Instructor with Jason Liu - Weaviate Podcast #88!

Video Description

Jason Liu is the creator of Instructor, one of the world's leading LLM frameworks, particularly focused on structured output parsing with LLMs, or as Jason puts it "making LLMs more backwards compatible". It is hard to understand the impact of Instructor, this is truly leading us to the next era of LLM programming. It was such an honor chatting with Jason, his experience currently as an independent consultant and previously engineering at StitchFix and Meta makes him truly one of the most unique guests we have featured on the Weaviate podcast! I hope you enjoy the podcast! Links: Pydantic is all you need: Jason Liu - https://www.youtube.com/watch?v=yj-wSRJwrrc Course trailer for Jason's new course with Weights & Biases: https://twitter.com/jxnlco/status/1756704806436040856 (amazing, one of the best I've ever seen) Instructor: https://github.com/jxnl/instructor Model embeddings: https://modal.com/blog/embedding-wikipedia Attention Heads podcast with Jason Liu: https://www.youtube.com/watch?v=5-5jf3_mvBg New Weaviate Blog Post related to what Jason describes on generating synthetic data for fine-tuning at 6:30: https://weaviate.io/blog/fine-tuning-coheres-reranker Chapters 0:00 Welcome Jason! 1:10 Overview of Instructor 3:30 How does Instructor work? 6:00 Making the IDE happy 7:15 Function Calling 9:30 Cleaning Prompting Code 13:45 Instructor for RAG 18:50 RAG Consulting 20:26 Jason’s thoughts on Ref2Vec 28:16 LLM Latency and Streaming 31:15 ETL for LLMs with Instructor 35:25 Modal Embeddings 41:18 Fine-Tuning Embedding models 50:20 Future Directions for Instructor