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
- [CI Failure]: V1 KV Offload - tests/v1/kv_offload/tiering/p2p/test_manager.py::TestBidirectionalManager::test_lookup_expiry_allows_zmq_reconnect
- [Perf][Attention] Larger prefill Q tile for Triton unified attention on sm80/sm89
- [Bugfix][Structured Output] Reject unproductive $ref cycles before backend selection
- [RFC]: Context-length-bounded FULL decode CUDA graphs in Model Runner V2
- [Bugfix] Don't initialize CUDA when importing the FLA ops
- [CLI] Add interactive chat recovery, commands, and line editing
- [Bugfix][Frontend] Reject nonzero priorities without priority scheduling
- [ROCm][MiniMax] Add MiniMax-M3.1 AMD compatibility via MXFP4 experts and FP8 KV
- [Bug][Anthropic API] /v1/messages/count_tokens requires --tool-call-parser when tools are present
- [Bug][Rust Frontend] Tool requests return 500 on every request when the `auto` tool parser cannot resolve a parser for the model
- Docs
- Python not yet supported