NVIDIA and IIT Madras Release 2.8T-Token Indic Pretraining Dataset

NVIDIA and IIT Madras' AI4Bharat lab released Nemotron-pretraining-Indic-v1, about 2.8 trillion tokens covering 22 constitutionally recognized Indian languages, plus an open GPU-accelerated data generation pipeline with prompts, LLM workflows and quality checks that developers can reuse to produce more data.

• Nvidia CEO on AI Market Turmoil: 'All of This Is Going to Pass'2026-10-16
• Reflection AI CEO says AI risks 'certainly worry me'2026-10-16
• Ex-Nvidia exec: robotics still in its 'Lego era'2026-10-16
• Huang and Tan See the AI Boom From Slightly Different Places2026-10-16
• Nvidia releases FP4 quant of Qwen3.8-Flash-Next2026-10-16