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 supported59 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [ROCm][Multimodal][MiniMax-M3] Sequence-parallel vision encoder for single-image batches
- [Feature]: Add opt-in in-flight request deduplication for chat completions
- [Bug][Spec Decode][Structured Output] Reverting #54442 reopens grammar-invalid proposal acceptance in async scheduling
- [Perf][MiniMax-M3] Raise the fused MoE finalize + all-reduce cap from 32 to 128 tokens
- [Perf][Sampling] Gather vocab-parallel logits into their final layout in one pass
- [Feature]: Replace CUTLASS MXFP4 W4A4 MoE (CutlassExpertsMxfp4) with FlashInfer kernels on Blackwell
- [CI Failure][ROCm]: MI355 spec decode tests OOM with the extensible KV cache
- [Perf][MiniMax-M3.1-Flash] Specialized decode on B200 and B300
- [Build] Only build FA2 by default when targeting SM8x/SM12x
- [Bugfix][Distributed] Copy misaligned packed weight-transfer entries on unpack
- Docs
- Python not yet supported