Fix dtype in Paligemma #40912
Merged
Fix dtype in Paligemma #40912
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
|
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. |
Cyrilvallez
approved these changes
Sep 16, 2025
Comment on lines
778
to
779
| self.pad_token_id = self.config.pad_token_id if self.config.pad_token_id is not None else -1 | ||
| self.text_config_dtype = self.config.get_text_config().dtype or self.dtype |
There was a problem hiding this comment.
Are those artefacts of modular in the two Gemma3? If so, let's quickly remove it with a del as it's not used anywhere
There was a problem hiding this comment.
it is the modular copy ig, Yeah, we can 'del' explictly
|
[For maintainers] Suggested jobs to run (before merge) run-slow: colpali, colqwen2, gemma3, gemma3n, paligemma |
|
you're the best @zucchini-nlp 💗 |
ErfanBaghaei
pushed a commit
to ErfanBaghaei/transformers
that referenced
this pull request
Sep 25, 2025
* fix dtypes * fix copies * delete unused attr
vijayabhaskar-ev
pushed a commit
to vijayabhaskar-ev/transformers
that referenced
this pull request
Oct 2, 2025
* fix dtypes * fix copies * delete unused attr
yuchenxie4645
pushed a commit
to yuchenxie4645/transformers
that referenced
this pull request
Oct 4, 2025
* fix dtypes * fix copies * delete unused attr
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Outdated suggestions cannot be applied.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?
Fixes #40875 and makes sure we have the same dtype before operations. The attention mask is used by LM and has to be the same dtype. Then we also need to cast VLM outputs because VLM and projection can be of different dtypes in configs
I still wonder if we're supposed to load the models with auto-dtype when doing
from_pretrained, I found that it changed between 4.53 and 4.54 for now