diff options
| author | nunzip <np.scarh@gmail.com> | 2019-03-04 21:38:17 +0000 |
|---|---|---|
| committer | nunzip <np.scarh@gmail.com> | 2019-03-04 21:38:17 +0000 |
| commit | 6529cc095c57e375f34d69fb6bfb36d058dd2192 (patch) | |
| tree | a2dd9c3d09b8d8d50ce754c1a65eb6869035d705 /cgan.py | |
| parent | f00bc97bcb820d30d73fed37eb5c0d5ffddcd9ca (diff) | |
| download | e4-gan-6529cc095c57e375f34d69fb6bfb36d058dd2192.tar.gz e4-gan-6529cc095c57e375f34d69fb6bfb36d058dd2192.tar.bz2 e4-gan-6529cc095c57e375f34d69fb6bfb36d058dd2192.zip | |
Improve one sided smoothing
Diffstat (limited to 'cgan.py')
| -rwxr-xr-x | cgan.py | 3 |
1 files changed, 2 insertions, 1 deletions
@@ -142,9 +142,10 @@ class CGAN(): # Train the discriminator if smooth == True: d_loss_real = self.discriminator.train_on_batch([imgs, labels], valid*0.9) + d_loss_fake = self.discriminator.train_on_batch([gen_imgs, labels], valid*0.1) else: d_loss_real = self.discriminator.train_on_batch([imgs, labels], valid) - d_loss_fake = self.discriminator.train_on_batch([gen_imgs, labels], fake) + d_loss_fake = self.discriminator.train_on_batch([gen_imgs, labels], fake) d_loss = 0.5 * np.add(d_loss_real, d_loss_fake) # --------------------- |