]	Speed: 8598.952658 samples/sec	accuracy=0.538039	lr=2.490978
Epoch[51] Batch [399]	Speed: 8736.324108 samples/sec	accuracy=0.537390	lr=2.490608
Epoch[51] Batch [449]	Speed: 8774.540392 samples/sec	accuracy=0.536877	lr=2.490239
Epoch[51] Batch [499]	Speed: 8587.973614 samples/sec	accuracy=0.536197	lr=2.489868
Epoch[51] Batch [549]	Speed: 8920.504029 samples/sec	accuracy=0.535844	lr=2.489497
Epoch[51] Batch [599]	Speed: 8489.476075 samples/sec	accuracy=0.535653	lr=2.489126
[Epoch 51] training: accuracy=0.535600
[Epoch 51] speed: 8680 samples/sec	time cost: 152.207259
[Epoch 51] validation: err-top1=0.425557 err-top5=0.183411
Epoch[52] Batch [49]	Speed: 8871.588543 samples/sec	accuracy=0.547227	lr=2.488567
Epoch[52] Batch [99]	Speed: 8766.076272 samples/sec	accuracy=0.544067	lr=2.488194
Epoch[52] Batch [149]	Speed: 8774.306461 samples/sec	accuracy=0.542783	lr=2.487820
Epoch[52] Batch [199]	Speed: 8770.552545 samples/sec	accuracy=0.541973	lr=2.487446
Epoch[52] Batch [249]	Speed: 8700.053959 samples/sec	accuracy=0.540877	lr=2.487071
Epoch[52] Batch [299]	Speed: 8657.595781 samples/sec	accuracy=0.540091	lr=2.486695
Epoch[52] Batch [349]	Speed: 8587.008483 samples/sec	accuracy=0.539747	lr=2.486319
Epoch[52] Batch [399]	Speed: 8791.676669 samples/sec	accuracy=0.539204	lr=2.485943
Epoch[52] Batch [449]	Speed: 8778.565600 samples/sec	accuracy=0.538733	lr=2.485565
Epoch[52] Batch [499]	Speed: 8514.769105 samples/sec	accuracy=0.538119	lr=2.485188
Epoch[52] Batch [549]	Speed: 8773.973422 samples/sec	accuracy=0.537434	lr=2.484809
Epoch[52] Batch [599]	Speed: 8680.621040 samples/sec	accuracy=0.537275	lr=2.484430
[Epoch 52] training: accuracy=0.537060
[Epoch 52] speed: 8727 samples/sec	time cost: 151.852243
[Epoch 52] validation: err-top1=0.424072 err-top5=0.183370
Epoch[53] Batch [49]	Speed: 8439.061224 samples/sec	accuracy=0.545937	lr=2.483853
Epoch[53] Batch [99]	Speed: 8768.179583 samples/sec	accuracy=0.543403	lr=2.483472
Epoch[53] Batch [149]	Speed: 8476.800165 samples/sec	accuracy=0.542207	lr=2.483091
Epoch[53] Batch [199]	Speed: 8783.298020 samples/sec	accuracy=0.541514	lr=2.482709
Epoch[53] Batch [249]	Speed: 8841.277979 samples/sec	accuracy=0.540656	lr=2.482327
Epoch[53] Batch [299]	Speed: 8590.468576 samples/sec	accuracy=0.539720	lr=2.481944
Epoch[53] Batch [349]	Speed: 8585.113021 samples/sec	accuracy=0.539215	lr=2.481560
Epoch[53] Batch [399]	Speed: 8849.321843 samples/sec	accuracy=0.539062	lr=2.481176
Epoch[53] Batch [449]	Speed: 8807.669519 samples/sec	accuracy=0.538816	lr=2.480792
Epoch[53] Batch [499]	Speed: 8677.205061 samples/sec	accuracy=0.537936	lr=2.480406
Epoch[53] Batch [549]	Speed: 8565.375604 samples/sec	accuracy=0.537776	lr=2.480020
Epoch[53] Batch [599]	Speed: 8661.739896 samples/sec	accuracy=0.537703	lr=2.479634
[Epoch 53] training: accuracy=0.537266
[Epoch 53] speed: 8670 samples/sec	time cost: 152.796027
[Epoch 53] validation: err-top1=0.445195 err-top5=0.197988
Epoch[54] Batch [49]	Speed: 8590.468576 samples/sec	accuracy=0.543066	lr=2.479053
Epoch[54] Batch [99]	Speed: 8830.431754 samples/sec	accuracy=0.542798	lr=2.478665
Epoch[54] Batch [149]	Speed: 8466.829933 samples/sec	accuracy=0.542500	lr=2.478277
Epoch[54] Batch [199]	Speed: 8463.532579 samples/sec	accuracy=0.541575	lr=2.477887
Epoch[54] Batch [249]	Speed: 8548.566199 samples/sec	accuracy=0.541277	lr=2.477498
Epoch[54] Batch [299]	Speed: 8940.096538 samples/sec	accuracy=0.540425	lr=2.477107
Epoch[54] Batch [349]	Speed: 8711.609997 samples/sec	accuracy=0.539615	lr=2.476717
Epoch[54] Batch [399]	Speed: 8794.771865 samples/sec	accuracy=0.539111	lr=2.476325
Epoch[54] Batch [449]	Speed: 8745.872626 samples/sec	accuracy=0.538668	lr=2.475933
Epoch[54] Batch [499]	Speed: 8679.463782 samples/sec	accuracy=0.538171	lr=2.475540
Epoch[54] Batch [549]	Speed: 8771.086114 samples/sec	accuracy=0.538094	lr=2.475147
Epoch[54] Batch [599]	Speed: 8601.417140 samples/sec	accuracy=0.537751	lr=2.474753
[Epoch 54] training: accuracy=0.537648
[Epoch 54] speed: 8668 samples/sec	time cost: 152.653620
[Epoch 54] validation: err-top1=0.421224 err-top5=0.179871
Epoch[55] Batch [49]	Speed: 8461.253313 samples/sec	accuracy=0.546396	lr=2.474153
Epoch[55] Batch [99]	Speed: 8752.822162 samples/sec	accuracy=0.543887	lr=2.473758
Epoch[55] Batch [149]	Speed: 8797.153297 samples/sec	accuracy=0.542542	lr=2.473362
Epoch[55] Batch [199]	Speed: 8721.647517 samples/sec	accuracy=0.540959	lr=2.472966
Epoch[55] Batch [249]	Speed: 8394.244112 samples/sec	accuracy=0.540965	lr=2.472568
Epoch[55] Batch [299]	Speed: 8651.938362 samples/sec	accuracy=0.540565	lr=2.472171
Epoch[55] Batch [349]	Speed: 8681.858983 samples/sec	accuracy=0.539997	lr=2.471772
Epoch[55] Batch [399]	Speed: 8764.606360 samples/sec	accuracy=0.539083	lr=2.471374
Epoch[55] Batch [449]	Speed: 8738.776936 samples/sec	accuracy=0.538725	lr=2.470974
Epoch[55] Batch [499]	Speed: 8613.739004 samples/sec	accuracy=0.538452	lr=2.470574
Epoch[55] Batch [549]	Speed: 8766.541658 samples/sec	accuracy=0.538311	lr=2.470173
Epoch[55] Batch [599]	Speed: 8854.422752 samples/sec	accuracy=0.538382	lr=2.469772
[Epoch 55] training: accuracy=0.537966
[Epoch 55] speed: 8680 samples/sec	time cost: 152.359088
[Epoch 55] validation: err-top1=0.445820 err-top5=0.199355
Epoch[56] Batch [49]	Speed: 8695.551808 samples/sec	accuracy=0.548545	lr=2.469169
Epoch[56] Batch [99]	Speed: 8412.966990 samples/sec	accuracy=0.545591	lr=2.468767
Epoch[56] Batch [149]	Speed: 8837.627666 samples/sec	accuracy=0.544004	lr=2.468363
Epoch[56] Batch [199]	Speed: 8723.069569 samples/sec	accuracy=0.542302	lr=2.467960
Epoch[56] Batch [249]	Speed: 8657.487582 samples/sec	accuracy=0.541561	lr=2.467555
Epoch[56] Batch [299]	Speed: 8749.982268 samples/sec	accuracy=0.541074	lr=2.467150
Epoch[56] Batch [349]	Speed: 9025.118021 samples/sec	accuracy=0.540773	lr=2.466744
Epoch[56] Batch [399]	Speed: 8554.490770 samples/sec	accuracy=0.540453	lr=2.466338
Epoch[56] Batch [449]	Speed: 8933.657233 samples/sec	accuracy=0.539625	lr=2.465932
Epoch[56] Batch [499]	Speed: 8692.427175 samples/sec	accuracy=0.539506	lr=2.465524
Epoch[56] Batch [549]	Speed: 8857.618922 samples/sec	accuracy=0.539080	lr=2.465116
Epoch[56] Batch [599]	Speed: 8681.344461 samples/sec	accuracy=0.538815	lr=2.464708
[Epoch 56] training: accuracy=0.538722
[Epoch 56] speed: 8728 samples/sec	time cost: 151.631768
[Epoch 56] validation: err-top1=0.433268 err-top5=0.190450
Epoch[57] Batch [49]	Speed: 8521.078653 samples/sec	accuracy=0.544258	lr=2.464086
Epoch[57] Batch [99]	Speed: 8667.649430 samples/sec	accuracy=0.544521	lr=2.463676
Epoch[57] Batch [149]	Speed: 8577.977406 samples/sec	accuracy=0.543060	lr=2.463265
Epoch[57] Batch [199]	Speed: 8874.717166 samples/sec	accuracy=0.542341	lr=2.462854
Epoch[57] Batch [249]	Speed: 8876.315600 samples/sec	accuracy=0.540994	lr=2.462442
Epoch[57] Batch [299]	Speed: 8651.110226 samples/sec	accuracy=0.540573	lr=2.462030
Epoch[57] Batch [349]	Speed: 8873.614280 samples/sec	accuracy=0.540371	lr=2.461617
Epoch[57] Batch [399]	Speed: 8604.329118 samples/sec	accuracy=0.539727	lr=2.461203
Epoch[57] Batch [449]	Speed: 8762.850519 samples/sec	accuracy=0.539393	lr=2.460789
Epoch[57] Batch [499]	Speed: 8589.623304 samples/sec	accuracy=0.538925	lr=2.460375
Epoch[57] Batch [549]	Speed: 8434.194326 samples/sec	accuracy=0.538668	lr=2.459959
Epoch[57] Batch [599]	Speed: 8681.669803 samples/sec	accuracy=0.538497	lr=2.459544
[Epoch 57] training: accuracy=0.538419
[Epoch 57] speed: 8672 samples/sec	time cost: 152.848195
[Epoch 57] validation: err-top1=0.422969 err-top5=0.180293
Epoch[58] Batch [49]	Speed: 8448.843468 samples/sec	accuracy=0.549180	lr=2.458910
Epoch[58] Batch [99]	Speed: 8800.406703 samples/sec	accuracy=0.546060	lr=2.458493
Epoch[58] Batch [149]	Speed: 8681.496776 samples/sec	accuracy=0.545579	lr=2.458075
Epoch[58] Batch [199]	Speed: 8565.802841 samples/sec	accuracy=0.544282	lr=2.457657
Epoch[58] Batch [249]	Speed: 8831.660314 samples/sec	accuracy=0.543119	lr=2.457238
Epoch[58] Batch [299]	Speed: 9115.800317 samples/sec	accuracy=0.542137	lr=2.456818
Epoch[58] Batch [349]	Speed: 8875.820877 samples/sec	accuracy=0.541539	lr=2.456398
Epoch[58] Batch [399]	Speed: 8780.322361 samples/sec	accuracy=0.540945	lr=2.455977
Epoch[58] Batch [449]	Speed: 8774.696712 samples/sec	accuracy=0.540459	lr=2.455556
Epoch[58] Batch [499]	Speed: 8751.542145 samples/sec	accuracy=0.539967	lr=2.455134
Epoch[58] Batch [549]	Speed: 8594.436874 samples/sec	accuracy=0.539780	lr=2.454711
Epoch[58] Batch [599]	Speed: 8765.021327 samples/sec	accuracy=0.539394	lr=2.454288
[Epoch 58] training: accuracy=0.539255
[Epoch 58] speed: 8738 samples/sec	time cost: 151.275617
[Epoch 58] validation: err-top1=0.450948 err-top5=0.200887
Epoch[59] Batch [49]	Speed: 8659.894764 samples/sec	accuracy=0.550771	lr=2.453653
Epoch[59] Batch [99]	Speed: 8491.946531 samples/sec	accuracy=0.547734	lr=2.453228
Epoch[59] Batch [149]	Speed: 8824.213217 samples/sec	accuracy=0.545176	lr=2.452803
Epoch[59] Batch [199]	Speed: 8560.538729 samples/sec	accuracy=0.543547	lr=2.452377
Epoch[59] Batch [249]	Speed: 8887.031246 samples/sec	accuracy=0.543199	lr=2.451951
Epoch[59] Batch [299]	Speed: 8865.894605 samples/sec	accuracy=0.542554	lr=2.451524
Epoch[59] Batch [349]	Speed: 8884.014299 samples/sec	accuracy=0.542317	lr=2.451097
Epoch[59] Batch [399]	Speed: 8727.628644 samples/sec	accuracy=0.542051	lr=2.450669
Epoch[59] Batch [449]	Speed: 8989.957118 samples/sec	accuracy=0.541682	lr=2.450240
Epoch[59] Batch [499]	Speed: 8825.910127 samples/sec	accuracy=0.541018	lr=2.449811
Epoch[59] Batch [549]	Speed: 8651.990997 samples/sec	accuracy=0.540671	lr=2.449381
Epoch[59] Batch [599]	Speed: 8903.923623 samples/sec	accuracy=0.540461	lr=2.448951
[Epoch 59] training: accuracy=0.540275
[Epoch 59] speed: 8761 samples/sec	time cost: 151.145681
[Epoch 59] validation: err-top1=0.449198 err-top5=0.203613
Epoch[60] Batch [49]	Speed: 8964.027295 samples/sec	accuracy=0.550557	lr=2.448296
Epoch[60] Batch [99]	Speed: 8822.563536 samples/sec	accuracy=0.546548	lr=2.447864
Epoch[60] Batch [149]	Speed: 9028.638483 samples/sec	accuracy=0.544404	lr=2.447432
Epoch[60] Batch [199]	Speed: 9050.032831 samples/sec	accuracy=0.543530	lr=2.446999
Epoch[60] Batch [249]	Speed: 8711.154135 samples/sec	accuracy=0.542814	lr=2.446565
Epoch[60] Batch [299]	Speed: 8783.184142 samples/sec	accuracy=0.542194	lr=2.446131
Epoch[60] Batch [349]	Speed: 8897.406356 samples/sec	accuracy=0.541412	lr=2.445697
Epoch[60] Batch [399]	Speed: 8761.691610 samples/sec	accuracy=0.540972	lr=2.445262
Epoch[60] Batch [449]	Speed: 8947.333875 samples/sec	accuracy=0.540536	lr=2.444826
Epoch[60] Batch [499]	Speed: 8894.858157 samples/sec	accuracy=0.540300	lr=2.444389
Epoch[60] Batch [549]	Speed: 8743.870431 samples/sec	accuracy=0.540047	lr=2.443953
Epoch[60] Batch [599]	Speed: 8866.684565 samples/sec	accuracy=0.539664	lr=2.443515
[Epoch 60] training: accuracy=0.539488
[Epoch 60] speed: 8869 samples/sec	time cost: 149.274626
[Epoch 60] validation: err-top1=0.428965 err-top5=0.185234
Epoch[61] Batch [49]	Speed: 8518.815770 samples/sec	accuracy=0.547598	lr=2.442858
Epoch[61] Batch [99]	Speed: 8808.903135 samples/sec	accuracy=0.546621	lr=2.442419
Epoch[61] Batch [149]	Speed: 8759.351665 samples/sec	accuracy=0.544544	lr=2.441979
Epoch[61] Batch [199]	Speed: 8707.519366 samples/sec	accuracy=0.544038	lr=2.441539
Epoch[61] Batch [249]	Speed: 8901.907274 samples/sec	accuracy=0.544172	lr=2.441099
Epoch[61] Batch [299]	Speed: 8928.203881 samples/sec	accuracy=0.543289	lr=2.440657
Epoch[61] Batch [349]	Speed: 8939.060134 samples/sec	accuracy=0.542759	lr=2.440216
Epoch[61] Batch [399]	Speed: 8432.613395 samples/sec	accuracy=0.542430	lr=2.439773
Epoch[61] Batch [449]	Speed: 8655.721538 samples/sec	accuracy=0.542017	lr=2.439330
Epoch[61] Batch [499]	Speed: 8862.520555 samples/sec	accuracy=0.541764	lr=2.438887
Epoch[61] Batch [549]	Speed: 9054.532621 samples/sec	accuracy=0.541384	lr=2.438443
Epoch[61] Batch [599]	Speed: 8632.039263 samples/sec	accuracy=0.541000	lr=2.437998
[Epoch 61] training: accuracy=0.540926
[Epoch 61] speed: 8760 samples/sec	time cost: 151.081631
[Epoch 61] validation: err-top1=0.416911 err-top5=0.177714
Epoch[62] Batch [49]	Speed: 8875.106313 samples/sec	accuracy=0.548213	lr=2.437321
Epoch[62] Batch [99]	Speed: 9267.562847 samples/sec	accuracy=0.547388	lr=2.436875
Epoch[62] Batch [149]	Speed: 8778.509081 samples/sec	accuracy=0.544414	lr=2.436428
Epoch[62] Batch [199]	Speed: 8748.150671 samples/sec	accuracy=0.543877	lr=2.435981
Epoch[62] Batch [249]	Speed: 9067.564757 samples/sec	accuracy=0.543941	lr=2.435533
Epoch[62] Batch [299]	Speed: 8559.836666 samples/sec	accuracy=0.543595	lr=2.435085
Epoch[62] Batch [349]	Speed: 8926.799141 samples/sec	accuracy=0.542948	lr=2.434636
Epoch[62] Batch [399]	Speed: 8703.675088 samples/sec	accuracy=0.542366	lr=2.434187
Epoch[62] Batch [449]	Speed: 8853.073796 samples/sec	accuracy=0.541895	lr=2.433737
Epoch[62] Batch [499]	Speed: 8618.967757 samples/sec	accuracy=0.541559	lr=2.433286
Epoch[62] Batch [549]	Speed: 8772.322762 samples/sec	accuracy=0.541127	lr=2.432835
Epoch[62] Batch [599]	Speed: 8655.338310 samples/sec	accuracy=0.540959	lr=2.432383
[Epoch 62] training: accuracy=0.540875
[Epoch 62] speed: 8813 samples/sec	time cost: 150.289970
[Epoch 62] validation: err-top1=0.431660 err-top5=0.188945
Epoch[63] Batch [49]	Speed: 9112.868533 samples/sec	accuracy=0.546611	lr=2.431704
Epoch[63] Batch [99]	Speed: 8862.107276 samples/sec	accuracy=0.544937	lr=2.431251
Epoch[63] Batch [149]	Speed: 8829.629180 samples/sec	accuracy=0.545036	lr=2.430797
Epoch[63] Batch [199]	Speed: 8665.199650 samples/sec	accuracy=0.544150	lr=2.430343
Epoch[63] Batch [249]	Speed: 8849.044344 samples/sec	accuracy=0.543490	lr=2.429888
Epoch[63] Batch [299]	Speed: 8665.294929 samples/sec	accuracy=0.542520	lr=2.429433
Epoch[63] Batch [349]	Speed: 8620.398736 samples/sec	accuracy=0.541804	lr=2.428977
Epoch[63] Batch [399]	Speed: 8865.135526 samples/sec	accuracy=0.541587	lr=2.428520
Epoch[63] Batch [449]	Speed: 8590.836802 samples/sec	accuracy=0.541232	lr=2.428063
Epoch[63] Batch [499]	Speed: 8663.295727 samples/sec	accuracy=0.540932	lr=2.427605
Epoch[63] Batch [549]	Speed: 8888.500019 samples/sec	accuracy=0.540692	lr=2.427147
Epoch[63] Batch [599]	Speed: 8805.822901 samples/sec	accuracy=0.540662	lr=2.426688
[Epoch 63] training: accuracy=0.540525
[Epoch 63] speed: 8782 samples/sec	time cost: 150.811002
[Epoch 63] validation: err-top1=0.433105 err-top5=0.188558
Epoch[64] Batch [49]	Speed: 8781.289067 samples/sec	accuracy=0.549795	lr=2.425989
Epoch[64] Batch [99]	Speed: 8724.651234 samples/sec	accuracy=0.547764	lr=2.425529
Epoch[64] Batch [149]	Speed: 8789.148194 samples/sec	accuracy=0.545479	lr=2.425068
Epoch[64] Batch [199]	Speed: 8459.767200 samples/sec	accuracy=0.544407	lr=2.424607
Epoch[64] Batch [249]	Speed: 8762.112915 samples/sec	accuracy=0.544580	lr=2.424145
Epoch[64] Batch [299]	Speed: 8719.302181 samples/sec	accuracy=0.544329	lr=2.423682
Epoch[64] Batch [349]	Speed: 8690.629439 samples/sec	accuracy=0.543927	lr=2.423219
Epoch[64] Batch [399]	Speed: 8713.486958 samples/sec	accuracy=0.543181	lr=2.422756
Epoch[64] Batch [449]	Speed: 8752.796476 samples/sec	accuracy=0.542959	lr=2.422292
Epoch[64] Batch [499]	Speed: 8882.247400 samples/sec	accuracy=0.542820	lr=2.421827
Epoch[64] Batch [549]	Speed: 8670.391140 samples/sec	accuracy=0.542512	lr=2.421361
Epoch[64] Batch [599]	Speed: 8800.453586 samples/sec	accuracy=0.542155	lr=2.420896
[Epoch 64] training: accuracy=0.542074
[Epoch 64] speed: 8724 samples/sec	time cost: 151.877130
[Epoch 64] validation: err-top1=0.459045 err-top5=0.209127
Epoch[65] Batch [49]	Speed: 8472.286223 samples/sec	accuracy=0.551855	lr=2.420186
Epoch[65] Batch [99]	Speed: 8980.285887 samples/sec	accuracy=0.549976	lr=2.419719
Epoch[65] Batch [149]	Speed: 8707.623699 samples/sec	accuracy=0.548320	lr=2.419251
Epoch[65] Batch [199]	Speed: 8868.168043 samples/sec	accuracy=0.547051	lr=2.418783
Epoch[65] Batch [249]	Speed: 8451.963389 samples/sec	accuracy=0.546334	lr=2.418314
Epoch[65] Batch [299]	Speed: 8913.320743 samples/sec	accuracy=0.544771	lr=2.417844
Epoch[65] Batch [349]	Speed: 8771.812510 samples/sec	accuracy=0.543998	lr=2.417374
Epoch[65] Batch [399]	Speed: 8558.574094 samples/sec	accuracy=0.543871	lr=2.416903
Epoch[65] Batch [449]	Speed: 8788.967799 samples/sec	accuracy=0.543513	lr=2.416432
Epoch[65] Batch [499]	Speed: 8742.533415 samples/sec	accuracy=0.542658	lr=2.415960
Epoch[65] Batch [549]	Speed: 8640.065461 samples/sec	accuracy=0.542542	lr=2.415488
Epoch[65] Batch [599]	Speed: 8710.817746 samples/sec	accuracy=0.542123	lr=2.415015
[Epoch 65] training: accuracy=0.542086
[Epoch 65] speed: 8716 samples/sec	time cost: 151.911192
[Epoch 65] validation: err-top1=0.442617 err-top5=0.195703
Epoch[66] Batch [49]	Speed: 8528.649084 samples/sec	accuracy=0.553379	lr=2.414305
Epoch[66] Batch [99]	Speed: 8591.166909 samples/sec	accuracy=0.549580	lr=2.413830
Epoch[66] Batch [149]	Speed: 8666.342090 samples/sec	accuracy=0.549482	lr=2.413355
Epoch[66] Batch [199]	Speed: 8756.077663 samples/sec	accuracy=0.548120	lr=2.412880
Epoch[66] Batch [249]	Speed: 8886.146648 samples/sec	accuracy=0.547318	lr=2.412404
Epoch[66] Batch [299]	Speed: 8566.709215 samples/sec	accuracy=0.546528	lr=2.411927
Epoch[66] Batch [349]	Speed: 8471.655540 samples/sec	accuracy=0.545604	lr=2.411450
Epoch[66] Batch [399]	Speed: 8699.339222 samples/sec	accuracy=0.545074	lr=2.410973
Epoch[66] Batch [449]	Speed: 8807.261159 samples/sec	accuracy=0.544501	lr=2.410494
Epoch[66] Batch [499]	Speed: 8578.784574 samples/sec	accuracy=0.543862	lr=2.410016
Epoch[66] Batch [549]	Speed: 8670.808261 samples/sec	accuracy=0.543547	lr=2.409536
Epoch[66] Batch [599]	Speed: 8487.513559 samples/sec	accuracy=0.543185	lr=2.409056
[Epoch 66] training: accuracy=0.543009
[Epoch 66] speed: 8638 samples/sec	time cost: 153.203294
[Epoch 66] validation: err-top1=0.417460 err-top5=0.176208
Epoch[67] Batch [49]	Speed: 8941.251938 samples/sec	accuracy=0.550254	lr=2.408326
Epoch[67] Batch [99]	Speed: 9146.670082 samples/sec	accuracy=0.549341	lr=2.407845
Epoch[67] Batch [149]	Speed: 8784.040094 samples/sec	accuracy=0.547406	lr=2.407363
Epoch[67] Batch [199]	Speed: 8805.962283 samples/sec	accuracy=0.546956	lr=2.406880
Epoch[67] Batch [249]	Speed: 8863.213340 samples/sec	accuracy=0.545924	lr=2.406397
Epoch[67] Batch [299]	Speed: 8603.262767 samples/sec	accuracy=0.544707	lr=2.405914
Epoch[67] Batch [349]	Speed: 8676.136870 samples/sec	accuracy=0.543963	lr=2.405430
Epoch[67] Batch [399]	Speed: 8667.192385 samples/sec	accuracy=0.543685	lr=2.404945
Epoch[67] Batch [449]	Speed: 8729.942618 samples/sec	accuracy=0.543247	lr=2.404460
Epoch[67] Batch [499]	Speed: 8569.901761 samples/sec	accuracy=0.543089	lr=2.403974
Epoch[67] Batch [549]	Speed: 8898.946597 samples/sec	accuracy=0.542849	lr=2.403488
Epoch[67] Batch [599]	Speed: 8656.759233 samples/sec	accuracy=0.542513	lr=2.403001
[Epoch 67] training: accuracy=0.542374
[Epoch 67] speed: 8766 samples/sec	time cost: 150.943017
[Epoch 67] validation: err-top1=0.435410 err-top5=0.186934
Epoch[68] Batch [49]	Speed: 8387.895684 samples/sec	accuracy=0.552090	lr=2.402270
Epoch[68] Batch [99]	Speed: 8882.460669 samples/sec	accuracy=0.549126	lr=2.401782
Epoch[68] Batch [149]	Speed: 8609.871438 samples/sec	accuracy=0.548021	lr=2.401293
Epoch[68] Batch [199]	Speed: 8924.227961 samples/sec	accuracy=0.546685	lr=2.400804
Epoch[68] Batch [249]	Speed: 8676.666901 samples/sec	accuracy=0.545869	lr=2.400314
Epoch[68] Batch [299]	Speed: 8642.369925 samples/sec	accuracy=0.545142	lr=2.399823
Epoch[68] Batch [349]	Speed: 8669.075974 samples/sec	accuracy=0.544982	lr=2.399332
Epoch[68] Batch [399]	Speed: 8638.550801 samples/sec	accuracy=0.544518	lr=2.398841
Epoch[68] Batch [449]	Speed: 8741.471319 samples/sec	accuracy=0.544098	lr=2.398349
Epoch[68] Batch [499]	Speed: 8659.626050 samples/sec	accuracy=0.543962	lr=2.397856
Epoch[68] Batch [549]	Speed: 8627.132544 samples/sec	accuracy=0.543644	lr=2.397363
Epoch[68] Batch [599]	Speed: 8837.590206 samples/sec	accuracy=0.543469	lr=2.396869
[Epoch 68] training: accuracy=0.543108
[Epoch 68] speed: 8690 samples/sec	time cost: 152.244942
[Epoch 68] validation: err-top1=0.482686 err-top5=0.229574
Epoch[69] Batch [49]	Speed: 9054.822394 samples/sec	accuracy=0.551836	lr=2.396118
Epoch[69] Batch [99]	Speed: 8768.175108 samples/sec	accuracy=0.551660	lr=2.395623
Epoch[69] Batch [149]	Speed: 9014.859931 samples/sec	accuracy=0.550049	lr=2.395127
Epoch[69] Batch [199]	Speed: 8678.294908 samples/sec	accuracy=0.549177	lr=2.394631
Epoch[69] Batch [249]	Speed: 8788.288190 samples/sec	accuracy=0.548332	lr=2.394134
Epoch[69] Batch [299]	Speed: 8606.483483 samples/sec	accuracy=0.547249	lr=2.393637
Epoch[69] Batch [349]	Speed: 8597.962854 samples/sec	accuracy=0.546376	lr=2.393139
Epoch[69] Batch [399]	Speed: 8864.867647 samples/sec	accuracy=0.545515	lr=2.392641
Epoch[69] Batch [449]	Speed: 8637.121258 samples/sec	accuracy=0.545031	lr=2.392142
Epoch[69] Batch [499]	Speed: 8833.993086 samples/sec	accuracy=0.544502	lr=2.391642
Epoch[69] Batch [549]	Speed: 8911.200849 samples/sec	accuracy=0.544078	lr=2.391142
Epoch[69] Batch [599]	Speed: 9141.951981 samples/sec	accuracy=0.543732	lr=2.390642
[Epoch 69] training: accuracy=0.543575
[Epoch 69] speed: 8816 samples/sec	time cost: 149.693118
[Epoch 69] validation: err-top1=0.442688 err-top5=0.198039
Epoch[70] Batch [49]	Speed: 8548.256371 samples/sec	accuracy=0.551025	lr=2.389890
Epoch[70] Batch [99]	Speed: 8553.184466 samples/sec	accuracy=0.550186	lr=2.389388
Epoch[70] Batch [149]	Speed: 8718.636311 samples/sec	accuracy=0.548434	lr=2.388885
Epoch[70] Batch [199]	Speed: 9036.872268 samples/sec	accuracy=0.548101	lr=2.388382
Epoch[70] Batch [249]	Speed: 8884.215341 samples/sec	accuracy=0.546557	lr=2.387879
Epoch[70] Batch [299]	Speed: 8984.271895 samples/sec	accuracy=0.545439	lr=2.387374
Epoch[70] Batch [349]	Speed: 8680.244375 samples/sec	accuracy=0.544901	lr=2.386870
Epoch[70] Batch [399]	Speed: 8764.385299 samples/sec	accuracy=0.544747	lr=2.386364
Epoch[70] Batch [449]	Speed: 8611.334792 samples/sec	accuracy=0.544344	lr=2.385859
Epoch[70] Batch [499]	Speed: 8698.465696 samples/sec	accuracy=0.543651	lr=2.385352
Epoch[70] Batch [549]	Speed: 8658.275050 samples/sec	accuracy=0.543366	lr=2.384846
Epoch[70] Batch [599]	Speed: 8657.748834 samples/sec	accuracy=0.543049	lr=2.384338
[Epoch 70] training: accuracy=0.543011
[Epoch 70] speed: 8725 samples/sec	time cost: 151.796766
[Epoch 70] validation: err-top1=0.435781 err-top5=0.188516
Epoch[71] Batch [49]	Speed: 8882.136269 samples/sec	accuracy=0.552168	lr=2.383566
Epoch[71] Batch [99]	Speed: 8747.185368 samples/sec	accuracy=0.550039	lr=2.383057
Epoch[71] Batch [149]	Speed: 8732.123070 samples/sec	accuracy=0.548923	lr=2.382548
Epoch[71] Batch [199]	Speed: 8642.051695 samples/sec	accuracy=0.548142	lr=2.382038
Epoch[71] Batch [249]	Speed: 8719.400247 samples/sec	accuracy=0.546904	lr=2.381528
Epoch[71] Batch [299]	Speed: 8709.338932 samples/sec	accuracy=0.546453	lr=2.381017
Epoch[71] Batch [349]	Speed: 8526.894917 samples/sec	accuracy=0.545982	lr=2.380505
Epoch[71] Batch [399]	Speed: 8484.659650 samples/sec	accuracy=0.545698	lr=2.379993
Epoch[71] Batch [449]	Speed: 8966.635690 samples/sec	accuracy=0.544907	lr=2.379481
Epoch[71] Batch [499]	Speed: 8866.717514 samples/sec	accuracy=0.544302	lr=2.378967
Epoch[71] Batch [549]	Speed: 8800.805951 samples/sec	accuracy=0.544138	lr=2.378454
Epoch[71] Batch [599]	Speed: 8762.924894 samples/sec	accuracy=0.543899	lr=2.377940
[Epoch 71] training: accuracy=0.544002
[Epoch 71] speed: 8733 samples/sec	time cost: 151.437474
[Epoch 71] validation: err-top1=0.415548 err-top5=0.173971
Epoch[72] Batch [49]	Speed: 8393.287258 samples/sec	accuracy=0.552998	lr=2.377157
Epoch[72] Batch [99]	Speed: 8540.926245 samples/sec	accuracy=0.550864	lr=2.376641
Epoch[72] Batch [149]	Speed: 8635.880933 samples/sec	accuracy=0.549495	lr=2.376125
Epoch[72] Batch [199]	Speed: 8826.732523 samples/sec	accuracy=0.547788	lr=2.375609
Epoch[72] Batch [249]	Speed: 8843.415170 samples/sec	accuracy=0.547422	lr=2.375092
Epoch[72] Batch [299]	Speed: 8511.400074 samples/sec	accuracy=0.546170	lr=2.374574
Epoch[72] Batch [349]	Speed: 8672.888866 samples/sec	accuracy=0.545558	lr=2.374056
Epoch[72] Batch [399]	Speed: 8652.394672 samples/sec	accuracy=0.545077	lr=2.373537
Epoch[72] Batch [449]	Speed: 8834.688142 samples/sec	accuracy=0.544263	lr=2.373017
Epoch[72] Batch [499]	Speed: 8775.466738 samples/sec	accuracy=0.544104	lr=2.372498
Epoch[72] Batch [549]	Speed: 8709.918245 samples/sec	accuracy=0.543958	lr=2.371977
Epoch[72] Batch [599]	Speed: 8727.741440 samples/sec	accuracy=0.543680	lr=2.371456
[Epoch 72] training: accuracy=0.543649
[Epoch 72] speed: 8684 samples/sec	time cost: 152.255026
[Epoch 72] validation: err-top1=0.459668 err-top5=0.209766
Epoch[73] Batch [49]	Speed: 8292.283571 samples/sec	accuracy=0.551025	lr=2.370674
Epoch[73] Batch [99]	Speed: 8939.714883 samples/sec	accuracy=0.550269	lr=2.370152
Epoch[73] Batch [149]	Speed: 8851.769765 samples/sec	accuracy=0.549460	lr=2.369629
Epoch[73] Batch [199]	Speed: 8861.739747 samples/sec	accuracy=0.548809	lr=2.369105
Epoch[73] Batch [249]	Speed: 8749.501526 samples/sec	accuracy=0.547875	lr=2.368582
Epoch[73] Batch [299]	Speed: 8670.568276 samples/sec	accuracy=0.547238	lr=2.368057
Epoch[73] Batch [349]	Speed: 8606.988996 samples/sec	accuracy=0.546839	lr=2.367532
Epoch[73] Batch [399]	Speed: 8375.904109 samples/sec	accuracy=0.546458	lr=2.367007
Epoch[73] Batch [449]	Speed: 8962.360836 samples/sec	accuracy=0.546406	lr=2.366481
Epoch[73] Batch [499]	Speed: 8677.838843 samples/sec	accuracy=0.546014	lr=2.365954
Epoch[73] Batch [549]	Speed: 8602.203226 samples/sec	accuracy=0.545571	lr=2.365427
Epoch[73] Batch [599]	Speed: 8671.754679 samples/sec	accuracy=0.545120	lr=2.364899
[Epoch 73] training: accuracy=0.544901
[Epoch 73] speed: 8691 samples/sec	time cost: 152.145939
[Epoch 73] validation: err-top1=0.439575 err-top5=0.195638
Epoch[74] Batch [49]	Speed: 8491.621653 samples/sec	accuracy=0.554004	lr=2.364096
Epoch[74] Batch [99]	Speed: 8733.584236 samples/sec	accuracy=0.552573	lr=2.363567
Epoch[74] Batch [149]	Speed: 8598.956273 samples/sec	accuracy=0.550459	lr=2.363038
Epoch[74] Batch [199]	Speed: 8837.518558 samples/sec	accuracy=0.549167	lr=2.362508
Epoch[74] Batch [249]	Speed: 8669.643292 samples/sec	accuracy=0.548277	lr=2.361977
Epoch[74] Batch [299]	Speed: 8961.136405 samples/sec	accuracy=0.548265	lr=2.361446
Epoch[74] Batch [349]	Speed: 8580.032891 samples/sec	accuracy=0.548105	lr=2.360915
Epoch[74] Batch [399]	Speed: 8930.346171 samples/sec	accuracy=0.547811	lr=2.360382
Epoch[74] Batch [449]	Speed: 8732.326351 samples/sec	accuracy=0.547050	lr=2.359850
Epoch[74] Batch [499]	Speed: 8869.589008 samples/sec	accuracy=0.546527	lr=2.359316
Epoch[74] Batch [549]	Speed: 8814.734293 samples/sec	accuracy=0.546050	lr=2.358783
Epoch[74] Batch [599]	Speed: 8755.162902 samples/sec	accuracy=0.545586	lr=2.358248
[Epoch 74] training: accuracy=0.545534
[Epoch 74] speed: 8734 samples/sec	time cost: 151.454294
[Epoch 74] validation: err-top1=0.437793 err-top5=0.196406
Epoch[75] Batch [49]	Speed: 8626.439094 samples/sec	accuracy=0.554561	lr=2.357446
Epoch[75] Batch [99]	Speed: 8548.420555 samples/sec	accuracy=0.552236	lr=2.356910
Epoch[75] Batch [149]	Speed: 8605.953197 samples/sec	accuracy=0.550462	lr=2.356374
Epoch[75] Batch [199]	Speed: 8798.660643 samples/sec	accuracy=0.549709	lr=2.355838
Epoch[75] Batch [249]	Speed: 8704.506792 samples/sec	accuracy=0.548820	lr=2.355300
Epoch[75] Batch [299]	Speed: 8630.813404 samples/sec	accuracy=0.548162	lr=2.354763
Epoch[75] Batch [349]	Speed: 8758.426396 samples/sec	accuracy=0.547437	lr=2.354224
Epoch[75] Batch [399]	Speed: 8619.028121 samples/sec	accuracy=0.546875	lr=2.353686
Epoch[75] Batch [449]	Speed: 8594.997389 samples/sec	accuracy=0.546243	lr=2.353146
Epoch[75] Batch [499]	Speed: 8714.626607 samples/sec	accuracy=0.545814	lr=2.352606
Epoch[75] Batch [549]	Speed: 8582.096220 samples/sec	accuracy=0.545669	lr=2.352066
Epoch[75] Batch [599]	Speed: 8594.874066 samples/sec	accuracy=0.545148	lr=2.351525
[Epoch 75] training: accuracy=0.544953
[Epoch 75] speed: 8648 samples/sec	time cost: 153.071005
[Epoch 75] validation: err-top1=0.462911 err-top5=0.213013
Epoch[76] Batch [49]	Speed: 8552.245190 samples/sec	accuracy=0.553740	lr=2.350702
Epoch[76] Batch [99]	Speed: 8648.585842 samples/sec	accuracy=0.552329	lr=2.350160
Epoch[76] Batch [149]	Speed: 8512.637629 samples/sec	accuracy=0.551110	lr=2.349617
Epoch[76] Batch [199]	Speed: 8571.789998 samples/sec	accuracy=0.550254	lr=2.349074
Epoch[76] Batch [249]	Speed: 8591.180657 samples/sec	accuracy=0.549137	lr=2.348530
Epoch[76] Batch [299]	Speed: 8743.789793 samples/sec	accuracy=0.548545	lr=2.347986
Epoch[76] Batch [349]	Speed: 8825.127415 samples/sec	accuracy=0.548185	lr=2.347441
Epoch[76] Batch [399]	Speed: 8547.126310 samples/sec	accuracy=0.547877	lr=2.346895
Epoch[76] Batch [449]	Speed: 8527.945129 samples/sec	accuracy=0.547475	lr=2.346349
Epoch[76] Batch [499]	Speed: 8606.891545 samples/sec	accuracy=0.546860	lr=2.345803
Epoch[76] Batch [549]	Speed: 8565.385683 samples/sec	accuracy=0.546716	lr=2.345256
Epoch[76] Batch [599]	Speed: 8575.586607 samples/sec	accuracy=0.546353	lr=2.344709
[Epoch 76] training: accuracy=0.546170
[Epoch 76] speed: 8609 samples/sec	time cost: 153.354444
[Epoch 76] validation: err-top1=0.414042 err-top5=0.178141
Epoch[77] Batch [49]	Speed: 8384.024951 samples/sec	accuracy=0.555352	lr=2.343886
Epoch[77] Batch [99]	Speed: 8778.563268 samples/sec	accuracy=0.553950	lr=2.343338
Epoch[77] Batch [149]	Speed: 8849.575108 samples/sec	accuracy=0.552227	lr=2.342788
Epoch[77] Batch [199]	Speed: 8707.132067 samples/sec	accuracy=0.551091	lr=2.342238
Epoch[77] Batch [249]	Speed: 8664.365302 samples/sec	accuracy=0.549875	lr=2.341688
Epoch[77] Batch [299]	Speed: 8528.616907 samples/sec	accuracy=0.549014	lr=2.341137
Epoch[77] Batch [349]	Speed: 8643.444428 samples/sec	accuracy=0.548375	lr=2.340586
Epoch[77] Batch [399]	Speed: 9007.332960 samples/sec	accuracy=0.547662	lr=2.340034
Epoch[77] Batch [449]	Speed: 8903.994136 samples/sec	accuracy=0.546905	lr=2.339481
Epoch[77] Batch [499]	Speed: 8683.005821 samples/sec	accuracy=0.546589	lr=2.338928
Epoch[77] Batch [549]	Speed: 8929.299402 samples/sec	accuracy=0.546381	lr=2.338375
Epoch[77] Batch [599]	Speed: 8723.503469 samples/sec	accuracy=0.546141	lr=2.337821
[Epoch 77] training: accuracy=0.545938
[Epoch 77] speed: 8732 samples/sec	time cost: 151.664706
[Epoch 77] validation: err-top1=0.431211 err-top5=0.187285
Epoch[78] Batch [49]	Speed: 8975.732957 samples/sec	accuracy=0.556914	lr=2.336978
Epoch[78] Batch [99]	Speed: 8834.436092 samples/sec	accuracy=0.554849	lr=2.336423
Epoch[78] Batch [149]	Speed: 8730.812006 samples/sec	accuracy=0.552952	lr=2.335867
Epoch[78] Batch [199]	Speed: 8410.435709 samples/sec	accuracy=0.551570	lr=2.335310
Epoch[78] Batch [249]	Speed: 8561.965903 samples/sec	accuracy=0.550488	lr=2.334753
Epoch[78] Batch [299]	Speed: 8536.621343 samples/sec	accuracy=0.549071	lr=2.334196
Epoch[78] Batch [349]	Speed: 8765.947990 samples/sec	accuracy=0.547900	lr=2.333638
Epoch[78] Batch [399]	Speed: 8281.741508 samples/sec	accuracy=0.547433	lr=2.333080
Epoch[78] Batch [449]	Speed: 8659.339719 samples/sec	accuracy=0.546850	lr=2.332521
Epoch[78] Batch [499]	Speed: 8581.617803 samples/sec	accuracy=0.546679	lr=2.331961
Epoch[78] Batch [549]	Speed: 8598.230681 samples/sec	accuracy=0.546382	lr=2.331401
Epoch[78] Batch [599]	Speed: 8545.246882 samples/sec	accuracy=0.546184	lr=2.330841
[Epoch 78] training: accuracy=0.546203
[Epoch 78] speed: 8623 samples/sec	time cost: 153.281976
[Epoch 78] validation: err-top1=0.421855 err-top5=0.178731
Epoch[79] Batch [49]	Speed: 8775.435182 samples/sec	accuracy=0.554697	lr=2.329988
Epoch[79] Batch [99]	Speed: 8589.931671 samples/sec	accuracy=0.552700	lr=2.329426
Epoch[79] Batch [149]	Speed: 8810.428607 samples/sec	accuracy=0.551592	lr=2.328864
Epoch[79] Batch [199]	Speed: 8954.659671 samples/sec	accuracy=0.551262	lr=2.328301
Epoch[79] Batch [249]	Speed: 8873.098960 samples/sec	accuracy=0.549676	lr=2.327738
Epoch[79] Batch [299]	Speed: 8605.569190 samples/sec	accuracy=0.549377	lr=2.327174
Epoch[79] Batch [349]	Speed: 8680.706483 samples/sec	accuracy=0.549022	lr=2.326609
Epoch[79] Batch [399]	Speed: 8681.649096 samples/sec	accuracy=0.548077	lr=2.326044
Epoch[79] Batch [449]	Speed: 8674.532274 samples/sec	accuracy=0.547910	lr=2.325479
Epoch[79] Batch [499]	Speed: 8920.950937 samples/sec	accuracy=0.547790	lr=2.324913
Epoch[79] Batch [549]	Speed: 8827.221789 samples/sec	accuracy=0.547502	lr=2.324347
Epoch[79] Batch [599]	Speed: 8800.439882 samples/sec	accuracy=0.547207	lr=2.323780
[Epoch 79] training: accuracy=0.547254
[Epoch 79] speed: 8768 samples/sec	time cost: 150.958050
[Epoch 79] validation: err-top1=0.413613 err-top5=0.175078
Epoch[80] Batch [49]	Speed: 8971.333241 samples/sec	accuracy=0.557090	lr=2.322928
Epoch[80] Batch [99]	Speed: 8692.987173 samples/sec	accuracy=0.554614	lr=2.322360
Epoch[80] Batch [149]	Speed: 8550.184953 samples/sec	accuracy=0.553242	lr=2.321791
Epoch[80] Batch [199]	Speed: 9002.376679 samples/sec	accuracy=0.551975	lr=2.321222
Epoch[80] Batch [249]	Speed: 8517.195359 samples/sec	accuracy=0.551484	lr=2.320652
Epoch[80] Batch [299]	Speed: 8611.615022 samples/sec	accuracy=0.550981	lr=2.320082
Epoch[80] Batch [349]	Speed: 8942.944819 samples/sec	accuracy=0.549980	lr=2.319511
Epoch[80] Batch [399]	Speed: 8786.871761 samples/sec	accuracy=0.549259	lr=2.318940
Epoch[80] Batch [449]	Speed: 9112.023467 samples/sec	accuracy=0.548697	lr=2.318368
Epoch[80] Batch [499]	Speed: 8840.239976 samples/sec	accuracy=0.547970	lr=2.317796
Epoch[80] Batch [549]	Speed: 8635.360040 samples/sec	accuracy=0.547958	lr=2.317223
Epoch[80] Batch [599]	Speed: 8785.842000 samples/sec	accuracy=0.547786	lr=2.316650
[Epoch 80] training: accuracy=0.547559
[Epoch 80] speed: 8788 samples/sec	time cost: 150.611761
[Epoch 80] validation: err-top1=0.417908 err-top5=0.180400
Epoch[81] Batch [49]	Speed: 9035.199525 samples/sec	accuracy=0.557441	lr=2.315777
Epoch[81] Batch [99]	Speed: 8351.665095 samples/sec	accuracy=0.553843	lr=2.315203
Epoch[81] Batch [149]	Speed: 8973.362232 samples/sec	accuracy=0.552891	lr=2.314628
Epoch[81] Batch [199]	Speed: 8604.100728 samples/sec	accuracy=0.552397	lr=2.314052
Epoch[81] Batch [249]	Speed: 8499.506641 samples/sec	accuracy=0.551621	lr=2.313476
Epoch[81] Batch [299]	Speed: 8818.068790 samples/sec	accuracy=0.550685	lr=2.312899
Epoch[81] Batch [349]	Speed: 8672.955767 samples/sec	accuracy=0.550403	lr=2.312322
Epoch[81] Batch [399]	Speed: 8859.095713 samples/sec	accuracy=0.549479	lr=2.311744
Epoch[81] Batch [449]	Speed: 8904.131105 samples/sec	accuracy=0.548483	lr=2.311166
Epoch[81] Batch [499]	Speed: 8895.324791 samples/sec	accuracy=0.548163	lr=2.310587
Epoch[81] Batch [549]	Speed: 8827.325926 samples/sec	accuracy=0.547915	lr=2.310008
Epoch[81] Batch [599]	Speed: 8780.859992 samples/sec	accuracy=0.547799	lr=2.309428
[Epoch 81] training: accuracy=0.547602
[Epoch 81] speed: 8768 samples/sec	time cost: 150.691318
[Epoch 81] validation: err-top1=0.440369 err-top5=0.197347
Epoch[82] Batch [49]	Speed: 9055.057967 samples/sec	accuracy=0.553604	lr=2.308558
Epoch[82] Batch [99]	Speed: 8628.672326 samples/sec	accuracy=0.554746	lr=2.307977
Epoch[82] Batch [149]	Speed: 8745.041368 samples/sec	accuracy=0.553581	lr=2.307396
Epoch[82] Batch [199]	Speed: 8585.518201 samples/sec	accuracy=0.552761	lr=2.306814
Epoch[82] Batch [249]	Speed: 9220.571722 samples/sec	accuracy=0.552115	lr=2.306231
Epoch[82] Batch [299]	Speed: 8731.980868 samples/sec	accuracy=0.551489	lr=2.305648
Epoch[82] Batch [349]	Speed: 8542.852200 samples/sec	accuracy=0.551289	lr=2.305065
Epoch[82] Batch [399]	Speed: 8692.471507 samples/sec	accuracy=0.550791	lr=2.304481
Epoch[82] Batch [449]	Speed: 8854.345538 samples/sec	accuracy=0.550192	lr=2.303896
Epoch[82] Batch [499]	Speed: 8882.090164 samples/sec	accuracy=0.549507	lr=2.303311
Epoch[82] Batch [549]	Speed: 8618.778368 samples/sec	accuracy=0.548737	lr=2.302726
Epoch[82] Batch [599]	Speed: 9076.466028 samples/sec	accuracy=0.548274	lr=2.302140
[Epoch 82] training: accuracy=0.548096
[Epoch 82] speed: 8797 samples/sec	time cost: 150.610576
[Epoch 82] validation: err-top1=0.446504 err-top5=0.198340
Epoch[83] Batch [49]	Speed: 9109.748116 samples/sec	accuracy=0.558652	lr=2.301248
Epoch[83] Batch [99]	Speed: 8551.046144 samples/sec	accuracy=0.554736	lr=2.300661
Epoch[83] Batch [149]	Speed: 8715.351109 samples/sec	accuracy=0.554017	lr=2.300073
Epoch[83] Batch [199]	Speed: 8703.924495 samples/sec	accuracy=0.552507	lr=2.299485
Epoch[83] Batch [249]	Speed: 8564.352699 samples/sec	accuracy=0.551869	lr=2.298896
Epoch[83] Batch [299]	Speed: 8880.521782 samples/sec	accuracy=0.551102	lr=2.298307
Epoch[83] Batch [349]	Speed: 8536.349536 samples/sec	accuracy=0.550259	lr=2.297717
Epoch[83] Batch [399]	Speed: 8807.471023 samples/sec	accuracy=0.549609	lr=2.297127
Epoch[83] Batch [449]	Speed: 8768.223081 samples/sec	accuracy=0.549263	lr=2.296536
Epoch[83] Batch [499]	Speed: 8588.981387 samples/sec	accuracy=0.549042	lr=2.295945
Epoch[83] Batch [549]	Speed: 8768.347669 samples/sec	accuracy=0.548909	lr=2.295353
Epoch[83] Batch [599]	Speed: 8574.553387 samples/sec	accuracy=0.548683	lr=2.294760
[Epoch 83] training: accuracy=0.548613
[Epoch 83] speed: 8709 samples/sec	time cost: 151.664557
[Epoch 83] validation: err-top1=0.469279 err-top5=0.218119
Epoch[84] Batch [49]	Speed: 9232.453928 samples/sec	accuracy=0.558467	lr=2.293871
Epoch[84] Batch [99]	Speed: 8691.420837 samples/sec	accuracy=0.555522	lr=2.293278
Epoch[84] Batch [149]	Speed: 8753.294170 samples/sec	accuracy=0.554010	lr=2.292684
Epoch[84] Batch [199]	Speed: 9013.962947 samples/sec	accuracy=0.553015	lr=2.292089
Epoch[84] Batch [249]	Speed: 8634.460779 samples/sec	accuracy=0.552109	lr=2.291494
Epoch[84] Batch [299]	Speed: 8758.186537 samples/sec	accuracy=0.551566	lr=2.290899
Epoch[84] Batch [349]	Speed: 8492.238689 samples/sec	accuracy=0.551196	lr=2.290303
Epoch[84] Batch [399]	Speed: 8699.936414 samples/sec	accuracy=0.550985	lr=2.289706
Epoch[84] Batch [449]	Speed: 8739.172212 samples/sec	accuracy=0.550545	lr=2.289109
Epoch[84] Batch [499]	Speed: 8890.819851 samples/sec	accuracy=0.550350	lr=2.288512
Epoch[84] Batch [549]	Speed: 8771.834367 samples/sec	accuracy=0.550037	lr=2.287914
Epoch[84] Batch [599]	Speed: 8824.118218 samples/sec	accuracy=0.549732	lr=2.287315
[Epoch 84] training: accuracy=0.549481
[Epoch 84] speed: 8784 samples/sec	time cost: 150.810985
[Epoch 84] validation: err-top1=0.415117 err-top5=0.177617
Epoch[85] Batch [49]	Speed: 9067.216359 samples/sec	accuracy=0.558398	lr=2.286405
Epoch[85] Batch [99]	Speed: 9106.026474 samples/sec	accuracy=0.556025	lr=2.285805
Epoch[85] Batch [149]	Speed: 8739.361595 samples/sec	accuracy=0.555218	lr=2.285205
Epoch[85] Batch [199]	Speed: 8695.724516 samples/sec	accuracy=0.554905	lr=2.284604
Epoch[85] Batch [249]	Speed: 8747.023079 samples/sec	accuracy=0.553525	lr=2.284003
Epoch[85] Batch [299]	Speed: 8917.951287 samples/sec	accuracy=0.552609	lr=2.283401
Epoch[85] Batch [349]	Speed: 8681.335512 samples/sec	accuracy=0.551821	lr=2.282799
Epoch[85] Batch [399]	Speed: 9076.128069 samples/sec	accuracy=0.551271	lr=2.282196
Epoch[85] Batch [449]	Speed: 8843.264586 samples/sec	accuracy=0.550667	lr=2.281593
Epoch[85] Batch [499]	Speed: 8657.945001 samples/sec	accuracy=0.550288	lr=2.280989
Epoch[85] Batch [549]	Speed: 9075.585125 samples/sec	accuracy=0.549909	lr=2.280385
Epoch[85] Batch [599]	Speed: 8583.426295 samples/sec	accuracy=0.549804	lr=2.279780
[Epoch 85] training: accuracy=0.549458
[Epoch 85] speed: 8838 samples/sec	time cost: 149.558207
[Epoch 85] validation: err-top1=0.459412 err-top5=0.212463
Epoch[86] Batch [49]	Speed: 8675.416420 samples/sec	accuracy=0.555156	lr=2.278860
Epoch[86] Batch [99]	Speed: 8841.262873 samples/sec	accuracy=0.554341	lr=2.278254
Epoch[86] Batch [149]	Speed: 8784.797386 samples/sec	accuracy=0.552992	lr=2.277648
Epoch[86] Batch [199]	Speed: 8612.813843 samples/sec	accuracy=0.552810	lr=2.277041
Epoch[86] Batch [249]	Speed: 8575.417268 samples/sec	accuracy=0.552602	lr=2.276434
Epoch[86] Batch [299]	Speed: 8605.343665 samples/sec	accuracy=0.552347	lr=2.275826
Epoch[86] Batch [349]	Speed: 8877.020452 samples/sec	accuracy=0.551864	lr=2.275218
Epoch[86] Batch [399]	Speed: 8785.411761 samples/sec	accuracy=0.551327	lr=2.274609
Epoch[86] Batch [449]	Speed: 8674.913700 samples/sec	accuracy=0.551025	lr=2.274000
Epoch[86] Batch [499]	Speed: 8622.521169 samples/sec	accuracy=0.550859	lr=2.273390
Epoch[86] Batch [549]	Speed: 8812.354362 samples/sec	accuracy=0.550426	lr=2.272779
Epoch[86] Batch [599]	Speed: 8774.386230 samples/sec	accuracy=0.549925	lr=2.272169
[Epoch 86] training: accuracy=0.549905
[Epoch 86] speed: 8721 samples/sec	time cost: 151.835125
[Epoch 86] validation: err-top1=0.420176 err-top5=0.179941
Epoch[87] Batch [49]	Speed: 9021.638393 samples/sec	accuracy=0.555039	lr=2.271252
Epoch[87] Batch [99]	Speed: 8323.129195 samples/sec	accuracy=0.554609	lr=2.270640
Epoch[87] Batch [149]	Speed: 8595.143077 samples/sec	accuracy=0.554121	lr=2.270027
Epoch[87] Batch [199]	Speed: 8720.172461 samples/sec	accuracy=0.553274	lr=2.269414
Epoch[87] Batch [249]	Speed: 8559.013274 samples/sec	accuracy=0.552424	lr=2.268801
Epoch[87] Batch [299]	Speed: 8546.314374 samples/sec	accuracy=0.552157	lr=2.268187
Epoch[87] Batch [349]	Speed: 8652.468404 samples/sec	accuracy=0.551297	lr=2.267572
Epoch[87] Batch [399]	Speed: 8323.715211 samples/sec	accuracy=0.550499	lr=2.266957
Epoch[87] Batch [449]	Speed: 8661.518054 samples/sec	accuracy=0.550367	lr=2.266342
Epoch[87] Batch [499]	Speed: 8399.766565 samples/sec	accuracy=0.549958	lr=2.265726
Epoch[87] Batch [549]	Speed: 8519.021406 samples/sec	accuracy=0.549577	lr=2.265110
Epoch[87] Batch [599]	Speed: 8636.239170 samples/sec	accuracy=0.549637	lr=2.264493
[Epoch 87] training: accuracy=0.549367
[Epoch 87] speed: 8579 samples/sec	time cost: 154.025070
[Epoch 87] validation: err-top1=0.415894 err-top5=0.174683
Epoch[88] Batch [49]	Speed: 8134.116285 samples/sec	accuracy=0.558994	lr=2.263554
Epoch[88] Batch [99]	Speed: 8581.001598 samples/sec	accuracy=0.557676	lr=2.262936
Epoch[88] Batch [149]	Speed: 8499.290845 samples/sec	accuracy=0.555205	lr=2.262318
Epoch[88] Batch [199]	Speed: 8691.188502 samples/sec	accuracy=0.554863	lr=2.261699
Epoch[88] Batch [249]	Speed: 8758.889364 samples/sec	accuracy=0.553736	lr=2.261079
Epoch[88] Batch [299]	Speed: 8569.270653 samples/sec	accuracy=0.552959	lr=2.260459
Epoch[88] Batch [349]	Speed: 8607.735044 samples/sec	accuracy=0.552249	lr=2.259839
Epoch[88] Batch [399]	Speed: 8618.721293 samples/sec	accuracy=0.551710	lr=2.259218
Epoch[88] Batch [449]	Speed: 8436.728662 samples/sec	accuracy=0.551699	lr=2.258596
Epoch[88] Batch [499]	Speed: 8868.639572 samples/sec	accuracy=0.551098	lr=2.257974
Epoch[88] Batch [549]	Speed: 8916.363183 samples/sec	accuracy=0.550765	lr=2.257352
Epoch[88] Batch [599]	Speed: 8602.438925 samples/sec	accuracy=0.550608	lr=2.256729
[Epoch 88] training: accuracy=0.550380
[Epoch 88] speed: 8602 samples/sec	time cost: 153.586274
[Epoch 88] validation: err-top1=0.409871 err-top5=0.175700
Epoch[89] Batch [49]	Speed: 8623.092452 samples/sec	accuracy=0.557490	lr=2.255794
Epoch[89] Batch [99]	Speed: 8516.191527 samples/sec	accuracy=0.558188	lr=2.255170
Epoch[89] Batch [149]	Speed: 8736.586407 samples/sec	accuracy=0.556436	lr=2.254545
Epoch[89] Batch [199]	Speed: 8692.686317 samples/sec	accuracy=0.554287	lr=2.253920
Epoch[89] Batch [249]	Speed: 8604.502014 samples/sec	accuracy=0.553785	lr=2.253295
Epoch[89] Batch [299]	Speed: 8415.328641 samples/sec	accuracy=0.552917	lr=2.252669
Epoch[89] Batch [349]	Speed: 8773.372833 samples/sec	accuracy=0.552101	lr=2.252042
Epoch[89] Batch [399]	Speed: 8469.594695 samples/sec	accuracy=0.551622	lr=2.251415
Epoch[89] Batch [449]	Speed: 8542.062652 samples/sec	accuracy=0.551366	lr=2.250788
Epoch[89] Batch [499]	Speed: 8686.946765 samples/sec	accuracy=0.551293	lr=2.250160
Epoch[89] Batch [549]	Speed: 8552.890313 samples/sec	accuracy=0.551124	lr=2.249532
Epoch[89] Batch [599]	Speed: 8903.079768 samples/sec	accuracy=0.550976	lr=2.248903
[Epoch 89] training: accuracy=0.550968
[Epoch 89] speed: 8633 samples/sec	time cost: 153.497862
[Epoch 89] validation: err-top1=0.414336 err-top5=0.174316
Epoch[90] Batch [49]	Speed: 8303.754495 samples/sec	accuracy=0.558232	lr=2.247946
Epoch[90] Batch [99]	Speed: 8501.960387 samples/sec	accuracy=0.558027	lr=2.247316
Epoch[90] Batch [149]	Speed: 8934.715804 samples/sec	accuracy=0.555348	lr=2.246685
Epoch[90] Batch [199]	Speed: 8774.667491 samples/sec	accuracy=0.554326	lr=2.246054
Epoch[90] Batch [249]	Speed: 8596.622420 samples/sec	accuracy=0.553256	lr=2.245423
Epoch[90] Batch [299]	Speed: 8524.880039 samples/sec	accuracy=0.553099	lr=2.244791
Epoch[90] Batch [349]	Speed: 8747.344277 samples/sec	accuracy=0.552705	lr=2.244159
Epoch[90] Batch [399]	Speed: 8491.140848 samples/sec	accuracy=0.552592	lr=2.243526
Epoch[90] Batch [449]	Speed: 8564.965832 samples/sec	accuracy=0.552314	lr=2.242892
Epoch[90] Batch [499]	Speed: 8587.438567 samples/sec	accuracy=0.551892	lr=2.242258
Epoch[90] Batch [549]	Speed: 8462.707767 samples/sec	accuracy=0.551351	lr=2.241624
Epoch[90] Batch [599]	Speed: 8498.704736 samples/sec	accuracy=0.551240	lr=2.240989
[Epoch 90] training: accuracy=0.551266
[Epoch 90] speed: 8586 samples/sec	time cost: 153.748230
[Epoch 90] validation: err-top1=0.418416 err-top5=0.180786
Epoch[91] Batch [49]	Speed: 8628.506432 samples/sec	accuracy=0.559756	lr=2.240036
Epoch[91] Batch [99]	Speed: 8703.465027 samples/sec	accuracy=0.560068	lr=2.239400
Epoch[91] Batch [149]	Speed: 8368.769114 samples/sec	accuracy=0.557507	lr=2.238764
Epoch[91] Batch [199]	Speed: 8470.479653 samples/sec	accuracy=0.556143	lr=2.238127
Epoch[91] Batch [249]	Speed: 8609.937371 samples/sec	accuracy=0.555176	lr=2.237490
Epoch[91] Batch [299]	Speed: 8642.617743 samples/sec	accuracy=0.554390	lr=2.236852
Epoch[91] Batch [349]	Speed: 8635.274099 samples/sec	accuracy=0.553657	lr=2.236213
Epoch[91] Batch [399]	Speed: 8734.461810 samples/sec	accuracy=0.553254	lr=2.235575
Epoch[91] Batch [449]	Speed: 8623.642340 samples/sec	accuracy=0.552798	lr=2.234935
Epoch[91] Batch [499]	Speed: 8495.654407 samples/sec	accuracy=0.552315	lr=2.234296
Epoch[91] Batch [549]	Speed: 8685.542437 samples/sec	accuracy=0.551884	lr=2.233656
Epoch[91] Batch [599]	Speed: 8586.233067 samples/sec	accuracy=0.551584	lr=2.233015
[Epoch 91] training: accuracy=0.551417
[Epoch 91] speed: 8606 samples/sec	time cost: 153.860611
[Epoch 91] validation: err-top1=0.425293 err-top5=0.183066
Epoch[92] Batch [49]	Speed: 8367.267546 samples/sec	accuracy=0.559277	lr=2.232040
Epoch[92] Batch [99]	Speed: 9003.564843 samples/sec	accuracy=0.557173	lr=2.231398
Epoch[92] Batch [149]	Speed: 8618.216995 samples/sec	accuracy=0.556191	lr=2.230756
Epoch[92] Batch [199]	Speed: 8486.058951 samples/sec	accuracy=0.555334	lr=2.230113
Epoch[92] Batch [249]	Speed: 8410.409523 samples/sec	accuracy=0.554691	lr=2.229470
Epoch[92] Batch [299]	Speed: 8351.904804 samples/sec	accuracy=0.553949	lr=2.228826
Epoch[92] Batch [349]	Speed: 8752.894583 samples/sec	accuracy=0.553539	lr=2.228182
Epoch[92] Batch [399]	Speed: 8790.805014 samples/sec	accuracy=0.553340	lr=2.227538
Epoch[92] Batch [449]	Speed: 8858.574587 samples/sec	accuracy=0.553074	lr=2.226892
Epoch[92] Batch [499]	Speed: 8711.423759 samples/sec	accuracy=0.553021	lr=2.226247
Epoch[92] Batch [549]	Speed: 8766.700019 samples/sec	accuracy=0.552766	lr=2.225601
Epoch[92] Batch [599]	Speed: 8743.794955 samples/sec	accuracy=0.552356	lr=2.224954
[Epoch 92] training: accuracy=0.552204
[Epoch 92] speed: 8653 samples/sec	time cost: 152.587606
[Epoch 92] validation: err-top1=0.417216 err-top5=0.175680
Epoch[93] Batch [49]	Speed: 8471.407237 samples/sec	accuracy=0.563516	lr=2.223984
Epoch[93] Batch [99]	Speed: 8716.731127 samples/sec	accuracy=0.560259	lr=2.223336
Epoch[93] Batch [149]	Speed: 8522.593321 samples/sec	accuracy=0.557936	lr=2.222688
Epoch[93] Batch [199]	Speed: 8410.659698 samples/sec	accuracy=0.556743	lr=2.222039
Epoch[93] Batch [249]	Speed: 8543.245584 samples/sec	accuracy=0.555670	lr=2.221390
Epoch[93] Batch [299]	Speed: 8565.801987 samples/sec	accuracy=0.555106	lr=2.220741
Epoch[93] Batch [349]	Speed: 8649.788360 samples/sec	accuracy=0.554625	lr=2.220091
Epoch[93] Batch [399]	Speed: 8704.521611 samples/sec	accuracy=0.554442	lr=2.219441
Epoch[93] Batch [449]	Speed: 8620.862453 samples/sec	accuracy=0.553961	lr=2.218790
Epoch[93] Batch [499]	Speed: 8510.311606 samples/sec	accuracy=0.553877	lr=2.218138
Epoch[93] Batch [549]	Speed: 8536.133222 samples/sec	accuracy=0.553493	lr=2.217487
Epoch[93] Batch [599]	Speed: 8597.091499 samples/sec	accuracy=0.553095	lr=2.216834
[Epoch 93] training: accuracy=0.553026
[Epoch 93] speed: 8580 samples/sec	time cost: 154.070312
[Epoch 93] validation: err-top1=0.415731 err-top5=0.175863
Epoch[94] Batch [49]	Speed: 8301.911238 samples/sec	accuracy=0.562900	lr=2.215842
Epoch[94] Batch [99]	Speed: 8491.908417 samples/sec	accuracy=0.559941	lr=2.215189
Epoch[94] Batch [149]	Speed: 8632.827486 samples/sec	accuracy=0.558750	lr=2.214535
Epoch[94] Batch [199]	Speed: 8710.039769 samples/sec	accuracy=0.557505	lr=2.213880
Epoch[94] Batch [249]	Speed: 8602.674465 samples/sec	accuracy=0.556156	lr=2.213226
Epoch[94] Batch [299]	Speed: 8770.174840 samples/sec	accuracy=0.555635	lr=2.212570
Epoch[94] Batch [349]	Speed: 8515.278084 samples/sec	accuracy=0.555059	lr=2.211915
Epoch[94] Batch [399]	Speed: 8476.040845 samples/sec	accuracy=0.554482	lr=2.211258
Epoch[94] Batch [449]	Speed: 8590.067223 samples/sec	accuracy=0.553664	lr=2.210602
Epoch[94] Batch [499]	Speed: 8574.167042 samples/sec	accuracy=0.553334	lr=2.209945
Epoch[94] Batch [549]	Speed: 8562.467225 samples/sec	accuracy=0.553120	lr=2.209287
Epoch[94] Batch [599]	Speed: 8734.047600 samples/sec	accuracy=0.552931	lr=2.208629
[Epoch 94] training: accuracy=0.552780
[Epoch 94] speed: 8583 samples/sec	time cost: 154.314882
[Epoch 94] validation: err-top1=0.402168 err-top5=0.164609
Epoch[95] Batch [49]	Speed: 8741.900468 samples/sec	accuracy=0.558174	lr=2.207628
Epoch[95] Batch [99]	Speed: 8514.833421 samples/sec	accuracy=0.559092	lr=2.206969
Epoch[95] Batch [149]	Speed: 8478.024834 samples/sec	accuracy=0.558594	lr=2.206309
Epoch[95] Batch [199]	Speed: 8344.230708 samples/sec	accuracy=0.557700	lr=2.205649
Epoch[95] Batch [249]	Speed: 8672.269640 samples/sec	accuracy=0.557680	lr=2.204989
Epoch[95] Batch [299]	Speed: 8434.174782 samples/sec	accuracy=0.556868	lr=2.204328
Epoch[95] Batch [349]	Speed: 8743.827886 samples/sec	accuracy=0.556189	lr=2.203667
Epoch[95] Batch [399]	Speed: 8321.261205 samples/sec	accuracy=0.555693	lr=2.203005
Epoch[95] Batch [449]	Speed: 8751.330301 samples/sec	accuracy=0.555691	lr=2.202342
Epoch[95] Batch [499]	Speed: 8690.042315 samples/sec	accuracy=0.555237	lr=2.201680
Epoch[95] Batch [549]	Speed: 8844.385074 samples/sec	accuracy=0.555064	lr=2.201016
Epoch[95] Batch [599]	Speed: 8692.605740 samples/sec	accuracy=0.554631	lr=2.200353
[Epoch 95] training: accuracy=0.554287
[Epoch 95] speed: 8597 samples/sec	time cost: 153.769930
[Epoch 95] validation: err-top1=0.453288 err-top5=0.209717
Epoch[96] Batch [49]	Speed: 8621.887999 samples/sec	accuracy=0.561260	lr=2.199356
Epoch[96] Batch [99]	Speed: 8419.409555 samples/sec	accuracy=0.559526	lr=2.198692
Epoch[96] Batch [149]	Speed: 8649.502679 samples/sec	accuracy=0.558558	lr=2.198026
Epoch[96] Batch [199]	Speed: 8355.588412 samples/sec	accuracy=0.557542	lr=2.197361
Epoch[96] Batch [249]	Speed: 8459.943999 samples/sec	accuracy=0.556826	lr=2.196694
Epoch[96] Batch [299]	Speed: 8666.720697 samples/sec	accuracy=0.555951	lr=2.196028
Epoch[96] Batch [349]	Speed: 8501.306769 samples/sec	accuracy=0.555225	lr=2.195361
Epoch[96] Batch [399]	Speed: 8736.083325 samples/sec	accuracy=0.554670	lr=2.194693
Epoch[96] Batch [449]	Speed: 8733.047051 samples/sec	accuracy=0.554307	lr=2.194025
Epoch[96] Batch [499]	Speed: 8463.438016 samples/sec	accuracy=0.553919	lr=2.193357
Epoch[96] Batch [549]	Speed: 8793.105989 samples/sec	accuracy=0.554018	lr=2.192688
Epoch[96] Batch [599]	Speed: 8553.196049 samples/sec	accuracy=0.553558	lr=2.192019
[Epoch 96] training: accuracy=0.553390
[Epoch 96] speed: 8580 samples/sec	time cost: 154.430876
[Epoch 96] validation: err-top1=0.422207 err-top5=0.181758
Epoch[97] Batch [49]	Speed: 8611.045086 samples/sec	accuracy=0.566406	lr=2.191001
Epoch[97] Batch [99]	Speed: 8663.491446 samples/sec	accuracy=0.563433	lr=2.190330
Epoch[97] Batch [149]	Speed: 8508.139213 samples/sec	accuracy=0.561139	lr=2.189659
Epoch[97] Batch [199]	Speed: 8752.833934 samples/sec	accuracy=0.559326	lr=2.188988
Epoch[97] Batch [249]	Speed: 8573.523154 samples/sec	accuracy=0.558299	lr=2.188316
Epoch[97] Batch [299]	Speed: 8768.302022 samples/sec	accuracy=0.556841	lr=2.187644
Epoch[97] Batch [349]	Speed: 8674.049452 samples/sec	accuracy=0.556391	lr=2.186972
Epoch[97] Batch [399]	Speed: 8792.855946 samples/sec	accuracy=0.555764	lr=2.186299
Epoch[97] Batch [449]	Speed: 8597.048134 samples/sec	accuracy=0.555284	lr=2.185625
Epoch[97] Batch [499]	Speed: 8838.850408 samples/sec	accuracy=0.554796	lr=2.184951
Epoch[97] Batch [549]	Speed: 8692.572665 samples/sec	accuracy=0.554528	lr=2.184277
Epoch[97] Batch [599]	Speed: 8599.913760 samples/sec	accuracy=0.554431	lr=2.183602
[Epoch 97] training: accuracy=0.554192
[Epoch 97] speed: 8667 samples/sec	time cost: 152.220735
[Epoch 97] validation: err-top1=0.421875 err-top5=0.181132
Epoch[98] Batch [49]	Speed: 8432.262084 samples/sec	accuracy=0.566885	lr=2.182589
Epoch[98] Batch [99]	Speed: 8807.162732 samples/sec	accuracy=0.561914	lr=2.181913
Epoch[98] Batch [149]	Speed: 8428.007784 samples/sec	accuracy=0.559964	lr=2.181236
Epoch[98] Batch [199]	Speed: 8620.617784 samples/sec	accuracy=0.558860	lr=2.180559
Epoch[98] Batch [249]	Speed: 8543.930652 samples/sec	accuracy=0.558055	lr=2.179882
Epoch[98] Batch [299]	Speed: 8648.947398 samples/sec	accuracy=0.557788	lr=2.179204
Epoch[98] Batch [349]	Speed: 8764.502267 samples/sec	accuracy=0.557698	lr=2.178526
Epoch[98] Batch [399]	Speed: 8715.228022 samples/sec	accuracy=0.556902	lr=2.177848
Epoch[98] Batch [449]	Speed: 8631.943846 samples/sec	accuracy=0.556388	lr=2.177169
Epoch[98] Batch [499]	Speed: 8614.428168 samples/sec	accuracy=0.556188	lr=2.176489
Epoch[98] Batch [549]	Speed: 8636.351353 samples/sec	accuracy=0.556007	lr=2.175809
Epoch[98] Batch [599]	Speed: 8615.653005 samples/sec	accuracy=0.555739	lr=2.175129
[Epoch 98] training: accuracy=0.555512
[Epoch 98] speed: 8618 samples/sec	time cost: 153.181463
[Epoch 98] validation: err-top1=0.430929 err-top5=0.186727
Epoch[99] Batch [49]	Speed: 8719.668435 samples/sec	accuracy=0.562480	lr=2.174094
Epoch[99] Batch [99]	Speed: 8613.882564 samples/sec	accuracy=0.561558	lr=2.173412
Epoch[99] Batch [149]	Speed: 8447.972995 samples/sec	accuracy=0.560312	lr=2.172730
Epoch[99] Batch [199]	Speed: 8422.056890 samples/sec	accuracy=0.558992	lr=2.172048
Epoch[99] Batch [249]	Speed: 8669.340900 samples/sec	accuracy=0.558168	lr=2.171365
Epoch[99] Batch [299]	Speed: 8501.802358 samples/sec	accuracy=0.557510	lr=2.170682
Epoch[99] Batch [349]	Speed: 8360.835981 samples/sec	accuracy=0.556762	lr=2.169998
Epoch[99] Batch [399]	Speed: 8422.979019 samples/sec	accuracy=0.556261	lr=2.169314
Epoch[99] Batch [449]	Speed: 8449.082639 samples/sec	accuracy=0.555859	lr=2.168630
Epoch[99] Batch [499]	Speed: 8536.417062 samples/sec	accuracy=0.555294	lr=2.167945
Epoch[99] Batch [549]	Speed: 8806.983944 samples/sec	accuracy=0.555114	lr=2.167259
Epoch[99] Batch [599]	Speed: 8620.485766 samples/sec	accuracy=0.554867	lr=2.166574
[Epoch 99] training: accuracy=0.554609
[Epoch 99] speed: 8546 samples/sec	time cost: 154.738189
[Epoch 99] validation: err-top1=0.448770 err-top5=0.202187
Epoch[100] Batch [49]	Speed: 8393.706029 samples/sec	accuracy=0.564131	lr=2.165544
Epoch[100] Batch [99]	Speed: 8731.123140 samples/sec	accuracy=0.562759	lr=2.164857
Epoch[100] Batch [149]	Speed: 8463.829791 samples/sec	accuracy=0.561094	lr=2.164170
Epoch[100] Batch [199]	Speed: 8476.632364 samples/sec	accuracy=0.560291	lr=2.163482
Epoch[100] Batch [249]	Speed: 8520.393695 samples/sec	accuracy=0.559145	lr=2.162794
Epoch[100] Batch [299]	Speed: 8809.629665 samples/sec	accuracy=0.558254	lr=2.162105
Epoch[100] Batch [349]	Speed: 8518.420408 samples/sec	accuracy=0.557849	lr=2.161416
Epoch[100] Batch [399]	Speed: 8701.574576 samples/sec	accuracy=0.557212	lr=2.160726
Epoch[100] Batch [449]	Speed: 8800.458094 samples/sec	accuracy=0.556701	lr=2.160036
Epoch[100] Batch [499]	Speed: 8823.125928 samples/sec	accuracy=0.556337	lr=2.159346
Epoch[100] Batch [549]	Speed: 8644.018487 samples/sec	accuracy=0.555992	lr=2.158655
Epoch[100] Batch [599]	Speed: 8564.649349 samples/sec	accuracy=0.555616	lr=2.157964
[Epoch 100] training: accuracy=0.555475
[Epoch 100] speed: 8626 samples/sec	time cost: 153.273600
[Epoch 100] validation: err-top1=0.429484 err-top5=0.186910
Epoch[101] Batch [49]	Speed: 8590.697618 samples/sec	accuracy=0.565859	lr=2.156913
Epoch[101] Batch [99]	Speed: 8281.608965 samples/sec	accuracy=0.564346	lr=2.156220
Epoch[101] Batch [149]	Speed: 8776.755916 samples/sec	accuracy=0.562845	lr=2.155527
Epoch[101] Batch [199]	Speed: 8615.139216 samples/sec	accuracy=0.561721	lr=2.154834
Epoch[101] Batch [249]	Speed: 8406.658179 samples/sec	accuracy=0.560275	lr=2.154141
Epoch[101] Batch [299]	Speed: 8439.817251 samples/sec	accuracy=0.559741	lr=2.153447
Epoch[101] Batch [349]	Speed: 8616.515854 samples/sec	accuracy=0.558581	lr=2.152752
Epoch[101] Batch [399]	Speed: 8734.894712 samples/sec	accuracy=0.557993	lr=2.152057
Epoch[101] Batch [449]	Speed: 8688.420269 samples/sec	accuracy=0.557521	lr=2.151362
Epoch[101] Batch [499]	Speed: 8710.006385 samples/sec	accuracy=0.556850	lr=2.150666
Epoch[101] Batch [549]	Speed: 8560.885112 samples/sec	accuracy=0.556426	lr=2.149970
Epoch[101] Batch [599]	Speed: 8702.147038 samples/sec	accuracy=0.556235	lr=2.149273
[Epoch 101] training: accuracy=0.556169
[Epoch 101] speed: 8590 samples/sec	time cost: 153.952101
[Epoch 101] validation: err-top1=0.421621 err-top5=0.182207
Epoch[102] Batch [49]	Speed: 8562.339030 samples/sec	accuracy=0.565342	lr=2.148214
Epoch[102] Batch [99]	Speed: 8888.669071 samples/sec	accuracy=0.563481	lr=2.147516
Epoch[102] Batch [149]	Speed: 8689.740784 samples/sec	accuracy=0.561751	lr=2.146818
Epoch[102] Batch [199]	Speed: 8683.476298 samples/sec	accuracy=0.560256	lr=2.146119
Epoch[102] Batch [249]	Speed: 8763.250301 samples/sec	accuracy=0.559639	lr=2.145421
Epoch[102] Batch [299]	Speed: 8538.790312 samples/sec	accuracy=0.559206	lr=2.144721
Epoch[102] Batch [349]	Speed: 8363.964526 samples/sec	accuracy=0.558418	lr=2.144021
Epoch[102] Batch [399]	Speed: 8435.981473 samples/sec	accuracy=0.557590	lr=2.143321
Epoch[102] Batch [449]	Speed: 8598.835763 samples/sec	accuracy=0.557347	lr=2.142621
Epoch[102] Batch [499]	Speed: 8736.484222 samples/sec	accuracy=0.556991	lr=2.141919
Epoch[102] Batch [549]	Speed: 8605.353665 samples/sec	accuracy=0.556686	lr=2.141218
Epoch[102] Batch [599]	Speed: 8605.760241 samples/sec	accuracy=0.556258	lr=2.140516
[Epoch 102] training: accuracy=0.556210
[Epoch 102] speed: 8620 samples/sec	time cost: 153.071934
[Epoch 102] validation: err-top1=0.416829 err-top5=0.178202
Epoch[103] Batch [49]	Speed: 8389.855008 samples/sec	accuracy=0.568369	lr=2.139462
Epoch[103] Batch [99]	Speed: 8516.471509 samples/sec	accuracy=0.565669	lr=2.138759
Epoch[103] Batch [149]	Speed: 8670.873555 samples/sec	accuracy=0.563590	lr=2.138056
Epoch[103] Batch [199]	Speed: 8536.733837 samples/sec	accuracy=0.562273	lr=2.137352
Epoch[103] Batch [249]	Speed: 8636.474480 samples/sec	accuracy=0.560818	lr=2.136648
Epoch[103] Batch [299]	Speed: 8673.640251 samples/sec	accuracy=0.560280	lr=2.135943
Epoch[103] Batch [349]	Speed: 8851.287806 samples/sec	accuracy=0.559636	lr=2.135238
Epoch[103] Batch [399]	Speed: 8765.009343 samples/sec	accuracy=0.558922	lr=2.134533
Epoch[103] Batch [449]	Speed: 8872.068133 samples/sec	accuracy=0.558402	lr=2.133827
Epoch[103] Batch [499]	Speed: 8913.305205 samples/sec	accuracy=0.557957	lr=2.133121
Epoch[103] Batch [549]	Speed: 8747.452061 samples/sec	accuracy=0.557684	lr=2.132414
Epoch[103] Batch [599]	Speed: 8471.042995 samples/sec	accuracy=0.557355	lr=2.131707
[Epoch 103] training: accuracy=0.557326
[Epoch 103] speed: 8665 samples/sec	time cost: 152.849920
[Epoch 103] validation: err-top1=0.418496 err-top5=0.178242
Epoch[104] Batch [49]	Speed: 8417.262697 samples/sec	accuracy=0.567080	lr=2.130631
Epoch[104] Batch [99]	Speed: 8863.681964 samples/sec	accuracy=0.565732	lr=2.129923
Epoch[104] Batch [149]	Speed: 8569.831481 samples/sec	accuracy=0.564102	lr=2.129214
Epoch[104] Batch [199]	Speed: 8553.145972 samples/sec	accuracy=0.563284	lr=2.128505
Epoch[104] Batch [249]	Speed: 8423.222839 samples/sec	accuracy=0.562086	lr=2.127796
Epoch[104] Batch [299]	Speed: 8387.194790 samples/sec	accuracy=0.561330	lr=2.127086
Epoch[104] Batch [349]	Speed: 8322.993873 samples/sec	accuracy=0.560472	lr=2.126376
Epoch[104] Batch [399]	Speed: 8634.972710 samples/sec	accuracy=0.559972	lr=2.125665
Epoch[104] Batch [449]	Speed: 8543.034019 samples/sec	accuracy=0.559404	lr=2.124954
Epoch[104] Batch [499]	Speed: 8393.810851 samples/sec	accuracy=0.558936	lr=2.124243
Epoch[104] Batch [549]	Speed: 8435.601219 samples/sec	accuracy=0.558568	lr=2.123531
Epoch[104] Batch [599]	Speed: 8591.752951 samples/sec	accuracy=0.557961	lr=2.122818
[Epoch 104] training: accuracy=0.557881
[Epoch 104] speed: 8517 samples/sec	time cost: 155.047257
[Epoch 104] validation: err-top1=0.440877 err-top5=0.195719
Epoch[105] Batch [49]	Speed: 8313.253874 samples/sec	accuracy=0.567002	lr=2.121749
Epoch[105] Batch [99]	Speed: 8546.226115 samples/sec	accuracy=0.563936	lr=2.121036
Epoch[105] Batch [149]	Speed: 8660.592207 samples/sec	accuracy=0.562464	lr=2.120322
Epoch[105] Batch [199]	Speed: 8465.953580 samples/sec	accuracy=0.562107	lr=2.119608
Epoch[105] Batch [249]	Speed: 8491.948545 samples/sec	accuracy=0.561361	lr=2.118893
Epoch[105] Batch [299]	Speed: 8459.805025 samples/sec	accuracy=0.560562	lr=2.118178
Epoch[105] Batch [349]	Speed: 8541.995716 samples/sec	accuracy=0.559867	lr=2.117463
Epoch[105] Batch [399]	Speed: 8514.609081 samples/sec	accuracy=0.559309	lr=2.116747
Epoch[105] Batch [449]	Speed: 8721.483342 samples/sec	accuracy=0.558901	lr=2.116031
Epoch[105] Batch [499]	Speed: 8543.395980 samples/sec	accuracy=0.558497	lr=2.115314
Epoch[105] Batch [549]	Speed: 8667.975146 samples/sec	accuracy=0.558030	lr=2.114597
Epoch[105] Batch [599]	Speed: 8569.622017 samples/sec	accuracy=0.557574	lr=2.113879
[Epoch 105] training: accuracy=0.557541
[Epoch 105] speed: 8548 samples/sec	time cost: 154.668658
[Epoch 105] validation: err-top1=0.410136 err-top5=0.175334
Epoch[106] Batch [49]	Speed: 8378.537064 samples/sec	accuracy=0.568701	lr=2.112788
Epoch[106] Batch [99]	Speed: 8730.861701 samples/sec	accuracy=0.564893	lr=2.112070
Epoch[106] Batch [149]	Speed: 8813.120524 samples/sec	accuracy=0.563320	lr=2.111351
Epoch[106] Batch [199]	Speed: 8786.963982 samples/sec	accuracy=0.562642	lr=2.110631
Epoch[106] Batch [249]	Speed: 8636.870108 samples/sec	accuracy=0.561805	lr=2.109912
Epoch[106] Batch [299]	Speed: 8289.854141 samples/sec	accuracy=0.561069	lr=2.109191
Epoch[106] Batch [349]	Speed: 8421.522171 samples/sec	accuracy=0.560767	lr=2.108471
Epoch[106] Batch [399]	Speed: 8610.573102 samples/sec	accuracy=0.560477	lr=2.107750
Epoch[106] Batch [449]	Speed: 8432.180966 samples/sec	accuracy=0.559927	lr=2.107029
Epoch[106] Batch [499]	Speed: 8530.808920 samples/sec	accuracy=0.559610	lr=2.106307
Epoch[106] Batch [549]	Speed: 8522.588417 samples/sec	accuracy=0.559574	lr=2.105585
Epoch[106] Batch [599]	Speed: 8604.702154 samples/sec	accuracy=0.559189	lr=2.104862
[Epoch 106] training: accuracy=0.559059
[Epoch 106] speed: 8566 samples/sec	time cost: 154.364264
[Epoch 106] validation: err-top1=0.435391 err-top5=0.193594
Epoch[107] Batch [49]	Speed: 8231.618495 samples/sec	accuracy=0.570479	lr=2.103778
Epoch[107] Batch [99]	Speed: 8442.210264 samples/sec	accuracy=0.567881	lr=2.103054
Epoch[107] Batch [149]	Speed: 8440.655688 samples/sec	accuracy=0.565290	lr=2.102330
Epoch[107] Batch [199]	Speed: 8775.436616 samples/sec	accuracy=0.564209	lr=2.101606
Epoch[107] Batch [249]	Speed: 8602.469939 samples/sec	accuracy=0.562479	lr=2.100881
Epoch[107] Batch [299]	Speed: 8403.977266 samples/sec	accuracy=0.561517	lr=2.100156
Epoch[107] Batch [349]	Speed: 8570.829356 samples/sec	accuracy=0.560739	lr=2.099430
Epoch[107] Batch [399]	Speed: 8771.721682 samples/sec	accuracy=0.560194	lr=2.098704
Epoch[107] Batch [449]	Speed: 8673.293617 samples/sec	accuracy=0.559448	lr=2.097978
Epoch[107] Batch [499]	Speed: 8685.605846 samples/sec	accuracy=0.559136	lr=2.097251
Epoch[107] Batch [549]	Speed: 8622.101411 samples/sec	accuracy=0.558793	lr=2.096524
Epoch[107] Batch [599]	Speed: 8761.117185 samples/sec	accuracy=0.558756	lr=2.095796
[Epoch 107] training: accuracy=0.558610
[Epoch 107] speed: 8575 samples/sec	time cost: 154.242849
[Epoch 107] validation: err-top1=0.434631 err-top5=0.188802
Epoch[108] Batch [49]	Speed: 8277.100423 samples/sec	accuracy=0.569805	lr=2.094690
Epoch[108] Batch [99]	Speed: 8722.180998 samples/sec	accuracy=0.566968	lr=2.093961
Epoch[108] Batch [149]	Speed: 8729.454495 samples/sec	accuracy=0.565719	lr=2.093232
Epoch[108] Batch [199]	Speed: 8489.157259 samples/sec	accuracy=0.563616	lr=2.092503
Epoch[108] Batch [249]	Speed: 8704.722902 samples/sec	accuracy=0.562875	lr=2.091773
Epoch[108] Batch [299]	Speed: 8554.094135 samples/sec	accuracy=0.562251	lr=2.091043
Epoch[108] Batch [349]	Speed: 8787.335583 samples/sec	accuracy=0.561589	lr=2.090312
Epoch[108] Batch [399]	Speed: 8883.299702 samples/sec	accuracy=0.560913	lr=2.089581
Epoch[108] Batch [449]	Speed: 8585.628384 samples/sec	accuracy=0.560622	lr=2.088850
Epoch[108] Batch [499]	Speed: 8995.017656 samples/sec	accuracy=0.559981	lr=2.088118
Epoch[108] Batch [549]	Speed: 8575.101895 samples/sec	accuracy=0.559398	lr=2.087386
Epoch[108] Batch [599]	Speed: 8541.860828 samples/sec	accuracy=0.559132	lr=2.086654
[Epoch 108] training: accuracy=0.559070
[Epoch 108] speed: 8652 samples/sec	time cost: 153.119542
[Epoch 108] validation: err-top1=0.407168 err-top5=0.172109
Epoch[109] Batch [49]	Speed: 8818.225940 samples/sec	accuracy=0.570000	lr=2.085540
Epoch[109] Batch [99]	Speed: 8424.675484 samples/sec	accuracy=0.567783	lr=2.084806
Epoch[109] Batch [149]	Speed: 8905.977267 samples/sec	accuracy=0.565960	lr=2.084072
Epoch[109] Batch [199]	Speed: 8622.060043 samples/sec	accuracy=0.564849	lr=2.083338
Epoch[109] Batch [249]	Speed: 8547.251158 samples/sec	accuracy=0.564203	lr=2.082603
Epoch[109] Batch [299]	Speed: 8535.646514 samples/sec	accuracy=0.563286	lr=2.081868
Epoch[109] Batch [349]	Speed: 8711.933901 samples/sec	accuracy=0.562907	lr=2.081132
Epoch[109] Batch [399]	Speed: 8700.155117 samples/sec	accuracy=0.562573	lr=2.080397
Epoch[109] Batch [449]	Speed: 8638.218606 samples/sec	accuracy=0.561708	lr=2.079660
Epoch[109] Batch [499]	Speed: 8569.918348 samples/sec	accuracy=0.561250	lr=2.078924
Epoch[109] Batch [549]	Speed: 8581.877411 samples/sec	accuracy=0.561110	lr=2.078187
Epoch[109] Batch [599]	Speed: 8705.033591 samples/sec	accuracy=0.560727	lr=2.077449
[Epoch 109] training: accuracy=0.560521
[Epoch 109] speed: 8650 samples/sec	time cost: 152.567678
[Epoch 109] validation: err-top1=0.409851 err-top5=0.173848
Epoch[110] Batch [49]	Speed: 8477.612667 samples/sec	accuracy=0.568223	lr=2.076342
Epoch[110] Batch [99]	Speed: 8574.225239 samples/sec	accuracy=0.566099	lr=2.075604
Epoch[110] Batch [149]	Speed: 9054.726373 samples/sec	accuracy=0.564980	lr=2.074865
Epoch[110] Batch [199]	Speed: 8377.209432 samples/sec	accuracy=0.563611	lr=2.074126
Epoch[110] Batch [249]	Speed: 8655.629435 samples/sec	accuracy=0.563326	lr=2.073386
Epoch[110] Batch [299]	Speed: 8543.194264 samples/sec	accuracy=0.562720	lr=2.072646
Epoch[110] Batch [349]	Speed: 8498.602491 samples/sec	accuracy=0.562323	lr=2.071906
Epoch[110] Batch [399]	Speed: 8571.638601 samples/sec	accuracy=0.561975	lr=2.071165
Epoch[110] Batch [449]	Speed: 8619.948736 samples/sec	accuracy=0.561119	lr=2.070424
Epoch[110] Batch [499]	Speed: 8716.533879 samples/sec	accuracy=0.560903	lr=2.069683
Epoch[110] Batch [549]	Speed: 8688.951098 samples/sec	accuracy=0.560453	lr=2.068941
Epoch[110] Batch [599]	Speed: 8715.276478 samples/sec	accuracy=0.560034	lr=2.068198
[Epoch 110] training: accuracy=0.559800
[Epoch 110] speed: 8628 samples/sec	time cost: 153.279355
[Epoch 110] validation: err-top1=0.418905 err-top5=0.178955
Epoch[111] Batch [49]	Speed: 8275.197390 samples/sec	accuracy=0.568867	lr=2.067069
Epoch[111] Batch [99]	Speed: 8764.334864 samples/sec	accuracy=0.566514	lr=2.066326
Epoch[111] Batch [149]	Speed: 8605.070912 samples/sec	accuracy=0.565283	lr=2.065582
Epoch[111] Batch [199]	Speed: 8874.461727 samples/sec	accuracy=0.564204	lr=2.064838
Epoch[111] Batch [249]	Speed: 8633.130113 samples/sec	accuracy=0.563176	lr=2.064094
Epoch[111] Batch [299]	Speed: 8485.840149 samples/sec	accuracy=0.562319	lr=2.063349
Epoch[111] Batch [349]	Speed: 8536.141705 samples/sec	accuracy=0.561772	lr=2.062604
Epoch[111] Batch [399]	Speed: 8836.460712 samples/sec	accuracy=0.561360	lr=2.061859
Epoch[111] Batch [449]	Speed: 9076.913353 samples/sec	accuracy=0.561069	lr=2.061113
Epoch[111] Batch [499]	Speed: 8695.165573 samples/sec	accuracy=0.560772	lr=2.060366
Epoch[111] Batch [549]	Speed: 8725.165584 samples/sec	accuracy=0.560566	lr=2.059620
Epoch[111] Batch [599]	Speed: 8642.787137 samples/sec	accuracy=0.560484	lr=2.058872
[Epoch 111] training: accuracy=0.560319
[Epoch 111] speed: 8675 samples/sec	time cost: 152.485039
[Epoch 111] validation: err-top1=0.428926 err-top5=0.185391
Epoch[112] Batch [49]	Speed: 8779.492617 samples/sec	accuracy=0.570176	lr=2.057751
Epoch[112] Batch [99]	Speed: 8517.169855 samples/sec	accuracy=0.569473	lr=2.057003
Epoch[112] Batch [149]	Speed: 8449.293233 samples/sec	accuracy=0.567331	lr=2.056255
Epoch[112] Batch [199]	Speed: 8542.361837 samples/sec	accuracy=0.566274	lr=2.055506
Epoch[112] Batch [249]	Speed: 8488.849541 samples/sec	accuracy=0.565229	lr=2.054757
Epoch[112] Batch [299]	Speed: 8703.280195 samples/sec	accuracy=0.564676	lr=2.054007
Epoch[112] Batch [349]	Speed: 8529.786970 samples/sec	accuracy=0.563754	lr=2.053257
Epoch[112] Batch [399]	Speed: 8666.784530 samples/sec	accuracy=0.562991	lr=2.052507
Epoch[112] Batch [449]	Speed: 8486.269213 samples/sec	accuracy=0.562330	lr=2.051756
Epoch[112] Batch [499]	Speed: 8719.188009 samples/sec	accuracy=0.561733	lr=2.051005
Epoch[112] Batch [549]	Speed: 8726.666977 samples/sec	accuracy=0.561385	lr=2.050254
Epoch[112] Batch [599]	Speed: 8698.326879 samples/sec	accuracy=0.561029	lr=2.049502
[Epoch 112] training: accuracy=0.560864
[Epoch 112] speed: 8611 samples/sec	time cost: 153.343076
[Epoch 112] validation: err-top1=0.463318 err-top5=0.216349
Epoch[113] Batch [49]	Speed: 8311.724709 samples/sec	accuracy=0.571602	lr=2.048359
Epoch[113] Batch [99]	Speed: 8797.343940 samples/sec	accuracy=0.569976	lr=2.047606
Epoch[113] Batch [149]	Speed: 8726.526194 samples/sec	accuracy=0.569238	lr=2.046853
Epoch[113] Batch [199]	Speed: 8527.756841 samples/sec	accuracy=0.567434	lr=2.046099
Epoch[113] Batch [249]	Speed: 8412.825271 samples/sec	accuracy=0.566115	lr=2.045345
Epoch[113] Batch [299]	Speed: 8747.157577 samples/sec	accuracy=0.565369	lr=2.044591
Epoch[113] Batch [349]	Speed: 8615.587849 samples/sec	accuracy=0.564276	lr=2.043836
Epoch[113] Batch [399]	Speed: 8579.174591 samples/sec	accuracy=0.563329	lr=2.043081
Epoch[113] Batch [449]	Speed: 8549.049277 samples/sec	accuracy=0.562655	lr=2.042326
Epoch[113] Batch [499]	Speed: 8911.774783 samples/sec	accuracy=0.562311	lr=2.041570
Epoch[113] Batch [549]	Speed: 8632.518807 samples/sec	accuracy=0.561774	lr=2.040814
Epoch[113] Batch [599]	Speed: 8542.201624 samples/sec	accuracy=0.561467	lr=2.040058
[Epoch 113] training: accuracy=0.561205
[Epoch 113] speed: 8614 samples/sec	time cost: 153.356188
[Epoch 113] validation: err-top1=0.460508 err-top5=0.211504
Epoch[114] Batch [49]	Speed: 8292.967890 samples/sec	accuracy=0.571953	lr=2.038922
Epoch[114] Batch [99]	Speed: 8719.057203 samples/sec	accuracy=0.570059	lr=2.038165
Epoch[114] Batch [149]	Speed: 8694.840803 samples/sec	accuracy=0.568066	lr=2.037407
Epoch[114] Batch [199]	Speed: 8719.611787 samples/sec	accuracy=0.567124	lr=2.036649
Epoch[114] Batch [249]	Speed: 8749.168764 samples/sec	accuracy=0.566199	lr=2.035890
Epoch[114] Batch [299]	Speed: 8599.141865 samples/sec	accuracy=0.565080	lr=2.035132
Epoch[114] Batch [349]	Speed: 8664.555826 samples/sec	accuracy=0.564622	lr=2.034372
Epoch[114] Batch [399]	Speed: 8508.074999 samples/sec	accuracy=0.564368	lr=2.033613
Epoch[114] Batch [449]	Speed: 8451.928628 samples/sec	accuracy=0.563950	lr=2.032853
Epoch[114] Batch [499]	Speed: 8544.608690 samples/sec	accuracy=0.563788	lr=2.032092
Epoch[114] Batch [549]	Speed: 8721.521950 samples/sec	accuracy=0.563439	lr=2.031332
Epoch[114] Batch [599]	Speed: 8650.238345 samples/sec	accuracy=0.562857	lr=2.030570
[Epoch 114] training: accuracy=0.562778
[Epoch 114] speed: 8610 samples/sec	time cost: 153.360672
[Epoch 114] validation: err-top1=0.410767 err-top5=0.170471
Epoch[115] Batch [49]	Speed: 8411.772741 samples/sec	accuracy=0.572617	lr=2.029413
Epoch[115] Batch [99]	Speed: 8512.506198 samples/sec	accuracy=0.570176	lr=2.028651
Epoch[115] Batch [149]	Speed: 8784.881119 samples/sec	accuracy=0.568558	lr=2.027888
Epoch[115] Batch [199]	Speed: 8732.197457 samples/sec	accuracy=0.567090	lr=2.027126
Epoch[115] Batch [249]	Speed: 8612.628351 samples/sec	accuracy=0.566916	lr=2.026363
Epoch[115] Batch [299]	Speed: 8585.894070 samples/sec	accuracy=0.566047	lr=2.025599
Epoch[115] Batch [349]	Speed: 8773.078931 samples/sec	accuracy=0.565017	lr=2.024835
Epoch[115] Batch [399]	Speed: 8661.903402 samples/sec	accuracy=0.564104	lr=2.024071
Epoch[115] Batch [449]	Speed: 8607.198049 samples/sec	accuracy=0.563206	lr=2.023307
Epoch[115] Batch [499]	Speed: 8649.948279 samples/sec	accuracy=0.563083	lr=2.022542
Epoch[115] Batch [549]	Speed: 8623.144044 samples/sec	accuracy=0.562718	lr=2.021776
Epoch[115] Batch [599]	Speed: 8499.815132 samples/sec	accuracy=0.562393	lr=2.021011
[Epoch 115] training: accuracy=0.562255
[Epoch 115] speed: 8627 samples/sec	time cost: 153.374283
[Epoch 115] validation: err-top1=0.445684 err-top5=0.203223
Epoch[116] Batch [49]	Speed: 8732.330789 samples/sec	accuracy=0.571025	lr=2.019846
Epoch[116] Batch [99]	Speed: 8567.522809 samples/sec	accuracy=0.568164	lr=2.019080
Epoch[116] Batch [149]	Speed: 8829.585071 samples/sec	accuracy=0.567988	lr=2.018313
Epoch[116] Batch [199]	Speed: 8650.002981 samples/sec	accuracy=0.566414	lr=2.017545
Epoch[116] Batch [249]	Speed: 8579.419141 samples/sec	accuracy=0.565859	lr=2.016778
Epoch[116] Batch [299]	Speed: 8509.743703 samples/sec	accuracy=0.564790	lr=2.016010
Epoch[116] Batch [349]	Speed: 8677.287456 samples/sec	accuracy=0.564728	lr=2.015241
Epoch[116] Batch [399]	Speed: 8633.491767 samples/sec	accuracy=0.564625	lr=2.014473
Epoch[116] Batch [449]	Speed: 8662.722599 samples/sec	accuracy=0.564338	lr=2.013704
Epoch[116] Batch [499]	Speed: 8672.782386 samples/sec	accuracy=0.563940	lr=2.012934
Epoch[116] Batch [549]	Speed: 8728.818828 samples/sec	accuracy=0.563782	lr=2.012164
Epoch[116] Batch [599]	Speed: 8468.082442 samples/sec	accuracy=0.563559	lr=2.011394
[Epoch 116] training: accuracy=0.563509
[Epoch 116] speed: 8633 samples/sec	time cost: 152.907007
[Epoch 116] validation: err-top1=0.401571 err-top5=0.165812
Epoch[117] Batch [49]	Speed: 8476.263324 samples/sec	accuracy=0.572295	lr=2.010238
Epoch[117] Batch [99]	Speed: 8664.511253 samples/sec	accuracy=0.571304	lr=2.009467
Epoch[117] Batch [149]	Speed: 8613.045633 samples/sec	accuracy=0.570270	lr=2.008696
Epoch[117] Batch [199]	Speed: 8795.839207 samples/sec	accuracy=0.569734	lr=2.007924
Epoch[117] Batch [249]	Speed: 8579.097819 samples/sec	accuracy=0.568025	lr=2.007152
Epoch[117] Batch [299]	Speed: 8873.907989 samples/sec	accuracy=0.566984	lr=2.006380
Epoch[117] Batch [349]	Speed: 8591.198529 samples/sec	accuracy=0.566085	lr=2.005607
Epoch[117] Batch [399]	Speed: 8490.043290 samples/sec	accuracy=0.565374	lr=2.004834
Epoch[117] Batch [449]	Speed: 8950.427748 samples/sec	accuracy=0.564623	lr=2.004060
Epoch[117] Batch [499]	Speed: 8641.022738 samples/sec	accuracy=0.564104	lr=2.003286
Epoch[117] Batch [549]	Speed: 8416.072175 samples/sec	accuracy=0.563939	lr=2.002512
Epoch[117] Batch [599]	Speed: 8408.597305 samples/sec	accuracy=0.563637	lr=2.001737
[Epoch 117] training: accuracy=0.563536
[Epoch 117] speed: 8626 samples/sec	time cost: 153.270996
[Epoch 117] validation: err-top1=0.409424 err-top5=0.172078
Epoch[118] Batch [49]	Speed: 8715.357122 samples/sec	accuracy=0.572920	lr=2.000559
Epoch[118] Batch [99]	Speed: 8508.668132 samples/sec	accuracy=0.571309	lr=1.999784
Epoch[118] Batch [149]	Speed: 8598.925973 samples/sec	accuracy=0.570267	lr=1.999008
Epoch[118] Batch [199]	Speed: 8587.033034 samples/sec	accuracy=0.569385	lr=1.998232
Epoch[118] Batch [249]	Speed: 8442.351482 samples/sec	accuracy=0.567873	lr=1.997456
Epoch[118] Batch [299]	Speed: 8735.695614 samples/sec	accuracy=0.567531	lr=1.996679
Epoch[118] Batch [349]	Speed: 8699.395431 samples/sec	accuracy=0.567263	lr=1.995901
Epoch[118] Batch [399]	Speed: 8509.766971 samples/sec	accuracy=0.566522	lr=1.995124
Epoch[118] Batch [449]	Speed: 8429.503605 samples/sec	accuracy=0.565896	lr=1.994346
Epoch[118] Batch [499]	Speed: 8771.274554 samples/sec	accuracy=0.565169	lr=1.993568
Epoch[118] Batch [549]	Speed: 8789.587074 samples/sec	accuracy=0.564994	lr=1.992789
Epoch[118] Batch [599]	Speed: 8677.907224 samples/sec	accuracy=0.564684	lr=1.992010
[Epoch 118] training: accuracy=0.564605
[Epoch 118] speed: 8626 samples/sec	time cost: 153.130990
[Epoch 118] validation: err-top1=0.406348 err-top5=0.169648
Epoch[119] Batch [49]	Speed: 8590.216351 samples/sec	accuracy=0.572959	lr=1.990841
Epoch[119] Batch [99]	Speed: 8634.334238 samples/sec	accuracy=0.569707	lr=1.990061
Epoch[119] Batch [149]	Speed: 8307.521720 samples/sec	accuracy=0.569769	lr=1.989281
Epoch[119] Batch [199]	Speed: 8665.549309 samples/sec	accuracy=0.569077	lr=1.988501
Epoch[119] Batch [249]	Speed: 8780.211611 samples/sec	accuracy=0.568654	lr=1.987720
Epoch[119] Batch [299]	Speed: 8627.961123 samples/sec	accuracy=0.567842	lr=1.986939
Epoch[119] Batch [349]	Speed: 8576.419013 samples/sec	accuracy=0.567550	lr=1.986157
Epoch[119] Batch [399]	Speed: 8702.666144 samples/sec	accuracy=0.566941	lr=1.985375
Epoch[119] Batch [449]	Speed: 8737.510089 samples/sec	accuracy=0.566083	lr=1.984593
Epoch[119] Batch [499]	Speed: 8696.120308 samples/sec	accuracy=0.565501	lr=1.983810
Epoch[119] Batch [549]	Speed: 8742.575946 samples/sec	accuracy=0.565056	lr=1.983027
Epoch[119] Batch [599]	Speed: 8351.543946 samples/sec	accuracy=0.564657	lr=1.982244
[Epoch 119] training: accuracy=0.564606
[Epoch 119] speed: 8618 samples/sec	time cost: 153.368637
[Epoch 119] validation: err-top1=0.406372 err-top5=0.172017
Epoch[120] Batch [49]	Speed: 8471.010081 samples/sec	accuracy=0.574063	lr=1.981053
Epoch[120] Batch [99]	Speed: 8594.573943 samples/sec	accuracy=0.571250	lr=1.980269
Epoch[120] Batch [149]	Speed: 8713.596560 samples/sec	accuracy=0.569919	lr=1.979484
Epoch[120] Batch [199]	Speed: 8613.021624 samples/sec	accuracy=0.568274	lr=1.978700
Epoch[120] Batch [249]	Speed: 8700.116698 samples/sec	accuracy=0.567422	lr=1.977914
Epoch[120] Batch [299]	Speed: 8564.941579 samples/sec	accuracy=0.567062	lr=1.977129
Epoch[120] Batch [349]	Speed: 8573.813764 samples/sec	accuracy=0.566610	lr=1.976343
Epoch[120] Batch [399]	Speed: 8618.728385 samples/sec	accuracy=0.565898	lr=1.975557
Epoch[120] Batch [449]	Speed: 8728.750353 samples/sec	accuracy=0.565741	lr=1.974770
Epoch[120] Batch [499]	Speed: 8632.698564 samples/sec	accuracy=0.565521	lr=1.973984
Epoch[120] Batch [549]	Speed: 8760.884147 samples/sec	accuracy=0.565370	lr=1.973196
Epoch[120] Batch [599]	Speed: 8558.388031 samples/sec	accuracy=0.565271	lr=1.972409
[Epoch 120] training: accuracy=0.565143
[Epoch 120] speed: 8630 samples/sec	time cost: 153.042906
[Epoch 120] validation: err-top1=0.413047 err-top5=0.173457
Epoch[121] Batch [49]	Speed: 8495.264553 samples/sec	accuracy=0.574150	lr=1.971227
Epoch[121] Batch [99]	Speed: 8779.747105 samples/sec	accuracy=0.573271	lr=1.970439
Epoch[121] Batch [149]	Speed: 8602.813175 samples/sec	accuracy=0.571113	lr=1.969650
Epoch[121] Batch [199]	Speed: 8774.085448 samples/sec	accuracy=0.570388	lr=1.968861
Epoch[121] Batch [249]	Speed: 8337.834658 samples/sec	accuracy=0.569541	lr=1.968072
Epoch[121] Batch [299]	Speed: 8450.154170 samples/sec	accuracy=0.568016	lr=1.967282
Epoch[121] Batch [349]	Speed: 8795.077848 samples/sec	accuracy=0.567319	lr=1.966492
Epoch[121] Batch [399]	Speed: 8849.780794 samples/sec	accuracy=0.566914	lr=1.965701
Epoch[121] Batch [449]	Speed: 8617.237274 samples/sec	accuracy=0.566301	lr=1.964911
Epoch[121] Batch [499]	Speed: 8731.992763 samples/sec	accuracy=0.566000	lr=1.964120
Epoch[121] Batch [549]	Speed: 8815.805762 samples/sec	accuracy=0.565692	lr=1.963328
Epoch[121] Batch [599]	Speed: 8838.294194 samples/sec	accuracy=0.565382	lr=1.962537
[Epoch 121] training: accuracy=0.565098
[Epoch 121] speed: 8671 samples/sec	time cost: 152.301880
[Epoch 121] validation: err-top1=0.415873 err-top5=0.177877
Epoch[122] Batch [49]	Speed: 8784.297900 samples/sec	accuracy=0.577598	lr=1.961333
Epoch[122] Batch [99]	Speed: 8613.218706 samples/sec	accuracy=0.574854	lr=1.960540
Epoch[122] Batch [149]	Speed: 8596.462230 samples/sec	accuracy=0.573350	lr=1.959747
Epoch[122] Batch [199]	Speed: 8626.434936 samples/sec	accuracy=0.572297	lr=1.958954
Epoch[122] Batch [249]	Speed: 8372.930995 samples/sec	accuracy=0.570916	lr=1.958161
Epoch[122] Batch [299]	Speed: 8572.132159 samples/sec	accuracy=0.570355	lr=1.957367
Epoch[122] Batch [349]	Speed: 8512.175697 samples/sec	accuracy=0.569930	lr=1.956573
Epoch[122] Batch [399]	Speed: 8838.741997 samples/sec	accuracy=0.569080	lr=1.955778
Epoch[122] Batch [449]	Speed: 8797.934299 samples/sec	accuracy=0.568669	lr=1.954983
Epoch[122] Batch [499]	Speed: 8786.673663 samples/sec	accuracy=0.568000	lr=1.954188
Epoch[122] Batch [549]	Speed: 8375.206688 samples/sec	accuracy=0.567556	lr=1.953392
Epoch[122] Batch [599]	Speed: 8737.591323 samples/sec	accuracy=0.566877	lr=1.952597
[Epoch 122] training: accuracy=0.566508
[Epoch 122] speed: 8641 samples/sec	time cost: 153.054009
[Epoch 122] validation: err-top1=0.458171 err-top5=0.210795
Epoch[123] Batch [49]	Speed: 8594.293274 samples/sec	accuracy=0.573564	lr=1.951386
Epoch[123] Batch [99]	Speed: 8493.436078 samples/sec	accuracy=0.572993	lr=1.950590
Epoch[123] Batch [149]	Speed: 8464.221436 samples/sec	accuracy=0.572122	lr=1.949793
Epoch[123] Batch [199]	Speed: 8739.121534 samples/sec	accuracy=0.571667	lr=1.948995
Epoch[123] Batch [249]	Speed: 8856.694328 samples/sec	accuracy=0.570551	lr=1.948198
Epoch[123] Batch [299]	Speed: 8800.612995 samples/sec	accuracy=0.570072	lr=1.947400
Epoch[123] Batch [349]	Speed: 8608.713467 samples/sec	accuracy=0.569487	lr=1.946602
Epoch[123] Batch [399]	Speed: 8545.743187 samples/sec	accuracy=0.568796	lr=1.945803
Epoch[123] Batch [449]	Speed: 8692.196370 samples/sec	accuracy=0.568269	lr=1.945004
Epoch[123] Batch [499]	Speed: 8533.926433 samples/sec	accuracy=0.567987	lr=1.944205
Epoch[123] Batch [549]	Speed: 8628.161836 samples/sec	accuracy=0.567385	lr=1.943405
Epoch[123] Batch [599]	Speed: 8617.674542 samples/sec	accuracy=0.567008	lr=1.942605
[Epoch 123] training: accuracy=0.566784
[Epoch 123] speed: 8632 samples/sec	time cost: 153.135974
[Epoch 123] validation: err-top1=0.440234 err-top5=0.199082
Epoch[124] Batch [49]	Speed: 8798.299258 samples/sec	accuracy=0.578154	lr=1.941405
Epoch[124] Batch [99]	Speed: 8599.193171 samples/sec	accuracy=0.573442	lr=1.940604
Epoch[124] Batch [149]	Speed: 8695.015066 samples/sec	accuracy=0.572943	lr=1.939803
Epoch[124] Batch [199]	Speed: 8778.130512 samples/sec	accuracy=0.571853	lr=1.939002
Epoch[124] Batch [249]	Speed: 8521.264280 samples/sec	accuracy=0.570705	lr=1.938200
Epoch[124] Batch [299]	Speed: 8505.908802 samples/sec	accuracy=0.569744	lr=1.937398
Epoch[124] Batch [349]	Speed: 8718.674186 samples/sec	accuracy=0.569715	lr=1.936596
Epoch[124] Batch [399]	Speed: 8442.471297 samples/sec	accuracy=0.568870	lr=1.935793
Epoch[124] Batch [449]	Speed: 8883.204161 samples/sec	accuracy=0.568298	lr=1.934990
Epoch[124] Batch [499]	Speed: 8546.603993 samples/sec	accuracy=0.567891	lr=1.934187
Epoch[124] Batch [549]	Speed: 8895.056005 samples/sec	accuracy=0.567536	lr=1.933384
Epoch[124] Batch [599]	Speed: 8620.959010 samples/sec	accuracy=0.567246	lr=1.932580
[Epoch 124] training: accuracy=0.567156
[Epoch 124] speed: 8659 samples/sec	time cost: 152.736092
[Epoch 124] validation: err-top1=0.407613 err-top5=0.172729
Epoch[125] Batch [49]	Speed: 8463.515735 samples/sec	accuracy=0.576484	lr=1.931357
Epoch[125] Batch [99]	Speed: 8951.229860 samples/sec	accuracy=0.574224	lr=1.930552
Epoch[125] Batch [149]	Speed: 8430.650104 samples/sec	accuracy=0.573330	lr=1.929747
Epoch[125] Batch [199]	Speed: 8570.979870 samples/sec	accuracy=0.572031	lr=1.928942
Epoch[125] Batch [249]	Speed: 8674.945939 samples/sec	accuracy=0.570652	lr=1.928137
Epoch[125] Batch [299]	Speed: 8532.261792 samples/sec	accuracy=0.569474	lr=1.927331
Epoch[125] Batch [349]	Speed: 8881.369080 samples/sec	accuracy=0.568898	lr=1.926524
Epoch[125] Batch [399]	Speed: 8556.101712 samples/sec	accuracy=0.568484	lr=1.925718
Epoch[125] Batch [449]	Speed: 8777.080198 samples/sec	accuracy=0.568006	lr=1.924911
Epoch[125] Batch [499]	Speed: 8612.910219 samples/sec	accuracy=0.567688	lr=1.924104
Epoch[125] Batch [549]	Speed: 8616.963767 samples/sec	accuracy=0.567343	lr=1.923296
Epoch[125] Batch [599]	Speed: 8725.923928 samples/sec	accuracy=0.567268	lr=1.922488
[Epoch 125] training: accuracy=0.567226
[Epoch 125] speed: 8646 samples/sec	time cost: 152.918663
[Epoch 125] validation: err-top1=0.454160 err-top5=0.208223
Epoch[126] Batch [49]	Speed: 8700.861880 samples/sec	accuracy=0.576172	lr=1.921276
Epoch[126] Batch [99]	Speed: 8588.370478 samples/sec	accuracy=0.575288	lr=1.920467
Epoch[126] Batch [149]	Speed: 8861.264381 samples/sec	accuracy=0.574082	lr=1.919658
Epoch[126] Batch [199]	Speed: 8575.262832 samples/sec	accuracy=0.572983	lr=1.918849
Epoch[126] Batch [249]	Speed: 8738.918471 samples/sec	accuracy=0.572385	lr=1.918040
Epoch[126] Batch [299]	Speed: 8464.324357 samples/sec	accuracy=0.571750	lr=1.917230
Epoch[126] Batch [349]	Speed: 8690.710507 samples/sec	accuracy=0.571006	lr=1.916420
Epoch[126] Batch [399]	Speed: 8773.913199 samples/sec	accuracy=0.569989	lr=1.915609
Epoch[126] Batch [449]	Speed: 8456.334819 samples/sec	accuracy=0.569488	lr=1.914798
Epoch[126] Batch [499]	Speed: 8559.592719 samples/sec	accuracy=0.569037	lr=1.913987
Epoch[126] Batch [549]	Speed: 8394.765200 samples/sec	accuracy=0.568599	lr=1.913176
Epoch[126] Batch [599]	Speed: 8751.308903 samples/sec	accuracy=0.568385	lr=1.912364
[Epoch 126] training: accuracy=0.568136
[Epoch 126] speed: 8620 samples/sec	time cost: 153.224882
[Epoch 126] validation: err-top1=0.398376 err-top5=0.163188
Epoch[127] Batch [49]	Speed: 8784.676822 samples/sec	accuracy=0.578672	lr=1.911130
Epoch[127] Batch [99]	Speed: 8670.509463 samples/sec	accuracy=0.575503	lr=1.910317
Epoch[127] Batch [149]	Speed: 8531.608422 samples/sec	accuracy=0.574235	lr=1.909505
Epoch[127] Batch [199]	Speed: 8378.591165 samples/sec	accuracy=0.573669	lr=1.908692
Epoch[127] Batch [249]	Speed: 8759.254306 samples/sec	accuracy=0.572107	lr=1.907878
Epoch[127] Batch [299]	Speed: 8570.025224 samples/sec	accuracy=0.571437	lr=1.907064
Epoch[127] Batch [349]	Speed: 8624.672184 samples/sec	accuracy=0.570953	lr=1.906250
Epoch[127] Batch [399]	Speed: 8862.795973 samples/sec	accuracy=0.570325	lr=1.905436
Epoch[127] Batch [449]	Speed: 8792.685298 samples/sec	accuracy=0.569761	lr=1.904622
Epoch[127] Batch [499]	Speed: 8306.147745 samples/sec	accuracy=0.569531	lr=1.903807
Epoch[127] Batch [549]	Speed: 8682.255094 samples/sec	accuracy=0.569085	lr=1.902991
Epoch[127] Batch [599]	Speed: 8499.558952 samples/sec	accuracy=0.568723	lr=1.902176
[Epoch 127] training: accuracy=0.568407
[Epoch 127] speed: 8626 samples/sec	time cost: 152.810888
[Epoch 127] validation: err-top1=0.423360 err-top5=0.181844
Epoch[128] Batch [49]	Speed: 9194.518107 samples/sec	accuracy=0.577617	lr=1.900952
Epoch[128] Batch [99]	Speed: 8642.026307 samples/sec	accuracy=0.577402	lr=1.900136
Epoch[128] Batch [149]	Speed: 8927.126070 samples/sec	accuracy=0.575827	lr=1.899319
Epoch[128] Batch [199]	Speed: 8305.364885 samples/sec	accuracy=0.575232	lr=1.898502
Epoch[128] Batch [249]	Speed: 8551.458161 samples/sec	accuracy=0.574002	lr=1.897685
Epoch[128] Batch [299]	Speed: 8507.128584 samples/sec	accuracy=0.573102	lr=1.896868
Epoch[128] Batch [349]	Speed: 8893.331669 samples/sec	accuracy=0.572461	lr=1.896050
Epoch[128] Batch [399]	Speed: 8736.383639 samples/sec	accuracy=0.571624	lr=1.895232
Epoch[128] Batch [449]	Speed: 8544.766954 samples/sec	accuracy=0.570935	lr=1.894413
Epoch[128] Batch [499]	Speed: 8766.125832 samples/sec	accuracy=0.570436	lr=1.893595
Epoch[128] Batch [549]	Speed: 8727.717142 samples/sec	accuracy=0.570092	lr=1.892776
Epoch[128] Batch [599]	Speed: 8575.966572 samples/sec	accuracy=0.569747	lr=1.891957
[Epoch 128] training: accuracy=0.569586
[Epoch 128] speed: 8693 samples/sec	time cost: 152.528385
[Epoch 128] validation: err-top1=0.450098 err-top5=0.204688
Epoch[129] Batch [49]	Speed: 8827.470525 samples/sec	accuracy=0.575068	lr=1.890711
Epoch[129] Batch [99]	Speed: 8788.819244 samples/sec	accuracy=0.574775	lr=1.889891
Epoch[129] Batch [149]	Speed: 8798.535191 samples/sec	accuracy=0.573887	lr=1.889070
Epoch[129] Batch [199]	Speed: 8980.033910 samples/sec	accuracy=0.573076	lr=1.888250
Epoch[129] Batch [249]	Speed: 8688.544006 samples/sec	accuracy=0.572311	lr=1.887429
Epoch[129] Batch [299]	Speed: 8698.163580 samples/sec	accuracy=0.571748	lr=1.886608
Epoch[129] Batch [349]	Speed: 8645.439524 samples/sec	accuracy=0.571303	lr=1.885786
Epoch[129] Batch [399]	Speed: 8800.107200 samples/sec	accuracy=0.570924	lr=1.884965
Epoch[129] Batch [449]	Speed: 8599.847464 samples/sec	accuracy=0.570507	lr=1.884142
Epoch[129] Batch [499]	Speed: 8590.111033 samples/sec	accuracy=0.570116	lr=1.883320
Epoch[129] Batch [549]	Speed: 8638.072149 samples/sec	accuracy=0.569999	lr=1.882497
Epoch[129] Batch [599]	Speed: 8853.032555 samples/sec	accuracy=0.569631	lr=1.881675
[Epoch 129] training: accuracy=0.569557
[Epoch 129] speed: 8735 samples/sec	time cost: 151.373192
[Epoch 129] validation: err-top1=0.400452 err-top5=0.163635
Epoch[130] Batch [49]	Speed: 8403.443360 samples/sec	accuracy=0.578184	lr=1.880423
Epoch[130] Batch [99]	Speed: 8731.222182 samples/sec	accuracy=0.577100	lr=1.879599
Epoch[130] Batch [149]	Speed: 8552.073706 samples/sec	accuracy=0.575482	lr=1.878775
Epoch[130] Batch [199]	Speed: 8656.184004 samples/sec	accuracy=0.574402	lr=1.877951
Epoch[130] Batch [249]	Speed: 8811.613643 samples/sec	accuracy=0.573504	lr=1.877127
Epoch[130] Batch [299]	Speed: 8797.531887 samples/sec	accuracy=0.572894	lr=1.876302
Epoch[130] Batch [349]	Speed: 8775.360594 samples/sec	accuracy=0.572592	lr=1.875477
Epoch[130] Batch [399]	Speed: 8781.301276 samples/sec	accuracy=0.572249	lr=1.874651
Epoch[130] Batch [449]	Speed: 8757.388827 samples/sec	accuracy=0.571713	lr=1.873826
Epoch[130] Batch [499]	Speed: 8851.453804 samples/sec	accuracy=0.571265	lr=1.873000
Epoch[130] Batch [549]	Speed: 8838.770373 samples/sec	accuracy=0.571150	lr=1.872173
Epoch[130] Batch [599]	Speed: 8876.773502 samples/sec	accuracy=0.570705	lr=1.871347
[Epoch 130] training: accuracy=0.570580
[Epoch 130] speed: 8730 samples/sec	time cost: 151.448490
[Epoch 130] validation: err-top1=0.407969 err-top5=0.168418
Epoch[131] Batch [49]	Speed: 8653.664853 samples/sec	accuracy=0.580068	lr=1.870106
Epoch[131] Batch [99]	Speed: 8889.161180 samples/sec	accuracy=0.578403	lr=1.869279
Epoch[131] Batch [149]	Speed: 8690.780673 samples/sec	accuracy=0.575876	lr=1.868452
Epoch[131] Batch [199]	Speed: 8505.027204 samples/sec	accuracy=0.575000	lr=1.867624
Epoch[131] Batch [249]	Speed: 8659.860541 samples/sec	accuracy=0.574104	lr=1.866796
Epoch[131] Batch [299]	Speed: 8667.854093 samples/sec	accuracy=0.573145	lr=1.865967
Epoch[131] Batch [349]	Speed: 8554.327716 samples/sec	accuracy=0.572446	lr=1.865138
Epoch[131] Batch [399]	Speed: 8388.027555 samples/sec	accuracy=0.572104	lr=1.864309
Epoch[131] Batch [449]	Speed: 8597.865263 samples/sec	accuracy=0.571471	lr=1.863480
Epoch[131] Batch [499]	Speed: 8928.659356 samples/sec	accuracy=0.570994	lr=1.862651
Epoch[131] Batch [549]	Speed: 8699.012554 samples/sec	accuracy=0.570925	lr=1.861821
Epoch[131] Batch [599]	Speed: 8603.841842 samples/sec	accuracy=0.570821	lr=1.860991
[Epoch 131] training: accuracy=0.570504
[Epoch 131] speed: 8649 samples/sec	time cost: 152.777276
[Epoch 131] validation: err-top1=0.427836 err-top5=0.186991
Epoch[132] Batch [49]	Speed: 8244.226792 samples/sec	accuracy=0.579248	lr=1.859728
Epoch[132] Batch [99]	Speed: 8761.990648 samples/sec	accuracy=0.579160	lr=1.858898
Epoch[132] Batch [149]	Speed: 8882.601017 samples/sec	accuracy=0.578298	lr=1.858066
Epoch[132] Batch [199]	Speed: 8744.040791 samples/sec	accuracy=0.576924	lr=1.857235
Epoch[132] Batch [249]	Speed: 8707.914644 samples/sec	accuracy=0.575879	lr=1.856403
Epoch[132] Batch [299]	Speed: 8524.170616 samples/sec	accuracy=0.574919	lr=1.855571
Epoch[132] Batch [349]	Speed: 8543.979092 samples/sec	accuracy=0.573947	lr=1.854739
Epoch[132] Batch [399]	Speed: 8550.244358 samples/sec	accuracy=0.573724	lr=1.853907
Epoch[132] Batch [449]	Speed: 8606.442265 samples/sec	accuracy=0.573013	lr=1.853074
Epoch[132] Batch [499]	Speed: 8297.507846 samples/sec	accuracy=0.572357	lr=1.852241
Epoch[132] Batch [549]	Speed: 8803.725855 samples/sec	accuracy=0.572086	lr=1.851408
Epoch[132] Batch [599]	Speed: 8515.148596 samples/sec	accuracy=0.571622	lr=1.850574
[Epoch 132] training: accuracy=0.571440
[Epoch 132] speed: 8602 samples/sec	time cost: 153.581880
[Epoch 132] validation: err-top1=0.406113 err-top5=0.167578
Epoch[133] Batch [49]	Speed: 8343.340165 samples/sec	accuracy=0.581953	lr=1.849323
Epoch[133] Batch [99]	Speed: 8616.083717 samples/sec	accuracy=0.579282	lr=1.848489
Epoch[133] Batch [149]	Speed: 8209.485524 samples/sec	accuracy=0.577054	lr=1.847654
Epoch[133] Batch [199]	Speed: 9017.950322 samples/sec	accuracy=0.575718	lr=1.846819
Epoch[133] Batch [249]	Speed: 8711.087526 samples/sec	accuracy=0.575068	lr=1.845984
Epoch[133] Batch [299]	Speed: 8501.190999 samples/sec	accuracy=0.574365	lr=1.845149
Epoch[133] Batch [349]	Speed: 8555.103855 samples/sec	accuracy=0.574208	lr=1.844313
Epoch[133] Batch [399]	Speed: 8732.552544 samples/sec	accuracy=0.573433	lr=1.843477
Epoch[133] Batch [449]	Speed: 8559.735162 samples/sec	accuracy=0.572887	lr=1.842641
Epoch[133] Batch [499]	Speed: 8582.606076 samples/sec	accuracy=0.572528	lr=1.841805
Epoch[133] Batch [549]	Speed: 8603.100777 samples/sec	accuracy=0.572227	lr=1.840968
Epoch[133] Batch [599]	Speed: 8639.872363 samples/sec	accuracy=0.571918	lr=1.840131
[Epoch 133] training: accuracy=0.571880
[Epoch 133] speed: 8596 samples/sec	time cost: 153.786358
[Epoch 133] validation: err-top1=0.395671 err-top5=0.161133
Epoch[134] Batch [49]	Speed: 8440.101853 samples/sec	accuracy=0.580352	lr=1.838858
Epoch[134] Batch [99]	Speed: 8705.918671 samples/sec	accuracy=0.579023	lr=1.838020
Epoch[134] Batch [149]	Speed: 8398.135291 samples/sec	accuracy=0.579046	lr=1.837182
Epoch[134] Batch [199]	Speed: 8395.922127 samples/sec	accuracy=0.576936	lr=1.836344
Epoch[134] Batch [249]	Speed: 8703.979881 samples/sec	accuracy=0.576443	lr=1.835506
Epoch[134] Batch [299]	Speed: 8233.805856 samples/sec	accuracy=0.575791	lr=1.834667
Epoch[134] Batch [349]	Speed: 8289.691099 samples/sec	accuracy=0.575304	lr=1.833828
Epoch[134] Batch [399]	Speed: 8635.777444 samples/sec	accuracy=0.574825	lr=1.832988
Epoch[134] Batch [449]	Speed: 8499.304973 samples/sec	accuracy=0.574433	lr=1.832149
Epoch[134] Batch [499]	Speed: 8589.408577 samples/sec	accuracy=0.573639	lr=1.831309
Epoch[134] Batch [549]	Speed: 8575.963147 samples/sec	accuracy=0.573165	lr=1.830469
Epoch[134] Batch [599]	Speed: 8795.761570 samples/sec	accuracy=0.572612	lr=1.829629
[Epoch 134] training: accuracy=0.572417
[Epoch 134] speed: 8524 samples/sec	time cost: 154.900485
[Epoch 134] validation: err-top1=0.427877 err-top5=0.184570
Epoch[135] Batch [49]	Speed: 8833.197494 samples/sec	accuracy=0.581846	lr=1.828367
Epoch[135] Batch [99]	Speed: 8512.641172 samples/sec	accuracy=0.580459	lr=1.827526
Epoch[135] Batch [149]	Speed: 8456.870638 samples/sec	accuracy=0.578346	lr=1.826685
Epoch[135] Batch [199]	Speed: 8441.773199 samples/sec	accuracy=0.577786	lr=1.825844
Epoch[135] Batch [249]	Speed: 8623.796273 samples/sec	accuracy=0.576855	lr=1.825002
Epoch[135] Batch [299]	Speed: 8700.429525 samples/sec	accuracy=0.575645	lr=1.824160
Epoch[135] Batch [349]	Speed: 8581.523498 samples/sec	accuracy=0.574876	lr=1.823317
Epoch[135] Batch [399]	Speed: 8711.426233 samples/sec	accuracy=0.574659	lr=1.822475
Epoch[135] Batch [449]	Speed: 8825.354089 samples/sec	accuracy=0.574215	lr=1.821632
Epoch[135] Batch [499]	Speed: 8460.436111 samples/sec	accuracy=0.574049	lr=1.820789
Epoch[135] Batch [549]	Speed: 8695.137583 samples/sec	accuracy=0.573575	lr=1.819945
Epoch[135] Batch [599]	Speed: 8617.669527 samples/sec	accuracy=0.573314	lr=1.819102
[Epoch 135] training: accuracy=0.573256
[Epoch 135] speed: 8621 samples/sec	time cost: 153.568711
[Epoch 135] validation: err-top1=0.446250 err-top5=0.201895
Epoch[136] Batch [49]	Speed: 8508.748707 samples/sec	accuracy=0.582441	lr=1.817819
Epoch[136] Batch [99]	Speed: 8891.884494 samples/sec	accuracy=0.580205	lr=1.816974
Epoch[136] Batch [149]	Speed: 8571.533909 samples/sec	accuracy=0.578180	lr=1.816130
Epoch[136] Batch [199]	Speed: 8821.092195 samples/sec	accuracy=0.577102	lr=1.815285
Epoch[136] Batch [249]	Speed: 8356.518317 samples/sec	accuracy=0.576295	lr=1.814440
Epoch[136] Batch [299]	Speed: 8572.203332 samples/sec	accuracy=0.575627	lr=1.813594
Epoch[136] Batch [349]	Speed: 8501.772739 samples/sec	accuracy=0.574764	lr=1.812749
Epoch[136] Batch [399]	Speed: 8475.572171 samples/sec	accuracy=0.574368	lr=1.811903
Epoch[136] Batch [449]	Speed: 8741.878760 samples/sec	accuracy=0.574027	lr=1.811057
Epoch[136] Batch [499]	Speed: 8346.554894 samples/sec	accuracy=0.573697	lr=1.810210
Epoch[136] Batch [549]	Speed: 8759.095321 samples/sec	accuracy=0.573337	lr=1.809364
Epoch[136] Batch [599]	Speed: 8617.211168 samples/sec	accuracy=0.573016	lr=1.808517
[Epoch 136] training: accuracy=0.572840
[Epoch 136] speed: 8593 samples/sec	time cost: 153.812787
[Epoch 136] validation: err-top1=0.409424 err-top5=0.170959
Epoch[137] Batch [49]	Speed: 8551.968640 samples/sec	accuracy=0.583359	lr=1.807229
Epoch[137] Batch [99]	Speed: 8740.705824 samples/sec	accuracy=0.582080	lr=1.806382
Epoch[137] Batch [149]	Speed: 8526.662662 samples/sec	accuracy=0.580592	lr=1.805534
Epoch[137] Batch [199]	Speed: 8628.145196 samples/sec	accuracy=0.579690	lr=1.804686
Epoch[137] Batch [249]	Speed: 8603.666905 samples/sec	accuracy=0.578475	lr=1.803837
Epoch[137] Batch [299]	Speed: 8617.504056 samples/sec	accuracy=0.577606	lr=1.802989
Epoch[137] Batch [349]	Speed: 8593.047853 samples/sec	accuracy=0.576745	lr=1.802140
Epoch[137] Batch [399]	Speed: 8451.890207 samples/sec	accuracy=0.575988	lr=1.801291
Epoch[137] Batch [449]	Speed: 8583.147039 samples/sec	accuracy=0.575075	lr=1.800442
Epoch[137] Batch [499]	Speed: 8587.298120 samples/sec	accuracy=0.574675	lr=1.799592
Epoch[137] Batch [549]	Speed: 8731.145326 samples/sec	accuracy=0.574192	lr=1.798742
Epoch[137] Batch [599]	Speed: 8840.493267 samples/sec	accuracy=0.573858	lr=1.797892
[Epoch 137] training: accuracy=0.573805
[Epoch 137] speed: 8614 samples/sec	time cost: 153.449510
[Epoch 137] validation: err-top1=0.381816 err-top5=0.154727
Epoch[138] Batch [49]	Speed: 8761.673557 samples/sec	accuracy=0.583477	lr=1.796617
Epoch[138] Batch [99]	Speed: 8536.271153 samples/sec	accuracy=0.583643	lr=1.795766
Epoch[138] Batch [149]	Speed: 8566.051413 samples/sec	accuracy=0.581009	lr=1.794915
Epoch[138] Batch [199]	Speed: 8489.534471 samples/sec	accuracy=0.580027	lr=1.794064
Epoch[138] Batch [249]	Speed: 8880.804381 samples/sec	accuracy=0.579555	lr=1.793213
Epoch[138] Batch [299]	Speed: 8530.751480 samples/sec	accuracy=0.578506	lr=1.792361
Epoch[138] Batch [349]	Speed: 8815.943831 samples/sec	accuracy=0.577610	lr=1.791509
Epoch[138] Batch [399]	Speed: 8371.202602 samples/sec	accuracy=0.576952	lr=1.790657
Epoch[138] Batch [449]	Speed: 8584.276009 samples/sec	accuracy=0.576155	lr=1.789804
Epoch[138] Batch [499]	Speed: 8493.119821 samples/sec	accuracy=0.575781	lr=1.788952
Epoch[138] Batch [549]	Speed: 8544.439043 samples/sec	accuracy=0.575354	lr=1.788099
Epoch[138] Batch [599]	Speed: 8713.840525 samples/sec	accuracy=0.574819	lr=1.787246
[Epoch 138] training: accuracy=0.574621
[Epoch 138] speed: 8610 samples/sec	time cost: 153.538174
[Epoch 138] validation: err-top1=0.408468 err-top5=0.171122
Epoch[139] Batch [49]	Speed: 8672.502365 samples/sec	accuracy=0.583203	lr=1.785948
Epoch[139] Batch [99]	Speed: 8670.914867 samples/sec	accuracy=0.582476	lr=1.785095
Epoch[139] Batch [149]	Speed: 8531.690956 samples/sec	accuracy=0.581331	lr=1.784241
Epoch[139] Batch [199]	Speed: 8584.533718 samples/sec	accuracy=0.580962	lr=1.783386
Epoch[139] Batch [249]	Speed: 8685.564569 samples/sec	accuracy=0.579934	lr=1.782532
Epoch[139] Batch [299]	Speed: 8640.127512 samples/sec	accuracy=0.579146	lr=1.781677
Epoch[139] Batch [349]	Speed: 8634.396727 samples/sec	accuracy=0.578465	lr=1.780822
Epoch[139] Batch [399]	Speed: 8586.032927 samples/sec	accuracy=0.577437	lr=1.779967
Epoch[139] Batch [449]	Speed: 8521.589231 samples/sec	accuracy=0.576911	lr=1.779112
Epoch[139] Batch [499]	Speed: 8662.033897 samples/sec	accuracy=0.576141	lr=1.778256
Epoch[139] Batch [549]	Speed: 8560.721984 samples/sec	accuracy=0.575730	lr=1.777400
Epoch[139] Batch [599]	Speed: 8548.415621 samples/sec	accuracy=0.575541	lr=1.776544
[Epoch 139] training: accuracy=0.575391
[Epoch 139] speed: 8611 samples/sec	time cost: 153.466944
[Epoch 139] validation: err-top1=0.404195 err-top5=0.168132
Epoch[140] Batch [49]	Speed: 8400.049787 samples/sec	accuracy=0.581885	lr=1.775259
Epoch[140] Batch [99]	Speed: 8738.657454 samples/sec	accuracy=0.581143	lr=1.774402
Epoch[140] Batch [149]	Speed: 8638.877287 samples/sec	accuracy=0.580749	lr=1.773545
Epoch[140] Batch [199]	Speed: 8741.890148 samples/sec	accuracy=0.579893	lr=1.772688
Epoch[140] Batch [249]	Speed: 8893.678619 samples/sec	accuracy=0.578660	lr=1.771831
Epoch[140] Batch [299]	Speed: 8384.515145 samples/sec	accuracy=0.577852	lr=1.770973
Epoch[140] Batch [349]	Speed: 8580.499646 samples/sec	accuracy=0.577158	lr=1.770115
Epoch[140] Batch [399]	Speed: 8523.117611 samples/sec	accuracy=0.576781	lr=1.769257
Epoch[140] Batch [449]	Speed: 8397.527749 samples/sec	accuracy=0.576475	lr=1.768398
Epoch[140] Batch [499]	Speed: 8476.103405 samples/sec	accuracy=0.576428	lr=1.767540
Epoch[140] Batch [549]	Speed: 8810.419571 samples/sec	accuracy=0.576278	lr=1.766681
Epoch[140] Batch [599]	Speed: 8794.441052 samples/sec	accuracy=0.575819	lr=1.765822
[Epoch 140] training: accuracy=0.575617
[Epoch 140] speed: 8611 samples/sec	time cost: 153.865437
[Epoch 140] validation: err-top1=0.404980 err-top5=0.165508
Epoch[141] Batch [49]	Speed: 8250.274107 samples/sec	accuracy=0.586631	lr=1.764515
Epoch[141] Batch [99]	Speed: 8645.829882 samples/sec	accuracy=0.585391	lr=1.763656
Epoch[141] Batch [149]	Speed: 8700.620228 samples/sec	accuracy=0.583441	lr=1.762796
Epoch[141] Batch [199]	Speed: 8567.838821 samples/sec	accuracy=0.581831	lr=1.761935
Epoch[141] Batch [249]	Speed: 8649.229558 samples/sec	accuracy=0.580836	lr=1.761075
Epoch[141] Batch [299]	Speed: 8464.658159 samples/sec	accuracy=0.579764	lr=1.760214
Epoch[141] Batch [349]	Speed: 8661.321027 samples/sec	accuracy=0.579169	lr=1.759354
Epoch[141] Batch [399]	Speed: 8497.451390 samples/sec	accuracy=0.578363	lr=1.758492
Epoch[141] Batch [449]	Speed: 8707.400737 samples/sec	accuracy=0.577912	lr=1.757631
Epoch[141] Batch [499]	Speed: 8689.779639 samples/sec	accuracy=0.577548	lr=1.756770
Epoch[141] Batch [549]	Speed: 8621.678232 samples/sec	accuracy=0.577124	lr=1.755908
Epoch[141] Batch [599]	Speed: 8695.291791 samples/sec	accuracy=0.576627	lr=1.755046
[Epoch 141] training: accuracy=0.576634
[Epoch 141] speed: 8594 samples/sec	time cost: 153.564435
[Epoch 141] validation: err-top1=0.383036 err-top5=0.151103
Epoch[142] Batch [49]	Speed: 8505.623450 samples/sec	accuracy=0.584570	lr=1.753752
Epoch[142] Batch [99]	Speed: 8554.069943 samples/sec	accuracy=0.582900	lr=1.752890
Epoch[142] Batch [149]	Speed: 8390.637977 samples/sec	accuracy=0.581917	lr=1.752027
Epoch[142] Batch [199]	Speed: 8503.741013 samples/sec	accuracy=0.581455	lr=1.751164
Epoch[142] Batch [249]	Speed: 8583.136919 samples/sec	accuracy=0.580004	lr=1.750301
Epoch[142] Batch [299]	Speed: 8336.210359 samples/sec	accuracy=0.579512	lr=1.749437
Epoch[142] Batch [349]	Speed: 8782.759910 samples/sec	accuracy=0.579135	lr=1.748573
Epoch[142] Batch [399]	Speed: 8742.633427 samples/sec	accuracy=0.578398	lr=1.747709
Epoch[142] Batch [449]	Speed: 8704.715316 samples/sec	accuracy=0.577954	lr=1.746845
Epoch[142] Batch [499]	Speed: 8749.535749 samples/sec	accuracy=0.577514	lr=1.745981
Epoch[142] Batch [549]	Speed: 8584.627232 samples/sec	accuracy=0.577061	lr=1.745116
Epoch[142] Batch [599]	Speed: 8652.684204 samples/sec	accuracy=0.576785	lr=1.744251
[Epoch 142] training: accuracy=0.576738
[Epoch 142] speed: 8598 samples/sec	time cost: 153.997623
[Epoch 142] validation: err-top1=0.389609 err-top5=0.157715
Epoch[143] Batch [49]	Speed: 8628.199622 samples/sec	accuracy=0.585820	lr=1.742936
Epoch[143] Batch [99]	Speed: 9008.933797 samples/sec	accuracy=0.582988	lr=1.742071
Epoch[143] Batch [149]	Speed: 8440.187436 samples/sec	accuracy=0.581217	lr=1.741205
Epoch[143] Batch [199]	Speed: 8425.865631 samples/sec	accuracy=0.581035	lr=1.740340
Epoch[143] Batch [249]	Speed: 8620.677307 samples/sec	accuracy=0.579941	lr=1.739473
Epoch[143] Batch [299]	Speed: 8803.573011 samples/sec	accuracy=0.579707	lr=1.738607
Epoch[143] Batch [349]	Speed: 8528.779660 samples/sec	accuracy=0.578943	lr=1.737741
Epoch[143] Batch [399]	Speed: 8484.909234 samples/sec	accuracy=0.578152	lr=1.736874
Epoch[143] Batch [449]	Speed: 8864.099414 samples/sec	accuracy=0.577823	lr=1.736007
Epoch[143] Batch [499]	Speed: 8531.368962 samples/sec	accuracy=0.577377	lr=1.735140
Epoch[143] Batch [549]	Speed: 8558.571194 samples/sec	accuracy=0.577069	lr=1.734273
Epoch[143] Batch [599]	Speed: 8675.540488 samples/sec	accuracy=0.576675	lr=1.733405
[Epoch 143] training: accuracy=0.576580
[Epoch 143] speed: 8632 samples/sec	time cost: 153.222892
[Epoch 143] validation: err-top1=0.418844 err-top5=0.179728
Epoch[144] Batch [49]	Speed: 8369.802753 samples/sec	accuracy=0.587910	lr=1.732086
Epoch[144] Batch [99]	Speed: 8787.161375 samples/sec	accuracy=0.585518	lr=1.731218
Epoch[144] Batch [149]	Speed: 8711.743938 samples/sec	accuracy=0.583105	lr=1.730349
Epoch[144] Batch [199]	Speed: 8597.645821 samples/sec	accuracy=0.582100	lr=1.729481
Epoch[144] Batch [249]	Speed: 8511.340027 samples/sec	accuracy=0.580941	lr=1.728612
Epoch[144] Batch [299]	Speed: 8641.138696 samples/sec	accuracy=0.579897	lr=1.727743
Epoch[144] Batch [349]	Speed: 8409.426751 samples/sec	accuracy=0.579711	lr=1.726874
Epoch[144] Batch [399]	Speed: 8717.008704 samples/sec	accuracy=0.579463	lr=1.726004
Epoch[144] Batch [449]	Speed: 8403.042194 samples/sec	accuracy=0.578808	lr=1.725135
Epoch[144] Batch [499]	Speed: 8671.081346 samples/sec	accuracy=0.578248	lr=1.724265
Epoch[144] Batch [549]	Speed: 8654.394600 samples/sec	accuracy=0.577861	lr=1.723395
Epoch[144] Batch [599]	Speed: 8682.524512 samples/sec	accuracy=0.577584	lr=1.722524
[Epoch 144] training: accuracy=0.577516
[Epoch 144] speed: 8595 samples/sec	time cost: 153.855133
[Epoch 144] validation: err-top1=0.408945 err-top5=0.169180
Epoch[145] Batch [49]	Speed: 8387.849489 samples/sec	accuracy=0.587852	lr=1.721219
Epoch[145] Batch [99]	Speed: 8750.255727 samples/sec	accuracy=0.586587	lr=1.720348
Epoch[145] Batch [149]	Speed: 8654.088213 samples/sec	accuracy=0.585218	lr=1.719477
Epoch[145] Batch [199]	Speed: 8859.044913 samples/sec	accuracy=0.584407	lr=1.718605
Epoch[145] Batch [249]	Speed: 8586.543423 samples/sec	accuracy=0.582727	lr=1.717734
Epoch[145] Batch [299]	Speed: 8480.662109 samples/sec	accuracy=0.581434	lr=1.716862
Epoch[145] Batch [349]	Speed: 8583.023884 samples/sec	accuracy=0.580845	lr=1.715991
Epoch[145] Batch [399]	Speed: 8568.950945 samples/sec	accuracy=0.580507	lr=1.715118
Epoch[145] Batch [449]	Speed: 8558.343009 samples/sec	accuracy=0.579782	lr=1.714246
Epoch[145] Batch [499]	Speed: 8728.687910 samples/sec	accuracy=0.579268	lr=1.713374
Epoch[145] Batch [549]	Speed: 8665.924874 samples/sec	accuracy=0.578814	lr=1.712501
Epoch[145] Batch [599]	Speed: 8599.317996 samples/sec	accuracy=0.578429	lr=1.711628
[Epoch 145] training: accuracy=0.578170
[Epoch 145] speed: 8614 samples/sec	time cost: 153.641884
[Epoch 145] validation: err-top1=0.408590 err-top5=0.170369
Epoch[146] Batch [49]	Speed: 8586.478706 samples/sec	accuracy=0.587617	lr=1.710301
Epoch[146] Batch [99]	Speed: 8521.205615 samples/sec	accuracy=0.585928	lr=1.709427
Epoch[146] Batch [149]	Speed: 8519.039148 samples/sec	accuracy=0.585020	lr=1.708554
Epoch[146] Batch [199]	Speed: 8572.560583 samples/sec	accuracy=0.584280	lr=1.707680
Epoch[146] Batch [249]	Speed: 8401.257143 samples/sec	accuracy=0.582449	lr=1.706806
Epoch[146] Batch [299]	Speed: 8503.481229 samples/sec	accuracy=0.581335	lr=1.705932
Epoch[146] Batch [349]	Speed: 8435.677101 samples/sec	accuracy=0.580819	lr=1.705057
Epoch[146] Batch [399]	Speed: 8455.167508 samples/sec	accuracy=0.580255	lr=1.704183
Epoch[146] Batch [449]	Speed: 8684.484661 samples/sec	accuracy=0.579897	lr=1.703308
Epoch[146] Batch [499]	Speed: 8779.071972 samples/sec	accuracy=0.579692	lr=1.702433
Epoch[146] Batch [549]	Speed: 8583.427667 samples/sec	accuracy=0.579331	lr=1.701558
Epoch[146] Batch [599]	Speed: 8715.400097 samples/sec	accuracy=0.578901	lr=1.700682
[Epoch 146] training: accuracy=0.578699
[Epoch 146] speed: 8568 samples/sec	time cost: 154.147003
[Epoch 146] validation: err-top1=0.389323 err-top5=0.156637
Epoch[147] Batch [49]	Speed: 8606.982615 samples/sec	accuracy=0.587217	lr=1.699369
Epoch[147] Batch [99]	Speed: 8575.883350 samples/sec	accuracy=0.585288	lr=1.698493
Epoch[147] Batch [149]	Speed: 8576.071373 samples/sec	accuracy=0.584635	lr=1.697616
Epoch[147] Batch [199]	Speed: 8483.646715 samples/sec	accuracy=0.583315	lr=1.696740
Epoch[147] Batch [249]	Speed: 8701.908135 samples/sec	accuracy=0.582799	lr=1.695864
Epoch[147] Batch [299]	Speed: 8798.148765 samples/sec	accuracy=0.582358	lr=1.694987
Epoch[147] Batch [349]	Speed: 8708.615252 samples/sec	accuracy=0.582037	lr=1.694110
Epoch[147] Batch [399]	Speed: 8472.954274 samples/sec	accuracy=0.581718	lr=1.693233
Epoch[147] Batch [449]	Speed: 8614.904203 samples/sec	accuracy=0.581026	lr=1.692355
Epoch[147] Batch [499]	Speed: 8524.671582 samples/sec	accuracy=0.580603	lr=1.691478
Epoch[147] Batch [549]	Speed: 8563.388601 samples/sec	accuracy=0.580297	lr=1.690600
Epoch[147] Batch [599]	Speed: 8552.643016 samples/sec	accuracy=0.580014	lr=1.689722
[Epoch 147] training: accuracy=0.579822
[Epoch 147] speed: 8600 samples/sec	time cost: 154.049953
[Epoch 147] validation: err-top1=0.391660 err-top5=0.159629
Epoch[148] Batch [49]	Speed: 8428.451860 samples/sec	accuracy=0.590703	lr=1.688387
Epoch[148] Batch [99]	Speed: 8626.061919 samples/sec	accuracy=0.589082	lr=1.687509
Epoch[148] Batch [149]	Speed: 8773.910331 samples/sec	accuracy=0.587041	lr=1.686630
Epoch[148] Batch [199]	Speed: 8760.817491 samples/sec	accuracy=0.586311	lr=1.685751
Epoch[148] Batch [249]	Speed: 8763.299114 samples/sec	accuracy=0.585234	lr=1.684872
Epoch[148] Batch [299]	Speed: 8612.183652 samples/sec	accuracy=0.583732	lr=1.683993
Epoch[148] Batch [349]	Speed: 8585.850818 samples/sec	accuracy=0.583118	lr=1.683114
Epoch[148] Batch [399]	Speed: 8524.196838 samples/sec	accuracy=0.582770	lr=1.682234
Epoch[148] Batch [449]	Speed: 8776.135399 samples/sec	accuracy=0.582464	lr=1.681355
Epoch[148] Batch [499]	Speed: 8627.917619 samples/sec	accuracy=0.581956	lr=1.680475
Epoch[148] Batch [549]	Speed: 8458.046080 samples/sec	accuracy=0.581467	lr=1.679594
Epoch[148] Batch [599]	Speed: 8698.451075 samples/sec	accuracy=0.580934	lr=1.678714
[Epoch 148] training: accuracy=0.580673
[Epoch 148] speed: 8636 samples/sec	time cost: 152.843353
[Epoch 148] validation: err-top1=0.424357 err-top5=0.183838
Epoch[149] Batch [49]	Speed: 8555.822187 samples/sec	accuracy=0.590361	lr=1.677393
Epoch[149] Batch [99]	Speed: 8651.713885 samples/sec	accuracy=0.589224	lr=1.676512
Epoch[149] Batch [149]	Speed: 8744.038477 samples/sec	accuracy=0.587663	lr=1.675631
Epoch[149] Batch [199]	Speed: 8918.185719 samples/sec	accuracy=0.586187	lr=1.674750
Epoch[149] Batch [249]	Speed: 8486.156535 samples/sec	accuracy=0.585438	lr=1.673869
Epoch[149] Batch [299]	Speed: 8639.502007 samples/sec	accuracy=0.584699	lr=1.672987
Epoch[149] Batch [349]	Speed: 8654.889887 samples/sec	accuracy=0.583806	lr=1.672105
Epoch[149] Batch [399]	Speed: 8896.133272 samples/sec	accuracy=0.583290	lr=1.671224
Epoch[149] Batch [449]	Speed: 8463.281250 samples/sec	accuracy=0.582752	lr=1.670341
Epoch[149] Batch [499]	Speed: 8647.304790 samples/sec	accuracy=0.582098	lr=1.669459
Epoch[149] Batch [549]	Speed: 8682.400069 samples/sec	accuracy=0.581515	lr=1.668577
Epoch[149] Batch [599]	Speed: 8450.178609 samples/sec	accuracy=0.581309	lr=1.667694
[Epoch 149] training: accuracy=0.581192
[Epoch 149] speed: 8645 samples/sec	time cost: 153.136764
[Epoch 149] validation: err-top1=0.444336 err-top5=0.200254
Epoch[150] Batch [49]	Speed: 8433.367770 samples/sec	accuracy=0.590918	lr=1.666352
Epoch[150] Batch [99]	Speed: 8651.092800 samples/sec	accuracy=0.589766	lr=1.665469
Epoch[150] Batch [149]	Speed: 8564.426475 samples/sec	accuracy=0.587949	lr=1.664585
Epoch[150] Batch [199]	Speed: 8638.074408 samples/sec	accuracy=0.586936	lr=1.663702
Epoch[150] Batch [249]	Speed: 8798.011974 samples/sec	accuracy=0.586025	lr=1.662818
Epoch[150] Batch [299]	Speed: 8695.147089 samples/sec	accuracy=0.584928	lr=1.661934
Epoch[150] Batch [349]	Speed: 8646.669541 samples/sec	accuracy=0.584537	lr=1.661050
Epoch[150] Batch [399]	Speed: 8461.947634 samples/sec	accuracy=0.584041	lr=1.660166
Epoch[150] Batch [449]	Speed: 8658.942903 samples/sec	accuracy=0.583066	lr=1.659282
Epoch[150] Batch [499]	Speed: 8659.349845 samples/sec	accuracy=0.582670	lr=1.658397
Epoch[150] Batch [549]	Speed: 8820.883131 samples/sec	accuracy=0.582234	lr=1.657512
Epoch[150] Batch [599]	Speed: 8638.288622 samples/sec	accuracy=0.581881	lr=1.656627
[Epoch 150] training: accuracy=0.581647
[Epoch 150] speed: 8639 samples/sec	time cost: 153.042878
[Epoch 150] validation: err-top1=0.392395 err-top5=0.160197
Epoch[151] Batch [49]	Speed: 8429.576400 samples/sec	accuracy=0.589766	lr=1.655282
Epoch[151] Batch [99]	Speed: 8601.322571 samples/sec	accuracy=0.588857	lr=1.654396
Epoch[151] Batch [149]	Speed: 8502.565460 samples/sec	accuracy=0.588014	lr=1.653511
Epoch[151] Batch [199]	Speed: 8781.914263 samples/sec	accuracy=0.587122	lr=1.652625
Epoch[151] Batch [249]	Speed: 8668.116145 samples/sec	accuracy=0.585887	lr=1.651739
Epoch[151] Batch [299]	Speed: 8808.352851 samples/sec	accuracy=0.584989	lr=1.650853
Epoch[151] Batch [349]	Speed: 8596.903070 samples/sec	accuracy=0.584438	lr=1.649967
Epoch[151] Batch [399]	Speed: 8666.703909 samples/sec	accuracy=0.584109	lr=1.649080
Epoch[151] Batch [449]	Speed: 8470.360378 samples/sec	accuracy=0.583592	lr=1.648194
Epoch[151] Batch [499]	Speed: 8813.482404 samples/sec	accuracy=0.583032	lr=1.647307
Epoch[151] Batch [549]	Speed: 8445.188786 samples/sec	accuracy=0.582433	lr=1.646420
Epoch[151] Batch [599]	Speed: 8756.348112 samples/sec	accuracy=0.582204	lr=1.645533
[Epoch 151] training: accuracy=0.582073
[Epoch 151] speed: 8626 samples/sec	time cost: 153.037480
[Epoch 151] validation: err-top1=0.422872 err-top5=0.184509
Epoch[152] Batch [49]	Speed: 8592.914443 samples/sec	accuracy=0.590703	lr=1.644201
Epoch[152] Batch [99]	Speed: 8742.505119 samples/sec	accuracy=0.589658	lr=1.643314
Epoch[152] Batch [149]	Speed: 8613.347738 samples/sec	accuracy=0.588092	lr=1.642426
Epoch[152] Batch [199]	Speed: 8534.676996 samples/sec	accuracy=0.587378	lr=1.641538
Epoch[152] Batch [249]	Speed: 8442.084981 samples/sec	accuracy=0.586047	lr=1.640650
Epoch[152] Batch [299]	Speed: 8725.143960 samples/sec	accuracy=0.585334	lr=1.639762
Epoch[152] Batch [349]	Speed: 8618.283055 samples/sec	accuracy=0.584801	lr=1.638873
Epoch[152] Batch [399]	Speed: 8545.109171 samples/sec	accuracy=0.584131	lr=1.637985
Epoch[152] Batch [449]	Speed: 8718.191570 samples/sec	accuracy=0.583357	lr=1.637096
Epoch[152] Batch [499]	Speed: 8877.478243 samples/sec	accuracy=0.582762	lr=1.636207
Epoch[152] Batch [549]	Speed: 8777.729551 samples/sec	accuracy=0.582237	lr=1.635318
Epoch[152] Batch [599]	Speed: 8712.399212 samples/sec	accuracy=0.582077	lr=1.634429
[Epoch 152] training: accuracy=0.581931
[Epoch 152] speed: 8655 samples/sec	time cost: 153.086795
[Epoch 152] validation: err-top1=0.402344 err-top5=0.169219
Epoch[153] Batch [49]	Speed: 8531.351338 samples/sec	accuracy=0.595059	lr=1.633076
Epoch[153] Batch [99]	Speed: 8537.837052 samples/sec	accuracy=0.592227	lr=1.632187
Epoch[153] Batch [149]	Speed: 8758.338702 samples/sec	accuracy=0.590319	lr=1.631297
Epoch[153] Batch [199]	Speed: 8607.329833 samples/sec	accuracy=0.589846	lr=1.630407
Epoch[153] Batch [249]	Speed: 8688.624507 samples/sec	accuracy=0.588666	lr=1.629517
Epoch[153] Batch [299]	Speed: 8673.910712 samples/sec	accuracy=0.587204	lr=1.628626
Epoch[153] Batch [349]	Speed: 8629.737530 samples/sec	accuracy=0.586210	lr=1.627736
Epoch[153] Batch [399]	Speed: 8757.161345 samples/sec	accuracy=0.585337	lr=1.626845
Epoch[153] Batch [449]	Speed: 8715.930692 samples/sec	accuracy=0.584647	lr=1.625954
Epoch[153] Batch [499]	Speed: 8589.775338 samples/sec	accuracy=0.584430	lr=1.625063
Epoch[153] Batch [549]	Speed: 8785.704153 samples/sec	accuracy=0.583914	lr=1.624172
Epoch[153] Batch [599]	Speed: 8439.857054 samples/sec	accuracy=0.583712	lr=1.623280
[Epoch 153] training: accuracy=0.583571
[Epoch 153] speed: 8649 samples/sec	time cost: 152.547376
[Epoch 153] validation: err-top1=0.385844 err-top5=0.157898
Epoch[154] Batch [49]	Speed: 8450.809091 samples/sec	accuracy=0.594521	lr=1.621943
Epoch[154] Batch [99]	Speed: 8473.869359 samples/sec	accuracy=0.592734	lr=1.621051
Epoch[154] Batch [149]	Speed: 8570.736827 samples/sec	accuracy=0.590059	lr=1.620159
Epoch[154] Batch [199]	Speed: 8444.302961 samples/sec	accuracy=0.588909	lr=1.619267
Epoch[154] Batch [249]	Speed: 8677.594787 samples/sec	accuracy=0.588197	lr=1.618375
Epoch[154] Batch [299]	Speed: 8746.070492 samples/sec	accuracy=0.587433	lr=1.617483
Epoch[154] Batch [349]	Speed: 8482.408528 samples/sec	accuracy=0.586512	lr=1.616590
Epoch[154] Batch [399]	Speed: 8724.503072 samples/sec	accuracy=0.586134	lr=1.615697
Epoch[154] Batch [449]	Speed: 8501.023745 samples/sec	accuracy=0.585734	lr=1.614804
Epoch[154] Batch [499]	Speed: 8759.656618 samples/sec	accuracy=0.585147	lr=1.613911
Epoch[154] Batch [549]	Speed: 8526.627791 samples/sec	accuracy=0.584727	lr=1.613018
Epoch[154] Batch [599]	Speed: 8658.290584 samples/sec	accuracy=0.584272	lr=1.612125
[Epoch 154] training: accuracy=0.584158
[Epoch 154] speed: 8583 samples/sec	time cost: 154.155863
[Epoch 154] validation: err-top1=0.390176 err-top5=0.161758
Epoch[155] Batch [49]	Speed: 8682.939642 samples/sec	accuracy=0.592549	lr=1.610767
Epoch[155] Batch [99]	Speed: 8552.825252 samples/sec	accuracy=0.590967	lr=1.609873
Epoch[155] Batch [149]	Speed: 8515.125806 samples/sec	accuracy=0.588864	lr=1.608979
Epoch[155] Batch [199]	Speed: 8422.985957 samples/sec	accuracy=0.588396	lr=1.608085
Epoch[155] Batch [249]	Speed: 8690.985726 samples/sec	accuracy=0.587516	lr=1.607191
Epoch[155] Batch [299]	Speed: 8591.488449 samples/sec	accuracy=0.586509	lr=1.606296
Epoch[155] Batch [349]	Speed: 8834.743751 samples/sec	accuracy=0.585586	lr=1.605402
Epoch[155] Batch [399]	Speed: 8714.674880 samples/sec	accuracy=0.585057	lr=1.604507
Epoch[155] Batch [449]	Speed: 8635.316288 samples/sec	accuracy=0.584883	lr=1.603612
Epoch[155] Batch [499]	Speed: 8683.360079 samples/sec	accuracy=0.584640	lr=1.602717
Epoch[155] Batch [549]	Speed: 8523.060951 samples/sec	accuracy=0.584449	lr=1.601822
Epoch[155] Batch [599]	Speed: 8699.726533 samples/sec	accuracy=0.583945	lr=1.600927
[Epoch 155] training: accuracy=0.583888
[Epoch 155] speed: 8632 samples/sec	time cost: 152.828429
[Epoch 155] validation: err-top1=0.407979 err-top5=0.172302
Epoch[156] Batch [49]	Speed: 8333.979578 samples/sec	accuracy=0.593682	lr=1.599584
Epoch[156] Batch [99]	Speed: 8646.974707 samples/sec	accuracy=0.592822	lr=1.598688
Epoch[156] Batch [149]	Speed: 8895.317974 samples/sec	accuracy=0.591227	lr=1.597792
Epoch[156] Batch [199]	Speed: 8384.542479 samples/sec	accuracy=0.590503	lr=1.596896
Epoch[156] Batch [249]	Speed: 8647.565601 samples/sec	accuracy=0.589609	lr=1.596000
Epoch[156] Batch [299]	Speed: 8630.755129 samples/sec	accuracy=0.589030	lr=1.595104
Epoch[156] Batch [349]	Speed: 8480.228086 samples/sec	accuracy=0.588133	lr=1.594208
Epoch[156] Batch [399]	Speed: 8435.806834 samples/sec	accuracy=0.587793	lr=1.593311
Epoch[156] Batch [449]	Speed: 8308.276701 samples/sec	accuracy=0.586984	lr=1.592414
Epoch[156] Batch [499]	Speed: 8619.160787 samples/sec	accuracy=0.586429	lr=1.591517
Epoch[156] Batch [549]	Speed: 8718.468709 samples/sec	accuracy=0.585827	lr=1.590620
Epoch[156] Batch [599]	Speed: 8884.585288 samples/sec	accuracy=0.585530	lr=1.589723
[Epoch 156] training: accuracy=0.585154
[Epoch 156] speed: 8584 samples/sec	time cost: 154.161539
[Epoch 156] validation: err-top1=0.391836 err-top5=0.158438
Epoch[157] Batch [49]	Speed: 8381.101022 samples/sec	accuracy=0.593760	lr=1.588359
Epoch[157] Batch [99]	Speed: 8651.790046 samples/sec	accuracy=0.592324	lr=1.587462
Epoch[157] Batch [149]	Speed: 8792.442299 samples/sec	accuracy=0.591230	lr=1.586564
Epoch[157] Batch [199]	Speed: 8705.274076 samples/sec	accuracy=0.590493	lr=1.585666
Epoch[157] Batch [249]	Speed: 8566.892051 samples/sec	accuracy=0.589893	lr=1.584768
Epoch[157] Batch [299]	Speed: 8586.078241 samples/sec	accuracy=0.589181	lr=1.583870
Epoch[157] Batch [349]	Speed: 8535.004668 samples/sec	accuracy=0.588327	lr=1.582972
Epoch[157] Batch [399]	Speed: 8470.622486 samples/sec	accuracy=0.587487	lr=1.582074
Epoch[157] Batch [449]	Speed: 8556.340687 samples/sec	accuracy=0.587137	lr=1.581175
Epoch[157] Batch [499]	Speed: 8601.941871 samples/sec	accuracy=0.586843	lr=1.580277
Epoch[157] Batch [549]	Speed: 8623.272682 samples/sec	accuracy=0.586547	lr=1.579378
Epoch[157] Batch [599]	Speed: 8508.419172 samples/sec	accuracy=0.585988	lr=1.578479
[Epoch 157] training: accuracy=0.585822
[Epoch 157] speed: 8584 samples/sec	time cost: 154.152979
[Epoch 157] validation: err-top1=0.382141 err-top5=0.151550
Epoch[158] Batch [49]	Speed: 8172.941941 samples/sec	accuracy=0.595928	lr=1.577112
Epoch[158] Batch [99]	Speed: 8527.254835 samples/sec	accuracy=0.594316	lr=1.576213
Epoch[158] Batch [149]	Speed: 8649.482125 samples/sec	accuracy=0.592790	lr=1.575314
Epoch[158] Batch [199]	Speed: 8709.956751 samples/sec	accuracy=0.591877	lr=1.574414
Epoch[158] Batch [249]	Speed: 8538.282593 samples/sec	accuracy=0.591187	lr=1.573514
Epoch[158] Batch [299]	Speed: 8597.939618 samples/sec	accuracy=0.590441	lr=1.572615
Epoch[158] Batch [349]	Speed: 8657.478682 samples/sec	accuracy=0.590172	lr=1.571715
Epoch[158] Batch [399]	Speed: 8614.176608 samples/sec	accuracy=0.589276	lr=1.570815
Epoch[158] Batch [449]	Speed: 8662.510141 samples/sec	accuracy=0.588599	lr=1.569914
Epoch[158] Batch [499]	Speed: 8834.186782 samples/sec	accuracy=0.587995	lr=1.569014
Epoch[158] Batch [549]	Speed: 8625.159744 samples/sec	accuracy=0.587578	lr=1.568113
Epoch[158] Batch [599]	Speed: 8812.983990 samples/sec	accuracy=0.587265	lr=1.567213
[Epoch 158] training: accuracy=0.587151
[Epoch 158] speed: 8613 samples/sec	time cost: 153.327472
[Epoch 158] validation: err-top1=0.387288 err-top5=0.154806
Epoch[159] Batch [49]	Speed: 8445.740465 samples/sec	accuracy=0.597744	lr=1.565862
Epoch[159] Batch [99]	Speed: 8813.469382 samples/sec	accuracy=0.594590	lr=1.564961
Epoch[159] Batch [149]	Speed: 8539.000479 samples/sec	accuracy=0.593180	lr=1.564059
Epoch[159] Batch [199]	Speed: 8687.459667 samples/sec	accuracy=0.591953	lr=1.563158
Epoch[159] Batch [249]	Speed: 8633.861781 samples/sec	accuracy=0.591021	lr=1.562257
Epoch[159] Batch [299]	Speed: 8690.900960 samples/sec	accuracy=0.589948	lr=1.561355
Epoch[159] Batch [349]	Speed: 8558.811501 samples/sec	accuracy=0.589308	lr=1.560454
Epoch[159] Batch [399]	Speed: 8682.273523 samples/sec	accuracy=0.588693	lr=1.559552
Epoch[159] Batch [449]	Speed: 8523.351026 samples/sec	accuracy=0.588292	lr=1.558650
Epoch[159] Batch [499]	Speed: 8546.061335 samples/sec	accuracy=0.587939	lr=1.557748
Epoch[159] Batch [549]	Speed: 8822.057934 samples/sec	accuracy=0.587710	lr=1.556846
Epoch[159] Batch [599]	Speed: 8828.348740 samples/sec	accuracy=0.587253	lr=1.555944
[Epoch 159] training: accuracy=0.587022
[Epoch 159] speed: 8644 samples/sec	time cost: 153.129969
[Epoch 159] validation: err-top1=0.384082 err-top5=0.152988
Epoch[160] Batch [49]	Speed: 8429.558366 samples/sec	accuracy=0.596885	lr=1.554572
Epoch[160] Batch [99]	Speed: 8508.680101 samples/sec	accuracy=0.594448	lr=1.553669
Epoch[160] Batch [149]	Speed: 8418.769227 samples/sec	accuracy=0.593382	lr=1.552767
Epoch[160] Batch [199]	Speed: 8456.074426 samples/sec	accuracy=0.591677	lr=1.551864
Epoch[160] Batch [249]	Speed: 8973.622647 samples/sec	accuracy=0.591238	lr=1.550961
Epoch[160] Batch [299]	Speed: 8635.476715 samples/sec	accuracy=0.590091	lr=1.550058
Epoch[160] Batch [349]	Speed: 8670.446275 samples/sec	accuracy=0.589939	lr=1.549154
Epoch[160] Batch [399]	Speed: 8586.627538 samples/sec	accuracy=0.589420	lr=1.548251
Epoch[160] Batch [449]	Speed: 8615.200736 samples/sec	accuracy=0.588940	lr=1.547348
Epoch[160] Batch [499]	Speed: 8455.818046 samples/sec	accuracy=0.588679	lr=1.546444
Epoch[160] Batch [549]	Speed: 8651.953874 samples/sec	accuracy=0.588244	lr=1.545540
Epoch[160] Batch [599]	Speed: 8813.158501 samples/sec	accuracy=0.588016	lr=1.544636
[Epoch 160] training: accuracy=0.587740
[Epoch 160] speed: 8601 samples/sec	time cost: 153.360941
[Epoch 160] validation: err-top1=0.400513 err-top5=0.163615
Epoch[161] Batch [49]	Speed: 8306.105338 samples/sec	accuracy=0.596777	lr=1.543280
Epoch[161] Batch [99]	Speed: 8634.130982 samples/sec	accuracy=0.593901	lr=1.542376
Epoch[161] Batch [149]	Speed: 8654.452322 samples/sec	accuracy=0.593099	lr=1.541472
Epoch[161] Batch [199]	Speed: 8740.892426 samples/sec	accuracy=0.592092	lr=1.540568
Epoch[161] Batch [249]	Speed: 8595.037122 samples/sec	accuracy=0.590959	lr=1.539663
Epoch[161] Batch [299]	Speed: 8653.505319 samples/sec	accuracy=0.590596	lr=1.538758
Epoch[161] Batch [349]	Speed: 8468.411699 samples/sec	accuracy=0.589874	lr=1.537854
Epoch[161] Batch [399]	Speed: 8456.702125 samples/sec	accuracy=0.589547	lr=1.536949
Epoch[161] Batch [449]	Speed: 8753.049419 samples/sec	accuracy=0.588990	lr=1.536044
Epoch[161] Batch [499]	Speed: 8648.973697 samples/sec	accuracy=0.588599	lr=1.535139
Epoch[161] Batch [549]	Speed: 8314.780061 samples/sec	accuracy=0.588335	lr=1.534233
Epoch[161] Batch [599]	Speed: 8646.386330 samples/sec	accuracy=0.588069	lr=1.533328
[Epoch 161] training: accuracy=0.587873
[Epoch 161] speed: 8573 samples/sec	time cost: 154.320498
[Epoch 161] validation: err-top1=0.413125 err-top5=0.174609
Epoch[162] Batch [49]	Speed: 8780.515146 samples/sec	accuracy=0.598760	lr=1.531952
Epoch[162] Batch [99]	Speed: 8781.957898 samples/sec	accuracy=0.596582	lr=1.531046
Epoch[162] Batch [149]	Speed: 8679.883706 samples/sec	accuracy=0.596035	lr=1.530140
Epoch[162] Batch [199]	Speed: 8692.219942 samples/sec	accuracy=0.595110	lr=1.529234
Epoch[162] Batch [249]	Speed: 8702.636520 samples/sec	accuracy=0.593705	lr=1.528328
Epoch[162] Batch [299]	Speed: 8899.765511 samples/sec	accuracy=0.592603	lr=1.527422
Epoch[162] Batch [349]	Speed: 8681.735963 samples/sec	accuracy=0.591936	lr=1.526516
Epoch[162] Batch [399]	Speed: 8934.301898 samples/sec	accuracy=0.591407	lr=1.525610
Epoch[162] Batch [449]	Speed: 8644.137483 samples/sec	accuracy=0.590354	lr=1.524703
Epoch[162] Batch [499]	Speed: 8652.464046 samples/sec	accuracy=0.589935	lr=1.523797
Epoch[162] Batch [549]	Speed: 8685.344315 samples/sec	accuracy=0.589669	lr=1.522890
Epoch[162] Batch [599]	Speed: 8838.939722 samples/sec	accuracy=0.589500	lr=1.521983
[Epoch 162] training: accuracy=0.589305
[Epoch 162] speed: 8741 samples/sec	time cost: 150.902923
[Epoch 162] validation: err-top1=0.381856 err-top5=0.152832
Epoch[163] Batch [49]	Speed: 8746.639027 samples/sec	accuracy=0.598291	lr=1.520623
Epoch[163] Batch [99]	Speed: 8776.822815 samples/sec	accuracy=0.597935	lr=1.519716
Epoch[163] Batch [149]	Speed: 8695.055553 samples/sec	accuracy=0.596475	lr=1.518809
Epoch[163] Batch [199]	Speed: 8600.903496 samples/sec	accuracy=0.594646	lr=1.517901
Epoch[163] Batch [249]	Speed: 8909.767260 samples/sec	accuracy=0.593975	lr=1.516994
Epoch[163] Batch [299]	Speed: 8308.420223 samples/sec	accuracy=0.592992	lr=1.516086
Epoch[163] Batch [349]	Speed: 8545.089280 samples/sec	accuracy=0.592739	lr=1.515179
Epoch[163] Batch [399]	Speed: 8871.739910 samples/sec	accuracy=0.592109	lr=1.514271
Epoch[163] Batch [449]	Speed: 8530.269452 samples/sec	accuracy=0.591632	lr=1.513363
Epoch[163] Batch [499]	Speed: 8640.573711 samples/sec	accuracy=0.591326	lr=1.512455
Epoch[163] Batch [549]	Speed: 8754.740659 samples/sec	accuracy=0.590818	lr=1.511547
Epoch[163] Batch [599]	Speed: 8843.075225 samples/sec	accuracy=0.590278	lr=1.510639
[Epoch 163] training: accuracy=0.590058
[Epoch 163] speed: 8682 samples/sec	time cost: 152.379294
[Epoch 163] validation: err-top1=0.386719 err-top5=0.155599
Epoch[164] Batch [49]	Speed: 8699.730057 samples/sec	accuracy=0.598242	lr=1.509259
Epoch[164] Batch [99]	Speed: 8687.994070 samples/sec	accuracy=0.598340	lr=1.508350
Epoch[164] Batch [149]	Speed: 9019.779774 samples/sec	accuracy=0.596995	lr=1.507442
Epoch[164] Batch [199]	Speed: 8903.301790 samples/sec	accuracy=0.595713	lr=1.506533
Epoch[164] Batch [249]	Speed: 8688.833853 samples/sec	accuracy=0.594686	lr=1.505624
Epoch[164] Batch [299]	Speed: 9003.404038 samples/sec	accuracy=0.593800	lr=1.504716
Epoch[164] Batch [349]	Speed: 8955.521176 samples/sec	accuracy=0.593513	lr=1.503807
Epoch[164] Batch [399]	Speed: 8597.722409 samples/sec	accuracy=0.592802	lr=1.502898
Epoch[164] Batch [449]	Speed: 8712.500127 samples/sec	accuracy=0.591906	lr=1.501988
Epoch[164] Batch [499]	Speed: 8647.718822 samples/sec	accuracy=0.591676	lr=1.501079
Epoch[164] Batch [549]	Speed: 8391.198947 samples/sec	accuracy=0.591010	lr=1.500170
Epoch[164] Batch [599]	Speed: 8763.123711 samples/sec	accuracy=0.590634	lr=1.499260
[Epoch 164] training: accuracy=0.590539
[Epoch 164] speed: 8755 samples/sec	time cost: 151.436526
[Epoch 164] validation: err-top1=0.407012 err-top5=0.171230
Epoch[165] Batch [49]	Speed: 8625.526447 samples/sec	accuracy=0.599414	lr=1.497878
Epoch[165] Batch [99]	Speed: 8450.039789 samples/sec	accuracy=0.597314	lr=1.496968
Epoch[165] Batch [149]	Speed: 8535.613097 samples/sec	accuracy=0.596318	lr=1.496058
Epoch[165] Batch [199]	Speed: 8533.732285 samples/sec	accuracy=0.594858	lr=1.495149
Epoch[165] Batch [249]	Speed: 8540.217545 samples/sec	accuracy=0.594123	lr=1.494239
Epoch[165] Batch [299]	Speed: 8642.486441 samples/sec	accuracy=0.593215	lr=1.493328
Epoch[165] Batch [349]	Speed: 8763.287849 samples/sec	accuracy=0.592755	lr=1.492418
Epoch[165] Batch [399]	Speed: 8796.078070 samples/sec	accuracy=0.592227	lr=1.491508
Epoch[165] Batch [449]	Speed: 8686.340287 samples/sec	accuracy=0.591862	lr=1.490598
Epoch[165] Batch [499]	Speed: 8612.379141 samples/sec	accuracy=0.591441	lr=1.489687
Epoch[165] Batch [549]	Speed: 8719.672861 samples/sec	accuracy=0.591000	lr=1.488777
Epoch[165] Batch [599]	Speed: 8619.940778 samples/sec	accuracy=0.590593	lr=1.487866
[Epoch 165] training: accuracy=0.590440
[Epoch 165] speed: 8628 samples/sec	time cost: 153.169185
[Epoch 165] validation: err-top1=0.405172 err-top5=0.172852
Epoch[166] Batch [49]	Speed: 8584.670300 samples/sec	accuracy=0.599346	lr=1.486500
Epoch[166] Batch [99]	Speed: 8766.803449 samples/sec	accuracy=0.599087	lr=1.485589
Epoch[166] Batch [149]	Speed: 8552.583067 samples/sec	accuracy=0.597458	lr=1.484678
Epoch[166] Batch [199]	Speed: 8848.276664 samples/sec	accuracy=0.596921	lr=1.483767
Epoch[166] Batch [249]	Speed: 8764.965161 samples/sec	accuracy=0.596023	lr=1.482856
Epoch[166] Batch [299]	Speed: 8424.723903 samples/sec	accuracy=0.595208	lr=1.481945
Epoch[166] Batch [349]	Speed: 8650.710505 samples/sec	accuracy=0.594247	lr=1.481033
Epoch[166] Batch [399]	Speed: 8698.181195 samples/sec	accuracy=0.593536	lr=1.480122
Epoch[166] Batch [449]	Speed: 8581.156927 samples/sec	accuracy=0.593096	lr=1.479210
Epoch[166] Batch [499]	Speed: 8712.348136 samples/sec	accuracy=0.592507	lr=1.478299
Epoch[166] Batch [549]	Speed: 8713.033903 samples/sec	accuracy=0.592280	lr=1.477387
Epoch[166] Batch [599]	Speed: 8700.255573 samples/sec	accuracy=0.591760	lr=1.476475
[Epoch 166] training: accuracy=0.591678
[Epoch 166] speed: 8666 samples/sec	time cost: 152.701028
[Epoch 166] validation: err-top1=0.383965 err-top5=0.154355
Epoch[167] Batch [49]	Speed: 8378.948970 samples/sec	accuracy=0.600498	lr=1.475089
Epoch[167] Batch [99]	Speed: 8682.777623 samples/sec	accuracy=0.599429	lr=1.474177
Epoch[167] Batch [149]	Speed: 8704.472745 samples/sec	accuracy=0.598522	lr=1.473265
Epoch[167] Batch [199]	Speed: 8813.584770 samples/sec	accuracy=0.597017	lr=1.472353
Epoch[167] Batch [249]	Speed: 8627.399246 samples/sec	accuracy=0.596285	lr=1.471440
Epoch[167] Batch [299]	Speed: 8770.320438 samples/sec	accuracy=0.595213	lr=1.470528
Epoch[167] Batch [349]	Speed: 8572.236182 samples/sec	accuracy=0.594378	lr=1.469616
Epoch[167] Batch [399]	Speed: 8690.747436 samples/sec	accuracy=0.593762	lr=1.468703
Epoch[167] Batch [449]	Speed: 8625.286190 samples/sec	accuracy=0.593288	lr=1.467790
Epoch[167] Batch [499]	Speed: 8881.536943 samples/sec	accuracy=0.593124	lr=1.466878
Epoch[167] Batch [549]	Speed: 8400.803604 samples/sec	accuracy=0.592783	lr=1.465965
Epoch[167] Batch [599]	Speed: 8497.035483 samples/sec	accuracy=0.592313	lr=1.465052
[Epoch 167] training: accuracy=0.592294
[Epoch 167] speed: 8638 samples/sec	time cost: 152.721124
[Epoch 167] validation: err-top1=0.372986 err-top5=0.146322
Epoch[168] Batch [49]	Speed: 8238.969329 samples/sec	accuracy=0.601475	lr=1.463683
Epoch[168] Batch [99]	Speed: 8951.043496 samples/sec	accuracy=0.599873	lr=1.462770
Epoch[168] Batch [149]	Speed: 8849.562162 samples/sec	accuracy=0.598197	lr=1.461856
Epoch[168] Batch [199]	Speed: 8749.852140 samples/sec	accuracy=0.596938	lr=1.460943
Epoch[168] Batch [249]	Speed: 8894.846736 samples/sec	accuracy=0.595385	lr=1.460030
Epoch[168] Batch [299]	Speed: 8369.414905 samples/sec	accuracy=0.595399	lr=1.459116
Epoch[168] Batch [349]	Speed: 8638.114366 samples/sec	accuracy=0.594743	lr=1.458203
Epoch[168] Batch [399]	Speed: 8629.174729 samples/sec	accuracy=0.594135	lr=1.457289
Epoch[168] Batch [449]	Speed: 8888.122020 samples/sec	accuracy=0.593677	lr=1.456376
Epoch[168] Batch [499]	Speed: 8802.066327 samples/sec	accuracy=0.593296	lr=1.455462
Epoch[168] Batch [549]	Speed: 8706.771631 samples/sec	accuracy=0.592832	lr=1.454548
Epoch[168] Batch [599]	Speed: 8559.966833 samples/sec	accuracy=0.592753	lr=1.453634
[Epoch 168] training: accuracy=0.592684
[Epoch 168] speed: 8691 samples/sec	time cost: 152.011575
[Epoch 168] validation: err-top1=0.380432 err-top5=0.152710
Epoch[169] Batch [49]	Speed: 8677.559372 samples/sec	accuracy=0.603203	lr=1.452245
Epoch[169] Batch [99]	Speed: 8694.710725 samples/sec	accuracy=0.601328	lr=1.451331
Epoch[169] Batch [149]	Speed: 8646.887838 samples/sec	accuracy=0.599883	lr=1.450417
Epoch[169] Batch [199]	Speed: 8516.493969 samples/sec	accuracy=0.598579	lr=1.449503
Epoch[169] Batch [249]	Speed: 8559.778151 samples/sec	accuracy=0.597293	lr=1.448588
Epoch[169] Batch [299]	Speed: 8359.608811 samples/sec	accuracy=0.596530	lr=1.447674
Epoch[169] Batch [349]	Speed: 8731.814706 samples/sec	accuracy=0.595781	lr=1.446760
Epoch[169] Batch [399]	Speed: 8764.349888 samples/sec	accuracy=0.595272	lr=1.445845
Epoch[169] Batch [449]	Speed: 9005.646397 samples/sec	accuracy=0.594900	lr=1.444931
Epoch[169] Batch [499]	Speed: 8857.493975 samples/sec	accuracy=0.594225	lr=1.444016
Epoch[169] Batch [549]	Speed: 8390.554871 samples/sec	accuracy=0.593809	lr=1.443101
Epoch[169] Batch [599]	Speed: 8753.579968 samples/sec	accuracy=0.593249	lr=1.442186
[Epoch 169] training: accuracy=0.593129
[Epoch 169] speed: 8666 samples/sec	time cost: 152.501530
[Epoch 169] validation: err-top1=0.382695 err-top5=0.154590
Epoch[170] Batch [49]	Speed: 8447.477514 samples/sec	accuracy=0.603203	lr=1.440814
Epoch[170] Batch [99]	Speed: 8815.210107 samples/sec	accuracy=0.600142	lr=1.439899
Epoch[170] Batch [149]	Speed: 8449.840462 samples/sec	accuracy=0.599521	lr=1.438984
Epoch[170] Batch [199]	Speed: 8737.119407 samples/sec	accuracy=0.598000	lr=1.438069
Epoch[170] Batch [249]	Speed: 8815.112045 samples/sec	accuracy=0.596994	lr=1.437154
Epoch[170] Batch [299]	Speed: 8489.899801 samples/sec	accuracy=0.596322	lr=1.436238
Epoch[170] Batch [349]	Speed: 8525.609888 samples/sec	accuracy=0.596070	lr=1.435323
Epoch[170] Batch [399]	Speed: 8851.875576 samples/sec	accuracy=0.595548	lr=1.434408
Epoch[170] Batch [449]	Speed: 8678.286666 samples/sec	accuracy=0.595216	lr=1.433492
Epoch[170] Batch [499]	Speed: 8666.322330 samples/sec	accuracy=0.594814	lr=1.432577
Epoch[170] Batch [549]	Speed: 8750.529383 samples/sec	accuracy=0.594465	lr=1.431661
Epoch[170] Batch [599]	Speed: 8656.317990 samples/sec	accuracy=0.594189	lr=1.430746
[Epoch 170] training: accuracy=0.594103
[Epoch 170] speed: 8654 samples/sec	time cost: 152.746670
[Epoch 170] validation: err-top1=0.364848 err-top5=0.144226
Epoch[171] Batch [49]	Speed: 8644.837262 samples/sec	accuracy=0.603799	lr=1.429354
Epoch[171] Batch [99]	Speed: 8552.045950 samples/sec	accuracy=0.602280	lr=1.428438
Epoch[171] Batch [149]	Speed: 8982.144983 samples/sec	accuracy=0.601064	lr=1.427522
Epoch[171] Batch [199]	Speed: 8655.480119 samples/sec	accuracy=0.599758	lr=1.426606
Epoch[171] Batch [249]	Speed: 8562.589449 samples/sec	accuracy=0.598377	lr=1.425690
Epoch[171] Batch [299]	Speed: 8726.180462 samples/sec	accuracy=0.597669	lr=1.424774
Epoch[171] Batch [349]	Speed: 8863.088235 samples/sec	accuracy=0.596782	lr=1.423858
Epoch[171] Batch [399]	Speed: 8539.253270 samples/sec	accuracy=0.596005	lr=1.422942
Epoch[171] Batch [449]	Speed: 8608.313859 samples/sec	accuracy=0.595774	lr=1.422025
Epoch[171] Batch [499]	Speed: 8674.573446 samples/sec	accuracy=0.595368	lr=1.421109
Epoch[171] Batch [549]	Speed: 8716.564129 samples/sec	accuracy=0.594961	lr=1.420193
Epoch[171] Batch [599]	Speed: 8728.987006 samples/sec	accuracy=0.594644	lr=1.419276
[Epoch 171] training: accuracy=0.594524
[Epoch 171] speed: 8688 samples/sec	time cost: 152.519705
[Epoch 171] validation: err-top1=0.391270 err-top5=0.157129
Epoch[172] Batch [49]	Speed: 8464.215931 samples/sec	accuracy=0.605537	lr=1.417883
Epoch[172] Batch [99]	Speed: 8819.823286 samples/sec	accuracy=0.603481	lr=1.416966
Epoch[172] Batch [149]	Speed: 8630.648815 samples/sec	accuracy=0.602380	lr=1.416050
Epoch[172] Batch [199]	Speed: 8834.019069 samples/sec	accuracy=0.600886	lr=1.415133
Epoch[172] Batch [249]	Speed: 8612.954435 samples/sec	accuracy=0.600025	lr=1.414216
Epoch[172] Batch [299]	Speed: 8740.546266 samples/sec	accuracy=0.599425	lr=1.413299
Epoch[172] Batch [349]	Speed: 8561.405421 samples/sec	accuracy=0.598205	lr=1.412383
Epoch[172] Batch [399]	Speed: 8619.006155 samples/sec	accuracy=0.597939	lr=1.411466
Epoch[172] Batch [449]	Speed: 8664.718565 samples/sec	accuracy=0.597284	lr=1.410549
Epoch[172] Batch [499]	Speed: 8756.789078 samples/sec	accuracy=0.596724	lr=1.409632
Epoch[172] Batch [549]	Speed: 8375.697811 samples/sec	accuracy=0.596390	lr=1.408714
Epoch[172] Batch [599]	Speed: 8753.784963 samples/sec	accuracy=0.595745	lr=1.407797
[Epoch 172] training: accuracy=0.595552
[Epoch 172] speed: 8655 samples/sec	time cost: 152.487591
[Epoch 172] validation: err-top1=0.385925 err-top5=0.156555
Epoch[173] Batch [49]	Speed: 8825.961091 samples/sec	accuracy=0.604277	lr=1.406421
Epoch[173] Batch [99]	Speed: 8453.942445 samples/sec	accuracy=0.602393	lr=1.405504
Epoch[173] Batch [149]	Speed: 8743.129077 samples/sec	accuracy=0.600716	lr=1.404587
Epoch[173] Batch [199]	Speed: 8696.966771 samples/sec	accuracy=0.600339	lr=1.403669
Epoch[173] Batch [249]	Speed: 8825.411032 samples/sec	accuracy=0.599203	lr=1.402752
Epoch[173] Batch [299]	Speed: 8719.758897 samples/sec	accuracy=0.598211	lr=1.401834
Epoch[173] Batch [349]	Speed: 8634.272271 samples/sec	accuracy=0.597987	lr=1.400917
Epoch[173] Batch [399]	Speed: 8552.740264 samples/sec	accuracy=0.597651	lr=1.399999
Epoch[173] Batch [449]	Speed: 8729.987157 samples/sec	accuracy=0.597310	lr=1.399082
Epoch[173] Batch [499]	Speed: 8902.072039 samples/sec	accuracy=0.596756	lr=1.398164
Epoch[173] Batch [549]	Speed: 8667.559871 samples/sec	accuracy=0.596388	lr=1.397246
Epoch[173] Batch [599]	Speed: 8598.805464 samples/sec	accuracy=0.596252	lr=1.396328
[Epoch 173] training: accuracy=0.596093
[Epoch 173] speed: 8694 samples/sec	time cost: 152.450116
[Epoch 173] validation: err-top1=0.398945 err-top5=0.165137
Epoch[174] Batch [49]	Speed: 8691.540438 samples/sec	accuracy=0.601611	lr=1.394933
Epoch[174] Batch [99]	Speed: 8711.729978 samples/sec	accuracy=0.603506	lr=1.394015
Epoch[174] Batch [149]	Speed: 8577.047064 samples/sec	accuracy=0.601592	lr=1.393097
Epoch[174] Batch [199]	Speed: 8648.995468 samples/sec	accuracy=0.600984	lr=1.392179
Epoch[174] Batch [249]	Speed: 8934.817846 samples/sec	accuracy=0.600191	lr=1.391261
Epoch[174] Batch [299]	Speed: 8624.887986 samples/sec	accuracy=0.599463	lr=1.390343
Epoch[174] Batch [349]	Speed: 8925.791415 samples/sec	accuracy=0.599015	lr=1.389425
Epoch[174] Batch [399]	Speed: 8475.815868 samples/sec	accuracy=0.598193	lr=1.388507
Epoch[174] Batch [449]	Speed: 8769.860917 samples/sec	accuracy=0.597717	lr=1.387588
Epoch[174] Batch [499]	Speed: 8961.703514 samples/sec	accuracy=0.597423	lr=1.386670
Epoch[174] Batch [549]	Speed: 8659.971593 samples/sec	accuracy=0.597246	lr=1.385752
Epoch[174] Batch [599]	Speed: 8643.355368 samples/sec	accuracy=0.596635	lr=1.384834
[Epoch 174] training: accuracy=0.596434
[Epoch 174] speed: 8715 samples/sec	time cost: 151.580967
[Epoch 174] validation: err-top1=0.369405 err-top5=0.145040
Epoch[175] Batch [49]	Speed: 8690.626450 samples/sec	accuracy=0.606484	lr=1.383456
Epoch[175] Batch [99]	Speed: 8958.565964 samples/sec	accuracy=0.604731	lr=1.382537
Epoch[175] Batch [149]	Speed: 8716.842580 samples/sec	accuracy=0.603656	lr=1.381619
Epoch[175] Batch [199]	Speed: 8682.891194 samples/sec	accuracy=0.602280	lr=1.380700
Epoch[175] Batch [249]	Speed: 8848.228358 samples/sec	accuracy=0.601506	lr=1.379782
Epoch[175] Batch [299]	Speed: 8731.497488 samples/sec	accuracy=0.600990	lr=1.378863
Epoch[175] Batch [349]	Speed: 8361.699821 samples/sec	accuracy=0.600671	lr=1.377944
Epoch[175] Batch [399]	Speed: 8679.451855 samples/sec	accuracy=0.600144	lr=1.377026
Epoch[175] Batch [449]	Speed: 8767.156166 samples/sec	accuracy=0.599268	lr=1.376107
Epoch[175] Batch [499]	Speed: 8697.268980 samples/sec	accuracy=0.598851	lr=1.375188
Epoch[175] Batch [549]	Speed: 8994.617170 samples/sec	accuracy=0.598399	lr=1.374269
Epoch[175] Batch [599]	Speed: 8620.827673 samples/sec	accuracy=0.598036	lr=1.373350
[Epoch 175] training: accuracy=0.597627
[Epoch 175] speed: 8719 samples/sec	time cost: 151.760326
[Epoch 175] validation: err-top1=0.379883 err-top5=0.150004
Epoch[176] Batch [49]	Speed: 8926.946275 samples/sec	accuracy=0.608027	lr=1.371954
Epoch[176] Batch [99]	Speed: 8824.874822 samples/sec	accuracy=0.603877	lr=1.371035
Epoch[176] Batch [149]	Speed: 8825.452198 samples/sec	accuracy=0.603477	lr=1.370116
Epoch[176] Batch [199]	Speed: 8782.210194 samples/sec	accuracy=0.602383	lr=1.369197
Epoch[176] Batch [249]	Speed: 8857.864806 samples/sec	accuracy=0.601705	lr=1.368278
Epoch[176] Batch [299]	Speed: 8726.722476 samples/sec	accuracy=0.600918	lr=1.367358
Epoch[176] Batch [349]	Speed: 8645.922647 samples/sec	accuracy=0.600075	lr=1.366439
Epoch[176] Batch [399]	Speed: 8460.229794 samples/sec	accuracy=0.599423	lr=1.365520
Epoch[176] Batch [449]	Speed: 8431.953180 samples/sec	accuracy=0.598983	lr=1.364601
Epoch[176] Batch [499]	Speed: 8658.781080 samples/sec	accuracy=0.598394	lr=1.363682
Epoch[176] Batch [549]	Speed: 8614.660045 samples/sec	accuracy=0.597955	lr=1.362762
Epoch[176] Batch [599]	Speed: 8660.466819 samples/sec	accuracy=0.597848	lr=1.361843
[Epoch 176] training: accuracy=0.597816
[Epoch 176] speed: 8700 samples/sec	time cost: 152.193545
[Epoch 176] validation: err-top1=0.359941 err-top5=0.140684
Epoch[177] Batch [49]	Speed: 8555.098914 samples/sec	accuracy=0.610352	lr=1.360464
Epoch[177] Batch [99]	Speed: 8708.752280 samples/sec	accuracy=0.608320	lr=1.359545
Epoch[177] Batch [149]	Speed: 8672.654369 samples/sec	accuracy=0.606745	lr=1.358625
Epoch[177] Batch [199]	Speed: 8999.813268 samples/sec	accuracy=0.605154	lr=1.357706
Epoch[177] Batch [249]	Speed: 8956.007643 samples/sec	accuracy=0.603768	lr=1.356786
Epoch[177] Batch [299]	Speed: 8688.920161 samples/sec	accuracy=0.603117	lr=1.355867
Epoch[177] Batch [349]	Speed: 8774.235836 samples/sec	accuracy=0.602121	lr=1.354947
Epoch[177] Batch [399]	Speed: 8581.958005 samples/sec	accuracy=0.601243	lr=1.354028
Epoch[177] Batch [449]	Speed: 8530.949729 samples/sec	accuracy=0.600650	lr=1.353108
Epoch[177] Batch [499]	Speed: 8986.023783 samples/sec	accuracy=0.599980	lr=1.352189
Epoch[177] Batch [549]	Speed: 8862.655701 samples/sec	accuracy=0.599547	lr=1.351269
Epoch[177] Batch [599]	Speed: 8736.959802 samples/sec	accuracy=0.599211	lr=1.350349
[Epoch 177] training: accuracy=0.598988
[Epoch 177] speed: 8747 samples/sec	time cost: 151.231805
[Epoch 177] validation: err-top1=0.388875 err-top5=0.157328
Epoch[178] Batch [49]	Speed: 9015.695587 samples/sec	accuracy=0.611406	lr=1.348951
Epoch[178] Batch [99]	Speed: 8855.124861 samples/sec	accuracy=0.609927	lr=1.348032
Epoch[178] Batch [149]	Speed: 8636.001789 samples/sec	accuracy=0.608385	lr=1.347112
Epoch[178] Batch [199]	Speed: 8869.080566 samples/sec	accuracy=0.606321	lr=1.346192
Epoch[178] Batch [249]	Speed: 8756.292771 samples/sec	accuracy=0.604508	lr=1.345272
Epoch[178] Batch [299]	Speed: 8608.345778 samples/sec	accuracy=0.603452	lr=1.344353
Epoch[178] Batch [349]	Speed: 8857.781321 samples/sec	accuracy=0.602778	lr=1.343433
Epoch[178] Batch [399]	Speed: 8871.394120 samples/sec	accuracy=0.602083	lr=1.342513
Epoch[178] Batch [449]	Speed: 8573.481395 samples/sec	accuracy=0.601668	lr=1.341593
Epoch[178] Batch [499]	Speed: 8957.287648 samples/sec	accuracy=0.601147	lr=1.340673
Epoch[178] Batch [549]	Speed: 8788.303115 samples/sec	accuracy=0.600599	lr=1.339753
Epoch[178] Batch [599]	Speed: 8971.096757 samples/sec	accuracy=0.600226	lr=1.338833
[Epoch 178] training: accuracy=0.600081
[Epoch 178] speed: 8812 samples/sec	time cost: 150.224644
[Epoch 178] validation: err-top1=0.366426 err-top5=0.142051
Epoch[179] Batch [49]	Speed: 9003.743586 samples/sec	accuracy=0.611143	lr=1.337453
Epoch[179] Batch [99]	Speed: 9038.981427 samples/sec	accuracy=0.609551	lr=1.336533
Epoch[179] Batch [149]	Speed: 8996.784858 samples/sec	accuracy=0.607370	lr=1.335613
Epoch[179] Batch [199]	Speed: 9070.787180 samples/sec	accuracy=0.605833	lr=1.334693
Epoch[179] Batch [249]	Speed: 8590.445552 samples/sec	accuracy=0.604045	lr=1.333773
Epoch[179] Batch [299]	Speed: 8695.225248 samples/sec	accuracy=0.603216	lr=1.332853
Epoch[179] Batch [349]	Speed: 8822.857682 samples/sec	accuracy=0.602386	lr=1.331933
Epoch[179] Batch [399]	Speed: 8915.438685 samples/sec	accuracy=0.602123	lr=1.331013
Epoch[179] Batch [449]	Speed: 8713.142081 samples/sec	accuracy=0.601638	lr=1.330093
Epoch[179] Batch [499]	Speed: 8876.449517 samples/sec	accuracy=0.601122	lr=1.329173
Epoch[179] Batch [549]	Speed: 8762.227498 samples/sec	accuracy=0.600797	lr=1.328253
Epoch[179] Batch [599]	Speed: 8778.914779 samples/sec	accuracy=0.600488	lr=1.327333
[Epoch 179] training: accuracy=0.600325
[Epoch 179] speed: 8841 samples/sec	time cost: 149.771609
[Epoch 179] validation: err-top1=0.404785 err-top5=0.170268
Epoch[180] Batch [49]	Speed: 9084.981791 samples/sec	accuracy=0.609326	lr=1.325934
Epoch[180] Batch [99]	Speed: 9119.694741 samples/sec	accuracy=0.607876	lr=1.325014
Epoch[180] Batch [149]	Speed: 8643.435557 samples/sec	accuracy=0.606901	lr=1.324094
Epoch[180] Batch [199]	Speed: 8880.462474 samples/sec	accuracy=0.605061	lr=1.323173
Epoch[180] Batch [249]	Speed: 9020.559315 samples/sec	accuracy=0.604529	lr=1.322253
Epoch[180] Batch [299]	Speed: 8942.041981 samples/sec	accuracy=0.603674	lr=1.321333
Epoch[180] Batch [349]	Speed: 8778.299160 samples/sec	accuracy=0.603217	lr=1.320413
Epoch[180] Batch [399]	Speed: 8735.478851 samples/sec	accuracy=0.602454	lr=1.319492
Epoch[180] Batch [449]	Speed: 8666.647946 samples/sec	accuracy=0.602071	lr=1.318572
Epoch[180] Batch [499]	Speed: 8579.270387 samples/sec	accuracy=0.601646	lr=1.317652
Epoch[180] Batch [549]	Speed: 8940.697652 samples/sec	accuracy=0.601204	lr=1.316732
Epoch[180] Batch [599]	Speed: 8524.857365 samples/sec	accuracy=0.601021	lr=1.315811
[Epoch 180] training: accuracy=0.600896
[Epoch 180] speed: 8823 samples/sec	time cost: 150.229938
[Epoch 180] validation: err-top1=0.379333 err-top5=0.150614
Epoch[181] Batch [49]	Speed: 8939.358192 samples/sec	accuracy=0.610947	lr=1.314413
Epoch[181] Batch [99]	Speed: 9151.527406 samples/sec	accuracy=0.608955	lr=1.313492
Epoch[181] Batch [149]	Speed: 8634.462342 samples/sec	accuracy=0.607223	lr=1.312572
Epoch[181] Batch [199]	Speed: 8690.532019 samples/sec	accuracy=0.605828	lr=1.311652
Epoch[181] Batch [249]	Speed: 8776.073711 samples/sec	accuracy=0.604416	lr=1.310731
Epoch[181] Batch [299]	Speed: 8935.831515 samples/sec	accuracy=0.603778	lr=1.309811
Epoch[181] Batch [349]	Speed: 8651.597817 samples/sec	accuracy=0.603437	lr=1.308891
Epoch[181] Batch [399]	Speed: 8792.180955 samples/sec	accuracy=0.602860	lr=1.307970
Epoch[181] Batch [449]	Speed: 8897.757311 samples/sec	accuracy=0.602611	lr=1.307050
Epoch[181] Batch [499]	Speed: 8778.521103 samples/sec	accuracy=0.602158	lr=1.306130
Epoch[181] Batch [549]	Speed: 8599.638768 samples/sec	accuracy=0.601843	lr=1.305209
Epoch[181] Batch [599]	Speed: 8726.084548 samples/sec	accuracy=0.601583	lr=1.304289
[Epoch 181] training: accuracy=0.601456
[Epoch 181] speed: 8799 samples/sec	time cost: 150.532238
[Epoch 181] validation: err-top1=0.374141 err-top5=0.145254
Epoch[182] Batch [49]	Speed: 9018.780490 samples/sec	accuracy=0.612139	lr=1.302908
Epoch[182] Batch [99]	Speed: 8864.346573 samples/sec	accuracy=0.610933	lr=1.301988
Epoch[182] Batch [149]	Speed: 8682.797634 samples/sec	accuracy=0.608981	lr=1.301068
Epoch[182] Batch [199]	Speed: 8912.358223 samples/sec	accuracy=0.608237	lr=1.300147
Epoch[182] Batch [249]	Speed: 8765.800212 samples/sec	accuracy=0.606984	lr=1.299227
Epoch[182] Batch [299]	Speed: 8773.499898 samples/sec	accuracy=0.605939	lr=1.298307
Epoch[182] Batch [349]	Speed: 8517.162761 samples/sec	accuracy=0.605232	lr=1.297386
Epoch[182] Batch [399]	Speed: 8441.190683 samples/sec	accuracy=0.604943	lr=1.296466
Epoch[182] Batch [449]	Speed: 8661.772212 samples/sec	accuracy=0.604563	lr=1.295545
Epoch[182] Batch [499]	Speed: 9039.993374 samples/sec	accuracy=0.604366	lr=1.294625
Epoch[182] Batch [549]	Speed: 8929.496372 samples/sec	accuracy=0.603972	lr=1.293705
Epoch[182] Batch [599]	Speed: 8972.243503 samples/sec	accuracy=0.603388	lr=1.292784
[Epoch 182] training: accuracy=0.603201
[Epoch 182] speed: 8787 samples/sec	time cost: 150.687355
[Epoch 182] validation: err-top1=0.367086 err-top5=0.143351
Epoch[183] Batch [49]	Speed: 8831.590942 samples/sec	accuracy=0.610605	lr=1.291386
Epoch[183] Batch [99]	Speed: 8984.785924 samples/sec	accuracy=0.611011	lr=1.290465
Epoch[183] Batch [149]	Speed: 8973.922264 samples/sec	accuracy=0.609635	lr=1.289545
Epoch[183] Batch [199]	Speed: 8871.431684 samples/sec	accuracy=0.607883	lr=1.288625
Epoch[183] Batch [249]	Speed: 8770.886219 samples/sec	accuracy=0.607113	lr=1.287704
Epoch[183] Batch [299]	Speed: 8954.860189 samples/sec	accuracy=0.606227	lr=1.286784
Epoch[183] Batch [349]	Speed: 8741.615431 samples/sec	accuracy=0.605144	lr=1.285864
Epoch[183] Batch [399]	Speed: 8604.596480 samples/sec	accuracy=0.604729	lr=1.284943
Epoch[183] Batch [449]	Speed: 8540.658239 samples/sec	accuracy=0.604295	lr=1.284023
Epoch[183] Batch [499]	Speed: 8671.309630 samples/sec	accuracy=0.603567	lr=1.283103
Epoch[183] Batch [549]	Speed: 8709.043300 samples/sec	accuracy=0.603080	lr=1.282182
Epoch[183] Batch [599]	Speed: 8672.163001 samples/sec	accuracy=0.603032	lr=1.281262
[Epoch 183] training: accuracy=0.602709
[Epoch 183] speed: 8766 samples/sec	time cost: 150.882775
[Epoch 183] validation: err-top1=0.396113 err-top5=0.160977
Epoch[184] Batch [49]	Speed: 9138.251232 samples/sec	accuracy=0.614131	lr=1.279882
Epoch[184] Batch [99]	Speed: 9156.607744 samples/sec	accuracy=0.611890	lr=1.278962
Epoch[184] Batch [149]	Speed: 8824.198714 samples/sec	accuracy=0.609938	lr=1.278041
Epoch[184] Batch [199]	Speed: 8622.175493 samples/sec	accuracy=0.608655	lr=1.277121
Epoch[184] Batch [249]	Speed: 8556.351256 samples/sec	accuracy=0.607430	lr=1.276201
Epoch[184] Batch [299]	Speed: 8801.626020 samples/sec	accuracy=0.606781	lr=1.275281
Epoch[184] Batch [349]	Speed: 8929.822756 samples/sec	accuracy=0.606378	lr=1.274360
Epoch[184] Batch [399]	Speed: 8755.314070 samples/sec	accuracy=0.606288	lr=1.273440
Epoch[184] Batch [449]	Speed: 8740.881931 samples/sec	accuracy=0.606033	lr=1.272520
Epoch[184] Batch [499]	Speed: 8891.286061 samples/sec	accuracy=0.605301	lr=1.271600
Epoch[184] Batch [549]	Speed: 8838.419327 samples/sec	accuracy=0.604705	lr=1.270680
Epoch[184] Batch [599]	Speed: 8845.104172 samples/sec	accuracy=0.604351	lr=1.269760
[Epoch 184] training: accuracy=0.604294
[Epoch 184] speed: 8840 samples/sec	time cost: 149.585942
[Epoch 184] validation: err-top1=0.382548 err-top5=0.155151
Epoch[185] Batch [49]	Speed: 9057.935664 samples/sec	accuracy=0.616953	lr=1.268361
Epoch[185] Batch [99]	Speed: 8831.321999 samples/sec	accuracy=0.613125	lr=1.267441
Epoch[185] Batch [149]	Speed: 8369.820532 samples/sec	accuracy=0.611442	lr=1.266521
Epoch[185] Batch [199]	Speed: 8583.236406 samples/sec	accuracy=0.610015	lr=1.265601
Epoch[185] Batch [249]	Speed: 8705.906318 samples/sec	accuracy=0.609389	lr=1.264681
Epoch[185] Batch [299]	Speed: 8544.392637 samples/sec	accuracy=0.608408	lr=1.263761
Epoch[185] Batch [349]	Speed: 8795.509034 samples/sec	accuracy=0.608025	lr=1.262841
Epoch[185] Batch [399]	Speed: 8613.754207 samples/sec	accuracy=0.607653	lr=1.261921
Epoch[185] Batch [449]	Speed: 8623.701731 samples/sec	accuracy=0.606913	lr=1.261001
Epoch[185] Batch [499]	Speed: 8894.656449 samples/sec	accuracy=0.606178	lr=1.260081
Epoch[185] Batch [549]	Speed: 8711.651345 samples/sec	accuracy=0.605811	lr=1.259161
Epoch[185] Batch [599]	Speed: 8551.091771 samples/sec	accuracy=0.605251	lr=1.258241
[Epoch 185] training: accuracy=0.605088
[Epoch 185] speed: 8692 samples/sec	time cost: 152.113802
[Epoch 185] validation: err-top1=0.357910 err-top5=0.139141
Epoch[186] Batch [49]	Speed: 8259.225459 samples/sec	accuracy=0.615898	lr=1.256862
Epoch[186] Batch [99]	Speed: 8936.682335 samples/sec	accuracy=0.613682	lr=1.255942
Epoch[186] Batch [149]	Speed: 8656.063280 samples/sec	accuracy=0.612568	lr=1.255022
Epoch[186] Batch [199]	Speed: 8478.738648 samples/sec	accuracy=0.612002	lr=1.254102
Epoch[186] Batch [249]	Speed: 8705.918142 samples/sec	accuracy=0.610314	lr=1.253182
Epoch[186] Batch [299]	Speed: 8886.582398 samples/sec	accuracy=0.609450	lr=1.252263
Epoch[186] Batch [349]	Speed: 8809.143793 samples/sec	accuracy=0.608763	lr=1.251343
Epoch[186] Batch [399]	Speed: 8685.238057 samples/sec	accuracy=0.608132	lr=1.250423
Epoch[186] Batch [449]	Speed: 8827.886022 samples/sec	accuracy=0.607804	lr=1.249503
Epoch[186] Batch [499]	Speed: 8861.556542 samples/sec	accuracy=0.606961	lr=1.248584
Epoch[186] Batch [549]	Speed: 8990.220755 samples/sec	accuracy=0.606338	lr=1.247664
Epoch[186] Batch [599]	Speed: 8417.895866 samples/sec	accuracy=0.606134	lr=1.246745
[Epoch 186] training: accuracy=0.605941
[Epoch 186] speed: 8697 samples/sec	time cost: 152.114843
[Epoch 186] validation: err-top1=0.383301 err-top5=0.153585
Epoch[187] Batch [49]	Speed: 8447.586840 samples/sec	accuracy=0.616768	lr=1.245347
Epoch[187] Batch [99]	Speed: 9058.940586 samples/sec	accuracy=0.614380	lr=1.244427
Epoch[187] Batch [149]	Speed: 8770.597678 samples/sec	accuracy=0.612598	lr=1.243508
Epoch[187] Batch [199]	Speed: 9021.011379 samples/sec	accuracy=0.610779	lr=1.242588
Epoch[187] Batch [249]	Speed: 8593.716343 samples/sec	accuracy=0.609701	lr=1.241669
Epoch[187] Batch [299]	Speed: 8682.789384 samples/sec	accuracy=0.609336	lr=1.240749
Epoch[187] Batch [349]	Speed: 8622.885570 samples/sec	accuracy=0.608537	lr=1.239830
Epoch[187] Batch [399]	Speed: 8514.126850 samples/sec	accuracy=0.607633	lr=1.238911
Epoch[187] Batch [449]	Speed: 8917.130320 samples/sec	accuracy=0.607241	lr=1.237991
Epoch[187] Batch [499]	Speed: 8805.492882 samples/sec	accuracy=0.607005	lr=1.237072
Epoch[187] Batch [549]	Speed: 8642.603830 samples/sec	accuracy=0.606767	lr=1.236153
Epoch[187] Batch [599]	Speed: 8952.582025 samples/sec	accuracy=0.606356	lr=1.235234
[Epoch 187] training: accuracy=0.606057
[Epoch 187] speed: 8747 samples/sec	time cost: 151.386666
[Epoch 187] validation: err-top1=0.409505 err-top5=0.173462
Epoch[188] Batch [49]	Speed: 8524.858888 samples/sec	accuracy=0.615225	lr=1.233836
Epoch[188] Batch [99]	Speed: 8633.201435 samples/sec	accuracy=0.613037	lr=1.232917
Epoch[188] Batch [149]	Speed: 8843.313566 samples/sec	accuracy=0.612354	lr=1.231998
Epoch[188] Batch [199]	Speed: 8773.787018 samples/sec	accuracy=0.611514	lr=1.231079
Epoch[188] Batch [249]	Speed: 8561.079815 samples/sec	accuracy=0.611203	lr=1.230160
Epoch[188] Batch [299]	Speed: 8596.139629 samples/sec	accuracy=0.610801	lr=1.229241
Epoch[188] Batch [349]	Speed: 8705.498341 samples/sec	accuracy=0.609967	lr=1.228322
Epoch[188] Batch [399]	Speed: 8784.123632 samples/sec	accuracy=0.609636	lr=1.227403
Epoch[188] Batch [449]	Speed: 8702.716401 samples/sec	accuracy=0.608722	lr=1.226484
Epoch[188] Batch [499]	Speed: 8737.287726 samples/sec	accuracy=0.608105	lr=1.225565
Epoch[188] Batch [549]	Speed: 8758.396391 samples/sec	accuracy=0.607613	lr=1.224647
Epoch[188] Batch [599]	Speed: 8774.356832 samples/sec	accuracy=0.607122	lr=1.223728
[Epoch 188] training: accuracy=0.606887
[Epoch 188] speed: 8692 samples/sec	time cost: 152.159846
[Epoch 188] validation: err-top1=0.364707 err-top5=0.143379
Epoch[189] Batch [49]	Speed: 8533.789256 samples/sec	accuracy=0.615049	lr=1.222350
Epoch[189] Batch [99]	Speed: 8817.699654 samples/sec	accuracy=0.612490	lr=1.221431
Epoch[189] Batch [149]	Speed: 8837.559655 samples/sec	accuracy=0.612845	lr=1.220512
Epoch[189] Batch [199]	Speed: 8828.169817 samples/sec	accuracy=0.612161	lr=1.219594
Epoch[189] Batch [249]	Speed: 8572.472979 samples/sec	accuracy=0.610879	lr=1.218675
Epoch[189] Batch [299]	Speed: 8687.777559 samples/sec	accuracy=0.609526	lr=1.217757
Epoch[189] Batch [349]	Speed: 8683.583216 samples/sec	accuracy=0.609417	lr=1.216838
Epoch[189] Batch [399]	Speed: 8880.936229 samples/sec	accuracy=0.608817	lr=1.215920
Epoch[189] Batch [449]	Speed: 8649.771462 samples/sec	accuracy=0.608250	lr=1.215001
Epoch[189] Batch [499]	Speed: 8722.699486 samples/sec	accuracy=0.607991	lr=1.214083
Epoch[189] Batch [549]	Speed: 8738.395741 samples/sec	accuracy=0.607555	lr=1.213164
Epoch[189] Batch [599]	Speed: 8641.342804 samples/sec	accuracy=0.607182	lr=1.212246
[Epoch 189] training: accuracy=0.607103
[Epoch 189] speed: 8713 samples/sec	time cost: 151.825651
[Epoch 189] validation: err-top1=0.361003 err-top5=0.141378
Epoch[190] Batch [49]	Speed: 8457.240655 samples/sec	accuracy=0.617207	lr=1.210851
Epoch[190] Batch [99]	Speed: 8641.081847 samples/sec	accuracy=0.616538	lr=1.209932
Epoch[190] Batch [149]	Speed: 8742.036231 samples/sec	accuracy=0.615501	lr=1.209014
Epoch[190] Batch [199]	Speed: 8723.865468 samples/sec	accuracy=0.614565	lr=1.208096
Epoch[190] Batch [249]	Speed: 8942.394232 samples/sec	accuracy=0.613158	lr=1.207178
Epoch[190] Batch [299]	Speed: 8735.116775 samples/sec	accuracy=0.612407	lr=1.206260
Epoch[190] Batch [349]	Speed: 8621.570064 samples/sec	accuracy=0.611503	lr=1.205342
Epoch[190] Batch [399]	Speed: 8658.868188 samples/sec	accuracy=0.610736	lr=1.204424
Epoch[190] Batch [449]	Speed: 8999.468359 samples/sec	accuracy=0.610077	lr=1.203506
Epoch[190] Batch [499]	Speed: 8612.621097 samples/sec	accuracy=0.609743	lr=1.202589
Epoch[190] Batch [549]	Speed: 9036.175264 samples/sec	accuracy=0.609550	lr=1.201671
Epoch[190] Batch [599]	Speed: 8522.557299 samples/sec	accuracy=0.609229	lr=1.200753
[Epoch 190] training: accuracy=0.609018
[Epoch 190] speed: 8723 samples/sec	time cost: 151.570054
[Epoch 190] validation: err-top1=0.368340 err-top5=0.145137
Epoch[191] Batch [49]	Speed: 8538.615124 samples/sec	accuracy=0.619521	lr=1.199377
Epoch[191] Batch [99]	Speed: 8567.183750 samples/sec	accuracy=0.616255	lr=1.198459
Epoch[191] Batch [149]	Speed: 8855.994162 samples/sec	accuracy=0.615049	lr=1.197542
Epoch[191] Batch [199]	Speed: 8599.332458 samples/sec	accuracy=0.613887	lr=1.196624
Epoch[191] Batch [249]	Speed: 8832.061677 samples/sec	accuracy=0.612680	lr=1.195707
Epoch[191] Batch [299]	Speed: 8345.508664 samples/sec	accuracy=0.612095	lr=1.194789
Epoch[191] Batch [349]	Speed: 8964.443773 samples/sec	accuracy=0.611498	lr=1.193872
Epoch[191] Batch [399]	Speed: 8675.508945 samples/sec	accuracy=0.611144	lr=1.192955
Epoch[191] Batch [449]	Speed: 8488.810952 samples/sec	accuracy=0.610811	lr=1.192038
Epoch[191] Batch [499]	Speed: 8785.898613 samples/sec	accuracy=0.610565	lr=1.191120
Epoch[191] Batch [549]	Speed: 8622.064370 samples/sec	accuracy=0.610213	lr=1.190203
Epoch[191] Batch [599]	Speed: 8421.925929 samples/sec	accuracy=0.609782	lr=1.189286
[Epoch 191] training: accuracy=0.609538
[Epoch 191] speed: 8640 samples/sec	time cost: 153.039089
[Epoch 191] validation: err-top1=0.382141 err-top5=0.156331
Epoch[192] Batch [49]	Speed: 8998.294660 samples/sec	accuracy=0.619902	lr=1.187893
Epoch[192] Batch [99]	Speed: 8987.253333 samples/sec	accuracy=0.617983	lr=1.186976
Epoch[192] Batch [149]	Speed: 8706.706325 samples/sec	accuracy=0.615951	lr=1.186059
Epoch[192] Batch [199]	Speed: 8859.317740 samples/sec	accuracy=0.614585	lr=1.185142
Epoch[192] Batch [249]	Speed: 8991.486277 samples/sec	accuracy=0.613115	lr=1.184225
Epoch[192] Batch [299]	Speed: 8615.335530 samples/sec	accuracy=0.612349	lr=1.183309
Epoch[192] Batch [349]	Speed: 8708.073189 samples/sec	accuracy=0.611565	lr=1.182392
Epoch[192] Batch [399]	Speed: 8574.684858 samples/sec	accuracy=0.611144	lr=1.181475
Epoch[192] Batch [449]	Speed: 8849.291394 samples/sec	accuracy=0.610968	lr=1.180559
Epoch[192] Batch [499]	Speed: 8566.356723 samples/sec	accuracy=0.610604	lr=1.179643
Epoch[192] Batch [549]	Speed: 8837.627485 samples/sec	accuracy=0.610165	lr=1.178726
Epoch[192] Batch [599]	Speed: 8753.839558 samples/sec	accuracy=0.609848	lr=1.177810
[Epoch 192] training: accuracy=0.609634
[Epoch 192] speed: 8784 samples/sec	time cost: 150.476710
[Epoch 192] validation: err-top1=0.371623 err-top5=0.147034
Epoch[193] Batch [49]	Speed: 8689.282812 samples/sec	accuracy=0.618330	lr=1.176435
Epoch[193] Batch [99]	Speed: 8628.444548 samples/sec	accuracy=0.617964	lr=1.175519
Epoch[193] Batch [149]	Speed: 8921.901414 samples/sec	accuracy=0.615837	lr=1.174603
Epoch[193] Batch [199]	Speed: 8777.349614 samples/sec	accuracy=0.615808	lr=1.173687
Epoch[193] Batch [249]	Speed: 8725.311996 samples/sec	accuracy=0.615150	lr=1.172771
Epoch[193] Batch [299]	Speed: 8794.086856 samples/sec	accuracy=0.614326	lr=1.171855
Epoch[193] Batch [349]	Speed: 8991.648916 samples/sec	accuracy=0.613881	lr=1.170939
Epoch[193] Batch [399]	Speed: 8886.872369 samples/sec	accuracy=0.613196	lr=1.170024
Epoch[193] Batch [449]	Speed: 8757.564179 samples/sec	accuracy=0.612588	lr=1.169108
Epoch[193] Batch [499]	Speed: 8539.855003 samples/sec	accuracy=0.612196	lr=1.168192
Epoch[193] Batch [549]	Speed: 8985.894436 samples/sec	accuracy=0.611616	lr=1.167277
Epoch[193] Batch [599]	Speed: 8748.055877 samples/sec	accuracy=0.611174	lr=1.166361
[Epoch 193] training: accuracy=0.610926
[Epoch 193] speed: 8785 samples/sec	time cost: 150.990655
[Epoch 193] validation: err-top1=0.364727 err-top5=0.140137
Epoch[194] Batch [49]	Speed: 9003.703194 samples/sec	accuracy=0.621025	lr=1.164970
Epoch[194] Batch [99]	Speed: 9204.911410 samples/sec	accuracy=0.618643	lr=1.164054
Epoch[194] Batch [149]	Speed: 8937.280571 samples/sec	accuracy=0.617995	lr=1.163139
Epoch[194] Batch [199]	Speed: 9001.590089 samples/sec	accuracy=0.615657	lr=1.162224
Epoch[194] Batch [249]	Speed: 8749.694209 samples/sec	accuracy=0.614691	lr=1.161309
Epoch[194] Batch [299]	Speed: 8704.890858 samples/sec	accuracy=0.614097	lr=1.160394
Epoch[194] Batch [349]	Speed: 8775.773532 samples/sec	accuracy=0.613330	lr=1.159479
Epoch[194] Batch [399]	Speed: 8754.179456 samples/sec	accuracy=0.612832	lr=1.158564
Epoch[194] Batch [449]	Speed: 8512.058956 samples/sec	accuracy=0.612081	lr=1.157649
Epoch[194] Batch [499]	Speed: 8776.905678 samples/sec	accuracy=0.611717	lr=1.156734
Epoch[194] Batch [549]	Speed: 8787.929275 samples/sec	accuracy=0.611157	lr=1.155819
Epoch[194] Batch [599]	Speed: 8934.443703 samples/sec	accuracy=0.610728	lr=1.154905
[Epoch 194] training: accuracy=0.610535
[Epoch 194] speed: 8828 samples/sec	time cost: 149.922677
[Epoch 194] validation: err-top1=0.372762 err-top5=0.145528
Epoch[195] Batch [49]	Speed: 8689.335902 samples/sec	accuracy=0.620322	lr=1.153515
Epoch[195] Batch [99]	Speed: 9026.455231 samples/sec	accuracy=0.620059	lr=1.152600
Epoch[195] Batch [149]	Speed: 8811.163162 samples/sec	accuracy=0.618281	lr=1.151686
Epoch[195] Batch [199]	Speed: 8739.271437 samples/sec	accuracy=0.617178	lr=1.150772
Epoch[195] Batch [249]	Speed: 8904.102308 samples/sec	accuracy=0.616029	lr=1.149857
Epoch[195] Batch [299]	Speed: 8752.249077 samples/sec	accuracy=0.615534	lr=1.148943
Epoch[195] Batch [349]	Speed: 8947.541894 samples/sec	accuracy=0.614870	lr=1.148029
Epoch[195] Batch [399]	Speed: 8973.306552 samples/sec	accuracy=0.614186	lr=1.147115
Epoch[195] Batch [449]	Speed: 8873.609696 samples/sec	accuracy=0.613418	lr=1.146201
Epoch[195] Batch [499]	Speed: 8577.439321 samples/sec	accuracy=0.613003	lr=1.145287
Epoch[195] Batch [549]	Speed: 8930.250916 samples/sec	accuracy=0.612634	lr=1.144373
Epoch[195] Batch [599]	Speed: 8580.904391 samples/sec	accuracy=0.612074	lr=1.143460
[Epoch 195] training: accuracy=0.612042
[Epoch 195] speed: 8815 samples/sec	time cost: 150.174911
[Epoch 195] validation: err-top1=0.416992 err-top5=0.183047
Epoch[196] Batch [49]	Speed: 8872.262769 samples/sec	accuracy=0.620371	lr=1.142089
Epoch[196] Batch [99]	Speed: 9130.229625 samples/sec	accuracy=0.619717	lr=1.141176
Epoch[196] Batch [149]	Speed: 8711.271983 samples/sec	accuracy=0.618949	lr=1.140262
Epoch[196] Batch [199]	Speed: 8551.278197 samples/sec	accuracy=0.618264	lr=1.139349
Epoch[196] Batch [249]	Speed: 8664.329995 samples/sec	accuracy=0.616984	lr=1.138436
Epoch[196] Batch [299]	Speed: 8778.782353 samples/sec	accuracy=0.616240	lr=1.137523
Epoch[196] Batch [349]	Speed: 8784.296283 samples/sec	accuracy=0.615745	lr=1.136610
Epoch[196] Batch [399]	Speed: 8829.702333 samples/sec	accuracy=0.615217	lr=1.135697
Epoch[196] Batch [449]	Speed: 8626.636791 samples/sec	accuracy=0.614586	lr=1.134784
Epoch[196] Batch [499]	Speed: 8766.388492 samples/sec	accuracy=0.613977	lr=1.133871
Epoch[196] Batch [549]	Speed: 8975.626039 samples/sec	accuracy=0.613801	lr=1.132958
Epoch[196] Batch [599]	Speed: 8630.349310 samples/sec	accuracy=0.613561	lr=1.132045
[Epoch 196] training: accuracy=0.613387
[Epoch 196] speed: 8765 samples/sec	time cost: 150.941175
[Epoch 196] validation: err-top1=0.353984 err-top5=0.135050
Epoch[197] Batch [49]	Speed: 8713.253973 samples/sec	accuracy=0.625039	lr=1.130658
Epoch[197] Batch [99]	Speed: 8822.158143 samples/sec	accuracy=0.622495	lr=1.129746
Epoch[197] Batch [149]	Speed: 8569.386232 samples/sec	accuracy=0.620957	lr=1.128833
Epoch[197] Batch [199]	Speed: 8720.878053 samples/sec	accuracy=0.619905	lr=1.127921
Epoch[197] Batch [249]	Speed: 8975.502430 samples/sec	accuracy=0.618783	lr=1.127009
Epoch[197] Batch [299]	Speed: 8504.406288 samples/sec	accuracy=0.617640	lr=1.126097
Epoch[197] Batch [349]	Speed: 8661.008211 samples/sec	accuracy=0.617172	lr=1.125185
Epoch[197] Batch [399]	Speed: 8920.450485 samples/sec	accuracy=0.616737	lr=1.124273
Epoch[197] Batch [449]	Speed: 8769.087935 samples/sec	accuracy=0.616083	lr=1.123361
Epoch[197] Batch [499]	Speed: 8721.381687 samples/sec	accuracy=0.615140	lr=1.122449
Epoch[197] Batch [549]	Speed: 8643.733883 samples/sec	accuracy=0.614514	lr=1.121537
Epoch[197] Batch [599]	Speed: 8742.954303 samples/sec	accuracy=0.614006	lr=1.120626
[Epoch 197] training: accuracy=0.613847
[Epoch 197] speed: 8730 samples/sec	time cost: 151.203676
[Epoch 197] validation: err-top1=0.353577 err-top5=0.133199
Epoch[198] Batch [49]	Speed: 9008.351248 samples/sec	accuracy=0.625693	lr=1.119258
Epoch[198] Batch [99]	Speed: 9009.757203 samples/sec	accuracy=0.623940	lr=1.118347
Epoch[198] Batch [149]	Speed: 8858.097003 samples/sec	accuracy=0.622246	lr=1.117436
Epoch[198] Batch [199]	Speed: 8805.324812 samples/sec	accuracy=0.620344	lr=1.116525
Epoch[198] Batch [249]	Speed: 8694.479095 samples/sec	accuracy=0.619314	lr=1.115614
Epoch[198] Batch [299]	Speed: 8364.052481 samples/sec	accuracy=0.618577	lr=1.114703
Epoch[198] Batch [349]	Speed: 8799.468053 samples/sec	accuracy=0.617937	lr=1.113792
Epoch[198] Batch [399]	Speed: 8757.211519 samples/sec	accuracy=0.616971	lr=1.112881
Epoch[198] Batch [449]	Speed: 8811.255894 samples/sec	accuracy=0.616604	lr=1.111970
Epoch[198] Batch [499]	Speed: 8960.206525 samples/sec	accuracy=0.616129	lr=1.111059
Epoch[198] Batch [549]	Speed: 8981.302763 samples/sec	accuracy=0.615756	lr=1.110149
Epoch[198] Batch [599]	Speed: 8853.150441 samples/sec	accuracy=0.615380	lr=1.109238
[Epoch 198] training: accuracy=0.615173
[Epoch 198] speed: 8819 samples/sec	time cost: 150.256077
[Epoch 198] validation: err-top1=0.360000 err-top5=0.137324
Epoch[199] Batch [49]	Speed: 8511.205768 samples/sec	accuracy=0.620889	lr=1.107855
Epoch[199] Batch [99]	Speed: 8760.890223 samples/sec	accuracy=0.620732	lr=1.106945
Epoch[199] Batch [149]	Speed: 8610.249270 samples/sec	accuracy=0.620645	lr=1.106034
Epoch[199] Batch [199]	Speed: 8648.117048 samples/sec	accuracy=0.619526	lr=1.105124
Epoch[199] Batch [249]	Speed: 8723.192878 samples/sec	accuracy=0.618971	lr=1.104215
Epoch[199] Batch [299]	Speed: 8736.213044 samples/sec	accuracy=0.618735	lr=1.103305
Epoch[199] Batch [349]	Speed: 8624.604640 samples/sec	accuracy=0.617969	lr=1.102395
Epoch[199] Batch [399]	Speed: 8617.298824 samples/sec	accuracy=0.617354	lr=1.101485
Epoch[199] Batch [449]	Speed: 8769.996834 samples/sec	accuracy=0.616939	lr=1.100576
Epoch[199] Batch [499]	Speed: 8623.587106 samples/sec	accuracy=0.616689	lr=1.099666
Epoch[199] Batch [549]	Speed: 9000.112939 samples/sec	accuracy=0.616159	lr=1.098757
Epoch[199] Batch [599]	Speed: 8912.634344 samples/sec	accuracy=0.615824	lr=1.097848
[Epoch 199] training: accuracy=0.615577
[Epoch 199] speed: 8712 samples/sec	time cost: 151.544420
[Epoch 199] validation: err-top1=0.388102 err-top5=0.158488
Epoch[200] Batch [49]	Speed: 8461.959638 samples/sec	accuracy=0.626465	lr=1.096484
Epoch[200] Batch [99]	Speed: 8819.466862 samples/sec	accuracy=0.624321	lr=1.095575
Epoch[200] Batch [149]	Speed: 8904.316074 samples/sec	accuracy=0.622591	lr=1.094666
Epoch[200] Batch [199]	Speed: 8726.132770 samples/sec	accuracy=0.621877	lr=1.093758
Epoch[200] Batch [249]	Speed: 8861.915462 samples/sec	accuracy=0.620975	lr=1.092849
Epoch[200] Batch [299]	Speed: 8637.760316 samples/sec	accuracy=0.619969	lr=1.091941
Epoch[200] Batch [349]	Speed: 8790.995201 samples/sec	accuracy=0.619554	lr=1.091032
Epoch[200] Batch [399]	Speed: 8636.976229 samples/sec	accuracy=0.619084	lr=1.090124
Epoch[200] Batch [449]	Speed: 8840.479619 samples/sec	accuracy=0.618413	lr=1.089216
Epoch[200] Batch [499]	Speed: 8833.855905 samples/sec	accuracy=0.617658	lr=1.088307
Epoch[200] Batch [549]	Speed: 8606.118225 samples/sec	accuracy=0.617066	lr=1.087399
Epoch[200] Batch [599]	Speed: 8547.711972 samples/sec	accuracy=0.616456	lr=1.086491
[Epoch 200] training: accuracy=0.616357
[Epoch 200] speed: 8715 samples/sec	time cost: 151.946780
[Epoch 200] validation: err-top1=0.353008 err-top5=0.135879
Epoch[201] Batch [49]	Speed: 9029.106542 samples/sec	accuracy=0.624316	lr=1.085112
Epoch[201] Batch [99]	Speed: 8778.515002 samples/sec	accuracy=0.624316	lr=1.084204
Epoch[201] Batch [149]	Speed: 8752.679820 samples/sec	accuracy=0.622822	lr=1.083296
Epoch[201] Batch [199]	Speed: 8844.511290 samples/sec	accuracy=0.621741	lr=1.082389
Epoch[201] Batch [249]	Speed: 8774.856802 samples/sec	accuracy=0.620920	lr=1.081482
Epoch[201] Batch [299]	Speed: 8678.046968 samples/sec	accuracy=0.620218	lr=1.080575
Epoch[201] Batch [349]	Speed: 8791.948422 samples/sec	accuracy=0.619757	lr=1.079667
Epoch[201] Batch [399]	Speed: 8458.441521 samples/sec	accuracy=0.619309	lr=1.078760
Epoch[201] Batch [449]	Speed: 8793.022099 samples/sec	accuracy=0.618860	lr=1.077854
Epoch[201] Batch [499]	Speed: 8815.221325 samples/sec	accuracy=0.618127	lr=1.076947
Epoch[201] Batch [549]	Speed: 8670.620963 samples/sec	accuracy=0.617843	lr=1.076040
Epoch[201] Batch [599]	Speed: 8426.798845 samples/sec	accuracy=0.617450	lr=1.075134
[Epoch 201] training: accuracy=0.617086
[Epoch 201] speed: 8726 samples/sec	time cost: 151.585978
[Epoch 201] validation: err-top1=0.343608 err-top5=0.129049
Epoch[202] Batch [49]	Speed: 8693.903242 samples/sec	accuracy=0.624639	lr=1.073756
Epoch[202] Batch [99]	Speed: 8514.550339 samples/sec	accuracy=0.624272	lr=1.072850
Epoch[202] Batch [149]	Speed: 8646.003753 samples/sec	accuracy=0.622852	lr=1.071943
Epoch[202] Batch [199]	Speed: 8373.719135 samples/sec	accuracy=0.622190	lr=1.071037
Epoch[202] Batch [249]	Speed: 8452.499320 samples/sec	accuracy=0.621322	lr=1.070132
Epoch[202] Batch [299]	Speed: 8743.096328 samples/sec	accuracy=0.620638	lr=1.069226
Epoch[202] Batch [349]	Speed: 8718.725513 samples/sec	accuracy=0.620007	lr=1.068320
Epoch[202] Batch [399]	Speed: 8684.954596 samples/sec	accuracy=0.619218	lr=1.067414
Epoch[202] Batch [449]	Speed: 8772.041650 samples/sec	accuracy=0.618801	lr=1.066509
Epoch[202] Batch [499]	Speed: 8691.133982 samples/sec	accuracy=0.618365	lr=1.065604
Epoch[202] Batch [549]	Speed: 8638.374623 samples/sec	accuracy=0.617977	lr=1.064698
Epoch[202] Batch [599]	Speed: 8746.780994 samples/sec	accuracy=0.617651	lr=1.063793
[Epoch 202] training: accuracy=0.617363
[Epoch 202] speed: 8639 samples/sec	time cost: 152.941638
[Epoch 202] validation: err-top1=0.371191 err-top5=0.147773
Epoch[203] Batch [49]	Speed: 9092.448585 samples/sec	accuracy=0.626572	lr=1.062436
Epoch[203] Batch [99]	Speed: 8884.699605 samples/sec	accuracy=0.626191	lr=1.061531
Epoch[203] Batch [149]	Speed: 8602.088311 samples/sec	accuracy=0.624899	lr=1.060626
Epoch[203] Batch [199]	Speed: 8674.223584 samples/sec	accuracy=0.623743	lr=1.059722
Epoch[203] Batch [249]	Speed: 8601.399742 samples/sec	accuracy=0.622551	lr=1.058817
Epoch[203] Batch [299]	Speed: 8532.455196 samples/sec	accuracy=0.622080	lr=1.057913
Epoch[203] Batch [349]	Speed: 8878.810051 samples/sec	accuracy=0.621345	lr=1.057009
Epoch[203] Batch [399]	Speed: 8711.212793 samples/sec	accuracy=0.620507	lr=1.056105
Epoch[203] Batch [449]	Speed: 8581.694106 samples/sec	accuracy=0.619830	lr=1.055201
Epoch[203] Batch [499]	Speed: 8619.140377 samples/sec	accuracy=0.619354	lr=1.054297
Epoch[203] Batch [549]	Speed: 8702.680251 samples/sec	accuracy=0.618920	lr=1.053393
Epoch[203] Batch [599]	Speed: 8692.622102 samples/sec	accuracy=0.618369	lr=1.052490
[Epoch 203] training: accuracy=0.618276
[Epoch 203] speed: 8714 samples/sec	time cost: 151.829688
[Epoch 203] validation: err-top1=0.352010 err-top5=0.134054
Epoch[204] Batch [49]	Speed: 8421.691265 samples/sec	accuracy=0.627051	lr=1.051117
Epoch[204] Batch [99]	Speed: 8811.408101 samples/sec	accuracy=0.625103	lr=1.050213
Epoch[204] Batch [149]	Speed: 8765.997549 samples/sec	accuracy=0.623939	lr=1.049310
Epoch[204] Batch [199]	Speed: 8904.168209 samples/sec	accuracy=0.622876	lr=1.048407
Epoch[204] Batch [249]	Speed: 8827.893643 samples/sec	accuracy=0.622865	lr=1.047504
Epoch[204] Batch [299]	Speed: 8833.927493 samples/sec	accuracy=0.622056	lr=1.046601
Epoch[204] Batch [349]	Speed: 8811.303616 samples/sec	accuracy=0.621352	lr=1.045699
Epoch[204] Batch [399]	Speed: 8640.093966 samples/sec	accuracy=0.620657	lr=1.044796
Epoch[204] Batch [449]	Speed: 8700.669051 samples/sec	accuracy=0.620233	lr=1.043894
Epoch[204] Batch [499]	Speed: 8517.790440 samples/sec	accuracy=0.619794	lr=1.042992
Epoch[204] Batch [549]	Speed: 8558.479612 samples/sec	accuracy=0.619305	lr=1.042090
Epoch[204] Batch [599]	Speed: 8704.095419 samples/sec	accuracy=0.618997	lr=1.041188
[Epoch 204] training: accuracy=0.618805
[Epoch 204] speed: 8704 samples/sec	time cost: 151.698778
[Epoch 204] validation: err-top1=0.339010 err-top5=0.126200
Epoch[205] Batch [49]	Speed: 8410.022842 samples/sec	accuracy=0.630469	lr=1.039835
Epoch[205] Batch [99]	Speed: 8577.304854 samples/sec	accuracy=0.627949	lr=1.038933
Epoch[205] Batch [149]	Speed: 8486.935949 samples/sec	accuracy=0.626839	lr=1.038032
Epoch[205] Batch [199]	Speed: 8810.778156 samples/sec	accuracy=0.625452	lr=1.037130
Epoch[205] Batch [249]	Speed: 8577.718720 samples/sec	accuracy=0.625182	lr=1.036229
Epoch[205] Batch [299]	Speed: 8813.116726 samples/sec	accuracy=0.624126	lr=1.035328
Epoch[205] Batch [349]	Speed: 8659.117826 samples/sec	accuracy=0.623548	lr=1.034427
Epoch[205] Batch [399]	Speed: 8540.104109 samples/sec	accuracy=0.622955	lr=1.033526
Epoch[205] Batch [449]	Speed: 8638.354990 samples/sec	accuracy=0.622332	lr=1.032625
Epoch[205] Batch [499]	Speed: 8733.540193 samples/sec	accuracy=0.621779	lr=1.031724
Epoch[205] Batch [549]	Speed: 8662.230958 samples/sec	accuracy=0.621179	lr=1.030824
Epoch[205] Batch [599]	Speed: 8527.640519 samples/sec	accuracy=0.620667	lr=1.029924
[Epoch 205] training: accuracy=0.620424
[Epoch 205] speed: 8620 samples/sec	time cost: 153.562409
[Epoch 205] validation: err-top1=0.349121 err-top5=0.131738
Epoch[206] Batch [49]	Speed: 9024.407849 samples/sec	accuracy=0.629668	lr=1.028555
Epoch[206] Batch [99]	Speed: 8541.507660 samples/sec	accuracy=0.627573	lr=1.027655
Epoch[206] Batch [149]	Speed: 8645.851986 samples/sec	accuracy=0.626999	lr=1.026756
Epoch[206] Batch [199]	Speed: 8655.876095 samples/sec	accuracy=0.625410	lr=1.025856
Epoch[206] Batch [249]	Speed: 8689.641801 samples/sec	accuracy=0.624750	lr=1.024956
Epoch[206] Batch [299]	Speed: 8782.323149 samples/sec	accuracy=0.624032	lr=1.024057
Epoch[206] Batch [349]	Speed: 8620.400813 samples/sec	accuracy=0.623019	lr=1.023157
Epoch[206] Batch [399]	Speed: 8680.237007 samples/sec	accuracy=0.622454	lr=1.022258
Epoch[206] Batch [449]	Speed: 8923.496868 samples/sec	accuracy=0.621875	lr=1.021359
Epoch[206] Batch [499]	Speed: 8616.090803 samples/sec	accuracy=0.621404	lr=1.020460
Epoch[206] Batch [549]	Speed: 8655.954248 samples/sec	accuracy=0.621143	lr=1.019562
Epoch[206] Batch [599]	Speed: 8706.555419 samples/sec	accuracy=0.620880	lr=1.018663
[Epoch 206] training: accuracy=0.620679
[Epoch 206] speed: 8706 samples/sec	time cost: 151.674901
[Epoch 206] validation: err-top1=0.347860 err-top5=0.129272
Epoch[207] Batch [49]	Speed: 8540.243866 samples/sec	accuracy=0.630908	lr=1.017315
Epoch[207] Batch [99]	Speed: 8861.190886 samples/sec	accuracy=0.628496	lr=1.016417
Epoch[207] Batch [149]	Speed: 8854.925864 samples/sec	accuracy=0.628092	lr=1.015519
Epoch[207] Batch [199]	Speed: 8727.392596 samples/sec	accuracy=0.627373	lr=1.014621
Epoch[207] Batch [249]	Speed: 8751.162866 samples/sec	accuracy=0.626256	lr=1.013723
Epoch[207] Batch [299]	Speed: 8743.143137 samples/sec	accuracy=0.625223	lr=1.012825
Epoch[207] Batch [349]	Speed: 8690.996805 samples/sec	accuracy=0.624505	lr=1.011928
Epoch[207] Batch [399]	Speed: 8641.760787 samples/sec	accuracy=0.623982	lr=1.011030
Epoch[207] Batch [449]	Speed: 8571.075996 samples/sec	accuracy=0.623383	lr=1.010133
Epoch[207] Batch [499]	Speed: 8620.487150 samples/sec	accuracy=0.622810	lr=1.009236
Epoch[207] Batch [549]	Speed: 8622.641478 samples/sec	accuracy=0.622365	lr=1.008339
Epoch[207] Batch [599]	Speed: 8804.972627 samples/sec	accuracy=0.621897	lr=1.007442
[Epoch 207] training: accuracy=0.621658
[Epoch 207] speed: 8694 samples/sec	time cost: 152.347802
[Epoch 207] validation: err-top1=0.366152 err-top5=0.141836
Epoch[208] Batch [49]	Speed: 8763.421953 samples/sec	accuracy=0.632637	lr=1.006079
Epoch[208] Batch [99]	Speed: 8894.703606 samples/sec	accuracy=0.631489	lr=1.005183
Epoch[208] Batch [149]	Speed: 8521.271549 samples/sec	accuracy=0.629258	lr=1.004287
Epoch[208] Batch [199]	Speed: 8594.472989 samples/sec	accuracy=0.627961	lr=1.003390
Epoch[208] Batch [249]	Speed: 8463.746396 samples/sec	accuracy=0.626744	lr=1.002494
Epoch[208] Batch [299]	Speed: 8582.165844 samples/sec	accuracy=0.626361	lr=1.001599
Epoch[208] Batch [349]	Speed: 8841.209912 samples/sec	accuracy=0.625562	lr=1.000703
Epoch[208] Batch [399]	Speed: 8351.970905 samples/sec	accuracy=0.624950	lr=0.999807
Epoch[208] Batch [449]	Speed: 8751.603310 samples/sec	accuracy=0.624603	lr=0.998912
Epoch[208] Batch [499]	Speed: 8787.743715 samples/sec	accuracy=0.624005	lr=0.998017
Epoch[208] Batch [549]	Speed: 8517.984033 samples/sec	accuracy=0.623447	lr=0.997122
Epoch[208] Batch [599]	Speed: 8594.616423 samples/sec	accuracy=0.623009	lr=0.996227
[Epoch 208] training: accuracy=0.622803
[Epoch 208] speed: 8638 samples/sec	time cost: 152.921320
[Epoch 208] validation: err-top1=0.354777 err-top5=0.133789
Epoch[209] Batch [49]	Speed: 8479.626188 samples/sec	accuracy=0.631328	lr=0.994867
Epoch[209] Batch [99]	Speed: 8692.658168 samples/sec	accuracy=0.630181	lr=0.993972
Epoch[209] Batch [149]	Speed: 8940.596034 samples/sec	accuracy=0.629922	lr=0.993078
Epoch[209] Batch [199]	Speed: 8771.467659 samples/sec	accuracy=0.628347	lr=0.992183
Epoch[209] Batch [249]	Speed: 8765.283206 samples/sec	accuracy=0.627254	lr=0.991289
Epoch[209] Batch [299]	Speed: 8636.354132 samples/sec	accuracy=0.627048	lr=0.990395
Epoch[209] Batch [349]	Speed: 8364.740552 samples/sec	accuracy=0.626427	lr=0.989501
Epoch[209] Batch [399]	Speed: 8749.212786 samples/sec	accuracy=0.625720	lr=0.988608
Epoch[209] Batch [449]	Speed: 8637.590598 samples/sec	accuracy=0.625012	lr=0.987714
Epoch[209] Batch [499]	Speed: 8741.754211 samples/sec	accuracy=0.624232	lr=0.986821
Epoch[209] Batch [549]	Speed: 8732.000929 samples/sec	accuracy=0.624085	lr=0.985928
Epoch[209] Batch [599]	Speed: 8666.020169 samples/sec	accuracy=0.623557	lr=0.985035
[Epoch 209] training: accuracy=0.623346
[Epoch 209] speed: 8678 samples/sec	time cost: 152.085845
[Epoch 209] validation: err-top1=0.338908 err-top5=0.127279
Epoch[210] Batch [49]	Speed: 8337.117992 samples/sec	accuracy=0.633086	lr=0.983696
Epoch[210] Batch [99]	Speed: 8544.645068 samples/sec	accuracy=0.631992	lr=0.982803
Epoch[210] Batch [149]	Speed: 8566.456334 samples/sec	accuracy=0.630176	lr=0.981910
Epoch[210] Batch [199]	Speed: 8699.006564 samples/sec	accuracy=0.629353	lr=0.981018
Epoch[210] Batch [249]	Speed: 8588.815641 samples/sec	accuracy=0.627998	lr=0.980126
Epoch[210] Batch [299]	Speed: 8613.755589 samples/sec	accuracy=0.627606	lr=0.979234
Epoch[210] Batch [349]	Speed: 8691.876747 samples/sec	accuracy=0.627019	lr=0.978342
Epoch[210] Batch [399]	Speed: 8564.498033 samples/sec	accuracy=0.626323	lr=0.977451
Epoch[210] Batch [449]	Speed: 8857.605769 samples/sec	accuracy=0.625992	lr=0.976559
Epoch[210] Batch [499]	Speed: 8684.326973 samples/sec	accuracy=0.625631	lr=0.975668
Epoch[210] Batch [549]	Speed: 8400.825623 samples/sec	accuracy=0.625420	lr=0.974777
Epoch[210] Batch [599]	Speed: 8415.972733 samples/sec	accuracy=0.624837	lr=0.973886
[Epoch 210] training: accuracy=0.624478
[Epoch 210] speed: 8582 samples/sec	time cost: 154.313167
[Epoch 210] validation: err-top1=0.347676 err-top5=0.130117
Epoch[211] Batch [49]	Speed: 8425.091278 samples/sec	accuracy=0.633789	lr=0.972532
Epoch[211] Batch [99]	Speed: 8513.196131 samples/sec	accuracy=0.632788	lr=0.971641
Epoch[211] Batch [149]	Speed: 8492.877479 samples/sec	accuracy=0.632142	lr=0.970751
Epoch[211] Batch [199]	Speed: 8427.415425 samples/sec	accuracy=0.630901	lr=0.969860
Epoch[211] Batch [249]	Speed: 8732.825271 samples/sec	accuracy=0.630016	lr=0.968970
Epoch[211] Batch [299]	Speed: 8704.913794 samples/sec	accuracy=0.629152	lr=0.968080
Epoch[211] Batch [349]	Speed: 8732.365055 samples/sec	accuracy=0.628239	lr=0.967190
Epoch[211] Batch [399]	Speed: 8638.881631 samples/sec	accuracy=0.627638	lr=0.966301
Epoch[211] Batch [449]	Speed: 8910.095346 samples/sec	accuracy=0.627374	lr=0.965411
Epoch[211] Batch [499]	Speed: 8607.019526 samples/sec	accuracy=0.626824	lr=0.964522
Epoch[211] Batch [549]	Speed: 8690.271775 samples/sec	accuracy=0.626547	lr=0.963633
Epoch[211] Batch [599]	Speed: 8792.417280 samples/sec	accuracy=0.625800	lr=0.962744
[Epoch 211] training: accuracy=0.625561
[Epoch 211] speed: 8638 samples/sec	time cost: 152.645987
[Epoch 211] validation: err-top1=0.367533 err-top5=0.143799
Epoch[212] Batch [49]	Speed: 8470.679955 samples/sec	accuracy=0.631709	lr=0.961411
Epoch[212] Batch [99]	Speed: 8649.140902 samples/sec	accuracy=0.631890	lr=0.960523
Epoch[212] Batch [149]	Speed: 8720.469557 samples/sec	accuracy=0.631051	lr=0.959634
Epoch[212] Batch [199]	Speed: 8814.826738 samples/sec	accuracy=0.630278	lr=0.958746
Epoch[212] Batch [249]	Speed: 8585.004396 samples/sec	accuracy=0.629615	lr=0.957858
Epoch[212] Batch [299]	Speed: 8466.150164 samples/sec	accuracy=0.629702	lr=0.956970
Epoch[212] Batch [349]	Speed: 8909.800714 samples/sec	accuracy=0.629220	lr=0.956083
Epoch[212] Batch [399]	Speed: 8566.878381 samples/sec	accuracy=0.628429	lr=0.955195
Epoch[212] Batch [449]	Speed: 8552.397265 samples/sec	accuracy=0.627771	lr=0.954308
Epoch[212] Batch [499]	Speed: 8464.792455 samples/sec	accuracy=0.627214	lr=0.953421
Epoch[212] Batch [549]	Speed: 8961.796823 samples/sec	accuracy=0.626769	lr=0.952534
Epoch[212] Batch [599]	Speed: 8596.647714 samples/sec	accuracy=0.626381	lr=0.951647
[Epoch 212] training: accuracy=0.626244
[Epoch 212] speed: 8645 samples/sec	time cost: 153.126756
[Epoch 212] validation: err-top1=0.346777 err-top5=0.130820
Epoch[213] Batch [49]	Speed: 8426.147144 samples/sec	accuracy=0.636309	lr=0.950299
Epoch[213] Batch [99]	Speed: 8644.041277 samples/sec	accuracy=0.635415	lr=0.949413
Epoch[213] Batch [149]	Speed: 8686.781785 samples/sec	accuracy=0.633307	lr=0.948527
Epoch[213] Batch [199]	Speed: 8516.418314 samples/sec	accuracy=0.632234	lr=0.947641
Epoch[213] Batch [249]	Speed: 8778.287139 samples/sec	accuracy=0.631334	lr=0.946755
Epoch[213] Batch [299]	Speed: 8591.837340 samples/sec	accuracy=0.631120	lr=0.945869
Epoch[213] Batch [349]	Speed: 8646.574671 samples/sec	accuracy=0.630356	lr=0.944984
Epoch[213] Batch [399]	Speed: 8674.760214 samples/sec	accuracy=0.629667	lr=0.944099
Epoch[213] Batch [449]	Speed: 8625.974950 samples/sec	accuracy=0.629098	lr=0.943213
Epoch[213] Batch [499]	Speed: 8743.358322 samples/sec	accuracy=0.628593	lr=0.942329
Epoch[213] Batch [549]	Speed: 8320.811587 samples/sec	accuracy=0.627861	lr=0.941444
Epoch[213] Batch [599]	Speed: 8779.140522 samples/sec	accuracy=0.627294	lr=0.940559
[Epoch 213] training: accuracy=0.627073
[Epoch 213] speed: 8624 samples/sec	time cost: 152.960576
[Epoch 213] validation: err-top1=0.346069 err-top5=0.128886
Epoch[214] Batch [49]	Speed: 8446.874607 samples/sec	accuracy=0.637725	lr=0.939233
Epoch[214] Batch [99]	Speed: 8481.450228 samples/sec	accuracy=0.635420	lr=0.938349
Epoch[214] Batch [149]	Speed: 8845.858184 samples/sec	accuracy=0.634027	lr=0.937465
Epoch[214] Batch [199]	Speed: 8941.047377 samples/sec	accuracy=0.633057	lr=0.936581
Epoch[214] Batch [249]	Speed: 8502.002125 samples/sec	accuracy=0.631902	lr=0.935697
Epoch[214] Batch [299]	Speed: 8634.744773 samples/sec	accuracy=0.631512	lr=0.934814
Epoch[214] Batch [349]	Speed: 8485.174255 samples/sec	accuracy=0.630819	lr=0.933931
Epoch[214] Batch [399]	Speed: 8464.269810 samples/sec	accuracy=0.629990	lr=0.933048
Epoch[214] Batch [449]	Speed: 8626.191337 samples/sec	accuracy=0.629378	lr=0.932165
Epoch[214] Batch [499]	Speed: 8526.303471 samples/sec	accuracy=0.628745	lr=0.931282
Epoch[214] Batch [549]	Speed: 8519.058918 samples/sec	accuracy=0.628248	lr=0.930400
Epoch[214] Batch [599]	Speed: 8587.609239 samples/sec	accuracy=0.627835	lr=0.929517
[Epoch 214] training: accuracy=0.627730
[Epoch 214] speed: 8580 samples/sec	time cost: 154.198032
[Epoch 214] validation: err-top1=0.357871 err-top5=0.139102
Epoch[215] Batch [49]	Speed: 8445.398355 samples/sec	accuracy=0.638984	lr=0.928177
Epoch[215] Batch [99]	Speed: 8822.024591 samples/sec	accuracy=0.636738	lr=0.927295
Epoch[215] Batch [149]	Speed: 8405.935885 samples/sec	accuracy=0.635342	lr=0.926413
Epoch[215] Batch [199]	Speed: 8615.689300 samples/sec	accuracy=0.634048	lr=0.925532
Epoch[215] Batch [249]	Speed: 8821.592795 samples/sec	accuracy=0.632887	lr=0.924651
Epoch[215] Batch [299]	Speed: 8537.628470 samples/sec	accuracy=0.632114	lr=0.923770
Epoch[215] Batch [349]	Speed: 8608.413757 samples/sec	accuracy=0.631179	lr=0.922889
Epoch[215] Batch [399]	Speed: 8580.449420 samples/sec	accuracy=0.630405	lr=0.922008
Epoch[215] Batch [449]	Speed: 8666.004782 samples/sec	accuracy=0.629765	lr=0.921127
Epoch[215] Batch [499]	Speed: 8531.611303 samples/sec	accuracy=0.628872	lr=0.920247
Epoch[215] Batch [549]	Speed: 8841.066501 samples/sec	accuracy=0.628445	lr=0.919367
Epoch[215] Batch [599]	Speed: 8437.966810 samples/sec	accuracy=0.628187	lr=0.918487
[Epoch 215] training: accuracy=0.628008
[Epoch 215] speed: 8607 samples/sec	time cost: 153.604324
[Epoch 215] validation: err-top1=0.347066 err-top5=0.131307
Epoch[216] Batch [49]	Speed: 8746.822855 samples/sec	accuracy=0.638213	lr=0.917150
Epoch[216] Batch [99]	Speed: 8704.767890 samples/sec	accuracy=0.635713	lr=0.916271
Epoch[216] Batch [149]	Speed: 8539.650568 samples/sec	accuracy=0.634867	lr=0.915391
Epoch[216] Batch [199]	Speed: 8684.669397 samples/sec	accuracy=0.633755	lr=0.914512
Epoch[216] Batch [249]	Speed: 8803.999978 samples/sec	accuracy=0.633312	lr=0.913633
Epoch[216] Batch [299]	Speed: 8742.873859 samples/sec	accuracy=0.632801	lr=0.912755
Epoch[216] Batch [349]	Speed: 8839.067056 samples/sec	accuracy=0.632394	lr=0.911876
Epoch[216] Batch [399]	Speed: 8542.789500 samples/sec	accuracy=0.631600	lr=0.910998
Epoch[216] Batch [449]	Speed: 8502.941003 samples/sec	accuracy=0.631012	lr=0.910120
Epoch[216] Batch [499]	Speed: 8363.467287 samples/sec	accuracy=0.630764	lr=0.909242
Epoch[216] Batch [549]	Speed: 8683.397648 samples/sec	accuracy=0.630510	lr=0.908364
Epoch[216] Batch [599]	Speed: 8715.438829 samples/sec	accuracy=0.629836	lr=0.907487
[Epoch 216] training: accuracy=0.629649
[Epoch 216] speed: 8657 samples/sec	time cost: 152.705599
[Epoch 216] validation: err-top1=0.342468 err-top5=0.129822
Epoch[217] Batch [49]	Speed: 8544.565852 samples/sec	accuracy=0.641260	lr=0.906171
Epoch[217] Batch [99]	Speed: 8596.285356 samples/sec	accuracy=0.637578	lr=0.905294
Epoch[217] Batch [149]	Speed: 8573.430567 samples/sec	accuracy=0.635677	lr=0.904417
Epoch[217] Batch [199]	Speed: 8573.333703 samples/sec	accuracy=0.635278	lr=0.903540
Epoch[217] Batch [249]	Speed: 8503.929590 samples/sec	accuracy=0.634062	lr=0.902664
Epoch[217] Batch [299]	Speed: 8562.733357 samples/sec	accuracy=0.633626	lr=0.901788
Epoch[217] Batch [349]	Speed: 8694.403589 samples/sec	accuracy=0.632602	lr=0.900912
Epoch[217] Batch [399]	Speed: 8527.738385 samples/sec	accuracy=0.631887	lr=0.900036
Epoch[217] Batch [449]	Speed: 8728.629903 samples/sec	accuracy=0.631447	lr=0.899160
Epoch[217] Batch [499]	Speed: 8522.689210 samples/sec	accuracy=0.630985	lr=0.898285
Epoch[217] Batch [549]	Speed: 8605.412287 samples/sec	accuracy=0.630633	lr=0.897410
Epoch[217] Batch [599]	Speed: 8738.865661 samples/sec	accuracy=0.630282	lr=0.896535
[Epoch 217] training: accuracy=0.630118
[Epoch 217] speed: 8596 samples/sec	time cost: 153.946135
[Epoch 217] validation: err-top1=0.356094 err-top5=0.134648
Epoch[218] Batch [49]	Speed: 8611.268148 samples/sec	accuracy=0.640557	lr=0.895205
Epoch[218] Batch [99]	Speed: 8563.757583 samples/sec	accuracy=0.638481	lr=0.894331
Epoch[218] Batch [149]	Speed: 8726.521939 samples/sec	accuracy=0.636950	lr=0.893456
Epoch[218] Batch [199]	Speed: 8740.578995 samples/sec	accuracy=0.635991	lr=0.892582
Epoch[218] Batch [249]	Speed: 8670.074168 samples/sec	accuracy=0.635354	lr=0.891708
Epoch[218] Batch [299]	Speed: 8826.046154 samples/sec	accuracy=0.634528	lr=0.890835
Epoch[218] Batch [349]	Speed: 8711.363154 samples/sec	accuracy=0.633168	lr=0.889961
Epoch[218] Batch [399]	Speed: 8690.662499 samples/sec	accuracy=0.632477	lr=0.889088
Epoch[218] Batch [449]	Speed: 8754.989253 samples/sec	accuracy=0.631967	lr=0.888215
Epoch[218] Batch [499]	Speed: 8640.561369 samples/sec	accuracy=0.631432	lr=0.887342
Epoch[218] Batch [549]	Speed: 8693.882476 samples/sec	accuracy=0.631067	lr=0.886469
Epoch[218] Batch [599]	Speed: 8760.415787 samples/sec	accuracy=0.630542	lr=0.885597
[Epoch 218] training: accuracy=0.630410
[Epoch 218] speed: 8702 samples/sec	time cost: 151.755273
[Epoch 218] validation: err-top1=0.345703 err-top5=0.130839
Epoch[219] Batch [49]	Speed: 8625.234572 samples/sec	accuracy=0.640234	lr=0.884289
Epoch[219] Batch [99]	Speed: 8726.637012 samples/sec	accuracy=0.639272	lr=0.883417
Epoch[219] Batch [149]	Speed: 8697.055353 samples/sec	accuracy=0.637910	lr=0.882545
Epoch[219] Batch [199]	Speed: 8600.170859 samples/sec	accuracy=0.637610	lr=0.881673
Epoch[219] Batch [249]	Speed: 8637.159123 samples/sec	accuracy=0.636553	lr=0.880802
Epoch[219] Batch [299]	Speed: 8550.570672 samples/sec	accuracy=0.635654	lr=0.879931
Epoch[219] Batch [349]	Speed: 8568.724428 samples/sec	accuracy=0.634867	lr=0.879060
Epoch[219] Batch [399]	Speed: 8757.156703 samples/sec	accuracy=0.634093	lr=0.878189
Epoch[219] Batch [449]	Speed: 8774.009450 samples/sec	accuracy=0.633172	lr=0.877319
Epoch[219] Batch [499]	Speed: 8693.391691 samples/sec	accuracy=0.632547	lr=0.876449
Epoch[219] Batch [549]	Speed: 8688.227112 samples/sec	accuracy=0.632102	lr=0.875579
Epoch[219] Batch [599]	Speed: 8694.727623 samples/sec	accuracy=0.631807	lr=0.874709
[Epoch 219] training: accuracy=0.631603
[Epoch 219] speed: 8666 samples/sec	time cost: 152.698536
[Epoch 219] validation: err-top1=0.341152 err-top5=0.123809
Epoch[220] Batch [49]	Speed: 8440.450334 samples/sec	accuracy=0.641748	lr=0.873387
Epoch[220] Batch [99]	Speed: 8732.651796 samples/sec	accuracy=0.641079	lr=0.872518
Epoch[220] Batch [149]	Speed: 8631.672527 samples/sec	accuracy=0.639710	lr=0.871649
Epoch[220] Batch [199]	Speed: 8648.371813 samples/sec	accuracy=0.639236	lr=0.870780
Epoch[220] Batch [249]	Speed: 8518.390504 samples/sec	accuracy=0.638100	lr=0.869911
Epoch[220] Batch [299]	Speed: 8605.103496 samples/sec	accuracy=0.637135	lr=0.869043
Epoch[220] Batch [349]	Speed: 8409.361878 samples/sec	accuracy=0.636264	lr=0.868175
Epoch[220] Batch [399]	Speed: 8718.895425 samples/sec	accuracy=0.635464	lr=0.867307
Epoch[220] Batch [449]	Speed: 8502.630265 samples/sec	accuracy=0.634934	lr=0.866439
Epoch[220] Batch [499]	Speed: 8726.239323 samples/sec	accuracy=0.634598	lr=0.865571
Epoch[220] Batch [549]	Speed: 8589.413730 samples/sec	accuracy=0.634099	lr=0.864704
Epoch[220] Batch [599]	Speed: 8625.748181 samples/sec	accuracy=0.633814	lr=0.863837
[Epoch 220] training: accuracy=0.633587
[Epoch 220] speed: 8594 samples/sec	time cost: 153.605252
[Epoch 220] validation: err-top1=0.351054 err-top5=0.135213
Epoch[221] Batch [49]	Speed: 8195.417832 samples/sec	accuracy=0.643828	lr=0.862537
Epoch[221] Batch [99]	Speed: 8615.118824 samples/sec	accuracy=0.643594	lr=0.861670
Epoch[221] Batch [149]	Speed: 8484.179632 samples/sec	accuracy=0.640706	lr=0.860804
Epoch[221] Batch [199]	Speed: 8576.016232 samples/sec	accuracy=0.639456	lr=0.859938
Epoch[221] Batch [249]	Speed: 8538.281914 samples/sec	accuracy=0.638273	lr=0.859072
Epoch[221] Batch [299]	Speed: 8521.777754 samples/sec	accuracy=0.637552	lr=0.858206
Epoch[221] Batch [349]	Speed: 8551.443859 samples/sec	accuracy=0.636473	lr=0.857340
Epoch[221] Batch [399]	Speed: 8575.397749 samples/sec	accuracy=0.635642	lr=0.856475
Epoch[221] Batch [449]	Speed: 8591.154020 samples/sec	accuracy=0.635257	lr=0.855610
Epoch[221] Batch [499]	Speed: 8284.912929 samples/sec	accuracy=0.634554	lr=0.854745
Epoch[221] Batch [549]	Speed: 8531.277283 samples/sec	accuracy=0.634138	lr=0.853881
Epoch[221] Batch [599]	Speed: 8451.888211 samples/sec	accuracy=0.633577	lr=0.853016
[Epoch 221] training: accuracy=0.633518
[Epoch 221] speed: 8500 samples/sec	time cost: 155.528516
[Epoch 221] validation: err-top1=0.339111 err-top5=0.125326
Epoch[222] Batch [49]	Speed: 8678.830990 samples/sec	accuracy=0.641992	lr=0.851703
Epoch[222] Batch [99]	Speed: 8536.743000 samples/sec	accuracy=0.641577	lr=0.850839
Epoch[222] Batch [149]	Speed: 8827.737600 samples/sec	accuracy=0.640352	lr=0.849976
Epoch[222] Batch [199]	Speed: 8672.877833 samples/sec	accuracy=0.639595	lr=0.849112
Epoch[222] Batch [249]	Speed: 8578.432287 samples/sec	accuracy=0.638441	lr=0.848249
Epoch[222] Batch [299]	Speed: 8718.728345 samples/sec	accuracy=0.637539	lr=0.847386
Epoch[222] Batch [349]	Speed: 8542.808021 samples/sec	accuracy=0.636975	lr=0.846524
Epoch[222] Batch [399]	Speed: 8538.654337 samples/sec	accuracy=0.636206	lr=0.845661
Epoch[222] Batch [449]	Speed: 8596.903758 samples/sec	accuracy=0.635656	lr=0.844799
Epoch[222] Batch [499]	Speed: 8606.902929 samples/sec	accuracy=0.635140	lr=0.843937
Epoch[222] Batch [549]	Speed: 8631.170527 samples/sec	accuracy=0.634776	lr=0.843075
Epoch[222] Batch [599]	Speed: 8703.951659 samples/sec	accuracy=0.634532	lr=0.842214
[Epoch 222] training: accuracy=0.634347
[Epoch 222] speed: 8637 samples/sec	time cost: 153.336925
[Epoch 222] validation: err-top1=0.353770 err-top5=0.135527
Epoch[223] Batch [49]	Speed: 8418.880122 samples/sec	accuracy=0.645742	lr=0.840905
Epoch[223] Batch [99]	Speed: 8629.563099 samples/sec	accuracy=0.643286	lr=0.840044
Epoch[223] Batch [149]	Speed: 8596.555315 samples/sec	accuracy=0.641768	lr=0.839183
Epoch[223] Batch [199]	Speed: 8490.149022 samples/sec	accuracy=0.640073	lr=0.838323
Epoch[223] Batch [249]	Speed: 8311.859021 samples/sec	accuracy=0.639377	lr=0.837462
Epoch[223] Batch [299]	Speed: 8532.122466 samples/sec	accuracy=0.638680	lr=0.836602
Epoch[223] Batch [349]	Speed: 8756.603403 samples/sec	accuracy=0.637903	lr=0.835743
Epoch[223] Batch [399]	Speed: 8587.944078 samples/sec	accuracy=0.637372	lr=0.834883
Epoch[223] Batch [449]	Speed: 8626.106444 samples/sec	accuracy=0.636845	lr=0.834024
Epoch[223] Batch [499]	Speed: 8375.121764 samples/sec	accuracy=0.636268	lr=0.833165
Epoch[223] Batch [549]	Speed: 8444.900519 samples/sec	accuracy=0.636091	lr=0.832306
Epoch[223] Batch [599]	Speed: 8590.952792 samples/sec	accuracy=0.635614	lr=0.831447
[Epoch 223] training: accuracy=0.635401
[Epoch 223] speed: 8526 samples/sec	time cost: 154.790950
[Epoch 223] validation: err-top1=0.336934 err-top5=0.123637
Epoch[224] Batch [49]	Speed: 8352.954101 samples/sec	accuracy=0.646387	lr=0.830160
Epoch[224] Batch [99]	Speed: 8644.597842 samples/sec	accuracy=0.644404	lr=0.829302
Epoch[224] Batch [149]	Speed: 8528.073651 samples/sec	accuracy=0.642865	lr=0.828444
Epoch[224] Batch [199]	Speed: 8565.104867 samples/sec	accuracy=0.642239	lr=0.827586
Epoch[224] Batch [249]	Speed: 8685.008336 samples/sec	accuracy=0.641121	lr=0.826729
Epoch[224] Batch [299]	Speed: 8575.015095 samples/sec	accuracy=0.640093	lr=0.825872
Epoch[224] Batch [349]	Speed: 8606.181858 samples/sec	accuracy=0.639283	lr=0.825015
Epoch[224] Batch [399]	Speed: 8791.272850 samples/sec	accuracy=0.638691	lr=0.824158
Epoch[224] Batch [449]	Speed: 8600.216495 samples/sec	accuracy=0.637900	lr=0.823302
Epoch[224] Batch [499]	Speed: 8719.600811 samples/sec	accuracy=0.637310	lr=0.822446
Epoch[224] Batch [549]	Speed: 8501.894078 samples/sec	accuracy=0.636917	lr=0.821590
Epoch[224] Batch [599]	Speed: 8626.746645 samples/sec	accuracy=0.636555	lr=0.820734
[Epoch 224] training: accuracy=0.636370
[Epoch 224] speed: 8589 samples/sec	time cost: 154.187932
[Epoch 224] validation: err-top1=0.330840 err-top5=0.120957
Epoch[225] Batch [49]	Speed: 8416.777244 samples/sec	accuracy=0.645469	lr=0.819434
Epoch[225] Batch [99]	Speed: 8685.776753 samples/sec	accuracy=0.643872	lr=0.818579
Epoch[225] Batch [149]	Speed: 8524.587153 samples/sec	accuracy=0.643444	lr=0.817724
Epoch[225] Batch [199]	Speed: 8448.093468 samples/sec	accuracy=0.642512	lr=0.816870
Epoch[225] Batch [249]	Speed: 8470.904157 samples/sec	accuracy=0.641348	lr=0.816016
Epoch[225] Batch [299]	Speed: 8645.822572 samples/sec	accuracy=0.640557	lr=0.815161
Epoch[225] Batch [349]	Speed: 8365.387025 samples/sec	accuracy=0.640028	lr=0.814308
Epoch[225] Batch [399]	Speed: 8573.576893 samples/sec	accuracy=0.639396	lr=0.813454
Epoch[225] Batch [449]	Speed: 8598.392315 samples/sec	accuracy=0.638416	lr=0.812601
Epoch[225] Batch [499]	Speed: 8378.460408 samples/sec	accuracy=0.638043	lr=0.811748
Epoch[225] Batch [549]	Speed: 8801.523210 samples/sec	accuracy=0.637487	lr=0.810895
Epoch[225] Batch [599]	Speed: 8678.612481 samples/sec	accuracy=0.637056	lr=0.810042
[Epoch 225] training: accuracy=0.637069
[Epoch 225] speed: 8543 samples/sec	time cost: 154.457540
[Epoch 225] validation: err-top1=0.345927 err-top5=0.132507
Epoch[226] Batch [49]	Speed: 8492.585276 samples/sec	accuracy=0.644932	lr=0.808764
Epoch[226] Batch [99]	Speed: 8877.429984 samples/sec	accuracy=0.644907	lr=0.807912
Epoch[226] Batch [149]	Speed: 8493.171381 samples/sec	accuracy=0.644473	lr=0.807060
Epoch[226] Batch [199]	Speed: 8541.564566 samples/sec	accuracy=0.643838	lr=0.806208
Epoch[226] Batch [249]	Speed: 8592.098942 samples/sec	accuracy=0.642639	lr=0.805357
Epoch[226] Batch [299]	Speed: 8513.469335 samples/sec	accuracy=0.641799	lr=0.804506
Epoch[226] Batch [349]	Speed: 8515.070602 samples/sec	accuracy=0.641004	lr=0.803655
Epoch[226] Batch [399]	Speed: 8690.852774 samples/sec	accuracy=0.640502	lr=0.802805
Epoch[226] Batch [449]	Speed: 8445.238935 samples/sec	accuracy=0.639855	lr=0.801955
Epoch[226] Batch [499]	Speed: 8440.916458 samples/sec	accuracy=0.639502	lr=0.801105
Epoch[226] Batch [549]	Speed: 8688.294953 samples/sec	accuracy=0.638952	lr=0.800255
Epoch[226] Batch [599]	Speed: 8509.167278 samples/sec	accuracy=0.638594	lr=0.799405
[Epoch 226] training: accuracy=0.638502
[Epoch 226] speed: 8568 samples/sec	time cost: 154.087463
[Epoch 226] validation: err-top1=0.321411 err-top5=0.112549
Epoch[227] Batch [49]	Speed: 8555.793042 samples/sec	accuracy=0.650400	lr=0.798115
Epoch[227] Batch [99]	Speed: 8533.962042 samples/sec	accuracy=0.646660	lr=0.797266
Epoch[227] Batch [149]	Speed: 8580.132306 samples/sec	accuracy=0.645863	lr=0.796417
Epoch[227] Batch [199]	Speed: 8575.618283 samples/sec	accuracy=0.644653	lr=0.795569
Epoch[227] Batch [249]	Speed: 8457.152561 samples/sec	accuracy=0.643258	lr=0.794721
Epoch[227] Batch [299]	Speed: 8357.789142 samples/sec	accuracy=0.642271	lr=0.793873
Epoch[227] Batch [349]	Speed: 8597.613292 samples/sec	accuracy=0.641346	lr=0.793025
Epoch[227] Batch [399]	Speed: 8583.149098 samples/sec	accuracy=0.640461	lr=0.792178
Epoch[227] Batch [449]	Speed: 8761.389375 samples/sec	accuracy=0.640073	lr=0.791331
Epoch[227] Batch [499]	Speed: 8721.165457 samples/sec	accuracy=0.639708	lr=0.790484
Epoch[227] Batch [549]	Speed: 8607.375372 samples/sec	accuracy=0.639457	lr=0.789637
Epoch[227] Batch [599]	Speed: 8419.885242 samples/sec	accuracy=0.638998	lr=0.788791
[Epoch 227] training: accuracy=0.638702
[Epoch 227] speed: 8573 samples/sec	time cost: 154.109729
[Epoch 227] validation: err-top1=0.338398 err-top5=0.128105
Epoch[228] Batch [49]	Speed: 8309.274391 samples/sec	accuracy=0.647188	lr=0.787522
Epoch[228] Batch [99]	Speed: 8703.330282 samples/sec	accuracy=0.646260	lr=0.786676
Epoch[228] Batch [149]	Speed: 8597.111633 samples/sec	accuracy=0.645524	lr=0.785831
Epoch[228] Batch [199]	Speed: 8504.056715 samples/sec	accuracy=0.644539	lr=0.784986
Epoch[228] Batch [249]	Speed: 8558.534015 samples/sec	accuracy=0.644086	lr=0.784141
Epoch[228] Batch [299]	Speed: 8529.991103 samples/sec	accuracy=0.643600	lr=0.783296
Epoch[228] Batch [349]	Speed: 8435.266391 samples/sec	accuracy=0.643348	lr=0.782452
Epoch[228] Batch [399]	Speed: 8566.145207 samples/sec	accuracy=0.642513	lr=0.781607
Epoch[228] Batch [449]	Speed: 8780.605439 samples/sec	accuracy=0.642043	lr=0.780763
Epoch[228] Batch [499]	Speed: 8545.902343 samples/sec	accuracy=0.641453	lr=0.779920
Epoch[228] Batch [549]	Speed: 8576.340920 samples/sec	accuracy=0.640794	lr=0.779076
Epoch[228] Batch [599]	Speed: 8715.228729 samples/sec	accuracy=0.640341	lr=0.778233
[Epoch 228] training: accuracy=0.639942
[Epoch 228] speed: 8568 samples/sec	time cost: 154.171349
[Epoch 228] validation: err-top1=0.346436 err-top5=0.128052
Epoch[229] Batch [49]	Speed: 8462.469992 samples/sec	accuracy=0.646934	lr=0.776952
Epoch[229] Batch [99]	Speed: 8826.059032 samples/sec	accuracy=0.645288	lr=0.776110
Epoch[229] Batch [149]	Speed: 8796.474043 samples/sec	accuracy=0.644479	lr=0.775268
Epoch[229] Batch [199]	Speed: 8752.603478 samples/sec	accuracy=0.644177	lr=0.774426
Epoch[229] Batch [249]	Speed: 8419.051587 samples/sec	accuracy=0.643641	lr=0.773584
Epoch[229] Batch [299]	Speed: 8636.548636 samples/sec	accuracy=0.643027	lr=0.772743
Epoch[229] Batch [349]	Speed: 8490.299568 samples/sec	accuracy=0.642725	lr=0.771902
Epoch[229] Batch [399]	Speed: 8677.380723 samples/sec	accuracy=0.642335	lr=0.771061
Epoch[229] Batch [449]	Speed: 8711.116325 samples/sec	accuracy=0.641695	lr=0.770220
Epoch[229] Batch [499]	Speed: 8641.452338 samples/sec	accuracy=0.641311	lr=0.769380
Epoch[229] Batch [549]	Speed: 8555.168952 samples/sec	accuracy=0.640973	lr=0.768540
Epoch[229] Batch [599]	Speed: 8795.640704 samples/sec	accuracy=0.640572	lr=0.767700
[Epoch 229] training: accuracy=0.640431
[Epoch 229] speed: 8632 samples/sec	time cost: 153.429318
[Epoch 229] validation: err-top1=0.339844 err-top5=0.126133
Epoch[230] Batch [49]	Speed: 8388.951095 samples/sec	accuracy=0.649873	lr=0.766424
Epoch[230] Batch [99]	Speed: 8616.655012 samples/sec	accuracy=0.648140	lr=0.765585
Epoch[230] Batch [149]	Speed: 8535.996484 samples/sec	accuracy=0.647262	lr=0.764746
Epoch[230] Batch [199]	Speed: 8713.225867 samples/sec	accuracy=0.646113	lr=0.763907
Epoch[230] Batch [249]	Speed: 8472.299593 samples/sec	accuracy=0.645771	lr=0.763069
Epoch[230] Batch [299]	Speed: 8567.122743 samples/sec	accuracy=0.645109	lr=0.762231
Epoch[230] Batch [349]	Speed: 8661.339192 samples/sec	accuracy=0.644294	lr=0.761393
Epoch[230] Batch [399]	Speed: 8495.429565 samples/sec	accuracy=0.643809	lr=0.760556
Epoch[230] Batch [449]	Speed: 8611.304405 samples/sec	accuracy=0.643204	lr=0.759718
Epoch[230] Batch [499]	Speed: 8687.525387 samples/sec	accuracy=0.642727	lr=0.758881
Epoch[230] Batch [549]	Speed: 8592.724822 samples/sec	accuracy=0.642526	lr=0.758045
Epoch[230] Batch [599]	Speed: 8688.005142 samples/sec	accuracy=0.642126	lr=0.757208
[Epoch 230] training: accuracy=0.641908
[Epoch 230] speed: 8586 samples/sec	time cost: 153.573283
[Epoch 230] validation: err-top1=0.332275 err-top5=0.121847
Epoch[231] Batch [49]	Speed: 8650.652832 samples/sec	accuracy=0.649600	lr=0.755954
Epoch[231] Batch [99]	Speed: 8850.137302 samples/sec	accuracy=0.648638	lr=0.755118
Epoch[231] Batch [149]	Speed: 8449.159263 samples/sec	accuracy=0.647946	lr=0.754283
Epoch[231] Batch [199]	Speed: 8769.071643 samples/sec	accuracy=0.646731	lr=0.753448
Epoch[231] Batch [249]	Speed: 8657.873968 samples/sec	accuracy=0.645277	lr=0.752613
Epoch[231] Batch [299]	Speed: 8573.473523 samples/sec	accuracy=0.645138	lr=0.751778
Epoch[231] Batch [349]	Speed: 8544.047588 samples/sec	accuracy=0.644354	lr=0.750944
Epoch[231] Batch [399]	Speed: 8569.085664 samples/sec	accuracy=0.643778	lr=0.750110
Epoch[231] Batch [449]	Speed: 8572.567085 samples/sec	accuracy=0.643394	lr=0.749276
Epoch[231] Batch [499]	Speed: 8681.152321 samples/sec	accuracy=0.643178	lr=0.748442
Epoch[231] Batch [549]	Speed: 8745.339271 samples/sec	accuracy=0.643076	lr=0.747609
Epoch[231] Batch [599]	Speed: 8775.979746 samples/sec	accuracy=0.642647	lr=0.746776
[Epoch 231] training: accuracy=0.642495
[Epoch 231] speed: 8654 samples/sec	time cost: 152.992591
[Epoch 231] validation: err-top1=0.328340 err-top5=0.117891
Epoch[232] Batch [49]	Speed: 8262.588358 samples/sec	accuracy=0.653916	lr=0.745510
Epoch[232] Batch [99]	Speed: 8547.323109 samples/sec	accuracy=0.650757	lr=0.744678
Epoch[232] Batch [149]	Speed: 8663.527446 samples/sec	accuracy=0.649909	lr=0.743846
Epoch[232] Batch [199]	Speed: 8827.274946 samples/sec	accuracy=0.649033	lr=0.743014
Epoch[232] Batch [249]	Speed: 8699.076864 samples/sec	accuracy=0.647955	lr=0.742183
Epoch[232] Batch [299]	Speed: 8637.940986 samples/sec	accuracy=0.647393	lr=0.741352
Epoch[232] Batch [349]	Speed: 8704.182559 samples/sec	accuracy=0.646498	lr=0.740521
Epoch[232] Batch [399]	Speed: 8653.310748 samples/sec	accuracy=0.645950	lr=0.739690
Epoch[232] Batch [449]	Speed: 8645.941618 samples/sec	accuracy=0.645243	lr=0.738860
Epoch[232] Batch [499]	Speed: 8609.066521 samples/sec	accuracy=0.644411	lr=0.738030
Epoch[232] Batch [549]	Speed: 8442.395458 samples/sec	accuracy=0.643999	lr=0.737200
Epoch[232] Batch [599]	Speed: 8543.412465 samples/sec	accuracy=0.643757	lr=0.736371
[Epoch 232] training: accuracy=0.643437
[Epoch 232] speed: 8593 samples/sec	time cost: 153.636587
[Epoch 232] validation: err-top1=0.346965 err-top5=0.130249
Epoch[233] Batch [49]	Speed: 8455.887300 samples/sec	accuracy=0.653086	lr=0.735127
Epoch[233] Batch [99]	Speed: 8391.950520 samples/sec	accuracy=0.651372	lr=0.734298
Epoch[233] Batch [149]	Speed: 8603.619854 samples/sec	accuracy=0.650088	lr=0.733470
Epoch[233] Batch [199]	Speed: 8709.038179 samples/sec	accuracy=0.648945	lr=0.732641
Epoch[233] Batch [249]	Speed: 8960.711448 samples/sec	accuracy=0.648252	lr=0.731813
Epoch[233] Batch [299]	Speed: 8731.918912 samples/sec	accuracy=0.648127	lr=0.730986
Epoch[233] Batch [349]	Speed: 8550.089465 samples/sec	accuracy=0.647327	lr=0.730158
Epoch[233] Batch [399]	Speed: 8488.934774 samples/sec	accuracy=0.646849	lr=0.729331
Epoch[233] Batch [449]	Speed: 8699.127784 samples/sec	accuracy=0.646331	lr=0.728504
Epoch[233] Batch [499]	Speed: 8598.559119 samples/sec	accuracy=0.645933	lr=0.727678
Epoch[233] Batch [549]	Speed: 8674.040868 samples/sec	accuracy=0.645210	lr=0.726852
Epoch[233] Batch [599]	Speed: 8931.410830 samples/sec	accuracy=0.644673	lr=0.726026
[Epoch 233] training: accuracy=0.644430
[Epoch 233] speed: 8645 samples/sec	time cost: 152.942750
[Epoch 233] validation: err-top1=0.324443 err-top5=0.115926
Epoch[234] Batch [49]	Speed: 8555.236265 samples/sec	accuracy=0.655889	lr=0.724771
Epoch[234] Batch [99]	Speed: 8782.910057 samples/sec	accuracy=0.653281	lr=0.723946
Epoch[234] Batch [149]	Speed: 8677.211723 samples/sec	accuracy=0.652259	lr=0.723121
Epoch[234] Batch [199]	Speed: 8723.760037 samples/sec	accuracy=0.650901	lr=0.722296
Epoch[234] Batch [249]	Speed: 8622.498492 samples/sec	accuracy=0.649820	lr=0.721472
Epoch[234] Batch [299]	Speed: 8889.748655 samples/sec	accuracy=0.649238	lr=0.720648
Epoch[234] Batch [349]	Speed: 8931.970281 samples/sec	accuracy=0.648403	lr=0.719824
Epoch[234] Batch [399]	Speed: 8964.312801 samples/sec	accuracy=0.647646	lr=0.719000
Epoch[234] Batch [449]	Speed: 8736.321087 samples/sec	accuracy=0.647070	lr=0.718177
Epoch[234] Batch [499]	Speed: 8796.743750 samples/sec	accuracy=0.646420	lr=0.717354
Epoch[234] Batch [549]	Speed: 8593.184191 samples/sec	accuracy=0.645871	lr=0.716532
Epoch[234] Batch [599]	Speed: 8703.867522 samples/sec	accuracy=0.645400	lr=0.715710
[Epoch 234] training: accuracy=0.645190
[Epoch 234] speed: 8744 samples/sec	time cost: 151.372500
[Epoch 234] validation: err-top1=0.330898 err-top5=0.123652
Epoch[235] Batch [49]	Speed: 8598.759155 samples/sec	accuracy=0.654033	lr=0.714477
Epoch[235] Batch [99]	Speed: 8712.170173 samples/sec	accuracy=0.653027	lr=0.713655
Epoch[235] Batch [149]	Speed: 8814.888248 samples/sec	accuracy=0.652516	lr=0.712834
Epoch[235] Batch [199]	Speed: 8609.120189 samples/sec	accuracy=0.651440	lr=0.712013
Epoch[235] Batch [249]	Speed: 8828.068019 samples/sec	accuracy=0.650658	lr=0.711192
Epoch[235] Batch [299]	Speed: 8513.622060 samples/sec	accuracy=0.649824	lr=0.710372
Epoch[235] Batch [349]	Speed: 8791.787708 samples/sec	accuracy=0.649602	lr=0.709552
Epoch[235] Batch [399]	Speed: 8455.231592 samples/sec	accuracy=0.648883	lr=0.708732
Epoch[235] Batch [449]	Speed: 8748.489236 samples/sec	accuracy=0.648316	lr=0.707912
Epoch[235] Batch [499]	Speed: 8751.027532 samples/sec	accuracy=0.647743	lr=0.707093
Epoch[235] Batch [549]	Speed: 8745.543879 samples/sec	accuracy=0.647369	lr=0.706274
Epoch[235] Batch [599]	Speed: 8804.553147 samples/sec	accuracy=0.647061	lr=0.705456
[Epoch 235] training: accuracy=0.646867
[Epoch 235] speed: 8693 samples/sec	time cost: 152.075070
[Epoch 235] validation: err-top1=0.342448 err-top5=0.129964
Epoch[236] Batch [49]	Speed: 8400.339271 samples/sec	accuracy=0.657217	lr=0.704212
Epoch[236] Batch [99]	Speed: 8739.830018 samples/sec	accuracy=0.655566	lr=0.703394
Epoch[236] Batch [149]	Speed: 8741.026736 samples/sec	accuracy=0.654082	lr=0.702576
Epoch[236] Batch [199]	Speed: 8770.290530 samples/sec	accuracy=0.652883	lr=0.701759
Epoch[236] Batch [249]	Speed: 8755.257314 samples/sec	accuracy=0.651582	lr=0.700942
Epoch[236] Batch [299]	Speed: 8713.797565 samples/sec	accuracy=0.650467	lr=0.700125
Epoch[236] Batch [349]	Speed: 8673.429360 samples/sec	accuracy=0.650020	lr=0.699309
Epoch[236] Batch [399]	Speed: 8818.049780 samples/sec	accuracy=0.649666	lr=0.698493
Epoch[236] Batch [449]	Speed: 8734.851366 samples/sec	accuracy=0.649025	lr=0.697677
Epoch[236] Batch [499]	Speed: 8616.155967 samples/sec	accuracy=0.648359	lr=0.696862
Epoch[236] Batch [549]	Speed: 8373.308232 samples/sec	accuracy=0.647794	lr=0.696047
Epoch[236] Batch [599]	Speed: 8644.459346 samples/sec	accuracy=0.647367	lr=0.695232
[Epoch 236] training: accuracy=0.647176
[Epoch 236] speed: 8660 samples/sec	time cost: 152.793268
[Epoch 236] validation: err-top1=0.320020 err-top5=0.113945
Epoch[237] Batch [49]	Speed: 8521.465807 samples/sec	accuracy=0.658271	lr=0.693994
Epoch[237] Batch [99]	Speed: 8718.042566 samples/sec	accuracy=0.655732	lr=0.693180
Epoch[237] Batch [149]	Speed: 8592.917365 samples/sec	accuracy=0.654251	lr=0.692366
Epoch[237] Batch [199]	Speed: 8750.009007 samples/sec	accuracy=0.653342	lr=0.691552
Epoch[237] Batch [249]	Speed: 8766.963430 samples/sec	accuracy=0.652395	lr=0.690739
Epoch[237] Batch [299]	Speed: 8726.166278 samples/sec	accuracy=0.651097	lr=0.689926
Epoch[237] Batch [349]	Speed: 8703.268202 samples/sec	accuracy=0.650364	lr=0.689114
Epoch[237] Batch [399]	Speed: 8463.812945 samples/sec	accuracy=0.649734	lr=0.688302
Epoch[237] Batch [449]	Speed: 8737.637539 samples/sec	accuracy=0.649468	lr=0.687490
Epoch[237] Batch [499]	Speed: 8858.434083 samples/sec	accuracy=0.649000	lr=0.686678
Epoch[237] Batch [549]	Speed: 8758.552850 samples/sec	accuracy=0.648657	lr=0.685867
Epoch[237] Batch [599]	Speed: 8630.062657 samples/sec	accuracy=0.648298	lr=0.685056
[Epoch 237] training: accuracy=0.648239
[Epoch 237] speed: 8690 samples/sec	time cost: 151.832861
[Epoch 237] validation: err-top1=0.324015 err-top5=0.117065
Epoch[238] Batch [49]	Speed: 8533.296204 samples/sec	accuracy=0.659453	lr=0.683840
Epoch[238] Batch [99]	Speed: 8775.636001 samples/sec	accuracy=0.657495	lr=0.683029
Epoch[238] Batch [149]	Speed: 8704.417176 samples/sec	accuracy=0.655570	lr=0.682219
Epoch[238] Batch [199]	Speed: 8530.683874 samples/sec	accuracy=0.655054	lr=0.681410
Epoch[238] Batch [249]	Speed: 8543.173702 samples/sec	accuracy=0.653922	lr=0.680601
Epoch[238] Batch [299]	Speed: 8393.431600 samples/sec	accuracy=0.652596	lr=0.679791
Epoch[238] Batch [349]	Speed: 8656.066246 samples/sec	accuracy=0.651772	lr=0.678983
Epoch[238] Batch [399]	Speed: 8604.524251 samples/sec	accuracy=0.651116	lr=0.678174
Epoch[238] Batch [449]	Speed: 8582.161557 samples/sec	accuracy=0.650611	lr=0.677366
Epoch[238] Batch [499]	Speed: 8823.442951 samples/sec	accuracy=0.649773	lr=0.676559
Epoch[238] Batch [549]	Speed: 8539.664151 samples/sec	accuracy=0.649334	lr=0.675751
Epoch[238] Batch [599]	Speed: 8586.302243 samples/sec	accuracy=0.648845	lr=0.674944
[Epoch 238] training: accuracy=0.648741
[Epoch 238] speed: 8613 samples/sec	time cost: 153.444394
[Epoch 238] validation: err-top1=0.322957 err-top5=0.115133
Epoch[239] Batch [49]	Speed: 8101.137705 samples/sec	accuracy=0.659043	lr=0.673718
Epoch[239] Batch [99]	Speed: 8436.548688 samples/sec	accuracy=0.657305	lr=0.672911
Epoch[239] Batch [149]	Speed: 8967.484335 samples/sec	accuracy=0.656038	lr=0.672105
Epoch[239] Batch [199]	Speed: 8729.076242 samples/sec	accuracy=0.654639	lr=0.671300
Epoch[239] Batch [249]	Speed: 8641.900761 samples/sec	accuracy=0.653445	lr=0.670494
Epoch[239] Batch [299]	Speed: 8620.675230 samples/sec	accuracy=0.652728	lr=0.669689
Epoch[239] Batch [349]	Speed: 8625.437063 samples/sec	accuracy=0.652453	lr=0.668884
Epoch[239] Batch [399]	Speed: 8411.075428 samples/sec	accuracy=0.652008	lr=0.668080
Epoch[239] Batch [449]	Speed: 8686.604864 samples/sec	accuracy=0.651645	lr=0.667276
Epoch[239] Batch [499]	Speed: 8586.266196 samples/sec	accuracy=0.650951	lr=0.666472
Epoch[239] Batch [549]	Speed: 8602.396884 samples/sec	accuracy=0.650435	lr=0.665668
Epoch[239] Batch [599]	Speed: 8529.216465 samples/sec	accuracy=0.650051	lr=0.664865
[Epoch 239] training: accuracy=0.649884
[Epoch 239] speed: 8581 samples/sec	time cost: 154.010437
[Epoch 239] validation: err-top1=0.323672 err-top5=0.116934
Epoch[240] Batch [49]	Speed: 8444.616258 samples/sec	accuracy=0.659551	lr=0.663661
Epoch[240] Batch [99]	Speed: 8926.310462 samples/sec	accuracy=0.658398	lr=0.662858
Epoch[240] Batch [149]	Speed: 8585.648636 samples/sec	accuracy=0.657393	lr=0.662056
Epoch[240] Batch [199]	Speed: 8613.967216 samples/sec	accuracy=0.656401	lr=0.661255
Epoch[240] Batch [249]	Speed: 8424.365483 samples/sec	accuracy=0.655414	lr=0.660453
Epoch[240] Batch [299]	Speed: 8655.156737 samples/sec	accuracy=0.654689	lr=0.659652
Epoch[240] Batch [349]	Speed: 8514.659383 samples/sec	accuracy=0.653864	lr=0.658851
Epoch[240] Batch [399]	Speed: 8687.730111 samples/sec	accuracy=0.653434	lr=0.658051
Epoch[240] Batch [449]	Speed: 8780.457884 samples/sec	accuracy=0.652850	lr=0.657251
Epoch[240] Batch [499]	Speed: 8746.075657 samples/sec	accuracy=0.652537	lr=0.656451
Epoch[240] Batch [549]	Speed: 8376.729732 samples/sec	accuracy=0.651909	lr=0.655651
Epoch[240] Batch [599]	Speed: 8627.816574 samples/sec	accuracy=0.651576	lr=0.654852
[Epoch 240] training: accuracy=0.651384
[Epoch 240] speed: 8618 samples/sec	time cost: 153.254630
[Epoch 240] validation: err-top1=0.339152 err-top5=0.125875
Epoch[241] Batch [49]	Speed: 8459.714045 samples/sec	accuracy=0.660723	lr=0.653638
Epoch[241] Batch [99]	Speed: 8562.300282 samples/sec	accuracy=0.658794	lr=0.652840
Epoch[241] Batch [149]	Speed: 8818.089972 samples/sec	accuracy=0.657790	lr=0.652042
Epoch[241] Batch [199]	Speed: 8650.386085 samples/sec	accuracy=0.657283	lr=0.651244
Epoch[241] Batch [249]	Speed: 8558.360916 samples/sec	accuracy=0.656426	lr=0.650446
Epoch[241] Batch [299]	Speed: 8591.129446 samples/sec	accuracy=0.655407	lr=0.649649
Epoch[241] Batch [349]	Speed: 8747.873423 samples/sec	accuracy=0.654920	lr=0.648853
Epoch[241] Batch [399]	Speed: 8697.265282 samples/sec	accuracy=0.654346	lr=0.648056
Epoch[241] Batch [449]	Speed: 8660.458786 samples/sec	accuracy=0.653873	lr=0.647260
Epoch[241] Batch [499]	Speed: 8605.429701 samples/sec	accuracy=0.653344	lr=0.646464
Epoch[241] Batch [549]	Speed: 8732.490224 samples/sec	accuracy=0.652668	lr=0.645669
Epoch[241] Batch [599]	Speed: 8731.531747 samples/sec	accuracy=0.652052	lr=0.644874
[Epoch 241] training: accuracy=0.651930
[Epoch 241] speed: 8656 samples/sec	time cost: 152.670686
[Epoch 241] validation: err-top1=0.318613 err-top5=0.114473
Epoch[242] Batch [49]	Speed: 8656.423194 samples/sec	accuracy=0.662197	lr=0.643682
Epoch[242] Batch [99]	Speed: 8747.601359 samples/sec	accuracy=0.660415	lr=0.642887
Epoch[242] Batch [149]	Speed: 8758.660910 samples/sec	accuracy=0.658477	lr=0.642093
Epoch[242] Batch [199]	Speed: 8780.910258 samples/sec	accuracy=0.657500	lr=0.641300
Epoch[242] Batch [249]	Speed: 8582.590984 samples/sec	accuracy=0.656457	lr=0.640507
Epoch[242] Batch [299]	Speed: 9012.958710 samples/sec	accuracy=0.655843	lr=0.639714
Epoch[242] Batch [349]	Speed: 8548.070928 samples/sec	accuracy=0.655278	lr=0.638921
Epoch[242] Batch [399]	Speed: 8887.309661 samples/sec	accuracy=0.654916	lr=0.638129
Epoch[242] Batch [449]	Speed: 8846.958920 samples/sec	accuracy=0.654320	lr=0.637337
Epoch[242] Batch [499]	Speed: 8709.863490 samples/sec	accuracy=0.653925	lr=0.636545
Epoch[242] Batch [549]	Speed: 8748.153700 samples/sec	accuracy=0.653535	lr=0.635754
Epoch[242] Batch [599]	Speed: 8640.006540 samples/sec	accuracy=0.653094	lr=0.634963
[Epoch 242] training: accuracy=0.652806
[Epoch 242] speed: 8733 samples/sec	time cost: 151.419772
[Epoch 242] validation: err-top1=0.323527 err-top5=0.116882
Epoch[243] Batch [49]	Speed: 8803.053705 samples/sec	accuracy=0.662461	lr=0.633761
Epoch[243] Batch [99]	Speed: 8639.558140 samples/sec	accuracy=0.660908	lr=0.632971
Epoch[243] Batch [149]	Speed: 8608.304542 samples/sec	accuracy=0.660758	lr=0.632181
Epoch[243] Batch [199]	Speed: 8664.002281 samples/sec	accuracy=0.659819	lr=0.631392
Epoch[243] Batch [249]	Speed: 8732.917604 samples/sec	accuracy=0.658449	lr=0.630603
Epoch[243] Batch [299]	Speed: 8626.007520 samples/sec	accuracy=0.657660	lr=0.629814
Epoch[243] Batch [349]	Speed: 8772.007968 samples/sec	accuracy=0.657323	lr=0.629025
Epoch[243] Batch [399]	Speed: 8756.686420 samples/sec	accuracy=0.656565	lr=0.628237
Epoch[243] Batch [449]	Speed: 8531.489114 samples/sec	accuracy=0.655816	lr=0.627449
Epoch[243] Batch [499]	Speed: 8510.544151 samples/sec	accuracy=0.655154	lr=0.626662
Epoch[243] Batch [549]	Speed: 8548.760343 samples/sec	accuracy=0.654948	lr=0.625875
Epoch[243] Batch [599]	Speed: 8456.236420 samples/sec	accuracy=0.654464	lr=0.625088
[Epoch 243] training: accuracy=0.654295
[Epoch 243] speed: 8644 samples/sec	time cost: 153.204677
[Epoch 243] validation: err-top1=0.326523 err-top5=0.117539
Epoch[244] Batch [49]	Speed: 8450.909525 samples/sec	accuracy=0.663770	lr=0.623893
Epoch[244] Batch [99]	Speed: 8638.623081 samples/sec	accuracy=0.662783	lr=0.623107
Epoch[244] Batch [149]	Speed: 8653.834332 samples/sec	accuracy=0.661064	lr=0.622321
Epoch[244] Batch [199]	Speed: 8625.341966 samples/sec	accuracy=0.660198	lr=0.621536
Epoch[244] Batch [249]	Speed: 8735.943127 samples/sec	accuracy=0.659127	lr=0.620751
Epoch[244] Batch [299]	Speed: 8374.777840 samples/sec	accuracy=0.658595	lr=0.619967
Epoch[244] Batch [349]	Speed: 8744.529835 samples/sec	accuracy=0.657662	lr=0.619182
Epoch[244] Batch [399]	Speed: 8755.084375 samples/sec	accuracy=0.657051	lr=0.618398
Epoch[244] Batch [449]	Speed: 8532.482995 samples/sec	accuracy=0.656546	lr=0.617615
Epoch[244] Batch [499]	Speed: 8631.562720 samples/sec	accuracy=0.655745	lr=0.616832
Epoch[244] Batch [549]	Speed: 8752.357339 samples/sec	accuracy=0.655493	lr=0.616049
Epoch[244] Batch [599]	Speed: 8650.119354 samples/sec	accuracy=0.655296	lr=0.615266
[Epoch 244] training: accuracy=0.655079
[Epoch 244] speed: 8626 samples/sec	time cost: 153.002517
[Epoch 244] validation: err-top1=0.316650 err-top5=0.112935
Epoch[245] Batch [49]	Speed: 8654.279158 samples/sec	accuracy=0.665781	lr=0.614093
Epoch[245] Batch [99]	Speed: 8794.101081 samples/sec	accuracy=0.664214	lr=0.613312
Epoch[245] Batch [149]	Speed: 8749.880482 samples/sec	accuracy=0.662070	lr=0.612530
Epoch[245] Batch [199]	Speed: 8741.451927 samples/sec	accuracy=0.661094	lr=0.611749
Epoch[245] Batch [249]	Speed: 8647.405595 samples/sec	accuracy=0.659705	lr=0.610969
Epoch[245] Batch [299]	Speed: 8762.931688 samples/sec	accuracy=0.658927	lr=0.610188
Epoch[245] Batch [349]	Speed: 8631.757702 samples/sec	accuracy=0.658271	lr=0.609408
Epoch[245] Batch [399]	Speed: 8620.509643 samples/sec	accuracy=0.657903	lr=0.608629
Epoch[245] Batch [449]	Speed: 8567.217416 samples/sec	accuracy=0.657518	lr=0.607850
Epoch[245] Batch [499]	Speed: 8468.292649 samples/sec	accuracy=0.657020	lr=0.607071
Epoch[245] Batch [549]	Speed: 8723.577886 samples/sec	accuracy=0.656629	lr=0.606292
Epoch[245] Batch [599]	Speed: 8854.994690 samples/sec	accuracy=0.656134	lr=0.605514
[Epoch 245] training: accuracy=0.656033
[Epoch 245] speed: 8687 samples/sec	time cost: 152.192607
[Epoch 245] validation: err-top1=0.325073 err-top5=0.118774
Epoch[246] Batch [49]	Speed: 8613.173105 samples/sec	accuracy=0.667930	lr=0.604332
Epoch[246] Batch [99]	Speed: 8761.678741 samples/sec	accuracy=0.664692	lr=0.603554
Epoch[246] Batch [149]	Speed: 8596.994100 samples/sec	accuracy=0.663610	lr=0.602778
Epoch[246] Batch [199]	Speed: 8792.492878 samples/sec	accuracy=0.662607	lr=0.602001
Epoch[246] Batch [249]	Speed: 8844.227719 samples/sec	accuracy=0.661596	lr=0.601225
Epoch[246] Batch [299]	Speed: 8483.826693 samples/sec	accuracy=0.660905	lr=0.600449
Epoch[246] Batch [349]	Speed: 8629.177676 samples/sec	accuracy=0.660181	lr=0.599673
Epoch[246] Batch [399]	Speed: 8400.830388 samples/sec	accuracy=0.659487	lr=0.598898
Epoch[246] Batch [449]	Speed: 8620.829576 samples/sec	accuracy=0.658867	lr=0.598123
Epoch[246] Batch [499]	Speed: 8597.402640 samples/sec	accuracy=0.658678	lr=0.597349
Epoch[246] Batch [549]	Speed: 8866.606039 samples/sec	accuracy=0.658104	lr=0.596574
Epoch[246] Batch [599]	Speed: 8586.634748 samples/sec	accuracy=0.657712	lr=0.595801
[Epoch 246] training: accuracy=0.657557
[Epoch 246] speed: 8650 samples/sec	time cost: 152.790729
[Epoch 246] validation: err-top1=0.317949 err-top5=0.113809
Epoch[247] Batch [49]	Speed: 8548.702490 samples/sec	accuracy=0.667803	lr=0.594641
Epoch[247] Batch [99]	Speed: 8605.348665 samples/sec	accuracy=0.666387	lr=0.593868
Epoch[247] Batch [149]	Speed: 8696.728857 samples/sec	accuracy=0.664971	lr=0.593095
Epoch[247] Batch [199]	Speed: 8844.253762 samples/sec	accuracy=0.663271	lr=0.592323
Epoch[247] Batch [249]	Speed: 8671.372130 samples/sec	accuracy=0.661855	lr=0.591551
Epoch[247] Batch [299]	Speed: 8742.244600 samples/sec	accuracy=0.661074	lr=0.590779
Epoch[247] Batch [349]	Speed: 8721.874752 samples/sec	accuracy=0.660544	lr=0.590008
Epoch[247] Batch [399]	Speed: 8721.977304 samples/sec	accuracy=0.659640	lr=0.589237
Epoch[247] Batch [449]	Speed: 8870.487167 samples/sec	accuracy=0.658867	lr=0.588467
Epoch[247] Batch [499]	Speed: 8526.774048 samples/sec	accuracy=0.658332	lr=0.587697
Epoch[247] Batch [549]	Speed: 8675.280439 samples/sec	accuracy=0.658058	lr=0.586927
Epoch[247] Batch [599]	Speed: 8755.348694 samples/sec	accuracy=0.657590	lr=0.586158
[Epoch 247] training: accuracy=0.657346
[Epoch 247] speed: 8700 samples/sec	time cost: 152.009896
[Epoch 247] validation: err-top1=0.313944 err-top5=0.111186
Epoch[248] Batch [49]	Speed: 8390.061185 samples/sec	accuracy=0.666562	lr=0.584989
Epoch[248] Batch [99]	Speed: 8623.982939 samples/sec	accuracy=0.666138	lr=0.584221
Epoch[248] Batch [149]	Speed: 8634.214817 samples/sec	accuracy=0.665234	lr=0.583453
Epoch[248] Batch [199]	Speed: 8569.333742 samples/sec	accuracy=0.664309	lr=0.582685
Epoch[248] Batch [249]	Speed: 8713.436400 samples/sec	accuracy=0.663660	lr=0.581917
Epoch[248] Batch [299]	Speed: 8544.428334 samples/sec	accuracy=0.662949	lr=0.581150
Epoch[248] Batch [349]	Speed: 8639.255584 samples/sec	accuracy=0.662309	lr=0.580384
Epoch[248] Batch [399]	Speed: 8796.922844 samples/sec	accuracy=0.661702	lr=0.579617
Epoch[248] Batch [449]	Speed: 8769.854829 samples/sec	accuracy=0.661039	lr=0.578851
Epoch[248] Batch [499]	Speed: 8684.785125 samples/sec	accuracy=0.660528	lr=0.578086
Epoch[248] Batch [549]	Speed: 8622.591450 samples/sec	accuracy=0.659802	lr=0.577321
Epoch[248] Batch [599]	Speed: 8405.891630 samples/sec	accuracy=0.659428	lr=0.576556
[Epoch 248] training: accuracy=0.659320
[Epoch 248] speed: 8616 samples/sec	time cost: 153.386314
[Epoch 248] validation: err-top1=0.309805 err-top5=0.108770
Epoch[249] Batch [49]	Speed: 8855.503162 samples/sec	accuracy=0.670107	lr=0.575409
Epoch[249] Batch [99]	Speed: 8347.608038 samples/sec	accuracy=0.666543	lr=0.574645
Epoch[249] Batch [149]	Speed: 8819.508696 samples/sec	accuracy=0.664984	lr=0.573882
Epoch[249] Batch [199]	Speed: 8526.497282 samples/sec	accuracy=0.664548	lr=0.573118
Epoch[249] Batch [249]	Speed: 8634.381626 samples/sec	accuracy=0.663412	lr=0.572356
Epoch[249] Batch [299]	Speed: 8866.053464 samples/sec	accuracy=0.662373	lr=0.571593
Epoch[249] Batch [349]	Speed: 8738.019201 samples/sec	accuracy=0.661808	lr=0.570831
Epoch[249] Batch [399]	Speed: 8360.133093 samples/sec	accuracy=0.661696	lr=0.570069
Epoch[249] Batch [449]	Speed: 8241.735920 samples/sec	accuracy=0.661484	lr=0.569308
Epoch[249] Batch [499]	Speed: 8603.105947 samples/sec	accuracy=0.660979	lr=0.568547
Epoch[249] Batch [549]	Speed: 8620.857435 samples/sec	accuracy=0.660291	lr=0.567786
Epoch[249] Batch [599]	Speed: 8518.395573 samples/sec	accuracy=0.659947	lr=0.567026
[Epoch 249] training: accuracy=0.659654
[Epoch 249] speed: 8588 samples/sec	time cost: 153.939426
[Epoch 249] validation: err-top1=0.310954 err-top5=0.109253
Epoch[250] Batch [49]	Speed: 8818.410616 samples/sec	accuracy=0.667881	lr=0.565871
Epoch[250] Batch [99]	Speed: 8439.132194 samples/sec	accuracy=0.666787	lr=0.565111
Epoch[250] Batch [149]	Speed: 8531.007341 samples/sec	accuracy=0.665537	lr=0.564352
Epoch[250] Batch [199]	Speed: 8553.768915 samples/sec	accuracy=0.664778	lr=0.563594
Epoch[250] Batch [249]	Speed: 8605.448322 samples/sec	accuracy=0.664123	lr=0.562836
Epoch[250] Batch [299]	Speed: 8437.947912 samples/sec	accuracy=0.663921	lr=0.562078
Epoch[250] Batch [349]	Speed: 8714.853653 samples/sec	accuracy=0.663341	lr=0.561320
Epoch[250] Batch [399]	Speed: 8526.294162 samples/sec	accuracy=0.663140	lr=0.560563
Epoch[250] Batch [449]	Speed: 8743.253131 samples/sec	accuracy=0.662551	lr=0.559806
Epoch[250] Batch [499]	Speed: 8642.798094 samples/sec	accuracy=0.662108	lr=0.559050
Epoch[250] Batch [549]	Speed: 8627.169629 samples/sec	accuracy=0.661490	lr=0.558294
Epoch[250] Batch [599]	Speed: 8499.714878 samples/sec	accuracy=0.660950	lr=0.557538
[Epoch 250] training: accuracy=0.660836
[Epoch 250] speed: 8598 samples/sec	time cost: 153.748409
[Epoch 250] validation: err-top1=0.329020 err-top5=0.119141
Epoch[251] Batch [49]	Speed: 8518.992174 samples/sec	accuracy=0.672705	lr=0.556390
Epoch[251] Batch [99]	Speed: 8480.030681 samples/sec	accuracy=0.670513	lr=0.555635
Epoch[251] Batch [149]	Speed: 8738.175111 samples/sec	accuracy=0.668512	lr=0.554881
Epoch[251] Batch [199]	Speed: 8802.875986 samples/sec	accuracy=0.667336	lr=0.554127
Epoch[251] Batch [249]	Speed: 8877.845611 samples/sec	accuracy=0.666590	lr=0.553373
Epoch[251] Batch [299]	Speed: 8810.260349 samples/sec	accuracy=0.665879	lr=0.552620
Epoch[251] Batch [349]	Speed: 8764.634262 samples/sec	accuracy=0.665248	lr=0.551867
Epoch[251] Batch [399]	Speed: 8623.086739 samples/sec	accuracy=0.664727	lr=0.551115
Epoch[251] Batch [449]	Speed: 8634.224364 samples/sec	accuracy=0.664182	lr=0.550363
Epoch[251] Batch [499]	Speed: 8555.740719 samples/sec	accuracy=0.663875	lr=0.549611
Epoch[251] Batch [549]	Speed: 8551.084620 samples/sec	accuracy=0.663251	lr=0.548860
Epoch[251] Batch [599]	Speed: 8609.137618 samples/sec	accuracy=0.662981	lr=0.548109
[Epoch 251] training: accuracy=0.662698
[Epoch 251] speed: 8661 samples/sec	time cost: 152.683639
[Epoch 251] validation: err-top1=0.315625 err-top5=0.111191
Epoch[252] Batch [49]	Speed: 8532.253995 samples/sec	accuracy=0.669893	lr=0.546983
Epoch[252] Batch [99]	Speed: 8586.956120 samples/sec	accuracy=0.670347	lr=0.546233
Epoch[252] Batch [149]	Speed: 8630.512327 samples/sec	accuracy=0.669736	lr=0.545483
Epoch[252] Batch [199]	Speed: 8505.264514 samples/sec	accuracy=0.668882	lr=0.544734
Epoch[252] Batch [249]	Speed: 8673.485234 samples/sec	accuracy=0.667803	lr=0.543985
Epoch[252] Batch [299]	Speed: 8319.836588 samples/sec	accuracy=0.666862	lr=0.543236
Epoch[252] Batch [349]	Speed: 8601.255909 samples/sec	accuracy=0.666030	lr=0.542488
Epoch[252] Batch [399]	Speed: 8668.731104 samples/sec	accuracy=0.665483	lr=0.541740
Epoch[252] Batch [449]	Speed: 8688.047848 samples/sec	accuracy=0.664730	lr=0.540993
Epoch[252] Batch [499]	Speed: 8832.914830 samples/sec	accuracy=0.664492	lr=0.540246
Epoch[252] Batch [549]	Speed: 8550.279933 samples/sec	accuracy=0.663924	lr=0.539499
Epoch[252] Batch [599]	Speed: 8594.626743 samples/sec	accuracy=0.663461	lr=0.538753
[Epoch 252] training: accuracy=0.663086
[Epoch 252] speed: 8595 samples/sec	time cost: 153.866286
[Epoch 252] validation: err-top1=0.315796 err-top5=0.111837
Epoch[253] Batch [49]	Speed: 8740.122408 samples/sec	accuracy=0.671836	lr=0.537620
Epoch[253] Batch [99]	Speed: 8494.092014 samples/sec	accuracy=0.671816	lr=0.536874
Epoch[253] Batch [149]	Speed: 8837.416363 samples/sec	accuracy=0.670684	lr=0.536129
Epoch[253] Batch [199]	Speed: 8520.487168 samples/sec	accuracy=0.669744	lr=0.535385
Epoch[253] Batch [249]	Speed: 8626.899996 samples/sec	accuracy=0.668961	lr=0.534641
Epoch[253] Batch [299]	Speed: 8625.297795 samples/sec	accuracy=0.667438	lr=0.533897
Epoch[253] Batch [349]	Speed: 8342.551737 samples/sec	accuracy=0.667266	lr=0.533154
Epoch[253] Batch [399]	Speed: 8616.707218 samples/sec	accuracy=0.666796	lr=0.532411
Epoch[253] Batch [449]	Speed: 8950.562978 samples/sec	accuracy=0.666288	lr=0.531668
Epoch[253] Batch [499]	Speed: 8431.965595 samples/sec	accuracy=0.665604	lr=0.530926
Epoch[253] Batch [549]	Speed: 8696.300434 samples/sec	accuracy=0.665092	lr=0.530184
Epoch[253] Batch [599]	Speed: 8719.699592 samples/sec	accuracy=0.664836	lr=0.529442
[Epoch 253] training: accuracy=0.664642
[Epoch 253] speed: 8637 samples/sec	time cost: 153.272414
[Epoch 253] validation: err-top1=0.306953 err-top5=0.106074
Epoch[254] Batch [49]	Speed: 8604.267581 samples/sec	accuracy=0.674580	lr=0.528331
Epoch[254] Batch [99]	Speed: 8743.603778 samples/sec	accuracy=0.673457	lr=0.527591
Epoch[254] Batch [149]	Speed: 8865.341204 samples/sec	accuracy=0.671973	lr=0.526850
Epoch[254] Batch [199]	Speed: 8918.606096 samples/sec	accuracy=0.670820	lr=0.526111
Epoch[254] Batch [249]	Speed: 8679.018643 samples/sec	accuracy=0.669854	lr=0.525371
Epoch[254] Batch [299]	Speed: 8673.056997 samples/sec	accuracy=0.668859	lr=0.524633
Epoch[254] Batch [349]	Speed: 8664.311817 samples/sec	accuracy=0.668297	lr=0.523894
Epoch[254] Batch [399]	Speed: 8692.126181 samples/sec	accuracy=0.667894	lr=0.523156
Epoch[254] Batch [449]	Speed: 8855.451125 samples/sec	accuracy=0.667661	lr=0.522418
Epoch[254] Batch [499]	Speed: 8775.644966 samples/sec	accuracy=0.667129	lr=0.521681
Epoch[254] Batch [549]	Speed: 8897.372258 samples/sec	accuracy=0.666329	lr=0.520944
Epoch[254] Batch [599]	Speed: 8671.406795 samples/sec	accuracy=0.665941	lr=0.520207
[Epoch 254] training: accuracy=0.665851
[Epoch 254] speed: 8758 samples/sec	time cost: 151.002905
[Epoch 254] validation: err-top1=0.326884 err-top5=0.119914
Epoch[255] Batch [49]	Speed: 8675.588329 samples/sec	accuracy=0.674775	lr=0.519088
Epoch[255] Batch [99]	Speed: 8833.100849 samples/sec	accuracy=0.672534	lr=0.518353
Epoch[255] Batch [149]	Speed: 8585.688797 samples/sec	accuracy=0.671465	lr=0.517617
Epoch[255] Batch [199]	Speed: 8660.364312 samples/sec	accuracy=0.670703	lr=0.516883
Epoch[255] Batch [249]	Speed: 8891.319193 samples/sec	accuracy=0.669900	lr=0.516148
Epoch[255] Batch [299]	Speed: 8637.784984 samples/sec	accuracy=0.669365	lr=0.515414
Epoch[255] Batch [349]	Speed: 8828.724757 samples/sec	accuracy=0.668832	lr=0.514681
Epoch[255] Batch [399]	Speed: 8946.620982 samples/sec	accuracy=0.668032	lr=0.513947
Epoch[255] Batch [449]	Speed: 8886.518412 samples/sec	accuracy=0.667560	lr=0.513214
Epoch[255] Batch [499]	Speed: 8754.684625 samples/sec	accuracy=0.667224	lr=0.512482
Epoch[255] Batch [549]	Speed: 8360.504295 samples/sec	accuracy=0.666831	lr=0.511750
Epoch[255] Batch [599]	Speed: 8634.939899 samples/sec	accuracy=0.666555	lr=0.511018
[Epoch 255] training: accuracy=0.666323
[Epoch 255] speed: 8726 samples/sec	time cost: 151.289389
[Epoch 255] validation: err-top1=0.317383 err-top5=0.113363
Epoch[256] Batch [49]	Speed: 8786.966140 samples/sec	accuracy=0.675664	lr=0.509921
Epoch[256] Batch [99]	Speed: 8844.312588 samples/sec	accuracy=0.675308	lr=0.509191
Epoch[256] Batch [149]	Speed: 8580.117050 samples/sec	accuracy=0.674063	lr=0.508460
Epoch[256] Batch [199]	Speed: 8683.122207 samples/sec	accuracy=0.673000	lr=0.507731
Epoch[256] Batch [249]	Speed: 8439.612933 samples/sec	accuracy=0.671770	lr=0.507001
Epoch[256] Batch [299]	Speed: 8483.724805 samples/sec	accuracy=0.670825	lr=0.506272
Epoch[256] Batch [349]	Speed: 8548.692451 samples/sec	accuracy=0.670212	lr=0.505543
Epoch[256] Batch [399]	Speed: 8609.319335 samples/sec	accuracy=0.669646	lr=0.504815
Epoch[256] Batch [449]	Speed: 8500.002694 samples/sec	accuracy=0.669073	lr=0.504087
Epoch[256] Batch [499]	Speed: 8581.229450 samples/sec	accuracy=0.668769	lr=0.503360
Epoch[256] Batch [549]	Speed: 8549.568999 samples/sec	accuracy=0.668477	lr=0.502632
Epoch[256] Batch [599]	Speed: 8562.807105 samples/sec	accuracy=0.667937	lr=0.501906
[Epoch 256] training: accuracy=0.667839
[Epoch 256] speed: 8596 samples/sec	time cost: 154.132029
[Epoch 256] validation: err-top1=0.316875 err-top5=0.114375
Epoch[257] Batch [49]	Speed: 8349.635100 samples/sec	accuracy=0.675732	lr=0.500802
Epoch[257] Batch [99]	Speed: 8286.361901 samples/sec	accuracy=0.675020	lr=0.500076
Epoch[257] Batch [149]	Speed: 8573.471469 samples/sec	accuracy=0.673844	lr=0.499351
Epoch[257] Batch [199]	Speed: 8843.971118 samples/sec	accuracy=0.673418	lr=0.498626
Epoch[257] Batch [249]	Speed: 8546.875944 samples/sec	accuracy=0.672178	lr=0.497902
Epoch[257] Batch [299]	Speed: 8726.862379 samples/sec	accuracy=0.671198	lr=0.497178
Epoch[257] Batch [349]	Speed: 8520.668713 samples/sec	accuracy=0.670566	lr=0.496454
Epoch[257] Batch [399]	Speed: 8733.821862 samples/sec	accuracy=0.669857	lr=0.495731
Epoch[257] Batch [449]	Speed: 8844.859542 samples/sec	accuracy=0.669339	lr=0.495008
Epoch[257] Batch [499]	Speed: 8613.827282 samples/sec	accuracy=0.668855	lr=0.494285
Epoch[257] Batch [549]	Speed: 8869.635899 samples/sec	accuracy=0.668408	lr=0.493563
Epoch[257] Batch [599]	Speed: 8652.910823 samples/sec	accuracy=0.668117	lr=0.492841
[Epoch 257] training: accuracy=0.668104
[Epoch 257] speed: 8629 samples/sec	time cost: 153.106642
[Epoch 257] validation: err-top1=0.327759 err-top5=0.119466
Epoch[258] Batch [49]	Speed: 8795.971607 samples/sec	accuracy=0.676953	lr=0.491745
Epoch[258] Batch [99]	Speed: 8555.782305 samples/sec	accuracy=0.674893	lr=0.491025
Epoch[258] Batch [149]	Speed: 8532.066364 samples/sec	accuracy=0.673919	lr=0.490304
Epoch[258] Batch [199]	Speed: 8709.618158 samples/sec	accuracy=0.673091	lr=0.489585
Epoch[258] Batch [249]	Speed: 8563.537658 samples/sec	accuracy=0.672756	lr=0.488865
Epoch[258] Batch [299]	Speed: 8806.315088 samples/sec	accuracy=0.672313	lr=0.488146
Epoch[258] Batch [349]	Speed: 8742.335709 samples/sec	accuracy=0.671685	lr=0.487427
Epoch[258] Batch [399]	Speed: 8811.583272 samples/sec	accuracy=0.671173	lr=0.486709
Epoch[258] Batch [449]	Speed: 8467.077634 samples/sec	accuracy=0.670557	lr=0.485991
Epoch[258] Batch [499]	Speed: 8704.831579 samples/sec	accuracy=0.669791	lr=0.485274
Epoch[258] Batch [549]	Speed: 9028.919389 samples/sec	accuracy=0.669523	lr=0.484557
Epoch[258] Batch [599]	Speed: 8586.357000 samples/sec	accuracy=0.669112	lr=0.483841
[Epoch 258] training: accuracy=0.669016
[Epoch 258] speed: 8697 samples/sec	time cost: 151.899921
[Epoch 258] validation: err-top1=0.319336 err-top5=0.115234
Epoch[259] Batch [49]	Speed: 8402.007232 samples/sec	accuracy=0.679668	lr=0.482766
Epoch[259] Batch [99]	Speed: 8609.120879 samples/sec	accuracy=0.678662	lr=0.482051
Epoch[259] Batch [149]	Speed: 8560.785631 samples/sec	accuracy=0.677376	lr=0.481336
Epoch[259] Batch [199]	Speed: 8737.275107 samples/sec	accuracy=0.676558	lr=0.480621
Epoch[259] Batch [249]	Speed: 8725.126235 samples/sec	accuracy=0.674988	lr=0.479907
Epoch[259] Batch [299]	Speed: 8794.994641 samples/sec	accuracy=0.674307	lr=0.479193
Epoch[259] Batch [349]	Speed: 8651.936096 samples/sec	accuracy=0.673489	lr=0.478479
Epoch[259] Batch [399]	Speed: 8640.067547 samples/sec	accuracy=0.672809	lr=0.477766
Epoch[259] Batch [449]	Speed: 8611.661469 samples/sec	accuracy=0.672493	lr=0.477053
Epoch[259] Batch [499]	Speed: 8530.602715 samples/sec	accuracy=0.672041	lr=0.476341
Epoch[259] Batch [549]	Speed: 8621.143130 samples/sec	accuracy=0.671588	lr=0.475629
Epoch[259] Batch [599]	Speed: 8805.145918 samples/sec	accuracy=0.671411	lr=0.474918
[Epoch 259] training: accuracy=0.671374
[Epoch 259] speed: 8639 samples/sec	time cost: 152.984578
[Epoch 259] validation: err-top1=0.320882 err-top5=0.115479
Epoch[260] Batch [49]	Speed: 8288.999802 samples/sec	accuracy=0.679160	lr=0.473837
Epoch[260] Batch [99]	Speed: 8448.786960 samples/sec	accuracy=0.679321	lr=0.473127
Epoch[260] Batch [149]	Speed: 8453.802004 samples/sec	accuracy=0.677633	lr=0.472417
Epoch[260] Batch [199]	Speed: 8506.802039 samples/sec	accuracy=0.676548	lr=0.471707
Epoch[260] Batch [249]	Speed: 8901.080771 samples/sec	accuracy=0.675895	lr=0.470998
Epoch[260] Batch [299]	Speed: 8753.993178 samples/sec	accuracy=0.675550	lr=0.470290
Epoch[260] Batch [349]	Speed: 8410.430933 samples/sec	accuracy=0.674646	lr=0.469581
Epoch[260] Batch [399]	Speed: 8725.225673 samples/sec	accuracy=0.674082	lr=0.468873
Epoch[260] Batch [449]	Speed: 8506.564644 samples/sec	accuracy=0.673469	lr=0.468166
Epoch[260] Batch [499]	Speed: 8847.528071 samples/sec	accuracy=0.672869	lr=0.467459
Epoch[260] Batch [549]	Speed: 8643.779286 samples/sec	accuracy=0.672338	lr=0.466752
Epoch[260] Batch [599]	Speed: 8637.754410 samples/sec	accuracy=0.671742	lr=0.466046
[Epoch 260] training: accuracy=0.671456
[Epoch 260] speed: 8594 samples/sec	time cost: 153.714021
[Epoch 260] validation: err-top1=0.303301 err-top5=0.103984
Epoch[261] Batch [49]	Speed: 8673.788091 samples/sec	accuracy=0.683066	lr=0.464987
Epoch[261] Batch [99]	Speed: 8497.936442 samples/sec	accuracy=0.679341	lr=0.464282
Epoch[261] Batch [149]	Speed: 8609.830878 samples/sec	accuracy=0.677985	lr=0.463577
Epoch[261] Batch [199]	Speed: 8462.760959 samples/sec	accuracy=0.677908	lr=0.462873
Epoch[261] Batch [249]	Speed: 8852.557210 samples/sec	accuracy=0.677088	lr=0.462169
Epoch[261] Batch [299]	Speed: 8774.865945 samples/sec	accuracy=0.675990	lr=0.461466
Epoch[261] Batch [349]	Speed: 8785.965652 samples/sec	accuracy=0.675293	lr=0.460762
Epoch[261] Batch [399]	Speed: 8540.932699 samples/sec	accuracy=0.674633	lr=0.460060
Epoch[261] Batch [449]	Speed: 8508.595314 samples/sec	accuracy=0.673892	lr=0.459358
Epoch[261] Batch [499]	Speed: 8617.641862 samples/sec	accuracy=0.673496	lr=0.458656
Epoch[261] Batch [549]	Speed: 8575.603900 samples/sec	accuracy=0.673077	lr=0.457954
Epoch[261] Batch [599]	Speed: 8624.983939 samples/sec	accuracy=0.672637	lr=0.457253
[Epoch 261] training: accuracy=0.672553
[Epoch 261] speed: 8616 samples/sec	time cost: 153.357468
[Epoch 261] validation: err-top1=0.326538 err-top5=0.119527
Epoch[262] Batch [49]	Speed: 8382.644207 samples/sec	accuracy=0.678916	lr=0.456189
Epoch[262] Batch [99]	Speed: 8883.902573 samples/sec	accuracy=0.679082	lr=0.455489
Epoch[262] Batch [149]	Speed: 8480.222393 samples/sec	accuracy=0.679417	lr=0.454789
Epoch[262] Batch [199]	Speed: 8867.324910 samples/sec	accuracy=0.678440	lr=0.454090
Epoch[262] Batch [249]	Speed: 8748.265780 samples/sec	accuracy=0.677008	lr=0.453391
Epoch[262] Batch [299]	Speed: 8693.193210 samples/sec	accuracy=0.676859	lr=0.452693
Epoch[262] Batch [349]	Speed: 8772.697067 samples/sec	accuracy=0.676197	lr=0.451995
Epoch[262] Batch [399]	Speed: 8549.468249 samples/sec	accuracy=0.676163	lr=0.451298
Epoch[262] Batch [449]	Speed: 8520.092328 samples/sec	accuracy=0.675536	lr=0.450601
Epoch[262] Batch [499]	Speed: 8534.536065 samples/sec	accuracy=0.675091	lr=0.449905
Epoch[262] Batch [549]	Speed: 8667.173845 samples/sec	accuracy=0.674466	lr=0.449209
Epoch[262] Batch [599]	Speed: 8598.390249 samples/sec	accuracy=0.673991	lr=0.448513
[Epoch 262] training: accuracy=0.673791
[Epoch 262] speed: 8646 samples/sec	time cost: 152.726228
[Epoch 262] validation: err-top1=0.312500 err-top5=0.109843
Epoch[263] Batch [49]	Speed: 8569.549519 samples/sec	accuracy=0.681807	lr=0.447470
Epoch[263] Batch [99]	Speed: 8582.257248 samples/sec	accuracy=0.681680	lr=0.446776
Epoch[263] Batch [149]	Speed: 8737.990936 samples/sec	accuracy=0.681543	lr=0.446081
Epoch[263] Batch [199]	Speed: 8588.505808 samples/sec	accuracy=0.680867	lr=0.445388
Epoch[263] Batch [249]	Speed: 8530.234721 samples/sec	accuracy=0.679992	lr=0.444694
Epoch[263] Batch [299]	Speed: 8586.332454 samples/sec	accuracy=0.679170	lr=0.444001
Epoch[263] Batch [349]	Speed: 8827.079376 samples/sec	accuracy=0.678295	lr=0.443309
Epoch[263] Batch [399]	Speed: 8668.668642 samples/sec	accuracy=0.677328	lr=0.442617
Epoch[263] Batch [449]	Speed: 8710.456828 samples/sec	accuracy=0.676587	lr=0.441925
Epoch[263] Batch [499]	Speed: 8640.337308 samples/sec	accuracy=0.676097	lr=0.441234
Epoch[263] Batch [549]	Speed: 8557.894520 samples/sec	accuracy=0.675684	lr=0.440543
Epoch[263] Batch [599]	Speed: 8680.426125 samples/sec	accuracy=0.675431	lr=0.439853
[Epoch 263] training: accuracy=0.675180
[Epoch 263] speed: 8634 samples/sec	time cost: 153.355368
[Epoch 263] validation: err-top1=0.315742 err-top5=0.111953
Epoch[264] Batch [49]	Speed: 8283.768177 samples/sec	accuracy=0.682725	lr=0.438805
Epoch[264] Batch [99]	Speed: 8652.785658 samples/sec	accuracy=0.682646	lr=0.438115
Epoch[264] Batch [149]	Speed: 8381.865346 samples/sec	accuracy=0.681689	lr=0.437427
Epoch[264] Batch [199]	Speed: 8551.998099 samples/sec	accuracy=0.680640	lr=0.436738
Epoch[264] Batch [249]	Speed: 8899.161591 samples/sec	accuracy=0.680168	lr=0.436050
Epoch[264] Batch [299]	Speed: 8741.339309 samples/sec	accuracy=0.679476	lr=0.435363
Epoch[264] Batch [349]	Speed: 8800.318887 samples/sec	accuracy=0.678743	lr=0.434676
Epoch[264] Batch [399]	Speed: 8755.096154 samples/sec	accuracy=0.678009	lr=0.433989
Epoch[264] Batch [449]	Speed: 8653.323823 samples/sec	accuracy=0.677543	lr=0.433303
Epoch[264] Batch [499]	Speed: 8659.783365 samples/sec	accuracy=0.677271	lr=0.432617
Epoch[264] Batch [549]	Speed: 8728.268217 samples/sec	accuracy=0.676817	lr=0.431932
Epoch[264] Batch [599]	Speed: 8627.005352 samples/sec	accuracy=0.676339	lr=0.431247
[Epoch 264] training: accuracy=0.676078
[Epoch 264] speed: 8642 samples/sec	time cost: 152.954458
[Epoch 264] validation: err-top1=0.307719 err-top5=0.106018
Epoch[265] Batch [49]	Speed: 8507.546828 samples/sec	accuracy=0.685225	lr=0.430207
Epoch[265] Batch [99]	Speed: 8705.404470 samples/sec	accuracy=0.683438	lr=0.429523
Epoch[265] Batch [149]	Speed: 8517.968322 samples/sec	accuracy=0.682822	lr=0.428840
Epoch[265] Batch [199]	Speed: 8583.156473 samples/sec	accuracy=0.681611	lr=0.428157
Epoch[265] Batch [249]	Speed: 8639.800931 samples/sec	accuracy=0.680832	lr=0.427474
Epoch[265] Batch [299]	Speed: 8464.064974 samples/sec	accuracy=0.680400	lr=0.426792
Epoch[265] Batch [349]	Speed: 8487.535364 samples/sec	accuracy=0.679704	lr=0.426111
Epoch[265] Batch [399]	Speed: 8793.376209 samples/sec	accuracy=0.679037	lr=0.425429
Epoch[265] Batch [449]	Speed: 8566.916999 samples/sec	accuracy=0.678416	lr=0.424749
Epoch[265] Batch [499]	Speed: 8744.689716 samples/sec	accuracy=0.677902	lr=0.424068
Epoch[265] Batch [549]	Speed: 8723.653014 samples/sec	accuracy=0.677612	lr=0.423389
Epoch[265] Batch [599]	Speed: 8495.683984 samples/sec	accuracy=0.677196	lr=0.422709
[Epoch 265] training: accuracy=0.677086
[Epoch 265] speed: 8589 samples/sec	time cost: 153.919548
[Epoch 265] validation: err-top1=0.308730 err-top5=0.107051
Epoch[266] Batch [49]	Speed: 8552.363716 samples/sec	accuracy=0.688193	lr=0.421691
Epoch[266] Batch [99]	Speed: 8533.222285 samples/sec	accuracy=0.686870	lr=0.421013
Epoch[266] Batch [149]	Speed: 8516.609649 samples/sec	accuracy=0.685684	lr=0.420335
Epoch[266] Batch [199]	Speed: 8570.413932 samples/sec	accuracy=0.684380	lr=0.419657
Epoch[266] Batch [249]	Speed: 8579.222061 samples/sec	accuracy=0.682607	lr=0.418980
Epoch[266] Batch [299]	Speed: 8711.360857 samples/sec	accuracy=0.681673	lr=0.418304
Epoch[266] Batch [349]	Speed: 8817.377070 samples/sec	accuracy=0.681378	lr=0.417628
Epoch[266] Batch [399]	Speed: 8787.789745 samples/sec	accuracy=0.680521	lr=0.416952
Epoch[266] Batch [449]	Speed: 8509.060566 samples/sec	accuracy=0.679929	lr=0.416277
Epoch[266] Batch [499]	Speed: 8573.970374 samples/sec	accuracy=0.679310	lr=0.415602
Epoch[266] Batch [549]	Speed: 8859.310248 samples/sec	accuracy=0.679004	lr=0.414928
Epoch[266] Batch [599]	Speed: 8705.742735 samples/sec	accuracy=0.678531	lr=0.414254
[Epoch 266] training: accuracy=0.678127
[Epoch 266] speed: 8643 samples/sec	time cost: 152.691940
[Epoch 266] validation: err-top1=0.301737 err-top5=0.104146
Epoch[267] Batch [49]	Speed: 8476.958604 samples/sec	accuracy=0.690957	lr=0.413230
Epoch[267] Batch [99]	Speed: 8714.111199 samples/sec	accuracy=0.688379	lr=0.412558
Epoch[267] Batch [149]	Speed: 8697.132842 samples/sec	accuracy=0.686706	lr=0.411885
Epoch[267] Batch [199]	Speed: 8895.369559 samples/sec	accuracy=0.685525	lr=0.411213
Epoch[267] Batch [249]	Speed: 8915.237708 samples/sec	accuracy=0.684193	lr=0.410542
Epoch[267] Batch [299]	Speed: 8703.185313 samples/sec	accuracy=0.683638	lr=0.409871
Epoch[267] Batch [349]	Speed: 8859.731672 samples/sec	accuracy=0.682874	lr=0.409200
Epoch[267] Batch [399]	Speed: 8755.779742 samples/sec	accuracy=0.682135	lr=0.408530
Epoch[267] Batch [449]	Speed: 8827.268777 samples/sec	accuracy=0.681581	lr=0.407861
Epoch[267] Batch [499]	Speed: 8483.877302 samples/sec	accuracy=0.681064	lr=0.407191
Epoch[267] Batch [549]	Speed: 8749.392979 samples/sec	accuracy=0.680408	lr=0.406523
Epoch[267] Batch [599]	Speed: 8512.844992 samples/sec	accuracy=0.679936	lr=0.405854
[Epoch 267] training: accuracy=0.679691
[Epoch 267] speed: 8707 samples/sec	time cost: 151.613123
[Epoch 267] validation: err-top1=0.302917 err-top5=0.106445
Epoch[268] Batch [49]	Speed: 8350.408469 samples/sec	accuracy=0.687822	lr=0.404853
Epoch[268] Batch [99]	Speed: 8753.055484 samples/sec	accuracy=0.686323	lr=0.404186
Epoch[268] Batch [149]	Speed: 8672.792544 samples/sec	accuracy=0.686536	lr=0.403519
Epoch[268] Batch [199]	Speed: 8833.383525 samples/sec	accuracy=0.685132	lr=0.402852
Epoch[268] Batch [249]	Speed: 8583.857221 samples/sec	accuracy=0.684553	lr=0.402187
Epoch[268] Batch [299]	Speed: 8534.649183 samples/sec	accuracy=0.683867	lr=0.401521
Epoch[268] Batch [349]	Speed: 8588.011736 samples/sec	accuracy=0.683181	lr=0.400856
Epoch[268] Batch [399]	Speed: 8356.900256 samples/sec	accuracy=0.682571	lr=0.400192
Epoch[268] Batch [449]	Speed: 8797.993051 samples/sec	accuracy=0.682153	lr=0.399528
Epoch[268] Batch [499]	Speed: 8762.594149 samples/sec	accuracy=0.681531	lr=0.398864
Epoch[268] Batch [549]	Speed: 8854.944850 samples/sec	accuracy=0.680981	lr=0.398201
Epoch[268] Batch [599]	Speed: 8814.724162 samples/sec	accuracy=0.680402	lr=0.397538
[Epoch 268] training: accuracy=0.680173
[Epoch 268] speed: 8654 samples/sec	time cost: 152.880434
[Epoch 268] validation: err-top1=0.302090 err-top5=0.104648
Epoch[269] Batch [49]	Speed: 8532.367731 samples/sec	accuracy=0.688398	lr=0.396532
Epoch[269] Batch [99]	Speed: 8739.637592 samples/sec	accuracy=0.686987	lr=0.395870
Epoch[269] Batch [149]	Speed: 8688.344341 samples/sec	accuracy=0.685550	lr=0.395209
Epoch[269] Batch [199]	Speed: 8749.403495 samples/sec	accuracy=0.684902	lr=0.394549
Epoch[269] Batch [249]	Speed: 8921.861011 samples/sec	accuracy=0.683850	lr=0.393889
Epoch[269] Batch [299]	Speed: 8485.410625 samples/sec	accuracy=0.683460	lr=0.393229
Epoch[269] Batch [349]	Speed: 8216.849840 samples/sec	accuracy=0.682520	lr=0.392570
Epoch[269] Batch [399]	Speed: 8601.501030 samples/sec	accuracy=0.681857	lr=0.391911
Epoch[269] Batch [449]	Speed: 8504.561719 samples/sec	accuracy=0.681851	lr=0.391252
Epoch[269] Batch [499]	Speed: 8514.992103 samples/sec	accuracy=0.681700	lr=0.390594
Epoch[269] Batch [549]	Speed: 8633.633035 samples/sec	accuracy=0.681411	lr=0.389937
Epoch[269] Batch [599]	Speed: 8716.487885 samples/sec	accuracy=0.680977	lr=0.389280
[Epoch 269] training: accuracy=0.680773
[Epoch 269] speed: 8610 samples/sec	time cost: 153.227330
[Epoch 269] validation: err-top1=0.307739 err-top5=0.106893
Epoch[270] Batch [49]	Speed: 8349.272165 samples/sec	accuracy=0.690830	lr=0.388295
Epoch[270] Batch [99]	Speed: 8335.396748 samples/sec	accuracy=0.690659	lr=0.387639
Epoch[270] Batch [149]	Speed: 8632.183433 samples/sec	accuracy=0.689333	lr=0.386984
Epoch[270] Batch [199]	Speed: 8425.441331 samples/sec	accuracy=0.688677	lr=0.386329
Epoch[270] Batch [249]	Speed: 8689.300567 samples/sec	accuracy=0.687123	lr=0.385675
Epoch[270] Batch [299]	Speed: 8598.803914 samples/sec	accuracy=0.685998	lr=0.385021
Epoch[270] Batch [349]	Speed: 8608.734346 samples/sec	accuracy=0.685352	lr=0.384367
Epoch[270] Batch [399]	Speed: 8488.684953 samples/sec	accuracy=0.684500	lr=0.383714
Epoch[270] Batch [449]	Speed: 8681.565214 samples/sec	accuracy=0.684055	lr=0.383061
Epoch[270] Batch [499]	Speed: 8723.048841 samples/sec	accuracy=0.683782	lr=0.382409
Epoch[270] Batch [549]	Speed: 8562.947436 samples/sec	accuracy=0.683271	lr=0.381757
Epoch[270] Batch [599]	Speed: 8725.213088 samples/sec	accuracy=0.682873	lr=0.381106
[Epoch 270] training: accuracy=0.682634
[Epoch 270] speed: 8568 samples/sec	time cost: 154.391470
[Epoch 270] validation: err-top1=0.310469 err-top5=0.109199
Epoch[271] Batch [49]	Speed: 8445.183804 samples/sec	accuracy=0.692305	lr=0.380117
Epoch[271] Batch [99]	Speed: 8628.490484 samples/sec	accuracy=0.690000	lr=0.379467
Epoch[271] Batch [149]	Speed: 8346.239262 samples/sec	accuracy=0.689495	lr=0.378817
Epoch[271] Batch [199]	Speed: 8915.685384 samples/sec	accuracy=0.688457	lr=0.378168
Epoch[271] Batch [249]	Speed: 8677.092340 samples/sec	accuracy=0.687934	lr=0.377519
Epoch[271] Batch [299]	Speed: 8709.118178 samples/sec	accuracy=0.686851	lr=0.376871
Epoch[271] Batch [349]	Speed: 8589.838729 samples/sec	accuracy=0.686482	lr=0.376223
Epoch[271] Batch [399]	Speed: 8692.263394 samples/sec	accuracy=0.685841	lr=0.375576
Epoch[271] Batch [449]	Speed: 8784.202860 samples/sec	accuracy=0.685361	lr=0.374929
Epoch[271] Batch [499]	Speed: 8655.036740 samples/sec	accuracy=0.685053	lr=0.374283
Epoch[271] Batch [549]	Speed: 8452.278752 samples/sec	accuracy=0.684714	lr=0.373637
Epoch[271] Batch [599]	Speed: 8689.572533 samples/sec	accuracy=0.684357	lr=0.372991
[Epoch 271] training: accuracy=0.683971
[Epoch 271] speed: 8632 samples/sec	time cost: 152.842310
[Epoch 271] validation: err-top1=0.301229 err-top5=0.102071
Epoch[272] Batch [49]	Speed: 8649.588033 samples/sec	accuracy=0.691455	lr=0.372024
Epoch[272] Batch [99]	Speed: 8799.121024 samples/sec	accuracy=0.690527	lr=0.371380
Epoch[272] Batch [149]	Speed: 8730.493264 samples/sec	accuracy=0.689756	lr=0.370736
Epoch[272] Batch [199]	Speed: 8651.310623 samples/sec	accuracy=0.688472	lr=0.370093
Epoch[272] Batch [249]	Speed: 8617.934952 samples/sec	accuracy=0.687779	lr=0.369450
Epoch[272] Batch [299]	Speed: 8670.896312 samples/sec	accuracy=0.687254	lr=0.368807
Epoch[272] Batch [349]	Speed: 8815.467213 samples/sec	accuracy=0.686719	lr=0.368165
Epoch[272] Batch [399]	Speed: 8465.694431 samples/sec	accuracy=0.686326	lr=0.367524
Epoch[272] Batch [449]	Speed: 8386.606189 samples/sec	accuracy=0.685821	lr=0.366883
Epoch[272] Batch [499]	Speed: 8512.615021 samples/sec	accuracy=0.685734	lr=0.366242
Epoch[272] Batch [549]	Speed: 8469.720796 samples/sec	accuracy=0.685503	lr=0.365602
Epoch[272] Batch [599]	Speed: 8500.197497 samples/sec	accuracy=0.685044	lr=0.364962
[Epoch 272] training: accuracy=0.684892
[Epoch 272] speed: 8610 samples/sec	time cost: 153.726911
[Epoch 272] validation: err-top1=0.302930 err-top5=0.104238
Epoch[273] Batch [49]	Speed: 8372.236190 samples/sec	accuracy=0.693115	lr=0.363991
Epoch[273] Batch [99]	Speed: 8789.623409 samples/sec	accuracy=0.690522	lr=0.363352
Epoch[273] Batch [149]	Speed: 8433.569136 samples/sec	accuracy=0.690319	lr=0.362714
Epoch[273] Batch [199]	Speed: 8604.846791 samples/sec	accuracy=0.689585	lr=0.362077
Epoch[273] Batch [249]	Speed: 8693.803637 samples/sec	accuracy=0.689375	lr=0.361440
Epoch[273] Batch [299]	Speed: 8743.010365 samples/sec	accuracy=0.688664	lr=0.360803
Epoch[273] Batch [349]	Speed: 8701.226410 samples/sec	accuracy=0.688334	lr=0.360167
Epoch[273] Batch [399]	Speed: 8503.184593 samples/sec	accuracy=0.687722	lr=0.359532
Epoch[273] Batch [449]	Speed: 8440.272192 samples/sec	accuracy=0.687297	lr=0.358896
Epoch[273] Batch [499]	Speed: 8878.130425 samples/sec	accuracy=0.686777	lr=0.358262
Epoch[273] Batch [549]	Speed: 8763.122102 samples/sec	accuracy=0.686276	lr=0.357627
Epoch[273] Batch [599]	Speed: 8589.190253 samples/sec	accuracy=0.685667	lr=0.356994
[Epoch 273] training: accuracy=0.685607
[Epoch 273] speed: 8624 samples/sec	time cost: 153.408482
[Epoch 273] validation: err-top1=0.295898 err-top5=0.100871
Epoch[274] Batch [49]	Speed: 8796.715824 samples/sec	accuracy=0.698037	lr=0.356031
Epoch[274] Batch [99]	Speed: 8766.420700 samples/sec	accuracy=0.694570	lr=0.355399
Epoch[274] Batch [149]	Speed: 8743.910840 samples/sec	accuracy=0.693343	lr=0.354767
Epoch[274] Batch [199]	Speed: 8581.701822 samples/sec	accuracy=0.692683	lr=0.354135
Epoch[274] Batch [249]	Speed: 8714.129056 samples/sec	accuracy=0.692168	lr=0.353504
Epoch[274] Batch [299]	Speed: 8794.375325 samples/sec	accuracy=0.691626	lr=0.352873
Epoch[274] Batch [349]	Speed: 9012.456581 samples/sec	accuracy=0.690596	lr=0.352243
Epoch[274] Batch [399]	Speed: 8535.662799 samples/sec	accuracy=0.689897	lr=0.351613
Epoch[274] Batch [449]	Speed: 8755.775280 samples/sec	accuracy=0.689219	lr=0.350984
Epoch[274] Batch [499]	Speed: 8629.602285 samples/sec	accuracy=0.688500	lr=0.350355
Epoch[274] Batch [549]	Speed: 8472.545609 samples/sec	accuracy=0.688086	lr=0.349727
Epoch[274] Batch [599]	Speed: 8765.113627 samples/sec	accuracy=0.687634	lr=0.349099
[Epoch 274] training: accuracy=0.687558
[Epoch 274] speed: 8719 samples/sec	time cost: 151.371761
[Epoch 274] validation: err-top1=0.294983 err-top5=0.100118
Epoch[275] Batch [49]	Speed: 8646.364398 samples/sec	accuracy=0.696396	lr=0.348158
Epoch[275] Batch [99]	Speed: 8385.090520 samples/sec	accuracy=0.695903	lr=0.347532
Epoch[275] Batch [149]	Speed: 8704.033151 samples/sec	accuracy=0.694528	lr=0.346906
Epoch[275] Batch [199]	Speed: 8436.177330 samples/sec	accuracy=0.693594	lr=0.346280
Epoch[275] Batch [249]	Speed: 8743.467966 samples/sec	accuracy=0.692906	lr=0.345655
Epoch[275] Batch [299]	Speed: 8699.062064 samples/sec	accuracy=0.692049	lr=0.345030
Epoch[275] Batch [349]	Speed: 8697.861484 samples/sec	accuracy=0.691122	lr=0.344406
Epoch[275] Batch [399]	Speed: 8550.784143 samples/sec	accuracy=0.690544	lr=0.343782
Epoch[275] Batch [449]	Speed: 8529.273376 samples/sec	accuracy=0.689882	lr=0.343159
Epoch[275] Batch [499]	Speed: 8651.473562 samples/sec	accuracy=0.689156	lr=0.342536
Epoch[275] Batch [549]	Speed: 8720.085708 samples/sec	accuracy=0.688629	lr=0.341914
Epoch[275] Batch [599]	Speed: 8647.978091 samples/sec	accuracy=0.688406	lr=0.341292
[Epoch 275] training: accuracy=0.688366
[Epoch 275] speed: 8618 samples/sec	time cost: 153.590919
[Epoch 275] validation: err-top1=0.294473 err-top5=0.101094
Epoch[276] Batch [49]	Speed: 8578.684505 samples/sec	accuracy=0.697041	lr=0.340348
Epoch[276] Batch [99]	Speed: 8458.547800 samples/sec	accuracy=0.695557	lr=0.339727
Epoch[276] Batch [149]	Speed: 8660.572997 samples/sec	accuracy=0.693864	lr=0.339107
Epoch[276] Batch [199]	Speed: 8582.103251 samples/sec	accuracy=0.692954	lr=0.338487
Epoch[276] Batch [249]	Speed: 8827.554891 samples/sec	accuracy=0.692652	lr=0.337868
Epoch[276] Batch [299]	Speed: 8589.890612 samples/sec	accuracy=0.692340	lr=0.337249
Epoch[276] Batch [349]	Speed: 8626.710258 samples/sec	accuracy=0.691470	lr=0.336631
Epoch[276] Batch [399]	Speed: 8781.568437 samples/sec	accuracy=0.691050	lr=0.336013
Epoch[276] Batch [449]	Speed: 8813.548960 samples/sec	accuracy=0.690726	lr=0.335396
Epoch[276] Batch [499]	Speed: 8560.196640 samples/sec	accuracy=0.690223	lr=0.334779
Epoch[276] Batch [549]	Speed: 8419.190050 samples/sec	accuracy=0.689779	lr=0.334163
Epoch[276] Batch [599]	Speed: 8543.805900 samples/sec	accuracy=0.689344	lr=0.333547
[Epoch 276] training: accuracy=0.689127
[Epoch 276] speed: 8620 samples/sec	time cost: 153.023094
[Epoch 276] validation: err-top1=0.293274 err-top5=0.100220
Epoch[277] Batch [49]	Speed: 8630.822943 samples/sec	accuracy=0.698945	lr=0.332624
Epoch[277] Batch [99]	Speed: 8727.966864 samples/sec	accuracy=0.698018	lr=0.332010
Epoch[277] Batch [149]	Speed: 8427.328778 samples/sec	accuracy=0.696673	lr=0.331396
Epoch[277] Batch [199]	Speed: 8691.507196 samples/sec	accuracy=0.695972	lr=0.330782
Epoch[277] Batch [249]	Speed: 8677.795361 samples/sec	accuracy=0.695287	lr=0.330169
Epoch[277] Batch [299]	Speed: 8468.716434 samples/sec	accuracy=0.694600	lr=0.329556
Epoch[277] Batch [349]	Speed: 8485.279027 samples/sec	accuracy=0.694148	lr=0.328944
Epoch[277] Batch [399]	Speed: 8748.309081 samples/sec	accuracy=0.693359	lr=0.328333
Epoch[277] Batch [449]	Speed: 8695.575398 samples/sec	accuracy=0.692976	lr=0.327721
Epoch[277] Batch [499]	Speed: 8572.611573 samples/sec	accuracy=0.692537	lr=0.327111
Epoch[277] Batch [549]	Speed: 8416.726442 samples/sec	accuracy=0.692081	lr=0.326500
Epoch[277] Batch [599]	Speed: 8618.804484 samples/sec	accuracy=0.691627	lr=0.325891
[Epoch 277] training: accuracy=0.691246
[Epoch 277] speed: 8604 samples/sec	time cost: 153.823048
[Epoch 277] validation: err-top1=0.290547 err-top5=0.097793
Epoch[278] Batch [49]	Speed: 8578.504250 samples/sec	accuracy=0.700000	lr=0.324965
Epoch[278] Batch [99]	Speed: 8977.650780 samples/sec	accuracy=0.699004	lr=0.324356
Epoch[278] Batch [149]	Speed: 8805.241412 samples/sec	accuracy=0.697783	lr=0.323748
Epoch[278] Batch [199]	Speed: 8563.097331 samples/sec	accuracy=0.696572	lr=0.323141
Epoch[278] Batch [249]	Speed: 8389.892047 samples/sec	accuracy=0.695436	lr=0.322534
Epoch[278] Batch [299]	Speed: 8539.038846 samples/sec	accuracy=0.694880	lr=0.321927
Epoch[278] Batch [349]	Speed: 8749.295128 samples/sec	accuracy=0.694184	lr=0.321321
Epoch[278] Batch [399]	Speed: 8574.663802 samples/sec	accuracy=0.693579	lr=0.320716
Epoch[278] Batch [449]	Speed: 8722.615518 samples/sec	accuracy=0.693101	lr=0.320111
Epoch[278] Batch [499]	Speed: 8696.760026 samples/sec	accuracy=0.692697	lr=0.319506
Epoch[278] Batch [549]	Speed: 8298.470083 samples/sec	accuracy=0.692434	lr=0.318902
Epoch[278] Batch [599]	Speed: 8725.289662 samples/sec	accuracy=0.692080	lr=0.318299
[Epoch 278] training: accuracy=0.691890
[Epoch 278] speed: 8633 samples/sec	time cost: 152.649709
[Epoch 278] validation: err-top1=0.294678 err-top5=0.100444
Epoch[279] Batch [49]	Speed: 8627.442918 samples/sec	accuracy=0.697578	lr=0.317394
Epoch[279] Batch [99]	Speed: 8610.847585 samples/sec	accuracy=0.698159	lr=0.316792
Epoch[279] Batch [149]	Speed: 8726.630274 samples/sec	accuracy=0.698096	lr=0.316190
Epoch[279] Batch [199]	Speed: 8459.064906 samples/sec	accuracy=0.696746	lr=0.315588
Epoch[279] Batch [249]	Speed: 8956.389944 samples/sec	accuracy=0.696779	lr=0.314988
Epoch[279] Batch [299]	Speed: 8624.758607 samples/sec	accuracy=0.696214	lr=0.314387
Epoch[279] Batch [349]	Speed: 8668.999508 samples/sec	accuracy=0.695811	lr=0.313787
Epoch[279] Batch [399]	Speed: 8658.928239 samples/sec	accuracy=0.695048	lr=0.313188
Epoch[279] Batch [449]	Speed: 8569.680324 samples/sec	accuracy=0.694776	lr=0.312589
Epoch[279] Batch [499]	Speed: 8588.499454 samples/sec	accuracy=0.694389	lr=0.311991
Epoch[279] Batch [549]	Speed: 8811.261136 samples/sec	accuracy=0.693676	lr=0.311393
Epoch[279] Batch [599]	Speed: 8913.552711 samples/sec	accuracy=0.693462	lr=0.310795
[Epoch 279] training: accuracy=0.693336
[Epoch 279] speed: 8687 samples/sec	time cost: 152.205913
[Epoch 279] validation: err-top1=0.291728 err-top5=0.098368
Epoch[280] Batch [49]	Speed: 8561.433409 samples/sec	accuracy=0.702148	lr=0.309888
Epoch[280] Batch [99]	Speed: 8579.944619 samples/sec	accuracy=0.701074	lr=0.309292
Epoch[280] Batch [149]	Speed: 8731.745473 samples/sec	accuracy=0.700166	lr=0.308696
Epoch[280] Batch [199]	Speed: 8887.241803 samples/sec	accuracy=0.700051	lr=0.308101
Epoch[280] Batch [249]	Speed: 8684.681163 samples/sec	accuracy=0.699146	lr=0.307507
Epoch[280] Batch [299]	Speed: 8752.990195 samples/sec	accuracy=0.697961	lr=0.306912
Epoch[280] Batch [349]	Speed: 8589.121030 samples/sec	accuracy=0.697111	lr=0.306319
Epoch[280] Batch [399]	Speed: 8504.620323 samples/sec	accuracy=0.696322	lr=0.305726
Epoch[280] Batch [449]	Speed: 8508.155562 samples/sec	accuracy=0.695405	lr=0.305133
Epoch[280] Batch [499]	Speed: 8779.029802 samples/sec	accuracy=0.695185	lr=0.304541
Epoch[280] Batch [549]	Speed: 8697.706305 samples/sec	accuracy=0.694983	lr=0.303949
Epoch[280] Batch [599]	Speed: 8721.707557 samples/sec	accuracy=0.694630	lr=0.303358
[Epoch 280] training: accuracy=0.694333
[Epoch 280] speed: 8663 samples/sec	time cost: 152.836892
[Epoch 280] validation: err-top1=0.298164 err-top5=0.103125
Epoch[281] Batch [49]	Speed: 8478.274027 samples/sec	accuracy=0.706084	lr=0.302460
Epoch[281] Batch [99]	Speed: 8736.004963 samples/sec	accuracy=0.703496	lr=0.301870
Epoch[281] Batch [149]	Speed: 8820.506332 samples/sec	accuracy=0.701348	lr=0.301281
Epoch[281] Batch [199]	Speed: 8651.708134 samples/sec	accuracy=0.700564	lr=0.300692
Epoch[281] Batch [249]	Speed: 8689.343989 samples/sec	accuracy=0.699408	lr=0.300103
Epoch[281] Batch [299]	Speed: 8458.242463 samples/sec	accuracy=0.699132	lr=0.299516
Epoch[281] Batch [349]	Speed: 8382.204945 samples/sec	accuracy=0.698491	lr=0.298928
Epoch[281] Batch [399]	Speed: 8490.672685 samples/sec	accuracy=0.697767	lr=0.298341
Epoch[281] Batch [449]	Speed: 8796.982663 samples/sec	accuracy=0.697624	lr=0.297755
Epoch[281] Batch [499]	Speed: 8696.974167 samples/sec	accuracy=0.696885	lr=0.297169
Epoch[281] Batch [549]	Speed: 8531.879249 samples/sec	accuracy=0.696479	lr=0.296583
Epoch[281] Batch [599]	Speed: 8821.275180 samples/sec	accuracy=0.696178	lr=0.295999
[Epoch 281] training: accuracy=0.695955
[Epoch 281] speed: 8626 samples/sec	time cost: 152.814515
[Epoch 281] validation: err-top1=0.295878 err-top5=0.103088
Epoch[282] Batch [49]	Speed: 8532.279590 samples/sec	accuracy=0.701953	lr=0.295122
Epoch[282] Batch [99]	Speed: 8734.079570 samples/sec	accuracy=0.701406	lr=0.294539
Epoch[282] Batch [149]	Speed: 8691.233526 samples/sec	accuracy=0.701038	lr=0.293955
Epoch[282] Batch [199]	Speed: 8539.812214 samples/sec	accuracy=0.700916	lr=0.293373
Epoch[282] Batch [249]	Speed: 8786.015437 samples/sec	accuracy=0.700230	lr=0.292791
Epoch[282] Batch [299]	Speed: 8827.618212 samples/sec	accuracy=0.699546	lr=0.292209
Epoch[282] Batch [349]	Speed: 8764.802213 samples/sec	accuracy=0.699021	lr=0.291628
Epoch[282] Batch [399]	Speed: 8692.244219 samples/sec	accuracy=0.698494	lr=0.291047
Epoch[282] Batch [449]	Speed: 8581.255339 samples/sec	accuracy=0.698018	lr=0.290467
Epoch[282] Batch [499]	Speed: 8652.640276 samples/sec	accuracy=0.697458	lr=0.289888
Epoch[282] Batch [549]	Speed: 8549.935259 samples/sec	accuracy=0.697258	lr=0.289308
Epoch[282] Batch [599]	Speed: 8476.375571 samples/sec	accuracy=0.696790	lr=0.288730
[Epoch 282] training: accuracy=0.696534
[Epoch 282] speed: 8645 samples/sec	time cost: 152.882929
[Epoch 282] validation: err-top1=0.299727 err-top5=0.103105
Epoch[283] Batch [49]	Speed: 8441.405695 samples/sec	accuracy=0.704014	lr=0.287851
Epoch[283] Batch [99]	Speed: 8521.696257 samples/sec	accuracy=0.703623	lr=0.287274
Epoch[283] Batch [149]	Speed: 8721.501229 samples/sec	accuracy=0.703454	lr=0.286697
Epoch[283] Batch [199]	Speed: 8720.147851 samples/sec	accuracy=0.702737	lr=0.286121
Epoch[283] Batch [249]	Speed: 8624.180350 samples/sec	accuracy=0.701846	lr=0.285545
Epoch[283] Batch [299]	Speed: 8528.309539 samples/sec	accuracy=0.701090	lr=0.284970
Epoch[283] Batch [349]	Speed: 8684.701885 samples/sec	accuracy=0.700579	lr=0.284395
Epoch[283] Batch [399]	Speed: 8525.076154 samples/sec	accuracy=0.700232	lr=0.283821
Epoch[283] Batch [449]	Speed: 8672.484153 samples/sec	accuracy=0.699529	lr=0.283247
Epoch[283] Batch [499]	Speed: 8659.922702 samples/sec	accuracy=0.699070	lr=0.282674
Epoch[283] Batch [549]	Speed: 8727.691071 samples/sec	accuracy=0.698550	lr=0.282101
Epoch[283] Batch [599]	Speed: 8611.847610 samples/sec	accuracy=0.698031	lr=0.281529
[Epoch 283] training: accuracy=0.697760
[Epoch 283] speed: 8617 samples/sec	time cost: 153.019516
[Epoch 283] validation: err-top1=0.289897 err-top5=0.099426
Epoch[284] Batch [49]	Speed: 8473.142323 samples/sec	accuracy=0.707041	lr=0.280671
Epoch[284] Batch [99]	Speed: 8516.897765 samples/sec	accuracy=0.704775	lr=0.280100
Epoch[284] Batch [149]	Speed: 8608.539022 samples/sec	accuracy=0.704811	lr=0.279530
Epoch[284] Batch [199]	Speed: 8759.422051 samples/sec	accuracy=0.704199	lr=0.278960
Epoch[284] Batch [249]	Speed: 8977.700322 samples/sec	accuracy=0.702852	lr=0.278391
Epoch[284] Batch [299]	Speed: 8618.360703 samples/sec	accuracy=0.702448	lr=0.277822
Epoch[284] Batch [349]	Speed: 8588.178138 samples/sec	accuracy=0.701811	lr=0.277253
Epoch[284] Batch [399]	Speed: 8630.715066 samples/sec	accuracy=0.700887	lr=0.276685
Epoch[284] Batch [449]	Speed: 8767.187127 samples/sec	accuracy=0.700426	lr=0.276118
Epoch[284] Batch [499]	Speed: 8736.031084 samples/sec	accuracy=0.699831	lr=0.275551
Epoch[284] Batch [549]	Speed: 8781.670243 samples/sec	accuracy=0.699369	lr=0.274985
Epoch[284] Batch [599]	Speed: 8875.813907 samples/sec	accuracy=0.699056	lr=0.274419
[Epoch 284] training: accuracy=0.699028
[Epoch 284] speed: 8682 samples/sec	time cost: 152.485739
[Epoch 284] validation: err-top1=0.297676 err-top5=0.103945
Epoch[285] Batch [49]	Speed: 8831.550445 samples/sec	accuracy=0.706514	lr=0.273560
Epoch[285] Batch [99]	Speed: 8550.670597 samples/sec	accuracy=0.706846	lr=0.272995
Epoch[285] Batch [149]	Speed: 8625.575990 samples/sec	accuracy=0.706732	lr=0.272431
Epoch[285] Batch [199]	Speed: 8543.105899 samples/sec	accuracy=0.705630	lr=0.271868
Epoch[285] Batch [249]	Speed: 8777.902669 samples/sec	accuracy=0.704627	lr=0.271305
Epoch[285] Batch [299]	Speed: 8747.049266 samples/sec	accuracy=0.703662	lr=0.270743
Epoch[285] Batch [349]	Speed: 8682.188049 samples/sec	accuracy=0.702917	lr=0.270181
Epoch[285] Batch [399]	Speed: 8831.343427 samples/sec	accuracy=0.702413	lr=0.269619
Epoch[285] Batch [449]	Speed: 8613.355684 samples/sec	accuracy=0.701918	lr=0.269058
Epoch[285] Batch [499]	Speed: 8868.389976 samples/sec	accuracy=0.701558	lr=0.268498
Epoch[285] Batch [549]	Speed: 8437.503000 samples/sec	accuracy=0.701257	lr=0.267938
Epoch[285] Batch [599]	Speed: 8669.661668 samples/sec	accuracy=0.700833	lr=0.267379
[Epoch 285] training: accuracy=0.700855
[Epoch 285] speed: 8676 samples/sec	time cost: 152.097322
[Epoch 285] validation: err-top1=0.288676 err-top5=0.098022
Epoch[286] Batch [49]	Speed: 8706.488705 samples/sec	accuracy=0.708682	lr=0.266540
Epoch[286] Batch [99]	Speed: 8490.897276 samples/sec	accuracy=0.709624	lr=0.265982
Epoch[286] Batch [149]	Speed: 8506.260717 samples/sec	accuracy=0.708311	lr=0.265425
Epoch[286] Batch [199]	Speed: 8528.569488 samples/sec	accuracy=0.707791	lr=0.264868
Epoch[286] Batch [249]	Speed: 8345.233810 samples/sec	accuracy=0.706475	lr=0.264311
Epoch[286] Batch [299]	Speed: 8588.760852 samples/sec	accuracy=0.705568	lr=0.263755
Epoch[286] Batch [349]	Speed: 8673.204642 samples/sec	accuracy=0.704817	lr=0.263200
Epoch[286] Batch [399]	Speed: 8901.993069 samples/sec	accuracy=0.703815	lr=0.262645
Epoch[286] Batch [449]	Speed: 8773.568361 samples/sec	accuracy=0.703368	lr=0.262090
Epoch[286] Batch [499]	Speed: 8675.604626 samples/sec	accuracy=0.703000	lr=0.261536
Epoch[286] Batch [549]	Speed: 8606.864639 samples/sec	accuracy=0.702478	lr=0.260983
Epoch[286] Batch [599]	Speed: 8404.351714 samples/sec	accuracy=0.702184	lr=0.260430
[Epoch 286] training: accuracy=0.701949
[Epoch 286] speed: 8592 samples/sec	time cost: 153.709130
[Epoch 286] validation: err-top1=0.290466 err-top5=0.098185
Epoch[287] Batch [49]	Speed: 8327.486203 samples/sec	accuracy=0.708828	lr=0.259591
Epoch[287] Batch [99]	Speed: 8575.636947 samples/sec	accuracy=0.709097	lr=0.259039
Epoch[287] Batch [149]	Speed: 8906.753330 samples/sec	accuracy=0.708389	lr=0.258488
Epoch[287] Batch [199]	Speed: 8677.535879 samples/sec	accuracy=0.707986	lr=0.257938
Epoch[287] Batch [249]	Speed: 8366.744487 samples/sec	accuracy=0.706924	lr=0.257388
Epoch[287] Batch [299]	Speed: 8603.116114 samples/sec	accuracy=0.706099	lr=0.256838
Epoch[287] Batch [349]	Speed: 8697.731317 samples/sec	accuracy=0.705456	lr=0.256289
Epoch[287] Batch [399]	Speed: 8712.659369 samples/sec	accuracy=0.705408	lr=0.255741
Epoch[287] Batch [449]	Speed: 8724.814642 samples/sec	accuracy=0.704842	lr=0.255193
Epoch[287] Batch [499]	Speed: 8963.498428 samples/sec	accuracy=0.704225	lr=0.254645
Epoch[287] Batch [549]	Speed: 8794.176708 samples/sec	accuracy=0.703929	lr=0.254099
Epoch[287] Batch [599]	Speed: 8743.849604 samples/sec	accuracy=0.703579	lr=0.253552
[Epoch 287] training: accuracy=0.703326
[Epoch 287] speed: 8677 samples/sec	time cost: 152.494564
[Epoch 287] validation: err-top1=0.288066 err-top5=0.096152
Epoch[288] Batch [49]	Speed: 8497.836401 samples/sec	accuracy=0.713496	lr=0.252723
Epoch[288] Batch [99]	Speed: 8764.493146 samples/sec	accuracy=0.711572	lr=0.252178
Epoch[288] Batch [149]	Speed: 8785.337722 samples/sec	accuracy=0.710199	lr=0.251633
Epoch[288] Batch [199]	Speed: 8602.960679 samples/sec	accuracy=0.708994	lr=0.251089
Epoch[288] Batch [249]	Speed: 8958.184972 samples/sec	accuracy=0.707648	lr=0.250546
Epoch[288] Batch [299]	Speed: 8540.767444 samples/sec	accuracy=0.706846	lr=0.250003
Epoch[288] Batch [349]	Speed: 8669.187437 samples/sec	accuracy=0.706348	lr=0.249461
Epoch[288] Batch [399]	Speed: 8852.784566 samples/sec	accuracy=0.705913	lr=0.248919
Epoch[288] Batch [449]	Speed: 8842.001668 samples/sec	accuracy=0.705476	lr=0.248377
Epoch[288] Batch [499]	Speed: 8625.173428 samples/sec	accuracy=0.705209	lr=0.247837
Epoch[288] Batch [549]	Speed: 8587.961422 samples/sec	accuracy=0.704761	lr=0.247296
Epoch[288] Batch [599]	Speed: 8763.117632 samples/sec	accuracy=0.704277	lr=0.246757
[Epoch 288] training: accuracy=0.704123
[Epoch 288] speed: 8704 samples/sec	time cost: 151.539077
[Epoch 288] validation: err-top1=0.298075 err-top5=0.102559
Epoch[289] Batch [49]	Speed: 8261.195199 samples/sec	accuracy=0.711777	lr=0.245948
Epoch[289] Batch [99]	Speed: 8615.990727 samples/sec	accuracy=0.711733	lr=0.245410
Epoch[289] Batch [149]	Speed: 8839.500565 samples/sec	accuracy=0.710934	lr=0.244872
Epoch[289] Batch [199]	Speed: 8294.510825 samples/sec	accuracy=0.709939	lr=0.244334
Epoch[289] Batch [249]	Speed: 8816.521486 samples/sec	accuracy=0.709158	lr=0.243797
Epoch[289] Batch [299]	Speed: 8654.625843 samples/sec	accuracy=0.708607	lr=0.243261
Epoch[289] Batch [349]	Speed: 8728.085345 samples/sec	accuracy=0.707939	lr=0.242725
Epoch[289] Batch [399]	Speed: 8606.147024 samples/sec	accuracy=0.707323	lr=0.242190
Epoch[289] Batch [449]	Speed: 8832.925547 samples/sec	accuracy=0.707033	lr=0.241655
Epoch[289] Batch [499]	Speed: 8639.441529 samples/sec	accuracy=0.706485	lr=0.241121
Epoch[289] Batch [549]	Speed: 8549.059997 samples/sec	accuracy=0.706153	lr=0.240587
Epoch[289] Batch [599]	Speed: 8686.872444 samples/sec	accuracy=0.705746	lr=0.240054
[Epoch 289] training: accuracy=0.705589
[Epoch 289] speed: 8627 samples/sec	time cost: 153.304765
[Epoch 289] validation: err-top1=0.282500 err-top5=0.094258
Epoch[290] Batch [49]	Speed: 8387.808373 samples/sec	accuracy=0.715098	lr=0.239245
Epoch[290] Batch [99]	Speed: 8623.165166 samples/sec	accuracy=0.713853	lr=0.238713
Epoch[290] Batch [149]	Speed: 8707.140187 samples/sec	accuracy=0.712839	lr=0.238182
Epoch[290] Batch [199]	Speed: 8724.778309 samples/sec	accuracy=0.711826	lr=0.237651
Epoch[290] Batch [249]	Speed: 8834.195140 samples/sec	accuracy=0.710662	lr=0.237121
Epoch[290] Batch [299]	Speed: 8615.153559 samples/sec	accuracy=0.709775	lr=0.236591
Epoch[290] Batch [349]	Speed: 8676.088848 samples/sec	accuracy=0.709283	lr=0.236062
Epoch[290] Batch [399]	Speed: 8638.810390 samples/sec	accuracy=0.708434	lr=0.235534
Epoch[290] Batch [449]	Speed: 8474.931133 samples/sec	accuracy=0.707892	lr=0.235006
Epoch[290] Batch [499]	Speed: 8556.994101 samples/sec	accuracy=0.707441	lr=0.234478
Epoch[290] Batch [549]	Speed: 8892.320070 samples/sec	accuracy=0.707006	lr=0.233951
Epoch[290] Batch [599]	Speed: 8554.696599 samples/sec	accuracy=0.706628	lr=0.233425
[Epoch 290] training: accuracy=0.706492
[Epoch 290] speed: 8636 samples/sec	time cost: 152.830789
[Epoch 290] validation: err-top1=0.289164 err-top5=0.097758
Epoch[291] Batch [49]	Speed: 8453.711984 samples/sec	accuracy=0.715576	lr=0.232636
Epoch[291] Batch [99]	Speed: 8421.188295 samples/sec	accuracy=0.715122	lr=0.232111
Epoch[291] Batch [149]	Speed: 8519.730141 samples/sec	accuracy=0.713486	lr=0.231586
Epoch[291] Batch [199]	Speed: 8677.128628 samples/sec	accuracy=0.712029	lr=0.231062
Epoch[291] Batch [249]	Speed: 8497.806809 samples/sec	accuracy=0.711047	lr=0.230539
Epoch[291] Batch [299]	Speed: 8586.971057 samples/sec	accuracy=0.710701	lr=0.230016
Epoch[291] Batch [349]	Speed: 8625.716826 samples/sec	accuracy=0.709953	lr=0.229493
Epoch[291] Batch [399]	Speed: 8680.342968 samples/sec	accuracy=0.709497	lr=0.228971
Epoch[291] Batch [449]	Speed: 8616.569442 samples/sec	accuracy=0.708831	lr=0.228450
Epoch[291] Batch [499]	Speed: 8671.707756 samples/sec	accuracy=0.708568	lr=0.227929
Epoch[291] Batch [549]	Speed: 8631.268008 samples/sec	accuracy=0.708103	lr=0.227409
Epoch[291] Batch [599]	Speed: 8683.019864 samples/sec	accuracy=0.707622	lr=0.226889
[Epoch 291] training: accuracy=0.707575
[Epoch 291] speed: 8589 samples/sec	time cost: 153.885830
[Epoch 291] validation: err-top1=0.286926 err-top5=0.096903
Epoch[292] Batch [49]	Speed: 8555.797303 samples/sec	accuracy=0.715195	lr=0.226100
Epoch[292] Batch [99]	Speed: 8689.258552 samples/sec	accuracy=0.714907	lr=0.225582
Epoch[292] Batch [149]	Speed: 8860.390937 samples/sec	accuracy=0.713721	lr=0.225064
Epoch[292] Batch [199]	Speed: 8636.445305 samples/sec	accuracy=0.712771	lr=0.224546
Epoch[292] Batch [249]	Speed: 8577.934919 samples/sec	accuracy=0.711939	lr=0.224030
Epoch[292] Batch [299]	Speed: 8493.133928 samples/sec	accuracy=0.711497	lr=0.223513
Epoch[292] Batch [349]	Speed: 8732.484898 samples/sec	accuracy=0.711161	lr=0.222998
Epoch[292] Batch [399]	Speed: 8480.306615 samples/sec	accuracy=0.711013	lr=0.222482
Epoch[292] Batch [449]	Speed: 8903.944482 samples/sec	accuracy=0.710592	lr=0.221968
Epoch[292] Batch [499]	Speed: 8803.108917 samples/sec	accuracy=0.710122	lr=0.221454
Epoch[292] Batch [549]	Speed: 8805.769100 samples/sec	accuracy=0.709688	lr=0.220940
Epoch[292] Batch [599]	Speed: 8866.759798 samples/sec	accuracy=0.709403	lr=0.220427
[Epoch 292] training: accuracy=0.709170
[Epoch 292] speed: 8704 samples/sec	time cost: 152.005200
[Epoch 292] validation: err-top1=0.295820 err-top5=0.101797
Epoch[293] Batch [49]	Speed: 8332.629493 samples/sec	accuracy=0.716104	lr=0.219659
Epoch[293] Batch [99]	Speed: 8635.777444 samples/sec	accuracy=0.715801	lr=0.219147
Epoch[293] Batch [149]	Speed: 8613.312155 samples/sec	accuracy=0.715094	lr=0.218636
Epoch[293] Batch [199]	Speed: 8867.010767 samples/sec	accuracy=0.714570	lr=0.218125
Epoch[293] Batch [249]	Speed: 8654.568642 samples/sec	accuracy=0.714133	lr=0.217615
Epoch[293] Batch [299]	Speed: 8796.863926 samples/sec	accuracy=0.713844	lr=0.217106
Epoch[293] Batch [349]	Speed: 8556.949093 samples/sec	accuracy=0.713135	lr=0.216597
Epoch[293] Batch [399]	Speed: 8587.313916 samples/sec	accuracy=0.712617	lr=0.216088
Epoch[293] Batch [449]	Speed: 8899.730103 samples/sec	accuracy=0.712370	lr=0.215581
Epoch[293] Batch [499]	Speed: 8581.755150 samples/sec	accuracy=0.711810	lr=0.215073
Epoch[293] Batch [549]	Speed: 8697.485964 samples/sec	accuracy=0.711384	lr=0.214567
Epoch[293] Batch [599]	Speed: 8587.733728 samples/sec	accuracy=0.710988	lr=0.214060
[Epoch 293] training: accuracy=0.710880
[Epoch 293] speed: 8644 samples/sec	time cost: 152.876006
[Epoch 293] validation: err-top1=0.279338 err-top5=0.092407
Epoch[294] Batch [49]	Speed: 8396.636792 samples/sec	accuracy=0.717500	lr=0.213292
Epoch[294] Batch [99]	Speed: 8560.439256 samples/sec	accuracy=0.716802	lr=0.212787
Epoch[294] Batch [149]	Speed: 8683.912238 samples/sec	accuracy=0.716296	lr=0.212283
Epoch[294] Batch [199]	Speed: 8545.307578 samples/sec	accuracy=0.715732	lr=0.211779
Epoch[294] Batch [249]	Speed: 8720.056673 samples/sec	accuracy=0.715154	lr=0.211276
Epoch[294] Batch [299]	Speed: 8755.405451 samples/sec	accuracy=0.714917	lr=0.210773
Epoch[294] Batch [349]	Speed: 8508.796917 samples/sec	accuracy=0.714033	lr=0.210271
Epoch[294] Batch [399]	Speed: 8654.661594 samples/sec	accuracy=0.713605	lr=0.209769
Epoch[294] Batch [449]	Speed: 8729.243720 samples/sec	accuracy=0.713009	lr=0.209268
Epoch[294] Batch [499]	Speed: 8708.656749 samples/sec	accuracy=0.712667	lr=0.208768
Epoch[294] Batch [549]	Speed: 8771.445267 samples/sec	accuracy=0.712340	lr=0.208268
Epoch[294] Batch [599]	Speed: 8709.479691 samples/sec	accuracy=0.711889	lr=0.207769
[Epoch 294] training: accuracy=0.711870
[Epoch 294] speed: 8646 samples/sec	time cost: 153.060697
[Epoch 294] validation: err-top1=0.278828 err-top5=0.090898
Epoch[295] Batch [49]	Speed: 8586.467891 samples/sec	accuracy=0.720566	lr=0.207010
Epoch[295] Batch [99]	Speed: 8698.796728 samples/sec	accuracy=0.719580	lr=0.206512
Epoch[295] Batch [149]	Speed: 8744.676896 samples/sec	accuracy=0.718486	lr=0.206015
Epoch[295] Batch [199]	Speed: 8915.468110 samples/sec	accuracy=0.717473	lr=0.205518
Epoch[295] Batch [249]	Speed: 8646.481892 samples/sec	accuracy=0.716848	lr=0.205022
Epoch[295] Batch [299]	Speed: 8713.754605 samples/sec	accuracy=0.716118	lr=0.204526
Epoch[295] Batch [349]	Speed: 8685.505552 samples/sec	accuracy=0.715629	lr=0.204031
Epoch[295] Batch [399]	Speed: 8576.256835 samples/sec	accuracy=0.714838	lr=0.203536
Epoch[295] Batch [449]	Speed: 8918.181830 samples/sec	accuracy=0.714434	lr=0.203042
Epoch[295] Batch [499]	Speed: 8478.726094 samples/sec	accuracy=0.714052	lr=0.202548
Epoch[295] Batch [549]	Speed: 8692.223637 samples/sec	accuracy=0.713577	lr=0.202055
Epoch[295] Batch [599]	Speed: 8713.604692 samples/sec	accuracy=0.713289	lr=0.201563
[Epoch 295] training: accuracy=0.713252
[Epoch 295] speed: 8695 samples/sec	time cost: 151.767480
[Epoch 295] validation: err-top1=0.282613 err-top5=0.093445
Epoch[296] Batch [49]	Speed: 8380.994063 samples/sec	accuracy=0.722627	lr=0.200825
Epoch[296] Batch [99]	Speed: 8922.934770 samples/sec	accuracy=0.720728	lr=0.200334
Epoch[296] Batch [149]	Speed: 8845.811544 samples/sec	accuracy=0.719740	lr=0.199843
Epoch[296] Batch [199]	Speed: 8624.911368 samples/sec	accuracy=0.718821	lr=0.199353
Epoch[296] Batch [249]	Speed: 8671.850277 samples/sec	accuracy=0.718152	lr=0.198863
Epoch[296] Batch [299]	Speed: 8461.839936 samples/sec	accuracy=0.717290	lr=0.198375
Epoch[296] Batch [349]	Speed: 8330.641218 samples/sec	accuracy=0.716483	lr=0.197886
Epoch[296] Batch [399]	Speed: 8645.513834 samples/sec	accuracy=0.715919	lr=0.197398
Epoch[296] Batch [449]	Speed: 8616.570134 samples/sec	accuracy=0.715470	lr=0.196911
Epoch[296] Batch [499]	Speed: 8598.055972 samples/sec	accuracy=0.714908	lr=0.196424
Epoch[296] Batch [549]	Speed: 8657.037192 samples/sec	accuracy=0.714801	lr=0.195938
Epoch[296] Batch [599]	Speed: 8522.013294 samples/sec	accuracy=0.714657	lr=0.195453
[Epoch 296] training: accuracy=0.714408
[Epoch 296] speed: 8603 samples/sec	time cost: 153.476559
[Epoch 296] validation: err-top1=0.278503 err-top5=0.089600
Epoch[297] Batch [49]	Speed: 8496.218414 samples/sec	accuracy=0.721533	lr=0.194715
Epoch[297] Batch [99]	Speed: 8717.327878 samples/sec	accuracy=0.721250	lr=0.194231
Epoch[297] Batch [149]	Speed: 8547.617050 samples/sec	accuracy=0.721068	lr=0.193748
Epoch[297] Batch [199]	Speed: 8764.239720 samples/sec	accuracy=0.720339	lr=0.193264
Epoch[297] Batch [249]	Speed: 8904.445299 samples/sec	accuracy=0.719592	lr=0.192782
Epoch[297] Batch [299]	Speed: 8994.410912 samples/sec	accuracy=0.718748	lr=0.192300
Epoch[297] Batch [349]	Speed: 8647.123032 samples/sec	accuracy=0.718372	lr=0.191818
Epoch[297] Batch [399]	Speed: 8628.563289 samples/sec	accuracy=0.717850	lr=0.191338
Epoch[297] Batch [449]	Speed: 8592.826422 samples/sec	accuracy=0.717564	lr=0.190857
Epoch[297] Batch [499]	Speed: 8594.062664 samples/sec	accuracy=0.717239	lr=0.190377
Epoch[297] Batch [549]	Speed: 8765.915607 samples/sec	accuracy=0.716745	lr=0.189898
Epoch[297] Batch [599]	Speed: 8855.396716 samples/sec	accuracy=0.716308	lr=0.189419
[Epoch 297] training: accuracy=0.716167
[Epoch 297] speed: 8714 samples/sec	time cost: 151.635199
[Epoch 297] validation: err-top1=0.279023 err-top5=0.090430
Epoch[298] Batch [49]	Speed: 8546.962176 samples/sec	accuracy=0.723779	lr=0.188703
Epoch[298] Batch [99]	Speed: 8753.831529 samples/sec	accuracy=0.722852	lr=0.188225
Epoch[298] Batch [149]	Speed: 8492.367648 samples/sec	accuracy=0.722327	lr=0.187749
Epoch[298] Batch [199]	Speed: 8467.075798 samples/sec	accuracy=0.721436	lr=0.187272
Epoch[298] Batch [249]	Speed: 8476.350980 samples/sec	accuracy=0.720398	lr=0.186797
Epoch[298] Batch [299]	Speed: 8473.924699 samples/sec	accuracy=0.719648	lr=0.186322
Epoch[298] Batch [349]	Speed: 8667.212149 samples/sec	accuracy=0.718998	lr=0.185847
Epoch[298] Batch [399]	Speed: 8522.573704 samples/sec	accuracy=0.718652	lr=0.185373
Epoch[298] Batch [449]	Speed: 8422.355490 samples/sec	accuracy=0.718383	lr=0.184900
Epoch[298] Batch [499]	Speed: 8591.272940 samples/sec	accuracy=0.718031	lr=0.184427
Epoch[298] Batch [549]	Speed: 8598.510747 samples/sec	accuracy=0.717401	lr=0.183955
Epoch[298] Batch [599]	Speed: 8538.182279 samples/sec	accuracy=0.717087	lr=0.183483
[Epoch 298] training: accuracy=0.716892
[Epoch 298] speed: 8540 samples/sec	time cost: 154.679811
[Epoch 298] validation: err-top1=0.278280 err-top5=0.089823
Epoch[299] Batch [49]	Speed: 8540.218903 samples/sec	accuracy=0.724951	lr=0.182767
Epoch[299] Batch [99]	Speed: 8466.522663 samples/sec	accuracy=0.723022	lr=0.182297
Epoch[299] Batch [149]	Speed: 8747.440659 samples/sec	accuracy=0.722526	lr=0.181827
Epoch[299] Batch [199]	Speed: 8304.703408 samples/sec	accuracy=0.722000	lr=0.181358
Epoch[299] Batch [249]	Speed: 8565.107087 samples/sec	accuracy=0.721389	lr=0.180889
Epoch[299] Batch [299]	Speed: 8915.100953 samples/sec	accuracy=0.721139	lr=0.180421
Epoch[299] Batch [349]	Speed: 8784.104229 samples/sec	accuracy=0.721013	lr=0.179954
Epoch[299] Batch [399]	Speed: 8626.289399 samples/sec	accuracy=0.720865	lr=0.179487
Epoch[299] Batch [449]	Speed: 8755.373860 samples/sec	accuracy=0.720637	lr=0.179021
Epoch[299] Batch [499]	Speed: 8674.205365 samples/sec	accuracy=0.720132	lr=0.178555
Epoch[299] Batch [549]	Speed: 8582.185908 samples/sec	accuracy=0.719812	lr=0.178090
Epoch[299] Batch [599]	Speed: 8612.782582 samples/sec	accuracy=0.719368	lr=0.177625
[Epoch 299] training: accuracy=0.719068
[Epoch 299] speed: 8629 samples/sec	time cost: 153.132982
[Epoch 299] validation: err-top1=0.277871 err-top5=0.091465
Epoch[300] Batch [49]	Speed: 8584.707878 samples/sec	accuracy=0.728506	lr=0.176929
Epoch[300] Batch [99]	Speed: 8682.509769 samples/sec	accuracy=0.726675	lr=0.176466
Epoch[300] Batch [149]	Speed: 8800.538880 samples/sec	accuracy=0.725837	lr=0.176003
Epoch[300] Batch [199]	Speed: 8893.662413 samples/sec	accuracy=0.724985	lr=0.175541
Epoch[300] Batch [249]	Speed: 8484.744631 samples/sec	accuracy=0.724186	lr=0.175079
Epoch[300] Batch [299]	Speed: 8721.347330 samples/sec	accuracy=0.723078	lr=0.174618
Epoch[300] Batch [349]	Speed: 8726.028702 samples/sec	accuracy=0.722095	lr=0.174158
Epoch[300] Batch [399]	Speed: 8538.792519 samples/sec	accuracy=0.721945	lr=0.173698
Epoch[300] Batch [449]	Speed: 8630.532444 samples/sec	accuracy=0.721492	lr=0.173239
Epoch[300] Batch [499]	Speed: 8703.531519 samples/sec	accuracy=0.720968	lr=0.172780
Epoch[300] Batch [549]	Speed: 8646.063627 samples/sec	accuracy=0.720555	lr=0.172322
Epoch[300] Batch [599]	Speed: 8757.480431 samples/sec	accuracy=0.720129	lr=0.171864
[Epoch 300] training: accuracy=0.719835
[Epoch 300] speed: 8681 samples/sec	time cost: 152.213636
[Epoch 300] validation: err-top1=0.272278 err-top5=0.088928
Epoch[301] Batch [49]	Speed: 8850.437119 samples/sec	accuracy=0.728691	lr=0.171169
Epoch[301] Batch [99]	Speed: 8589.774479 samples/sec	accuracy=0.727734	lr=0.170713
Epoch[301] Batch [149]	Speed: 8824.010169 samples/sec	accuracy=0.726543	lr=0.170258
Epoch[301] Batch [199]	Speed: 8586.515442 samples/sec	accuracy=0.725793	lr=0.169803
Epoch[301] Batch [249]	Speed: 8817.378518 samples/sec	accuracy=0.725090	lr=0.169348
Epoch[301] Batch [299]	Speed: 8391.286328 samples/sec	accuracy=0.724002	lr=0.168894
Epoch[301] Batch [349]	Speed: 8498.184958 samples/sec	accuracy=0.723062	lr=0.168441
Epoch[301] Batch [399]	Speed: 8593.006764 samples/sec	accuracy=0.722410	lr=0.167988
Epoch[301] Batch [449]	Speed: 8894.648713 samples/sec	accuracy=0.721947	lr=0.167536
Epoch[301] Batch [499]	Speed: 8622.635939 samples/sec	accuracy=0.721681	lr=0.167084
Epoch[301] Batch [549]	Speed: 8426.033412 samples/sec	accuracy=0.721199	lr=0.166633
Epoch[301] Batch [599]	Speed: 8546.418621 samples/sec	accuracy=0.720868	lr=0.166182
[Epoch 301] training: accuracy=0.720824
[Epoch 301] speed: 8630 samples/sec	time cost: 153.434358
[Epoch 301] validation: err-top1=0.273652 err-top5=0.089355
Epoch[302] Batch [49]	Speed: 8815.988528 samples/sec	accuracy=0.730059	lr=0.165499
Epoch[302] Batch [99]	Speed: 8752.801113 samples/sec	accuracy=0.727876	lr=0.165050
Epoch[302] Batch [149]	Speed: 8740.894561 samples/sec	accuracy=0.727396	lr=0.164601
Epoch[302] Batch [199]	Speed: 8569.825667 samples/sec	accuracy=0.725625	lr=0.164153
Epoch[302] Batch [249]	Speed: 8612.213354 samples/sec	accuracy=0.725570	lr=0.163706
Epoch[302] Batch [299]	Speed: 8615.704509 samples/sec	accuracy=0.725234	lr=0.163259
Epoch[302] Batch [349]	Speed: 8718.216876 samples/sec	accuracy=0.724699	lr=0.162813
Epoch[302] Batch [399]	Speed: 8627.732862 samples/sec	accuracy=0.724187	lr=0.162367
Epoch[302] Batch [449]	Speed: 8446.065827 samples/sec	accuracy=0.723953	lr=0.161922
Epoch[302] Batch [499]	Speed: 8819.290109 samples/sec	accuracy=0.723484	lr=0.161477
Epoch[302] Batch [549]	Speed: 8396.703603 samples/sec	accuracy=0.723411	lr=0.161033
Epoch[302] Batch [599]	Speed: 8818.738708 samples/sec	accuracy=0.722791	lr=0.160590
[Epoch 302] training: accuracy=0.722570
[Epoch 302] speed: 8657 samples/sec	time cost: 152.647203
[Epoch 302] validation: err-top1=0.276408 err-top5=0.090800
Epoch[303] Batch [49]	Speed: 8698.756206 samples/sec	accuracy=0.728896	lr=0.159926
Epoch[303] Batch [99]	Speed: 8839.616454 samples/sec	accuracy=0.728467	lr=0.159484
Epoch[303] Batch [149]	Speed: 8824.012525 samples/sec	accuracy=0.728317	lr=0.159043
Epoch[303] Batch [199]	Speed: 8581.292030 samples/sec	accuracy=0.727881	lr=0.158602
Epoch[303] Batch [249]	Speed: 8814.681829 samples/sec	accuracy=0.727180	lr=0.158161
Epoch[303] Batch [299]	Speed: 8754.468168 samples/sec	accuracy=0.726885	lr=0.157722
Epoch[303] Batch [349]	Speed: 8614.712918 samples/sec	accuracy=0.726528	lr=0.157283
Epoch[303] Batch [399]	Speed: 8773.287886 samples/sec	accuracy=0.726097	lr=0.156844
Epoch[303] Batch [449]	Speed: 8785.047690 samples/sec	accuracy=0.725583	lr=0.156406
Epoch[303] Batch [499]	Speed: 8651.841111 samples/sec	accuracy=0.725228	lr=0.155969
Epoch[303] Batch [549]	Speed: 8812.410776 samples/sec	accuracy=0.724901	lr=0.155532
Epoch[303] Batch [599]	Speed: 8495.826324 samples/sec	accuracy=0.724561	lr=0.155096
[Epoch 303] training: accuracy=0.724323
[Epoch 303] speed: 8717 samples/sec	time cost: 151.760826
[Epoch 303] validation: err-top1=0.272481 err-top5=0.088847
Epoch[304] Batch [49]	Speed: 8663.856522 samples/sec	accuracy=0.732197	lr=0.154434
Epoch[304] Batch [99]	Speed: 8904.250909 samples/sec	accuracy=0.731079	lr=0.153999
Epoch[304] Batch [149]	Speed: 8470.078075 samples/sec	accuracy=0.730508	lr=0.153565
Epoch[304] Batch [199]	Speed: 8495.823467 samples/sec	accuracy=0.729434	lr=0.153131
Epoch[304] Batch [249]	Speed: 8783.795237 samples/sec	accuracy=0.728971	lr=0.152698
Epoch[304] Batch [299]	Speed: 8830.137829 samples/sec	accuracy=0.728411	lr=0.152266
Epoch[304] Batch [349]	Speed: 8970.574175 samples/sec	accuracy=0.727817	lr=0.151834
Epoch[304] Batch [399]	Speed: 8810.094989 samples/sec	accuracy=0.727439	lr=0.151402
Epoch[304] Batch [449]	Speed: 8810.016197 samples/sec	accuracy=0.727000	lr=0.150971
Epoch[304] Batch [499]	Speed: 8583.013936 samples/sec	accuracy=0.726581	lr=0.150541
Epoch[304] Batch [549]	Speed: 8793.896535 samples/sec	accuracy=0.725938	lr=0.150112
Epoch[304] Batch [599]	Speed: 8634.508168 samples/sec	accuracy=0.725700	lr=0.149683
[Epoch 304] training: accuracy=0.725691
[Epoch 304] speed: 8723 samples/sec	time cost: 151.592155
[Epoch 304] validation: err-top1=0.276484 err-top5=0.090938
Epoch[305] Batch [49]	Speed: 8769.499567 samples/sec	accuracy=0.733057	lr=0.149040
Epoch[305] Batch [99]	Speed: 8725.584802 samples/sec	accuracy=0.731987	lr=0.148613
Epoch[305] Batch [149]	Speed: 8642.818268 samples/sec	accuracy=0.731283	lr=0.148185
Epoch[305] Batch [199]	Speed: 8578.099560 samples/sec	accuracy=0.730911	lr=0.147759
Epoch[305] Batch [249]	Speed: 8706.225209 samples/sec	accuracy=0.730141	lr=0.147333
Epoch[305] Batch [299]	Speed: 8758.572319 samples/sec	accuracy=0.729520	lr=0.146908
Epoch[305] Batch [349]	Speed: 8717.276568 samples/sec	accuracy=0.728902	lr=0.146483
Epoch[305] Batch [399]	Speed: 8668.322930 samples/sec	accuracy=0.728698	lr=0.146059
Epoch[305] Batch [449]	Speed: 8596.937141 samples/sec	accuracy=0.727992	lr=0.145635
Epoch[305] Batch [499]	Speed: 8708.881011 samples/sec	accuracy=0.727421	lr=0.145212
Epoch[305] Batch [549]	Speed: 8481.172712 samples/sec	accuracy=0.727288	lr=0.144790
Epoch[305] Batch [599]	Speed: 8431.710840 samples/sec	accuracy=0.726851	lr=0.144368
[Epoch 305] training: accuracy=0.726694
[Epoch 305] speed: 8651 samples/sec	time cost: 152.666578
[Epoch 305] validation: err-top1=0.273763 err-top5=0.089193
Epoch[306] Batch [49]	Speed: 8466.450564 samples/sec	accuracy=0.732930	lr=0.143728
Epoch[306] Batch [99]	Speed: 8529.857441 samples/sec	accuracy=0.733638	lr=0.143308
Epoch[306] Batch [149]	Speed: 8541.007431 samples/sec	accuracy=0.732676	lr=0.142888
Epoch[306] Batch [199]	Speed: 8731.499795 samples/sec	accuracy=0.731560	lr=0.142469
Epoch[306] Batch [249]	Speed: 8712.686058 samples/sec	accuracy=0.731328	lr=0.142050
Epoch[306] Batch [299]	Speed: 8775.569120 samples/sec	accuracy=0.730614	lr=0.141632
Epoch[306] Batch [349]	Speed: 8830.019103 samples/sec	accuracy=0.730280	lr=0.141215
Epoch[306] Batch [399]	Speed: 8491.620310 samples/sec	accuracy=0.730016	lr=0.140798
Epoch[306] Batch [449]	Speed: 8765.745288 samples/sec	accuracy=0.729772	lr=0.140382
Epoch[306] Batch [499]	Speed: 8458.208983 samples/sec	accuracy=0.729591	lr=0.139966
Epoch[306] Batch [549]	Speed: 8716.768277 samples/sec	accuracy=0.729464	lr=0.139551
Epoch[306] Batch [599]	Speed: 8580.132991 samples/sec	accuracy=0.729205	lr=0.139136
[Epoch 306] training: accuracy=0.728880
[Epoch 306] speed: 8628 samples/sec	time cost: 153.083530
[Epoch 306] validation: err-top1=0.271191 err-top5=0.087324
Epoch[307] Batch [49]	Speed: 8332.705312 samples/sec	accuracy=0.735449	lr=0.138515
Epoch[307] Batch [99]	Speed: 8676.938953 samples/sec	accuracy=0.735776	lr=0.138102
Epoch[307] Batch [149]	Speed: 8701.560648 samples/sec	accuracy=0.734554	lr=0.137690
Epoch[307] Batch [199]	Speed: 8562.757086 samples/sec	accuracy=0.733362	lr=0.137278
Epoch[307] Batch [249]	Speed: 8883.695298 samples/sec	accuracy=0.732379	lr=0.136866
Epoch[307] Batch [299]	Speed: 8508.371808 samples/sec	accuracy=0.732020	lr=0.136456
Epoch[307] Batch [349]	Speed: 8762.336364 samples/sec	accuracy=0.731540	lr=0.136046
Epoch[307] Batch [399]	Speed: 8559.221537 samples/sec	accuracy=0.731312	lr=0.135636
Epoch[307] Batch [449]	Speed: 8694.803838 samples/sec	accuracy=0.730751	lr=0.135227
Epoch[307] Batch [499]	Speed: 8525.541010 samples/sec	accuracy=0.730396	lr=0.134818
Epoch[307] Batch [549]	Speed: 8638.550106 samples/sec	accuracy=0.730282	lr=0.134411
Epoch[307] Batch [599]	Speed: 8573.401815 samples/sec	accuracy=0.729779	lr=0.134003
[Epoch 307] training: accuracy=0.729601
[Epoch 307] speed: 8616 samples/sec	time cost: 153.339411
[Epoch 307] validation: err-top1=0.269572 err-top5=0.086466
Epoch[308] Batch [49]	Speed: 8608.539712 samples/sec	accuracy=0.736709	lr=0.133385
Epoch[308] Batch [99]	Speed: 8868.486663 samples/sec	accuracy=0.735747	lr=0.132980
Epoch[308] Batch [149]	Speed: 8691.867952 samples/sec	accuracy=0.734997	lr=0.132574
Epoch[308] Batch [199]	Speed: 8569.623556 samples/sec	accuracy=0.734937	lr=0.132170
Epoch[308] Batch [249]	Speed: 8403.783394 samples/sec	accuracy=0.734607	lr=0.131766
Epoch[308] Batch [299]	Speed: 8673.834862 samples/sec	accuracy=0.734103	lr=0.131362
Epoch[308] Batch [349]	Speed: 8395.481637 samples/sec	accuracy=0.733705	lr=0.130959
Epoch[308] Batch [399]	Speed: 8568.810418 samples/sec	accuracy=0.733450	lr=0.130557
Epoch[308] Batch [449]	Speed: 8610.381147 samples/sec	accuracy=0.733047	lr=0.130155
Epoch[308] Batch [499]	Speed: 8775.366690 samples/sec	accuracy=0.732514	lr=0.129754
Epoch[308] Batch [549]	Speed: 8529.368569 samples/sec	accuracy=0.731854	lr=0.129354
Epoch[308] Batch [599]	Speed: 8505.326833 samples/sec	accuracy=0.731597	lr=0.128954
[Epoch 308] training: accuracy=0.731363
[Epoch 308] speed: 8601 samples/sec	time cost: 153.694991
[Epoch 308] validation: err-top1=0.272339 err-top5=0.086304
Epoch[309] Batch [49]	Speed: 8706.432404 samples/sec	accuracy=0.739131	lr=0.128347
Epoch[309] Batch [99]	Speed: 8550.016446 samples/sec	accuracy=0.738369	lr=0.127949
Epoch[309] Batch [149]	Speed: 8704.061374 samples/sec	accuracy=0.737031	lr=0.127551
Epoch[309] Batch [199]	Speed: 8641.203718 samples/sec	accuracy=0.736516	lr=0.127154
Epoch[309] Batch [249]	Speed: 8533.164812 samples/sec	accuracy=0.735400	lr=0.126757
Epoch[309] Batch [299]	Speed: 8605.868530 samples/sec	accuracy=0.734849	lr=0.126361
Epoch[309] Batch [349]	Speed: 8866.177737 samples/sec	accuracy=0.734654	lr=0.125965
Epoch[309] Batch [399]	Speed: 8823.854988 samples/sec	accuracy=0.734237	lr=0.125570
Epoch[309] Batch [449]	Speed: 8921.303010 samples/sec	accuracy=0.733866	lr=0.125176
Epoch[309] Batch [499]	Speed: 8575.417953 samples/sec	accuracy=0.733522	lr=0.124782
Epoch[309] Batch [549]	Speed: 8658.697639 samples/sec	accuracy=0.732994	lr=0.124389
Epoch[309] Batch [599]	Speed: 8670.173230 samples/sec	accuracy=0.732824	lr=0.123996
[Epoch 309] training: accuracy=0.732801
[Epoch 309] speed: 8690 samples/sec	time cost: 152.156011
[Epoch 309] validation: err-top1=0.267168 err-top5=0.086484
Epoch[310] Batch [49]	Speed: 8582.460813 samples/sec	accuracy=0.742275	lr=0.123409
Epoch[310] Batch [99]	Speed: 8436.460361 samples/sec	accuracy=0.739648	lr=0.123018
Epoch[310] Batch [149]	Speed: 8502.149558 samples/sec	accuracy=0.738561	lr=0.122627
Epoch[310] Batch [199]	Speed: 8720.211765 samples/sec	accuracy=0.737734	lr=0.122237
Epoch[310] Batch [249]	Speed: 8706.497706 samples/sec	accuracy=0.737080	lr=0.121848
Epoch[310] Batch [299]	Speed: 8683.226132 samples/sec	accuracy=0.736400	lr=0.121459
Epoch[310] Batch [349]	Speed: 8655.241679 samples/sec	accuracy=0.736182	lr=0.121071
Epoch[310] Batch [399]	Speed: 8649.509821 samples/sec	accuracy=0.735942	lr=0.120683
Epoch[310] Batch [449]	Speed: 8698.749335 samples/sec	accuracy=0.735409	lr=0.120296
Epoch[310] Batch [499]	Speed: 8679.439401 samples/sec	accuracy=0.735112	lr=0.119910
Epoch[310] Batch [549]	Speed: 8699.859404 samples/sec	accuracy=0.734546	lr=0.119524
Epoch[310] Batch [599]	Speed: 8820.666468 samples/sec	accuracy=0.734064	lr=0.119139
[Epoch 310] training: accuracy=0.734070
[Epoch 310] speed: 8656 samples/sec	time cost: 152.722706
[Epoch 310] validation: err-top1=0.263143 err-top5=0.083903
Epoch[311] Batch [49]	Speed: 8510.654610 samples/sec	accuracy=0.741201	lr=0.118555
Epoch[311] Batch [99]	Speed: 8672.127105 samples/sec	accuracy=0.740928	lr=0.118171
Epoch[311] Batch [149]	Speed: 8571.043669 samples/sec	accuracy=0.740179	lr=0.117788
Epoch[311] Batch [199]	Speed: 8636.784658 samples/sec	accuracy=0.739097	lr=0.117405
Epoch[311] Batch [249]	Speed: 8559.333093 samples/sec	accuracy=0.738691	lr=0.117024
Epoch[311] Batch [299]	Speed: 8617.118326 samples/sec	accuracy=0.738369	lr=0.116642
Epoch[311] Batch [349]	Speed: 8744.388653 samples/sec	accuracy=0.737952	lr=0.116261
Epoch[311] Batch [399]	Speed: 8571.986567 samples/sec	accuracy=0.737405	lr=0.115881
Epoch[311] Batch [449]	Speed: 8747.152411 samples/sec	accuracy=0.737025	lr=0.115502
Epoch[311] Batch [499]	Speed: 8837.596207 samples/sec	accuracy=0.736481	lr=0.115123
Epoch[311] Batch [549]	Speed: 8551.897973 samples/sec	accuracy=0.736299	lr=0.114744
Epoch[311] Batch [599]	Speed: 8604.912647 samples/sec	accuracy=0.735749	lr=0.114367
[Epoch 311] training: accuracy=0.735559
[Epoch 311] speed: 8640 samples/sec	time cost: 152.812473
[Epoch 311] validation: err-top1=0.266738 err-top5=0.085059
Epoch[312] Batch [49]	Speed: 8383.000231 samples/sec	accuracy=0.741494	lr=0.113801
Epoch[312] Batch [99]	Speed: 8506.653771 samples/sec	accuracy=0.740825	lr=0.113425
Epoch[312] Batch [149]	Speed: 8558.322886 samples/sec	accuracy=0.740446	lr=0.113049
Epoch[312] Batch [199]	Speed: 8472.800832 samples/sec	accuracy=0.739739	lr=0.112674
Epoch[312] Batch [249]	Speed: 8752.751704 samples/sec	accuracy=0.739080	lr=0.112300
Epoch[312] Batch [299]	Speed: 8432.594190 samples/sec	accuracy=0.739085	lr=0.111926
Epoch[312] Batch [349]	Speed: 8634.122303 samples/sec	accuracy=0.738929	lr=0.111552
Epoch[312] Batch [399]	Speed: 8660.380378 samples/sec	accuracy=0.738782	lr=0.111180
Epoch[312] Batch [449]	Speed: 8719.637101 samples/sec	accuracy=0.738454	lr=0.110808
Epoch[312] Batch [499]	Speed: 8537.370853 samples/sec	accuracy=0.738224	lr=0.110436
Epoch[312] Batch [549]	Speed: 8647.353712 samples/sec	accuracy=0.737844	lr=0.110065
Epoch[312] Batch [599]	Speed: 8676.283043 samples/sec	accuracy=0.737476	lr=0.109695
[Epoch 312] training: accuracy=0.737281
[Epoch 312] speed: 8585 samples/sec	time cost: 153.768827
[Epoch 312] validation: err-top1=0.272685 err-top5=0.089254
Epoch[313] Batch [49]	Speed: 8723.746570 samples/sec	accuracy=0.741895	lr=0.109133
Epoch[313] Batch [99]	Speed: 8565.897314 samples/sec	accuracy=0.742627	lr=0.108764
Epoch[313] Batch [149]	Speed: 8655.861965 samples/sec	accuracy=0.742321	lr=0.108396
Epoch[313] Batch [199]	Speed: 8505.595152 samples/sec	accuracy=0.741321	lr=0.108028
Epoch[313] Batch [249]	Speed: 8420.145557 samples/sec	accuracy=0.740859	lr=0.107661
Epoch[313] Batch [299]	Speed: 8485.975956 samples/sec	accuracy=0.740610	lr=0.107295
Epoch[313] Batch [349]	Speed: 8905.157025 samples/sec	accuracy=0.740201	lr=0.106929
Epoch[313] Batch [399]	Speed: 8561.278623 samples/sec	accuracy=0.739711	lr=0.106564
Epoch[313] Batch [449]	Speed: 8640.944855 samples/sec	accuracy=0.739329	lr=0.106199
Epoch[313] Batch [499]	Speed: 8668.725155 samples/sec	accuracy=0.738942	lr=0.105835
Epoch[313] Batch [549]	Speed: 8874.156429 samples/sec	accuracy=0.738651	lr=0.105472
Epoch[313] Batch [599]	Speed: 8615.401374 samples/sec	accuracy=0.738287	lr=0.105109
[Epoch 313] training: accuracy=0.738258
[Epoch 313] speed: 8634 samples/sec	time cost: 152.980907
[Epoch 313] validation: err-top1=0.264961 err-top5=0.083086
Epoch[314] Batch [49]	Speed: 8477.080909 samples/sec	accuracy=0.745713	lr=0.104566
Epoch[314] Batch [99]	Speed: 8635.143714 samples/sec	accuracy=0.745420	lr=0.104204
Epoch[314] Batch [149]	Speed: 8570.381609 samples/sec	accuracy=0.744984	lr=0.103844
Epoch[314] Batch [199]	Speed: 8668.945966 samples/sec	accuracy=0.743574	lr=0.103484
Epoch[314] Batch [249]	Speed: 8881.740995 samples/sec	accuracy=0.743189	lr=0.103124
Epoch[314] Batch [299]	Speed: 8732.400031 samples/sec	accuracy=0.743000	lr=0.102765
Epoch[314] Batch [349]	Speed: 8564.386342 samples/sec	accuracy=0.742850	lr=0.102407
Epoch[314] Batch [399]	Speed: 8830.929783 samples/sec	accuracy=0.742096	lr=0.102049
Epoch[314] Batch [449]	Speed: 8645.713798 samples/sec	accuracy=0.741705	lr=0.101692
Epoch[314] Batch [499]	Speed: 8554.606463 samples/sec	accuracy=0.741287	lr=0.101335
Epoch[314] Batch [549]	Speed: 8750.538475 samples/sec	accuracy=0.740895	lr=0.100979
Epoch[314] Batch [599]	Speed: 8820.927153 samples/sec	accuracy=0.740639	lr=0.100624
[Epoch 314] training: accuracy=0.740329
[Epoch 314] speed: 8675 samples/sec	time cost: 152.245460
[Epoch 314] validation: err-top1=0.264587 err-top5=0.083679
Epoch[315] Batch [49]	Speed: 8688.230803 samples/sec	accuracy=0.745820	lr=0.100085
Epoch[315] Batch [99]	Speed: 8394.305471 samples/sec	accuracy=0.745420	lr=0.099731
Epoch[315] Batch [149]	Speed: 8775.709697 samples/sec	accuracy=0.744691	lr=0.099378
Epoch[315] Batch [199]	Speed: 8469.222593 samples/sec	accuracy=0.743975	lr=0.099025
Epoch[315] Batch [249]	Speed: 8563.711651 samples/sec	accuracy=0.743607	lr=0.098673
Epoch[315] Batch [299]	Speed: 8483.218754 samples/sec	accuracy=0.743348	lr=0.098322
Epoch[315] Batch [349]	Speed: 8740.051443 samples/sec	accuracy=0.743015	lr=0.097971
Epoch[315] Batch [399]	Speed: 8578.600374 samples/sec	accuracy=0.742905	lr=0.097621
Epoch[315] Batch [449]	Speed: 8639.986031 samples/sec	accuracy=0.742452	lr=0.097271
Epoch[315] Batch [499]	Speed: 8622.357761 samples/sec	accuracy=0.742186	lr=0.096922
Epoch[315] Batch [549]	Speed: 8384.042299 samples/sec	accuracy=0.741916	lr=0.096574
Epoch[315] Batch [599]	Speed: 8947.708353 samples/sec	accuracy=0.741566	lr=0.096226
[Epoch 315] training: accuracy=0.741219
[Epoch 315] speed: 8605 samples/sec	time cost: 153.624398
[Epoch 315] validation: err-top1=0.266764 err-top5=0.084757
Epoch[316] Batch [49]	Speed: 8399.747837 samples/sec	accuracy=0.744951	lr=0.095698
Epoch[316] Batch [99]	Speed: 8813.077303 samples/sec	accuracy=0.745527	lr=0.095352
Epoch[316] Batch [149]	Speed: 8479.928214 samples/sec	accuracy=0.745605	lr=0.095007
Epoch[316] Batch [199]	Speed: 8517.135906 samples/sec	accuracy=0.745686	lr=0.094661
Epoch[316] Batch [249]	Speed: 8864.771588 samples/sec	accuracy=0.745699	lr=0.094317
Epoch[316] Batch [299]	Speed: 8707.354133 samples/sec	accuracy=0.744771	lr=0.093973
Epoch[316] Batch [349]	Speed: 8871.756220 samples/sec	accuracy=0.744302	lr=0.093630
Epoch[316] Batch [399]	Speed: 8607.152685 samples/sec	accuracy=0.744176	lr=0.093287
Epoch[316] Batch [449]	Speed: 8611.351540 samples/sec	accuracy=0.743854	lr=0.092945
Epoch[316] Batch [499]	Speed: 8809.829703 samples/sec	accuracy=0.743413	lr=0.092604
Epoch[316] Batch [549]	Speed: 8872.591031 samples/sec	accuracy=0.743139	lr=0.092263
Epoch[316] Batch [599]	Speed: 8598.271132 samples/sec	accuracy=0.742734	lr=0.091923
[Epoch 316] training: accuracy=0.742687
[Epoch 316] speed: 8679 samples/sec	time cost: 152.249061
[Epoch 316] validation: err-top1=0.269746 err-top5=0.085918
Epoch[317] Batch [49]	Speed: 8769.994685 samples/sec	accuracy=0.747207	lr=0.091413
Epoch[317] Batch [99]	Speed: 8489.926316 samples/sec	accuracy=0.747241	lr=0.091075
Epoch[317] Batch [149]	Speed: 8582.677938 samples/sec	accuracy=0.747018	lr=0.090737
Epoch[317] Batch [199]	Speed: 8494.079415 samples/sec	accuracy=0.746155	lr=0.090399
Epoch[317] Batch [249]	Speed: 8736.716141 samples/sec	accuracy=0.745525	lr=0.090062
Epoch[317] Batch [299]	Speed: 8448.876875 samples/sec	accuracy=0.744842	lr=0.089726
Epoch[317] Batch [349]	Speed: 8550.936847 samples/sec	accuracy=0.744593	lr=0.089390
Epoch[317] Batch [399]	Speed: 8680.538757 samples/sec	accuracy=0.744478	lr=0.089055
Epoch[317] Batch [449]	Speed: 8749.327210 samples/sec	accuracy=0.744284	lr=0.088721
Epoch[317] Batch [499]	Speed: 8686.039715 samples/sec	accuracy=0.744228	lr=0.088387
Epoch[317] Batch [549]	Speed: 8585.994822 samples/sec	accuracy=0.743952	lr=0.088053
Epoch[317] Batch [599]	Speed: 8580.207040 samples/sec	accuracy=0.743663	lr=0.087721
[Epoch 317] training: accuracy=0.743612
[Epoch 317] speed: 8613 samples/sec	time cost: 153.354845
[Epoch 317] validation: err-top1=0.260376 err-top5=0.082967
Epoch[318] Batch [49]	Speed: 8438.783822 samples/sec	accuracy=0.752051	lr=0.087216
Epoch[318] Batch [99]	Speed: 8678.402224 samples/sec	accuracy=0.750801	lr=0.086885
Epoch[318] Batch [149]	Speed: 8810.038244 samples/sec	accuracy=0.749163	lr=0.086555
Epoch[318] Batch [199]	Speed: 8558.741403 samples/sec	accuracy=0.748923	lr=0.086225
Epoch[318] Batch [249]	Speed: 8588.973486 samples/sec	accuracy=0.748238	lr=0.085895
Epoch[318] Batch [299]	Speed: 8503.272807 samples/sec	accuracy=0.747703	lr=0.085567
Epoch[318] Batch [349]	Speed: 8716.998266 samples/sec	accuracy=0.747134	lr=0.085239
Epoch[318] Batch [399]	Speed: 8635.195624 samples/sec	accuracy=0.746855	lr=0.084911
Epoch[318] Batch [449]	Speed: 8869.151078 samples/sec	accuracy=0.746682	lr=0.084584
Epoch[318] Batch [499]	Speed: 8507.905283 samples/sec	accuracy=0.746410	lr=0.084258
Epoch[318] Batch [549]	Speed: 8962.135671 samples/sec	accuracy=0.745987	lr=0.083932
Epoch[318] Batch [599]	Speed: 8715.747275 samples/sec	accuracy=0.745596	lr=0.083607
[Epoch 318] training: accuracy=0.745448
[Epoch 318] speed: 8666 samples/sec	time cost: 152.500115
[Epoch 318] validation: err-top1=0.261270 err-top5=0.083555
Epoch[319] Batch [49]	Speed: 8637.899466 samples/sec	accuracy=0.752656	lr=0.083121
Epoch[319] Batch [99]	Speed: 8656.159231 samples/sec	accuracy=0.750308	lr=0.082797
Epoch[319] Batch [149]	Speed: 8691.179709 samples/sec	accuracy=0.749938	lr=0.082475
Epoch[319] Batch [199]	Speed: 8487.063238 samples/sec	accuracy=0.749377	lr=0.082152
Epoch[319] Batch [249]	Speed: 8630.411741 samples/sec	accuracy=0.749238	lr=0.081831
Epoch[319] Batch [299]	Speed: 8841.227565 samples/sec	accuracy=0.748660	lr=0.081510
Epoch[319] Batch [349]	Speed: 8867.764674 samples/sec	accuracy=0.748154	lr=0.081189
Epoch[319] Batch [399]	Speed: 8652.034222 samples/sec	accuracy=0.747928	lr=0.080869
Epoch[319] Batch [449]	Speed: 8588.950985 samples/sec	accuracy=0.747632	lr=0.080550
Epoch[319] Batch [499]	Speed: 8974.426297 samples/sec	accuracy=0.747401	lr=0.080231
Epoch[319] Batch [549]	Speed: 8931.708378 samples/sec	accuracy=0.747132	lr=0.079913
Epoch[319] Batch [599]	Speed: 8617.125587 samples/sec	accuracy=0.746944	lr=0.079596
[Epoch 319] training: accuracy=0.746828
[Epoch 319] speed: 8716 samples/sec	time cost: 151.687633
[Epoch 319] validation: err-top1=0.260335 err-top5=0.083130
Epoch[320] Batch [49]	Speed: 8557.955907 samples/sec	accuracy=0.752793	lr=0.079115
Epoch[320] Batch [99]	Speed: 8781.787673 samples/sec	accuracy=0.752036	lr=0.078799
Epoch[320] Batch [149]	Speed: 8598.749686 samples/sec	accuracy=0.751315	lr=0.078484
Epoch[320] Batch [199]	Speed: 8562.579548 samples/sec	accuracy=0.750728	lr=0.078169
Epoch[320] Batch [249]	Speed: 8683.342348 samples/sec	accuracy=0.750510	lr=0.077855
Epoch[320] Batch [299]	Speed: 8790.862411 samples/sec	accuracy=0.749849	lr=0.077541
Epoch[320] Batch [349]	Speed: 8738.398764 samples/sec	accuracy=0.749668	lr=0.077229
Epoch[320] Batch [399]	Speed: 8797.069511 samples/sec	accuracy=0.749479	lr=0.076916
Epoch[320] Batch [449]	Speed: 8571.223782 samples/sec	accuracy=0.749070	lr=0.076605
Epoch[320] Batch [499]	Speed: 8693.042596 samples/sec	accuracy=0.748798	lr=0.076294
Epoch[320] Batch [549]	Speed: 8803.445795 samples/sec	accuracy=0.748643	lr=0.075984
Epoch[320] Batch [599]	Speed: 8482.426788 samples/sec	accuracy=0.748597	lr=0.075674
[Epoch 320] training: accuracy=0.748591
[Epoch 320] speed: 8669 samples/sec	time cost: 152.416326
[Epoch 320] validation: err-top1=0.262024 err-top5=0.083496
Epoch[321] Batch [49]	Speed: 8899.795017 samples/sec	accuracy=0.753779	lr=0.075210
Epoch[321] Batch [99]	Speed: 8886.786865 samples/sec	accuracy=0.753984	lr=0.074902
Epoch[321] Batch [149]	Speed: 8677.874085 samples/sec	accuracy=0.752751	lr=0.074595
Epoch[321] Batch [199]	Speed: 8525.863241 samples/sec	accuracy=0.752751	lr=0.074288
Epoch[321] Batch [249]	Speed: 8744.190863 samples/sec	accuracy=0.752387	lr=0.073981
Epoch[321] Batch [299]	Speed: 8655.025926 samples/sec	accuracy=0.752202	lr=0.073676
Epoch[321] Batch [349]	Speed: 9004.989129 samples/sec	accuracy=0.752026	lr=0.073370
Epoch[321] Batch [399]	Speed: 8808.171847 samples/sec	accuracy=0.751654	lr=0.073066
Epoch[321] Batch [449]	Speed: 8879.713199 samples/sec	accuracy=0.751292	lr=0.072762
Epoch[321] Batch [499]	Speed: 8751.106878 samples/sec	accuracy=0.751070	lr=0.072459
Epoch[321] Batch [549]	Speed: 8734.741228 samples/sec	accuracy=0.750732	lr=0.072156
Epoch[321] Batch [599]	Speed: 8909.764303 samples/sec	accuracy=0.750405	lr=0.071854
[Epoch 321] training: accuracy=0.750283
[Epoch 321] speed: 8780 samples/sec	time cost: 151.028994
[Epoch 321] validation: err-top1=0.264629 err-top5=0.084766
Epoch[322] Batch [49]	Speed: 8848.941882 samples/sec	accuracy=0.758096	lr=0.071396
Epoch[322] Batch [99]	Speed: 8812.010293 samples/sec	accuracy=0.755957	lr=0.071096
Epoch[322] Batch [149]	Speed: 8733.888996 samples/sec	accuracy=0.755671	lr=0.070796
Epoch[322] Batch [199]	Speed: 8804.934721 samples/sec	accuracy=0.755542	lr=0.070496
Epoch[322] Batch [249]	Speed: 8638.038620 samples/sec	accuracy=0.754725	lr=0.070198
Epoch[322] Batch [299]	Speed: 8623.858610 samples/sec	accuracy=0.754233	lr=0.069900
Epoch[322] Batch [349]	Speed: 8663.313201 samples/sec	accuracy=0.753584	lr=0.069602
Epoch[322] Batch [399]	Speed: 8585.295099 samples/sec	accuracy=0.753214	lr=0.069306
Epoch[322] Batch [449]	Speed: 8762.676565 samples/sec	accuracy=0.752620	lr=0.069009
Epoch[322] Batch [499]	Speed: 8804.879486 samples/sec	accuracy=0.752437	lr=0.068714
Epoch[322] Batch [549]	Speed: 8685.560705 samples/sec	accuracy=0.752315	lr=0.068419
Epoch[322] Batch [599]	Speed: 8587.183602 samples/sec	accuracy=0.752160	lr=0.068124
[Epoch 322] training: accuracy=0.752016
[Epoch 322] speed: 8715 samples/sec	time cost: 151.614738
[Epoch 322] validation: err-top1=0.257019 err-top5=0.081543
Epoch[323] Batch [49]	Speed: 8700.390927 samples/sec	accuracy=0.758506	lr=0.067678
Epoch[323] Batch [99]	Speed: 8820.382430 samples/sec	accuracy=0.757363	lr=0.067385
Epoch[323] Batch [149]	Speed: 8868.634262 samples/sec	accuracy=0.756758	lr=0.067093
Epoch[323] Batch [199]	Speed: 8651.540830 samples/sec	accuracy=0.755991	lr=0.066802
Epoch[323] Batch [249]	Speed: 8678.907629 samples/sec	accuracy=0.755584	lr=0.066511
Epoch[323] Batch [299]	Speed: 8508.671672 samples/sec	accuracy=0.754946	lr=0.066221
Epoch[323] Batch [349]	Speed: 8636.217984 samples/sec	accuracy=0.754528	lr=0.065931
Epoch[323] Batch [399]	Speed: 8634.300043 samples/sec	accuracy=0.753995	lr=0.065642
Epoch[323] Batch [449]	Speed: 8669.624217 samples/sec	accuracy=0.753447	lr=0.065353
Epoch[323] Batch [499]	Speed: 8673.680714 samples/sec	accuracy=0.753009	lr=0.065066
Epoch[323] Batch [549]	Speed: 8536.477632 samples/sec	accuracy=0.752840	lr=0.064778
Epoch[323] Batch [599]	Speed: 8637.083567 samples/sec	accuracy=0.752673	lr=0.064492
[Epoch 323] training: accuracy=0.752617
[Epoch 323] speed: 8666 samples/sec	time cost: 152.410178
[Epoch 323] validation: err-top1=0.259219 err-top5=0.082148
Epoch[324] Batch [49]	Speed: 8496.225977 samples/sec	accuracy=0.759424	lr=0.064063
Epoch[324] Batch [99]	Speed: 8569.899538 samples/sec	accuracy=0.759463	lr=0.063778
Epoch[324] Batch [149]	Speed: 8579.495234 samples/sec	accuracy=0.758711	lr=0.063493
Epoch[324] Batch [199]	Speed: 9020.052361 samples/sec	accuracy=0.757827	lr=0.063210
Epoch[324] Batch [249]	Speed: 8760.373082 samples/sec	accuracy=0.757240	lr=0.062927
Epoch[324] Batch [299]	Speed: 8452.091128 samples/sec	accuracy=0.756899	lr=0.062644
Epoch[324] Batch [349]	Speed: 8489.207765 samples/sec	accuracy=0.756777	lr=0.062362
Epoch[324] Batch [399]	Speed: 8514.905671 samples/sec	accuracy=0.756296	lr=0.062081
Epoch[324] Batch [449]	Speed: 8810.912995 samples/sec	accuracy=0.755932	lr=0.061800
Epoch[324] Batch [499]	Speed: 8510.609920 samples/sec	accuracy=0.755497	lr=0.061520
Epoch[324] Batch [549]	Speed: 8743.755437 samples/sec	accuracy=0.755318	lr=0.061240
Epoch[324] Batch [599]	Speed: 8756.894595 samples/sec	accuracy=0.755002	lr=0.060962
[Epoch 324] training: accuracy=0.754891
[Epoch 324] speed: 8640 samples/sec	time cost: 152.928832
[Epoch 324] validation: err-top1=0.256755 err-top5=0.080424
Epoch[325] Batch [49]	Speed: 8426.438264 samples/sec	accuracy=0.759668	lr=0.060539
Epoch[325] Batch [99]	Speed: 8815.690316 samples/sec	accuracy=0.757925	lr=0.060262
Epoch[325] Batch [149]	Speed: 8594.121648 samples/sec	accuracy=0.758148	lr=0.059985
Epoch[325] Batch [199]	Speed: 8623.259523 samples/sec	accuracy=0.757676	lr=0.059709
Epoch[325] Batch [249]	Speed: 8656.739517 samples/sec	accuracy=0.758035	lr=0.059434
Epoch[325] Batch [299]	Speed: 8663.965054 samples/sec	accuracy=0.757560	lr=0.059159
Epoch[325] Batch [349]	Speed: 8654.129889 samples/sec	accuracy=0.757274	lr=0.058885
Epoch[325] Batch [399]	Speed: 8451.380297 samples/sec	accuracy=0.756942	lr=0.058611
Epoch[325] Batch [449]	Speed: 8492.755053 samples/sec	accuracy=0.756610	lr=0.058338
Epoch[325] Batch [499]	Speed: 8843.871140 samples/sec	accuracy=0.756283	lr=0.058066
Epoch[325] Batch [549]	Speed: 8815.534704 samples/sec	accuracy=0.756179	lr=0.057794
Epoch[325] Batch [599]	Speed: 8503.881603 samples/sec	accuracy=0.756080	lr=0.057523
[Epoch 325] training: accuracy=0.755923
[Epoch 325] speed: 8632 samples/sec	time cost: 152.703500
[Epoch 325] validation: err-top1=0.256571 err-top5=0.080811
Epoch[326] Batch [49]	Speed: 8562.926778 samples/sec	accuracy=0.760820	lr=0.057118
Epoch[326] Batch [99]	Speed: 8463.626811 samples/sec	accuracy=0.761045	lr=0.056848
Epoch[326] Batch [149]	Speed: 8798.046757 samples/sec	accuracy=0.760586	lr=0.056579
Epoch[326] Batch [199]	Speed: 8703.627114 samples/sec	accuracy=0.760020	lr=0.056311
Epoch[326] Batch [249]	Speed: 8554.454990 samples/sec	accuracy=0.759074	lr=0.056043
Epoch[326] Batch [299]	Speed: 8468.449601 samples/sec	accuracy=0.758644	lr=0.055776
Epoch[326] Batch [349]	Speed: 8711.947861 samples/sec	accuracy=0.758809	lr=0.055510
Epoch[326] Batch [399]	Speed: 8930.649034 samples/sec	accuracy=0.758531	lr=0.055244
Epoch[326] Batch [449]	Speed: 8923.524122 samples/sec	accuracy=0.758329	lr=0.054979
Epoch[326] Batch [499]	Speed: 8731.980868 samples/sec	accuracy=0.758071	lr=0.054715
Epoch[326] Batch [549]	Speed: 8666.756549 samples/sec	accuracy=0.757783	lr=0.054451
Epoch[326] Batch [599]	Speed: 8829.202084 samples/sec	accuracy=0.757637	lr=0.054187
[Epoch 326] training: accuracy=0.757520
[Epoch 326] speed: 8696 samples/sec	time cost: 152.221044
[Epoch 326] validation: err-top1=0.260000 err-top5=0.081562
Epoch[327] Batch [49]	Speed: 8628.734040 samples/sec	accuracy=0.763418	lr=0.053788
Epoch[327] Batch [99]	Speed: 8598.417619 samples/sec	accuracy=0.762583	lr=0.053527
Epoch[327] Batch [149]	Speed: 8629.685166 samples/sec	accuracy=0.761390	lr=0.053266
Epoch[327] Batch [199]	Speed: 8769.737002 samples/sec	accuracy=0.760662	lr=0.053005
Epoch[327] Batch [249]	Speed: 8929.295504 samples/sec	accuracy=0.760186	lr=0.052745
Epoch[327] Batch [299]	Speed: 8822.393908 samples/sec	accuracy=0.759880	lr=0.052486
Epoch[327] Batch [349]	Speed: 8944.794631 samples/sec	accuracy=0.759775	lr=0.052228
Epoch[327] Batch [399]	Speed: 8714.623601 samples/sec	accuracy=0.759387	lr=0.051970
Epoch[327] Batch [449]	Speed: 8697.265105 samples/sec	accuracy=0.759110	lr=0.051712
Epoch[327] Batch [499]	Speed: 8671.855355 samples/sec	accuracy=0.759049	lr=0.051456
Epoch[327] Batch [549]	Speed: 8790.207336 samples/sec	accuracy=0.758786	lr=0.051200
Epoch[327] Batch [599]	Speed: 8515.306277 samples/sec	accuracy=0.758427	lr=0.050944
[Epoch 327] training: accuracy=0.758347
[Epoch 327] speed: 8724 samples/sec	time cost: 151.180690
[Epoch 327] validation: err-top1=0.256490 err-top5=0.081075
Epoch[328] Batch [49]	Speed: 8560.074314 samples/sec	accuracy=0.763437	lr=0.050562
Epoch[328] Batch [99]	Speed: 8390.302448 samples/sec	accuracy=0.764180	lr=0.050308
Epoch[328] Batch [149]	Speed: 8558.012521 samples/sec	accuracy=0.763047	lr=0.050055
Epoch[328] Batch [199]	Speed: 8726.765741 samples/sec	accuracy=0.762190	lr=0.049803
Epoch[328] Batch [249]	Speed: 8698.661599 samples/sec	accuracy=0.761600	lr=0.049551
Epoch[328] Batch [299]	Speed: 8631.068018 samples/sec	accuracy=0.761395	lr=0.049299
Epoch[328] Batch [349]	Speed: 8707.074346 samples/sec	accuracy=0.761095	lr=0.049048
Epoch[328] Batch [399]	Speed: 8757.392756 samples/sec	accuracy=0.760840	lr=0.048798
Epoch[328] Batch [449]	Speed: 8587.874189 samples/sec	accuracy=0.760492	lr=0.048549
Epoch[328] Batch [499]	Speed: 8774.371172 samples/sec	accuracy=0.760469	lr=0.048300
Epoch[328] Batch [549]	Speed: 8841.283985 samples/sec	accuracy=0.760193	lr=0.048052
Epoch[328] Batch [599]	Speed: 8796.681772 samples/sec	accuracy=0.759930	lr=0.047804
[Epoch 328] training: accuracy=0.759720
[Epoch 328] speed: 8669 samples/sec	time cost: 152.576536
[Epoch 328] validation: err-top1=0.254941 err-top5=0.080371
Epoch[329] Batch [49]	Speed: 8575.458703 samples/sec	accuracy=0.765078	lr=0.047429
Epoch[329] Batch [99]	Speed: 8715.285851 samples/sec	accuracy=0.764268	lr=0.047183
Epoch[329] Batch [149]	Speed: 8655.940466 samples/sec	accuracy=0.764144	lr=0.046938
Epoch[329] Batch [199]	Speed: 8405.913017 samples/sec	accuracy=0.763606	lr=0.046693
Epoch[329] Batch [249]	Speed: 8528.256196 samples/sec	accuracy=0.763041	lr=0.046449
Epoch[329] Batch [299]	Speed: 8619.578702 samples/sec	accuracy=0.763060	lr=0.046205
Epoch[329] Batch [349]	Speed: 8735.710539 samples/sec	accuracy=0.763221	lr=0.045962
Epoch[329] Batch [399]	Speed: 8513.719773 samples/sec	accuracy=0.762870	lr=0.045720
Epoch[329] Batch [449]	Speed: 8711.683328 samples/sec	accuracy=0.762766	lr=0.045478
Epoch[329] Batch [499]	Speed: 8714.224000 samples/sec	accuracy=0.762282	lr=0.045237
Epoch[329] Batch [549]	Speed: 8690.781376 samples/sec	accuracy=0.762357	lr=0.044997
Epoch[329] Batch [599]	Speed: 8618.655918 samples/sec	accuracy=0.762177	lr=0.044757
[Epoch 329] training: accuracy=0.762225
[Epoch 329] speed: 8611 samples/sec	time cost: 153.403506
[Epoch 329] validation: err-top1=0.256124 err-top5=0.080221
Epoch[330] Batch [49]	Speed: 8448.281745 samples/sec	accuracy=0.767490	lr=0.044394
Epoch[330] Batch [99]	Speed: 8682.385326 samples/sec	accuracy=0.766338	lr=0.044156
Epoch[330] Batch [149]	Speed: 8446.223119 samples/sec	accuracy=0.765260	lr=0.043918
Epoch[330] Batch [199]	Speed: 8501.140856 samples/sec	accuracy=0.764590	lr=0.043682
Epoch[330] Batch [249]	Speed: 8756.157814 samples/sec	accuracy=0.764719	lr=0.043445
Epoch[330] Batch [299]	Speed: 8722.601346 samples/sec	accuracy=0.764419	lr=0.043210
Epoch[330] Batch [349]	Speed: 8634.702589 samples/sec	accuracy=0.764004	lr=0.042975
Epoch[330] Batch [399]	Speed: 8648.999997 samples/sec	accuracy=0.763811	lr=0.042740
Epoch[330] Batch [449]	Speed: 8422.372171 samples/sec	accuracy=0.763308	lr=0.042507
Epoch[330] Batch [499]	Speed: 8537.415655 samples/sec	accuracy=0.763225	lr=0.042273
Epoch[330] Batch [549]	Speed: 8755.390459 samples/sec	accuracy=0.762828	lr=0.042041
Epoch[330] Batch [599]	Speed: 8757.784360 samples/sec	accuracy=0.762693	lr=0.041809
[Epoch 330] training: accuracy=0.762764
[Epoch 330] speed: 8610 samples/sec	time cost: 153.442003
[Epoch 330] validation: err-top1=0.255117 err-top5=0.079980
Epoch[331] Batch [49]	Speed: 8464.607111 samples/sec	accuracy=0.767578	lr=0.041463
Epoch[331] Batch [99]	Speed: 8510.386478 samples/sec	accuracy=0.767090	lr=0.041232
Epoch[331] Batch [149]	Speed: 8538.091981 samples/sec	accuracy=0.766559	lr=0.041003
Epoch[331] Batch [199]	Speed: 8673.862889 samples/sec	accuracy=0.766418	lr=0.040774
Epoch[331] Batch [249]	Speed: 8578.648008 samples/sec	accuracy=0.765998	lr=0.040545
Epoch[331] Batch [299]	Speed: 8791.255575 samples/sec	accuracy=0.765754	lr=0.040317
Epoch[331] Batch [349]	Speed: 8570.201532 samples/sec	accuracy=0.765550	lr=0.040090
Epoch[331] Batch [399]	Speed: 8560.957293 samples/sec	accuracy=0.765167	lr=0.039864
Epoch[331] Batch [449]	Speed: 8473.322357 samples/sec	accuracy=0.765037	lr=0.039638
Epoch[331] Batch [499]	Speed: 8632.427891 samples/sec	accuracy=0.764721	lr=0.039413
Epoch[331] Batch [549]	Speed: 8608.133046 samples/sec	accuracy=0.764441	lr=0.039188
Epoch[331] Batch [599]	Speed: 8596.185910 samples/sec	accuracy=0.764358	lr=0.038964
[Epoch 331] training: accuracy=0.764183
[Epoch 331] speed: 8585 samples/sec	time cost: 153.891803
[Epoch 331] validation: err-top1=0.254130 err-top5=0.079753
Epoch[332] Batch [49]	Speed: 8543.546382 samples/sec	accuracy=0.767520	lr=0.038625
Epoch[332] Batch [99]	Speed: 8686.019865 samples/sec	accuracy=0.766826	lr=0.038403
Epoch[332] Batch [149]	Speed: 8377.369726 samples/sec	accuracy=0.767054	lr=0.038181
Epoch[332] Batch [199]	Speed: 8476.212805 samples/sec	accuracy=0.766980	lr=0.037960
Epoch[332] Batch [249]	Speed: 8568.603055 samples/sec	accuracy=0.766570	lr=0.037739
Epoch[332] Batch [299]	Speed: 8736.232236 samples/sec	accuracy=0.766242	lr=0.037520
Epoch[332] Batch [349]	Speed: 8743.422399 samples/sec	accuracy=0.766218	lr=0.037300
Epoch[332] Batch [399]	Speed: 8764.398713 samples/sec	accuracy=0.766178	lr=0.037082
Epoch[332] Batch [449]	Speed: 8749.417576 samples/sec	accuracy=0.766215	lr=0.036864
Epoch[332] Batch [499]	Speed: 8629.546281 samples/sec	accuracy=0.766157	lr=0.036647
Epoch[332] Batch [549]	Speed: 8785.916766 samples/sec	accuracy=0.765819	lr=0.036430
Epoch[332] Batch [599]	Speed: 8639.182077 samples/sec	accuracy=0.765608	lr=0.036214
[Epoch 332] training: accuracy=0.765548
[Epoch 332] speed: 8646 samples/sec	time cost: 152.546820
[Epoch 332] validation: err-top1=0.258036 err-top5=0.081421
Epoch[333] Batch [49]	Speed: 8357.990981 samples/sec	accuracy=0.770371	lr=0.035891
Epoch[333] Batch [99]	Speed: 8552.996765 samples/sec	accuracy=0.769790	lr=0.035677
Epoch[333] Batch [149]	Speed: 8768.409249 samples/sec	accuracy=0.769525	lr=0.035463
Epoch[333] Batch [199]	Speed: 8713.009334 samples/sec	accuracy=0.769236	lr=0.035250
Epoch[333] Batch [249]	Speed: 8618.155950 samples/sec	accuracy=0.768768	lr=0.035037
Epoch[333] Batch [299]	Speed: 8360.156851 samples/sec	accuracy=0.768483	lr=0.034825
Epoch[333] Batch [349]	Speed: 8578.844891 samples/sec	accuracy=0.768068	lr=0.034614
Epoch[333] Batch [399]	Speed: 8707.814718 samples/sec	accuracy=0.767794	lr=0.034403
Epoch[333] Batch [449]	Speed: 8730.929677 samples/sec	accuracy=0.767509	lr=0.034193
Epoch[333] Batch [499]	Speed: 8823.012828 samples/sec	accuracy=0.767149	lr=0.033984
Epoch[333] Batch [549]	Speed: 8668.110897 samples/sec	accuracy=0.767085	lr=0.033775
Epoch[333] Batch [599]	Speed: 8835.732658 samples/sec	accuracy=0.766938	lr=0.033567
[Epoch 333] training: accuracy=0.766930
[Epoch 333] speed: 8638 samples/sec	time cost: 153.157567
[Epoch 333] validation: err-top1=0.253672 err-top5=0.079492
Epoch[334] Batch [49]	Speed: 8589.095094 samples/sec	accuracy=0.770742	lr=0.033252
Epoch[334] Batch [99]	Speed: 8879.714668 samples/sec	accuracy=0.770361	lr=0.033045
Epoch[334] Batch [149]	Speed: 9000.355105 samples/sec	accuracy=0.769756	lr=0.032839
Epoch[334] Batch [199]	Speed: 8758.012054 samples/sec	accuracy=0.769336	lr=0.032634
Epoch[334] Batch [249]	Speed: 8681.578551 samples/sec	accuracy=0.769283	lr=0.032429
Epoch[334] Batch [299]	Speed: 8553.562449 samples/sec	accuracy=0.769310	lr=0.032226
Epoch[334] Batch [349]	Speed: 8671.575568 samples/sec	accuracy=0.769230	lr=0.032022
Epoch[334] Batch [399]	Speed: 8720.094383 samples/sec	accuracy=0.769178	lr=0.031819
Epoch[334] Batch [449]	Speed: 8795.753284 samples/sec	accuracy=0.769246	lr=0.031617
Epoch[334] Batch [499]	Speed: 8750.089047 samples/sec	accuracy=0.769053	lr=0.031416
Epoch[334] Batch [549]	Speed: 8669.959532 samples/sec	accuracy=0.768850	lr=0.031215
Epoch[334] Batch [599]	Speed: 8626.879549 samples/sec	accuracy=0.768722	lr=0.031015
[Epoch 334] training: accuracy=0.768666
[Epoch 334] speed: 8721 samples/sec	time cost: 151.302222
[Epoch 334] validation: err-top1=0.251038 err-top5=0.079692
Epoch[335] Batch [49]	Speed: 8415.668978 samples/sec	accuracy=0.772344	lr=0.030716
Epoch[335] Batch [99]	Speed: 8916.330605 samples/sec	accuracy=0.772334	lr=0.030517
Epoch[335] Batch [149]	Speed: 8956.758082 samples/sec	accuracy=0.771432	lr=0.030320
Epoch[335] Batch [199]	Speed: 8650.990515 samples/sec	accuracy=0.770864	lr=0.030122
Epoch[335] Batch [249]	Speed: 8850.620776 samples/sec	accuracy=0.770674	lr=0.029926
Epoch[335] Batch [299]	Speed: 8494.480249 samples/sec	accuracy=0.770169	lr=0.029730
Epoch[335] Batch [349]	Speed: 8675.818427 samples/sec	accuracy=0.770099	lr=0.029534
Epoch[335] Batch [399]	Speed: 8861.330380 samples/sec	accuracy=0.769797	lr=0.029339
Epoch[335] Batch [449]	Speed: 8838.897884 samples/sec	accuracy=0.769732	lr=0.029145
Epoch[335] Batch [499]	Speed: 8746.587549 samples/sec	accuracy=0.769437	lr=0.028952
Epoch[335] Batch [549]	Speed: 8723.606059 samples/sec	accuracy=0.769588	lr=0.028759
Epoch[335] Batch [599]	Speed: 8858.927784 samples/sec	accuracy=0.769609	lr=0.028567
[Epoch 335] training: accuracy=0.769474
[Epoch 335] speed: 8745 samples/sec	time cost: 151.359280
[Epoch 335] validation: err-top1=0.252500 err-top5=0.079453
Epoch[336] Batch [49]	Speed: 8426.088955 samples/sec	accuracy=0.770918	lr=0.028276
Epoch[336] Batch [99]	Speed: 8663.270738 samples/sec	accuracy=0.771221	lr=0.028085
Epoch[336] Batch [149]	Speed: 8793.087086 samples/sec	accuracy=0.771774	lr=0.027895
Epoch[336] Batch [199]	Speed: 8735.406540 samples/sec	accuracy=0.771987	lr=0.027706
Epoch[336] Batch [249]	Speed: 8817.110621 samples/sec	accuracy=0.771896	lr=0.027517
Epoch[336] Batch [299]	Speed: 8741.069253 samples/sec	accuracy=0.771528	lr=0.027329
Epoch[336] Batch [349]	Speed: 8739.158698 samples/sec	accuracy=0.771380	lr=0.027142
Epoch[336] Batch [399]	Speed: 8622.021618 samples/sec	accuracy=0.771274	lr=0.026955
Epoch[336] Batch [449]	Speed: 8877.202829 samples/sec	accuracy=0.771261	lr=0.026769
Epoch[336] Batch [499]	Speed: 8785.109325 samples/sec	accuracy=0.771222	lr=0.026583
Epoch[336] Batch [549]	Speed: 8583.530591 samples/sec	accuracy=0.771118	lr=0.026399
Epoch[336] Batch [599]	Speed: 8714.143024 samples/sec	accuracy=0.771046	lr=0.026214
[Epoch 336] training: accuracy=0.771065
[Epoch 336] speed: 8703 samples/sec	time cost: 152.012682
[Epoch 336] validation: err-top1=0.251465 err-top5=0.078410
Epoch[337] Batch [49]	Speed: 8523.279309 samples/sec	accuracy=0.775654	lr=0.025936
Epoch[337] Batch [99]	Speed: 8661.410282 samples/sec	accuracy=0.774780	lr=0.025753
Epoch[337] Batch [149]	Speed: 8401.054523 samples/sec	accuracy=0.773682	lr=0.025571
Epoch[337] Batch [199]	Speed: 8725.072706 samples/sec	accuracy=0.774170	lr=0.025390
Epoch[337] Batch [249]	Speed: 8980.587640 samples/sec	accuracy=0.773971	lr=0.025209
Epoch[337] Batch [299]	Speed: 8582.418453 samples/sec	accuracy=0.773369	lr=0.025029
Epoch[337] Batch [349]	Speed: 8559.667266 samples/sec	accuracy=0.773170	lr=0.024850
Epoch[337] Batch [399]	Speed: 8590.583181 samples/sec	accuracy=0.773138	lr=0.024671
Epoch[337] Batch [449]	Speed: 8511.947111 samples/sec	accuracy=0.773044	lr=0.024493
Epoch[337] Batch [499]	Speed: 8730.403289 samples/sec	accuracy=0.773055	lr=0.024315
Epoch[337] Batch [549]	Speed: 8647.031634 samples/sec	accuracy=0.772771	lr=0.024138
Epoch[337] Batch [599]	Speed: 8848.372184 samples/sec	accuracy=0.772528	lr=0.023962
[Epoch 337] training: accuracy=0.772559
[Epoch 337] speed: 8645 samples/sec	time cost: 152.636059
[Epoch 337] validation: err-top1=0.251160 err-top5=0.079447
Epoch[338] Batch [49]	Speed: 8572.579405 samples/sec	accuracy=0.776094	lr=0.023699
Epoch[338] Batch [99]	Speed: 8657.349023 samples/sec	accuracy=0.775459	lr=0.023524
Epoch[338] Batch [149]	Speed: 8792.625717 samples/sec	accuracy=0.775534	lr=0.023350
Epoch[338] Batch [199]	Speed: 8569.209103 samples/sec	accuracy=0.775398	lr=0.023177
Epoch[338] Batch [249]	Speed: 8482.648765 samples/sec	accuracy=0.775033	lr=0.023004
Epoch[338] Batch [299]	Speed: 8970.437216 samples/sec	accuracy=0.774735	lr=0.022832
Epoch[338] Batch [349]	Speed: 8607.726246 samples/sec	accuracy=0.774706	lr=0.022661
Epoch[338] Batch [399]	Speed: 8719.268018 samples/sec	accuracy=0.774435	lr=0.022490
Epoch[338] Batch [449]	Speed: 8561.036812 samples/sec	accuracy=0.774465	lr=0.022320
Epoch[338] Batch [499]	Speed: 8558.210674 samples/sec	accuracy=0.774216	lr=0.022150
Epoch[338] Batch [549]	Speed: 8442.149530 samples/sec	accuracy=0.774122	lr=0.021982
Epoch[338] Batch [599]	Speed: 8767.996288 samples/sec	accuracy=0.774007	lr=0.021813
[Epoch 338] training: accuracy=0.773895
[Epoch 338] speed: 8639 samples/sec	time cost: 153.180546
[Epoch 338] validation: err-top1=0.249746 err-top5=0.078438
Epoch[339] Batch [49]	Speed: 8263.335034 samples/sec	accuracy=0.778359	lr=0.021559
Epoch[339] Batch [99]	Speed: 8698.409676 samples/sec	accuracy=0.777104	lr=0.021392
Epoch[339] Batch [149]	Speed: 8573.747186 samples/sec	accuracy=0.776523	lr=0.021226
Epoch[339] Batch [199]	Speed: 8655.917788 samples/sec	accuracy=0.775945	lr=0.021061
Epoch[339] Batch [249]	Speed: 8744.586629 samples/sec	accuracy=0.775820	lr=0.020896
Epoch[339] Batch [299]	Speed: 8686.398436 samples/sec	accuracy=0.775552	lr=0.020732
Epoch[339] Batch [349]	Speed: 8504.406457 samples/sec	accuracy=0.775293	lr=0.020569
Epoch[339] Batch [399]	Speed: 8553.685442 samples/sec	accuracy=0.775106	lr=0.020406
Epoch[339] Batch [449]	Speed: 8810.931251 samples/sec	accuracy=0.774971	lr=0.020244
Epoch[339] Batch [499]	Speed: 8510.498282 samples/sec	accuracy=0.774800	lr=0.020083
Epoch[339] Batch [549]	Speed: 8648.478391 samples/sec	accuracy=0.774759	lr=0.019922
Epoch[339] Batch [599]	Speed: 8855.733043 samples/sec	accuracy=0.774579	lr=0.019762
[Epoch 339] training: accuracy=0.774531
[Epoch 339] speed: 8629 samples/sec	time cost: 152.899440
[Epoch 339] validation: err-top1=0.249247 err-top5=0.078267
Epoch[340] Batch [49]	Speed: 8396.683576 samples/sec	accuracy=0.777969	lr=0.019523
Epoch[340] Batch [99]	Speed: 8642.676004 samples/sec	accuracy=0.778247	lr=0.019364
Epoch[340] Batch [149]	Speed: 8550.664128 samples/sec	accuracy=0.777109	lr=0.019206
Epoch[340] Batch [199]	Speed: 8577.214411 samples/sec	accuracy=0.776467	lr=0.019049
Epoch[340] Batch [249]	Speed: 8617.867341 samples/sec	accuracy=0.776508	lr=0.018892
Epoch[340] Batch [299]	Speed: 8560.852861 samples/sec	accuracy=0.776361	lr=0.018736
Epoch[340] Batch [349]	Speed: 8519.562833 samples/sec	accuracy=0.776251	lr=0.018581
Epoch[340] Batch [399]	Speed: 8930.709014 samples/sec	accuracy=0.776118	lr=0.018426
Epoch[340] Batch [449]	Speed: 8903.722981 samples/sec	accuracy=0.776168	lr=0.018272
Epoch[340] Batch [499]	Speed: 8801.104780 samples/sec	accuracy=0.775990	lr=0.018118
Epoch[340] Batch [549]	Speed: 8748.299458 samples/sec	accuracy=0.775620	lr=0.017966
Epoch[340] Batch [599]	Speed: 8794.582954 samples/sec	accuracy=0.775497	lr=0.017813
[Epoch 340] training: accuracy=0.775491
[Epoch 340] speed: 8667 samples/sec	time cost: 152.694619
[Epoch 340] validation: err-top1=0.250762 err-top5=0.078223
Epoch[341] Batch [49]	Speed: 8440.947811 samples/sec	accuracy=0.776602	lr=0.017583
Epoch[341] Batch [99]	Speed: 8676.563835 samples/sec	accuracy=0.776563	lr=0.017433
Epoch[341] Batch [149]	Speed: 8690.074843 samples/sec	accuracy=0.776598	lr=0.017283
Epoch[341] Batch [199]	Speed: 8523.008351 samples/sec	accuracy=0.776211	lr=0.017134
Epoch[341] Batch [249]	Speed: 8645.204421 samples/sec	accuracy=0.776600	lr=0.016985
Epoch[341] Batch [299]	Speed: 8552.200913 samples/sec	accuracy=0.776847	lr=0.016837
Epoch[341] Batch [349]	Speed: 8638.152240 samples/sec	accuracy=0.777139	lr=0.016690
Epoch[341] Batch [399]	Speed: 8612.352373 samples/sec	accuracy=0.777260	lr=0.016543
Epoch[341] Batch [449]	Speed: 8786.465688 samples/sec	accuracy=0.777057	lr=0.016397
Epoch[341] Batch [499]	Speed: 8749.649468 samples/sec	accuracy=0.777065	lr=0.016252
Epoch[341] Batch [549]	Speed: 8592.476762 samples/sec	accuracy=0.777052	lr=0.016107
Epoch[341] Batch [599]	Speed: 8680.765960 samples/sec	accuracy=0.776985	lr=0.015963
[Epoch 341] training: accuracy=0.776977
[Epoch 341] speed: 8640 samples/sec	time cost: 152.605960
[Epoch 341] validation: err-top1=0.247599 err-top5=0.077209
Epoch[342] Batch [49]	Speed: 8692.266384 samples/sec	accuracy=0.781260	lr=0.015748
Epoch[342] Batch [99]	Speed: 8883.727454 samples/sec	accuracy=0.780361	lr=0.015605
Epoch[342] Batch [149]	Speed: 8677.910906 samples/sec	accuracy=0.779300	lr=0.015464
Epoch[342] Batch [199]	Speed: 8932.007432 samples/sec	accuracy=0.779961	lr=0.015322
Epoch[342] Batch [249]	Speed: 8827.850277 samples/sec	accuracy=0.780080	lr=0.015182
Epoch[342] Batch [299]	Speed: 8689.473027 samples/sec	accuracy=0.779025	lr=0.015042
Epoch[342] Batch [349]	Speed: 8847.629043 samples/sec	accuracy=0.779030	lr=0.014903
Epoch[342] Batch [399]	Speed: 8539.803724 samples/sec	accuracy=0.778838	lr=0.014764
Epoch[342] Batch [449]	Speed: 8563.016407 samples/sec	accuracy=0.778736	lr=0.014626
Epoch[342] Batch [499]	Speed: 8405.603901 samples/sec	accuracy=0.778550	lr=0.014489
Epoch[342] Batch [549]	Speed: 8849.758729 samples/sec	accuracy=0.778462	lr=0.014352
Epoch[342] Batch [599]	Speed: 8524.081121 samples/sec	accuracy=0.778308	lr=0.014216
[Epoch 342] training: accuracy=0.778254
[Epoch 342] speed: 8700 samples/sec	time cost: 152.089512
[Epoch 342] validation: err-top1=0.247812 err-top5=0.077500
Epoch[343] Batch [49]	Speed: 8584.417730 samples/sec	accuracy=0.779561	lr=0.014010
Epoch[343] Batch [99]	Speed: 8904.505851 samples/sec	accuracy=0.779229	lr=0.013876
Epoch[343] Batch [149]	Speed: 8770.322766 samples/sec	accuracy=0.779408	lr=0.013742
Epoch[343] Batch [199]	Speed: 8754.541687 samples/sec	accuracy=0.779329	lr=0.013609
Epoch[343] Batch [249]	Speed: 8518.749874 samples/sec	accuracy=0.778961	lr=0.013476
Epoch[343] Batch [299]	Speed: 8747.404849 samples/sec	accuracy=0.778895	lr=0.013345
Epoch[343] Batch [349]	Speed: 8539.048184 samples/sec	accuracy=0.778631	lr=0.013213
Epoch[343] Batch [399]	Speed: 8732.616108 samples/sec	accuracy=0.778799	lr=0.013083
Epoch[343] Batch [449]	Speed: 8655.601699 samples/sec	accuracy=0.778625	lr=0.012953
Epoch[343] Batch [499]	Speed: 8458.105544 samples/sec	accuracy=0.778496	lr=0.012824
Epoch[343] Batch [549]	Speed: 8406.334694 samples/sec	accuracy=0.778773	lr=0.012695
Epoch[343] Batch [599]	Speed: 8530.322819 samples/sec	accuracy=0.778820	lr=0.012567
[Epoch 343] training: accuracy=0.778729
[Epoch 343] speed: 8639 samples/sec	time cost: 152.913939
[Epoch 343] validation: err-top1=0.248352 err-top5=0.077108
Epoch[344] Batch [49]	Speed: 8423.398115 samples/sec	accuracy=0.780117	lr=0.012374
Epoch[344] Batch [99]	Speed: 8743.933803 samples/sec	accuracy=0.781182	lr=0.012247
Epoch[344] Batch [149]	Speed: 8768.271770 samples/sec	accuracy=0.780811	lr=0.012122
Epoch[344] Batch [199]	Speed: 8831.841377 samples/sec	accuracy=0.780291	lr=0.011997
Epoch[344] Batch [249]	Speed: 8582.882895 samples/sec	accuracy=0.780031	lr=0.011872
Epoch[344] Batch [299]	Speed: 8449.973623 samples/sec	accuracy=0.779901	lr=0.011748
Epoch[344] Batch [349]	Speed: 8582.434917 samples/sec	accuracy=0.779791	lr=0.011625
Epoch[344] Batch [399]	Speed: 8736.968866 samples/sec	accuracy=0.779563	lr=0.011503
Epoch[344] Batch [449]	Speed: 8965.473534 samples/sec	accuracy=0.779453	lr=0.011381
Epoch[344] Batch [499]	Speed: 8750.774706 samples/sec	accuracy=0.779622	lr=0.011260
Epoch[344] Batch [549]	Speed: 8937.325019 samples/sec	accuracy=0.779537	lr=0.011139
Epoch[344] Batch [599]	Speed: 8799.505913 samples/sec	accuracy=0.779468	lr=0.011019
[Epoch 344] training: accuracy=0.779417
[Epoch 344] speed: 8713 samples/sec	time cost: 151.560953
[Epoch 344] validation: err-top1=0.247721 err-top5=0.077799
Epoch[345] Batch [49]	Speed: 8634.969064 samples/sec	accuracy=0.779922	lr=0.010841
Epoch[345] Batch [99]	Speed: 8423.362101 samples/sec	accuracy=0.780659	lr=0.010722
Epoch[345] Batch [149]	Speed: 8576.719411 samples/sec	accuracy=0.780745	lr=0.010605
Epoch[345] Batch [199]	Speed: 8585.400127 samples/sec	accuracy=0.780752	lr=0.010488
Epoch[345] Batch [249]	Speed: 8610.735891 samples/sec	accuracy=0.780826	lr=0.010371
Epoch[345] Batch [299]	Speed: 8682.022372 samples/sec	accuracy=0.780285	lr=0.010256
Epoch[345] Batch [349]	Speed: 8712.589733 samples/sec	accuracy=0.779890	lr=0.010141
Epoch[345] Batch [399]	Speed: 8583.838350 samples/sec	accuracy=0.779908	lr=0.010026
Epoch[345] Batch [449]	Speed: 8958.510280 samples/sec	accuracy=0.779887	lr=0.009912
Epoch[345] Batch [499]	Speed: 8642.263324 samples/sec	accuracy=0.779769	lr=0.009799
Epoch[345] Batch [549]	Speed: 8642.519309 samples/sec	accuracy=0.779963	lr=0.009687
Epoch[345] Batch [599]	Speed: 8632.039263 samples/sec	accuracy=0.779998	lr=0.009575
[Epoch 345] training: accuracy=0.780001
[Epoch 345] speed: 8644 samples/sec	time cost: 153.036141
[Epoch 345] validation: err-top1=0.248184 err-top5=0.077207
Epoch[346] Batch [49]	Speed: 8520.429191 samples/sec	accuracy=0.782773	lr=0.009406
Epoch[346] Batch [99]	Speed: 8560.283311 samples/sec	accuracy=0.782656	lr=0.009296
Epoch[346] Batch [149]	Speed: 8549.362055 samples/sec	accuracy=0.782288	lr=0.009187
Epoch[346] Batch [199]	Speed: 8758.236901 samples/sec	accuracy=0.781599	lr=0.009078
Epoch[346] Batch [249]	Speed: 8689.599255 samples/sec	accuracy=0.781738	lr=0.008969
Epoch[346] Batch [299]	Speed: 8643.193257 samples/sec	accuracy=0.781398	lr=0.008862
Epoch[346] Batch [349]	Speed: 8852.892409 samples/sec	accuracy=0.781304	lr=0.008755
Epoch[346] Batch [399]	Speed: 8728.537660 samples/sec	accuracy=0.781110	lr=0.008648
Epoch[346] Batch [449]	Speed: 8518.256023 samples/sec	accuracy=0.780832	lr=0.008543
Epoch[346] Batch [499]	Speed: 8461.707901 samples/sec	accuracy=0.780661	lr=0.008438
Epoch[346] Batch [549]	Speed: 8658.904847 samples/sec	accuracy=0.780509	lr=0.008333
Epoch[346] Batch [599]	Speed: 8803.944574 samples/sec	accuracy=0.780760	lr=0.008230
[Epoch 346] training: accuracy=0.780859
[Epoch 346] speed: 8643 samples/sec	time cost: 152.641756
[Epoch 346] validation: err-top1=0.246908 err-top5=0.077820
Epoch[347] Batch [49]	Speed: 8519.954244 samples/sec	accuracy=0.782168	lr=0.008075
Epoch[347] Batch [99]	Speed: 8661.290984 samples/sec	accuracy=0.783037	lr=0.007973
Epoch[347] Batch [149]	Speed: 8688.817506 samples/sec	accuracy=0.782435	lr=0.007872
Epoch[347] Batch [199]	Speed: 8660.412684 samples/sec	accuracy=0.782017	lr=0.007771
Epoch[347] Batch [249]	Speed: 8777.342080 samples/sec	accuracy=0.781873	lr=0.007671
Epoch[347] Batch [299]	Speed: 8598.936130 samples/sec	accuracy=0.782249	lr=0.007571
Epoch[347] Batch [349]	Speed: 8827.096066 samples/sec	accuracy=0.782189	lr=0.007472
Epoch[347] Batch [399]	Speed: 8580.947594 samples/sec	accuracy=0.782290	lr=0.007374
Epoch[347] Batch [449]	Speed: 8737.513111 samples/sec	accuracy=0.782263	lr=0.007277
Epoch[347] Batch [499]	Speed: 8973.004538 samples/sec	accuracy=0.782401	lr=0.007180
Epoch[347] Batch [549]	Speed: 8588.520234 samples/sec	accuracy=0.782568	lr=0.007083
Epoch[347] Batch [599]	Speed: 8521.301304 samples/sec	accuracy=0.782476	lr=0.006988
[Epoch 347] training: accuracy=0.782440
[Epoch 347] speed: 8680 samples/sec	time cost: 152.329693
[Epoch 347] validation: err-top1=0.247812 err-top5=0.077129
Epoch[348] Batch [49]	Speed: 8723.612792 samples/sec	accuracy=0.784531	lr=0.006844
Epoch[348] Batch [99]	Speed: 8620.636645 samples/sec	accuracy=0.783765	lr=0.006750
Epoch[348] Batch [149]	Speed: 8631.492813 samples/sec	accuracy=0.782952	lr=0.006656
Epoch[348] Batch [199]	Speed: 8902.876211 samples/sec	accuracy=0.782961	lr=0.006564
Epoch[348] Batch [249]	Speed: 8529.705319 samples/sec	accuracy=0.783061	lr=0.006472
Epoch[348] Batch [299]	Speed: 8553.190258 samples/sec	accuracy=0.782484	lr=0.006380
Epoch[348] Batch [349]	Speed: 8581.492634 samples/sec	accuracy=0.782659	lr=0.006290
Epoch[348] Batch [399]	Speed: 8775.287800 samples/sec	accuracy=0.782371	lr=0.006199
Epoch[348] Batch [449]	Speed: 8904.440868 samples/sec	accuracy=0.782478	lr=0.006110
Epoch[348] Batch [499]	Speed: 8408.686531 samples/sec	accuracy=0.782166	lr=0.006021
Epoch[348] Batch [549]	Speed: 8614.907140 samples/sec	accuracy=0.782369	lr=0.005933
Epoch[348] Batch [599]	Speed: 8546.512837 samples/sec	accuracy=0.782326	lr=0.005846
[Epoch 348] training: accuracy=0.782334
[Epoch 348] speed: 8652 samples/sec	time cost: 152.482419
[Epoch 348] validation: err-top1=0.246908 err-top5=0.077474
Epoch[349] Batch [49]	Speed: 8595.956403 samples/sec	accuracy=0.783672	lr=0.005715
Epoch[349] Batch [99]	Speed: 8667.086745 samples/sec	accuracy=0.784687	lr=0.005630
Epoch[349] Batch [149]	Speed: 8774.572660 samples/sec	accuracy=0.784163	lr=0.005544
Epoch[349] Batch [199]	Speed: 8556.569447 samples/sec	accuracy=0.783655	lr=0.005460
Epoch[349] Batch [249]	Speed: 8726.749251 samples/sec	accuracy=0.783287	lr=0.005376
Epoch[349] Batch [299]	Speed: 8525.171253 samples/sec	accuracy=0.782990	lr=0.005293
Epoch[349] Batch [349]	Speed: 8614.041158 samples/sec	accuracy=0.782919	lr=0.005210
Epoch[349] Batch [399]	Speed: 8394.873166 samples/sec	accuracy=0.783198	lr=0.005128
Epoch[349] Batch [449]	Speed: 8601.831096 samples/sec	accuracy=0.783199	lr=0.005047
Epoch[349] Batch [499]	Speed: 8617.868897 samples/sec	accuracy=0.783295	lr=0.004966
Epoch[349] Batch [549]	Speed: 8751.814275 samples/sec	accuracy=0.783204	lr=0.004886
Epoch[349] Batch [599]	Speed: 8583.684640 samples/sec	accuracy=0.783185	lr=0.004806
[Epoch 349] training: accuracy=0.783087
[Epoch 349] speed: 8623 samples/sec	time cost: 153.300639
[Epoch 349] validation: err-top1=0.247498 err-top5=0.077271
Epoch[350] Batch [49]	Speed: 8784.575845 samples/sec	accuracy=0.786758	lr=0.004687
Epoch[350] Batch [99]	Speed: 8748.534856 samples/sec	accuracy=0.784536	lr=0.004609
Epoch[350] Batch [149]	Speed: 8740.673094 samples/sec	accuracy=0.784466	lr=0.004532
Epoch[350] Batch [199]	Speed: 8784.801160 samples/sec	accuracy=0.784541	lr=0.004456
Epoch[350] Batch [249]	Speed: 8780.111455 samples/sec	accuracy=0.784488	lr=0.004380
Epoch[350] Batch [299]	Speed: 8553.862611 samples/sec	accuracy=0.783711	lr=0.004305
Epoch[350] Batch [349]	Speed: 8622.167531 samples/sec	accuracy=0.783763	lr=0.004230
Epoch[350] Batch [399]	Speed: 8587.618511 samples/sec	accuracy=0.783789	lr=0.004156
Epoch[350] Batch [449]	Speed: 8610.111700 samples/sec	accuracy=0.783582	lr=0.004083
Epoch[350] Batch [499]	Speed: 8701.559062 samples/sec	accuracy=0.783555	lr=0.004011
Epoch[350] Batch [549]	Speed: 8618.823336 samples/sec	accuracy=0.783518	lr=0.003939
Epoch[350] Batch [599]	Speed: 8885.893125 samples/sec	accuracy=0.783579	lr=0.003867
[Epoch 350] training: accuracy=0.783549
[Epoch 350] speed: 8707 samples/sec	time cost: 152.044863
[Epoch 350] validation: err-top1=0.246172 err-top5=0.077187
Epoch[351] Batch [49]	Speed: 8546.051302 samples/sec	accuracy=0.784902	lr=0.003760
Epoch[351] Batch [99]	Speed: 8554.587890 samples/sec	accuracy=0.785444	lr=0.003691
Epoch[351] Batch [149]	Speed: 8588.971940 samples/sec	accuracy=0.784863	lr=0.003622
Epoch[351] Batch [199]	Speed: 8618.539698 samples/sec	accuracy=0.785042	lr=0.003553
Epoch[351] Batch [249]	Speed: 8661.090998 samples/sec	accuracy=0.785268	lr=0.003486
Epoch[351] Batch [299]	Speed: 8519.063988 samples/sec	accuracy=0.785228	lr=0.003419
Epoch[351] Batch [349]	Speed: 8507.620471 samples/sec	accuracy=0.784946	lr=0.003352
Epoch[351] Batch [399]	Speed: 8767.664444 samples/sec	accuracy=0.784797	lr=0.003287
Epoch[351] Batch [449]	Speed: 8445.874161 samples/sec	accuracy=0.785055	lr=0.003221
Epoch[351] Batch [499]	Speed: 9003.908367 samples/sec	accuracy=0.785039	lr=0.003157
Epoch[351] Batch [549]	Speed: 8552.472368 samples/sec	accuracy=0.784873	lr=0.003093
Epoch[351] Batch [599]	Speed: 8350.284597 samples/sec	accuracy=0.784718	lr=0.003030
[Epoch 351] training: accuracy=0.784649
[Epoch 351] speed: 8592 samples/sec	time cost: 153.506294
[Epoch 351] validation: err-top1=0.246338 err-top5=0.076782
Epoch[352] Batch [49]	Speed: 8506.804230 samples/sec	accuracy=0.787041	lr=0.002937
Epoch[352] Batch [99]	Speed: 8585.126749 samples/sec	accuracy=0.785884	lr=0.002875
Epoch[352] Batch [149]	Speed: 8517.781994 samples/sec	accuracy=0.784948	lr=0.002814
Epoch[352] Batch [199]	Speed: 8611.893888 samples/sec	accuracy=0.784973	lr=0.002754
Epoch[352] Batch [249]	Speed: 8901.887716 samples/sec	accuracy=0.784492	lr=0.002695
Epoch[352] Batch [299]	Speed: 8727.941679 samples/sec	accuracy=0.784512	lr=0.002636
Epoch[352] Batch [349]	Speed: 8748.445934 samples/sec	accuracy=0.784410	lr=0.002577
Epoch[352] Batch [399]	Speed: 9014.461460 samples/sec	accuracy=0.784539	lr=0.002520
Epoch[352] Batch [449]	Speed: 8576.372431 samples/sec	accuracy=0.784506	lr=0.002463
Epoch[352] Batch [499]	Speed: 8685.118102 samples/sec	accuracy=0.784497	lr=0.002407
Epoch[352] Batch [549]	Speed: 8580.722500 samples/sec	accuracy=0.784553	lr=0.002351
Epoch[352] Batch [599]	Speed: 8546.977823 samples/sec	accuracy=0.784411	lr=0.002296
[Epoch 352] training: accuracy=0.784443
[Epoch 352] speed: 8666 samples/sec	time cost: 152.658316
[Epoch 352] validation: err-top1=0.246445 err-top5=0.077051
Epoch[353] Batch [49]	Speed: 8432.182952 samples/sec	accuracy=0.785957	lr=0.002214
Epoch[353] Batch [99]	Speed: 8686.430059 samples/sec	accuracy=0.785278	lr=0.002160
Epoch[353] Batch [149]	Speed: 8583.082374 samples/sec	accuracy=0.785762	lr=0.002107
Epoch[353] Batch [199]	Speed: 8802.653351 samples/sec	accuracy=0.785317	lr=0.002055
Epoch[353] Batch [249]	Speed: 8706.294744 samples/sec	accuracy=0.785480	lr=0.002004
Epoch[353] Batch [299]	Speed: 8601.607834 samples/sec	accuracy=0.785374	lr=0.001953
Epoch[353] Batch [349]	Speed: 8390.787802 samples/sec	accuracy=0.785240	lr=0.001903
Epoch[353] Batch [399]	Speed: 8601.747372 samples/sec	accuracy=0.784985	lr=0.001854
Epoch[353] Batch [449]	Speed: 8703.671384 samples/sec	accuracy=0.785204	lr=0.001805
Epoch[353] Batch [499]	Speed: 8687.061851 samples/sec	accuracy=0.785287	lr=0.001757
Epoch[353] Batch [549]	Speed: 8680.826316 samples/sec	accuracy=0.785362	lr=0.001709
Epoch[353] Batch [599]	Speed: 8661.256401 samples/sec	accuracy=0.785425	lr=0.001662
[Epoch 353] training: accuracy=0.785371
[Epoch 353] speed: 8636 samples/sec	time cost: 152.623002
[Epoch 353] validation: err-top1=0.245890 err-top5=0.077128
Epoch[354] Batch [49]	Speed: 8397.956796 samples/sec	accuracy=0.784238	lr=0.001593
Epoch[354] Batch [99]	Speed: 8717.825794 samples/sec	accuracy=0.784785	lr=0.001548
Epoch[354] Batch [149]	Speed: 8636.489763 samples/sec	accuracy=0.784430	lr=0.001504
Epoch[354] Batch [199]	Speed: 8531.500468 samples/sec	accuracy=0.784951	lr=0.001460
Epoch[354] Batch [249]	Speed: 8700.673458 samples/sec	accuracy=0.784908	lr=0.001416
Epoch[354] Batch [299]	Speed: 8452.542403 samples/sec	accuracy=0.784928	lr=0.001374
Epoch[354] Batch [349]	Speed: 8668.018530 samples/sec	accuracy=0.785107	lr=0.001332
Epoch[354] Batch [399]	Speed: 8544.334844 samples/sec	accuracy=0.785326	lr=0.001290
Epoch[354] Batch [449]	Speed: 8736.258536 samples/sec	accuracy=0.785344	lr=0.001250
Epoch[354] Batch [499]	Speed: 8930.090490 samples/sec	accuracy=0.785430	lr=0.001210
Epoch[354] Batch [549]	Speed: 8520.473815 samples/sec	accuracy=0.785450	lr=0.001170
Epoch[354] Batch [599]	Speed: 8674.812952 samples/sec	accuracy=0.785620	lr=0.001132
[Epoch 354] training: accuracy=0.785574
[Epoch 354] speed: 8636 samples/sec	time cost: 152.809645
[Epoch 354] validation: err-top1=0.246318 err-top5=0.076884
Epoch[355] Batch [49]	Speed: 8467.066618 samples/sec	accuracy=0.786270	lr=0.001074
Epoch[355] Batch [99]	Speed: 8578.757158 samples/sec	accuracy=0.786138	lr=0.001037
Epoch[355] Batch [149]	Speed: 9131.688640 samples/sec	accuracy=0.784860	lr=0.001001
Epoch[355] Batch [199]	Speed: 8430.272481 samples/sec	accuracy=0.785015	lr=0.000965
Epoch[355] Batch [249]	Speed: 8842.806493 samples/sec	accuracy=0.785457	lr=0.000930
Epoch[355] Batch [299]	Speed: 8701.792656 samples/sec	accuracy=0.785480	lr=0.000895
Epoch[355] Batch [349]	Speed: 8551.534440 samples/sec	accuracy=0.785272	lr=0.000861
Epoch[355] Batch [399]	Speed: 8655.519715 samples/sec	accuracy=0.785381	lr=0.000828
Epoch[355] Batch [449]	Speed: 8779.301671 samples/sec	accuracy=0.785477	lr=0.000796
Epoch[355] Batch [499]	Speed: 8638.738975 samples/sec	accuracy=0.785477	lr=0.000764
Epoch[355] Batch [549]	Speed: 8850.126542 samples/sec	accuracy=0.785328	lr=0.000733
Epoch[355] Batch [599]	Speed: 8859.367629 samples/sec	accuracy=0.785404	lr=0.000702
[Epoch 355] training: accuracy=0.785359
[Epoch 355] speed: 8677 samples/sec	time cost: 152.335988
[Epoch 355] validation: err-top1=0.246641 err-top5=0.076777
Epoch[356] Batch [49]	Speed: 8489.646903 samples/sec	accuracy=0.787256	lr=0.000657
Epoch[356] Batch [99]	Speed: 8838.551375 samples/sec	accuracy=0.786309	lr=0.000628
Epoch[356] Batch [149]	Speed: 8797.645237 samples/sec	accuracy=0.785794	lr=0.000600
Epoch[356] Batch [199]	Speed: 8436.976096 samples/sec	accuracy=0.786013	lr=0.000572
Epoch[356] Batch [249]	Speed: 8614.778925 samples/sec	accuracy=0.786078	lr=0.000545
Epoch[356] Batch [299]	Speed: 8825.280283 samples/sec	accuracy=0.785996	lr=0.000519
Epoch[356] Batch [349]	Speed: 8541.488975 samples/sec	accuracy=0.786229	lr=0.000493
Epoch[356] Batch [399]	Speed: 8613.815016 samples/sec	accuracy=0.786243	lr=0.000468
Epoch[356] Batch [449]	Speed: 8674.561007 samples/sec	accuracy=0.786457	lr=0.000444
Epoch[356] Batch [499]	Speed: 8611.518847 samples/sec	accuracy=0.786519	lr=0.000420
Epoch[356] Batch [549]	Speed: 8906.379318 samples/sec	accuracy=0.786365	lr=0.000397
Epoch[356] Batch [599]	Speed: 8591.242695 samples/sec	accuracy=0.786257	lr=0.000375
[Epoch 356] training: accuracy=0.786200
[Epoch 356] speed: 8666 samples/sec	time cost: 152.580998
[Epoch 356] validation: err-top1=0.246257 err-top5=0.077209
Epoch[357] Batch [49]	Speed: 8420.012674 samples/sec	accuracy=0.785000	lr=0.000342
Epoch[357] Batch [99]	Speed: 8745.540140 samples/sec	accuracy=0.784756	lr=0.000321
Epoch[357] Batch [149]	Speed: 8693.884060 samples/sec	accuracy=0.784967	lr=0.000301
Epoch[357] Batch [199]	Speed: 8754.237090 samples/sec	accuracy=0.784832	lr=0.000282
Epoch[357] Batch [249]	Speed: 8454.219513 samples/sec	accuracy=0.785021	lr=0.000263
Epoch[357] Batch [299]	Speed: 8869.459877 samples/sec	accuracy=0.784857	lr=0.000245
Epoch[357] Batch [349]	Speed: 8607.418669 samples/sec	accuracy=0.785027	lr=0.000227
Epoch[357] Batch [399]	Speed: 8796.915276 samples/sec	accuracy=0.785131	lr=0.000210
Epoch[357] Batch [449]	Speed: 8928.223368 samples/sec	accuracy=0.785263	lr=0.000194
Epoch[357] Batch [499]	Speed: 8707.162428 samples/sec	accuracy=0.785516	lr=0.000178
Epoch[357] Batch [549]	Speed: 8766.317279 samples/sec	accuracy=0.785467	lr=0.000163
Epoch[357] Batch [599]	Speed: 8691.327092 samples/sec	accuracy=0.785427	lr=0.000149
[Epoch 357] training: accuracy=0.785606
[Epoch 357] speed: 8706 samples/sec	time cost: 151.722018
[Epoch 357] validation: err-top1=0.246953 err-top5=0.076719
Epoch[358] Batch [49]	Speed: 8433.061434 samples/sec	accuracy=0.788730	lr=0.000129
Epoch[358] Batch [99]	Speed: 8589.044767 samples/sec	accuracy=0.787598	lr=0.000116
Epoch[358] Batch [149]	Speed: 8839.705783 samples/sec	accuracy=0.787090	lr=0.000104
Epoch[358] Batch [199]	Speed: 8639.002745 samples/sec	accuracy=0.787002	lr=0.000093
Epoch[358] Batch [249]	Speed: 8800.131362 samples/sec	accuracy=0.786789	lr=0.000082
Epoch[358] Batch [299]	Speed: 8728.477349 samples/sec	accuracy=0.786336	lr=0.000072
Epoch[358] Batch [349]	Speed: 8648.889052 samples/sec	accuracy=0.786159	lr=0.000063
Epoch[358] Batch [399]	Speed: 8569.353234 samples/sec	accuracy=0.786044	lr=0.000054
Epoch[358] Batch [449]	Speed: 8687.202418 samples/sec	accuracy=0.786163	lr=0.000046
Epoch[358] Batch [499]	Speed: 8757.134384 samples/sec	accuracy=0.785978	lr=0.000039
Epoch[358] Batch [549]	Speed: 8756.460403 samples/sec	accuracy=0.785914	lr=0.000032
Epoch[358] Batch [599]	Speed: 8856.875688 samples/sec	accuracy=0.785868	lr=0.000026
[Epoch 358] training: accuracy=0.785864
[Epoch 358] speed: 8696 samples/sec	time cost: 151.900014
[Epoch 358] validation: err-top1=0.246216 err-top5=0.076843
Epoch[359] Batch [49]	Speed: 8409.183729 samples/sec	accuracy=0.788896	lr=0.000018
Epoch[359] Batch [99]	Speed: 8698.270684 samples/sec	accuracy=0.787764	lr=0.000013
Epoch[359] Batch [149]	Speed: 8851.730177 samples/sec	accuracy=0.787878	lr=0.000009
Epoch[359] Batch [199]	Speed: 8721.962603 samples/sec	accuracy=0.787424	lr=0.000006
Epoch[359] Batch [249]	Speed: 8743.550200 samples/sec	accuracy=0.787207	lr=0.000004
Epoch[359] Batch [299]	Speed: 8599.619311 samples/sec	accuracy=0.786898	lr=0.000002
Epoch[359] Batch [349]	Speed: 8415.039607 samples/sec	accuracy=0.786925	lr=0.000001
Epoch[359] Batch [399]	Speed: 8464.746911 samples/sec	accuracy=0.786896	lr=0.000000
Epoch[359] Batch [449]	Speed: 8706.293155 samples/sec	accuracy=0.786973	lr=0.000000
Epoch[359] Batch [499]	Speed: 8593.486968 samples/sec	accuracy=0.786822	lr=0.000000
Epoch[359] Batch [549]	Speed: 8821.380083 samples/sec	accuracy=0.786580	lr=0.000000
Epoch[359] Batch [599]	Speed: 8474.946685 samples/sec	accuracy=0.786501	lr=0.000000
[Epoch 359] training: accuracy=0.786476
[Epoch 359] speed: 8623 samples/sec	time cost: 153.288852
[Epoch 359] validation: err-top1=0.246699 err-top5=0.077578