Loading…
Nemotron-Personas-India: Synthesized Data for Sovereign AI
Hugging FaceKiran Praveen, Utkarsh Vaidya, Evan A, Lipika Ramaswamy, Dhruv Nathawani, Dane Corneil, Yev Meyer
Summary
Most open artificial intelligence datasets reflect Western norms and English-only contexts, limiting model effectiveness across India's multilingual and multi-script landscape. To address this shortage of culturally grounded training data, NVIDIA released Nemotron-Personas-India under a CC BY 4.0 license. The synthetic dataset contains 21 million personas across 3 million records spanning 7.7 billion tokens in English, Devanagari Hindi, and Latin-script Hindi. The generation pipeline utilized NeMo Data Designer, an Apache-2.0 licensed Probabilistic Graphical Model for demographic grounding against 2011 Census distributions, and GPT-OSS-120B for narrative text. All records are entirely synthetic to eliminate re-identification risks while enabling fine-tuning for region-aware AI agents and domain-specific copilots.
Context
Most open datasets reflect Western norms and English-only contexts, creating a data gap that limits AI adoption in India's multilingual, multi-script environment. Without representative datasets, AI systems struggle with code-switching, fail to understand regional occupations, and miss cultural context.
Approach / What changed
NVIDIA generated Nemotron-Personas-India using NeMo Data Designer with Jinja templating, Pydantic validation, structured outputs, automated retries, an Apache-2.0 licensed Probabilistic Graphical Model for statistical grounding to official 2011 Census and Electoral Rolls distributions, and GPT-OSS-120B for narrative generation.
Takeaways
- Nemotron-Personas-India contains 21 million total synthetic personas across 3 million records and 7.7 billion tokens, licensed under CC BY 4.0.
- The dataset covers English, Hindi in Devanagari script (4.7B tokens), and Hindi in Latin script (2B tokens) across all 36 Indian states and 640 districts.
- The pipeline relies on a Probabilistic Graphical Model for demographic statistical alignment and GPT-OSS-120B for synthetic narrative generation without real personal data.
Related reading
How to Build a Healthcare Robot from Simulation to Deployment with NVIDIA Isaac for Healthcare
Simulation in healthcare robotics has historically been slow, siloed, and difficult to translate to physical hardware. NVIDIA introduced the Isaac for Healthcare v0.4 developer framework to streamline this lifecycle through an end-to-end SO-ARM starter workflow. Developers collect mixed real-world and synthetic datasets using LeRobot and SO-101 teleoperation, with synthetic generation accounting for over 93 percent of training data. The pipeline post-trains GR00T N1.5 models using dual-camera vision inputs, evaluates policies in Isaac Lab, and deploys them to physical hardware with RTI DDS communication. This integrated approach allows surgical assistant robots to interpret natural language instructions and execute physical manipulation tasks reliably.
AsawareeSynthetic Data Generation for Financial AI Research with NVIDIA NeMo