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
- [KVConnector][NIXL] Support attention-HMA layouts in pipeline-parallel push prefill
- [Bugfix][Frontend] Enforce parallel_tool_calls=false in the required-tool grammar
- [Bug]: Inkling tool call leaks as content in multi-turn streaming (turn-initial tool call, tool_calls empty)
- [Bugfix] Preserve FlashInfer CUTLASS MoE constant storage across weight reload
- [Bugfix] Zero-fill rematerialized weights so reload preserves padding
- [XPU] Alias is_current_stream_capturing to XPU in cuda wrapper
- [Bugfix][Gemma4] Keep image bidirectional attention within the sliding window on the V2 model runner
- [CPU] check some more implementation assumptions on kv layout
- [Bug]: [XPU] Multi-GPU TP serving hangs on Intel Arc Pro B60 with torch 2.13 wheels (oneCCL 2022.x): warmup allreduce never returns, GuC timeouts, DEVICE_LOST`
- [Bugfix][Core] Don't cache the reader's ring slot across acquire_read retries
- Docs
- Python not yet supported