How to train for 20 epochs

switchablenorms / DeepFashion_Try_On

Official code for "Towards Photo-Realistic Virtual Try-On by Adaptively Generating↔Preserving Image Content"，CVPR‘20 https://arxiv.org/abs/2003.05863

826 stars 251 forks source link

How to train for 20 epochs #83

Closed Pritam-N closed 2 years ago

Pritam-N commented 3 years ago

The results are pretty good and the adaptive preservation/ generation module that you used is helping a lot. I am trying to train the model in atr dataset for my academic research. Can you please help me in understanding the training process. You have mentioned training each network separately. I am not able to completely understand how that will work. Also, can you let me know if it will be possible to train all the networks together.

LZQhardworker commented 3 years ago

Hi, You only need to download the dataset and the training code, and run python train.py Then everything works fine. The networks are trained respectively. (Gradient are not flowed end2end)