Generate: replace breaks by a loop condition #29662
Merged
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
gante
commented
Mar 14, 2024
|
|
||
| if this_peer_finished and not synced_gpus: | ||
| break | ||
| this_peer_finished = unfinished_sequences.max() == 0 |
There was a problem hiding this comment.
The previous version is also a data-dependent control flow, so this change is for torch.compile readiness :)
|
FYI @zucchini-nlp (the stopping criteria solution did not preserve ZeRO stage 3 support) |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
amyeroberts
approved these changes
Mar 15, 2024
src/transformers/generation/utils.py
Outdated
Comment on lines
1795
to
1798
| else: | ||
| if this_peer_finished: | ||
| return False | ||
| return True |
There was a problem hiding this comment.
Or actually, we can just do
Suggested change
| else: | |
| if this_peer_finished: | |
| return False | |
| return True | |
| return not this_peer_finished |
There was a problem hiding this comment.
This solution can return False when synced_gpus is True and this_peer_finished is True, which is not intended -- this_peer_finished has to be True in all distributed devices when synced_gpus is True 🤗
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
itazap
pushed a commit
that referenced
this pull request
May 14, 2024
* replace breaks by a loop condition * Update src/transformers/generation/utils.py Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com> --------- Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
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?
Pulled from the
torch.compile(..., fullgraph=True)draft PR: #29374It replaces the
breaksthat exit the endless generation loop with an equivalent function that returnsFalsewhen it should stop generating, while preserving ZeRO stage 3 support. It is not only an improvement in terms of code reuse, but also a hard requirement to enabletorch.compile(..., fullgraph=True):breakand data-dependent control flow is not supported.