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
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Perf][Qwen3.5] Avoid GDN decode gate copies
- [Misc] Point the unquantized-MoE backend error at the speculative-config fix
- [Test] Add Thai coverage to detokenizer tests
- [Bugfix][Model] Reuse CUDA segments when loading Inkling expert weights
- [Bugfix][Model] Fix CUDA graph capture in GLM4-MoE-Lite MTP layer
- [Bug]: DeepSeek-V4-Flash: illegal memory access during FULL cudagraph capture (all-zero indexer seq_lens hit an OOB read in DeepGEMM's paged-MQA metadata kernel)
- [Core] Add per-request prefix-cache write policy
- [Performance]: Erratic/poor DeepSeek V4 Pro MTP results on MI325X (gfx942) — ROCm kernels untuned
- Cached blocks never hit eviction first
- [Bug] FlashInferMLASparseSM120Impl missing masked_mha_available: engine crash on sm_121 (GB10) at startup
- Docs
- Python not yet supported