Repository navigation
Add cuDNN deterministic env variable (only for convolution) - #24301
Closed
duncanriach wants to merge 2 commits into
Closed
duncanriach wants to merge 2 commits into
duncanriach wants to merge 2 commits into
Conversation
Contributor
Author
|
I will to close this pull request for now and open another one with master as base. However, at the time of writing this comment, master is not building for me. |
Contributor
Author
|
See follow-up pull request based on master. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This change is a major component of the recipe for making TensorFlow training reproducible on GPUs.
Setting the environment variable TF_CUDNN_DETERMINISTIC=1 (or true) will ensure that both forward and backwards convolution algorithms are both fixed and deterministic. It overrides autotune and selects deterministic back-prop algorithms.