gladzhang/ART

Train code

kimtaehyeong opened this issue · 6 comments

kimtaehyeong commented

Hi @gladzhang,

Thanks for sharing and great work.

I want to custom train my data. When will the train code be completed?

gladzhang commented

Hi @kimtaehyeong

Thanks for your interest in our work. We will update the new version including training codes as soon as possible. The expected date is close to the end of November, after the deadline of CVPR submission. We will provide the complete version based on BasicSR so that everyone can clone the repository to proceed training and testing. The old version will be pushed to another branch.

Best wishes.

kimtaehyeong commented

Hi @gladzhang

I'll be waiting, Thank you so much!

yulunzhang commented

Hi @kimtaehyeong

Thanks for your interest in our work. The full version of our code is ready now.

If you have any other problem, please let us know. Thanks.

Best,
Yulun

kimtaehyeong commented

Hello @yulunzhang
Thanks for the quick work.
I want to train in my custom data, but is there anything I need to pay attention to or change?

Thank you.

yulunzhang commented

Hi @kimtaehyeong

For your own paired data, I think it's easy for you to use them. You can mainly follow a similar naming strategy as ours when preparing the data.

kimtaehyeong commented

Hello @yulunzhang
Let's try it.
Thank you for kindly letting us know.