Metadata-Version: 2.4
Name: perihelion
Version: 0.0.1
Summary: A deep-learning framework built on JAX and Equinox
Project-URL: Homepage, https://github.com/djoshea/perihelion
Project-URL: Repository, https://github.com/djoshea/perihelion
Author-email: Dan O'Shea <dan@djoshea.com>
License-Expression: Apache-2.0
License-File: LICENSE
Requires-Python: >=3.12
Requires-Dist: dill>=0.3.9
Requires-Dist: equinox>=0.11.5
Requires-Dist: jax>=0.6.0
Requires-Dist: jaxtyping
Requires-Dist: loguru>=0.7.3
Requires-Dist: matplotlib<4.0.0,>=3.9.0
Requires-Dist: msgpack-types>=0.5.0
Requires-Dist: msgpack>=1.1.0
Requires-Dist: numpy
Requires-Dist: optax<0.3.0,>=0.2.2
Requires-Dist: optree>=0.18.0
Requires-Dist: penzai>=0.2.5
Requires-Dist: plotext>=5.3.2
Requires-Dist: pyyaml>=6.0
Requires-Dist: rich[jupyter]>=13.9.2
Requires-Dist: safetensors>=0.7.0
Requires-Dist: textual-plotext>=1.0.1
Requires-Dist: textual>=0.61.0
Requires-Dist: treescope>=0.1.8
Requires-Dist: wandb>=0.23.1
Provides-Extra: cuda
Requires-Dist: jax[cuda12]>=0.6.0; extra == 'cuda'
Description-Content-Type: text/markdown

# perihelion

A deep-learning framework built on [JAX](https://github.com/jax-ml/jax) and
[Equinox](https://github.com/patrick-kidger/equinox).

> **Status: early scaffolding.** The framework is being extracted from the
> `foraging_dis_rnn` research repository. Public API is not yet stable.

## Install

```bash
pip install perihelion
```

Plain `perihelion` installs a CPU `jaxlib` on every platform. For NVIDIA GPU
support (Linux), install the CUDA extra:

```bash
pip install "perihelion[cuda]"
```

## Usage

The intended import convention is:

```python
import perihelion as ph
```

## License

Apache-2.0. See [LICENSE](LICENSE) and [NOTICE](NOTICE).
