Some simple python scripts highlighting:
- How LangChain can be used to decorate LLM queries with Tree of Thoughts prompts
- How multiple models of the same type, but with different parameters and different ToT decorators, can be combined to yeild better ToT outcomes 'monoculture'
- How different models of different types can be combined to yield more diverse outcomes 'polyculture'
The code in this repo is based om the content of videos: https://youtu.be/y6SVA3aAfco
in this playlist: https://www.youtube.com/playlist?list=PLaJCKi8Nk1hwZcuLliEwPlz4kngXMDdGI