Naan Binary

Recipe generation and nutrition prediction project with:

1. Notebook training/inference pipeline.
2. FastAPI backend for model serving.
3. React frontend for interactive UI.

Current Status

1. Notebook pipeline is available in Naan_binary_v2.ipynb.
2. Backend is implemented under backend/ with /api/health and /api/predict.
3. Frontend is implemented under frontend/ and connected to the backend.
4. Repository is configured to ignore large/generated artifacts for GitHub.

Project Structure

1. Naan_binary_v2.ipynb: Main updated notebook pipeline.
2. backend/: FastAPI app and inference service.
3. frontend/: React + Vite UI.
4. requirements.txt: Root notebook/training dependencies.
5. backend/requirements.txt: Backend API/runtime dependencies.
6. DATASET.md: Dataset notes.
7. docs/TRAINING_AND_PREPROCESSING.md: Additional training details.

Environment Setup (Windows)

python -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip

Notebook Setup

pip install -r requirements.txt
jupyter notebook Naan_binary_v2.ipynb

Backend Setup and Run

cd backend
..\dl_env\Scripts\python -m pip install -r requirements.txt
..\dl_env\Scripts\python run.py

API docs: http://127.0.0.1:8000/docs

Frontend Setup and Run

cd frontend
npm install
npm run dev

Frontend URL: http://localhost:5173

Model and Data Files

Large model artifacts and datasets are intentionally ignored from git.

1. Keep cleaned_recipes.csv local.
2. Keep generated checkpoints local (recipe_multitask_ckpt/, recipe_checkpoints/, recipe_model_final/).
3. Keep deployed backend model files local under backend/assets/model_final/.

If you need to version large files, use Git LFS or a cloud artifact store.

Pre-Push Checklist

1. Confirm .gitignore is applied.
2. Run backend health check: GET /api/health.
3. Run one prediction through UI or POST /api/predict.
4. Verify no large artifacts are staged.

Push to GitHub

git status
git add .
git commit -m "Prepare repository for GitHub: ignore large artifacts, update docs and deps"
git push -u origin deployment

If your branch name is different, replace deployment with your branch.