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
- [Feature]: SM8x (Ampere A100/A800) support for DeepSeek-V4.1-Flash
- [Bugfix][Attention] Zero the null KV block after CUDA graph capture
- [Bug]: OffloadingConnector KV events advertise block hashes that are not independently retrievable
- [Bugfix][Quantization] Stop sleep(level=2) from zeroing compressed-tensors KV scales
- [Feature]: lora support for deepseek v4.1 flash
- [Bug]: Multi-turn benchmark drops unfinished active conversations when the task queue is exhausted
- [RFC]: Retain MRV2 DBO CUDA Graph replay under DP imbalance through real-token staging
- [Feature][Core] Warn when estimated decode coverage exceeds CUDA graph capture size
- [Rust Frontend] Support chat completion output logging
- Cap multimodal content-part uuid length (cache_salt parity)
- Docs
- Python not yet supported