Disable nondeterministic implementation for max_pool2d_backward
#619
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
We disable the nondeterministic implementation of
max_pool2d_backward
by default, so that to avoid potential accuracy issues.The background is that we encountered eager two-run-differ problems with
alexnet
(without enabling deterministic). However, CUDA could run without enabling it. To align with the CUDA's behavior, we decided to always go to the deterministic path in this kernel.This PR adds a flag
XPU_ALLOW_UNDETERMINISTIC
. In the future, we may enable this path for performance consideration.