Metadata-Version: 2.4
Name: energnn
Version: 0.1.0
Summary: A Graph Neural Network library for real-life energy networks.
Author-email: Balthazar DONON <balthazar.donon@rte-france.com>, Geoffroy JAMGOTCHIAN <geoffroy.jamgotchian@rte-france.com>, Hugo KULESZA <hugo.kulesza@rte-france.com>, Steve NOUATIN <steve.nouatin_externe@rte-france.com>
License-File: LICENSE
Requires-Python: >=3.10
Requires-Dist: chex==0.1.87
Requires-Dist: cloudpickle<3
Requires-Dist: diffrax==0.6.0
Requires-Dist: flatdict==4.0.1
Requires-Dist: flax==0.8.5
Requires-Dist: furo==2024.8.6
Requires-Dist: hydra-core==1.3.2
Requires-Dist: jax==0.4.30
Requires-Dist: neptune-scale==0.13.0
Requires-Dist: neptune==1.13.0
Requires-Dist: optax>=0.2.3
Requires-Dist: pandas>=2.2.2
Requires-Dist: sphinx==8.1.3
Requires-Dist: tqdm==4.67.0
Provides-Extra: gpu
Requires-Dist: jax[cuda12]; extra == 'gpu'
Description-Content-Type: text/markdown

# EnerGNN
A Graph Neural Network library for real-life energy networks.

## Build

To build this package locally, you can use one of the following commands at the root of the project:

```cmd
uv sync
```

```cmd
uv sync --extra gpu
```

The first one will install the `energnn` package with only CPU support.

Use the second one to also install the GPU extra dependencies (obtained from `jax[cuda12]`).

## Documentation

To build and access the documentation, run the following:

```shell
cd docs
make html
open _build/html/index.html
```