gansformer
gansformer copied to clipboard
Two typos in pytorch_version/training/loss.py
loss_D_real = 0
if D_main:
if self.d_loss == "logistic":
loss_D_real = torch.nn.functional.softplus(-real_logits) # -log(sigmoid(real_logits))
elif self.d_loss == "hinge":
loss_D_real = torch.clamp(1.0 - real_logits, min = 0)
elif self.d_loss == "wgan":
loss_D_real = -real_logits + tf.square(real_logits) * wgan_epsilon
training_stats.report("Loss/D/loss", loss_D_gen + loss_D_real)
In line 142 of loss.py, it should be loss_D_real = -real_logits + torch.square(real_logits) * self.wgan_epsilon
instead.