Faceswap Forum

I'm not entirely sure I understand what you mean.

You can just point a new model at the same B folder you used before and a new A folder for the new identity.

No. The values are just summed for the final loss figure.

Hello and welcome to the forum

Grab yourself a coffee and have look around.

Honestly, I'm not sure what to suggest for this. I have heard of one other person having this issue, but it seems spotty at best.

Maybe try enabling/disabling conv-aware/icnr inititialization and restarting the model? Would be interested to hear if this makes any difference.

This looks like a Cuda version conflict. In the first instance do this (do not skip any steps):
app.php/faqpage#f1r1

If that does not solve your issue, post back following these instructions:
app.php/rules#rule-4b

GTX1650 is listed there under compute capability 7.5

I don't think its as simple as that, so know I don't think your missing anything. My main takeaway (iirc) was that Lr needs to be lowered if BS is lowered. I didn't really look too closely into ratios/equations etc, as I would imagine that this would vary wildly on a case by case basis.

I would say that ms-ssim and ssim used together would probably be redundant.

I am (in a future update) going to lower the strength of the lpips output to allow more fine-grained control (as it is not possible to go below 1 for the weight), but need to think how best to do this

How are you checking model settings?

The specific model settings that are used for any model can be found in the state.json file located in the model folder.

No, they were trained with a mask.

To be honest, it has been a long time since I trained without a mask (back when Faceswap still only had 64px models), so it may be worth a test to see how things perform.