Open Models Almanac
Methodspost-trainingreinforcement learning algorithm

single mention · post-training

SFT checkpoint initialization for RL

Starts reinforcement-learning training from a supervised fine-tuning checkpoint.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See reinforcement learning algorithm for the methods filed beside it.

Also called starting from an SFT checkpoint.

Filed under post-training › reinforcement learning algorithm