2 ms·Starling LM 7B Beta1 points by tosh 3y agotosh 3y agoMost interesting here imho: > Also introducing Starling-RM-34B, a Yi-34B-based reward model trained on our Nectar dataset