Skip to main content

Easy to use distributions, bijections and normalizing flows in JAX.

Project description

FlowJAX

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 modules, making them compatible with JAX transformations.
  • Includes many state of the art normalizing flow models.
  • First class support for conditional distributions and density estimation.

Documentation

Available here.

Short example

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

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 modules.
  • coupling_flow (Dinh et al., 2017) and masked_autoregressive_flow (Kingma et al., 2016, Papamakarios et al., 2017) 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).
  • block_neural_autoregressive_flow, as introduced by De Cao et al., 2019.
  • planar_flow, as introduced by Rezende and Mohamed, 2015.
  • 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; Durkan et al., 2020).
  • 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

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

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

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 if desired.

@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},
}

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

flowjax-16.0.0.tar.gz (488.2 kB view details)

Uploaded Source

Built Distribution

flowjax-16.0.0-py3-none-any.whl (62.3 kB view details)

Uploaded Python 3

File details

Details for the file flowjax-16.0.0.tar.gz.

File metadata

  • Download URL: flowjax-16.0.0.tar.gz
  • Upload date:
  • Size: 488.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.1 CPython/3.12.7

File hashes

Hashes for flowjax-16.0.0.tar.gz
Algorithm Hash digest
SHA256 9fed2a3bcda03474346738e23ba3ca9d9cbdff8e0070558904222d17547053eb
MD5 f08897311230e3a94eb6fbdfe883ff72
BLAKE2b-256 7168ff9115e6dc4b6b92e77fe0177ea9ef4830280d4b82428ec21ab58ba99bad

See more details on using hashes here.

File details

Details for the file flowjax-16.0.0-py3-none-any.whl.

File metadata

  • Download URL: flowjax-16.0.0-py3-none-any.whl
  • Upload date:
  • Size: 62.3 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.1 CPython/3.12.7

File hashes

Hashes for flowjax-16.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 dfb13cdeef80fd3805bf70cb94fd3155926278040639c50d2b8b7ad6e3315b6e
MD5 b1264b07e504ca76a718f295c54a28ee
BLAKE2b-256 488a10cf920d7594fbd2c9c2fead56cb21cd22e11516fcd92352520927ac26b1

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page