Consider saving input ARange with trained model

matlab-deep-learning / pix2pix

Image to Image Translation Using Generative Adversarial Networks

Other

30 stars 12 forks source link

Consider saving input ARange with trained model #9

Open justinpinkney opened 4 years ago

justinpinkney commented 4 years ago

An easy gotcha is to train a model with ARange set to a non-default value, then forget to set this same value when using p2p.translate (as mentioned in #4).

Saving this ARange value along with the saved model would probably help avoid this error.