Sanjoy Das
Sanjoy Das
> I think I'm doing that already beeing even more conservative. The relevant parts from my above command also reformatted: Thanks, the reformat made it much easier to read. :)...
> How can a test fail when a GPU is available? In this specific case it looks like the test needs more than 1GB to run. > Hence I would...
The test tags could use some cleaning up, but that's more complicated than it sounds since there are several bazel macros involved here (and more blaze macros internally at Google)....
Hi, We expect this to be a non-issue once we're using the CUDA malloc async allocator by default. Can you give it a try? You can enable it by adding...
I believe this is fixed by https://github.com/tensorflow/tensorflow/pull/49173 (CC @nouiz )
Hi @WangTuoxyty, The benchmark you're using is very small so tf32 or not will not make a big difference, do you see the same issue when you try larger convolutions?
This hangs inside `ParallelMapIterator::~ParallelMapIterator`. Assigning to @jsimsa for further triaging.