architectures.py 文件源码

def convnet(input, output, dropout_rate=0., input_shape=(1, 28, 28), batch_size=100,
            l2_rate=0.001, nb_epoch=12, img_rows=28, img_cols=28, nb_filters=64,
            pool_size=(2, 2), kernel_size=(3, 3), activations='relu', constrain_norm=False):
    '''
    Helper function for building a Keras convolutional network.

    input:  Keras Input object appropriate for the data. e.g. input=Input(shape=(20,))
    output: Function representing final layer for the network that maps from the last
            hidden layer to output.
            e.g. if output = Dense(10, activation='softmax') if we're doing 10 class
            classification or output = Dense(1, activation='linear') if we're doing
            regression.
    '''
    const = maxnorm(2) if constrain_norm else None

    state = Convolution2D(nb_filters, kernel_size, padding='valid',
                          input_shape=input_shape, activation=activations,
                          kernel_regularizer=l2(l2_rate), kernel_constraint=const)(input)

    state = Convolution2D(nb_filters, kernel_size,
                          activation=activations, kernel_regularizer=l2(l2_rate),
                          kernel_constraint=const)(state)

    state = MaxPooling2D(pool_size=pool_size)(state)

    state = Flatten()(state)

    if dropout_rate > 0.:
        state = Dropout(dropout_rate)(state)
    state = Dense(128, activation=activations, kernel_regularizer=l2(l2_rate), kernel_constraint=const)(state)

    if dropout_rate > 0.:
        state = Dropout(dropout_rate)(state)
    return output(state)