Closed DickHanselRyan closed 9 months ago
This seems to be the same issue as #51 and it's related to multiprocessing issues with Windows. Follow this tutorial or disable multiprocessing in the DataLoader by setting num_workers=0
.
Hi, I've set the num_workers=0 and it works, but now I am facing another issue as shown on the screenshot.
It seems to be the exact same issue but with train_loader
instead of test_loader
. Did you set num_workers=0
for both?
Thanks, I forgot to set the other num_workers=0 and got the same problem. It's fixed now. Thanks for the help.
Problem I was trying the first few shot learning classifier by re-writing the whole code onto my PC, and when I got into the part of plotting the images, my VSCode showed an issue with Pickle.
Considered solutions I can run it on Google Colab, but I want to try and run it on my local environment. I use Python 3.10.13 with the latest EasyFSL version, torch=2.2 and cuda=12.1