Text set of images about training data

I'm glad to see such excellent work.
I hope to use this opportunity to conduct more in-depth research. We are working on multimodal pathology tasks, but we lack a more professional image text description set, like CLIP. I would like to know on which text set CONCH was completed, and if it is possible to release the text set or provide a download link.
Thank you in advance.