imagebart

imagebart copied to clipboard

The pretrained model was not found on this server

1

Could you please check the pretrained model: ffhq_4_scales_geometric in your server? It is not avaiable now. Thanks.

KangLiao929

questions with inpainting input in inpaint_imagebart.py

1

 Should the input of first stage's vqgan encoder be the masked input? Because there should be no unmasked image in inference.

mlyarthur

Awesome job on the code release; will the text-conditioned model be public?

1

Hi, awesome work on the paper and code, just a question: Will the text-conditioned model be available for download sometime in the future? 

johnpaulbin

Thanks for the awesome work! Really amazing generation results! It seems some related works like [1,2], which employs bi-directional transformer for generation, are not mentioned or compared in the paper....

fxyang9