InternVL2.5-MPO Collection Enhancing the Reasoning Ability of MLLMs via Mixed Preference Optimization • 16 items • Updated 17 days ago • 26
Tulu 3 Datasets Collection All datasets released with Tulu 3 -- state of the art open post-training recipes. • 32 items • Updated 20 days ago • 64