PygmalionAI is an open-source effort to create large language models (LLMs) for chatting and role-playing purposes. Our aim is to create LLMs that can offer interactions mimicking human behaviour as closely as possible.

Who are we? Link to this heading

The PygmalionAI team is comprised of three individuals:

  • TearGosling (Models & Training Code)
  • Alpin (Models & Misc.)
  • Tav (Community/PR Manager)

Learn More Link to this heading

You can refer to our HuggingFace to access our models and dataset.