import os import pandas as pd import tensorflow from keras.applications.densenet import layers def main(EPOCHS): if EPOCHS == 0: EPOCHS = 500 train_data_x = pd.read_csv('./X_train.csv') adults_train = train_data_x.copy() adults_predict = train_data_x.pop('age') normalize = layers.Normalization() normalize.adapt(adults_train) adult_model = tensorflow.keras.Sequential([ normalize, layers.Dense(64), layers.Dense(1) ]) adult_model.compile( loss=tensorflow.keras.losses.MeanSquaredError(), optimizer=tensorflow.keras.optimizers.Adam()) adult_model.fit(adults_train, adults_predict, epochs=EPOCHS) adult_model.save('model') if __name__ == "__main__": EPOCHS = int(os.environ['EPOCHS']) main(EPOCHS)