Closed thedirtyyugo closed 5 years ago
torzdf
commented
5 years ago 1) check https://github.com/deepfakes/faceswap/blob/master/INSTALL.md#windows-install-guide for detailed instructions on getting cuda/cudnn correctly installed 2) Use the latest commit and provide the generated crashlog if you are getting crashes 3) Python 3.7 is not supported by Tensorflow 4) GAN is unreliable at best, broken at worst. Not recommended. A new port is in the pipeline which may or may not be an improvement.
thedirtyyugo
commented
5 years ago I still cant convert videos. I use the original converter and installed everything from that link and my python still crashes. no error message pops up other that, couldn't create frames. what goes in output video, input video, and model. I feel as if im not accessing the right folders to convert the frames.
torzdf
commented
5 years ago Use the latest commit and provide the generated crashlog if you are getting crashes
CUDA/cuDNN version: cudnn-9.0-windows10-x64-v7.4.1. GPU model and memory:NVIDIA GeForce GTX 1070
issues 2018-12-01 22:15:38.357265: E C:\tf_jenkins\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\stream_executor\cuda\cuda_dnn.cc:378] Loaded runtime CuDNN library: 7401 (compatibility version 7400) but source was compiled with 7003 (compatibility version 7000). If using a binary install, upgrade your CuDNN library to match. If building from sources, make sure the library loaded at runtime matches a compatible version specified during compile configuration.
When trying to use GPU as my processor this is the error I receive. I understand that it says that I have the wrong cudnn file. I downloaded 7.0.4 and that didn't work and I downloaded version 6 but don't know how to upgrade to version 7. what can I do to be able to use GPU processor?
issues another issue is after finishing the training step using cpu I cant convert frames. it wont allow me to at all and shuts down python. why cant I create the frames on the original video?
I use GAN converter for both training and converting.