About the pretrained model used in the building process of PMC-VQA

[SunSain]

Thank you for your marvelous work! And would you open more information about the building process of PMC_VQA dataset in the future, such as the LLaMA-7B trained with text data and finetuned with the manually annotated 2k question-answer pairs？ Thak you very much and any reply would be appreciated.