MNIST学習 
 深層学習

 機能・要件 
 構成・方式
 タスク
 ライブラリ
 導入
 sample

  MNIST学習
 ・DeepConvNetによる学習 をフレームワーク化
 ・Dataset と DataLoader の機能を用いてもちいてミニバッチを制御
 ・datasets
torch/torchvision の ImageFolder 関数で画像を読み込みを行い、
DataLoader 関数でミニバッチの制御を行う。

 MNIST深層学習
 ・サンプル
mnistTorch.py （データセット）
mnistDeepConvNetTorch.py （MNIST深層学習）
 ・カスタムデータセット
datasets
 ・ローダー
ロード
 ・ディープなニューラルネットのモデル定義
 ・optimizer
 ・学習
 ・cuda（GPU使用）
 ・sample実行経過
 (base)> python mnistDeepConvNetTorch.py  cuda  Namespace(verbose=False, batch_size=64, test_batch_size=1000, no_shuffle=False,\   epochs=20, lr=0.001, seed=1, no_cuda=False, dry_run=False, log_interval=10,\   save_model=None, datadir='')  MyNet(   (conv1): Conv2d(1, 32, kernel_size=(3, 3), stride=(1, 1))   (conv2): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1))   (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)   (dropout1): Dropout(p=0.25, inplace=False)   (dropout2): Dropout(p=0.5, inplace=False)   (fc1): Linear(in_features=9216, out_features=128, bias=True)   (fc2): Linear(in_features=128, out_features=10, bias=True)  )  2022-09-11 17:37:03,594 INFO *** epoch=1/20 ***  2022-09-11 17:37:12,719 INFO test: total=9834/10000 (98.34%)  2022-09-11 17:37:12,719 INFO *** epoch=2/20 ***  2022-09-11 17:37:19,239 INFO test: total=9871/10000 (98.71%)  2022-09-11 17:37:19,239 INFO *** epoch=3/20 ***  2022-09-11 17:37:25,846 INFO test: total=9874/10000 (98.74%)  2022-09-11 17:37:25,846 INFO *** epoch=4/20 ***  2022-09-11 17:37:32,454 INFO test: total=9886/10000 (98.86%)  2022-09-11 17:37:32,454 INFO *** epoch=5/20 ***  2022-09-11 17:37:39,062 INFO test: total=9910/10000 (99.10%)  2022-09-11 17:37:39,062 INFO *** epoch=6/20 ***  2022-09-11 17:37:45,669 INFO test: total=9914/10000 (99.14%)  2022-09-11 17:37:45,669 INFO *** epoch=7/20 ***  2022-09-11 17:37:52,324 INFO test: total=9909/10000 (99.09%)  2022-09-11 17:37:52,324 INFO *** epoch=8/20 ***  2022-09-11 17:37:58,976 INFO test: total=9911/10000 (99.11%)  2022-09-11 17:37:58,976 INFO *** epoch=9/20 ***  2022-09-11 17:38:05,615 INFO test: total=9928/10000 (99.28%)  2022-09-11 17:38:05,615 INFO *** epoch=10/20 ***  2022-09-11 17:38:12,223 INFO test: total=9913/10000 (99.13%)  2022-09-11 17:38:12,223 INFO *** epoch=11/20 ***  2022-09-11 17:38:18,948 INFO test: total=9918/10000 (99.18%)  2022-09-11 17:38:18,948 INFO *** epoch=12/20 ***  2022-09-11 17:38:25,594 INFO test: total=9925/10000 (99.25%)  2022-09-11 17:38:25,610 INFO *** epoch=13/20 ***  2022-09-11 17:38:32,327 INFO test: total=9925/10000 (99.25%)  2022-09-11 17:38:32,327 INFO *** epoch=14/20 ***  2022-09-11 17:38:39,044 INFO test: total=9924/10000 (99.24%)  2022-09-11 17:38:39,044 INFO *** epoch=15/20 ***  2022-09-11 17:38:45,761 INFO test: total=9929/10000 (99.29%)  2022-09-11 17:38:45,761 INFO *** epoch=16/20 ***  2022-09-11 17:38:52,510 INFO test: total=9937/10000 (99.37%)  2022-09-11 17:38:52,510 INFO *** epoch=17/20 ***  2022-09-11 17:38:59,263 INFO test: total=9933/10000 (99.33%)  2022-09-11 17:38:59,263 INFO *** epoch=18/20 ***  2022-09-11 17:39:06,002 INFO test: total=9917/10000 (99.17%)  2022-09-11 17:39:06,002 INFO *** epoch=19/20 ***  2022-09-11 17:39:12,719 INFO test: total=9929/10000 (99.29%)  2022-09-11 17:39:12,719 INFO *** epoch=20/20 ***  2022-09-11 17:39:19,514 INFO test: total=9929/10000 (99.29%)