How do I know the network architecture of the model?

[Alex-fishred]

How do I know the network architecture of the model? I try to read the pth model with netron I have tried torch.save to store the whole model instead of the model parameters only but can't get the answer I want

I want to get the input and output dimensions of each layer, of course I can get it from the program code, but I don't believe my observation.

[Cogito2012]

You can debug the code line by line, and you can print all the input / output dimensions at each layer, as well as the complete network arch.