Tag

indic-languages.

2 writings found

Latest Archives

NVIDIA and AI4Bharat Release 2.8T Token Indic Dataset

Nemotron-pretraining-Indic-v1 brings synthetic training data across 22 Indian languages, with an open pipeline for developers to generate more.

NVIDIA and AI4Bharat Release 2.8T Token Indic Dataset

NVIDIA and IIT Madras release Nemotron-pretraining-Indic-v1, a 2.8 trillion token synthetic dataset across 22 Indian languages with open-source generation pipeline.
View all rollups →