How AI Is Built cover image

Maxime Labonne on Model Merging, AI Trends, and Beyond

How AI Is Built

00:00

Navigating Model Merging Techniques

This chapter explores the intricate process of merging AI models for skill transfer, focusing on the importance of different neural network layers. The discussion highlights the strategic manipulation of layers during fine-tuning and emphasizes the need for careful evaluation of merging strategies to enhance performance. Additionally, it addresses the complexities of using tokenizers and the implications of creating larger 'franken models' while maintaining adherence to safety protocols.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app