model = Sequential()
model.add(Dense(5, input_shape=(x_train.shape[1:])))
for i in range(6):
    model.add(Dense(512, activation='relu'))
model.add(Dense(1))