--- library_name: transformers license: mit language: - ban - my - en - id - jv - km - lo - ms - zh - su - ta - th - vi tags: - aisingapore - nvidia - pytorch arxiv: "2609.18310" datasets: - AI-MO/NuminaMath-CoT - aisingapore/SEA-Instruct-2602 - HuggingFaceFW/finetranslations-edu - jhu-clsp/megawika-2 - nvidia/Nemotron-SFT-OpenCode-v1 - nvidia/Nemotron-SFT-SWE-v2 - nvidia/Nemotron-SFT-Agentic-v2 - nvidia/Nemotron-SFT-Competitive-Programming-v2 - nvidia/Nemotron-SFT-Math-v3 - nvidia/OpenMathInstruct-2 - nvidia/OpenScienceReasoning-2 - KingNish/reasoning-base-20k - ZombitX64/Medical-o1-Reasoning-SFT-Thai pipeline_tag: text-generation track_downloads: true base_model: - nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16 --- ![Banner!](Nemotron-SEA-LION-v4.8-30B-A3B-Base.png "Nemotron-SEA-LION-v4.8-30B-A3B-Base")

Technical Report 👁️

# Nemotron-SEA-LION-v4.8-30B-A3B-Base *Last updated: 2026-09-18* **SEA-LION** is a collection of Large Language Models (LLMs) which have been pretrained and instruct-tuned for the Southeast Asian (SEA) region. **Nemotron-SEA-LION-v4.8-30B-A3B-Base** is built upon the [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16) architecture. This model underwent continued pretraining on 150B high-quality tokens across QA, reasoning, code, and translation data using the [NVIDIA NeMo Megatron Bridge](https://github.com/NVIDIA-NeMo/Megatron-Bridge) library. ## Model Details ### Model Description SEA-LION stands for Southeast Asian Languages In One Network. We performed CPT for 150B tokens in English and 10 SEA languages. For tokenization, the model employs the default tokenizer used in [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16). - **Developed by:** AI Products Pillar, AI Singapore - **Funded by:** National Research Foundation Singapore - **Shared by:** AI Products Pillar, AI Singapore - **Model type:** Base language model - **Architecture:** LatentMoE Hybrid - **Context length:** 262,144 tokens - **Language(s):** Balinese, Burmese, English, Indonesian, Javanese, Khmer, Lao, Malay, Mandarin, Sundanese, Tamil, Thai, and Vietnamese - **License:** [MIT](https://mit-license.org/) - **Parent model:** [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16) ### Model Sources - **Collection:** [SEA-LION v4.8 - an aisingapore Collection](https://huggingface.co/collections/aisingapore/sea-lion-v48) ## Training Details ### Training Data The continued pretraining dataset comprised 150B tokens on QA, CoT, Reasoning and parallel (translation) datasets adapted from - [AI-MO/NuminaMath-CoT](https://huggingface.co/datasets/AI-MO/NuminaMath-CoT) - [aisingapore/SEA-Instruct-2602](https://huggingface.co/datasets/aisingapore/SEA-Instruct-2602) - [HuggingFaceFW/finetranslations-edu](https://huggingface.co/datasets/HuggingFaceFW/finetranslations-edu) - [jhu-clsp/megawika-2](https://huggingface.co/datasets/jhu-clsp/megawika-2) - [nvidia/Nemotron-SFT-OpenCode-v1](https://huggingface.co/datasets/nvidia/Nemotron-SFT-OpenCode-v1) - [nvidia/Nemotron-SFT-SWE-v2](https://huggingface.co/datasets/nvidia/Nemotron-SFT-SWE-v2) - [nvidia/Nemotron-SFT-Agentic-v2](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Agentic-v2) - [nvidia/Nemotron-SFT-Competitive-Programming-v2](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Competitive-Programming-v2) - [nvidia/Nemotron-SFT-Math-v3](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Math-v3) - [nvidia/OpenMathInstruct-2](https://huggingface.co/datasets/nvidia/OpenMathInstruct-2) - [nvidia/OpenScienceReasoning-2](https://huggingface.co/datasets/nvidia/OpenScienceReasoning-2) - [KingNish/reasoning-base-20k](https://huggingface.co/datasets/KingNish/reasoning-base-20k) - [ZombitX64/Medical-o1-Reasoning-SFT-Thai](https://huggingface.co/datasets/ZombitX64/Medical-o1-Reasoning-SFT-Thai) ### Training Regime For training details, see the [SEA-LION-v4.8 Technical Report](https://arxiv.org/abs/2609.18310). ## Environmental Impact - **Hardware type:** H200 - **GPU-hours:** 114.23 - **Cloud provider:** SMC H200 - **Compute region:** Singapore - **Carbon emissions:** approximately 0.003 – 0.062 MT ## Technical Specifications ### Technical Report For training details, see the [SEA-LION-v4.8 Technical Report](https://arxiv.org/abs/2609.18310). ### Model Architecture The architecture is based on the highly efficient Nemotron-3-Nano foundation. The detailed architecture can be found at [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16 documentation](https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Nano-Technical-Report.pdf). ## Uses ### Out-of-Scope Use The model has not been aligned for safety. Developers and users should perform their own safety fine-tuning and related security measures. In no event shall the authors be held liable for any claims, damages, or other liabilities arising from the use of the released weights and codes. ### Bias, Risks, and Limitations The model was not tested for robustness against adversarial prompting. It is important for users to be aware that our model exhibits certain limitations that warrant consideration. Like many LLMs, the model can hallucinate and occasionally generates irrelevant content, introducing fictional elements that are not grounded in the provided context. Users should also exercise caution in interpreting and validating the model's responses due to the potential inconsistencies. ## Citation **BibTeX:** ```bibtex @misc{aisingapore2026sealionv48technicalreport, title={SEA-LION-v4.8: A Technical Report}, author={Adila Aulia and Ahmed Dabeer and Ahn Jeongmi and Antonyrex Sajeban and Chan Hok Teng Adwin and Cheng Zi Yi Nicholas and Choa Hsueh Mei Esther and Heng Jonathan and Jann Railey Estrada Montalan and Lee Chwan Ren and Leong Wai Yi and Leong Wei Qi and Liew Rachel and Limkonchotiwat Peerat and Muhammad Ridzuan Bin Mokhtar and Nagarajan Karthik and Ng Boon Cheong Raymond and Ngee Chia Tai and Ngui Jian Gang and Nguyen Thanh Ngan and Ong Tat-Wee David and Pereira Mark and Phang Shi Wei Benjamin and Poon Joseph and Rengarajan Hamsawardhini and Susanto Yosephine and Sutaveephamochanon Anocha and Tan Choon Meng and Tan Chor Phin Evelyn and Tan Le Min Sheryl and Tan Siao Wei Jessica and Tan Yixian and Tasawong Panuthep and Tee Jun Yun and Teng Kok Wai Walter and Teo Eng Sipp Leslie and Tjhi William and Tuchinda Pume and Wu Donghang and Yong Xianbin and Zhang Zhou}, year={2026}, eprint={2609.18310}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2609.18310}, } ``` ## Team AI Products Pillar, AI Singapore ## Acknowledgement This project is supported by the National Research Foundation Singapore and Infocomm Media Development Authority (IMDA), Singapore under its National Large Language Model Funding Initiative. ## Contact [**sealion@aisingapore.org**](mailto:sealion@aisingapore.org)