Skip to content

Fix ConvNeXt classifier activation bug #17544

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.

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
merged 10 commits into from
Feb 14, 2023
10 changes: 10 additions & 0 deletions keras/applications/applications_test.py
Original file line number Diff line number Diff line change
Expand Up @@ -191,6 +191,16 @@ def test_application_pooling(self, app, last_dim):
)
self.assertShapeEqual(output_shape, (None, last_dim))

@parameterized.parameters(MODEL_LIST)
def test_application_classifier_activation(self, app, _):
if "RegNet" in app.__name__:
self.skipTest("RegNet models do not support classifier activation")
model = app(
weights=None, include_top=True, classifier_activation="softmax"
)
last_layer_act = model.layers[-1].activation.__name__
self.assertEqual(last_layer_act, "softmax")

@parameterized.parameters(*MODEL_LIST_NO_NASNET)
def test_application_variable_input_channels(self, app, last_dim):
if backend.image_data_format() == "channels_first":
Expand Down
15 changes: 12 additions & 3 deletions keras/applications/convnext.py
Original file line number Diff line number Diff line change
Expand Up @@ -324,11 +324,12 @@ def apply(x):
return apply


def Head(num_classes=1000, name=None):
def Head(num_classes=1000, classifier_activation=None, name=None):
"""Implementation of classification head of RegNet.

Args:
num_classes: number of classes for Dense layer
classifier_activation: activation function for the Dense layer
name: name prefix

Returns:
Expand All @@ -342,7 +343,11 @@ def apply(x):
x = layers.LayerNormalization(
epsilon=1e-6, name=name + "_head_layernorm"
)(x)
x = layers.Dense(num_classes, name=name + "_head_dense")(x)
x = layers.Dense(
num_classes,
activation=classifier_activation,
name=name + "_head_dense",
)(x)
return x

return apply
Expand Down Expand Up @@ -522,8 +527,12 @@ def ConvNeXt(
cur += depths[i]

if include_top:
x = Head(num_classes=classes, name=model_name)(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = Head(
num_classes=classes,
classifier_activation=classifier_activation,
name=model_name,
)(x)

else:
if pooling == "avg":
Expand Down