Problems with my Project

Deep Learning with PyTorch: Zero to GANs Lecture 4 - Image Classification with Convolutiona
#1

Would anyone be willing to take a look at my project and help me figure out what I’m doing wrong? For some reason my initial accuracy is always 0.0, my graphs don’t seem to be coming up correctly and I’m getting dim errors when I try to test individual images.
Here is my notebook - a work in progress.

UPDATE
Well I managed to figure out that the dim errors I was getting on individual test images, was because I forgot to resize them to the same size as the training and validation images. I’m still a little unsure on the initial accuracy always showing 0.0, that just seems off to me. I’m still having issues with the graphs as well, specifically my accuracy and validation graphs. I’m only showing one data point each time I plot.

#3

I’m trying to understand how it even works, because you have only 2 classes, and your model has 10 outputs.

I suppose you have leftover cells from previous executions, back when you’ve started with CIFAR10 and tried to adjust it.

I suggest restarting the notebook, and running it again to make sure there is no leftover state from previous runs. Also changing the number of outputs to 2 should help - number of outputs of the model has to be equal (usually) to the number of classes you try to classify.

#4

Thank you for the help. I completely overlooked that I hadn’t changed the number of outputs in my model. That seems to have fixed my initial accuracy being 0.0 every time.