Install¶
Installation¶
EmpiricML is available on PyPI and can be installed using pip:
Requirements: Python 3.11 or higher.
Dependencies¶
When you install EmpiricML, the following dependencies will be automatically installed:
numpy>=1.26.4pandas>=2.2.2pyarrow>=15.0.2polars>=1.31.0matplotlib>=3.9.0scikit-learn>=1.7.1lightgbm>=4.6.0xgboost>=3.1.2catboost>=1.2.8skorch>=1.3.1
PyTorch Support¶
EmpiricML supports PyTorch models through the TorchWrapper class. However, PyTorch is not installed automatically to allow users to choose the specific version that matches their hardware configuration (CPU vs GPU, CUDA version, etc.).
To use PyTorch models with EmpiricML:
- Install PyTorch separately by following the instructions on the official PyTorch website.
- Use the
TorchWrapperclass to integrate your PyTorch models into the EmpiricML workflow.