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
- [RFC]: Universal Triton kernel autotuning at warmup
- [Bugfix][Spec Decode] Fix DeepSeek V4 DFlash input and cache formats
- [Frontend] Place Anthropic inline system messages per renderer
- [Bugfix][Spec Decode] Keep MLA-only KV cache dtypes away from non-MLA drafters
- [Rust Frontend] Add OpenAI Responses API endpoint
- [Bug]: moe_permute can divide by zero with no tokens
- [Bugfix][MoE] Handle empty routing in moe_permute
- [Perf][MoE] RoutedExperts: O(1) Pre-Indexed Weight Mapping, Bulk 3D Loading Interface, and Dynamic Kwargs Filtering
- [Bugfix][Frontend] Accept thinking content parts in the Inkling renderer
- [ROCm][Perf] Kimi-K3: O(1) Pre-Indexed Weight Mapping and Scale Key Collision Avoidance in AMD Linear Loader
- Docs
- Python not yet supported