Why serverless endpoint download sdxl1.0 from hugging Face hub so slow?
My project need download sdxl 1.0 form hugging face hub, about 7G,but after waiting 50minutes, job is still in queued, and log nothing. 48G GPU High Avaliability. Is there anything wrong ? detail see attachment.
5 Replies
Why are you downloading models from huggingface within your workers? You should either put them on network storage or bake them into your docker image.
I just wanted to see result as soon as possible. Now I see, I will try image again.
You will see the result faster if your worker doesn't need to download any models first.
The Dockerfile generated by runpodctl project build includes a base image specified as FROM runpod/base:0.6.1-cuda11.8.0. During the image build process, content related to CUDA 12 is being downloaded , why ?
Its normal, ignore it