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
- [RFC]: [Draft] Split FusedMoEQuantConfig into static schema and runtime state
- [Triton] Enable tunable BLOCK_SIZEs based on current plarform
- [Bugfix][MoRIIO] Expire write_done notifications no request matches
- [ROCm][Perf][Kimi-K3] Add HipCustomCommunicator, a communicator to support complex fusions.
- More gfx1250 models, di, and perf fixes
- [ROCm][Perf][MLA] Fold the decode RoPE + KV-cache write into the AITER fused decode query
- [ROCm] Guard FlashInfer import and add portable rsqrt fallback in Flint
- [Bugfix][CPU] Compute beta in float32 in the GDN decode kernel
- [Bug]: /v1/completions top_logprobs overwrites the sampled token's entry when candidates decode to the same string (byte fragments -> "")
- [Refactor][Spec Decode] Use the DFlash parent helper for Muse DSpark context KV init
- Docs
- Python not yet supported