How to apply a simple model's parameters to a complicated model(two different model))

1

I have written a simple model(only the horizontal part of above model) and trained the parameters. Now, I want to apply the simple model’s parameters to the above model(a different one). The complicated model can’t inherit from the simple one(or use simple one as a part), So,is there any way to do this work?

2

This will only work if each of the horizontal layers in your diagram are identical.

If so then you can save the trained model parameters:

torch.save(simple_model.state_dict(), "path/to/file.pth")

and then load them into each of the horizontal layers:

filepath = "path/to/file.pth"

complex_model = [network() for i in range(num_layers)]

for layer in complex_model:
    layer.load_state_dict( torch.load( filepath ))

where network is the model you originally trained.