Fix Qwen3Next dtype API usage #41735
Merged
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
vasqu
reviewed
Oct 20, 2025
src/transformers/generation/utils.py
Outdated
Comment on lines
1848
to
1850
| # Safety: if the model is sharded across multiple devices (hf_device_map/device_map) and we are | ||
| # doing sampling, enable `remove_invalid_values` by default to avoid NaN/Inf logits causing CUDA | ||
| # asserts during multinomial sampling. Users can still override this by passing the flag explicitly. |
There was a problem hiding this comment.
Unrelated changes? Also in logits_process.py
4 tasks
|
Yeah, if you could revert the unrelated changes this PR is good! |
Replace torch.get_current_dtype() with torch.get_default_dtype() to fix FLA compatibility
SrijanUpadhyay
force-pushed
the
fix-qwen3next-dtype-api
branch
from
15485b7 to
6717030
Compare
|
[For maintainers] Suggested jobs to run (before merge) run-slow: qwen3_next |
|
Thanks for pointing it out, i have made this pr clean and added other changes to #41734. Sorry for the delay. Please check and provide me the feedback or any further improvements. |
|
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. |
ngazagna-qc
pushed a commit
to ngazagna-qc/transformers
that referenced
this pull request
Oct 24, 2025
Replace torch.get_current_dtype() with torch.get_default_dtype() to fix FLA compatibility
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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Outdated suggestions cannot be applied.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.
This PR fixes an invalid PyTorch API usage in the Qwen3Next model.
Changes
torch.get_current_dtype()withtorch.get_default_dtype()in both modular and modeling filesTechnical Details
get_current_dtype()get_default_dtype()which returns the global default dtype settingTesting
The changes have been tested with
make fixupand pass the repository consistency checks.Fix #41732