For tf.gradients(), do not backpropagate through integer tensors.
All integer tensors are now considered constant with respect to all `xs`. This fixes a bug in gradients through tf.while_loop. PiperOrigin-RevId: 194438529
Loading
Please sign in to comment