UAE's G42 Introduces Open Source Arabic language AI Model By CIOTechOutlook Team

UAE's G42 Introduces Open Source Arabic language AI Model

CIOTechOutlook Team | Monday, 06 May 2024, 03:23 IST

A team of engineers, researchers, and a Silicon Valley semiconductor manufacturer partnered to create powerful Arabic language software for generative AI applications. The new massive language model known as Jais has 13 billion parameters and was created from a large batch of data combining Arabic and English, some of which is derived from computer code.

The group, which included academics and engineers, took on the project in part because they believed there were few large multilingual language models. The new language model was generated using supercomputers built by Silicon Valley-based Cerebras Systems, which creates dinner plate-sized chips that compete with Nvidia's strong AI hardware. Nvidia's processors are in low supply, prompting companies throughout the world to seek alternatives.

Jais, named after the tallest mountain in the UAE, is a cooperation between Cerebras, the Mohamed bin Zayed University of Artificial Intelligence, and Inception, an AI-focused subsidiary of Abu Dhabi-based tech giant G42.

According to Timothy Baldwin, a lecturer at Mohamed bin Zayed University of Artificial Intelligence, because there is insufficient Arabic data to train a model of Jais' scale, the computer code within the English language data helped train the model's reasoning capacity. "(Code) gives the model a big leg up in terms of reasoning abilities, because it spells out the (logical) steps," Baldwin added.