[Video] AI model distillation

Hi Community!

We're happy to share the next video in the series dedicated to Gen AI on our InterSystems Developers YouTube:

⏯ AI model distillation

https://www.youtube.com/embed/Ko0jM3YjprY
[This is an embedded link, but you cannot view embedded content directly on the site because you have declined the cookies necessary to access it. To view embedded content, you would need to accept all cookies in your Cookies Settings]

AI model distillation is a technique used to transfer knowledge from large AI models (teacher models) to smaller, more efficient models (student models). Large models, while highly capable, require significant computational resources, making them costly to train and deploy. Smaller models, on the other hand, are faster, cheaper, and more energy-efficient, making them suitable for applications on devices like phones and laptops.

Watch this video to get acquainted with the process of model distillation, which involves running data through both the teacher and student models, comparing their outputs (logits), and using algorithms to adjust the student model to better match the teacher’s responses. This iterative process helps the smaller model approximate the performance of the larger one while maintaining efficiency. Model distillation is widely used in the AI industry, as seen in versions like ChatGPT Turbo, which are distilled from larger models to optimize performance while reducing resource consumption.

🗣  Presenter: @Nicholai Mitchko, Manager of Solution Partner Sales Engineering, InterSystems

Enjoy watching, and look forward to more videos! 👍