Cohere Confirms Community Quants for Arabic Speech Model on Hugging Face
cohere · x · 2026-07-23
Cohere replied on X that their Hugging Face community members have already provided GGUF and ONNX quantizations for the recently released Arabic Speech Tashkeel 2B model. The model focuses on converting Arabic text into fully diacritized text.
Related event: Cohere Unveils Arabic Speech Model Tashkeel 2B(2 posts)→
More from Models
- OpenRouter Data Shows Gemini Becomes the API Market Leader — PsychicorAI · 2026-07-23
- MiniMax M3 weights use a community license, not MIT or Apache — SprinklesLeather7515 · 2026-07-23
- Emad Mostaque says prompt crafting is overrated, just ask the model for the result — mattshumer_ · 2026-07-23
- ARK Invest Analyzes Kimi K3: Narrows Intelligence Gap, Not Compute Bill — downingARK · 2026-07-23
- Developer Teases New Local LLM 'Ocelot' Arriving Mid-September — gnukeith · 2026-07-23
- Rumor: Google Delayed Gemini 3.5 Pro, Falling Further Behind OpenAI's GPT-6 — haider1 · 2026-07-23