Model:

The model has been trained using mosaicml/mpt-7b as base model.

Dataset:

The model has been trained on Katonic/extraction_data which consists of around 80k records.

Date:

18 july 2023