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 supported59 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Feature][Kernel] Add opt-in CuTe DSL compile and AOT caches
- [Perf][Rust Frontend] Serialize SSE chunks before building the event
- [Perf][Frontend] Append Responses streaming text in place
- [Doc]: Docs use CLI flag names that `vllm serve` does not register
- [Doc] Add Apple Silicon (Metal) weight compatibility note to the Apple install page
- [ModelRunner V2] Accumulate chunked prompt logprobs on the host
- [BI] Add the batch-invariant CUDA op library
- [BI][DSv4.1] Make dense projections, norms and logprobs batch invariant
- [Feature]: GDN prefill on SM110 (Jetson Thor): use the CuteDSL and FlashInfer kernels
- [Frontend][Reference only] Expose min_tokens and ignore_eos in Messages API
- Docs
- Python not yet supported