Thank you for the excellent paper. I was wondering if it would be possible for you to share the .pth file of the model trained only up to the L1 training stage via the model zoo. I would like to continue training it to fit other datasets. Additionally, I’m curious to know whether the model performs well when trained from scratch on other image datasets. Lastly, for image super-resolution quality, which approach do you think works best among GRL, DAT, and RRDB?
Thank you for the excellent paper. I was wondering if it would be possible for you to share the .pth file of the model trained only up to the L1 training stage via the model zoo. I would like to continue training it to fit other datasets. Additionally, I’m curious to know whether the model performs well when trained from scratch on other image datasets. Lastly, for image super-resolution quality, which approach do you think works best among GRL, DAT, and RRDB?