Loss difference between models

If training is failing to start, and you are not receiving an error message telling you what to do, tell us about it here


Forum rules

Read the FAQs and search the forum before posting a new topic.

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.

Locked
User avatar
Yarindadon
Posts: 1
Joined: Mon Mar 15, 2021 4:40 pm

Loss difference between models

Post by Yarindadon »

Hi guys I’m new here
I’ve been running faceswap on dfaker 256 and bs 4
Around 200 k iterations and the loss is about 0.024 for model a and 0.018 for model b, why is there such big difference?

User avatar
torzdf
Posts: 2649
Joined: Fri Jul 12, 2019 12:53 am
Answers: 159
Has thanked: 128 times
Been thanked: 622 times

Re: Loss difference between models

Post by torzdf »

Numerous reasons. Quality and variety of data, impact that the opposite side of the model is having on the weights.

Ultimately, this is quite normal.

My word is final

Locked