Technology

Watad Energy & Communications Unveils “Mulhem,” First Locally Trained Arabic-Language Large Language Model in Saudi Arabia

Published

on

RIYADH, Saudi Arabia, March 6, 2024 /PRNewswire/ — In a significant development for the Arabic-speaking AI sector, Watad Energy & Communications today announced the launch of “Mulhem,” the first Saudi domain-specific AI model, trained exclusively on Saudi data, embodying the nation’s commitment to technological advancement and digital transformation.

“Today marks a significant milestone for the AI industry in Saudi Arabia,” said Muhammad Al-Mutairi, CEO of Watad. “Mulhem’s exceptional performance capabilities signal a new era for customer applications in the region. This achievement aligns with our commitment to contributing to Saudi Arabia’s Vision 2030 by pushing the boundaries of technological innovation and digital transformation.”

Watad CTO Anton Alexander commented, “The development of Mulhem was both a challenge and a triumph. We built one of the largest Arabic databases in the world and developed our own tokenizer, as part of our commitment to creating a model that not only understands the Arabic language but also embodies the rich cultural heritage of the Arab world. While leveraging NVIDIA’s cutting-edge technology has been instrumental in our development, our philosophy is to remain agile, always ready to embrace the best solutions available to meet our objectives.”

Key Highlights of Mulhem:

Unparalleled Dataset: Mulhem has been trained on an extensive and diverse dataset, comprising more than 70,000 Saudi-specific Q&A data points and over half a million Arabic single turn, multi-turn, preference data sets, along with specialized datasets for context retrieval and offline data systems. This rich dataset ensures that Mulhem offers unparalleled accuracy and relevance in its responses and insights, specifically tailored to the Saudi context.

Bilingual Capabilities: With a training foundation of 90 billion Arabic and 90 billion English tokens, Mulhem is not just the first of its kind in terms of regional specificity but also in linguistic versatility. This dual-language capability ensures that Mulhem can serve a wide range of applications, from international business to local cultural projects.

Local Training and Development: Proudly trained and developed in Saudi Arabia, Mulhem leverages the latest in AI research and development practices, including Supervised Fine-tuning and Direct Preference Optimization. This local focus not only underscores Watad Energy and Communication’s commitment to national talent and resources but also ensures that Mulhem is perfectly aligned with the needs and nuances of the Saudi market.

Innovative Technology: Embracing cutting-edge technologies, Mulhem is equipped with advanced AI features that enable it to perform a wide array of tasks, from natural language processing and context-aware communication to complex data analysis and retrieval tasks, all with exceptional accuracy and efficiency.

Commitment to Open Innovation: In the spirit of innovation and community collaboration, Watad Energy and Communication is exploring avenues to make Mulhem accessible to a broader audience, including potential open-source initiatives. This approach aims to foster a vibrant ecosystem of developers, researchers, and businesses, all contributing to and benefiting from Mulhem’s capabilities.

A Milestone for Saudi Arabia: Mulhem’s launch underscores Saudi Arabia’s emerging role as a hub for technological innovation and AI development, showcasing the nation’s commitment to investing in homegrown technologies with the potential to influence the global tech landscape.

About Watad Energy & Communications

Watad Energy and Communication is at the forefront of technological innovation in Saudi Arabia, committed to driving innovation and digital transformation in line with Saudi Arabia’s Vision 2030. Specializing in the development of advanced AI technologies, Watad is dedicated to enhancing the technological landscape of the region and beyond.

Photo – https://mma.prnewswire.com/media/2355087/Watad.jpg

Contact:  
Teodora Pantu,
PR Manager,  
teodora@dizrupt.us 

View original content:https://www.prnewswire.co.uk/news-releases/watad-energy–communications-unveils-mulhem-first-locally-trained-arabic-language-large-language-model-in-saudi-arabia-302080084.html

Leave a Reply

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

Trending

Exit mobile version