Search found 1423 matches

by torzdf
Sun Jul 25, 2021 10:15 pm
Forum: Extract Discussion
Topic: Feature Request: Make Use of NVIDIA Hardware Video Decoding/Encoding with FFmpeg
Replies: 6
Views: 458

Re: Feature Request: Make Use of NVIDIA Hardware Video Decoding/Encoding with FFmpeg

Python itself is single threaded, but it can (and we do) use multiprocessing libs.

As to FFMPEG, as we have to read the frames linearly, it may just not be possible to run it more optimally. To be honest, I have not dug into it in that much detail

by torzdf
Sun Jul 25, 2021 2:00 pm
Forum: Training Support
Topic: crash report while training: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize
Replies: 9
Views: 679

Re: crash report while training: Failed to get convolution algorithm. This is probably because cuDNN failed to initializ

Glad to hear it.

You shouldn't have any global Cuda installed though. Whilst it may work now, it will probably break again in future if/when we upgrade.

We install Cuda locally in the Faceswap environment. Any global install may conflict.

by torzdf
Sun Jul 25, 2021 11:55 am
Forum: Training Support
Topic: crash report while training: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize
Replies: 9
Views: 679

Re: crash report while training: Failed to get convolution algorithm. This is probably because cuDNN failed to initializ

You still have Cuda installed: gpu_cuda: 9.1 The version listed by nvidia-smi is the maximum version supported by your driver. The version that should be being used is the conda package cudatoolkit 10.1.243 h6bb024c_0 Having a system install + a conda install can conflict. Work on getting 9.1 fully ...
by torzdf
Sun Jul 25, 2021 11:25 am
Forum: Training Support
Topic: crash report while training: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize
Replies: 9
Views: 679

Re: crash report while training: Failed to get convolution algorithm. This is probably because cuDNN failed to initializ

The next thing to do is to uninstall the global Cuda 9.1 you have installed. It is incompatible.

Reboot and try again.

by torzdf
Sun Jul 25, 2021 10:39 am
Forum: Swap Showcase
Topic: Phaze-A - 256px Demo
Replies: 9
Views: 8262

Re: Phaze-A - 256px Demo

I'll add to the above, it's not just legacy faces. The input resolution to the model is separate from the size of the extracted faces Try not to think of the extracted faces as the training images, rather, they contain the training images. Whether you use Face or legacy centering, the actual images ...
by torzdf
Fri Jul 23, 2021 10:14 am
Forum: Extract Discussion
Topic: Feature Request: Make Use of NVIDIA Hardware Video Decoding/Encoding with FFmpeg
Replies: 6
Views: 458

Re: Feature Request: Make Use of NVIDIA Hardware Video Decoding/Encoding with FFmpeg

There is, but it means using the cli. One day I would like to add the possibility of batch processing to Faceswap, but see my earlier explanation of the massive to do list ;) I believe you are using Windows, so this wouldn't work for you (I use Linux), however a quick google around for Windows scrip...
by torzdf
Thu Jul 22, 2021 5:14 pm
Forum: Training Support
Topic: Crash while trying to train - OSError: Unable to open file (truncated file)
Replies: 3
Views: 2391

Re: Crash while trying to train - OSError: Unable to open file (truncated file)

I was going to say "please search prior to posting support requests". But.... This questions was already asked... by you... just a month ago.

by torzdf
Thu Jul 22, 2021 10:24 am
Forum: Convert Support
Topic: CPU temperature problem
Replies: 2
Views: 304

Re: CPU temperature problem

Converting is CPU intensive. That's just the way it is.

You can try limiting them number of jobs, or run in "single-process" mode within the settings section

by torzdf
Thu Jul 22, 2021 10:20 am
Forum: Training Support
Topic: Failure to create Tensorboard logging folder
Replies: 1
Views: 88

Re: Failure to create Tensorboard logging folder

This is a problem with creating the Tensorboard log files. It may be a permissions issue, or a path location issue, so check those first. Secondly, on Windows based machines it may be that it doesn't like the forward slash ' / ' in the path for the model folder. This may just be how the error report...
by torzdf
Thu Jul 22, 2021 10:13 am
Forum: Extract Discussion
Topic: How important is it for the jawline and eyebrows to be perfect, assuming other landmarks are good?
Replies: 1
Views: 141

Re: How important is it for the jawline and eyebrows to be perfect, assuming other landmarks are good?

If you are using a landmarks based mask (components/extended), then it matters a lot more, as the mask will be wrong if the landmarks are wrong. If you are not using a landmarks based mask, you can disregard the jaw entirely. Only the core area landmarks are used for alignment (i.e. all landmarks ex...
by torzdf
Tue Jul 20, 2021 9:53 am
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 109
Views: 57497

Re: Process Exited No Clear Error Message

I know it has to do with the card because when I try using the CPU I get a similar readout but it works instead of saying "process exited" Many thanks in advance Unfortunately that output doesn't show any indication of why it might be exiting. Can you post your system info (output from He...
by torzdf
Tue Jul 20, 2021 9:50 am
Forum: Extract Discussion
Topic: Feature Request: Make Use of NVIDIA Hardware Video Decoding/Encoding with FFmpeg
Replies: 6
Views: 458

Re: Feature Request: Make Use of NVIDIA Hardware Video Decoding/Encoding with FFmpeg

Thanks for the idea.... for the reasons you mention (the complication when competing with GPU resources), as well as the sheer size of my to do list, this isn't something I can see myself implementing any time soon. I believe ffmpeg is multithreaded by default (unless you pass a parameter limiting t...
by torzdf
Sun Jul 18, 2021 11:47 am
Forum: Training Support
Topic: CRITICAL Error caught! Exiting...ERROR Caught exception in thread: '_training_0'
Replies: 1
Views: 557

Re: CRITICAL Error caught! Exiting...ERROR Caught exception in thread: '_training_0'

That looks like a problem with (at least) one of your training images not being an extracted faceswap face

by torzdf
Fri Jul 16, 2021 10:01 am
Forum: Installation Support
Topic: [Guide] Using Faceswap on Nvidia RTX 30xx cards
Replies: 109
Views: 57497

Re: [Guide] Using Faceswap on Nvidia RTX 30xx cards

The solid colors mean model corruption. You can try lowering the learning rate. That preset is particularly complex, so can be susceptible to collapse at higher learning rate.

Also, if using mixed precision, I highly recommend setting the Epsilon Exponent to -5 for that preset.

by torzdf
Fri Jul 16, 2021 9:57 am
Forum: Extract Support
Topic: Extraction Crash
Replies: 1
Views: 1505

Re: Extraction Crash

This happens for 1 of a number of reasons....

1) There are special characters within the meta data of the video file. You can use something like MP3Tag

2) It's a false positive derived from some kind of memory error. If this happens, normally just running the process again remedies the problem.

by torzdf
Fri Jul 16, 2021 9:54 am
Forum: Cloud Discussion
Topic: Extraction on Google colab
Replies: 1
Views: 945

Re: Extraction on Google colab

It's possible to do any of the Faceswap processes on Google Colab. I don't use it, personally, so can't help you with how, but it shouldn't be too involved.

by torzdf
Tue Jul 13, 2021 12:01 am
Forum: Swap Showcase
Topic: Phaze-A - 256px Demo
Replies: 9
Views: 8262

Re: Phaze-A - 256px Demo

ianstephens wrote: Mon Jul 12, 2021 8:31 pm

Quick question @torzdf -

Did you use "no warp" at the end of training for this video? Or was it simply a run of 2M iterations with the StoJo preset model?

Yes. Can't remember when I disabled. Probably about halfway through

by torzdf
Tue Jul 13, 2021 12:01 am
Forum: Swap Showcase
Topic: Phaze-A - 256px Demo
Replies: 9
Views: 8262

Re: Phaze-A - 256px Demo

ianstephens wrote: Tue Jul 06, 2021 1:52 pm

Is this simple change below all it takes to train at 384px for example (or is there more to it)?

Short answer. You could just do that, but there's more to it.

Long answer: A lot of maths.... I will come back and articulate at some point in the (hopefully) not too distant future.

by torzdf
Mon Jul 12, 2021 11:48 am
Forum: Training Discussion
Topic: [Guide] Introducing - Phaze-A
Replies: 12
Views: 9275

Re: [Guide] Introducing - Phaze-A

That speed is about correct for StoJo. It is a heavy and complex model.