DigitalUmuganda Mbaza-ASR-Afrivoice-660h
[ingest-scout] Rwanda-based NLP organization Digital Umuganda's Conformer-CTC ASR model trained on 660 hours of Kinyarwanda speech (Afrivoice dataset); last updated December 2 2025, 52 monthly downloads, CC-BY-4.0 license, built on NVIDIA NeMo. Distinct from the already-inventoried mbaza-whisper-small-kinyarwanda — this is a larger NeMo-based model; fills Rwanda AI infrastructure representation.
- Category
- AI Resources
- Pricing
- Not listed
- Country
- 🇳🇬 Nigeria
- Last verified
- Not recorded
{ "name": "DigitalUmuganda Mbaza-ASR-Afrivoice-660h", "slug": "digitalumuganda-mbaza-asr-afrivoice-660h", "category": "AI", "country": "NG", "docs_status": "LIVE", "licensing_required": "NONE", "verified": false, "last_verified": null, "website": "https://huggingface.co/DigitalUmuganda/Mbaza-ASR-Afrivoice-660h", "documentation_url": null}Related in AI Resources
YarnGPT2b
Nigerian-accented English TTS and ASR model (Jan 2025) with 11 voices, trained on Nollywood and podcast audio.
NaijaML
Production-ready NLP library (v0.2.1, Mar 2026) covering all 4 Nigerian languages, running in 4GB RAM with no GPU, with PII masking for NIN/phone. The most practical Nigerian AI resource for developers.
SERENGETI
A massively multilingual masked language model covering 517 African languages and varieties across five scripts, achieving state-of-the-art results on the AfroNLU benchmark. Developed by the UBC Deep Learning and NLP Lab as an Afrocentric resource.