Mason Chang comments

Repositories
Issues
Comments

Results 2 comments of


                                            Mason Chang

tf.concat (and tf.transpose) inside a for loop with tf.range in the context of a GradientTape while using XLA dosn't work

I'm not sure this is expected behavior or not. I was able to simplify the program a bit: ``` class Model(): def __init__(self, batchsize): self.batchsize = batchsize self.g = tf.Variable(initial_value=tf.ones([self.batchsize,1,1]),...

Automatically move CUDA non XLA Tensors to XLA Device and back to CUDA device

> Is this an experimental pr or you want to merge this? Ideally we would merge this, or is there a reason not to?