refrakt_cli is a command-line interface for the refrakt_core deep learning and machine learning research framework. It enables rapid, reproducible, and flexible pipeline training by converting CLI arguments into YAML-based experiment configurations, seamlessly integrating with refrakt_core's modular system.
Install the latest release from PyPI:
pip install refrakt_cli
- A CLI tool to launch, manage, and override ML/DL training pipelines using YAML configs.
- Bridges user-friendly command-line workflows with the powerful abstractions of refrakt_core.
- Supports dynamic hyperparameter overrides, modular experiment configs, and robust logging.
git clone https://github.com/refrakt-hub/refrakt_cli.git
cd refrakt_cli
# (Recommended) Create and activate a virtual environment
python3 -m venv .venv
source .venv/bin/activate
# Install dependencies
pip install .
refrakt --config path/to/your_config.yaml
refrakt --config path/to/your_config.yaml model.name=ResNet optimizer.lr=0.0005 trainer.epochs=20
Flag | Description |
---|---|
--config |
Path to YAML config file |
--log_type |
Logging backend: tensorboard , wandb , or both |
--debug |
Enable debug mode with extra verbosity |
src/refrakt_cli/
├── cli.py # Main CLI entry point
├── helpers/ # Argument parsing, config overrides, pipeline orchestration
│ ├── argument_parser.py
│ ├── config_overrides.py
│ └── pipeline_orchestrator.py
├── hooks/ # Custom hooks (e.g., hyperparameter overrides)
│ └── hyperparameter_override.py
├── utils/ # Utility functions for pipeline management
│ └── pipeline_utils.py
└── __init__.py
- cli.py: Main CLI logic and entry point.
- helpers/: Argument parsing, config override logic, and pipeline orchestration.
- hooks/: Custom hooks for advanced config/hyperparameter handling.
- utils/: Utility functions for pipeline and config management.
Test files are located in the tests/
directory, mirroring the main module structure.
Contributions, issues, and feature requests are welcome!
Please read our CONTRIBUTING.md for guidelines on how to get started and best practices for contributing to refrakt_cli.
This repository is licensed under the MIT License. See LICENSE for full details.
Akshath Mangudi
If you find issues, raise them. If you learn from this, share it. Built with love and curiosity :)