Open batizhou opened 5 years ago
In the example of Variational Auto-Encoder
kl_div_loss = 1 + z_std - tf.square(z_mean) - tf.exp(z_std) kl_div_loss = -0.5 * tf.reduce_sum(kl_div_loss, 1)
I don't understand how this formula is come. Can someone help explain it?
In the example of Variational Auto-Encoder
KL Divergence loss
kl_div_loss = 1 + z_std - tf.square(z_mean) - tf.exp(z_std) kl_div_loss = -0.5 * tf.reduce_sum(kl_div_loss, 1)
I don't understand how this formula is come. Can someone help explain it?