This dataset is intended to be the final training part, after Base-1 and Base-2.
This dataset is under the Andy license, which means it is open source, and very similar to the MIT or Apache 2.0 License, however Andy must be included in the final name somewhere, and that I, Sweaterdog, have to be cited in the model repo.
This dataset can be used in any terms. It is under then ShareGPT format, and uses conversations as the column… See the full description on the dataset page:
https://huggingface.co/datasets/Sweaterdog/Andy-4-FT.
