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 supported2 Subscribers
Add a CodeTriage badge to vllm
Help out
- Issues
- [Kernel] Add Fused Layernorm + Asymmetric int8 Quant
- [Bug] [ROCm]: ROCm fails to stop generating tokens on multiple GPTQ models
- [Usage]: I updated VLLM to the latest one and I discover that when I launch serve, I cann't see the ouput of the special token like <bos>. How can I get them?
- [Doc] Proofreading documentation
- [Draft][MISC] Use torch.frombuffer(array(list)) in prepare_input
- [Performance]: In tokens processing performance. Bad scaling with tensor-parallel
- [Usage]: how to abort request and stop inference?
- [Bug]: subprocess.CalledProcessError: Command '['cmake', '--build', '.', '-j=40', '--target=_moe_C', '--target=_C']' returned non-zero exit status 1.
- [Feature]: Is it possible to control whether to use speculative decoding when making a request?
- [RFC]: More rigorous compilation warnings
- Docs
- Python not yet supported