Yunhe Li
Yunhe Li
Hi, I guess it is not a bug. ``` python self.d_bn1 = batch_norm(name='d_bn1') self.d_bn2 = batch_norm(name='d_bn2') t_vars = tf.trainable_variables() self.d_vars = [var for var in t_vars if 'd_' in var.name]...
Oh I see. It may be a small bug. Because `updates_collections=None` force the updates for the moving mean and the moving variance in place. We may need to set `updates_collections`...