To train own model using author's model

Lotayou / Face-Renovation

Official repository of the paper "HiFaceGAN: Face Renovation via Collaborative Suppression and Replenishment".

https://arxiv.org/abs/2005.05005

GNU General Public License v3.0

284 stars 49 forks source link

To train own model using author's model #35

Closed WeibinDai closed 3 years ago

WeibinDai commented 3 years ago

First of all, thank you for your research and I learn much. Now, I want to use your model, 'latest_net_G.pth', as a pre-training model to train my own model with my own dataset, but I don't find anything like 'load_model(...)'. Is this interface provided in the source code? I'm looking forward to your reply. Thanks.

Lotayou commented 3 years ago

You can try specify continue_train = True in the config script, and delete the iter.txt file inside checkpoints/XXX/. This will clean the cache file for recording which epoch to resume training, and the program will automatically restart from epoch 1.