Fix several based models' pipeline parallel support#44699
Merged
hmellor merged 6 commits intohuggingface:mainfrom Mar 20, 2026
Merged
Fix several based models' pipeline parallel support#44699hmellor merged 6 commits intohuggingface:mainfrom
hmellor merged 6 commits intohuggingface:mainfrom
Conversation
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
|
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. |
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
hmellor
changed the title
ArthurZucker
approved these changes
Mar 19, 2026
Collaborator
ArthurZucker
left a comment
ArthurZucker
left a comment
There was a problem hiding this comment.
Yeah makes sense, we need to enforce with linter IMO!
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: afmoe, bamba, cohere2, cwm, dots1, gemma2, gemma3, gemma3n, gpt_oss, granitemoehybrid, lfm2, lfm2_moe, llama4, minimax, ministral |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
These models have
base_model_pp_plans but currently do not work because the base model's forward pass depends on all thelayersbeingQwen2VLDecoderLayer.i.e. if one of the layers is removed/replaced with
Identity,decoder_layer.attention_typewill fail with an attribute error.This PR fixes the issue in
/src/transformers/models/qwen2_vl/modeling_qwen2_vl.pyand runspython utils/modular_model_converter.pyto propagate it to the other models inherit the behaviour.