SHERKALA by Inception and MBZUAI revolutionizes Kazakhstan’s LLM sector

NewsSHERKALA by Inception and MBZUAI revolutionizes Kazakhstan's LLM sector

Abu Dhabi Unveils SHERKALA: A Breakthrough in Kazakh Language AI

In a significant development for the Kazakh-speaking community, Inception, a company under the G42 umbrella, has partnered with the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) to introduce SHERKALA. This cutting-edge Kazakh Large Language Model (LLM) aims to bring the power of generative artificial intelligence to over 13 million Kazakh speakers worldwide.

SHERKALA stands out with its impressive 8-billion-parameter model, which has been meticulously trained on a massive dataset consisting of 45 billion words. While the primary focus of the model is on the Kazakh language, it also incorporates English, Russian, and Turkish. The model builds on the foundation of Llama 3.1, adapting it specifically for Kazakh, and includes a 25% expansion in its tokenizer. This expansion is crucial for enhancing the model’s ability to understand and generate the Kazakh language more effectively. The training process utilized the Condor Galaxy, one of the most powerful AI supercomputers globally, developed by G42 and Cerebras.

Dr. Andrew Jackson, CEO of Inception, expressed excitement about the launch, stating that SHERKALA represents a significant step towards addressing the needs of underserved linguistic communities through AI. By working with MBZUAI, Inception is proud to introduce a model that not only empowers Kazakh speakers but also redefines the landscape of LLMs with scalable, efficient, and inclusive AI solutions. Following the success of JAIS for Arabic speakers and NANDA for Hindi speakers, SHERKALA’s introduction marks another milestone in ensuring that underserved languages are well-represented in the AI ecosystem. Dr. Jackson emphasized that this advancement brings us closer to a future where technology amplifies every voice, contributing to a more equitable world.

SHERKALA: Setting New Standards for Kazakh Language Models

SHERKALA has set a new benchmark in the realm of Kazakh LLMs, excelling in both understanding and generating responses in the Kazakh language. It surpasses larger models in terms of efficient token generation and boasts state-of-the-art conversational capabilities. These have been rigorously tested against a variety of human-curated queries related to Kazakh culture, history, and general knowledge. Remarkably, SHERKALA is recognized as the best-performing open-source model of its size, even outshining some 70-billion-parameter models in its generative abilities.

Professor Preslav Nakov, Chair of the Natural Language Processing Department at MBZUAI, highlighted the significance of the project, stating that the collaboration with Inception on SHERKALA reflects a shared vision of crafting impactful AI solutions for underserved markets. SHERKALA represents a considerable advancement in democratizing AI access, preserving linguistic heritage, and empowering communities to thrive in the digital age. According to Professor Nakov, this partnership is transforming the LLM landscape, setting a precedent for innovative, inclusive, and responsible AI development.

Availability and Accessibility

For those interested in exploring and building upon its capabilities, SHERKALA is now accessible as an open-source model on the Hugging Face platform. This availability allows researchers, developers, and businesses to delve into its potential and contribute to its ongoing development. You can find SHERKALA at Hugging Face.

Understanding Large Language Models

Large Language Models (LLMs) like SHERKALA are a type of artificial intelligence designed to understand and generate human language. These models are trained on vast amounts of text data and use complex algorithms to predict and generate text that is contextually relevant and coherent. The "parameter" in LLMs refers to the number of variables the model uses to make predictions. For instance, SHERKALA’s 8-billion-parameter framework allows it to handle complex linguistic tasks effectively.

LLMs have numerous applications, ranging from chatbots and virtual assistants to content creation and language translation. They represent a significant leap in AI capabilities, enabling more natural and intuitive interactions between humans and machines.

About Inception

Inception is a forward-thinking company under G42, dedicated to creating AI-native products that leverage advanced AI research and models to solve business challenges. They specialize in developing domain-specific AI applications, offering AI-driven solutions across various languages and sectors. More information about their work can be found on their website at Inception AI. You can also follow them on social media platforms like LinkedIn, Instagram, and X.

About Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)

As the first graduate research university focused on artificial intelligence, MBZUAI is at the forefront of AI development and innovation. They offer a range of programs aimed at advancing AI knowledge and applications. For those interested in pursuing an education in AI, detailed information about admissions can be accessed at MBZUAI.

Concluding Thoughts

The launch of SHERKALA is a testament to the potential of AI in bridging linguistic gaps and empowering underserved communities. By advancing the capabilities of AI in the Kazakh language, Inception and MBZUAI are not only preserving cultural heritage but also paving the way for a more inclusive digital future. This initiative highlights the importance of developing AI solutions that are not only technologically advanced but also socially responsible and culturally sensitive.

For more Information, Refer to this article.

Neil S
Neil S
Neil is a highly qualified Technical Writer with an M.Sc(IT) degree and an impressive range of IT and Support certifications including MCSE, CCNA, ACA(Adobe Certified Associates), and PG Dip (IT). With over 10 years of hands-on experience as an IT support engineer across Windows, Mac, iOS, and Linux Server platforms, Neil possesses the expertise to create comprehensive and user-friendly documentation that simplifies complex technical concepts for a wide audience.
Watch & Subscribe Our YouTube Channel
YouTube Subscribe Button

Latest From Hawkdive

You May like these Related Articles

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.