Added number of features in FrozenBatchNorm2d __repr__ (#2168) (#2431)

Summary:
* feat: Added number of features in FrozenBatchNorm2d repr

While BatchNorm layers have extensive information in their repr, FrozenBatchNorm2d has one

* refactor: Refactored FrozenBatchNorm2d __repr__

* test: Added unittest for FrozenBatchNorm2d __repr__

* style: Removed blank lines in test_ops

* refactor: Avoids creating an extra attribute for __repr__

* style: Switched __repr__ to f-string

Since support of Python version ealier than 3.6 have been dropped, f-string can be used.

* fix: Fixed typo in __repr__

* style: Switched unittest .format to f-string

Pull Request resolved: #2431

Reviewed By: zhangguanheng66

Differential Revision: D22437793

Pulled By: fmassa

fbshipit-source-id: 76fd6a42f0b6cabe47c96fcb5a343fe580a0005e

test/test_ops.py

@@ -530,5 +530,15 @@ def script_func(x_, offset_, weight_, bias_, stride_, pad_, dilation_):
                   (x, offset, weight, bias), nondet_tol=1e-5)
 
 
+class FrozenBNTester(unittest.TestCase):
+    def test_frozenbatchnorm2d_repr(self):
+        num_features = 32
+        t = ops.misc.FrozenBatchNorm2d(num_features)
+
+        # Check integrity of object __repr__ attribute
+        expected_string = f"FrozenBatchNorm2d({num_features})"
+        self.assertEqual(t.__repr__(), expected_string)
+
+
 if __name__ == '__main__':
     unittest.main()

torchvision/ops/misc.py

@@ -177,3 +177,6 @@ def forward(self, x):
         scale = w * rv.rsqrt()
         bias = b - rm * scale
         return x * scale + bias
+
+    def __repr__(self):
+        return f"{self.__class__.__name__}({self.weight.shape[0]})"