trying out the LLM interface

tryout different setup

added to try parse json

more structure to the request and reply: need for non-emptiness

setting up a 2-model strategy

app development tryout

replace gpt2 with google tflan5 and json graph with networkx

update for errors in current version app

increase max tokens

work with 3 models

updating requirements file

update requirements file

adding a requirements file

first version of app

modification in README

initial commit