vllm
https://github.com/vllm-project/vllm
Python
A high-throughput and memory-efficient inference and serving engine for LLMs
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported23 Subscribers
Add a CodeTriage badge to vllm
Help out
- Issues
- [Feature][V1][P/D]: Support P/D num_cached_tokens in response usage
- [Bug] Pass correct ep_rank_start and ep_rank_end
- [Bug]: torch._dynamo hit recompile_limit using FlexAttention backend
- [Bug]: VLLM Hang when DP=2, USE_EP=TRUE, backend = pplx
- Experimental attention backend in helion
- [Bugfix] Actually enable serialize_messages for harmony Responses (related to #26185)
- [Bug]: `guided_choice` and `guided_decoding_backend` don't work when `enable_thinking=False` for Qwen3
- [RFC]: Consolidated tool call parser implementations by type (JSON, Python, XML, Harmony)
- [Usage]: qwenvl2.5如何指定max_pixels
- [RFC]: SageMaker Day-0 Compatibility with vLLM
- Docs
- Python not yet supported