# GPT4-Vision synthetic data

Math-LLaVA-13B is an open-source multimodal large language model fine-tuned on the MathV360K dataset based on LLaVA-1.5-13B, suitable for scenarios such as multimodal reasoning and Q&A.