Sambanova Launches 1 Trillion-Parameter Language Model with Innovative “Composition of Experts” Architecture

Sambanova Launches 1 Trillion-Parameter Language Model with Innovative “Composition of Experts” Architecture

The post Sambanova Launches 1 Trillion-Parameter Language Model with Innovative “Composition of Experts” Architecture appeared first on News Directory 3.
Rodrygo Liang, CEO of Sambanova (Photo = Sambanova) Artificial intelligence (AI) chip startup Sambanova has launched a large language model (LLM) with over 1 trillion parameters. This model features a “composition of experts” (CoE) architecture that combines 56 independently trained models and directs them to the optimal model based on suggestions and requests. Venture Beat […]

Share this post :

Facebook
Twitter
LinkedIn
Pinterest

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest News