MonsoonASR-ro-RO / README.md
manasdhir04's picture
Remove MOS field
dfb7e97 verified
|
Raw History Blame Contribute Delete
1.3 kB
---
pretty_name: MonsoonASR-Test-ro-RO
language:
- ro
license: cc-by-nc-4.0
task_categories:
- automatic-speech-recognition
size_categories:
- n<1K
---
# MonsoonASR-Test-ro-RO
## Data fields
| Field | Type | Description |
|---|---|---|
| `audio` | Audio | The speech clip (FLAC). |
| `id` | string | Unique clip identifier. |
| `text` | string | Human reference transcript, verbatim (disfluencies such as "um", "uh", repetitions and false starts are kept), normalized with the scorer's normalizer: lowercased, punctuation removed; apostrophes and number separators kept. |
| `audio_length_s` | float | Clip duration in seconds. |
| `speaker_id` | int | Anonymised speaker identifier, consistent across all clips from the same speaker. |
| `gender` | string | Speaker gender: `male` or `female`. |
| `age` | int | Speaker age in years at the time of recording. |
| `current_city` | string | Speaker's city of residence as "city, region, district". Empty when not provided. |
## Scoring
Results on the Voice Arena Conversational ASR Leaderboard use LLM-based semantic WER.
To score your own model on this set, or to contribute, use the official scorer:
- GitHub: https://github.com/VoiceArenaAI/Semantic-WER-VA
- PyPI: https://pypi.org/project/va-semantic-wer/ (`pip install va-semantic-wer`)