I’m working on the Week 3 Assignment (M2NIST segmentation) and ran into a problem when trying to submit my model. I’ve built the full model using a custom encoder and FCN-8 decoder — everything trains and runs fine. But when I save the model it ends up being 106 MB, and the upload always gets stuck the end. How do I fix this issue
Hi @Abd_12
I think the problem might be with your network connection, so try uploading multiple times using different networks and browsers, as that sometimes solves the hang. If it still fails, you might need to reduce the submission size (see if you can reduce the model size while maintaining the same accuracy—or slightly lower, but still acceptable).
Hope it helps! Feel free to ask if you need further assistance.
hi @Abd_12
did you try to make any changes in your model, to reduce the model file size?
can you share a screenshot where it provides this information because even 100mb also is pretty high. check your model algorithm or batch size.