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 supported51 Subscribers
Add a CodeTriage badge to vllm
Help out
- Issues
- [Bug]: MiniMax-M3 fp8 KV cache on SM80 (A100/A800): coherent with --enforce-eager, garbage under CUDA graphs — root cause + working fix
- [Bugfix] MiniMax-M3 fp8_e5m2 KV cache on SM80 (A100/A800): admit the format Triton can read, fix the CUDA-graph KV corruption
- [Core] Mix per-prompt seeds into seeded SamplingParams in the offline API
- [Bug]: AOT compile cache key ignores limit_mm_per_prompt, so image:0 and image:1 runs collide and crash with AttributeError: 'NoneType' object has no attribute 'size'
- [Bugfix] Reject --renderer-num-workers > 1 with mm cache for generation models
- [Core][Spec Decode] Add lookahead-aware prefix cache hashing for EAGLE-style draft models
- [Core] Support partial layer loading natively using hf_overrides and ignore_patterns
- [RFC]: Cutting P/D-disaggregation TTFT: first-token seeding, fast KV side channels, and concurrent router dispatch (NixlConnector)
- [HelionLinearBackend][5/N] Add HelionFP8BlockScaledMMLinearKernel
- [Bug]: CUDA misaligned address crash on GB10 (sm_121) after ~10 days uptime — Nemotron NVFP4 + Marlin MoE + MTP speculative decoding
- Docs
- Python not yet supported