Hello. I was fine tuning LayoutlmV3 and after the training, It has generated a .pth file. How can I implement this .pth model to predict the labels of the text in an image?
Could you describe how exactly this file was generated? I.e. does it contain a state_dict or did you save the model directly?