Fig.1

Concept

Continual pre-training

Continual pre-training (sometimes "continued pre-training" or "domain-adaptive pre-training") means taking an already-pretrained base model and running more of the same next-token objective on a new corpus, before any instruction tuning. Think of it like the original…

The rest of “Continual pre-training” is a premium feature: every concept in the library gets a precise, practitioner-focused write-up like this one, cross-linked straight from the paper summaries that use it.

Log in to unlock

← Back to the library