exlnet6.log

cuda
112120
{'abnormal': 0, 'normal': 1}
SqueezeNet(
  (features): Sequential(
    (0): Conv2d(3, 96, kernel_size=(7, 7), stride=(2, 2))
    (1): ReLU(inplace=True)
    (2): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=True)
    (3): Fire(
      (squeeze): Conv2d(96, 16, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(16, 64, kernel_size=(1, 1), stride=(1, 1), groups=16)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(16, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=16)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (4): Fire(
      (squeeze): Conv2d(128, 16, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(16, 64, kernel_size=(1, 1), stride=(1, 1), groups=16)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(16, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=16)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (5): Fire(
      (squeeze): Conv2d(128, 32, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(32, 128, kernel_size=(1, 1), stride=(1, 1), groups=32)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(32, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (6): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=True)
    (7): Fire(
      (squeeze): Conv2d(256, 32, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(32, 128, kernel_size=(1, 1), stride=(1, 1), groups=32)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(32, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (8): Fire(
      (squeeze): Conv2d(256, 48, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(48, 192, kernel_size=(1, 1), stride=(1, 1), groups=48)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(48, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=48)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (9): Fire(
      (squeeze): Conv2d(384, 48, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(48, 192, kernel_size=(1, 1), stride=(1, 1), groups=48)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(48, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=48)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (10): Fire(
      (squeeze): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), groups=64)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(64, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64)
      (expand3x3_activation): ReLU(inplace=True)
    )
    (11): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=True)
    (12): Fire(
      (squeeze): Conv2d(512, 64, kernel_size=(1, 1), stride=(1, 1))
      (squeeze_activation): ReLU(inplace=True)
      (expand1x1): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), groups=64)
      (expand1x1_activation): ReLU(inplace=True)
      (expand3x3): Conv2d(64, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64)
      (expand3x3_activation): ReLU(inplace=True)
    )
  )
  (classifier): Sequential(
    (0): Dropout(p=0.3, inplace=False)
    (1): Conv2d(512, 2, kernel_size=(1, 1), stride=(1, 1))
    (2): AdaptiveAvgPool2d(output_size=(1, 1))
  )
)
Epoch 0/29
----------
Loss after 51136 examples: 0.591
Train Accuracy tensor(0.7314, dtype=torch.float64)
Validation Loss is 0.6159495762825012
Validation Accuracy is 0.6792

One of the best validation accuracy found.

Epoch 1/29
----------
Loss after 102304 examples: 0.542
Loss after 153504 examples: 0.514
Train Accuracy tensor(0.7334, dtype=torch.float64)
Validation Loss is 0.6165332942962647
Validation Accuracy is 0.6788000000000001

Epoch 2/29
----------
Loss after 204672 examples: 0.556
Loss after 255872 examples: 0.624
Train Accuracy tensor(0.7340, dtype=torch.float64)
Validation Loss is 0.6169411102294922
Validation Accuracy is 0.6754

Epoch 3/29
----------
Loss after 307040 examples: 0.620
Loss after 358240 examples: 0.530
Train Accuracy tensor(0.7343, dtype=torch.float64)
Validation Loss is 0.6175694991111755
Validation Accuracy is 0.6816

One of the best validation accuracy found.

Epoch 4/29
----------
Loss after 409408 examples: 0.520
Loss after 460608 examples: 0.572
Train Accuracy tensor(0.7346, dtype=torch.float64)
Validation Loss is 0.6181410458564758
Validation Accuracy is 0.6758000000000001

Epoch 5/29
----------
Loss after 511776 examples: 0.541
Loss after 562976 examples: 0.564
Train Accuracy tensor(0.7352, dtype=torch.float64)
Validation Loss is 0.6192948205947876
Validation Accuracy is 0.677

Epoch 6/29
----------
Loss after 614144 examples: 0.466
Loss after 665344 examples: 0.494
Train Accuracy tensor(0.7346, dtype=torch.float64)
Validation Loss is 0.6212577075481415
Validation Accuracy is 0.676

Epoch 7/29
----------
Loss after 716512 examples: 0.526
Loss after 767712 examples: 0.501
Train Accuracy tensor(0.7360, dtype=torch.float64)
Validation Loss is 0.6205282316207886
Validation Accuracy is 0.677

Epoch 8/29
----------
Loss after 818880 examples: 0.561
Loss after 870080 examples: 0.554
Train Accuracy tensor(0.7367, dtype=torch.float64)
Validation Loss is 0.6205138698577881
Validation Accuracy is 0.6796

Epoch 9/29
----------
Loss after 921248 examples: 0.481
Loss after 972448 examples: 0.613
Train Accuracy tensor(0.7353, dtype=torch.float64)
Validation Loss is 0.6211043663024902
Validation Accuracy is 0.6818000000000001

One of the best validation accuracy found.

Epoch 10/29
----------
Loss after 1023616 examples: 0.556
Loss after 1074816 examples: 0.570
Train Accuracy tensor(0.7364, dtype=torch.float64)
Validation Loss is 0.6218779594421386
Validation Accuracy is 0.6794

Epoch 11/29
----------
Loss after 1125984 examples: 0.541
Loss after 1177184 examples: 0.449
Train Accuracy tensor(0.7372, dtype=torch.float64)
Validation Loss is 0.621931196975708
Validation Accuracy is 0.6776

Epoch 12/29
----------
Loss after 1228352 examples: 0.510
Loss after 1279552 examples: 0.513
Train Accuracy tensor(0.7376, dtype=torch.float64)
Validation Loss is 0.6232577262878418
Validation Accuracy is 0.6736000000000001

Epoch 13/29
----------
Loss after 1330720 examples: 0.503
Loss after 1381920 examples: 0.462
Train Accuracy tensor(0.7389, dtype=torch.float64)
Validation Loss is 0.6239038917541504
Validation Accuracy is 0.6764

Epoch 14/29
----------
Loss after 1433088 examples: 0.532
Loss after 1484288 examples: 0.609
Train Accuracy tensor(0.7388, dtype=torch.float64)
Validation Loss is 0.6238823066711425
Validation Accuracy is 0.6738000000000001

Epoch 15/29
----------
Loss after 1535456 examples: 0.587
Loss after 1586656 examples: 0.577
Train Accuracy tensor(0.7392, dtype=torch.float64)
Validation Loss is 0.625021470451355
Validation Accuracy is 0.6762

Epoch 16/29
----------
Loss after 1637824 examples: 0.515
Loss after 1689024 examples: 0.632
Train Accuracy tensor(0.7389, dtype=torch.float64)
Validation Loss is 0.6255401782989501
Validation Accuracy is 0.6748000000000001

Epoch 17/29
----------
Loss after 1740192 examples: 0.491
Loss after 1791392 examples: 0.526
Train Accuracy tensor(0.7405, dtype=torch.float64)
Validation Loss is 0.6256977388381958
Validation Accuracy is 0.6744

Epoch 18/29
----------
Loss after 1842560 examples: 0.578
Loss after 1893760 examples: 0.547
Train Accuracy tensor(0.7405, dtype=torch.float64)
Validation Loss is 0.6260391654968261
Validation Accuracy is 0.6746

Epoch 19/29
----------
Loss after 1944928 examples: 0.486
Loss after 1996128 examples: 0.529
Train Accuracy tensor(0.7406, dtype=torch.float64)
Validation Loss is 0.6265597964286804
Validation Accuracy is 0.6758000000000001

Epoch 20/29
----------
Loss after 2047296 examples: 0.459
Loss after 2098496 examples: 0.592
Train Accuracy tensor(0.7416, dtype=torch.float64)
Validation Loss is 0.6271675033569336
Validation Accuracy is 0.6742

Epoch 21/29
----------
Loss after 2149664 examples: 0.635
Train Accuracy tensor(0.7413, dtype=torch.float64)
Validation Loss is 0.6284614952087403
Validation Accuracy is 0.6764

Epoch 22/29
----------
Loss after 2200832 examples: 0.593
Loss after 2252032 examples: 0.448
Train Accuracy tensor(0.7417, dtype=torch.float64)
Validation Loss is 0.6287846712112427
Validation Accuracy is 0.6734

Epoch 23/29
----------
Loss after 2303200 examples: 0.520
Loss after 2354400 examples: 0.484
Train Accuracy tensor(0.7419, dtype=torch.float64)
Validation Loss is 0.6285801776885986
Validation Accuracy is 0.6742

Epoch 24/29
----------
Loss after 2405568 examples: 0.486
Loss after 2456768 examples: 0.429
Train Accuracy tensor(0.7427, dtype=torch.float64)
Validation Loss is 0.6304812844276428
Validation Accuracy is 0.6736000000000001

Epoch 25/29
----------
Loss after 2507936 examples: 0.457
Loss after 2559136 examples: 0.549
Train Accuracy tensor(0.7436, dtype=torch.float64)
Validation Loss is 0.6309348356246949
Validation Accuracy is 0.672

Epoch 26/29
----------
Loss after 2610304 examples: 0.483
Loss after 2661504 examples: 0.554
Train Accuracy tensor(0.7435, dtype=torch.float64)
Validation Loss is 0.6309467897415161
Validation Accuracy is 0.6734

Epoch 27/29
----------
Loss after 2712672 examples: 0.501
Loss after 2763872 examples: 0.583
Train Accuracy tensor(0.7438, dtype=torch.float64)
Validation Loss is 0.6320569865226746
Validation Accuracy is 0.6714

Epoch 28/29
----------
Loss after 2815040 examples: 0.688
Loss after 2866240 examples: 0.514
Train Accuracy tensor(0.7447, dtype=torch.float64)
Validation Loss is 0.6316720613479614
Validation Accuracy is 0.675

Epoch 29/29
----------
Loss after 2917408 examples: 0.446
Loss after 2968608 examples: 0.479
Train Accuracy tensor(0.7444, dtype=torch.float64)
Validation Loss is 0.6324887269973755
Validation Accuracy is 0.6728000000000001

Training complete in 86m 20s
Test Loss is 0.6288784625824918
Test Accuracy is 0.6799157303370787