Pretrained model for VQGAN

[choyingw]

Thank you for the implementation! I would like to tune the second-stage transformer, but my VQGAN trained on Flickr landscape dataset is not so good. I see there is a load function of 'vq_flickr.pt' and you have much better landscape results. Could you kindly share that? Thanks!

[dome272]

I dont have a pretrained flickr lying around anymore, but you may wanna take a look at concurrent work for the paper Make-A-Scene where we have released some ckpt. In the colab I also use rudalle and taming-transformer. All of these are general purpose models and should work pretty good for flickr. https://colab.research.google.com/drive/1SPyQ-epTsAOAu8BEohUokN4-b5RM_TnE?usp=sharing

[choyingw]

Thank you for the quick response!