What does global or local means?

[Nise-2-meet-U]

self.parser.add_argument('--netG', type=str, default='global', help='selects model to use for netG')

[syfbme]

global means only G1 will be trained. You could refer to figure3 in the paper. G1 means global while G2 means local

[watershadow1212]

global means only G1 will be trained. You could refer to figure3 in the paper. G1 means global while G2 means local

so my bro， How do you do local training? Do you train globally first and then locally? My training instructions are: python train.py --name label2LWIR_214_mini_global_then_local --label_nc 3 --gpu_ids 0 --resize_or_crop none --dataroot D:\shuangguangxiagnmuyiqi_Dataset\tule --continue_train --lr 0.000010 I used annotated images for training, not real images，Unfortunately, my local training was also poor