Linear

class Linear(output_size: int, weight_initializer: bluebird.weight_initializers.WeightInitializer = <bluebird.weight_initializers.HeWeightInitializer object>, bias_initializer: bluebird.weight_initializers.WeightInitializer = <bluebird.weight_initializers.ZerosWeightInitializer object>)

Bases: bluebird.layers.layer.Layer

It calculates the output based on formula:

output = input @ weights + bias

Example:

linear = Lainear(50)
net = NeuralNet([
        ...
        linear,
        ...
    ])

__init__(output_size: int, weight_initializer: bluebird.weight_initializers.WeightInitializer = <bluebird.weight_initializers.HeWeightInitializer object>, bias_initializer: bluebird.weight_initializers.WeightInitializer = <bluebird.weight_initializers.ZerosWeightInitializer object>) → None

Initializes the object.

Parameters

• output_size (int) – dimension of the output

• activation (Activation) – activation function

• weight_initializer (WeightInitializer, optional) – defines how weights are initialized, defaults to HeWeightInitializer

• bias_initializer (WeightInitializer, optional) – defines how weights are initialized, defaults to ZerosWeightInitializer

Methods

__init__(output_size[, weight_initializer, …])	Initializes the object.
backward(grad)	Used to calculate the gradients of weights and biases.
build(input_size)	Called by the model, before its training step.
forward(inputs[, training])	Called each time the data passes throughout the nework.
get_params()	Returns params of layer.

backward(grad: numpy.ndarray) → numpy.ndarray

Used to calculate the gradients of weights and biases.

Parameters

grad (Tensor) – gradient from previous layer or loss function.

Returns

Gradient

Return type

Tensor

build(input_size: int) → None

Called by the model, before its training step.

It sets the input size and initializes the weights.

Parameters

input_size (int) – output size from previous layer

forward(inputs: numpy.ndarray, training: bool = False) → numpy.ndarray

Called each time the data passes throughout the nework.

Parameters

• inputs (Tensor) – output from the previous layer

• training (bool, optional) – set to true during training, and is false when network predicts

Returns

processed input data

Return type

Tensor

get_params() → Dict

Returns params of layer.

Returns

all params of layer

Return type

dict