Installation

Requirements

Chimera ML currently supports Python >=3.12,<3.13. The package declares PyTorch >=2.2,<3.0 and the other runtime dependencies in pyproject.toml.

Install from PyPI

python -m pip install chimera-ml

The package installs the chimera-ml CLI. The underscore alias chimera_ml is also available, but the documentation consistently uses the hyphenated name.

Install from source

For development, install the project and its development tools with Poetry:

poetry install --with dev,docs

The docs group contains MkDocs, Material for MkDocs, and the Python handler for mkdocstrings. It is intentionally separate from the production runtime dependencies.

Install an example plugin

Task-specific datamodules, models, losses, metrics, and callbacks are commonly provided by external plugins. From the repository root, an example plugin can be installed in editable mode:

python -m pip install -e examples/va_estimation

Example plugin dependencies and dataset paths are documented in each example's README. Installing a plugin is not required for the core CLI diagnostics or the generated API reference.