wecncodecrew/La-Lango

Low-Resource Language Translation API, A community-driven NLP platform focused on translating regional dialects \# 🌍 La Lango AI > **Low-Resource Language Translation API** — A community-driven NLP platform for translating > regional dialects, built from scratch by students, for the world. --- ## What is La Lango AI? Hundreds of regional dialects and low-resource languages have little to no machine translation support. Commercial AI APIs ignore them because they are not profitable. **La Lango AI is our answer to that.** We are an open, community-built translation platform where every language deserves a model. > **No external AI APIs. No black boxes. Everything is implemented from scratch.** --- ## Project structure ``` la-lango-ai/ │ ├── 📁 backend/ ← Translation engine + REST API (Python) │ ├── lalango/ ← Core package: models, tokenizers, data, API │ ├── scripts/ ← CLI tools: train, evaluate, preprocess │ ├── tests/ ← Automated tests │ └── experiments/ ← Jupyter notebooks (start here!) │ ├── 📁 frontend/ ← Web UI (plain HTML / CSS / JavaScript) │ └── index.html ← Entire UI in one file, no framework needed │ ├── 📁 languages/ ← Community language registry (add yours here!) ├── 📁 data/ ← Your datasets go here (not tracked by git) ├── 📁 docs/ ← Guides: architecture, data format, adding a language │ ├── README.md ├── CONTRIBUTING.md ├── ROADMAP.md └── LICENSE ``` --- ## How the system works ``` Browser (frontend/index.html) │ HTTP fetch ▼ FastAPI (backend/lalango/api/) │ ▼ Tokenizer (backend/lalango/tokenizers/) │ splits text into characters/subwords ▼ Translation Model (backend/lalango/models/) │ predicts the translation ▼ Tokenizer ← decodes output back to text │ ▼ Browser receives translation ``` --- ## Quickstart ### 1. Clone and install ```bash git clone github.com cd La-Lango pip install -r backend/requirements.txt ``` ### 2. Start the backend API ```bash uvicorn backend.lalango.api.main:app --reload # Swagger U …