I’m currently running the TrOCR model on my Jetson Nano, but I’ve encountered an issue with inference times. After some investigation, I discovered that CUDA is not available on my Jetson Nano, as indicated by the following command returning False
:
import torch
print(torch.cuda.is_available())
I attempted to install CUDA by using Jetpack:
$ sudo apt-get install nvidia-jetpack
However, the installation failed due to space constraints on the device. To resolve this, I need to expand my storage. Could anyone please confirm if it’s possible to use the SATA cables provided with the A205 carrier board alongside the Jetson Nano eMMC module?
Thank you.