Question about code and results

[intyeger]

Hello,
Thank you for your great work.

I have a question regarding the implementation.
In the released code, limit_train_batches=0.2 is used during training.
I was wondering whether the performance reported in the paper and the PSNR: 29.15 stated in the README were obtained using this setting.

Thank you in advance.

[Alone-gao]

I noticed that you have also been following Teacher Anpei's project recently and have any questions to consult. I tried to directly evaluate lara's results, but the rendered image was not very ideal. May I ask if you have encountered such problems?

[apchenstu]

Hello, Thank you for your great work.

I have a question regarding the implementation. In the released code, limit_train_batches=0.2 is used during training. I was wondering whether the performance reported in the paper and the PSNR: 29.15 stated in the README were obtained using this setting.

Thank you in advance.

Yes, I used this setting to limit the number of samples per epoch

[intyeger]

I noticed that you have also been following Teacher Anpei's project recently and have any questions to consult. I tried to directly evaluate lara's results, but the rendered image was not very ideal. May I ask if you have encountered such problems?

When trained for 2 epochs, the PSNR was around 23~24.
I think that training for all 30 epochs will achieve the performance reported in the paper.