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 supported37 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [RFC]: Add a fused MoE routing Backend for out-of-tree accelerators
- [Bug]: DSpark acceptance collapses after position 0 on DeepSeek-V4-Flash-0731 (0.26.1rc1)
- [Bugfix] Propagate decode SWA index width for DSpark + FLASHINFER_MLA_SPARSE_DSV4
- [Bug]: int8_per_token_head KV with prefix caching DISABLED corrupts the FIRST generated tokens (Gemma-4 hybrid, Triton, idle KV pool)
- [Bug]: Responses API Returns Raw Malformed Kimi Tool Marker as Assistant Text Instead of function_call
- [Misc] Upgrade fastsafetensors version, fix metadata is null
- [Bug]: NVFP4 MoE falls back to Marlin on sm_121 (GB10) in published builds, making the model unservable on a unified 121 GiB pool
- [Bugfix] Resolve $ref/$defs in tool schemas and preserve definitions
- perf: skip re-validating reused sampling params in streaming input loop
- [Bug]: Qwen parser routes grammar-constrained JSON to reasoning when `enable_in_reasoning=true`
- Docs
- Python not yet supported