eqxvision 0.1.10

Root package info.

Eqxvision

Eqxvision is a package of popular computer vision model architectures built using Equinox.

Installation

Use the package manager pip to install eqxvision.

pip install eqxvision

requires: python>=3.7

optional: torch, only if pretrained models are required.

Documentation

Available at https://eqxvision.readthedocs.io/en/latest/.

Usage

Picking a model and doing a forward pass is as simple as ...

    import jax
    import jax.random as jr
    import equinox as eqx
    from eqxvision.models import alexnet
    
    @eqx.filter_jit
    def forward(net, images, key):
        keys = jax.random.split(key, images.shape[0])
        output = jax.vmap(net, axis_name=('batch'))(images, key=keys)
        ...
        
    net = alexnet(num_classes=1000)
    
    images = jr.uniform(jr.PRNGKey(0), shape=(1,3,224,224))
    output = forward(net, images, jr.PRNGKey(0))

What's New?

• All torchvision classification models are now ported to eqxvision!

  !!! note Only DINO pretrained weights are available for ViTs.

Get Started!

Start with any one of these easy to follow tutorials.

Tips

• Better to use @equinox.jit_filter instead of @jax.jit
• Advisable to use jax.{v,p}map with axis_name='batch' for all models
• Don't forget to switch to inference mode for evaluations
• Wrap with eqx.filter(net, eqx.is_array) for Optax initialisation.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Development Process

If you plan to modify the code or documentation, please follow the steps below:

1. Fork the repository and create your branch from dev.
2. If you have modified the code (new feature or bug-fix), please add unit tests.
3. If you have changed APIs, update the documentation. Make sure the documentation builds. mkdocs serve
4. Ensure the test suite passes. pytest tests -vvv
5. Make sure your code passes the formatting checks. Automatically checked with a pre-commit hook.

Acknowledgements

• Equinox
• Patrick Kidger
• Torchvision

License

MIT