Question:
Do you need more iterations to get a quality model if you use a higher number of faces?
Does a 1.000.000 Iterations with 2000 faces give the same result as 1.000.000 with 4000 ?
Just curious...
Read the FAQs and search the forum before posting a new topic.
This forum is for discussing tips and understanding the process involved with Training a Faceswap model.
If you have found a bug are having issues with the Training process not working, then you should post in the Training Support forum.
Please mark any answers that fixed your problems so others can find the solutions.
Question:
Do you need more iterations to get a quality model if you use a higher number of faces?
Does a 1.000.000 Iterations with 2000 faces give the same result as 1.000.000 with 4000 ?
Just curious...
If you add too many similar faces, yes it does take longer. It becomes harder for a batch to represent a significant amount of the variations in your dataset and causes it to overlearn the similar images. That can even double total training time. This is why we recommend using as much variety in generating your data as possible.