Thank you for providing this solution. We are interested in conducting a more fine-grained analysis of the results from top-performing systems on the BIRD benchmark set. We are wondering if you might be willing to share the SQL queries generated by ContextualAI on the BIRD-dev benchmark set? This would be extremely helpful for our analysis. Thank you very much for considering our request, and we greatly appreciate your time and help.