Metadata-Version: 2.4
Name: jaxonmodels
Version: 0.5.1
Summary: JAX models for deep learning in Equinox
Author-email: "Artur A. Galstyan" <mail@arturgalstyan.dev>
License: MIT License
        
        Copyright (c) 2024 Artur A. Galstyan
        
        Permission is hereby granted, free of charge, to any person obtaining a copy
        of this software and associated documentation files (the "Software"), to deal
        in the Software without restriction, including without limitation the rights
        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
        copies of the Software, and to permit persons to whom the Software is
        furnished to do so, subject to the following conditions:
        
        The above copyright notice and this permission notice shall be included in all
        copies or substantial portions of the Software.
        
        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
        SOFTWARE.
        
Requires-Python: >=3.13
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: jax
Requires-Dist: equinox
Requires-Dist: pydantic
Requires-Dist: numpy
Requires-Dist: jaxlib
Requires-Dist: tqdm
Requires-Dist: beartype
Requires-Dist: typing_extensions
Requires-Dist: loguru
Requires-Dist: optax
Requires-Dist: jaxtyping
Requires-Dist: pytest
Requires-Dist: ftfy
Requires-Dist: regex
Requires-Dist: statedict2pytree
Requires-Dist: jaxonlayers
Provides-Extra: dev
Requires-Dist: pre-commit; extra == "dev"
Requires-Dist: pytest; extra == "dev"
Requires-Dist: mkdocs; extra == "dev"
Requires-Dist: vulture; extra == "dev"
Provides-Extra: examples
Requires-Dist: tensorflow-cpu; extra == "examples"
Requires-Dist: tensorflow_datasets; extra == "examples"
Requires-Dist: clu; extra == "examples"
Requires-Dist: requests; extra == "examples"
Provides-Extra: tests
Requires-Dist: torch; extra == "tests"
Requires-Dist: torchvision; extra == "tests"
Requires-Dist: esm>=3.2.3; extra == "tests"
Provides-Extra: esm
Requires-Dist: tokenizers; extra == "esm"
Requires-Dist: huggingface_hub; extra == "esm"
Requires-Dist: transformers; extra == "esm"
Dynamic: license-file

# Jaxonmodels

🚨 This library is still under HEAVY development and won't reach version 1.0.0 in a long time!

This library consists of deep learning model implementations in JAX using Equinox as the neural network library.

The goal of this library is to provide simple, yet performant and easy to understand implementations with the aim to give *exactly* the same output as their Pytorch counterparts. As such, great emphasis is placed on making sure that the layers and the models behave accordingly.

Using `statedict2pytree` we can also load the Pytorch model weights into the JAX models.

Some models will have inadvertently repeated code, but this is fine so long as the model remains self contained for the most part.

## Implemented Models

These models have been implemented:
- [x] AlexNet
- [x] CLIP
- [x] EfficientNet
- [x] ResNet
- [x] ViT
- [x] Mamba
- [x] ConvNext
- [x] Swin Transformer
- [ ] Siglip (in progress)
- [x] VQ-VAE 
- [x] ESMC
- [x] ESM3


## Contributing

If you have a model that you would like to include, then just open up a PR. It should contain your model and ideally a few tests showcasing that the model (and its components) behave like their Pytorch versions.
