Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/danielward27/flowjax


https://github.com/danielward27/flowjax

jax normalising-flows normalizing-flows

Last synced: 14 days ago
JSON representation

Awesome Lists containing this project

README 

        

![FlowJAX](/docs/_static/logo_light.svg)



Distributions, bijections and normalizing flows using Equinox and JAX

-----------------------------------------------------------------------

- Includes a wide range of distributions and bijections.

- Distributions and bijections are PyTrees, registered through 

  [Equinox](https://github.com/patrick-kidger/equinox/) modules, making them

  compatible with [JAX](https://github.com/google/jax) transformations.

- Includes many state of the art normalizing flow models.

- First class support for conditional distributions and density estimation.



## Documentation

Available [here](https://danielward27.github.io/flowjax/index.html).



## Short example

As an example we will create and train a normalizing flow model to toy data in just a few lines of code:



```python

from flowjax.flows import block_neural_autoregressive_flow

from flowjax.train import fit_to_data

from flowjax.distributions import Normal

import jax.random as jr

import jax.numpy as jnp



data_key, flow_key, train_key, sample_key = jr.split(jr.key(0), 4)



x = jr.uniform(data_key, (5000, 2))  # Toy data



flow = block_neural_autoregressive_flow(

    key=flow_key,

    base_dist=Normal(jnp.zeros(x.shape[1])),

)



flow, losses = fit_to_data(

    key=train_key,

    dist=flow,

    x=x,

    learning_rate=5e-3,

    max_epochs=200,

    )



# We can now evaluate the log-probability of arbitrary points

log_probs = flow.log_prob(x)



# And sample the distribution

samples = flow.sample(sample_key, (1000, ))

```



The package currently includes:

- Many simple bijections and distributions, implemented as [Equinox](https://arxiv.org/abs/2111.00254) modules.

- `coupling_flow` ([Dinh et al., 2017](https://arxiv.org/abs/1605.08803)) and `masked_autoregressive_flow` ([Kingma et al., 2016](https://arxiv.org/abs/1606.04934), [Papamakarios et al., 2017](https://arxiv.org/abs/1705.07057v4)) normalizing flow architectures.

    - These can be used with arbitrary bijections as transformers, such as `Affine` or `RationalQuadraticSpline` (the latter used in neural spline flows; [Durkan et al., 2019](https://arxiv.org/abs/1906.04032)). 

- `block_neural_autoregressive_flow`, as introduced by [De Cao et al., 2019](https://arxiv.org/abs/1904.04676).

- `planar_flow`, as introduced by [Rezende and Mohamed, 2015](https://arxiv.org/pdf/1505.05770.pdf).

- `triangular_spline_flow`, introduced here.

- Training scripts for fitting by maximum likelihood, variational inference, or using contrastive learning for sequential neural posterior estimation ([Greenberg et al., 2019](https://arxiv.org/abs/1905.07488); [Durkan et al., 2020](https://arxiv.org/abs/2002.03712])).

- A bisection search algorithm that allows inverting some bijections without a

known inverse, allowing for example both sampling and density evaluation to be

performed with block neural autoregressive flows.



## Installation

```bash

pip install flowjax

```



## Warning

This package is in its early stages of development and may undergo significant changes, including breaking changes, between major releases. Whilst ideally we should be on version 0.y.z to indicate its state, we have already progressed beyond that stage. Any breaking changes will be in the release notes for each major release.



## Development

We can install a version for development as follows

```bash

git clone https://github.com/danielward27/flowjax.git

cd flowjax

pip install -e .[dev]

sudo apt-get install pandoc  # Required for building documentation

```



## Related

- We make use of the [Equinox](https://arxiv.org/abs/2111.00254) package, which

  facilitates defining models using a PyTorch-like syntax with Jax.

- For applying parameterizations, we use

  [paramax](https://github.com/danielward27/paramax).



## Citation

If you found this package useful in academic work, please consider citing it using the

template below, filling in ``[version number]`` and ``[release year of version]`` to the

appropriate values. Version specific DOIs

can be obtained from [zenodo](https://zenodo.org/records/10402073) if desired.



```bibtex

@software{ward2023flowjax,

  title = {FlowJAX: Distributions and Normalizing Flows in Jax},

  author = {Daniel Ward},

  url = {https://github.com/danielward27/flowjax},

  version = {[version number]},

  year = {[release year of version]},

  doi = {10.5281/zenodo.10402073},

}

```