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 supported62 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Bugfix][ROCm][MLA] Normalize FlashAttention LSE layout in mixed decode/prefill scheduling
- Fix duplicated word in cohere_asr processor comment
- [CI] Use readable titles for nightly perf-eval builds
- [ROCm] Fall back from unsupported AITER FP8 MoE and custom all-reduce
- [Core] Add --kv-cache-host-pinned: host-pinned KV cache on integrated GPUs (GB10, Thor)
- [Feature]: Zero-copy RDMA for the KV cache on DGX Spark (GB10): device memory is not registrable, host-pinned is
- [Perf][MoE] Add tuned Triton FusedMoE config for Gemma 4 26B-A4B on H20 (E=128,N=704)
- [Kernel] Parallelize CUTLASS FP4 expert offset scan
- Add TurboQuant variant support for triton_attn
- [ROCm] Fuse Kimi-K3 DSpark QK RoPE with the MLA cache write
- Docs
- Python not yet supported