Deprecate reparameterization module (#1316)

apex/reparameterization/__init__.py

+import warnings
+
 from .weight_norm import WeightNorm
 from .reparameterization import Reparameterization
 
+
 def apply_weight_norm(module, name='', dim=0, hook_child=True):
     r"""
     Applies weight normalization to a parameter in the given module.
@@ -44,6 +47,7 @@ def apply_weight_norm(module, name='', dim=0, hook_child=True):
         torch.Size([40, 20])
 
     """
+    warnings.warn("`apply_weight_norm` will be removed by the end of June, 2022.", FutureWarning)
     return apply_reparameterization(module, reparameterization=WeightNorm, hook_child=hook_child,
                                     name=name, dim=dim)
 
@@ -58,6 +62,7 @@ def remove_weight_norm(module, name='', remove_all=False):
         >>> m = apply_weight_norm(nn.Linear(20, 40))
         >>> remove_weight_norm(m)
     """
+    warnings.warn("`remove_weight_norm` will be removed by the end of June, 2022.", FutureWarning)
     return remove_reparameterization(module, reparameterization=WeightNorm,
                                     name=name, remove_all=remove_all)
 
@@ -84,6 +89,10 @@ def apply_reparameterization(module, reparameterization=None, name='', dim=0, ho
         Linear (20 -> 40)
 
     """
+    warnings.warn(
+        "`apply_reparameterization` will be removed by the end of June, 2022.",
+        FutureWarning,
+    )
     assert reparameterization is not None
     if name != '':
         Reparameterization.apply(module, name, dim, reparameterization, hook_child)
@@ -107,6 +116,10 @@ def remove_reparameterization(module, reparameterization=Reparameterization,
         >>> m = apply_reparameterization(nn.Linear(20, 40),WeightNorm)
         >>> remove_reparameterization(m)
     """
+    warnings.warn(
+        "`remove_reparameterization` will be removed by the end of June, 2022.",
+        FutureWarning,
+    )
     if name != '' or remove_all:
         to_remove = []
         for k, hook in module._forward_pre_hooks.items():

apex/reparameterization/weight_norm.py

+import warnings
+
 import torch
 from torch.nn.parameter import Parameter
 from ..fp16_utils import Fused_Weight_Norm