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 supported61 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [CI][Test] Derive the expected FP8 block scale dtype from the DeepGEMM oracle
- [Doc]: MooncakeStoreConnector guide does not mention kv_load_failure_policy; with the default "fail" a failed load returns an empty response
- [Structured Outputs] Make `auto` pick one backend per server instead of per request
- [Bugfix][Scheduler] Preserve forward capacity for hybrid decode transitions
- [Bugfix][DeepSeek-V4.1] Honor continue_final_message: render the last assistant turn without EOS
- [Kernel][MoE] Accept the single-variant MXFP4 moe_backend aliases
- [ROCm][Ray] Refresh cached AITER env toggles after applying driver env
- [Bugfix][ROCm] Inkling: zero CUDA-graph padding rows in decode attention
- [ROCm][Kernel] Tune the GatedDeltaNet Triton prefill kernels on gfx942
- [Feature][KV Offload] Enable canonical cross-TP offload for GDN hybrids
- Docs
- Python not yet supported