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
- test: parallel sampling output kinds
- [Distributed][MoonEP] Sharded symmetric-memory expert weights and MoonEP release pin
- [Bugfix][KV Offload] Use the group's own DCP factor for the SWA store mask
- [Bug]: Silent data loss when typo in `role`, server deletes the whole message and request returns 200
- [Bugfix] Secure /scale_elastic_ep: API-key auth, bool rejection, 400 when elastic EP disabled
- [Bugfix][ROCm] Fix ROCM_AITER_FA context loss on out-of-order batches
- [Bugfix][NIXL] Re-save the block that straddles a chunked-prefill boundary in host-buffer mode
- PP on docker: gloo control plane resolves to bridge IP (535ms per send) and presents as a long-context decode collapse — suggest defaulting GLOO_SOCKET_IFNAME=lo for same-netns ranks
- [Attention] Support FlashInfer re-paging for packed BLHNC KV caches
- Restrict v1 engine zmq IPC sockets to a per-user private directory
- Docs
- Python not yet supported