Fix detector build issues

Summary: Made changes to make some previous components compatible with vizard. - Modified rotated box to accept additional arguments to support forward compatibility with vizard (arguments are unused) - Created new config file for testing, don't need to change and revert changes golden config file anymore during devloop - Added registering my test train datasets to setup function - Added imports to load components into registries - **[IMPORTANT TO REVIEW] Modified serializer class for datasets (no longer uses SharedList, read more on what SharedList does here: https://fburl.com/code/23u0ci1z)** Reviewed By: srihari2761 Differential Revision: D64119684 fbshipit-source-id: 662cf1a271cd5db6686c03d6ed93dd85e5287f9a

Author: Leon Dong

detectron2/structures/rotated_boxes.py

@@ -1,6 +1,6 @@
 # Copyright (c) Facebook, Inc. and its affiliates.
 import math
-from typing import List, Tuple
+from typing import Any, List, Tuple
 import torch
 
 from detectron2.layers.rotated_boxes import pairwise_iou_rotated
@@ -229,7 +229,7 @@ def clone(self) -> "RotatedBoxes":
  """
  return RotatedBoxes(self.tensor.clone())
 
- def to(self, device: torch.device):
+ def to(self, device: torch.device, *args: Any, **kwargs: Any):
  # Boxes are assumed float32 and does not support to(dtype)
  return RotatedBoxes(self.tensor.to(device=device))