CUDA_VISIBLE_DEVICES can not be found or is misspelled about thin-plate-spline-motion-model

[unknownentity123]

I always use "set CUDA_VISIBLE_DEVICES=1 & python......" when using local repos in an Anaconda env....the usage of the "&" was recommended on another repo and since using that I haven't had any issues

Where I use 1 for my second 1070ti card instead of my first card

[eizoxx]

Now is says:

Original:
Der Befehl "CUDA_VISIBLE_DEVICES" ist entweder falsch geschrieben oder
konnte nicht gefunden werden.
Traceback (most recent call last):
File "D:\AI\Thin-Plate-Spline-Motion-Model\demo.py", line 14, in
from modules.keypoint_detector import KPDetector
File "D:\AI\Thin-Plate-Spline-Motion-Model\modules\keypoint_detector.py", line 3, in
from torchvision import models
ModuleNotFoundError: No module named 'torchvision'

English:
The command "CUDA_VISIBLE_DEVICES" is either misspelled or could not be found
Traceback (most recent call last):
File "D:\AI\Thin-Plate-Spline-Motion-Model\demo.py", line 14, in
from modules.keypoint_detector import KPDetector
File "D:\AI\Thin-Plate-Spline-Motion-Model\modules\keypoint_detector.py", line 3, in
from torchvision import models
ModuleNotFoundError: No module named 'torchvision'

However, the first line is in German. But actually this should not make a difference, because all other commands had also worked

[yoyo-nb]

You got the error: ModuleNotFoundError: No module named 'torchvision'
You can follow the instructions on pytorch.org to install the appropriate version of torch and torchvision

[spider853]

General advices for Linux commands on Windows:
When you see something like XXX=NUMBER python ...
it doesn't work on Windows cmd
to make it work, separate the first part before python and set the enviornament variables manually
set CUDA_VISIBLE_DEVICES=0
and hit enter
then enter the second part from and including pyton ...