Thank you very much for this outstanding work.
However, I have encountered an issue where the test metrics on the ImageNet dataset are significantly lower than those reported in your paper when using the weights you provided.
This shows the metrics I obtained from testing on a set of 5000 images from the ImageNet dataset.
fid: 23.56819492, colorfulness: 29.00809186, psnr: 25.03597266, lpips: 0.181861103, ssim: 0.999294341.
I followed the method suggested for supplying the prompt information. And I also tried providing larger hint points. for example,
Could you please offer some guidance or insights regarding this discrepancy?
Thank you so much for your help.
Thank you very much for this outstanding work.
However, I have encountered an issue where the test metrics on the ImageNet dataset are significantly lower than those reported in your paper when using the weights you provided.
This shows the metrics I obtained from testing on a set of 5000 images from the ImageNet dataset.
fid: 23.56819492, colorfulness: 29.00809186, psnr: 25.03597266, lpips: 0.181861103, ssim: 0.999294341.
I followed the method suggested for supplying the prompt information. And I also tried providing larger hint points. for example,
Could you please offer some guidance or insights regarding this discrepancy?
Thank you so much for your help.