LibRecommender icon indicating copy to clipboard operation
LibRecommender copied to clipboard

New Model

Open Masoumeh-Alinia opened this issue 1 year ago • 1 comments

Hello, first of all thanks for your great work with the library! I had a question. I want to use the library's defined model from LightGCN and add another series of layers and then train the model. It is possible? How can I do this because the input for the next layer must be tensor.

Masoumeh-Alinia avatar Mar 13 '24 07:03 Masoumeh-Alinia

Hi, you can inherit from the LightGCN class. Internally it uses a pytorch module self.torch_model to perform the forward propagation. So you can add layers after this module.

massquantity avatar Mar 14 '24 02:03 massquantity