asr-africa/wav2vec2-pretrained-asr-africa-naijavoices-hausa-20hr-v1
Automatic Speech Recognition
•
Updated
•
1
You need to register at NaijaVoices to use this dataset. It'll only take a minute
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
Click here to register on NaijaVoices in the tier that corresponds to your use case.
Log in or Sign Up to review the conditions and access this dataset content.
Important Information: please be aware that this version of the dataset is huge (500+ GB) and can therefore be challenging to use. To alleviate this and facilitate adoption, we’ve provided a compressed version (84GB) here. We suggest using that instead if you have compute/storage constraints. They are both the exact same data.
Introduction
Welcome to the NaijaVoices dataset. The NaijaVoices dataset consists of 1,800 hours of authentic speech (from over 5,000 diverse speakers!) and expert curated text in Igbo, Hausa, and Yoruba. ~600 hours for each of the three languages. It also boasts of adequate female representation and balanced age-range distribution (young to old speakers). For more about the dataset info visit our website: https://naijavoices.com/. By using this dataset, you acknowledge reading and accepting to use this dataset accoridng to the data usage terms and conditions.
Metadata
Each row contains the audio data, the transcript, the language of the transcript, the speaker id, along with their gender and age-range. Use the “Dataset Viewer” to look at the dataset.
Usage
To ease usage, we divided the dataset into batches. Specifically, we batched the dataset into their languages, and further into three uniform batches for each language. Each batch should contain approximately 167 hours. We obtained the following batches:
igbo-batch-0igbo-batch-1igbo-batch-2yoruba-batch-0yoruba-batch-1yoruba-batch-2hausa-batch-0hausa-batch-1hausa-batch-2
To load a specific batch run (e.g. “igbo-batch-1”:
from datasets import load_dataset
# If the dataset is gated/private, make sure you have run "huggingface-cli login"
dataset = load_dataset("naijavoices/naijavoices-dataset","igbo-batch-1")
If you want to load the whole dataset, then just do
dataset = load_dataset("naijavoices/naijavoices-dataset")
It is advisable to load the batch in streaming mode by setting streaming=True
# Streaming mode
dataset = load_dataset("naijavoices/naijavoices-dataset", streaming=True)
License
The NaijaVoices dataset is licensed under the CC BY-NC-SA 4.0 license. Basically, It's free for research (non-commercial) use, with proper credit to the community. For commercial interests, please check out our membership tiers which offer special commercial waivers.
- Downloads last month
- 64
Models trained or fine-tuned on naijavoices/naijavoices-dataset