Menu Close
Zsnack
vreP
1235
#30084Posted 11 months ago

finally had the time to get around to this since users have reported success with RTX, but when installing faceswap with python I’m getting an error that says that cuDNN is not found:

I’m using CUDA 10.0 and cuDNN 7.3 as recommended by another RTX user. I’ve extracted the 3 cudnn files into bin, include, and lib/x64, and I made sure that the environment variables are pointing to the right place. Any ideas?

2 Replies

kpopdeepfakes
kpopdeepfakes
Jennie's Master
Creator
#30085Posted 11 months ago

A user from the repo has recommended to edit the code for the cudnn check method in setup.py here:

def cudnn_check(self):
        """ Check Linux or Windows cuDNN Version from cudnn.h """
        if self.env.os_version[0] == "Linux":
            cudnn_checkfiles = self.cudnn_checkfiles_linux()
        elif self.env.os_version[0] == "Windows":
             if self.env.cuda_path == "":
                 cuda_keys = [key
                         for key in os.environ.keys()
                         if key.lower().startswith("cuda_path_v")]
                 self.env.cuda_path = os.environ[cuda_keys[0]]
        cudnn_checkfiles = self.cudnn_checkfiles_windows()
Zsnack
vreP
1235
#30086Posted 11 months ago

The new method worked! But now I’m running into another problem; when I try to run the line to extract a faceset, it tells me that Tensorflow is not installed on my system even though I used “pip install tensorflow-gpu==1.12.0” to install it.

When I run the line again (the tensorflow installation one), it shows “Requirement already satisfied: tensorflow-gpu==1.12.0 in c:\users\user\appdata\local\programs\python\python36\lib\site-packages (1.12.0)”

You must log in to reply to this post

Disclaimer:

You must be 18+ years of age or the legal age in your area to view pornographic materials. Videos displayed on this site are deepfakes. Deepfakes are produced with software using machine learning. All models and faces of celebrities or idols used in our videos are at least 18+ years of age. These deepfakes are created for entertainment and learning purposes only.