Google With AI Singapore To Enhance LLMs: What Is Project SEALD?

AI Singapore (AISG) and Google Research have embarked on a joint mission called Project SEALD. This collaboration aims to strengthen cultural context understanding and competencies in SEA LLMs and expand their regional application to benefit society.

So, let’s briefly learn about the collaboration of Google with AI Singapore to enhance LLMs. Furthermore, we will also divulge some exciting facts about the meaning of SEALD.

What Is SEALD?

SEALD stands for Southeast Asian Languages in One Network Data. The research cooperation aims to improve datasets for training, fine-tuning, and evaluating large language models (LLMs) in Southeast Asian languages.

VISTEC is thrilled to be part of this pan-ASEAN natural language processing (NLP) development afforded by Project SEALD. This critical cooperation mechanism sets our different NLP communities on a single collective and strategic path.

Similarly, project SEALD will address the resource constraints of incorporating SEA languages into AI advances by providing new pre-trained language models, datasets, and benchmarks.

Furthermore, the project will also assist in building a diverse and high-quality data corpus of languages spoken in Southeast Asia. It will start with Thai, Filipino, Burmese, and Tamil to support the models’ training under SEA languages in one network.

Google With AI Singapore to Enhance LLMs – What’s the Mission?

Under Project SEALD, AISG and Google Research Asia Pacific (APAC) will collaborate to develop translocalization and translation models.

They aim to establish best practices for instruction tuning datasets, create tools to enable translocalization at scale and publish pre-training recipes for SEA languages.

Similarly, the collaboration will release datasets and output from Project SEALD, which will be available in open source to help the SEA LLM ecosystem grow and promote regional solid expertise.

Google With AI Singapore To Enhance LLMs. Image Credit: Social Media.

SEALD will interact with ecosystem partners (academic, industry, and government) in various ways. These include collaborating with industry players on data collection, curation, and quality checks.

Furthermore, there is collaboration with academia in various SEA countries to implement cutting-edge evaluation and benchmarking techniques, and government stakeholders in Singapore and across the region are being partnered to advance use cases that benefit society as a whole.

ALSO READ: NASA SpaceX Crew-7 Splashdown: ISS Crew Members Earth Returned

Advancement of SEA LLMs For the Region

Google Cloud and AI Singapore AI will work shoulder to shoulder to make SEA-LION LLMs available on Google Cloud’s Model Garden on Vertex AI.

It will give enterprises access to first-party, third-party, and open models that meet Google Cloud’s rigorous enterprise security and quality criteria.

Organizations may use Vertex AI’s enterprise-grade tools to tweak and incorporate these models into their applications.

Google With AI Singapore To Enhance LLMs. Image Credit: Social Media.

In addition to the collaboration of Google with AI Singapore to enhance LLMs, AISG has already begun partnerships across SEA countries and Singapore.

Memorandums of Understanding (MOUs) or Letters of Intent (LOIS) have already been signed with Indonesian, Vietnamese, and Malaysian entities to create datasets and applications for regional LLMs.

What Do the Key Partners Have To Say About the Collaboration?

Yolyn Ang – Google APAC

Ang mentioned that Google is exceptionally proud to partner with AI Singapore to put Southeast Asia, including Singapore, on the AI development map.

Further, she said that the collaboration will provide new opportunities and make AI more inclusive and accessible to individuals and businesses in the region.

Sarana Nutanong – Vidyasirimedhi Institute of Science and Technology, Thailand

The VISTEC professor excitedly announced that the institution is very excited to join the PA-ASEAN NLP development offered by Project SEALD. He mentioned that VISTEC feels proud to be a partner in the project.

Dr Leslie Teo – AI Singapore

Teo mentioned that the SEA-LION LLM project has always been about building an ecosystem and community that will work hand-in-hand to increase SEA-LION data corpus quality.