Qcompiler/vllm-mixed-precision

Support mixed-precsion inference with vllm

Python