███╗ ███╗██╗ ██████╗██╗ ██╗
████╗ ████║██║ ██╔════╝██║ ██║
██╔████╔██║██║ ██║ ██║ ██║
██║╚██╔╝██║██║ ██║ ██║ ██║
██║ ╚═╝ ██║███████╗╚██████╗███████╗██║
╚═╝ ╚═╝╚══════╝ ╚═════╝╚══════╝╚═╝
A powerful, modular CLI tool for training, evaluating, and tracking ML/DL models
- ML Models: Logistic Regression, SVM, Random Forest, XGBoost
- DL Models: TensorFlow DNN, CNN, RNN/LSTM/GRU
- Hyperparameter Tuning: Grid Search, Random Search, Bayesian (Optuna)
- Model Explainability: SHAP & LIME
- Preprocessing: Scalers, Normalizers, Encoders, Feature Selection
- Experiment Tracking: Built-in tracker with JSON storage
- Interactive TUI: Terminal-based user interface
pip install mlcli-toolkitmlcli --helpmlcli train --config configs/rf_config.jsonmlcli ui| Command | Description |
|---|---|
mlcli list-models |
List available model trainers |
mlcli train -c <config> |
Train a model |
mlcli eval -m <model> -d <data> -t <type> |
Evaluate a model |
mlcli tune -c <config> -m <method> |
Hyperparameter tuning |
mlcli explain -m <model> -d <data> -e <method> |
Model explainability |
mlcli preprocess -d <data> -o <output> -m <method> |
Preprocess data |
mlcli list-runs |
List experiment runs |
mlcli ui |
Launch interactive TUI |
{
"model": {
"type": "random_forest",
"params": {
"n_estimators": 100,
"max_depth": null,
"random_state": 42
}
},
"dataset": {
"path": "data/train.csv",
"type": "csv",
"target_column": "target"
},
"training": {
"test_size": 0.2,
"random_state": 42
},
"output": {
"model_dir": "artifacts",
"save_formats": ["pickle", "joblib"]
}
}For complete documentation including:
- All configuration options
- Hyperparameter tuning guides
- Model explainability (SHAP/LIME)
- Data preprocessing pipeline
- Extending MLCLI with custom trainers
- Troubleshooting
Contributions are welcome! Please feel free to submit a Pull Request.
This project is licensed under the MIT License.