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
- [Bugfix][Model] GraniteMoeHybrid: skip other ranks' weights under pipeline parallelism
- [Feature]: shared expert fusion via Flashinfer kernels
- [Bugfix][Gemma4] Align enable_thinking default with chat template
- [Bugfix] Fix get_open_port() livelock on DP-reserved ports and cover get_open_ports_list
- [RFC]: Pluggable structured-output backends + GRID as a third contract point (declared outcomes for schemas you don't control)
- [Bug]: Kimi-K2.6 TP=4 segfaults in gptq_marlin_repack on GB200 ARM64 with vLLM 0.26.0
- [BugFix] RayExecutorV2: choose the TCPStore port on the node that binds it
- [Bugfix][Core] Add VLLM_FORCE_FP32_ALL_REDUCE env var to fix PCIe non-english logit drift (#40725)
- [compile] Remove layer_name from unified_kv_cache_update
- profiler: add PrivateUse1 activity support for custom backends
- Docs
- Python not yet supported