You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Running FastGan with latest CUDA and torchvision required for 3090x series card gives the following warnings.
/home/ml1/anaconda3/lib/python3.9/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and will be removed in 0.15, please use 'weights' instead.
/home/ml1/anaconda3/lib/python3.9/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
I would switch to weights=VGG16_Weights.DEFAULT for latest version or to weights=VGG16_Weights.IMAGENET1K_V1 for a stable fixed version.
The text was updated successfully, but these errors were encountered:
Running FastGan with latest CUDA and torchvision required for 3090x series card gives the following warnings.
I would switch to weights=VGG16_Weights.DEFAULT for latest version or to weights=VGG16_Weights.IMAGENET1K_V1 for a stable fixed version.
The text was updated successfully, but these errors were encountered: