Hoar012/RAP-260K
收藏官方服务：
资源简介：
RAP Instruct 260K数据集是一个用于训练RAP-Phi3-V模型的完整数据集，包含了260K条数据。此外，还有一个210K的子集用于训练RAP-LLaVA模型。该数据集主要用于多模态大型语言模型（MLLMs）个性化研究。数据集遵循cc-by-nc-4.0许可，仅限于非商业研究使用。
The RAP Instruct 260K dataset is a full dataset containing 260K entries used for training the RAP-Phi3-V model. Additionally, there is a subset of 210K entries for training the RAP-LLaVA model. The dataset is primarily used for research on personalizing multimodal large language models (MLLMs). It is licensed under cc-by-nc-4.0 and is intended for non-commercial research purposes only.
提供机构：
Hoar012
