نماء · شبكة لتطوير معالجة اللغة العربية والذكاء الاصطناعي
NAMAA is a community dedicated to empowering research and building open‑source models and datasets for Arabic natural language processing and artificial intelligence. We advance the field through collaboration, innovation and knowledge sharing.
Website namaa.space · Report namaa.space/report · Models & datasets huggingface.co/NAMAA-Space
Every project lives in a team. Mention the team to reach the people who own it.
| Team | Mention | What it does |
|---|---|---|
| Board | @NAMAA-ORG/board |
Steers every project and chapter |
| TTS | @NAMAA-ORG/tts |
Text‑to‑speech and speech for Arabic dialects |
| Multimodality | @NAMAA-ORG/multimodality |
Multimodal and vision‑language models for Arabic |
| OCR | @NAMAA-ORG/ocr |
Arabic OCR: Qari‑OCR and manuscript recognition |
| Classical Arabic | @NAMAA-ORG/classical-arabic |
Classical and Modern Standard Arabic resources |
| LLM Security | @NAMAA-ORG/llm-security |
Safety and prompt‑guard models for Arabic LLMs |
| NAMAA Syria | @NAMAA-ORG/namaa-syria |
Syrian chapter: Modern Standard Arabic and Syrian dialects |
| └ Syrian Dialect Encyclopedia | @NAMAA-ORG/syrian-dialect-encyclopedia |
MSA ↔ Syrian dialect lexicon and parallel corpus |
- Every piece of work is an issue in the repo it belongs to, tracked on the team's project board: Backlog → In progress → Review → Done.
- Changes arrive as pull requests. The owning team reviews them;
CODEOWNERSrequests the review automatically. - Datasets and models are versioned with GitHub releases and mirrored on Hugging Face.
- Announcements and questions live in Discussions.
Read the contributing guide, pick an issue labelled good first issue, or write to us through namaa.space/contact.