pytorch: command not found

[kylanj7]

after setting up :~/Desktop/ztm-pytorch-course$ and installing pytorch 2.3.1, conda, and cuda 12.5.0 I am recieving this error. I tried installing the cuda toolkit inside of /Desktop/ztm-pytorch-course/env twice now and the error is still returning. Any advice?

after checking the env with this script, I get a tensor output but my GPU comes back as false.

import pandas as pd
import numpy as np
import torch
import sklearn
import matplotlib
import torchinfo, torchmetrics

Check PyTorch access (should print out a tensor)

print(torch.randn(3, 3))

Check for GPU (should return True)

print(torch.cuda.is_available())

[aldragan0]

So you mention that you've installed cuda toolkit but your gpu is not recognized in python.
Have you checked that you have the latest nvidia driver installed? If you open a terminal and type nvidia-smi what do you get?

Also, have you tried installing it the recommended way from the pytorch website?