Custom_Net is one of my own network architecture.I trained this network in my own computer GPU. But in order to test it I just had to transfer to Google Colab, then I had to face the above Error.
I figured out the problem and I just got resolved it. This topic is for helping someone who is going to struggle with this error.
Since I have used 'torch.save(PATH)' when I'm saving my model after training, this problem has occurred. This code snippet just keeps path to class module, which is having the network architecture, not the class. That was the reason for the problem I faced.
So I trained my model again in Google Colab, then I got resolved from the problem. Since my network is small it got trained very quickly. Thanks.
Think this will help someone!!!!!!! :smiley: :smiley: :smiley: