This repository has been archived by the owner on Aug 7, 2024. It is now read-only.
Checkpoint to reduce fp8_weight tensor saved for backwards #193
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
facebook-github-bot
added
the
CLA Signed
drisspg
force-pushed
the
checkpointing_weight_cast
branch
from
e9cc745 to
c07f1d9
Compare
|
I figure this is worth landing over the other since we anticipate that this will be supported by min-cut-partioner @bdhirsh I need to add the same for delayed scaling and ensure that that does indeed checkpoint correctly |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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.
Summary
Take two of doing this: #186
Issue: #185
I have verified that the the transpose_cast in backwards is fusable with inductor, now we just gotta coax it into doing what we want..
cc @bdhirsh
Eager Logging mode for this PR, using this gist:
https://gist.github.com/drisspg/1b334d851ae0caea22eebb07d3455239