About the pre-trained model on 256x256 places2?

[tanbuzheng]

Hello, your work is very amazing! We hope to compare with this work in our work. However, due to the limited computing resources, we can only implement it on 256x256 resolution. So, I have two questions about the provided pre-trained models.
Firstly, is the pre-trained model of MAE you provided trained on 256x256 resolution?
Secondly, I found the provided pre-trained model of FAR is on 512x512 resolution, it is can be directly used to inpaint the 256x256 corrupted images? If not, would you like to provide the pre-trained models on places2 with 256x256 resolution?

Looking forward to your reply！