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 supported52 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Bug]: int8_per_token_head KV + prefix caching corrupts output when the KV pool is pinned at 100% (Gemma-4 hybrid, Triton)
- [Bug]: With qwen3.5-35b-a3b, the performance is relatively poor both when using dflash and when not using it, but the accepted length of dflash is around 5–6.
- [Bug]: DeepSeek-V4-Flash-0731 + DSpark fails on RTX PRO 6000 (SM120) with FlashInfer sparse MLA decode kernel routing
- [Feature]: Kimi K3 Performance Optimization
- [Bugfix] Pull runai_streamer non-tensor model files on every node
- [Bugfix][Benchmarks] Count tokens, not chunks, in peak output throughput
- Hybrid multi-group KV: _update_requests_with_invalid_blocks crashes (too many values to unpack) on connector load-error blocks
- [Bug]: gfx1100 (RDNA3): first-call non-determinism + long-seq corruption from Triton paged-attention fallback (ROCm 7.14)
- [Bugfix] Fix NVFP4 shape mismatch in flashinfer_scaled_fp4_mm #50557
- [Quantization][ModelOpt] Serve modelopt_mixed checkpoints with FP8_PB layers
- Docs
- Python not yet supported