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]: Clarification on Base Model Inference Count with Multiple LoRA Models in vLLM Deployment
- [Feature]: Allow partial context in speculative decoding when using draft models with smaller context than target model
- [Usage]: KV cache memory utilization tracking
- [Performance]: reproducing vLLM performance benchmark
- [Bug]: watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
- [Bug]: AssertionError: Logits Processors are not supported in multi-step decoding
- [Usage]: Q1why the docker image is empty? Q2 do I have to input the hugging face token? Q3 any python script to start vllm openai servering with codellama(llama2)?
- [Bug]: (OOM) Find two places that cause a significant increase in GPU memory usage (probably lead to memory leak)
- [Misc]: HELPPP! Implement vLLM Library in FastAPI using MultiGPUS got Force Shutdown after some warning
- [Usage]: What's the minimum VRAM needed to use entire context length for Llama 3.1 70B and 405B
- Docs
- Python not yet supported