Collection of datasets for training and evaluating LVLMs for the Italian language, used in the paper "LLaVA-NDiNO: Empowering LLMs with Multimodality"