DeepFaceLab_Linux
DeepFaceLab_Linux copied to clipboard
runing 7_merge_SAEHD, the model didn't work
Runing 7_merge_SAEHD, turns out like training a model, have to setting lots of thing, is this right?
Preview of result face is fine in the 6_train_SAEHD training process, but when i ran 7_merge_SAEHD the dst face didn't show correctly in the interactive merger previewer, looks like a bad image on the face,.
Is it because the model or I did wrong somewhere?