Request for full evaluation and GPT-based evaluation scripts

[raytrun]

Thank you for open-sourcing this amazing project and your great work on DriveAgent-R1!

To better understand the benchmarks and facilitate community reproduction/further research, would it be possible to release the evaluation scripts mentioned in your paper? Specifically, we are looking forward to:

1. Full Evaluation Scripts: The complete scripts needed to reproduce the main benchmark results reported in the paper.
2. GPT Evaluation Scripts: The specific scripts, prompts, or pipeline used for the GPT-based evaluation.