Hello Ankit. This is quite interesting.
Can you provide some supplementary commentary on how the GPT was trained -- or is it using, for example, a vector database at the back end to go through the corpus of Modelica documentation and so on.
Looking forward to learning more about the process of setting this up.