I was trying to run a GPU based code. I followed the PyTorch official website to install PyTorch using ” conda install pytorch torchvision pytorch-cuda=11.6 -c pytorch -c nvidia”
Not only is a CPU-only binary installed but also an older one (1.10.2 while the current stable release is 1.13.0 as shown in your screenshot).
Uninstall all older PyTorch binaries and builds, and reinstall the new desired one. Creating a new and clean virtual conda environment might also help.