TrafficSafetyGPT: Tuning a Pre-trained Large Language Model to a Domain-Specific Expert in Transportation Safety

Large Language Models (LLMs) have shown remarkable effectiveness in various generaldomain natural language processing (NLP) tasks. However, their performance in transportation safety domain tasks has been suboptimal, primarily attributed to the requirement for specialized transportation safety expertise in generating accurate responses [1]. To address this challenge, we introduce TrafficSafetyGPT, a novel LLaMA-based model, which has undergone supervised fine-tuning using TrafficSafety-2K dataset which has human labels from government produced guiding books and ChatGPT-generated instruction-output pairs. Keywords: ChatGPT, Natural Language Processing, Deep Learning, Traffic Safety, Large Language Models, Generative Pre-trained Transformers 1. Introduction In the realm of natural language processing (NLP) and large language models, a surge in advancements has unfolded a plethora of potential applications. This rapid development, spearheaded by pre-trained large language models like OpenAI's ChatGPT and its derivatives, has drastically augmented our capabilities in language comprehension, generation, and interactivity. The foundational strength of these models lies in their pre-training on extensive and diverse datasets, empowering them to decipher intricate language patterns and contextual interconnections. Nevertheless, while these pre-trained models exhibit commendable proficiency across an array of tasks, their generic nature could constrain their efficacy in niche applications, such as transportation safety.