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
- Print request metrics to stdout
- Neuron cache blocks must be 1 more than max num seqs
- [Core][Kernel][Misc] Support external swapper for vllm
- [Usage]: how can I use vllm to generate logits instead of text
- [Bug]: name 'vllm_ops' is not defined using offline build from source vllm 0.4.3
- [Usage]: how to get GPU virtual block utilization log in LLM class(offline inference)
- [Bug]: Can't find safetensors model weights
- [Performance]: some questions regarding the performance degradation between asyncLLMEngine and the LLM class when TP = 4
- [Usage]: How to specify certain GPUs for Tensor Parallelism and Pipeline Parallelism
- [Bug]: Multistep with n>1 Fails
- Docs
- Python not yet supported