{
 "id": "nvidia/Granary-v2",
 "owner": "nvidia",
 "name": "Granary-v2",
 "downloads": 3859,
 "likes": 37,
 "last_modified": "2026-10-28T14:09:30.000Z",
 "stated_licence": "openmdw-1.1",
 "stated_from": "tag",
 "licence_name": null,
 "licence_link": null,
 "file_licence": null,
 "licence_files": [
  "LICENSE"
 ],
 "sources": [],
 "gated": false,
 "tasks": [
  "automatic-speech-recognition",
  "translation",
  "audio-text-to-text",
  "question-answering"
 ],
 "citation": "@techreport{koluguri2026granaryv2,\n  title       = {Granary-v2: Scalable Multilingual Speech Data Curation for Predictable and Controllable Model Outputs},\n  author      = {Nithin Rao Koluguri and Meline Mkrtchyan and Nune Tadevosyan and Aaftab V and Sushmitha Deva and Fan Qian and Davit Karamyan and\n                 Kunal Dhawan and Taejin Park and Piotr {\\.Z}elasko and Arushi Dhingra and Viraj Karandikar and Shuoyang Ding and Slyne Deng and\n                 Onur Yilmaz and Praateek Mahajan and Utkarsh Vaidya and Nikolay Karpov and Jagadeesh Balam and Boris Ginsburg},\n  institution = {NVIDIA},\n  type        = {Technical Report},\n  year        = {2026},\n  url         = {https://research.nvidia.com/labs/nemotron/files/granary-v2.pdf}\n}",
 "signal": "unclear",
 "reason": "The checks could not identify a recognised licence ID in the dataset's metadata.",
 "revision": "81402b751b451b5fe4f4c29e8b40f92892f1e440",
 "stated_evidence": "https://huggingface.co/datasets/nvidia/Granary-v2/blob/81402b751b451b5fe4f4c29e8b40f92892f1e440/README.md",
 "file_evidence": "https://huggingface.co/datasets/nvidia/Granary-v2/blob/81402b751b451b5fe4f4c29e8b40f92892f1e440/LICENSE",
 "card_nc": false
}