I trained a model using villain using 85% coverage. After 250k iterations I did a test conversion and the results are reasonable, but there is a shadow on the temple/side of the forehead. I was thinking I was too aggressive with the coverage in the model, so is it possible to reduce coverage mid training, or is the coverage locked from the start and I need to train again from scratch?
Changing coverage mid training
Forum rules
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.
Re: Changing coverage mid training
Coverage is locked, as it basically dictates how much of the image is fed into the model. Changing coverage mid-train would basically destroy your model, so you will need to start again.
My word is final