A Preliminary Attempt of Deploying AIGC in Autodriving Scenario Generation: Via LLM

With huge progress in Large Language Model (LLM) by Artificial-Intelligence-Generated-Content (AIGC), a technical demand from the industry of autodriving with respect to scenario generation needs sees a new possible solution. Thus a new powerful tools of computational modeling, and simulation is now at hand. In this paper, we explore the use of prompt engineering and the fine-tuning applied to open-source model. This paper studies the use about OpenSCENARIO. The theoretical descriptions comes with practical experiment in comparisons of prompt feedbacks and performance estimation of long conceptual Q&A and code snippet generation.