The translated data: InternLM-XComposer2 is an advanced multimodal large model that achieves exceptional performance by freely combining text and images. Utilizing a partial LoRA approach, it maintains the integrity of linguistic knowledge while allowing for highly customized creation. It has demonstrated outstanding performance in multiple experiments, emerging as one of the leading vision-language models, and providing superior performance for tasks across various domains.