3 ms·
Most interesting here imho: > Also introducing Starling-RM-34B, a Yi-34B-based reward model trained on our Nectar dataset
by tosh 3y ago
Most interesting here imho:
> Also introducing Starling-RM-34B, a Yi-34B-based reward model trained on our Nectar dataset