Skip to content

[Misc] Adjust max_position_embeddings for LoRA compatibility #8957

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 1 commit into from
Sep 30, 2024
Merged

[Misc] Adjust max_position_embeddings for LoRA compatibility #8957

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 9 additions & 2 deletions vllm/worker/model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1037,9 +1037,17 @@ def load_model(self) -> None:
assert supports_lora(
self.model
), f"{self.model.__class__.__name__} does not support LoRA yet."

if supports_multimodal(self.model):
logger.warning("Regarding multimodal models, vLLM currently "
"only supports adding LoRA to language model.")
# It's necessary to distinguish between the max_position_embeddings
# of VLMs and LLMs.
if hasattr(self.model.config, "max_position_embeddings"):
max_pos_embeddings = self.model.config.max_position_embeddings
else:
max_pos_embeddings = (
self.model.config.text_config.max_position_embeddings)

self.lora_manager = LRUCacheWorkerLoRAManager(
self.scheduler_config.max_num_seqs,
Expand All @@ -1049,8 +1057,7 @@ def load_model(self) -> None:
self.device,
self.model.embedding_modules,
self.model.embedding_padding_modules,
max_position_embeddings=self.model.config.
max_position_embeddings,
max_position_embeddings=max_pos_embeddings,
)
self.model = self.lora_manager.create_lora_manager(self.model)

Expand Down
Loading