Loading datasets...
Release Date: 10/20/2026
Format: FLAC, TSV
Size: 40.13 GB
Synthetic Chichewa (ny, nya) speech for automatic speech recognition, created by CLEAR Global. It contains 374,612 utterances totalling about 550 hours at 24 kHz mono, published as a single train split. The sentences were written by Claude 3.5 Sonnet (deduplicated corpus), and the audio was synthesised by a YourTTS model fine-tuned on Chichewa Bible recordings (open.bible). The voice is a single male voice. Each clip ships with a normalised transcript, the original LLM sentence and its duration. Audio is FLAC (lossless re-encoding of the original 24 kHz WAV).
Licensing
Creative Commons Attribution Non Commercial 4.0 International (CC-BY-NC-4.0)
https://spdx.org/licenses/CC-BY-NC-4.0.htmlRestrictions/Special Constraints
The synthetic text was generated with Claude 3.5 Sonnet. Use of the content must comply with Anthropic's Terms of Service and Usage Policy, and must not be used to develop products or services that compete with Anthropic's offerings.
Forbidden Usage
None
Ethical Review
The TTS model was fine-tuned on the open.bible Chichewa recordings from a single male narrator. The narrator is not identified in this release, and no personal data is included. The text is generated and does not contain personal information.
Intended Use
Research and development of automatic speech recognition for Chichewa, including training with real data.
This is the Chichewa synthetic speech from DeRenzi et al. (2025). In that study, a 1:2 mix of real and synthetic speech reduced Chichewa word error rate by about 6.5% relative, with only 34 hours of real data.
The sentences were generated with Claude 3.5 Sonnet. The audio was synthesised by a YourTTS model fine-tuned by CLEAR Global on Chichewa Bible recordings from the BibleTTS dataset.
clips/train/ — 374,612 FLAC files, 24 kHz mono, 16-bit, named by the MD5 of the original file name.metadata.tsv — one row per clip with the columns path, split, text, original_text, duration.text is lowercased with punctuation removed and is the transcript to use for ASR training. original_text keeps the LLM's capitalisation and punctuation.
DeRenzi, B., Dixon, A., Farhi, M.A., Resch, C. (2025) Synthetic Voice Data for Automatic Speech Recognition in African Languages. arXiv:2507.17578.
@misc{derenzi2025synthetic,
title = {Synthetic Voice Data for Automatic Speech Recognition in African Languages},
author = {DeRenzi, Brian and Dixon, Anna and Farhi, Mohamed Aymane and Resch, Christian},
year = {2025},
eprint = {2507.17578},
archivePrefix = {arXiv},
primaryClass = {cs.CL},
url = {https://arxiv.org/abs/2507.17578}
}
This dataset is synthetic: it was not recorded from human speakers. The text was generated by a large language model and the audio by a text-to-speech model, so it may contain model-induced biases. The voice is male only. The work was supported by the Gates Foundation; the conclusions and opinions expressed are those of the authors alone and shall not be attributed to the Foundation.
Please check README.md for more information.