search

facebookresearch / VMZ

VMZ: Model Zoo for Video Modeling
Apache License 2.0
1.04k stars 156 forks source link

torchvision video_reader video backend is not available #130

Open richardkxu opened 3 years ago

richardkxu commented 3 years ago

Hi,

Thanks for making this open source! I have followed the pytorch installation steps here: https://github.com/facebookresearch/VMZ/blob/master/pt/INSTALL.md. However, I have encountered the following error when I tried to do training on ucf 101 or simply load " torchvision.set_video_backend('video_reader')":

/home/richardkxu/anaconda3/envs/csn/lib/python3.7/site-packages/torchvision/__init__.py:64: UserWarning: video_reader video backend is not available
  warnings.warn("video_reader video backend is not available")

av                        7.0.1            py37h82f89c2_2    conda-forge
ffmpeg                    4.2.2                h20bf706_0  
torch                     1.4.0                    pypi_0    pypi
torchvision               0.5.0                    pypi_0    pypi
cudatoolkit               10.2.89              hfd86e86_1

I have CUDA=10.2 on my system. I am wondering how to solve the above error? Do I have to use torchvision==0.5.0 or newer version is fine? Do I need to compile torchvision from source (but usually av + torchvision should work)? Thank you!