AI & ML interests

None defined yet.

Recent Activity

demelewwย  updated a Space about 1 month ago
ethiopian-Demtse/README
demelewwย  published a Space about 1 month ago
ethiopian-Demtse/README
View all activity

Organization Card

แ‹ตแˆแ„ โ€” Ethiopian Languages AI ๐Ÿ‡ช๐Ÿ‡น

แˆˆแŠขแ‰ตแ‹ฎแŒตแ‹ซ 80+ แ‰‹แŠ•แ‰‹แ‹Žแ‰ฝ แŠญแแ‰ต แˆแŠ•แŒญ แ‹จแ‹ตแˆแŒฝ AI
Open-source speech AI for Ethiopia's 80+ languages

๐Ÿค– Amharic TTS Model โ€ข ๐ŸŽ™๏ธ Live Demo โ€ข ๐ŸŒ Voices For All


๐ŸŽฏ Mission / แ‰ฐแˆแ‹•แŠฎ

แ‹ตแˆแ„ (Demtse โ€” "my voice") builds free, open-source speech AI so 120M+ Ethiopians can access technology in their mother tongue.

Ethiopia has 80+ living languages. Big Tech supports almost none of them. We're changing that โ€” one language at a time, starting with the highest-quality open Amharic TTS ever built.


๐Ÿ—ฃ๏ธ Models

OmniVoice Amharic โ€” TTS + Voice Cloning

What it does Converts Amharic text to natural speech; clones any voice from 10s audio
Architecture Non-autoregressive discrete diffusion (612.6M params)
Training data 331 hours across 4 datasets
Best loss 3.9518
License Apache 2.0 โ€” completely free
Try it โ–ถ๏ธ Live Demo
Download ๐Ÿ“ฆ Model
# Quick start โ€” runs on free Colab T4
!pip install -q omnivoice soundfile
from omnivoice import OmniVoice, OmniVoiceGenerationConfig
import torch

model = OmniVoice.from_pretrained("ethiopian-Demtse/omnivoice-amharic", device_map="cuda:0", dtype=torch.float16)
audio = model.generate(text="แˆฐแˆ‹แˆแฃ แŠฅแŠ•แŠณแŠ• แ‹ฐแˆ…แŠ“ แˆ˜แŒฃแ‰ฝแˆ!", language="Amharic",
    generation_config=OmniVoiceGenerationConfig(num_step=32, guidance_scale=2.0))

๐Ÿ‡ช๐Ÿ‡น Roadmap โ€” Ethiopian Languages

แ‰‹แŠ•แ‰‹ / Language แ‰ฐแŠ“แŒ‹แˆช / Speakers แˆแŠ”แ‰ณ / Status
แŠ แˆ›แˆญแŠ› Amharic 60M+ โœ… TTS + Voice Cloning
แŠฆแˆฎแˆแŠ› Afaan Oromoo 40M+ ๐Ÿ”œ Next
แ‰ตแŒแˆญแŠ› Tigrinya 10M+ ๐Ÿ”œ Planned
แˆถแˆ›แˆŠแŠ› Somali 7M+ ๐Ÿ”œ Planned
แˆฒแ‹ณแˆแŠ› Sidamo 4M+ ๐Ÿ“‹ Future
แ‹ˆแˆ‹แ‹ญแ‰ตแŠ› Wolaytta 3M+ ๐Ÿ“‹ Future
แŒ‰แˆซแŒแŠ› Gurage 2M+ ๐Ÿ“‹ Future
แˆแ‹ฒแ‹ญแŠ› Hadiyya 2M+ ๐Ÿ“‹ Future
แŠ แ‹แˆญแŠ› Afar 2M+ ๐Ÿ“‹ Future
แŒˆแˆžแŠ› Gamo 1.5M+ ๐Ÿ“‹ Future

Goal: TTS + voice cloning for every Ethiopian language with 50h+ available audio.


๐Ÿ“š Training Data

Dataset Hours Language
google/WaxalNLP ~200h Amharic
gheero-Leyu/leyu-amharic-addis-ababa-dialect ~50h Amharic (Addis)
surafelabebe/amharic_clear_audio_tts ~40h Amharic
chappM/amharic-bdu-asr ~41h Amharic
Total ~331h

We actively collect data for Oromo, Tigrinya, and other Ethiopian languages. If you have audio recordings in any Ethiopian language, please reach out.


๐Ÿ› ๏ธ What We Build

Project Description Status
Text-to-Speech (TTS) แŒฝแˆ‘แ โ†’ แŠ•แŒแŒแˆญ โœ… Amharic done
Voice Cloning 10 แˆฐแŠจแŠ•แ‹ต แŠ“แˆ™แŠ“ โ†’ แˆ›แŠ•แŠ›แ‹แˆ แ‹ตแˆแŒฝ โœ… Amharic done
Speech Recognition (ASR) แŠ•แŒแŒแˆญ โ†’ แŒฝแˆ‘แ ๐Ÿ“‹ Planned
Language Models (NLP) แ‹จแ‰‹แŠ•แ‰‹ แˆžแ‹ดแˆŽแ‰ฝ ๐Ÿ“‹ Planned
Machine Translation แ‰ตแˆญแŒ‰แˆ แ‰ แŠขแ‰ตแ‹ฎแŒตแ‹ซ แ‰‹แŠ•แ‰‹แ‹Žแ‰ฝ ๐Ÿ“‹ Planned

๐Ÿค How to Contribute

We welcome contributions from anyone:

Way to help Description
๐ŸŽ™๏ธ Donate audio Record or share speech in any Ethiopian language (any dialect, any speaker)
๐Ÿ’ป Code Open Issues, submit PRs, improve training pipelines
๐Ÿ“ Data Text transcriptions, translations, text-audio pairs
๐Ÿซ Institutions Universities and organizations: partner with us for data collection
๐Ÿงช Evaluate Native speakers: test our models and give feedback on naturalness
๐Ÿ’ฐ Fund Support compute costs and data collection

๐ŸŒ Part of Voices For All

แ‹ตแˆแ„ is the Ethiopian chapter of Voices For All (african-low-resource) โ€” a pan-African initiative building speech AI for languages left behind by Big Tech.

Initiative Focus
african-low-resource Pan-African: Amharic, Wolof, Hausa, Swahili, Somali
ethiopian-Demtse Ethiopia-specific: all 80+ Ethiopian languages

๐Ÿ“ฌ Contact


๐Ÿ“ Citation

@software{ethiopian_demtse_2026,
  author = {demeleww and ethiopian-Demtse},
  title = {แ‹ตแˆแ„: Open Speech AI for Ethiopian Languages},
  year = {2026},
  url = {https://huggingface.co/ethiopian-Demtse},
  license = {Apache-2.0}
}

โค๏ธ แˆˆ120 แˆšแˆŠแ‹ฎแŠ•+ แŠขแ‰ตแ‹ฎแŒตแ‹ซแ‹แ‹ซแŠ• โ€” แ‰ แˆซแˆณแ‰ธแ‹ แ‰‹แŠ•แ‰‹ แ‹จ AI แ‹ตแˆแŒฝ แ‹ญแŒˆแ‰ฃแ‰ธแ‹‹แˆแข
For 120M+ Ethiopians who deserve a voice in AI.

datasets 0

None public yet