def single_layer_model(input_shape):
    model = Sequential()

    # 2D convolution layer (e.g. spatial convolution over images).
    model.add(Conv2D(32, (3, 3), input_shape=input_shape))

    # Flattens the input. Does not affect the batch size.
    model.add(Flatten())

    # Regular densely-connected NN layer.
    model.add(Dense(10, activation='softmax'))

    model.compile(loss='categorical_crossentropy',
                  optimizer='adam',
                  metrics=['accuracy'])
    return model


def evaluate():
    with Timer("Summary time: "):
        img_width, img_height = 50, 50
        train_data_dir = 'nn_labs/exp1/train'
        validation_data_dir = 'nn_labs/exp1/validation'
        epochs = 50
        batch_size = 32
        # nb_train_samples = 200
        nb_train_samples = 500
        nb_validation_samples = 500

        if K.image_data_format() == 'channels_first':
            input_shape = (3, img_width, img_height)
        else:
            input_shape = (img_width, img_height, 3)

        # this is the augmentation configuration we will use for training
        train_datagen = ImageDataGenerator(rescale=1./255)
        test_datagen = ImageDataGenerator(rescale=1./255)

        train_generator = train_datagen.flow_from_directory(
            train_data_dir,
            target_size=(img_width, img_height),
            batch_size=batch_size)

        validation_generator = test_datagen.flow_from_directory(
            validation_data_dir,
            target_size=(img_width, img_height),
            batch_size=batch_size)

        model = single_layer_model(input_shape)

        with Timer("Fit time: "):
            csv_logger = CSVLogger('exp7.log')
            model.fit_generator(
                train_generator,
                steps_per_epoch=nb_train_samples // batch_size,
                epochs=epochs,
                validation_data=validation_generator,
                validation_steps=nb_validation_samples // batch_size,
                callbacks=[csv_logger])