If you want to see if it’s using the GPU, you can go to the Jupyter notebook homepage. In the upper right, there should be a dropdown where you can select a Terminal. The terminal will open up in a new tab and from there you can treat it like any other terminal.
In your case, when the model is training, go over to the terminal tab and run an nvidia-smi command - see this fastai thread for a few different options if you haven’t used it before