(Reuters) – Google Cloud and Mistral AI are partnering to permit the Paris-based generative AI startup to distribute its language fashions on the tech big’s infrastructure, the businesses introduced on Wednesday.
“As a part of the settlement, Mistral AI will use Google Cloud’s AI-optimized infrastructure, together with TPU Accelerators, to additional take a look at, construct, and scale up its LLMs (giant language fashions), all whereas benefiting from Google Cloud’s safety and privateness requirements,” the joint assertion stated.
A big language mannequin is a statistical language mannequin, skilled on an enormous quantity of knowledge, that can be utilized to generate and translate textual content and different content material, and carry out different language processing duties.
(Reporting by Olivier Sorgho; Enhancing by Mark Potter)
