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 supported62 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Kimi-K3][ROCm] Reuse one attention metadata build across the KV cache groups
- Fix boolean parsing for LMCACHE_FORCE_SKIP_SAVE environment variable
- [Spec Decode] Log adaptive verification metrics and align request accounting
- [Bugfix][Tool Parser] Keep newlines inside Olmo3 tool call arguments
- [Perf][Gemma4][V1] Enable and optimize KV-sharing fast prefill and decode metadata
- [Bugfix][MoE] Fail fast when batched_triton lacks batched prepare/finalize
- [Frontend] Add CLI configuration for HTTP keep-alive timeout
- [Bug]: CUTLASS MoE kernel benchmarks fail with ImportError: cannot import name 'fused_topk'
- [Bugfix] Keep the AOT load-or-compile decision uniform across ranks
- [Perf] Pick the SM90 block-FP8 CUTLASS kernel by M size, not M % 4
- Docs
- Python not yet supported