{"slug": "kokoro-yoruba-native-tonal-yoruba-tts-on-styletts2-kokoro-82m", "title": "Kokoro Yoruba – Native Tonal Yoruba TTS on StyleTTS2 / Kokoro-82M", "summary": "A developer release published as Akintunde102/kokoro-yoruba-v1 provides native tone-aware Yorùbá text-to-speech built on StyleTTS2 and Kokoro-82M, with 526 MB model weights under CC BY-SA 4.0 and a free Gradio demo on Hugging Face Spaces. The model was trained on 2.70 hours of curated tone-marked Yorùbá audio from BibleTTS and WaxalNLP, compressed from 2091.5 MB to 526.0 MB, and ships with two voices: spk8 (Speaker 8, BibleTTS, the flagship) and spk2 (Speaker 2, WaxalNLP, quieter in v1 due to low training clip count). It relies on the MIT-licensed oro-g2p grapheme-to-phoneme engine and kokoro-tonal-kit, and the author notes rare 4–6 syllable compounds such as Ṣàwòròidẹ and ògbóǹtarìgì can still miss tones because of the small dataset and the English-pretrained PLBERT.", "body_md": "Native tone-aware **Yorùbá Text-to-Speech (TTS)** built on **StyleTTS2 / Kokoro-82M**, powered by **[`oro-g2p`](https://github.com/Akintunde102/oro-g2p)** (` pip install oro-g2p`) and **[` kokoro-tonal-kit`](https://github.com/Akintunde102/kokoro-tonal-kit)**.\n\n- **Interactive Gradio Demo (Free Google Colab):**- **Instant Audio Showcase:**[huggingface.co/spaces/Akintunde102/kokoro-yoruba-demo](https://huggingface.co/spaces/Akintunde102/kokoro-yoruba-demo)\n- **Model Weights (`526 MB`, CC BY-SA 4.0):**[huggingface.co/Akintunde102/kokoro-yoruba-v1](https://huggingface.co/Akintunde102/kokoro-yoruba-v1)\n\n```\ngit clone https://github.com/yl4579/StyleTTS2.git\npip install -r requirements.txt\npython\nfrom tts_yoruba import YorubaTTS\n\ntts = YorubaTTS(styletts2_dir=\"./StyleTTS2\")  # Auto-downloads Akintunde102/kokoro-yoruba-v1\ntts.speak(\"Ẹ kú àárọ̀ o! Báwo ni àwọn ará ilé?\", voice=\"spk8\", out_path=\"greeting.wav\")\n```\n\n*(Pass Yorùbá text with tone marks and underdots such as `ọkọ̀` vs `ọkọ` so `oro-g2p` assigns the exact tones.)*\n\n- Trained on **2.70 hours** of curated tone-marked Yorùbá (`BibleTTS` +`WaxalNLP` ), compressed from`2091.5 MB` →**`526.0 MB`** .\n- **Voices:**`spk8` (Speaker 8 studio BibleTTS — flagship quality) and`spk2` (Speaker 2 WaxalNLP — quieter in v1 due to low training clip count).\n- Strong on short and in-domain Yorùbá clauses; rare 4–6 syllable compounds (*Ṣàwòròidẹ* ,*ògbóǹtarìgì* ) can still miss tones due to the small dataset and English-pretrained`PLBERT` .\n\nContributors are welcome to build directly on `Akintunde102/kokoro-yoruba-v1`:\n\n1. **Expand Clean Audio (`2.7h → 15h+`):** Split longer` 1.2s–5.8s` studio clips and balance multiple speakers cleanly.\n2. **Pre-train a Native Yorùbá `PLBERT`:** Train` Utils/PLBERT/` on Yorùbá text corpora so multi-syllable compounds hold crisp tones.\n3. **Add Auto-Diacritization:** Wire a Yorùbá tone-restoration model in front of`oro-g2p` for unaccented text input.\n4. **Add New Consenting Voicepacks:** Train with unique`speaker_id` tags (`15+ mins` per voice) and extract`[1, 256]` voicepacks after`spectral_norm` warmup.\n\n- **Yorùbá G2P Engine (MIT):**[Akintunde102/oro-g2p](https://github.com/Akintunde102/oro-g2p)\n- **Tonal Kokoro Starter Kit (MIT):**[Akintunde102/kokoro-tonal-kit](https://github.com/Akintunde102/kokoro-tonal-kit)\n- **Wrapper Code:** MIT License ([`LICENSE`](https://github.com/Akintunde102/kokoro-yoruba/blob/main/LICENSE) ).**Model Weights (`kokoro_yoruba_gold.pth`):** CC BY-SA 4.0 (all audio outputs are AI-generated synthetic speech).", "url": "https://wpnews.pro/news/kokoro-yoruba-native-tonal-yoruba-tts-on-styletts2-kokoro-82m", "canonical_source": "https://github.com/Akintunde102/kokoro-yoruba", "published_at": "2026-11-20 10:50:35+00:00", "updated_at": "2026-11-20 11:16:01.869581+00:00", "lang": "en", "topics": ["artificial-intelligence", "natural-language-processing", "ai-research", "ai-tools", "large-language-models"], "entities": ["Akintunde102/kokoro-yoruba-v1", "StyleTTS2", "Kokoro-82M", "oro-g2p", "kokoro-tonal-kit", "BibleTTS", "WaxalNLP", "Hugging Face"], "also_reported_by": [], "takeaway": "Pass Yorùbá text with tone marks and underdots (ọkọ̀ vs ọkọ) so oro-g2p assigns exact tones; the 2.70-hour model handles short in-domain clauses but can miss tones on rare 4–6 syllable compounds.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/kokoro-yoruba-native-tonal-yoruba-tts-on-styletts2-kokoro-82m", "markdown": "https://wpnews.pro/news/kokoro-yoruba-native-tonal-yoruba-tts-on-styletts2-kokoro-82m.md", "text": "https://wpnews.pro/news/kokoro-yoruba-native-tonal-yoruba-tts-on-styletts2-kokoro-82m.txt", "jsonld": "https://wpnews.pro/news/kokoro-yoruba-native-tonal-yoruba-tts-on-styletts2-kokoro-82m.jsonld"}}