Base model: [gpt2-xl](https://huggingface.co/gpt2-xl)

Domain-adapted for dialogue response and narrative generation on a [narrative-aligned variant](https://github.com/AbrahamSanders/gutenberg-dialog#download-narrative-aligned-datasets) of the [Gutenberg Dialogue Dataset (Csaky & Recski, 2021)](https://aclanthology.org/2021.eacl-main.11.pdf)

Fine-tuned for dialogue response generation on [Persuasion For Good (Wang et al., 2019)](https://aclanthology.org/P19-1566.pdf) ([dataset](https://gitlab.com/ucdavisnlp/persuasionforgood))