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
- [CPU] UltraQuant 4-bit KV cache on CPU
- [Bug]: custom all-reduce creates its NVLink multicast reduce-scatter buffer mid-serving (not at startup); the run then crashes (Xid + CUBLAS_STATUS_EXECUTION_FAILED)
- [Bugfix][Audio] Honor VLLM_MAX_AUDIO_CLIP_FILESIZE_MB=0 as documented
- [KV Connector][Mooncake] Add DFS support to MooncakeStoreConnector
- [ROCm][DSv4] Enable EPLB on AITER MegaMoE
- [Bugfix][Spec Decode] Relay Qwen3.5 auxiliary states over PP
- [Bug]: mamba_config backend is silently overridden by the --mamba-backend default
- [Bug]: /v1/systemone reads score levels named "1".."5" one level high, and letter choices out of order as the wrong option
- [Bugfix][Frontend] systemone: take option labels from option names
- [Bugfix][Mamba] Keep mamba_config backend unless --mamba-backend is set
- Docs
- Python not yet supported