The Stack Overflow Podcast cover image

Generating text with diffusion (and ROI with LLMs)

The Stack Overflow Podcast

00:00

Training denoisers instead of next-token prediction

Stefano details training transformers to reconstruct clean text by adding mistakes and teaching the model to correct them.

Play episode from 02:48
Transcript

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