Project Structure (CCDS)
The repository follows the Cookiecutter Data Science layout to keep data, code, and artifacts cleanly separated.
emotion_detection_MLOps_pipeline/
├── data/
│ ├── raw/ # Train/test splits
│ ├── interim/ # Cleaned & normalized text
│ └── processed/ # Feature-engineered datasets
├── src/
│ ├── data/
│ │ ├── data_ingestion.py
│ │ └── data_preprocessing.py
│ ├── features/
│ │ └── feature_engineering.py
│ └── model/
│ ├── model_building.py
│ └── model_evaluation.py
├── models/ # Trained model artifacts
├── reports/ # Metrics and evaluation results
├── notebooks/ # Exploratory work
├── docs/ # Documentation
├── dvc.yaml
├── dvc.lock
├── params.yaml
├── requirements.txt
└── README.md
This structure makes it clear what is code, what is data, and what is output.