π Kinyarwanda ASR Dataset
This dataset contains transcribed Kinyarwanda audio, designed to support training and evaluation of Automatic Speech Recognition (ASR) systems. It is part of a study on how varying training data volumes affect model performance using Whisper-large-v3.
π Data Overview
The full dataset consists of approximately 263,000 audio samples covering 5 key domains:
- π₯ Health
- ποΈ Government
- π° Financial Services
- π Education
- πΎ Agriculture
To enable analysis of scale effects, the dataset is available in subsets of different durations:
audio_1h
,audio_50h
,audio_100h
,audio_150h
,audio_200h
,audio_500h
,audio_1000h
, andtrain_cleaned
(~1400h)
Each sample includes:
id
: a unique identifier for the sampleaudio
: the audio recordingaudio_language
: the language code (kin
for Kinyarwanda)text
: the corresponding transcription in Kinyarwandaprompt
: the domain or context of the audio (e.g., Health, Education, etc.)duration
: the length of the audio clip in secondsspeaker_id
: an anonymized identifier for the speaker
π οΈ Usage Example
You can load and explore the dataset using the π€ Datasets library:
from datasets import load_dataset
# Load a specific subset, e.g., 100 hours
dataset = load_dataset("evie-8/kinyarwanda-speech-hackathon", "audio_100h")
- Downloads last month
- 1,329