diff options
| author | nunzip <np.scarh@gmail.com> | 2019-02-27 22:49:16 +0000 |
|---|---|---|
| committer | nunzip <np.scarh@gmail.com> | 2019-02-27 22:49:16 +0000 |
| commit | 367167680e156ac611c5f1db9f9ff7e66d51a8fe (patch) | |
| tree | d57ac7e74ff5e07d9a66ffe2238378e8cc3946c2 /cgan.py | |
| parent | c7c740caff2afc4d615e289ef147c6228cca4a0e (diff) | |
| download | e4-gan-367167680e156ac611c5f1db9f9ff7e66d51a8fe.tar.gz e4-gan-367167680e156ac611c5f1db9f9ff7e66d51a8fe.tar.bz2 e4-gan-367167680e156ac611c5f1db9f9ff7e66d51a8fe.zip | |
Reshape labels after predict
Diffstat (limited to 'cgan.py')
| -rw-r--r-- | cgan.py | 7 |
1 files changed, 4 insertions, 3 deletions
@@ -207,13 +207,14 @@ class CGAN(): labels_test[i*1000:] = i labels_val[i*500:] = i + train_data = self.generator.predict([noise_train, labels_train]) + test_data = self.generator.predict([noise_test, labels_test]) + val_data = self.generator.predict([noise_val, labels_val]) + labels_train = keras.utils.to_categorical(labels_train, 10) labels_test = keras.utils.to_categorical(labels_test, 10) labels_val = keras.utils.to_categorical(labels_val, 10) - train_data = self.generator.predict([noise_train, labels_train]) - test_data = self.generator.predict([noise_test, labels_test]) - val_data = self.generator.predict([noise_val, labels_val]) return train_data, test_data, val_data, labels_train, labels_test, labels_val |