Training:
Epoch 0/29
----------
train Loss: 1.9273
Real Acc: 25.1708%
val Loss: 1.5132
Real Acc: 42.2750%

Epoch 1/29
----------
train Loss: 1.2974
Real Acc: 50.5125%
val Loss: 1.2198
Real Acc: 53.9750%

Epoch 2/29
----------
train Loss: 0.9807
Real Acc: 62.8292%
val Loss: 1.0250
Real Acc: 61.4375%

Epoch 3/29
----------
train Loss: 0.7659
Real Acc: 70.9625%
val Loss: 0.9823
Real Acc: 64.7750%

Epoch 4/29
----------
train Loss: 0.5967
Real Acc: 77.8833%
val Loss: 0.7307
Real Acc: 73.4625%

Epoch 5/29
----------
train Loss: 0.4705
Real Acc: 82.9167%
val Loss: 0.5932
Real Acc: 78.9125%

Epoch 6/29
----------
train Loss: 0.3634
Real Acc: 87.2875%
val Loss: 0.7798
Real Acc: 74.1875%

Epoch 7/29
----------
train Loss: 0.2835
Real Acc: 89.8625%
val Loss: 0.7363
Real Acc: 76.4000%

Epoch 8/29
----------
train Loss: 0.2345
Real Acc: 91.7417%
val Loss: 0.6380
Real Acc: 79.7375%

Epoch 9/29
----------
train Loss: 0.1861
Real Acc: 93.6458%
val Loss: 0.6116
Real Acc: 80.7750%

Epoch 10/29
----------
train Loss: 0.0657
Real Acc: 98.0000%
val Loss: 0.5543
Real Acc: 83.5000%

Epoch 11/29
----------
train Loss: 0.0402
Real Acc: 98.9250%
val Loss: 0.6225
Real Acc: 83.2500%

Epoch 12/29
----------
train Loss: 0.0466
Real Acc: 98.7000%
val Loss: 0.6727
Real Acc: 82.1250%

Epoch 13/29
----------
train Loss: 0.0462
Real Acc: 98.5750%
val Loss: 0.7329
Real Acc: 81.7750%

Epoch 14/29
----------
train Loss: 0.0469
Real Acc: 98.5083%
val Loss: 0.6769
Real Acc: 82.4000%

Epoch 15/29
----------
train Loss: 0.0337
Real Acc: 98.9042%
val Loss: 0.7130
Real Acc: 81.8250%

Epoch 16/29
----------
train Loss: 0.0418
Real Acc: 98.7000%
val Loss: 0.9583
Real Acc: 77.6750%

Epoch 17/29
----------
train Loss: 0.0333
Real Acc: 98.9833%
val Loss: 0.6982
Real Acc: 83.0500%

Epoch 18/29
----------
train Loss: 0.0423
Real Acc: 98.5792%
val Loss: 0.7301
Real Acc: 82.8375%

Epoch 19/29
----------
train Loss: 0.0347
Real Acc: 98.8250%
val Loss: 0.7470
Real Acc: 82.8000%

Epoch 20/29
----------
train Loss: 0.0114
Real Acc: 99.7458%
val Loss: 0.6386
Real Acc: 84.9375%

Epoch 21/29
----------
train Loss: 0.0050
Real Acc: 99.9167%
val Loss: 0.6368
Real Acc: 85.5750%

Epoch 22/29
----------
train Loss: 0.0045
Real Acc: 99.9333%
val Loss: 0.6238
Real Acc: 85.2375%

Epoch 23/29
----------
train Loss: 0.0045
Real Acc: 99.9167%
val Loss: 0.6749
Real Acc: 85.1375%

Epoch 24/29
----------
train Loss: 0.0149
Real Acc: 99.5958%
val Loss: 0.6842
Real Acc: 84.0750%

Epoch 25/29
----------
train Loss: 0.0071
Real Acc: 99.8083%
val Loss: 0.7630
Real Acc: 83.8875%

Epoch 26/29
----------
train Loss: 0.0074
Real Acc: 99.8042%
val Loss: 0.7658
Real Acc: 83.5625%

Epoch 27/29
----------
train Loss: 0.0107
Real Acc: 99.7083%
val Loss: 0.7606
Real Acc: 83.4875%

Epoch 28/29
----------
train Loss: 0.0094
Real Acc: 99.7333%
val Loss: 0.7134
Real Acc: 84.3000%

Epoch 29/29
----------
train Loss: 0.0088
Real Acc: 99.7542%
val Loss: 0.7597
Real Acc: 84.1250%

Training complete in 710m 54s
Best val Acc: 85.575000
[42.275, 53.97500000000001, 61.4375, 64.775, 73.46249999999999, 78.9125, 74.1875, 76.4, 79.73750000000001, 80.77499999999999, 83.5, 83.25, 82.125, 81.77499999999999, 82.4, 81.825, 77.67500000000001, 83.05, 82.83749999999999, 82.80000000000001, 84.9375, 85.575, 85.2375, 85.1375, 84.075, 83.8875, 83.5625, 83.4875, 84.3, 84.125]

test C: 75.85%
test 2x2: 46.05%
test 3x3: 43.6%
test LR: 89.65%
test UD: 88.85%
test inC: 84.6%
test in2x2: 65.85%