search

tatsy / keras-generative

Deep generative networks, coded with Keras.
122 stars 28 forks source link

Zero_loss function for EBGAN #12

Closed adnan1306 closed 6 years ago

adnan1306 commented 6 years ago

Can you please explain how the zero loss function exactly works? def zero_loss(y_true, y_pred): return K.zeros_like(y_true)

Isn't the loss always zero if we use this function?

tatsy commented 6 years ago

You are right. For Keras, we have to give the loss for each batch. The zero_loss function returns single zeros for each batch.