TFHubSample
TFHubSample copied to clipboard
Trained from scratch or on pretrained I3D?
Was the video classification example trained from scratch? Or were the weights from I3d module used?