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 supported60 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [ROCm][UMBP] Support UMBP standalone mode for DeepSeek-V4.1-Flash
- [ROCm][MoRIIO] Support DCP for decode instance in PD disaggregated serving with MoRI WRITE mode
- [Bugfix][Tool Parser] LFM2: constrain forced tool_choice to the native format and parse hyphenated tool names
- [Feature]: Add ID-only request logging while preserving lifecycle events
- [Models][XPU] Port eager Qwen3.5 models and fuse MoE decode
- [ROCm][MoE] Add opt-in AITER FlyDSL clamped MXFP4 BF16 experts
- [Misc] Watermarking: read the key from a file
- [KV Connector][NIXL] Add connector checkpoint hooks and restore NIXL P/D workers independently
- [RFC]: Bounded, retryable NCCL weight updates
- [Bugfix] Close the early-PDL completion race for small FlashInfer one-shot all-reduce calls on the trtllm IPC backend
- Docs
- Python not yet supported