---
title: Granary-v2
canonical_url: "https://www.modelscope.ai/datasets/nv-community/Granary-v2"
md_url: "https://www.modelscope.ai/datasets/nv-community/Granary-v2.md"
repository: nv-community/Granary-v2
last_updated: 2026-10-28
license: CC-BY-4.0
storage_size: "104 GB"
downloads: 162
stars: 0
---

# Granary-v2

> Granary-v2 - An open-source dataset by nv-community on ModelScope. Granary-v2: Multilingual Speech Recognition, Translation and Instruction Data

nv-community/Granary-v2 is a dataset on ModelScope. totalling 104 GB. licensed under CC-BY-4.0.

- **Repository**: nv-community/Granary-v2
- **License**: CC-BY-4.0
- **Storage size**: 104 GB
- **Downloads**: 162
- **Stars**: 0
- **Last updated**: 2026-10-28

Source: https://www.modelscope.ai/datasets/nv-community/Granary-v2

---

<div align="center">
<img src="figures/granary-v2-icon.png" alt="Granary-v2" width="420"/>
</div>

# Granary-v2: Multilingual Speech Recognition, Translation and Instruction Data

<div align="center" style="line-height: 1;">
  <a href="https://developer.nvidia.com/nemotron" target="_blank" style="margin: 2px;">
    <img alt="Homepage" src="https://img.shields.io/badge/%F0%9F%8F%A0Nemotron%20Developer%20Page-Learn%20More%20Here%21-536af5?color=76B900&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
  </a>
  <a href="https://github.com/NVIDIA-NeMo/Curator/tree/main/tutorials/audio/granary_v2/" target="_blank" style="margin: 2px;">
    <img alt="Granary Pipeline" src="https://img.shields.io/badge/%F0%9F%9B%A0%EF%B8%8FGranary%20Pipeline-View%20Code-536af5?color=76B900&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
  </a>
  <a href="https://research.nvidia.com/labs/nemotron/files/granary-v2.pdf" target="_blank" style="margin: 2px;">
    <img alt="Tech Report" src="https://img.shields.io/badge/%F0%9F%93%9DTech%20Report-Read%20Now%21-536af5?color=76B900&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
  </a>
  <a href="https://discord.gg/9xpKQtVvrk" target="_blank" style="margin: 2px;">
    <img alt="Discord" src="https://img.shields.io/badge/Discord-NVIDIA%20AI%20Developer-7289da?logo=discord&logoColor=white&color=7289da" style="display: inline-block; vertical-align: middle;"/>
  </a>
</div>

<div align="center" style="line-height: 1;">
  <a href="#granary-v2-license" style="margin: 2px;">
    <img alt="License: OpenMDW-1.1 / CC BY-SA 4.0" src="https://img.shields.io/badge/License-OpenMDW--1.1%20%2F%20CC%20BY--SA%204.0-f5de53" style="display: inline-block; vertical-align: middle;"/>
  </a>
</div>

**Granary-v2** is NVIDIA's second release of generated labels and instruction pairs
for multilingual speech. It re-processes a broad set of source corpora through
an entirely new labelling pipeline that produces not just transcripts and translations, but
inverse text normalization, captioning, code-switching and speech-QA targets, plus
ready-to-train instruction pairs.

<div align="center">
<img src="figures/granary-v2-pipeline.gif" alt="Animated walkthrough: one audio sample moves through segmentation, two ASR backbones, and text enrichment" width="1364"/>
</div>

> [!NOTE]
> **The pipeline source is already public.** All of it lives in
> [NeMo Curator](https://github.com/NVIDIA-NeMo/Curator/tree/main/tutorials/audio/granary_v2).
> The full end-to-end runs and the per-stage scripts will follow soon.

## Overview

- **48 languages** across **16 source corpora**
- Labels for **463,062 hours** of source speech
- **approximately 165,300 hours** flagged high-quality by the pipeline's own gate
- **approximately 92,496 hours** of X→English speech translation
- **approximately 39,375 hours** of English→X speech translation, each record fanned out
  to 31 targets at once (`ar`, `bg`, `cs`, `da`, `de`, `el`, `es`, `et`, `fi`, `fr`, `he`, `hi`, `hr`, `hu`, `it`, `ja`, `ko`, `lt`, `lv`, `mt`, `nl`, `pl`, `pt`, `ro`, `ru`, `sk`, `sl`, `sv`, `th`, `uk`, `zh`)
- **Six text targets per utterance**: normalized transcript, ITN, captioning,
  code-switched, speech-QA, and translation

### Supported Languages

Arabic, Assamese, Bengali, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Filipino, Finnish, French, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Kannada, Korean, Latvian, Lithuanian, Malayalam, Maltese, Marathi, Odia, Persian, Polish, Portuguese, Punjabi, Romanian, Russian, Slovak, Slovenian, Spanish, Swedish, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, Vietnamese.

## Hours by corpus

| Corpus | Languages | Hours | High-quality h | AST (X→en) h | AST (en→X) h | Source corpus license |
|---|---:|---:|---:|---:|---:|---|
| **[VoxPopuli (MOSEL labels)](https://mt.fbk.eu/mosel/)** (`voxpopuli`) | 23 | 380,711 | 123,192 | 84,999 | 8,990 | CC0 / CC-BY-4.0 labels |
| **[LibriLight](https://github.com/facebookresearch/libri-light)** (`librilight`) | 1 | 49,334 | 25,898 | 0 | 25,725 | MIT |
| **[People's Speech](https://mlcommons.org/en/peoples-speech/)** (`people_speech`) | 1 | 8,598 | 3,807 | 0 | 826 | [CC BY (`clean` partition)](https://huggingface.co/datasets/MLCommons/peoples_speech#licensing-information) |
| **[Multilingual LibriSpeech](https://www.openslr.org/94/)** (`mls`) | 8 | 7,805 | 3,073 | 1,848 | 103 | CC-BY-4.0 |
| **[FalAR](https://huggingface.co/datasets/inesc-id/FalAR)** (`falar`) | 1 | 4,715 | 1,605 | 1,602 | — | CC-BY-4.0 |
| **[Common Voice 26.0](https://commonvoice.mozilla.org/en/datasets)** (`mcv26`) | 39 | 3,871 | 2,837 | 2,372 | 496 | CC0-1.0 |
| **[National Speech Corpus](https://www.imda.gov.sg/NationalSpeechCorpus)** (`nsc`) | 1 | 2,785 | 1,647 | 0 | 1,689 | Singapore Open Data Licence |
| **[VoxPopuli (re-labelled)](https://github.com/facebookresearch/voxpopuli)** (`voxpopuli_improved`) | 16 | 1,277 | 527 | 562 | 160 | CC0 / CC-BY-4.0 labels |
| **[Golos](https://github.com/salute-developers/golos)** (`golos`) | 1 | 1,208 | 715 | 744 | — | Custom (Salute Developers) |
| **[Europarl-ASR](https://www.mllp.upv.es/git-pub/ggarces/Europarl-ASR/)** (`europarl`) | 1 | 1,023 | 605 | 0 | 613 | CC-BY-4.0 |
| **[LibriSpeech-PC](https://www.openslr.org/145/)** (`librispeech_pc`) | 1 | 953 | 953 | 0 | 686 | CC-BY-4.0 |
| **[FLEURS](https://huggingface.co/datasets/google/fleurs)** (`fleurs`) | 48 | 412 | 213 | 226 | 5 | CC-BY-4.0 |
| **[Dusha](https://github.com/salute-developers/golos/tree/master/dusha)** (`dusha`) | 1 | 204 | 104 | 143 | — | Custom (Salute Developers) |
| **[VCTK](https://datashare.ed.ac.uk/handle/10283/3443)** (`vctk`) | 1 | 82 | 76 | 0 | 76 | ODC-By 1.0 |
| **[AMI Meeting Corpus](https://lhotse.readthedocs.io/en/stable/_modules/lhotse/recipes/ami.html)** (`ami`) | 1 | 70 | 24 | 0 | 4 | CC-BY-4.0 |
| **[Google Speech Commands](https://www.tensorflow.org/datasets/catalog/speech_commands)** (`google_speech_commands`) | 1 | 15 | 15 | 0 | 2 | CC-BY-4.0 |
| **Total** | **48** | **463,062** | **~165,300** | **92,496** | **39,375** | |

Totals describe only the manifests in this public release. ASR hours and translation into English are calculated from the reported corpus durations. High-quality and translation-out totals are approximate because the source values are rounded. Translation targets do not add audio hours.

The table lists source-corpus licenses. The released labels and instruction pairs are governed by [OpenMDW-1.1](LICENSE), except for the `people_speech` subset, which is governed by [CC BY-SA 4.0](LICENSES/CC-BY-SA-4.0.txt). See [License](#license) for the subset mapping.

This repository contains label manifests only. Obtain audio from the linked source distributors under their applicable licenses and access terms.

## 🎧 Getting the audio

The commands below obtain source corpora from their distributors. Check the source licenses and access terms before downloading. Loading `nvidia/Granary-v2` downloads label manifests only; obtain source audio separately.

This repository ships **manifests only** — no audio travels with it. Every
`audio_filepath` is instead a self-describing identifier:

```
data_<lang>_<corpus>_<source path, with "/" replaced by "_">
```

So `data_de_mcv26_cv-corpus-26.0-2026-06-12_de_clips_common_voice_de_21137340.opus`
is `cv-corpus-26.0-2026-06-12/de/clips/common_voice_de_21137340.opus` in the Common
Voice 26.0 release. Fetch a corpus from its own distributor, then join on that
suffix.

Source audio remains subject to its distributor's terms. See the Source corpus license column in [Hours by corpus](#hours-by-corpus). See [License](#license) for the terms that apply to the released labels and instruction pairs.

The four routes below are ordered by how much work the download takes. Open the one
your corpus is in.

<table>
<tr>
<td valign="top" width="25%">
<b>1 · One HF command</b><br><br>
<code>people_speech</code> · 8,598 h<br>
<code>falar</code> · 4,715 h<br>
<code>fleurs</code> · 412 h
</td>
<td valign="top" width="25%">
<b>2 · Direct download</b><br><br>
<code>librilight</code> · 49,334 h<br>
<code>mls</code> · 7,805 h<br>
<code>golos</code> · 1,208 h<br>
<code>librispeech_pc</code> · 953 h<br>
<code>dusha</code> · 204 h<br>
<code>vctk</code> · 82 h<br>
<code>google_speech_commands</code> · 15 h
</td>
<td valign="top" width="25%">
<b>3 · Corpus tooling</b><br><br>
<code>voxpopuli</code> · 380,711 h<br>
<code>voxpopuli_improved</code> · 1,277 h<br>
<code>europarl</code> · 1,023 h<br>
<code>ami</code> · 70 h
</td>
<td valign="top" width="25%">
<b>4 · Gated access</b><br><br>
<code>mcv26</code> · 3,871 h<br>
<code>nsc</code> · 2,785 h
</td>
</tr>
</table>

<details>
<summary><b>1 · One Hugging Face command</b></summary>

```bash
pip install -U "huggingface_hub[cli]" datasets
```

| Corpus | Hours | How |
|---|---:|---|
| **People's Speech** (`people_speech`) | 8,598 | `hf download MLCommons/peoples_speech --repo-type dataset --include "clean/*"` — we release labels for the government-origin subset of `clean`. The released manifests are governed by [CC BY-SA 4.0](LICENSES/CC-BY-SA-4.0.txt). Obtain source audio under the distributor's terms. |
| **FalAR** (`falar`) | 4,715 | `load_dataset("inesc-id/FalAR")` — the audio is embedded in the Parquet shards, so load it rather than fetching files |
| **FLEURS** (`fleurs`) | 412 | `load_dataset("google/fleurs", "ar_eg")` — the config is `<lang>_<region>` |

</details>

<details>
<summary><b>2 · Direct download</b></summary>

| Corpus | Hours | Archive |
|---|---:|---|
| **LibriLight** (`librilight`) | 49,334 | `https://dl.fbaipublicfiles.com/librilight/data/large.tar` (3.06 TB) — we use the **`large`** 60k-hour segmentation; `small` and `medium` sit alongside it in the [repo](https://github.com/facebookresearch/libri-light) |
| **Multilingual LibriSpeech** (`mls`) | 7,805 | `https://dl.fbaipublicfiles.com/mls/mls_<lang>.tar.gz` for `english`, `german`, `dutch`, `french`, `spanish`, `italian`, `portuguese`, `polish` — an `_opus` variant of each exists; full index at [OpenSLR 94](https://www.openslr.org/94/) |
| **Golos** (`golos`) | 1,208 | `https://cdn.chatwm.opensmodel.sberdevices.ru/golos/train_crowd{0..9}.tar` (127 GB total) — we use the **crowd** partition, and `crowd9` carries the manifests. Whole set in Opus: `golos_opus.tar` |
| **LibriSpeech-PC** (`librispeech_pc`) | 953 | The audio is plain LibriSpeech: `https://www.openslr.org/resources/12/train-{clean-100,clean-360,other-500}.tar.gz`. [OpenSLR 145](https://www.openslr.org/145/) adds only the punctuated and capitalised text |
| **Dusha** (`dusha`) | 204 | `https://cdn.chatwm.opensmodel.sberdevices.ru/dusha/crowd.tar` (30 GB) — **crowd only**; the podcast partition ships no audio for licensing reasons and we use none of it |
| **VCTK** (`vctk`) | 82 | `https://datashare.ed.ac.uk/bitstream/handle/10283/3443/VCTK-Corpus-0.92.zip` (11.7 GB) |
| **Google Speech Commands** (`google_speech_commands`) | 15 | `http://download.tensorflow.org/data/speech_commands_v0.02.tar.gz` (2.4 GB) |

</details>

<details>
<summary><b>3 · Needs the corpus's own tooling</b></summary>

**VoxPopuli** — the two VoxPopuli leaves come from the same repository but
**different subsets**, so they take different commands. `voxpopuli` (380,711 h) is
the **v2 unlabelled** release, carrying the MOSEL pseudo-labels that seeded ours
([`FBK-MT/mosel`](https://huggingface.co/datasets/FBK-MT/mosel)); `voxpopuli_improved`
(1,277 h) is the transcribed ASR subset, re-labelled by us.

> [!IMPORTANT]
> Mind the `_v2` suffix. `--subset bg` is the v1 release and is roughly **4× smaller**
> than the `bg_v2` our manifests were built from.

```bash
git clone https://github.com/facebookresearch/voxpopuli && cd voxpopuli
pip install -r requirements.txt

# `voxpopuli` -> the v2 UNLABELLED subset, per language ("400k" fetches all 23 at once)
python -m voxpopuli.download_audios     --root ./data --subset bg_v2
python -m voxpopuli.get_unlabelled_data --root ./data --subset bg_v2

# `voxpopuli_improved` -> the transcribed ASR subset
python -m voxpopuli.download_audios --root ./data --subset asr
python -m voxpopuli.get_asr_data    --root ./data --lang cs
```

**Europarl-ASR** (`europarl`, 1,023 h) — clone the MLLP repository and run its fetch
scripts:

```bash
git clone https://www.mllp.upv.es/git-pub/ggarces/Europarl-ASR.git
```

**AMI** (`ami`, 70 h) — download through the
[Lhotse recipe](https://lhotse.readthedocs.io/en/stable/_modules/lhotse/recipes/ami.html)
with the **IHM** (individual headset microphone) setting; every record we ship is an
individual headset channel, so SDM/MDM audio will not align.

```bash
pip install lhotse && lhotse download ami --mic ihm ./ami
```

</details>

<details>
<summary><b>4 · Gated access</b></summary>

| Corpus | Hours | What it takes |
|---|---:|---|
| **Common Voice 26.0** (`mcv26`) | 3,871 | Download the per-language `cv-corpus-26.0-2026-06-12` archive from the [Common Voice portal](https://commonvoice.mozilla.org/en/datasets) (email-gated); the `mozilla-foundation/common_voice_*` mirrors on the Hub lag the latest release |
| **National Speech Corpus** (`nsc`) | 2,785 | Request access from [IMDA Singapore](https://www.imda.gov.sg/how-we-can-help/national-speech-corpus) — there is no direct download |

</details>

## Repository Structure

```
nvidia/Granary-v2/
└── <language>/                 # ISO 639-1 codes
    └── <corpus>/
        └── <lang>_asr.jsonl.gz
```

One or more manifest files per (language, corpus). Large manifests can be split into shards.

Unlike Granary-v1, there is no separate `<lang>_ast-en.jsonl`: every translation travels
inline in each record's `translations` field, alongside its quality-estimation score. The
manifest files for each (language, corpus) carry both the ASR and the AST targets.

`translations` is keyed by target language, and the release runs in **both directions**.
Non-English audio is translated into English, which is the `AST (X→en) h` column. English
audio is translated the other way, into 31 targets at once (`ar`, `bg`, `cs`, `da`, `de`, `el`, `es`, `et`, `fi`, `fr`, `he`, `hi`, `hr`, `hu`, `it`, `ja`, `ko`, `lt`, `lv`, `mt`, `nl`, `pl`, `pt`, `ro`, `ru`, `sk`, `sl`, `sv`, `th`, `uk`, `zh`) — that is the
`AST (en→X) h` column, and it is why corpora like LibriLight and VCTK show no X→English
hours yet still carry translation.

Coverage is per-record, and a present key is **not** a usable translation: the pipeline
writes a stub with an empty `translation_raw` where its quality estimator rejected the
candidate. **Test `translation_raw` for non-emptiness** rather than assuming a key means
a translation. Both table columns count distinct audio, once per record however many
targets it carries, so neither double-counts and the two never sum past a corpus's hours.

## 🚀 Quick Start

These examples load the released label manifests. Obtain the source audio separately as described in [Getting the audio](#-getting-the-audio). The `people_speech` records retain their CC BY-SA 4.0 license when loaded through the combined `en` configuration; see [License](#license).

```python
from datasets import load_dataset

# 🌍 Language-level access (every corpus for that language)
ds = load_dataset("nvidia/Granary-v2", "de")

# 🎯 Corpus-specific access, named <lang>_<corpus>
ds = load_dataset("nvidia/Granary-v2", "de_mcv26")
ds = load_dataset("nvidia/Granary-v2", "en_librilight")
ds = load_dataset("nvidia/Granary-v2", "ja_mcv26")

# 📡 Streaming for large configs
ds = load_dataset("nvidia/Granary-v2", "en", streaming=True)
```

## 📊 Data Sample Structure

```python
{
    "audio_filepath": str,            # Identifier used to locate source audio
    "duration": float,                # Duration in seconds
    "corpus": str,                    # source corpus name
    "source_lang": str,               # Source language code, e.g. "de"

    "tn_raw": str,                    # Primary transcript, text-normalized with P&C
    "itn_raw": str,                   # Inverse-text-normalized transcript
    "itn_clean": str,                 # ITN transcript after cleanup
    "captioning_text": str,           # Natural-language description of the audio
    "code_switched_text": str,        # Transcript with code-switching preserved
    "speech_qa_text": str,            # Q/A pair grounded in the audio ("SKIP" if N/A)

    "context_asr": dict | None,       # Coarse/fine context terms and entity categories
    "translations": dict,             # {"en": {"translation_raw": str,
                                      #          "translation_quality_score": float}}
    "preference_instructions": list,  # [{"prompt": str, "target": str, "tags": {}}, ...]
    "high_quality": bool | None,      # pipeline quality gate -- see note below
}
```

The pipeline uses 16 kHz mono audio for labeling. Source distributors can provide audio in other formats.

> **On `high_quality`.** The flag records the pipeline's quality check.
> Approximately 165,300 hours in this public release pass it. Context and
> translation coverage vary by record. Early translation runs also included
> records outside the high-quality subset. Check each target for a usable value.

`preference_instructions` is a **derived training view**, not a core target: it packs
the fields above into ready-to-train prompt/target pairs for instruction tuning. The
pairs add no information beyond the other fields, so ignore them if you are training a
plain ASR or AST model.

<a id="granary-v2-license"></a>

## 📄 License

This dataset is governed by the OpenMDW License Agreement, version 1.1, except for the People_Speech subset, which is governed by the Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).

- **OpenMDW-1.1:** See [LICENSE](LICENSE) and the [official agreement](https://openmdw.ai/license/1-1/).
- **People_Speech exception:** See [CC BY-SA 4.0](LICENSES/CC-BY-SA-4.0.txt) and the [official license](https://creativecommons.org/licenses/by-sa/4.0/).

`People_Speech` refers to the released `people_speech` corpus, stored at `en/people_speech/en_asr.jsonl.gz` and available through configuration `en_people_speech`. CC BY-SA 4.0 also applies to these records when they are accessed through the combined `en` configuration. OpenMDW-1.1 applies to the other released manifests. See [NOTICE](NOTICE) for the subset mapping.

Obtain source audio separately from its distributors under the applicable upstream terms. The licenses above describe the dataset materials released in this repository.

## 📚 Citation

If you use Granary-v2, please cite the [NVIDIA technical report](https://research.nvidia.com/labs/nemotron/files/granary-v2.pdf) below.
The report describes the curation system and distinguishes the full processed
collection from the public release subset.

```bibtex
@techreport{koluguri2026granaryv2,
  title       = {Granary-v2: Scalable Multilingual Speech Data Curation for Predictable and Controllable Model Outputs},
  author      = {Nithin Rao Koluguri and Meline Mkrtchyan and Nune Tadevosyan and Aaftab V and Sushmitha Deva and Fan Qian and Davit Karamyan and
                 Kunal Dhawan and Taejin Park and Piotr {\.Z}elasko and Arushi Dhingra and Viraj Karandikar and Shuoyang Ding and Slyne Deng and
                 Onur Yilmaz and Praateek Mahajan and Utkarsh Vaidya and Nikolay Karpov and Jagadeesh Balam and Boris Ginsburg},
  institution = {NVIDIA},
  type        = {Technical Report},
  year        = {2026},
  url         = {https://research.nvidia.com/labs/nemotron/files/granary-v2.pdf}
}
```

For background on Granary-v1, see the Granary-v1 paper in the References section.

## Dataset Description:

Granary-v2 provides synthetic text labels and instruction pairs for multilingual speech. The repository contains compressed JSONL manifests. Obtain the source audio from the distributors listed in the [Hours by corpus](#hours-by-corpus) table. The Overview and data sample sections above describe the available targets and fields.

## Dataset Owner(s):

NVIDIA Corporation

## Dataset Creation Date:

October 28, 2026

## Version: 2.0

Previous Version(s): [Granary-v1 (1.0)](https://huggingface.co/datasets/nvidia/Granary).

Relationship to Previous Version(s): Granary-v2 replaces the labelling pipeline and expands the release to 48 languages and 16 source entries. It adds inverse text normalization, captioning, code-switching, speech-QA, context terms, and instruction pairs. Translations are stored in each record's `translations` field, keyed by target language, and run in both directions: non-English audio into English, and English audio into 31 target languages.

## License/Terms of Use:

This dataset is governed by the OpenMDW License Agreement, version 1.1, except for the People_Speech subset, which is governed by the Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).

See [License](#license), [LICENSE](LICENSE), [CC BY-SA 4.0](LICENSES/CC-BY-SA-4.0.txt), and [NOTICE](NOTICE) for the license texts and subset mapping. Obtain source audio separately from its distributors under their applicable terms.

## Intended Usage:

Use the dataset for automatic speech recognition, speech translation in both directions (non-English audio into English, and English audio into 31 target languages), and speech instruction tuning. Additional targets support inverse text normalization, audio captioning, code-switching, speech question answering, and contextual speech recognition. Target availability varies by record; see the Data Sample Structure section above.

## Dataset Characterization

- **Data Collection Method:** Manually-Collected. This release uses existing source corpora. The repository distributes manifests; obtain the audio from the source distributors.
- **Labeling Method:** Synthetic. The pipeline produces labels with automatic speech and text models.

## Dataset Format

The repository contains compressed JSONL manifests (`.jsonl.gz`). Each line contains one utterance record. The manifest files for each language and corpus contain the available ASR and speech translation targets. Translations are stored in the `translations` field, keyed by target language.

See [Repository Structure](#repository-structure), Quick Start, and Data Sample Structure above for file paths, loading examples, and record fields.

## Dataset Quantification

| Measure | Value |
|---|---:|
| Languages | 48 |
| Source entries | 16 |
| ASR hours | 463,062 |
| Hours marked high-quality by the pipeline | approximately 165,300 |
| Speech translation hours (X→English) | 92,496 |
| Speech translation hours (English→X) | 39,375 |

Speech translation uses audio already included in the ASR total. See [Hours by corpus](#hours-by-corpus) above for the corpus distribution.

## Sources

The [Hours by corpus](#hours-by-corpus) table lists all 16 source entries, with links to their sources and their source-corpus licenses. This repository distributes manifests. Obtain the audio from each source distributor under that corpus's terms.

## Reference(s):

- Nithin Rao Koluguri, Meline Mkrtchyan, Nune Tadevosyan, et al. (2026). [*Granary-v2: Scalable Multilingual Speech Data Curation for Predictable and Controllable Model Outputs*](https://research.nvidia.com/labs/nemotron/files/granary-v2.pdf). NVIDIA technical report.
- [Granary-v1 dataset](https://huggingface.co/datasets/nvidia/Granary).
- [Granary-v1 paper: Speech Recognition and Translation Dataset in 25 European Languages](https://arxiv.org/abs/2505.13404) — describes the v1 pipeline, which this release supersedes.
- [Granary-v2 curation pipeline in NeMo Curator](https://github.com/NVIDIA-NeMo/Curator/tree/main/tutorials/audio/granary_v2).
- [NVIDIA NeMo Toolkit](https://github.com/NVIDIA/NeMo) — for training on these manifests.

The Citation section above provides the Granary-v2 NVIDIA technical report citation. The Granary-v1 paper describes the earlier release.

## Ethical Considerations:

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. Developers should work with their internal developer teams to ensure this dataset meets requirements for the relevant industry and use case and addresses unforeseen product misuse.

Please report quality, risk, security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).