Remove one TODO, the number of leading dims is correctly tested

Author: NicolasHug

torchvision/tv_tensors/_keypoints.py

@@ -8,9 +8,6 @@
 from ._tv_tensor import TVTensor
 
 
-# TODOKP do we support arbitrary number of leading dimension, as claimed in the
-# docs?
-# This doesn't seem to be tested?
 class KeyPoints(TVTensor):
     """:class:`torch.Tensor` subclass for tensors with shape ``[..., 2]`` that represent points in an image.