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 supported3 Subscribers
Add a CodeTriage badge to vllm
Help out
- Issues
- [Usage]: With the vLLM engine, the inference speed of the GPU A800 is not as fast as RTX3090 block, is this normal?
- [Frontend]: enable generator interface for offline inference
- [Performance]: Sampler account for most of time comparing to prefill and decode
- [Usage]: Running Phi3.5 on Intel x86 MacBook Pro?
- Prototyping `LLMEngineCore`
- [Bug]: Nonsense output for Qwen2.5 72B after upgrading to latest vllm 0.6.3.post1 [REPROs]
- [Bug]: v6.3 Gibberish produced with long ctx (Machete + W4A16 + Spec Decoding)
- [Feature]: Support a `flush_cache` API to clean the kvcache after `load_weights`
- [Usage]: Can I get the loss of model directly?
- [Misc]: Unable to load Llama 3B model in A10 GPU
- Docs
- Python not yet supported