Seed data
Collection
8 items • Updated
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
A 100-document sample from CaML's synthetic document finetuning (SDF) pipeline, with its companion reasoning set.
All text is fully synthetic (gemini-2.5-flash) with fictional people and organisations. This is a viewable sample for review, not a training-scale corpus.