fix iteration timing used in autotuning when gradient_accumulation_steps > 1 #2888
Conversation
This file contains 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
tjruwase
reviewed
Feb 24, 2023
auto-merge was automatically disabled
July 6, 2023 17:12
Merge queue setting changed
tjruwase
approved these changes
Sep 6, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
the number of flops (flops profiler) is captured at global step boundary, thus when gradient_accumulation_steps > 1, the iteration time (from fwd, bwd and step timers capturing micro steps) used in the autotuner needs to be multiplied by the gradient_accumulation_steps.