Target is to further fix bugs, ensure robust benchmarks, better structuring of the interpreter and first complete example of transformer based models. The docstrings and docu need to stay updated always.