vllm?

#10
by prudant - opened

Hi, question:

ocker run
--rm
--gpus all
--network host
ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddlex-genai-vllm-server

it uses a custom image of vllm or the model is supported by the latest official image of vllm?

NICE&GREAT JOB !!!!!!

PaddlePaddle org

Thank you for your question. As the official vLLM doesn’t support this yet, we are currently using a custom image that we built in-house to include the necessary components.

thanks, great JOB!

Sign up or log in to comment