I’m trying to create and save a basic TFLite model backed by a HashTable. Following the page, I was able to create the table, but I don’t really know enough about Tensorflow in order to turn it into a model. My goal is to create and save something I can later load with the tf.lite.Interpreter, does anyone know how I can accomplish this?
It would be really great to create a model that can take in two values, look each up in its own table, and return both results. Or even for the sake of example, fetching the values from the hashtable and then returning the average or something like that.
Edit: I was able to get this to work, but the Hashtable op isn’t available as a builtin op, despite doc here claiming that it is. Is the doc correct? Is there any plan to add HashTable to the builtin ops?