Datasets:
Northeast India Voices
A multilingual transcribed speech corpus covering five indigenous and regional languages of Northeast India: Khasi, Garo, Mizo, Nagamese, and Kokborok.
Dataset Summary
| Language | Family | Utterances |
|---|---|---|
| Khasi | Austroasiatic | 14,974 |
| Nagamese | Indo-Aryan (Creole) | 7,386 |
| Mizo | Tibeto-Burman | 7,250 |
| Kokborok | Tibeto-Burman | 3,278 |
| Garo | Tibeto-Burman | 1,612 |
| Total | 34,500 |
Data Collection
Recorded by native speakers across Northeast India with full informed consent provided in their native language prior to recording.
Contributors
We thank the native speaker contributors from Meghalaya, Mizoram, Nagaland, and Tripura for their time and generosity in making this dataset possible.
Models
- ASR: MWirelabs/ne-asr
- Speech-Text Retrieval: MWirelabs/ne-speechembed
Citation
@dataset{mwirelabs2026northeast,
title = {Northeast India Voices},
author = {MWire Labs},
year = {2026},
publisher = {Hugging Face},
url = {/datasets/Khrawsynth/northeast-india-voices}
}
License
CC-BY-4.0 — MWire Labs
- Downloads last month
- -