Faceswap Forum

NB: This guide was correct at time of writing, but things do change. I will try to keep it updated. Last edited: 28 September 2023 Contents --- Introduction Settings Alignments Preview Tool Settings Menu Convert Introduction --- If you've landed on this page and you don't know how to extract a face ...

I can only assume that this is hardware related, as this is not a widely reported issue.

There is a missing DLL file on your system. Unfortunately, I won't know what it is, but I would guess it is fairly fundamental if win32console is unable to import.

I can't really give you any more help than that, other than to try to google the error.

Reading through this thread should help:
viewtopic.php?f=7&t=728

1) There are some minor issues with the stats optimizations I implemented (I did not test as thoroughly as perhaps I should have). I am aware of these (albeit, not your particular "stops graphing" issue) and will look to fix. It is merely a display issue, not a training issue. 2) I have no...

They are not pre-trained.

There are various options for initialization (ICNR and Convolutional-Aware) which can be used, otherwise the are generally initialized with He Uniform distribution.

Yes and no. It's more about model complexity. More complex models require more VRAM.

It's a long time since I've used that model, but I seem to remember I got good results from it.

I would put Villain at the top of the list, as long as you're prepares to wait a week or so for it to train.

You're connecting to a headless server which doesn't have an X-Server installed.

It's outside the scope of Faceswap to go into configuration for this, but look into vnc4server for Ubuntu. You should find plenty of guides.

I agree.... To be honest, the whole of sort needs an overhaul, but it is not a priority for me, as (imho) there is only one sort method worth using and that is "face", which works incredibly well. Blurry images are required, as often there are blurry faces that need to be swapped so the AI...

Firstly make sure the file C:\Users\Gaël\Documents\faceswap-master\plugins\extract\detect\.cache\resnet_ssd_v1.prototext exists.

If it does, it may be that the download corrupted. If so, delete the file and try again.

You get speed up by increasing your batchsize. The same batchsize on a single GPU or multi GPUs is likely to run at about the same speed, or maybe slightly slower. EDIT Misread your message. I don't have a multi-gpu setup, so can't compare, and don't know where the "sweet-spot" is. Hopeful...

Yes it's possible, but you'll need to do some coding, and unravel the process a little. Ultimately, the Convert script is responsible for this. You would need to look to intercept the part where it runs the prediction through the model, and replaces the source face with the output from the model, an...

1660 will train faster, 1070 will train bigger models. It really depends where your priorities are. If it were me, and all things being equal, I'd go for the 1070, but it is subjective.

I'll take a look. Thanks. There are some bugs in the latest stats code.

None of the things you list should be causing issues, and afaik, there have been no recent updates to FS which should cause these issues, but I will double check.... Sometimes the thumbnails can go screwy, which can manifest itself in the way you describe, so try the regen-thumbs option when launchi...

This should be fixed in latest commit. Please update.

SAE with Multi-Scale output enabled?

Those are the outputs from the multi-scale output

Ok. my advice would be "don't worry about it too much"..... We don't save optimizer weights in the model (as it will double the size of the model file), so it effectively means the model "starts again" every time you stop and start it. However, as the weights have already been up...

When did you start training (what date approx.)?

Which model?

Did you change any of the loss multipliers for any of the sessions?

Faceswap Forum

Search found 2666 matches

[Guide] A Rough Guide to Converting

Re: Distributed with Dual 2060 supers

Re: ImportError: DLL load failed while importing win32console

Re: Converting outputting original.

Re: 3 small training abnormalities and questions.

Re: Autoencoder model weights

Re: GPU upgrade on a budge: 1070 8GB vs. 1660 6GB

Re: The realface model

Re: [Guide]Linux Install Guide: Installer Method

Re: Request: Blur sort that uses masks.

Re: cv2 fails to load prototext for extractor model

Re: Log and graph weirdness

Re: Use faceswap scripts to replace the same face (no swapping)?

Re: GPU upgrade on a budge: 1070 8GB vs. 1660 6GB

Re: Log and graph weirdness

Re: Manual Tool Issues since latest updates

Re: Conversion error involving multiple faces per frame

Re: Face A012 B012 ??

Re: Strange Training Graph

Re: Strange Training Graph