04-ResNET

#!/usr/bin/env python

from tensorflow.keras import Sequential
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.layers import Conv2D, Flatten, Dense, AveragePooling2D
from tensorflow.keras.optimizers import Adam
import numpy as np


def load_train(path):
    datagen = ImageDataGenerator(rescale=1./255, horizontal_flip=True, vertical_flip=True)
    train_datagen_flow = datagen.flow_from_directory(
        path,
        target_size=(150, 150),
        batch_size=50,
        class_mode='sparse',
    seed=12345)

    return train_datagen_flow


def create_model(input_shape):
    optimizer = Adam()
    model = Sequential()

    model.add(Conv2D(filters=6, kernel_size=(3, 3), padding='same', activation='relu', input_shape=input_shape))
    model.add(AveragePooling2D())
    model.add(Conv2D(filters=16, kernel_size=(3, 3), activation='relu'))
    model.add(AveragePooling2D())
    model.add(Flatten())
    model.add(Dense(units=120, activation='relu'))
    model.add(Dense(units=84, activation='relu'))
    model.add(Dense(units=12, activation = 'softmax'))

    model.compile(loss='sparse_categorical_crossentropy', optimizer=optimizer, metrics=['acc'])
    return model


def train_model(model, train_data, test_data, batch_size=None, epochs=5,
               steps_per_epoch=None, validation_steps=None):
    model.fit(train_data,
        batch_size=batch_size,
        epochs=epochs,
        steps_per_epoch=steps_per_epoch,
        validation_steps=validation_steps,
        validation_data=test_data,
        verbose=2, shuffle=True)
    return model


def main():
    path = "/datasets/fruits_small/"

    train_flow = load_train(path)
    # features_test, target_test = load_test(path)

    model = create_model((150,150,3))
    # print(model.summary())
    model = train_model(model, train_flow, None )


if __name__ == '__main__':
    main()