Using device: cuda:0
Epoch 1/100
----------
Learning rate: 0.0001
train Loss: 6.9024 Top-1: 0.0017 Top-5: 0.0078
val Loss: 6.8617 Top-1: 0.0023 Top-5: 0.0110
Epoch time: 3325s

Epoch 2/100
----------
Learning rate: 0.02008
train Loss: 6.0329 Top-1: 0.0423 Top-5: 0.1235
val Loss: 5.0223 Top-1: 0.1274 Top-5: 0.3129
Epoch time: 3313s

Epoch 3/100
----------
Learning rate: 0.04006
train Loss: 4.9776 Top-1: 0.1502 Top-5: 0.3346
val Loss: 4.5452 Top-1: 0.2136 Top-5: 0.4376
Epoch time: 3314s

Epoch 4/100
----------
Learning rate: 0.060039999999999996
train Loss: 4.3521 Top-1: 0.2469 Top-5: 0.4753
val Loss: 3.6762 Top-1: 0.3574 Top-5: 0.6238
Epoch time: 3312s

Epoch 5/100
----------
Learning rate: 0.08002
train Loss: 4.0247 Top-1: 0.3054 Top-5: 0.5470
val Loss: 3.4750 Top-1: 0.4055 Top-5: 0.6726
Epoch time: 3306s

Epoch 6/100
----------
Learning rate: 0.1
train Loss: 3.8462 Top-1: 0.3390 Top-5: 0.5854
val Loss: 3.3772 Top-1: 0.4169 Top-5: 0.6864
Epoch time: 3304s

Epoch 7/100
----------
Learning rate: 0.0999726656007596
train Loss: 3.6901 Top-1: 0.3694 Top-5: 0.6179
val Loss: 3.4853 Top-1: 0.3980 Top-5: 0.6652
Epoch time: 3301s

Epoch 8/100
----------
Learning rate: 0.09989069229280263
train Loss: 3.5934 Top-1: 0.3889 Top-5: 0.6378
val Loss: 3.1712 Top-1: 0.4630 Top-5: 0.7291
Epoch time: 3302s

Epoch 9/100
----------
Learning rate: 0.09975416971273784
train Loss: 3.5243 Top-1: 0.4032 Top-5: 0.6519
val Loss: 3.0799 Top-1: 0.4861 Top-5: 0.7485
Epoch time: 3302s

Epoch 10/100
----------
Learning rate: 0.09956324714600212
train Loss: 3.4687 Top-1: 0.4142 Top-5: 0.6633
val Loss: 3.0626 Top-1: 0.4882 Top-5: 0.7518
Epoch time: 3301s

Epoch 11/100
----------
Learning rate: 0.0993181333636191
train Loss: 3.4281 Top-1: 0.4233 Top-5: 0.6709
val Loss: 2.9913 Top-1: 0.4995 Top-5: 0.7645
Epoch time: 3300s

Epoch 12/100
----------
Learning rate: 0.09901909639391109
train Loss: 3.3952 Top-1: 0.4296 Top-5: 0.6773
val Loss: 3.0287 Top-1: 0.4940 Top-5: 0.7558
Epoch time: 3300s

Epoch 13/100
----------
Learning rate: 0.09866646322941404
train Loss: 3.3643 Top-1: 0.4359 Top-5: 0.6836
val Loss: 2.9753 Top-1: 0.5044 Top-5: 0.7660
Epoch time: 3301s

Epoch 14/100
----------
Learning rate: 0.09826061946931602
train Loss: 3.3425 Top-1: 0.4409 Top-5: 0.6877
val Loss: 2.8872 Top-1: 0.5274 Top-5: 0.7821
Epoch time: 3300s

Epoch 15/100
----------
Learning rate: 0.0978020088978102
train Loss: 3.3255 Top-1: 0.4450 Top-5: 0.6911
val Loss: 2.9545 Top-1: 0.5109 Top-5: 0.7709
Epoch time: 3300s

Epoch 16/100
----------
Learning rate: 0.09729113299882322
train Loss: 3.3056 Top-1: 0.4489 Top-5: 0.6948
val Loss: 2.9247 Top-1: 0.5171 Top-5: 0.7761
Epoch time: 3300s

Epoch 17/100
----------
Learning rate: 0.09672855040765005
train Loss: 3.2873 Top-1: 0.4526 Top-5: 0.6981
val Loss: 2.9355 Top-1: 0.5149 Top-5: 0.7741
Epoch time: 3299s

Epoch 18/100
----------
Learning rate: 0.09611487630009435
train Loss: 3.2753 Top-1: 0.4552 Top-5: 0.7005
val Loss: 2.9580 Top-1: 0.5095 Top-5: 0.7687
Epoch time: 3300s

Epoch 19/100
----------
Learning rate: 0.09545078171978288
train Loss: 3.2627 Top-1: 0.4579 Top-5: 0.7026
val Loss: 2.8543 Top-1: 0.5335 Top-5: 0.7881
Epoch time: 3300s

Epoch 20/100
----------
Learning rate: 0.09473699284438906
train Loss: 3.2499 Top-1: 0.4605 Top-5: 0.7050
val Loss: 2.8353 Top-1: 0.5375 Top-5: 0.7939
Epoch time: 3300s

Epoch 21/100
----------
Learning rate: 0.09397429019156842
train Loss: 3.2402 Top-1: 0.4628 Top-5: 0.7069
val Loss: 2.9278 Top-1: 0.5162 Top-5: 0.7774
Epoch time: 3299s

Epoch 22/100
----------
Learning rate: 0.0931635077654739
train Loss: 3.2268 Top-1: 0.4656 Top-5: 0.7095
val Loss: 2.8520 Top-1: 0.5345 Top-5: 0.7864
Epoch time: 3300s

Epoch 23/100
----------
Learning rate: 0.09230553214478465
train Loss: 3.2178 Top-1: 0.4674 Top-5: 0.7114
val Loss: 2.8013 Top-1: 0.5453 Top-5: 0.7981
Epoch time: 3300s

Epoch 24/100
----------
Learning rate: 0.09140130151324526
train Loss: 3.2073 Top-1: 0.4697 Top-5: 0.7134
val Loss: 2.8957 Top-1: 0.5223 Top-5: 0.7811
Epoch time: 3299s

Epoch 25/100
----------
Learning rate: 0.09045180463377549
train Loss: 3.1963 Top-1: 0.4723 Top-5: 0.7155
val Loss: 2.8247 Top-1: 0.5385 Top-5: 0.7926
Epoch time: 3299s

Epoch 26/100
----------
Learning rate: 0.08945807976727269
train Loss: 3.1858 Top-1: 0.4743 Top-5: 0.7174
val Loss: 2.8841 Top-1: 0.5283 Top-5: 0.7842
Epoch time: 3299s

Epoch 27/100
----------
Learning rate: 0.08842121353728866
train Loss: 3.1785 Top-1: 0.4765 Top-5: 0.7188
val Loss: 2.7812 Top-1: 0.5499 Top-5: 0.8009
Epoch time: 3300s

Epoch 28/100
----------
Learning rate: 0.08734233974182276
train Loss: 3.1672 Top-1: 0.4789 Top-5: 0.7204
val Loss: 2.7865 Top-1: 0.5495 Top-5: 0.8024
Epoch time: 3299s

Epoch 29/100
----------
Learning rate: 0.08622263811353048
train Loss: 3.1589 Top-1: 0.4803 Top-5: 0.7222
val Loss: 2.7490 Top-1: 0.5595 Top-5: 0.8066
Epoch time: 3298s

Epoch 30/100
----------
Learning rate: 0.08506333302970304
train Loss: 3.1515 Top-1: 0.4823 Top-5: 0.7234
val Loss: 2.7857 Top-1: 0.5509 Top-5: 0.8021
Epoch time: 3300s

Epoch 31/100
----------
Learning rate: 0.0838656921734289
train Loss: 3.1417 Top-1: 0.4837 Top-5: 0.7250
val Loss: 2.7208 Top-1: 0.5646 Top-5: 0.8122
Epoch time: 3299s

Epoch 32/100
----------
Learning rate: 0.08263102514740078
train Loss: 3.1298 Top-1: 0.4866 Top-5: 0.7273
val Loss: 2.8828 Top-1: 0.5284 Top-5: 0.7823
Epoch time: 3299s

Epoch 33/100
----------
Learning rate: 0.08136068204188446
train Loss: 3.1208 Top-1: 0.4889 Top-5: 0.7292
val Loss: 2.8212 Top-1: 0.5422 Top-5: 0.7928
Epoch time: 3298s

Epoch 34/100
----------
Learning rate: 0.08005605195841482
train Loss: 3.1136 Top-1: 0.4903 Top-5: 0.7300
val Loss: 2.7230 Top-1: 0.5670 Top-5: 0.8122
Epoch time: 3299s

Epoch 35/100
----------
Learning rate: 0.07871856149083375
train Loss: 3.1041 Top-1: 0.4917 Top-5: 0.7321
val Loss: 2.8575 Top-1: 0.5315 Top-5: 0.7841
Epoch time: 3299s

Epoch 36/100
----------
Learning rate: 0.07734967316533073
train Loss: 3.0933 Top-1: 0.4949 Top-5: 0.7342
val Loss: 2.7002 Top-1: 0.5688 Top-5: 0.8157
Epoch time: 3299s

Epoch 37/100
----------
Learning rate: 0.07595088384119184
train Loss: 3.0830 Top-1: 0.4968 Top-5: 0.7358
val Loss: 2.6861 Top-1: 0.5722 Top-5: 0.8207
Epoch time: 3299s

Epoch 38/100
----------
Learning rate: 0.07452372307400623
train Loss: 3.0733 Top-1: 0.4993 Top-5: 0.7381
val Loss: 2.6832 Top-1: 0.5733 Top-5: 0.8195
Epoch time: 3299s

Epoch 39/100
----------
Learning rate: 0.07306975144311949
train Loss: 3.0636 Top-1: 0.5009 Top-5: 0.7395
val Loss: 2.6857 Top-1: 0.5723 Top-5: 0.8196
Epoch time: 3299s

Epoch 40/100
----------
Learning rate: 0.07159055884516294
train Loss: 3.0548 Top-1: 0.5032 Top-5: 0.7410
val Loss: 2.7333 Top-1: 0.5640 Top-5: 0.8077
Epoch time: 3299s

Epoch 41/100
----------
Learning rate: 0.0700877627555252
train Loss: 3.0405 Top-1: 0.5063 Top-5: 0.7434
val Loss: 2.6546 Top-1: 0.5758 Top-5: 0.8255
Epoch time: 3298s

Epoch 42/100
----------
Learning rate: 0.06856300645966643
train Loss: 3.0323 Top-1: 0.5080 Top-5: 0.7451
val Loss: 2.6197 Top-1: 0.5911 Top-5: 0.8292
Epoch time: 3299s

Epoch 43/100
----------
Learning rate: 0.06701795725620992
train Loss: 3.0235 Top-1: 0.5095 Top-5: 0.7468
val Loss: 2.6413 Top-1: 0.5799 Top-5: 0.8274
Epoch time: 3297s

Epoch 44/100
----------
Learning rate: 0.06545430463377548
train Loss: 3.0118 Top-1: 0.5130 Top-5: 0.7487
val Loss: 2.5964 Top-1: 0.5958 Top-5: 0.8354
Epoch time: 3298s

Epoch 45/100
----------
Learning rate: 0.06387375842354842
train Loss: 3.0017 Top-1: 0.5145 Top-5: 0.7510
val Loss: 2.5904 Top-1: 0.5980 Top-5: 0.8351
Epoch time: 3298s

Epoch 46/100
----------
Learning rate: 0.06227804692960425
train Loss: 2.9871 Top-1: 0.5179 Top-5: 0.7530
val Loss: 2.6205 Top-1: 0.5896 Top-5: 0.8288
Epoch time: 3298s

Epoch 47/100
----------
Learning rate: 0.060668915039033634
train Loss: 2.9779 Top-1: 0.5200 Top-5: 0.7548
val Loss: 2.6571 Top-1: 0.5804 Top-5: 0.8227
Epoch time: 3298s

Epoch 48/100
----------
Learning rate: 0.059048122313933825
train Loss: 2.9640 Top-1: 0.5233 Top-5: 0.7572
val Loss: 2.5503 Top-1: 0.6063 Top-5: 0.8420
Epoch time: 3298s

Epoch 49/100
----------
Learning rate: 0.05741744106735353
train Loss: 2.9527 Top-1: 0.5254 Top-5: 0.7592
val Loss: 2.5702 Top-1: 0.6025 Top-5: 0.8362
Epoch time: 3297s

Epoch 50/100
----------
Learning rate: 0.05577865442529446
train Loss: 2.9406 Top-1: 0.5277 Top-5: 0.7612
val Loss: 2.5281 Top-1: 0.6112 Top-5: 0.8446
Epoch time: 3297s

Epoch 51/100
----------
Learning rate: 0.054133554376889256
train Loss: 2.9287 Top-1: 0.5308 Top-5: 0.7634
val Loss: 2.5609 Top-1: 0.6015 Top-5: 0.8377
Epoch time: 3298s

Epoch 52/100
----------
Learning rate: 0.052483939814887755
train Loss: 2.9162 Top-1: 0.5338 Top-5: 0.7655
val Loss: 2.5342 Top-1: 0.6102 Top-5: 0.8444
Epoch time: 3297s

Epoch 53/100
----------
Learning rate: 0.050831614568594086
train Loss: 2.9019 Top-1: 0.5371 Top-5: 0.7680
val Loss: 2.5266 Top-1: 0.6136 Top-5: 0.8439
Epoch time: 3297s

Epoch 54/100
----------
Learning rate: 0.04917838543140591
train Loss: 2.8892 Top-1: 0.5392 Top-5: 0.7703
val Loss: 2.4665 Top-1: 0.6256 Top-5: 0.8558
Epoch time: 3299s

Epoch 55/100
----------
Learning rate: 0.04752606018511226
train Loss: 2.8746 Top-1: 0.5428 Top-5: 0.7727
val Loss: 2.4895 Top-1: 0.6222 Top-5: 0.8511
Epoch time: 3298s

Epoch 56/100
----------
Learning rate: 0.04587644562311075
train Loss: 2.8644 Top-1: 0.5456 Top-5: 0.7747
val Loss: 2.4678 Top-1: 0.6239 Top-5: 0.8531
Epoch time: 3297s

Epoch 57/100
----------
Learning rate: 0.04423134557470556
train Loss: 2.8501 Top-1: 0.5488 Top-5: 0.7767
val Loss: 2.4704 Top-1: 0.6255 Top-5: 0.8527
Epoch time: 3297s

Epoch 58/100
----------
Learning rate: 0.04259255893264646
train Loss: 2.8310 Top-1: 0.5527 Top-5: 0.7804
val Loss: 2.4681 Top-1: 0.6273 Top-5: 0.8539
Epoch time: 3297s

Epoch 59/100
----------
Learning rate: 0.04096187768606618
train Loss: 2.8145 Top-1: 0.5560 Top-5: 0.7830
val Loss: 2.4100 Top-1: 0.6387 Top-5: 0.8640
Epoch time: 3298s

Epoch 60/100
----------
Learning rate: 0.03934108496096639
train Loss: 2.8029 Top-1: 0.5589 Top-5: 0.7850
val Loss: 2.4442 Top-1: 0.6300 Top-5: 0.8571
Epoch time: 3296s

Epoch 61/100
----------
Learning rate: 0.03773195307039576
train Loss: 2.7888 Top-1: 0.5627 Top-5: 0.7870
val Loss: 2.4177 Top-1: 0.6363 Top-5: 0.8603
Epoch time: 3297s

Epoch 62/100
----------
Learning rate: 0.036136241576451594
train Loss: 2.7737 Top-1: 0.5659 Top-5: 0.7896
val Loss: 2.3401 Top-1: 0.6554 Top-5: 0.8745
Epoch time: 3296s

Epoch 63/100
----------
Learning rate: 0.03455569536622452
train Loss: 2.7580 Top-1: 0.5698 Top-5: 0.7923
val Loss: 2.4094 Top-1: 0.6372 Top-5: 0.8626
Epoch time: 3296s

Epoch 64/100
----------
Learning rate: 0.032992042743790057
train Loss: 2.7422 Top-1: 0.5734 Top-5: 0.7949
val Loss: 2.3923 Top-1: 0.6434 Top-5: 0.8672
Epoch time: 3295s

Epoch 65/100
----------
Learning rate: 0.031446993540333555
train Loss: 2.7248 Top-1: 0.5774 Top-5: 0.7977
val Loss: 2.3931 Top-1: 0.6410 Top-5: 0.8641
Epoch time: 3295s

Epoch 66/100
----------
Learning rate: 0.0299222372444748
train Loss: 2.7057 Top-1: 0.5818 Top-5: 0.8012
val Loss: 2.3514 Top-1: 0.6522 Top-5: 0.8718
Epoch time: 3296s

Epoch 67/100
----------
Learning rate: 0.02841944115483705
train Loss: 2.6894 Top-1: 0.5847 Top-5: 0.8039
val Loss: 2.3139 Top-1: 0.6616 Top-5: 0.8770
Epoch time: 3295s

Epoch 68/100
----------
Learning rate: 0.02694024855688051
train Loss: 2.6720 Top-1: 0.5892 Top-5: 0.8067
val Loss: 2.3048 Top-1: 0.6633 Top-5: 0.8785
Epoch time: 3298s

Epoch 69/100
----------
Learning rate: 0.025486276925993746
train Loss: 2.6527 Top-1: 0.5931 Top-5: 0.8095
val Loss: 2.2934 Top-1: 0.6688 Top-5: 0.8794
Epoch time: 3296s

Epoch 70/100
----------
Learning rate: 0.024059116158808146
train Loss: 2.6371 Top-1: 0.5973 Top-5: 0.8123
val Loss: 2.2586 Top-1: 0.6762 Top-5: 0.8851
Epoch time: 3296s

Epoch 71/100
----------
Learning rate: 0.02266032683466928
train Loss: 2.6168 Top-1: 0.6019 Top-5: 0.8156
val Loss: 2.2268 Top-1: 0.6812 Top-5: 0.8902
Epoch time: 3297s

Epoch 72/100
----------
Learning rate: 0.021291438509166236
train Loss: 2.5969 Top-1: 0.6070 Top-5: 0.8186
val Loss: 2.2390 Top-1: 0.6806 Top-5: 0.8880
Epoch time: 3296s

Epoch 73/100
----------
Learning rate: 0.01995394804158516
train Loss: 2.5779 Top-1: 0.6108 Top-5: 0.8219
val Loss: 2.2048 Top-1: 0.6900 Top-5: 0.8926
Epoch time: 3297s

Epoch 74/100
----------
Learning rate: 0.018649317958115516
train Loss: 2.5577 Top-1: 0.6158 Top-5: 0.8250
val Loss: 2.2123 Top-1: 0.6857 Top-5: 0.8922
Epoch time: 3298s

Epoch 75/100
----------
Learning rate: 0.017378974852599188
train Loss: 2.5378 Top-1: 0.6208 Top-5: 0.8282
val Loss: 2.1793 Top-1: 0.6949 Top-5: 0.8975
Epoch time: 3297s

Epoch 76/100
----------
Learning rate: 0.016144307826571086
train Loss: 2.5153 Top-1: 0.6254 Top-5: 0.8316
val Loss: 2.1813 Top-1: 0.6926 Top-5: 0.8968
Epoch time: 3296s

Epoch 77/100
----------
Learning rate: 0.014946666970296934
train Loss: 2.4945 Top-1: 0.6306 Top-5: 0.8349
val Loss: 2.1552 Top-1: 0.7014 Top-5: 0.8994
Epoch time: 3297s

Epoch 78/100
----------
Learning rate: 0.013787361886469509
train Loss: 2.4743 Top-1: 0.6353 Top-5: 0.8383
val Loss: 2.1618 Top-1: 0.6986 Top-5: 0.8990
Epoch time: 3297s

Epoch 79/100
----------
Learning rate: 0.012667660258177242
train Loss: 2.4496 Top-1: 0.6411 Top-5: 0.8423
val Loss: 2.1570 Top-1: 0.7017 Top-5: 0.8984
Epoch time: 3298s

Epoch 80/100
----------
Learning rate: 0.011588786462711328
train Loss: 2.4282 Top-1: 0.6459 Top-5: 0.8451
val Loss: 2.1118 Top-1: 0.7107 Top-5: 0.9051
Epoch time: 3297s

Epoch 81/100
----------
Learning rate: 0.010551920232727309
train Loss: 2.4057 Top-1: 0.6510 Top-5: 0.8490
val Loss: 2.1160 Top-1: 0.7084 Top-5: 0.9042
Epoch time: 3297s

Epoch 82/100
----------
Learning rate: 0.009558195366224508
train Loss: 2.3789 Top-1: 0.6579 Top-5: 0.8529
val Loss: 2.0859 Top-1: 0.7182 Top-5: 0.9091
Epoch time: 3299s

Epoch 83/100
----------
Learning rate: 0.008608698486754737
train Loss: 2.3552 Top-1: 0.6631 Top-5: 0.8564
val Loss: 2.0485 Top-1: 0.7272 Top-5: 0.9131
Epoch time: 3299s

Epoch 84/100
----------
Learning rate: 0.007704467855215338
train Loss: 2.3321 Top-1: 0.6691 Top-5: 0.8599
val Loss: 2.0218 Top-1: 0.7319 Top-5: 0.9175
Epoch time: 3298s

Epoch 85/100
----------
Learning rate: 0.006846492234526091
train Loss: 2.3059 Top-1: 0.6754 Top-5: 0.8637
val Loss: 2.0064 Top-1: 0.7354 Top-5: 0.9197
Epoch time: 3299s

Epoch 86/100
----------
Learning rate: 0.006035709808431583
train Loss: 2.2815 Top-1: 0.6812 Top-5: 0.8674
val Loss: 2.0134 Top-1: 0.7363 Top-5: 0.9174
Epoch time: 3299s

Epoch 87/100
----------
Learning rate: 0.005273007155610929
train Loss: 2.2546 Top-1: 0.6878 Top-5: 0.8713
val Loss: 1.9825 Top-1: 0.7451 Top-5: 0.9219
Epoch time: 3299s

Epoch 88/100
----------
Learning rate: 0.004559218280217131
train Loss: 2.2328 Top-1: 0.6929 Top-5: 0.8746
val Loss: 1.9718 Top-1: 0.7485 Top-5: 0.9232
Epoch time: 3299s

Epoch 89/100
----------
Learning rate: 0.00389512369990565
train Loss: 2.2040 Top-1: 0.7004 Top-5: 0.8785
val Loss: 1.9565 Top-1: 0.7485 Top-5: 0.9254
Epoch time: 3297s

Epoch 90/100
----------
Learning rate: 0.0032814495923499496
train Loss: 2.1790 Top-1: 0.7059 Top-5: 0.8821
val Loss: 1.9418 Top-1: 0.7515 Top-5: 0.9279
Epoch time: 3297s

Epoch 91/100
----------
Learning rate: 0.0027188670011767716
train Loss: 2.1571 Top-1: 0.7118 Top-5: 0.8850
val Loss: 1.9178 Top-1: 0.7599 Top-5: 0.9311
Epoch time: 3300s

Epoch 92/100
----------
Learning rate: 0.00220799110218979
train Loss: 2.1341 Top-1: 0.7179 Top-5: 0.8881
val Loss: 1.9111 Top-1: 0.7616 Top-5: 0.9309
Epoch time: 3299s

Epoch 93/100
----------
Learning rate: 0.0017493805306839588
train Loss: 2.1110 Top-1: 0.7232 Top-5: 0.8918
val Loss: 1.8929 Top-1: 0.7654 Top-5: 0.9330
Epoch time: 3299s

Epoch 94/100
----------
Learning rate: 0.0013435367705859531
train Loss: 2.0887 Top-1: 0.7291 Top-5: 0.8948
val Loss: 1.8867 Top-1: 0.7673 Top-5: 0.9340
Epoch time: 3299s

Epoch 95/100
----------
Learning rate: 0.0009909036060889063
train Loss: 2.0730 Top-1: 0.7333 Top-5: 0.8967
val Loss: 1.8770 Top-1: 0.7693 Top-5: 0.9355
Epoch time: 3299s

Epoch 96/100
----------
Learning rate: 0.0006918666363808975
train Loss: 2.0574 Top-1: 0.7372 Top-5: 0.8990
val Loss: 1.8732 Top-1: 0.7713 Top-5: 0.9353
Epoch time: 3300s

Epoch 97/100
----------
Learning rate: 0.00044675285399787515
train Loss: 2.0459 Top-1: 0.7402 Top-5: 0.9005
val Loss: 1.8664 Top-1: 0.7726 Top-5: 0.9366
Epoch time: 3298s

Epoch 98/100
----------
Learning rate: 0.00025583028726215424
train Loss: 2.0334 Top-1: 0.7434 Top-5: 0.9020
val Loss: 1.8635 Top-1: 0.7732 Top-5: 0.9368
Epoch time: 3300s

Epoch 99/100
----------
Learning rate: 0.00011930770719736712
train Loss: 2.0290 Top-1: 0.7445 Top-5: 0.9027
val Loss: 1.8592 Top-1: 0.7746 Top-5: 0.9373
Epoch time: 3299s

Epoch 100/100
----------
Learning rate: 3.733439924040218e-05
train Loss: 2.0238 Top-1: 0.7462 Top-5: 0.9035
val Loss: 1.8581 Top-1: 0.7748 Top-5: 0.9375
Epoch time: 3294s

Training complete in 5498m 31s
Best val Top-1 Acc: 0.774800
