Gemma2 9B CPT Sahabat-AI v1
Sahabat-AI (Indonesian language for “close friends”) is a collection of Large Language Models (LLMs) which has been pretrained and instruct-tuned for Indonesian language and its various dialects. Sahabat-AI ecosystem is co-initiated by Indonesian tech and telecommunication companies: GoTo Group and Indosat Ooredoo Hutchison.
This is the card for the Gemma2 9B CPT Sahabat-AI v1 base model which has undergone continued pre-training from the
Gemma2 9B CPT SEA-Lionv3 base model.
Model Details
Model Description
The continued pre-training data for Gemma2 9B CPT Sahabat-AI v1 base model encompasses approximately 50B tokens.
- Co-initiated by: PT GoTo Gojek Tokopedia Tbk, Indosat Ooredoo Hutchison
- Developed by: PT GoTo Gojek Tokopedia Tbk, AI Singapore
- Model type: Decoder
- Languages: English, Indonesian, Javanese, Sundanese
- License: Gemma Community License
For tokenisation, the model employs the default tokenizer used in Gemma-2-9B. The model has a context length of 8192.
Benchmark Performance
We evaluated Gemma2 9B CPT Sahabat-AI v1 base model on general language capabilities.
General Language Capabilities
For the evaluation of general language capabilities, we employed the
- SEA HELM (also known as BHASA) evaluation benchmark across a variety of tasks.
- These tasks include Question Answering (QA), Sentiment Analysis (Sentiment), Toxicity Detection (Toxicity), Translation in both directions (Eng>Lang & Lang>Eng), Abstractive Summarization (Summ), Causal Reasoning (Causal) and Natural Language Inference (NLI).
- We also added support for Javanese and Sundanese for the BHASA tasks whenever applicable
- and the common English tasks from the HuggingFace LLM Leaderboard.
- These tasks consist of IFEval, BBH, Math Lvl 5, GPQA, MuSR, and MMLU-PRO.
- Caveat: Our results differ from the HuggingFace LLM Leaderboard because we have used VLLM as our inference platform. VLLM caps the context size at 4096 tokens while HuggingFace was set to 8192 tokens.
Note: SEA HELM is implemented using prompts to elicit answers in a strict format. For all tasks, the model is expected to provide an answer tag from which the answer is automatically extracted. For tasks where options are provided, the answer should comprise one of the pre-defined options. The scores for each task is normalised to account for baseline performance due to random chance.
The evaluation was done five-shot with native prompts on a sample of 100-1000 instances for each dataset.
Results
SEA HELM (also known as BHASA)
| Language / Model Name [Base] | Qwen2-7B | Qwen2.5-7B | Llama-3-8B | Llama-3.1-8B | sea-lionv2.1-8B | gemma-2-9B | sea-lionv3-9B | sahabatai-v1-8B | sahabatai-v1-9B |
|---|
| Overall (Bahasa Indonesia + Javanese + Sundanese) | 42.776 | 46.245 | 49.160 | 49.577 | 48.602 | 58.972 | 60.913 | 59.437 | 64.123 |
| Bahasa Indonesia | 49.341 | 55.913 | 47.865 | 48.110 | 49.154 | 58.572 | 62.437 | 53.454 | 60.040 |
| Javanese | 42.774 | 45.917 | 54.627 | 55.215 | 52.728 | 63.760 | 63.363 | 65.048 | 69.882 |
| Sundanese | 36.213 | 36.905 | 44.988 | 45.407 | 43.925 | 54.583 | 56.939 | 59.809 | 62.446 |
English Results
| Model Name [BASE] | Qwen2-7B | Qwen2.5-7B | Llama-3-8B | Llama-3.1-8B | sea-lionv2.1-8B | gemma-2-9B | sea-lionv3-9B | sahabatai-v1-8B | sahabatai-v1-9B |
|---|
| Average | 23.68 | 24.65 | 13.56 | 13.69 | 12.77 | 13.34 | 21.99 | 13.92 | 19.62 |
Training Details
Data
Gemma2 9B CPT Sahabat-AI v1 base model was continued pre-trained on 50B tokens of the following data:
| Data Source | Unique Tokens (B) | Multiplier | Total Tokens (B) | Percentage (%) |
|---|
| Dolma Refined Web | 9.5 | 1 | 9.5 | 18.7 |
| Dolma arXiv | 0.6 | 1 | 0.6 | 1.18 |
| Stack V2 | 5.5 | 1 | 5.5 | 10.85 |
| Dolma Semantic Scholar | 1.2 | 1 | 1.2 | 2.37 |
| Dolma Reddit | 1.7 | 1 | 1.7 | 3.36 |
| Dolma Pes2o | 1.2 | 1 | 1.2 | 2.37 |
| Wiki* + News* - Indonesian | 1.0 | 1 | 1.0 | 1.97 |
| SEA-LION Pile - Indonesian | 27.0 | 1 | 27.0 | 53.3 |
| JV Pile - Javanese | 0.92 | 1.6 | 1.5 | 3.0 |
| SU Pile - Sundanese | 0.39 | 3.8 | 1.5 | 3.0 |
Note:
- All token counts are counted using Gemma2 tokenizer
- Wiki* sources includes Wikipedia, Wiki Books, Wiki Source, Wiki Voyage and Fandom Wiki
- News* sources includes VOA, Global Voices
Infrastructure
Gemma2 9B CPT Sahabat-AI v1 was trained using
MosaicML Composer
on the following hardware:
| Training Details | Gemma2 9B CPT Sahabat-AI v1 |
|---|
| Nvidia H100 80GB GPU | 32 |
| Training Duration | 7 days |
Configuration
| HyperParameter | Gemma2 9B CPT Sahabat-AI v1 |
|---|
| Precision | bfloat16 |
| Optimizer | decoupled_adamw |
| Scheduler | weight_stable_decay |
| Learning Rate | 1.0e-5 |
| Global Batch Size | 256 |
| Micro Batch Size | 1 |
Call for Collaboration
Sahabat-AI (Indonesian language for “close friends”) a local open source Large Language Model (LLM) ecosystem in Indonesian language, co-initiated by Indonesian tech and telecommunication companies: GoTo Group and Indosat Ooredoo Hutchison.
Sahabat-AI ecosystem aims to empower Indonesians who want to develop AI-based services and applications using Bahasa Indonesia and its various local dialects.
We are supported by research centers and global tech experts such as AI Singapore and Tech Mahendra to train the model to gain general language understanding.
We also collaborate with key top Indonesia universities such as University of Indonesia, Gadjah Mada University, Bogor Institute of Agriculture, Bandung Institute of Technology, including top Indonesia media groups, such as Kompas Gramedia Group and Republika to train and enrich the model in Bahasa Indonesia, ensuring optimum provision of local context and cultural relevance.
We would like to invite researchers, developers, and language enthusiasts to actively contribute to the enhancement and expansion of Sahabat-AI.
Your collaborations can involve:
- Identifying and reporting technical issues
- Sharing pre-training, instruction, and preference data
- Improving documentation usability
- Proposing and implementing new model evaluation tasks and metrics
Join us in shaping the future of Sahabat-AI by sharing your expertise and insights to make these models more accessible, accurate, and versatile.
You can contribute your ideas through
this form.The Development Team (in ascending alphabetical order)
AI Singapore
Chan Adwin
Cheng Nicholas
Choa Esther
Huang Yuli
Lau Wayne
Lee Chwan Ren
Leong Wai Yi
Leong Wei Qi
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
PT GoTo Gojek Tokopedia Tbk
Anissa Dininta
Chau Shiau Ching
Choiri Hendra Hadhil
Goel Priyank
Saini Ajay Kumar
Shalev Ofir
Tan Daryl
Tep Kilian Rithi
Tiwari Anupam
Widjojo Daniel
Acknowledgements
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 National Research Foundation, Singapore.
Contact
For more info, please contact us using this
Sahabat-AI Inquiry Form.Disclaimer
This is the repository for the base 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 claim, damages, or other liability arising from the use of the released weights and codes.
