Faceswap Forum

This is the same issue as reported above, where I tested the exact settings used and couldn't replicate. I know of many who have updated and not hit this issue, which leads me to believe it is a false positive and, potentially, a system/vram issue. Unfortunately your output has not provided me with ...

ianstephens wrote: ↑Wed May 18, 2022 10:58 am
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.

The short answer is "a lot of work". Creating a bad swap is not hard. Creating a decent swap requires a fair amount of experience and a lot of mundane work cleaning up datasets etc. You may be better placed to ask this on our Discord server, as you will more likely get more eyes on it: htt...

Please help I am installing on Linux Mint 20.3 but this happens chupz22@KikoPredatorMint:~/Downloads$ bash faceswap_setup_x64.sh faceswap_setup_x64.sh: line 2: $'\r': command not found faceswap_setup_x64.sh: line 6: $'\r': command not found faceswap_setup_x64.sh: line 16: $'\r': command not found f...

Ok, sure, but that's not in the scope of what Faceswap does. AI requires processing power. A LOT of it. To add options into Faceswap to throttle hardware in some way would not be a trivial task, and I'm sure there are third party applications which allow you to limit CPU/GPU allocation per process, ...

They refer to the video that you plan to swap faces onto and the associated alignments file.

I may move this to training discussion, as these are generated for training previews... but I can see an argument for both locations :/ Also, probably worth adding (may need to check, but pretty sure this is correct), that if you're within the faceswap environment (i.e. conda activate faceswap ) the...

Ok, this one is going to be hard to debug, as I expect there are special characters in your path. In the first instance please try to do this: Start > Anaconda Prompt In the terminal that pops up: conda activate faceswap cd faceswap python -c "from lib.sysinfo import sysinfo ; print(sysinfo)&qu...

Yep, getting positive feedback for sure. It's good to know it was worth putting the time in to get the code updated

My mistake, I put the imports in the wrong way around. Should work now.

Please provide the following info to better diagnose your problem:
app.php/rules#rule-4b

Honestly, nothing I updated yesterday would have impacted this. The commits were basically a refactoring of some of the code so that we can (in the future) support Apple Silicon.

I guess it was just one of those things, but all's well that ends well!

This was down to an update I pushed today not supporting Python 3.7 (the Colab default). I have now fixed this, so hopefully the notebook should be working again now.

Honestly, my suspicion would be learning rate too. Why this happens on TF 2,8, I'm not sure. Just to rule something out, does changing the loss to MAE also cause this issue? I think it's highly unlikely, but as part of the TF 2.8 update I had to change the SSIM loss implementation. I actually change...

Yeah, this is a weird one, and it has not been reported elsewhere. You can install an earlier version of Tensorflow (the last version we used was 2.6) as Faceswap supports Tensorflow 2.2 - 2.8. The easiest way to do it (without messing up your conda environment) would be to do the following (not tes...

I honestly don't know then. There is nothing implicitly different about that model than any of the others which would cause this issue. The loss is calculated the same regardless of whichever model you use. The only other suggestions I have is to: 1) make sure that you have no overclocks (even stock...

Ok, well that's a different error, which leads me to believe that the initial error was a false positive. That error means you have run out of GPU memory. Try enabling "Mixed Precision" and/or lowering your batch size.

I have just tested training on latest faceswap with your exact setup (command line options + Model config) and I cannot recreate this bug. Unfortunately I cannot solve what I cannot recreate, which tends to suggest that the issue is with your setup somewhere. Not hugely helpful, I know. I know that ...

I have seen this before and I don't understand it. A print statement should not be raising an OS Error :/

I have literally never seen this before... looks like it's being generated from TF itself, which switching around in the GUI doesn't directly impact. Would be useful to get the full text of the error please.

Faceswap Forum

Search found 1675 matches

Re: Bug: ValueError: Please use `tf.keras.losses.Reduction.SUM` or `tf.keras.losses.Reduction.NONE`

Re: Bug: Updated to Latest faceswap version now crashing when starting training

Re: Open Discussion about Deepfakes in Phishing

Re: [Guide]Linux Install Guide: Installer Method

Re: Slider for how much gpu/cpu to use

Re: [Guide] A Rough Guide to Converting

Re: [Script] Create video from timelapse images

Re: I cant open Faceswap successfully ,UnicodeDecodeError

Re: Bug: Updated to Latest faceswap version now crashing when starting training

Re: [Resource] Google Colab Notebook

Re: Fresh Install - RTX 3060 Ti still Sleeping

Re: Bug: Updated to Latest faceswap version now crashing when starting training

Re: [Resource] Google Colab Notebook

Re: Bug: Updated to Latest faceswap version now crashing when starting training

Re: Bug: ValueError: Please use `tf.keras.losses.Reduction.SUM` or `tf.keras.losses.Reduction.NONE`

Re: Bug: ValueError: Please use `tf.keras.losses.Reduction.SUM` or `tf.keras.losses.Reduction.NONE`

Re: Bug: ValueError: Please use `tf.keras.losses.Reduction.SUM` or `tf.keras.losses.Reduction.NONE`

Re: Bug: ValueError: Please use `tf.keras.losses.Reduction.SUM` or `tf.keras.losses.Reduction.NONE`

Re: Bug: Updated to Latest faceswap version now crashing when starting training

Re: Bug: Updated to Latest faceswap version now crashing when starting training