I wonder if anyone was successful to fully train the model. Based on my experience, It takes very long time to finish even for one epoch. I used my personal laptop using one GPU. I wonder if this is related with my code or with the data. I see the images are considerably have high resolution that might require longer time to process.