Does TensorFlow 2.0 have any support for making a computation as "recompute during the gradient pass" to save memory? TensorFlow 1.x has tf.contrib.layers.recompute_grad , but contrib
is gone in TF 2.0 and it doesn't look like anyone moved recompute_grad
.
Ya, they didn't move it. This is a big error on Google's part. I won't move to Tensorflow 2 for this reason. Recomputing gradients and creating memory efficient graphs is EXTREMELY helpful, if not vital.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.