Skip to content

[LoRA][BugFix] Fix skipped LoRA adapters for Mistral3 #36928

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
robertgshaw2-redhat merged 3 commits into from
Mar 18, 2026
Merged

[LoRA][BugFix] Fix skipped LoRA adapters for Mistral3 #36928

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
d11c2c9
[LoRA][BugFix] Fix skipped LoRA adapters for Mistral3
WoosukKwon Mar 12, 2026
646c319
reorder mapping
njhill Mar 18, 2026
ea6f585
Merge remote-tracking branch 'refs/remotes/origin/main' into woosuk/f…
njhill Mar 18, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions vllm/model_executor/models/mistral3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -429,6 +429,9 @@ class Mistral3ForConditionalGeneration(
"model.vision_tower.": "vision_tower.",
"model.multi_modal_projector.": "multi_modal_projector.",
"lm_head.": "language_model.lm_head.",
# Some PEFT LoRAs are trained against the text submodule directly
# and produce names like `base_model.model.model.layers.*`.
"model.": "language_model.model.",
}
)

Expand Down
Loading