BOREAL-2B
A 2-billion-parameter dense hybrid language model pretrained from scratch on 500B–2T tokens. BOREAL-2B is the first model in...
A 2-billion-parameter dense hybrid language model pretrained from scratch on 500B–2T tokens. BOREAL-2B is the first model in...
Цель. Гибридная языковая модель Mixture of Experts с ~10 миллиардами параметров и ~2 миллиардами активных параметров на токен....