About DRCT-L x2 model

[Yanruyu-Beta]

Thank you for your great work. I followed the paper's setup using the ImageNet pre-trained DRCT-L x2 model and fine-tuned it with the DF2K dataset. However, I couldn't reproduce the results stated in the paper. My reproduced results are as follows: SET5 PSNR: 38.58, SSIM: 0.9635; SET14 PSNR: 34.96, SSIM: 0.9277.