Faceswap Forum

Hi

which way is better for training:

-train a model with different sets( images/vids data).

-put all the the sets in one huge set and train from there.

is there any difference at all?

Extract images from multiple video sources in to one training folder. The training folder should contain several thousand images.