RuntimeError: CUDA error: no kernel image is available for execution on the device

[mlmlh]

I followed the docker instructure and trying to run the zero123: python launch.py --config configs/zero123.yaml --train --gpu 0 data.image_path=./load/images/dog_rgba.png
and get this error...
I run this on an EC2 instance. Here are some more information:

• ubuntu 20.04
• Cuda version: 11.8
• Driver version:525.85.12
• Python 3.10.6
• torch: 2.0.1+cu118
• pytorch-lightning:2.0.6