lyndonzheng / Pluralistic-Inpainting

[CVPR 2019]: Pluralistic Image Completion
676 stars 145 forks source link

fine-tune #37

Closed XiaoYangon closed 3 years ago

XiaoYangon commented 4 years ago

I want to train my dataset using pre_trained model( imagenet_radom),How I change the code

lyndonzheng commented 4 years ago

@Purdandelion For fine-tune, you just need to add the --continue_train command in the training and add the pre-trained model in the checkpoint.