Generative AI Simplified - tokens, embeddings, vectors and similarity search

preview_player
Показать описание
In this video learn about some important concepts in Text based generative AI like tokens, vectors, and similarity.

Humans are vague! One of the big advances in these generative AI technologies is due to their ability to infer meaning from vague instructions, to ignore our spelling and grammatical mistakes and not behave like command lines. They are fuzzy and to be fuzzy they need all this apparatus to find similar meanings and text to predict what the answer should be.

If you're coding with generative AI like Open AI you'll need to know what tokens are, how to work with them and what vectors mean in embedding models.
Рекомендации по теме
Комментарии
Автор

Layman who has done his research making videos for laymen.
I'm also a Python programmer who is more interested in use cases and applications (for general research purposes) than how the models work, but ofc, we still need a basic idea of how they work. Keep up the videos, they will become more popular as ore laymen start to strap LLMs onto their projects.
+1 Like.

mandelbro
Автор

Fantastic video! Kudos to the presenter for demystifying these AI complex topics with such a clarity 👏

wassimchegham
Автор

Great video! I really got so many answers to my questions! Thank you so much!

fightinamrah
Автор

Very nice and clear explanations, thank you !

harrykaradimas