b80fb3bdcfeac64d91b7365be1196392e013dcb8,autokeras/generator.py,DefaultClassifierGenerator,generate,#DefaultClassifierGenerator#,49

Before Change

    def generate(self):
        return one Sequential model that has been compiled
        model = Sequential()
        pool = self._get_pool_layer_func()
        conv = get_conv_layer_func(len(self._get_shape(3)))
        model.add(conv(32, kernel_size=self._get_shape(3),
                       activation="relu",
                       padding="same",
                       input_shape=self.input_shape))
        model.add(conv(64, self._get_shape(3),
                       padding="same",
                       activation="relu"))
        model.add(pool(pool_size=self._get_shape(2)))
        model.add(Dropout(0.25))
        model.add(Flatten())
        model.add(Dense(128, activation="relu"))
        model.add(Dropout(0.5))
        model.add(Dense(self.n_classes, activation="softmax"))

        model.compile(loss=categorical_crossentropy,
                      optimizer=Adadelta(),
                      metrics=["accuracy"])

After Change

        pool = self._get_pool_layer_func()
        conv = get_conv_layer_func(len(self._get_shape(3)))

        input_tensor = Input(shape=self.input_shape)
        output_tensor = conv(32, kernel_size=self._get_shape(3),
                             padding="same")(input_tensor)
        output_tensor = BatchNormalization()(output_tensor)
        output_tensor = Activation("relu")(output_tensor)

        output_tensor = conv(64, self._get_shape(3),
                             padding="same")(output_tensor)
        output_tensor = BatchNormalization()(output_tensor)
        output_tensor = Activation("relu")(output_tensor)

        output_tensor = pool(pool_size=self._get_shape(2), padding="same")(output_tensor)
        output_tensor = Dropout(0.25)(output_tensor)
        output_tensor = Flatten()(output_tensor)
        output_tensor = Dense(128, activation="relu")(output_tensor)
        output_tensor = Dropout(0.5)(output_tensor)
        output_tensor = Dense(self.n_classes, activation="softmax")(output_tensor)

        model = Model(input_tensor, output_tensor)
        model.compile(loss=categorical_crossentropy,
                      optimizer=Adadelta(),
                      metrics=["accuracy"])
        return model

Instances