Problems encountered when using self-trained models for inference

[liuyongkang1-tal]

I set the training parameters image resolution to 448, up_scale to 1, and train my own model. I encountered the following problem when using the model I trained for inference. I checked that the input parameters of the network are all consistent in size. I debugged it several times and found no problem. The error content is “forward() missing 1 required positional argument: 'emb'”