pre_train question

[liyouxing]

hello,I'm a student, and i want to run the method and cite this paper,but the pre_train file decompress fail,can you reupload the file in github?thanks!

[liyouxing]

I solved this problem after debug half a day. The pre train file don't need to decompress, ".tar" only a name extension.
So you only torch.load this file, but you should add "encoding='ISO-8859-1'".
i.e. Base.py-->load_gan--> ckpt = torch.load(ckpt_path) change to ckpt = torch.load(ckpt_path,encoding='ISO-8859-1')