Skip to content
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

[multi-step] Raise error if not using async engine #7703

Merged
merged 2 commits into from
Aug 21, 2024
Merged

[multi-step] Raise error if not using async engine #7703

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
71f21e1
Raise error if not using async engine
SolitaryThinker Aug 20, 2024
ed704ac
Trigger CI
SolitaryThinker Aug 20, 2024
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
5 changes: 5 additions & 0 deletions vllm/engine/llm_engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1291,6 +1291,11 @@ def step(self) -> List[Union[RequestOutput, EmbeddingRequestOutput]]:
raise NotImplementedError(
"Pipeline parallelism is only supported through AsyncLLMEngine "
"as performance will be severely degraded otherwise.")

if self.scheduler_config.num_scheduler_steps > 1:
raise NotImplementedError(
"Multiple scheduler steps (multi-step) are only supported "
"through AsyncLLMEngine. ")
seq_group_metadata_list, scheduler_outputs = self.scheduler[
0].schedule()

Expand Down
Loading