※現在、ブログ記事を移行中のため一部表示が崩れる場合がございます。
順次修正対応にあたっておりますので何卒ご了承いただけますよう、お願い致します。
CIFAR-10$ python train_cifar5.py -g 0 -u 500 500 -p 3 GPU: 0 # number: 50000 # units: [500, 500] # pooling: 3 # Minibatch-size: 100 # epoch: 20 epoch main/loss validation/main/loss main/accuracy validation/main/accuracy elapsed_time 1 1.36881 1.05861 0.50406 0.6211 9.16741 2 0.941233 0.887284 0.66934 0.6877 18.1766 3 0.765903 0.862215 0.73098 0.6993 27.2039 4 0.628062 0.834987 0.7782 0.715 36.3343 5 0.491332 0.892278 0.82718 0.7122 45.5346 6 0.369885 0.987892 0.86914 0.7141 54.6321 7 0.258368 1.06508 0.909741 0.7145 63.7151 8 0.185255 1.28405 0.935921 0.7027 72.7915 9 0.134805 1.41083 0.953181 0.7086 81.8067 10 0.102211 1.61051 0.964881 0.7063 90.8323 11 0.0954875 1.59982 0.966341 0.7073 99.89 12 0.0744035 1.87134 0.974921 0.6999 108.908 13 0.0799358 1.87029 0.972621 0.7059 117.916 14 0.064477 1.79121 0.977761 0.706 126.942 15 0.0539574 1.99034 0.98168 0.6957 135.992 16 0.0564298 2.12325 0.981061 0.7012 145.031 17 0.0636167 1.98566 0.978781 0.6953 154.082 18 0.0543167 2.06661 0.98182 0.6956 163.078 19 0.0525986 2.02431 0.982701 0.7004 172.108 20 0.0473705 2.20115 0.98364 0.6995 181.123 CIFAR-10$
CIFAR-10$ python train_cifar5.py -g 0 -u 500 500 -p 4 GPU: 0 # number: 50000 # units: [500, 500] # pooling: 4 # Minibatch-size: 100 # epoch: 20 epoch main/loss validation/main/loss main/accuracy validation/main/accuracy elapsed_time 1 1.3321 1.0314 0.51792 0.6355 8.92798 2 0.918795 0.892485 0.67598 0.6891 17.7368 3 0.739756 0.828365 0.74008 0.7125 26.4688 4 0.599331 0.809332 0.78932 0.7276 35.2266 5 0.462065 0.8635 0.8369 0.7233 44.0019 6 0.334985 0.901586 0.88288 0.7274 52.7741 7 0.235922 1.03643 0.916901 0.7221 61.5884 8 0.16752 1.19928 0.941761 0.7251 70.3582 9 0.127554 1.30587 0.955101 0.7188 79.1021 10 0.106585 1.38816 0.962142 0.7147 87.8739 11 0.0812968 1.61297 0.971281 0.7076 96.6838 12 0.0796875 1.69275 0.972361 0.7081 105.472 13 0.0682185 1.78928 0.97666 0.7108 114.232 14 0.0667448 1.75986 0.977441 0.7132 123.013 15 0.054462 1.89674 0.981441 0.7137 131.803 16 0.0661604 1.76279 0.977181 0.7202 140.547 17 0.0535431 1.84795 0.981821 0.7185 149.297 18 0.0534156 1.9057 0.982001 0.7144 158.044 19 0.0432006 2.0893 0.98536 0.7146 166.812 20 0.0517301 2.09486 0.9826 0.7156 175.534 CIFAR-10$