Skip to content

Fix BatchNorm docstring to cover 4D input and test the NHWC path #3782

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
zcbenz merged 1 commit into from
Jul 1, 2026
+11 −1
Merged

Fix BatchNorm docstring to cover 4D input and test the NHWC path #3782

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion python/mlx/nn/layers/normalization.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -256,7 +256,7 @@ def __call__(self, x):


class BatchNorm(Module):
r"""Applies Batch Normalization over a 2D or 3D input.
r"""Applies Batch Normalization over a 2D, 3D or 4D input.

Computes

Expand Down
10 changes: 10 additions & 0 deletions python/tests/test_nn.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -770,6 +770,16 @@ def test_batch_norm(self):
self.assertIn("weight", bn_trainable)
self.assertIn("bias", bn_trainable)

# test with 4D input (NHWC)
mx.random.seed(42)
x = mx.random.normal((2, 3, 3, 6), dtype=mx.float32)
bn = nn.BatchNorm(num_features=6, affine=True)
y = bn(x)
self.assertTrue(x.shape == y.shape)
# batch norm over an NHWC input normalizes each channel across N, H, W
self.assertTrue(mx.allclose(y.mean(axis=(0, 1, 2)), mx.zeros((6,)), atol=1e-5))
self.assertTrue(mx.allclose(y.var(axis=(0, 1, 2)), mx.ones((6,)), atol=1e-2))

def test_batch_norm_stats(self):
batch_size = 2
num_features = 4
Expand Down
Loading