A Complementary French Question-Answering Dataset Generated by GPT-4
This dataset extends the work presented in the original CQuAE corpus by generating additional question-answer items automatically.
Specifically, GPT-4 was used to create new French questions (in four distinct categories) from the same collection of educational documents (Histoire, Géographie, SVT, etc.) used in CQuAE.
The goal is to broaden the scope of available training data for small… See the full description on the dataset page:
https://huggingface.co/datasets/LsTam/CQuAE_synthetic.
