Metadata-Version: 2.4
Name: hydra-sweeper-explicit
Version: 0.0.1
Summary: Hydra sweeper for explicit parameter combinations without Cartesian product
Project-URL: Documentation, https://github.com/quadbio/hydra-sweeper-explicit#readme
Project-URL: Issues, https://github.com/quadbio/hydra-sweeper-explicit/issues
Project-URL: Source, https://github.com/quadbio/hydra-sweeper-explicit
Author: Marius Lange
Maintainer: Marius Lange
License-Expression: MIT
License-File: LICENSE
Keywords: hydra,hyperparameter,machine-learning,sweeper
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Scientific/Engineering
Requires-Python: >=3.12
Requires-Dist: hydra-core>=1.3
Requires-Dist: omegaconf>=2.3
Provides-Extra: dev
Requires-Dist: pre-commit; extra == 'dev'
Provides-Extra: test
Requires-Dist: coverage>=7.10; extra == 'test'
Requires-Dist: pytest-cov; extra == 'test'
Requires-Dist: pytest>=8; extra == 'test'
Description-Content-Type: text/markdown

# hydra-sweeper-explicit

[![Tests][badge-tests]][tests]
[![PyPI][badge-pypi]][pypi]

[badge-tests]: https://img.shields.io/github/actions/workflow/status/quadbio/hydra-sweeper-explicit/test.yaml?branch=main&label=tests
[badge-pypi]: https://img.shields.io/pypi/v/hydra-sweeper-explicit

[tests]: https://github.com/quadbio/hydra-sweeper-explicit/actions/workflows/test.yaml
[pypi]: https://pypi.org/project/hydra-sweeper-explicit

A Hydra sweeper for running explicit parameter combinations without Cartesian product.

## Installation

```bash
pip install hydra-sweeper-explicit
```

## Usage

```yaml
hydra:
  sweeper:
    _target_: hydra_sweeper_explicit.ExplicitSweeper
    combinations:
      - {model: small, lr: 0.01}
      - {model: large, lr: 0.001}
      - {model: large, lr: 0.0001, dropout: 0.5}
```

```bash
python train.py --multirun
```

Runs exactly 3 jobs—no Cartesian product.
