Parse positional arguments to tf.function

[khatchad]

If so, we need to parse the positional arguments as well.

Related to #30.

[khatchad]

From https://www.tensorflow.org/versions/r2.9/api_docs/python/tf/function:

tf.function(
    func=None,
    input_signature=None,
    autograph=True,
    jit_compile=None,
    reduce_retracing=False,
    experimental_implements=None,
    experimental_autograph_options=None,
    experimental_relax_shapes=None,
    experimental_compile=None,
    experimental_follow_type_hints=None
) -> [tf.types.experimental.GenericFunction](https://www.tensorflow.org/versions/r2.9/api_docs/python/tf/types/experimental/GenericFunction)

[khatchad]

Need 10 test cases.

[khatchad]

@tatianacv Use the positions of the arguments, not the count.

[khatchad]

Must find the definition of the tf.function signature we are using to lookup the number of order of parameters. Ensure that it matches what we are expecting (up to v2.9), otherwise, throw an exception.