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
- [Bug]: topk=1 and temperature=0 cause different output in vllm
- [Bugfix]Fix evict v2 with long context length
- hidden-states from final (or middle layers)
- [Performance]: Qwen2-72B-Instruction-GPTQ-Int4 Openai Server Request Problem
- [Usage]: How do you specify a specific branch on huggingface to use when downloading a model?
- [Bug]: 8 GPU setup - vLLM can only start with --tensor-parallel-size=2 but not 4 or 8
- [Bug]: RuntimeError: CUDA error: an illegal memory access was encountered
- [Bug]: Multi GPU setup for VLLM in Openshift still does not work
- [Bugfix] Take the VRAM usage of prompt_logprobs into account
- [Usage]: Function calling for mistral v0.3
- Docs
- Python not yet supported