Buckets:
112 MB
4 files
Updated about 1 month ago
Ctrl+K
| Name | Size | Uploaded | Xet hash |
|---|---|---|---|
| data | 2 items | ||
| .gitattributes | 2.5 kB xet | 738f1125 | |
| README.md | 1.19 kB xet | 348e0f7a |
NE-TTS Idu Mishmi (clk)
Cleaned TTS dataset for Idu Mishmi (clk), a North East Indian language. Derived from the Vaani dataset with SNR filtering, LUFS normalization, and text cleaning.
Stats
| Metric | Value |
|---|---|
| Total clips | 959 |
| Total hours | 1.0h |
| High SNR (>=20dB) | 602 clips |
| Medium SNR (15-20dB) | 357 clips |
| Sample rate | 16kHz |
| Audio format | WAV, 16-bit PCM |
Schema
| Column | Type | Description |
|---|---|---|
| audio | Audio | 16kHz WAV audio |
| text | string | Cleaned transcript (Latin script) |
| language | string | ISO 639-3 code (clk) |
| gender | string | Speaker gender |
| district | string | Recording district |
| speaker_group | string | {gender}_{district} |
| snr | float | Estimated SNR in dB |
| duration | float | Clip duration in seconds |
| snr_tier | string | "high" (>=20dB) or "medium" (15-20dB) |
License
CC-BY-4.0 (inherited from Vaani dataset).
- Total size
- 112 MB
- Files
- 4
- Last updated
- Aug 25
- Pre-warmed CDN
- US EU US EU