1Dense layer does the below operation on the input and return the output.
2output = activation(dot(input, kernel) + bias)
1>>> # Create a `Sequential` model and add a Dense layer as the first layer.
2>>> model = tf.keras.models.Sequential()
3>>> model.add(tf.keras.Input(shape=(16,)))
4>>> model.add(tf.keras.layers.Dense(32, activation='relu'))
5>>> # Now the model will take as input arrays of shape (None, 16)
6>>> # and output arrays of shape (None, 32).
7>>> # Note that after the first layer, you don't need to specify
8>>> # the size of the input anymore:
9>>> model.add(tf.keras.layers.Dense(32))
10>>> model.output_shape
11(None, 32)
12
1tf.keras.applications.DenseNet169(
2 include_top=True,
3 weights="imagenet",
4 input_tensor=None,
5 input_shape=None,
6 pooling=None,
7 classes=1000,
8)
9