Understanding vector embeddings: how to build a ChatGPT bot for your own documents
In this video, Marcus Hellberg explains how you can use vector embeddings to provide custom context to ChatGPT to make it aware of your documents.
0:00 - The problem with LLMs, controlling the context
2:04 - What are vector embeddings? (simple explanation)
3:39 - Crafting a prompt with context
4:25 - Processing and splitting documents
5:13 - Asking questions with context
Blog post with code examples: https://marcushellberg.dev/how-to-bui...