LoginSignup
0

More than 5 years have passed since last update.

TensorFlowを使ってDir en greyの顔分類器を作ってみた - ⑧学習実行編

Last updated at Posted at 2017-05-31

はじめに

  • ここまでくれば大方完成したようなものです。

学習

実行

python main.py

結果

2017-05-30 17:59:11.750584: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2017-05-30 17:59:11.750621: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2017-05-30 17:59:11.750631: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2017-05-30 17:59:11.750639: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2017-05-30 17:59:11.750655: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.
step 0, training accuracy 0.588235
step 1, training accuracy 0.588235
step 2, training accuracy 0.568627
step 3, training accuracy 0.578431
step 4, training accuracy 0.705882
step 5, training accuracy 0.647059
step 6, training accuracy 0.637255
step 7, training accuracy 0.656863
step 8, training accuracy 0.77451
step 9, training accuracy 0.784314
step 10, training accuracy 0.784314
step 11, training accuracy 0.803922
step 12, training accuracy 0.813725
step 13, training accuracy 0.813725
step 14, training accuracy 0.852941
step 15, training accuracy 0.833333
step 16, training accuracy 0.823529
step 17, training accuracy 0.823529
step 18, training accuracy 0.882353
step 19, training accuracy 0.892157
step 20, training accuracy 0.862745
step 21, training accuracy 0.823529
step 22, training accuracy 0.892157
step 23, training accuracy 0.901961
step 24, training accuracy 0.931373
step 25, training accuracy 0.882353
step 26, training accuracy 0.862745
step 27, training accuracy 0.882353
step 28, training accuracy 0.931373
step 29, training accuracy 0.95098
step 30, training accuracy 0.970588
step 31, training accuracy 0.95098
step 32, training accuracy 0.95098
step 33, training accuracy 0.95098
step 34, training accuracy 0.960784
step 35, training accuracy 0.970588
step 36, training accuracy 0.941176
step 37, training accuracy 0.941176
step 38, training accuracy 0.960784
step 39, training accuracy 0.980392
step 40, training accuracy 0.980392
step 41, training accuracy 0.970588
step 42, training accuracy 0.980392
step 43, training accuracy 0.990196
step 44, training accuracy 0.970588
step 45, training accuracy 0.980392
step 46, training accuracy 0.970588
step 47, training accuracy 0.970588
step 48, training accuracy 0.980392
step 49, training accuracy 0.990196
step 50, training accuracy 1
step 51, training accuracy 0.990196
step 52, training accuracy 0.990196
step 53, training accuracy 0.990196
step 54, training accuracy 0.990196
step 55, training accuracy 1
step 56, training accuracy 1
step 57, training accuracy 1
step 58, training accuracy 1
step 59, training accuracy 0.990196
step 60, training accuracy 0.990196
step 61, training accuracy 0.990196
step 62, training accuracy 0.990196
step 63, training accuracy 1
step 64, training accuracy 0.990196
step 65, training accuracy 0.990196
step 66, training accuracy 0.990196
step 67, training accuracy 0.990196
step 68, training accuracy 0.990196
step 69, training accuracy 0.990196
step 70, training accuracy 0.990196
step 71, training accuracy 0.990196
step 72, training accuracy 0.990196
step 73, training accuracy 0.990196
step 74, training accuracy 1
step 75, training accuracy 1
step 76, training accuracy 0.990196
step 77, training accuracy 0.990196
step 78, training accuracy 0.990196
step 79, training accuracy 0.990196
step 80, training accuracy 0.990196
step 81, training accuracy 0.990196
step 82, training accuracy 0.990196
step 83, training accuracy 1
step 84, training accuracy 1
step 85, training accuracy 0.990196
step 86, training accuracy 0.990196
step 87, training accuracy 1
step 88, training accuracy 0.990196
step 89, training accuracy 1
step 90, training accuracy 0.990196
step 91, training accuracy 0.990196
step 92, training accuracy 0.990196
step 93, training accuracy 0.990196
step 94, training accuracy 0.990196
step 95, training accuracy 0.990196
step 96, training accuracy 0.990196
step 97, training accuracy 0.990196
step 98, training accuracy 0.990196
step 99, training accuracy 0.990196
test accuracy 0.791209

補足と言い訳

  • ここまでいくのに色々大変だった....。
  • 実行時の最初にエラーっぽいものでてますがあえてだしてます。気にしないでください。
  • accuracyが1に近づくほど精度が高いらしい
  • 実行後、main.pyと同じ階層に以下のようなファイルができていることを確認してください
    • model2.ckpt.index
    • model2.ckpt.meta
    • model2.ckpt.data-00000-of-00001
    • checkpoint

全ページリンク

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
0