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
- [Performance]: large rate of decrease in generation throughput when SamplingParams.logprobs increases
- [Feature]: Supporting a version of Consistency LLM
- [Bug]: export failed when kv cache fp8 quantizing Qwen1.5-72B-Chat-GPTQ-Int4
- [Bug]: why the logits is different between 0.4.1 and 0.4.2
- [Bug]: squeezeLLM with sparse could not work.
- [RFC]: Support specifying quant_config details in the LLM or Server entrypoints
- [Usage]: Vllm AutoAWQ with 4-GPU doesnt utilize GPU
- [Usage]: How to change the batch size when testing the throughput of VLLM by running benchmark_throughput
- [Bug]: RAM OOM Error Loading 480GB MoE Model Despite Fix in PR #1395
- [Performance]: Qwen 7b chat model, under 128 concurrency, the CPU utilization rate is 100%, and the GPU SM utilization rate is only about 60%-75%. Is it a CPU bottleneck?
- Docs
- Python not yet supported