Difference between tf.saved_model.save(model, path_to_dir) and tf.keras.model.save

I have a keras model and want to save it as a tensorflow graph. Is there a difference between tf.saved_model.save(model, path_to_dir) and tf.keras.model.save.

In both of these, I want to save in a tensorflow saved format and will not be using h5 format. I understand tf.saved_model.save is more generic but if I am using a keras model are these two different in anyways.

https://stackoverflow.com/questions/66827371/difference-between-tf-saved-model-savemodel-path-to-dir-and-tf-keras-model-sa March 27, 2021 at 10:02AM