Open Models Almanac
Methodsdata curationdata mixture & curriculum

single mention · data curation

Joint vision-language and action training

Co-training a model on a high-quality vision-language dataset under a next-token prediction objective.

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.

Filed under data curation › data mixture & curriculum