How to run sample.py using multiple GPUs ?

[skittoo]

I am using 8 GPUs instance but the script uses only GPU:0 not all the GPUs . I get CUDA memory error when i try to make clip_guidance ofViT-L/14 model. How can i utilize all GPUs so i don't have any limitation ?

[limiteinductive]

You can modify the script to load the clip model on the GPU you want (and push back and forth all the tensors you need for inference)

[skittoo]

Thanks a lot for your response .
But it's also rises the same error when putting --cpu argument and changing the tensor inside the script is not - as i tried - only in sample.py , i should change things in other files . This is a little bit exhausting :)

Kind Regards ,