Remove device_count for TPU launcher to avoid initializing runtime #3587
Conversation
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
SunMarc
reviewed
May 26, 2025
Comment on lines
-901
to
-904
| if args.num_processes and args.num_processes != device_count(): | ||
| raise ValueError( | ||
| f"Number of processes ({args.num_processes}) must match the number of TPU devices ({device_count()})" | ||
| ) |
There was a problem hiding this comment.
if you are removing that, maybe we should put somewhere that we can running the script on all tpu cores available. -> print("Launching a training on all TPU cores.")
There was a problem hiding this comment.
IIRC, now XLA expects no arguments at all, meaning it will use all TPU cores. If that is correct, the change seems correct.
tengomucho
approved these changes
May 26, 2025
Comment on lines
-901
to
-904
| if args.num_processes and args.num_processes != device_count(): | ||
| raise ValueError( | ||
| f"Number of processes ({args.num_processes}) must match the number of TPU devices ({device_count()})" | ||
| ) |
There was a problem hiding this comment.
IIRC, now XLA expects no arguments at all, meaning it will use all TPU cores. If that is correct, the change seems correct.
sorgfresser
commented
May 27, 2025
| @@ -898,10 +897,6 @@ def tpu_launcher(args): | |||
| f"Your training script should have a function named {args.main_training_function}, or you should pass a " | |||
| "different value to `--main_training_function`." | |||
| ) | |||
There was a problem hiding this comment.
Suggested change
| ) | |
| ) | |
| logger.warning("Launching training on all TPU cores.") |
Like that? @SunMarc
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Outdated suggestions cannot be applied.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.
What does this PR do?
A similar issue to PR #3541 remains for the TPU launcher. With this, we avoid prematurely initializing the XLA runtime.
currently fails. This fixes it. Tested on gcloud v6e TPUs.
Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
@SunMarc @zach-huggingface