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 supported21 Subscribers
Add a CodeTriage badge to vllm
Help out
- Issues
- [Bug]: kimi-k2-thinking on AMD MI300X not working
- [DO NOT LAND] Prototype Helion kernel in vLLM
- [CI][RL]: add end-to-end Verl+vLLM FSDP PPO test script for GSM8K fine-tuning
- [Bug]: Qwen3-Next-80B-A3B-Instruct-FP8 crashes with CUDA illegal memory access on vLLM 0.11.2 (even with cudagraph_mode=PIECEWISE)
- [PCP] PCP support ray
- Generalize pooling model support with multi-task, multi-layer, multi-label classification that can be pooled from both hidden states and LM head's logits.
- [CI] Speed up fusion tests
- [core] [DO NOT REVIEW] Enabling Zero-Copy Video with PyNvVideoCodec and IPC
- [Frontend] Add `reasoning_effort` and `parallel_tool_calls` to `extra_args` of `SamplingParams`.
- [Core] Optimize top-k + top-p sampling by avoiding full vocabulary sort
- Docs
- Python not yet supported