Why can't we run vis_w.py without training first?

[ucalyptus2]

Isn't there a way to use pre-trained weights and do inference?

[KelestZ]

Isn't there a way to use pre-trained weights and do inference?

Thanks for your interest in our work! Could you please mention a bit more details? I'm assuming it's perhaps because vis_w.py does not find the opt.yml file. If so, once you get one opt.yml, you may want to change the attribute in it and load the corresponding pre-trained weights. ``train.py" could generate an opt.yml file automatically at the very beginning.

[ucalyptus2]

sure @KelestZ , I tried to save the yml file while a training process was ongoing. However I used it for inference later (by aborting the training), I came up with an list index error on line 76 of vis_w.py

will try to reproduce it

On Thu, 15 Jul 2021, 22:25 Peiye Zhuang, @.***> wrote:

Isn't there a way to use pre-trained weights and do inference?

Thanks for your interest in our work! Could you please mention a bit more details? I'm assuming it's perhaps because vis_w.py does not find the opt.yml file. If so, once you get one opt.yml, you may want to change the attribute in it and load the corresponding pre-trained weights. ``train.py" could generate an opt.yml file automatically at the very beginning.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/KelestZ/Latent2im/issues/3#issuecomment-880860915, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB5KHMKDFIDUCGFB2AIIX4DTX4HGNANCNFSM5AN4DKOQ .