On the importance of post-training in modern LLM development

This site does not allow itself to be embedded. You can still read it on the original site — the toolbar below keeps your place in the directory.

p]:m-0">🚧 Note: This article is under construction. Post-training is all the rage. Are we going from training to teaching? What does a curriculum look like for LLMs? Let's find out. For most of the deep learning era, the core recipe for building intelligent models has been relatively straightforward: take a transformer, scale up the parameters, feed it vast amounts of text, predict the next token…

Read on dtsbourg.me ↗