README.md

---
license:
  - cc-by-sa-4.0
  - cc-by-4.0
annotation_creators:
  - human-annotated
  - crowdsourced
language_creators:
  - creator_1
tags:
  - audio
  - automatic-speech-recognition
  - text-to-speech
language:
  - ach
  - aka
  - dag
  - dga
  - ewe
  - fat
  - ful
  - hau
  - ibo
  - kpo
  - lin
  - lug
  - mas
  - mlg
  - nyn
  - sna
  - sog
  - swa
  - twi
  - yor
multilinguality:
  - multilingual
pretty_name: Waxal NLP Datasets
task_categories:
  - automatic-speech-recognition
  - text-to-speech
source_datasets:
  - UGSpeechData
  - DigitalUmuganda/AfriVoice
  - original
configs:
  - config_name: asr
    data_files:
      - split: train
        path: "data/ASR/**/*-train-*"
      - split: validation
        path: "data/ASR/**/*-validation-*"
      - split: test
        path: "data/ASR/**/*-test-*"
      - split: unlabeled
        path: "data/ASR/**/*-unlabeled-*"
  - config_name: tts
    data_files:
      - split: train
        path: "data/TTS/**/*-train-*"
      - split: validation
        path: "data/TTS/**/*-validation-*"
      - split: test
        path: "data/TTS/**/*-test-*"
dataset_info:
  - config_name: asr
    features:
      - name: id
        dtype: string
      - name: speaker_id
        dtype: string
      - name: transcription
        dtype: string
      - name: language
        dtype: string
      - name: gender
        dtype: string
      - name: audio
        dtype: audio
  - config_name: tts
    features:
      - name: id
        dtype: string
      - name: speaker_id
        dtype: string
      - name: transcription
        dtype: string
      - name: locale
        dtype: string
      - name: gender
        dtype: string
      - name: audio
        dtype: audio
---

# Waxal Datasets

## Table of Contents

-   [Dataset Description](#dataset-description)
    -   [ASR Dataset](#asr-dataset)
    -   [TTS Dataset](#tts-dataset)
    -   [How to Use](#how-to-use)
-   [Dataset Structure](#dataset-structure)
    -   [ASR Data Fields](#asr-data-fields)
    -   [TTS Data Fields](#tts-data-fields)
    -   [Data Splits](#data-splits)
-   [Dataset Curation](#dataset-curation)
-   [Considerations for Using the Data](#considerations-for-using-the-data)
-   [Additional Information](#additional-information)

## Dataset Description

The Waxal project provides datasets for both Automated Speech Recognition (ASR)
and Text-to-Speech (TTS) for African languages. The goal of this dataset's
creation and release is to facilitate research that improves the accuracy and
fluency of speech and language technology for these underserved languages, and
to serve as a repository for digital preservation.

The Waxal datasets are collections acquired through partnerships with Makerere
University, The University of Ghana, Digital Umuganda, and Media Trust.
Acquisition was funded by Google and the Gates Foundation under an agreement to
make the dataset openly accessible.

### ASR Dataset

The Waxal ASR dataset is a collection of data in 14 African languages. It
consists of approximately 1,250 hours of transcribed natural speech from a wide
variety of voices. The 14 languages in this dataset represent over 100 million
speakers across 40 Sub-Saharan African countries.

Provider            | Languages                                | License
:------------------ | :--------------------------------------- | :------------:
Makerere University | Acholi, Luganda, Masaaba, Nyankole, Soga | `CC-BY-4.0`
University of Ghana | Akan, Ewe, Dagbani, Dagaare, Ikposo      | `CC-BY-NC-4.0`
Digital Umuganda    | Fula, Lingala, Shona, Malagasy           | `CC-BY-4.0`

### TTS Dataset

The Waxal TTS dataset is a collection of text-to-speech data in 10 African
languages. It consists of approximately 240 hours of scripted natural speech
from a wide variety of voices.

Provider            | Languages                            | License
:------------------ | :----------------------------------- | :------------:
Makerere University | Acholi, Luganda, Kiswahili, Nyankole | `CC-BY-4.0`
University of Ghana | Akan (Fante, Twi)                    | `CC-BY-NC-4.0`
Media Trust         | Fula, Igbo, Hausa, Yoruba            | `CC-BY-4.0`

### How to Use

The `datasets` library allows you to load and pre-process your dataset in pure
Python, at scale.

First, ensure you have the necessary dependencies installed to handle audio
data:

```bash
pip install datasets[audio]
```

**Loading ASR Data**

To load ASR data, point to the `data/ASR` directory.

```python
from datasets import load_dataset, Audio

# Load Shona (sna) ASR dataset
asr_data = load_dataset("google/WaxalNLP", "sna", data_dir="data/ASR")

# Access splits
train = asr_data['train']
val = asr_data['validation']
test = asr_data['test']

# Example: Accessing audio bytes and other fields
example = train[0]
print(f"Transcription: {example['transcription']}")
print(f"Sampling Rate: {example['audio']['sampling_rate']}")
# 'array' contains the decoded audio bytes as a numpy array
print(f"Audio Array Shape: {example['audio']['array'].shape}")
```

**Loading TTS Data**

To load TTS data, point to the `data/TTS` directory.

```python
from datasets import load_dataset

# Load Swahili (swa) TTS dataset
tts_data = load_dataset("google/WaxalNLP", "swa", data_dir="data/TTS")

# Access splits
train = tts_data['train']
```

## Dataset Structure

### ASR Data Fields

```python
{
  'id': 'sna_0',
  'speaker_id': '...',
  'audio': {
    'array': [...],
    'sample_rate': 16_000
  },
  'transcription': '...',
  'language': 'sna',
  'gender': 'Female',
}
```

*   **id**: Unique identifier.
*   **speaker_id**: Unique identifier for the speaker.
*   **audio**: Audio data.
*   **transcription**: Transcription of the audio.
*   **language**: ISO 639-2 language code.
*   **gender**: Speaker gender ('Male', 'Female', or empty).

### TTS Data Fields

```python
{
  'id': 'swa_0',
  'speaker_id': '...',
  'audio': {
    'array': [...],
    'sample_rate': 16_000
  },
  'transcription': '...',
  'locale': 'swa',
  'gender': 'Female',
}
```

*   **id**: Unique identifier.
*   **speaker_id**: Unique identifier for the speaker.
*   **audio**: Audio data.
*   **transcription**: Transcription.
*   **locale**: ISO 639-2 language code.
*   **gender**: Speaker gender.

### Data Splits

For the **ASR Dataset**, the data with transcriptions is split as follows: *
**train**: 80% of labeled data. * **validation**: 10% of labeled data. *
**test**: 10% of labeled data.

The **unlabeled** split contains all samples that do not have a corresponding
transcription.

The **TTS Dataset** follows a similar structure, with data split into `train`,
`validation`, and `test` sets.

## Dataset Curation

The data was gathered by multiple partners:

Provider            | Dataset                                                  | License
:------------------ | :------------------------------------------------------- | :------
University of Ghana | [UGSpeechData](https://doi.org/10.57760/sciencedb.22298) | `CC BY 4.0`
Digital Umuganda    | [AfriVoice](DigitalUmuganda/AfriVoice)                   | `CC-BY 4.0`
Makerere University | [Yogera Dataset](https://doi.org/10.7910/DVN/BEROE0)     | `CC-BY 4.0`
Media Trust         |                                                          | `CC-BY 4.0`

## Considerations for Using the Data

Please check the license for the specific languages you are using, as they may
differ between providers.

**Affiliation:** Google Research

## Version and Maintenance

-   **Current Version:** 1.0.0
-   **Last Updated:** 01/2026