This finetuning dataset has been used to train MonadGPT, a chatGPT-like model for the early modern period.
It contains 10,797 excerpts of texts in English, French and Latin, mostly published in the 17th century, as well as synthetic questions generated by Mistral-Hermes.
The instructions use the chatML format with a unique system prompt (to help with consistency), user questions and assistant answers.
All the excerpts are in the public domain and so are the synthetic instructions (in… See the full description on the dataset page:
https://huggingface.co/datasets/Pclanglais/MonadGPT.
