Skip to content

[Bugfix][v1] fixed llava-hf/llava-1.5-7b-hf is broken on V1 #14554

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
merged 2 commits into from
Mar 10, 2025
Merged

[Bugfix][v1] fixed llava-hf/llava-1.5-7b-hf is broken on V1 #14554

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a0f91ce
[Bugfix][v1] fixed llava-hf/llava-1.5-7b-hf is broken on V1
chaunceyjiang Mar 10, 2025
da48540
Add note
DarkLight1337 Mar 10, 2025
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
20 changes: 12 additions & 8 deletions vllm/model_executor/models/llava.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -783,15 +783,19 @@ def get_multimodal_embeddings(self, **kwargs) -> Optional[NestedTensors]:
if image_input is None:
return None
vision_embeddings = self._process_image_input(image_input)
if kwargs.get("v0_path", False):

if kwargs.get("v0_path", False) or \
image_input.get("feat_is_patch") is None or \
image_input.get("embed_is_patch") is None:
# The path is used for pixtral (V0 only) and llava (V0/V1)
return vision_embeddings
else:
nested_emb = [
self._get_mm_embeds(*args) for args in zip(
vision_embeddings, image_input["feat_is_patch"],
image_input["num_crops"], image_input["embed_is_patch"])
]
return flatten_2d_lists(nested_emb)

nested_emb = [
self._get_mm_embeds(*args) for args in zip(
vision_embeddings, image_input["feat_is_patch"],
image_input["num_crops"], image_input["embed_is_patch"])
]
return flatten_2d_lists(nested_emb)

def get_input_embeddings(
self,
Expand Down