choosewhatulike/trainable-agents
A trainable agent system using LLMs for role-playing as specific historical characters like Beethoven and Cleopatra.
Not currently ranked — collecting fresh signals.
star history
This repository provides a trainable agent framework for role-playing using large language models. The system learns from generated experience data to embody specific characters with their knowledge and personalities. It introduces Experience Reconstruction, a data generation process that creates detailed experience data for training. Model weights for nine characters are released, trained via supervised fine-tuning.
Frequently asked
- What is choosewhatulike/trainable-agents?
- A trainable agent system using LLMs for role-playing as specific historical characters like Beethoven and Cleopatra.
- Is trainable-agents open source?
- Yes — choosewhatulike/trainable-agents is open source, released under the Apache-2.0 license.
- What language is trainable-agents written in?
- choosewhatulike/trainable-agents is primarily written in Python.
- How popular is trainable-agents?
- choosewhatulike/trainable-agents has 641 stars on GitHub.
- Where can I find trainable-agents?
- choosewhatulike/trainable-agents is on GitHub at https://github.com/choosewhatulike/trainable-agents.