Auto-log parallelism info to wandb.config using HF Accelerate#40055
Open
WoosungMyung wants to merge 1 commit intohuggingface:mainfrom
Open
Auto-log parallelism info to wandb.config using HF Accelerate#40055WoosungMyung wants to merge 1 commit intohuggingface:mainfrom
WoosungMyung wants to merge 1 commit intohuggingface:mainfrom
Conversation
Signed-off-by: WoosungMyung <dntjd517@naver.com>
Member
|
cc @SunMarc |
MekkCyber
approved these changes
Aug 26, 2025
| # Auto log Accelerate parallelism info to wandb.config | ||
| if self._initialized and state.is_world_process_zero and getattr(self._wandb, "run", None) is not None: | ||
| acc = getattr(model, "accelerator", None) | ||
| pc = getattr(acc, "parallelism_config", None) if acc is not None else None |
Contributor
There was a problem hiding this comment.
I don't think the if check is necessary since we can use getattr with NoneType as well in python
Suggested change
| pc = getattr(acc, "parallelism_config", None) if acc is not None else None | |
| pc = getattr(acc, "parallelism_config", None) |
| if self._initialized and state.is_world_process_zero and getattr(self._wandb, "run", None) is not None: | ||
| acc = getattr(model, "accelerator", None) | ||
| pc = getattr(acc, "parallelism_config", None) if acc is not None else None | ||
| sizes = getattr(pc, "_sizes", None) if pc is not None else None |
Contributor
There was a problem hiding this comment.
Suggested change
| sizes = getattr(pc, "_sizes", None) if pc is not None else None | |
| sizes = getattr(pc, "_sizes", None) |
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
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
This PR adds parallelism info non intrusively to
WandbCallbackso that, when Hugging Face Accelerate is in use, the callback automatically logs the parallelism sizes towandb.configat train start.Why
Users frequently want their distributed setup (e.g.,
tp_size,dp_replicate_size) with Weights & Biases for reproducibility and experiment analysis. Today, this is typically done manually on the user side. This change standardizes and automates that step.This PR is from Issue #39882
@Rocketknight1 @MekkCyber
Thanks for you precious time for reviewing this PR.