Training:
Epoch 0/29
----------
train Loss: 1.9287
Real Acc: 24.6267%
val Loss: 1.6386
Real Acc: 34.5400%

Epoch 1/29
----------
train Loss: 1.4799
Real Acc: 42.6233%
val Loss: 1.3241
Real Acc: 48.5100%

Epoch 2/29
----------
train Loss: 1.2372
Real Acc: 52.4600%
val Loss: 1.3469
Real Acc: 48.7200%

Epoch 3/29
----------
train Loss: 1.0676
Real Acc: 59.1667%
val Loss: 1.1670
Real Acc: 54.9000%

Epoch 4/29
----------
train Loss: 0.9408
Real Acc: 63.5467%
val Loss: 1.0797
Real Acc: 58.4300%

Epoch 5/29
----------
train Loss: 0.7925
Real Acc: 69.6367%
val Loss: 1.0977
Real Acc: 60.0000%

Epoch 6/29
----------
train Loss: 0.6345
Real Acc: 76.2467%
val Loss: 1.2527
Real Acc: 57.5600%

Epoch 7/29
----------
train Loss: 0.4954
Real Acc: 81.7333%
val Loss: 1.0395
Real Acc: 63.6400%

Epoch 8/29
----------
train Loss: 0.3902
Real Acc: 85.8633%
val Loss: 1.1722
Real Acc: 61.4700%

Epoch 9/29
----------
train Loss: 0.3005
Real Acc: 89.2300%
val Loss: 1.3053
Real Acc: 62.8000%

Epoch 10/29
----------
train Loss: 0.1094
Real Acc: 96.6533%
val Loss: 1.3075
Real Acc: 65.2600%

Epoch 11/29
----------
train Loss: 0.0616
Real Acc: 98.2400%
val Loss: 1.5537
Real Acc: 63.5300%

Epoch 12/29
----------
train Loss: 0.0604
Real Acc: 98.1267%
val Loss: 1.9208
Real Acc: 60.0700%

Epoch 13/29
----------
train Loss: 0.0594
Real Acc: 98.1600%
val Loss: 1.6385
Real Acc: 63.9500%

Epoch 14/29
----------
train Loss: 0.0481
Real Acc: 98.5067%
val Loss: 1.6947
Real Acc: 64.2600%

Epoch 15/29
----------
train Loss: 0.0470
Real Acc: 98.5733%
val Loss: 1.8204
Real Acc: 63.2900%

Epoch 16/29
----------
train Loss: 0.0466
Real Acc: 98.4367%
val Loss: 1.7697
Real Acc: 64.1300%

Epoch 17/29
----------
train Loss: 0.0414
Real Acc: 98.6867%
val Loss: 1.9378
Real Acc: 62.2400%

Epoch 18/29
----------
train Loss: 0.0515
Real Acc: 98.3333%
val Loss: 1.7849
Real Acc: 63.2500%

Epoch 19/29
----------
train Loss: 0.0403
Real Acc: 98.7300%
val Loss: 1.8947
Real Acc: 63.1500%

Epoch 20/29
----------
train Loss: 0.0139
Real Acc: 99.6233%
val Loss: 1.7406
Real Acc: 65.3900%

Epoch 21/29
----------
train Loss: 0.0090
Real Acc: 99.7933%
val Loss: 1.7468
Real Acc: 65.4300%

Epoch 22/29
----------
train Loss: 0.0076
Real Acc: 99.8300%
val Loss: 1.8313
Real Acc: 64.0100%

Epoch 23/29
----------
train Loss: 0.0103
Real Acc: 99.7000%
val Loss: 1.8608
Real Acc: 65.0600%

Epoch 24/29
----------
train Loss: 0.0120
Real Acc: 99.6567%
val Loss: 1.8617
Real Acc: 65.5100%

Epoch 25/29
----------
train Loss: 0.0113
Real Acc: 99.6867%
val Loss: 1.8835
Real Acc: 65.5200%

Epoch 26/29
----------
train Loss: 0.0090
Real Acc: 99.7600%
val Loss: 1.9464
Real Acc: 64.9400%

Epoch 27/29
----------
train Loss: 0.0129
Real Acc: 99.5967%
val Loss: 2.0267
Real Acc: 64.4200%

Epoch 28/29
----------
train Loss: 0.0093
Real Acc: 99.7233%
val Loss: 1.9199
Real Acc: 65.4600%

Epoch 29/29
----------
train Loss: 0.0082
Real Acc: 99.7900%
val Loss: 2.0192
Real Acc: 63.5700%

Training complete in 886m 21s
Best val Acc: 65.520000
[34.540000000000006, 48.510000000000005, 48.72, 54.900000000000006, 58.43000000000001, 60.0, 57.56, 63.64000000000001, 61.47, 62.8, 65.26, 63.53, 60.07, 63.95000000000001, 64.26, 63.29, 64.13, 62.24000000000001, 63.25000000000001, 63.150000000000006, 65.39, 65.42999999999999, 64.01, 65.06, 65.51, 65.52, 64.94, 64.42, 65.46000000000001, 63.57000000000001]

test C: 69.75%
test 2x2: 51.05%
test 3x3: 55.45%
test LR: 63.95%
test UD: 68.4%
test inC: 75.7%
test in2x2: 69.6%