Original generator loss for GANs. L = -log(sigmoid(D(G(z)))) See `Generative Adversarial Nets` (https://arxiv.org/abs/1406.2661) for more details. Args: discriminator_gen_outputs: Discriminator output on generated data. Expected to be in the range of (-inf, inf). Returns:
(discriminator_gen_outputs)
| 174 | |
| 175 | |
| 176 | def generator_loss(discriminator_gen_outputs): |
| 177 | """Original generator loss for GANs. |
| 178 | |
| 179 | L = -log(sigmoid(D(G(z)))) |
| 180 | |
| 181 | See `Generative Adversarial Nets` (https://arxiv.org/abs/1406.2661) |
| 182 | for more details. |
| 183 | |
| 184 | Args: |
| 185 | discriminator_gen_outputs: Discriminator output on generated data. Expected |
| 186 | to be in the range of (-inf, inf). |
| 187 | |
| 188 | Returns: |
| 189 | A scalar loss Tensor. |
| 190 | """ |
| 191 | loss = tf.losses.sigmoid_cross_entropy( |
| 192 | tf.ones_like(discriminator_gen_outputs), discriminator_gen_outputs) |
| 193 | tf.contrib.summary.scalar('generator_loss', loss) |
| 194 | return loss |
| 195 | |
| 196 | |
| 197 | def train_one_epoch(generator, discriminator, generator_optimizer, |
no test coverage detected