EfficientNetV2 really is something. As advertised for sure.
If training is failing to start, and you are not receiving an error message telling you what to do, tell us about it here
This forum is for reporting errors with the Training process. If you want to get tips, or better understand the Training process, then you should look in the Training Discussion forum.
Please mark any answers that fixed your problems so others can find the solutions.
Yep, getting positive feedback for sure. It's good to know it was worth putting the time in to get the code updated
My word is final
Eh, all this sarcasm, I do get pretty great results in shorter time with it. So thanks @torzdf for adding it. Also thanks in general that you still try to improve everything here and there, even just a little bit. I will keep donating via Patreon for a while, even if you would stop immediately, because what you done so far was worth it.
Where was the sarcasm?
The EfficientNetV2 really is something and the results are fantastic. I would agree with the quoted 5x-11x learning rate improvements.
So the white box preview continues when starting brand new models.
However, I have found the strangest workaround.
Before starting a brand new model, I simply update Faceswap (Help-->Update Faceswap). There doesn't even need to be an update available, I just simply need to perform this action.
Once done, I then start a training session, and boom - immediate previews.
It's consistent too - works every time.
I have no idea what's going on or why this works but perhaps it's loading something from the codebase/files that weren't present/loaded before.
Just thought I'd mention it - I'm sure you'll make more sense of it than me
Guess again! That makes absolutely no sense to me whatsoever . But, if it works, it works.
My word is final