A talk from MediaParty 2023
I promise this is actually about data journalism and generative AI, even if you reaaaally don't believe it! Based on my post Multi-language document Q&A with LangChain and GPT-3.5-turbo from March 2023.
Material:
Data sources:
- Eredeti népmesék on Project Gutenberg (a nice text file!)
- Eredeti népmesék on Google Books (the original!)
Tools and tech:
- langchain for anything and everything
- Our embeddings on HuggingFace
- Chroma vector database
- LlamaIndex for an alternative to langchain
Contact me: