wideresnet

WideResidualNetwork

Creates a Wide Residual Network with specified parameters.

Parameters:

Name	Type	Description	Default
input_shape	Tuple[int, int, int]	The size of the input tensor (height, width, channels).	required
depth	int	Depth of the network. Compute N = (n - 4) / 6. For a depth of 16, n = 16, N = (16 - 4) / 6 = 2 For a depth of 28, n = 28, N = (28 - 4) / 6 = 4 For a depth of 40, n = 40, N = (40 - 4) / 6 = 6	28
widen_factor	int	Width of the network.	10
dropout	float	Adds dropout if value is greater than 0.0.	0.0
classes	int	The number of outputs the model should generate.	10
activation	Optional[str]	activation function for last dense layer.	'softmax'

Returns:

Type	Description
Model	A Keras Model.

Raises:

Type	Description
ValueError	If (depth - 4) is not divisible by 6.

Source code in fastestimator/fastestimator/architecture/tensorflow/wideresnet.py

def WideResidualNetwork(input_shape: Tuple[int, int, int],
                        depth: int = 28,
                        widen_factor: int = 10,
                        dropout: float = 0.0,
                        classes: int = 10,
                        activation: Optional[str] = 'softmax') -> tf.keras.Model:
    """Creates a Wide Residual Network with specified parameters.

    Args:
        input_shape: The size of the input tensor (height, width, channels).
        depth: Depth of the network. Compute N = (n - 4) / 6.
               For a depth of 16, n = 16, N = (16 - 4) / 6 = 2
               For a depth of 28, n = 28, N = (28 - 4) / 6 = 4
               For a depth of 40, n = 40, N = (40 - 4) / 6 = 6
        widen_factor: Width of the network.
        dropout: Adds dropout if value is greater than 0.0.
        classes: The number of outputs the model should generate.
        activation: activation function for last dense layer.

    Returns:
        A Keras Model.

    Raises:
        ValueError: If (depth - 4) is not divisible by 6.
    """
    if (depth - 4) % 6 != 0:
        raise ValueError('Depth of the network must be such that (depth - 4)' 'should be divisible by 6.')

    img_input = layers.Input(shape=input_shape)
    inputs = img_input

    x = __create_wide_residual_network(classes, img_input, depth, widen_factor, dropout, activation)
    # Create model.
    model = Model(inputs, x)
    return model