该系列模型在通用中文基座模型（如 Chinese-LLaMA、ChatGLM 等）的基础上扩充法律领域专有词表、大规模中文法律语料预训练，增强了大模型在法律领域的基础语义理解能力。在此基础上，构造法律领域对话问答数据集、中国司法考试数据集进行指令精调，提升了模型对法律内容的理解和执行能力。
https://github.com/pengxiao-song/LaWGPT
LaWGPT is a large language model based on Chinese legal knowledge. This model is built upon a foundation of general Chinese language models such as Chinese-LLaMA and ChatGLM. However, it adds a domain-specific legal vocabulary and trains on a large corpus of Chinese legal language to enhance the model's understanding of legal semantics.
Additionally, LaWGPT is fine-tuned on legal dialogue question-answering datasets and the Chinese judicial examination dataset to improve its ability to understand and execute legal content.
LaWGPT can be found on GitHub at https://github.com/pengxiao-song/LaWGPT.