Ask for the prompt tuned llm

Hi there,

Tnanks for opensource this awesome project. I was wondering if it's possible for you to also provide the prompt-tuned llm weights as mentioned in the paper. 

It would be easier to reproduce and experiment with other ideas on top of GNN-RAG.

Thanks.