先G后D 训练 G: ——S:pred1, pred2 = model(images) ; loss_seg1 = loss_calc(pred1, labels) ; loss.backward() ; ——T:pred_target1, pred_target2 = model(images) Read More
posted @ 2022-05-01 20:49 ethan178 Views(294) Comments(0) Diggs(0) Edit