Open Models Almanac
Methodsdata curationdata mixture & curriculum

single mention · data curation

Dynamic sampling for mixed-task reinforcement learning

Dynamically sampling mixed-task data alongside a scheduler targeting a specified training distribution.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See data mixture & curriculum for the methods filed beside it.

Also called Dynamic sampling for mixed-task RL, dynamic sampler.

Filed under data curation › data mixture & curriculum