---

library_name: transformers

pipeline_tag: text-generation

base_model:

• aisingapore/Llama-SEA-LION-v3-70B-IT

new_version: aisingapore/Qwen-SEA-LION-v4.5-27B-IT

language:

• en
• zh
• vi
• id
• th
• fil
• ta
• ms
• km
• lo
• my
• jv
• su

license: llama3.1

---

<div>

<img src="llama_sea_lion_3.5_70b_r_banner.png"/>

</div>

Last updated: 2025-14-04

Llama-SEA-LION-v3.5-70B-R-GGUF

SEA-LION is a collection of Large Language Models (LLMs) which have been pretrained and instruct-tuned

for the Southeast Asia (SEA) region.

Model Description

<!-- Provide a longer summary of what this model is. -->

SEA-LION stands for Southeast Asian Languages In One Network.

Quantization was performed on Llama-SEA-LION-v3.5-70B-R to produce optimized variants that reduce memory requirements

while maintaining model quality. These quantized models support inference on a range of consumer-grade GPUs

and are compatible with various inference engines.

For tokenization, the model employs the default tokenizer used in Llama 3.1-70B-Instruct.

• Developed by: Products Pillar, AI Singapore
• Funded by: Singapore NRF
• Model type: Decoder
• Context length: 128k tokens
• Language(s): Burmese, Chinese, English, Filipino, Indonesian, Javanese, Khmer, Lao, Malay, Sundanese, Tamil, Thai, Vietnamese
• License: Llama 3.1 Community License
• Quantized from model: Llama-SEA-LION-v3.5-70B-R

This repo contains GGUF format models files for aisingapore/Llama-SEA-LION-v3.5-70B-R

Model Weights included in this repository:

• Llama-SEA-LION-v3.5-70B-R-F16
• Llama-SEA-LION-v3.5-70B-R-Q2_K
• Llama-SEA-LION-v3.5-70B-R-Q3_K_M
• Llama-SEA-LION-v3.5-70B-R-Q4_0
• Llama-SEA-LION-v3.5-70B-R-Q4_K_M
• Llama-SEA-LION-v3.5-70B-R-Q5_0
• Llama-SEA-LION-v3.5-70B-R-Q5_K_M
• Llama-SEA-LION-v3.5-70B-R-Q6_K
• Llama-SEA-LION-v3.5-70B-R-Q8_0

> [!NOTE]

> Take note that some GGUFs are split into parts. Most tools such as llama.cpp and those built on it do support split GGUFs,

> pointing the platform to the first split will be sufficient for it to function. In the event where a merge is necessary,

> it can be done using llama.cpp's gguf-split: ./gguf-split --merge ./path/to/first-split ./path/to/output-gguf More details:

> gguf-split guide & README

Lineage & Versioning

This model card serves as an immutable record of the final released checkpoint. Consequently, the hardware specifications, compute hours, and precise dataset volumes (such as final filtered instruction counts) reported here reflect the exact production run used to generate this specific artifact. These figures may differ from the aggregate totals, pre-filtered data pools, or preliminary experimental runs (e.g., initial H100 benchmarks) documented in our accompanying research papers.

The training counts and dataset mixture details reported in this model card reflect the exact constructed training pool consumed during this specific model run. Figures may differ slightly from public dataset releases, which represent downloadable open-source subsets of the broader corpus.

Evaluation

<!-- This section describes the evaluation protocols and provides the results. -->

Test Results

For details on Llama-SEA-LION-v3.5-70B-R performance, please refer to the SEA-HELM leaderboard, Leaderboard results on SEA-HELM.

Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

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

<!-- This section is meant to convey both technical and sociotechnical 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.

More Information

This is the repository for the commercial instruction-tuned model.

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.

AI Singapore is a national programme supported by the National Research Foundation, Singapore and hosted by the National University of Singapore.

Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and

do not reflect the views of the National Research Foundation or the National University of Singapore.

Link to SEA-LION's GitHub repository

For more info, please contact us at sealion@aisingapore.org

Team

Antonyrex Sajeban, Chan Adwin, Cheng Nicholas, Choa Esther, Huang Yuli, Hulagadri Adithya Venkatadri, Lau Wayne, Lee Chwan Ren, Leong Wai Yi, Leong Wei Qi, Liew Rachel, Limkonchotiwat Peerat, Liu Bing Jie Darius,

Montalan Jann Railey, Ng Boon Cheong Raymond, Ngui Jian Gang, Nguyen Thanh Ngan, Ong Brandon, Ong Tat-Wee David,

Ong Zhi Hao, Rengarajan Hamsawardhini, Siow Bryan, Susanto Yosephine, Tai Ngee Chia, Tan Choon Meng, Teng Walter,

Teo Eng Sipp Leslie, Teo Wei Yi, Tjhi William, Yeo Yeow Tong, Yong Xianbin

Contact

sealion@aisingapore.org