fleqx 0.0.2__tar.gz

fleqx-0.0.2/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 fleqx contributors
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

fleqx-0.0.2/PKG-INFO

@@ -0,0 +1,92 @@
+Metadata-Version: 2.4
+Name: fleqx
+Version: 0.0.2
+Summary: Normalizing flows for JAX, with a distreqx-native API.
+Author: fleqx contributors
+License: Apache-2.0
+Project-URL: homepage, https://gvcallen.github.io/fleqx
+Project-URL: documentation, https://gvcallen.github.io/fleqx
+Project-URL: repository, https://github.com/gvcallen/fleqx
+Keywords: jax,normalizing-flows,distreqx,equinox,machine-learning
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Scientific/Engineering :: Mathematics
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: jax>=0.7.0
+Requires-Dist: equinox>=0.11
+Requires-Dist: jaxtyping
+Requires-Dist: optax
+Requires-Dist: tqdm
+Requires-Dist: distreqx>=0.0.3
+Provides-Extra: test
+Requires-Dist: pytest; extra == "test"
+Requires-Dist: jax[cpu]; extra == "test"
+Provides-Extra: docs
+Requires-Dist: mkdocs>=1.6; extra == "docs"
+Requires-Dist: mkdocs-material>=9.5; extra == "docs"
+Requires-Dist: mkdocstrings[python]>=0.26; extra == "docs"
+Requires-Dist: pymdown-extensions>=10.7; extra == "docs"
+Dynamic: license-file
+
+# fleqx
+
+Normalizing flows for JAX, built directly on [distreqx](https://github.com/lockwo/distreqx).
+
+A flow from `fleqx` is a plain `distreqx.distributions.Transformed` — there's no
+flow-specific wrapper, so `log_prob`, `sample`, `optax` training, etc. all work
+exactly as they would for any other distreqx distribution. As with any distreqx
+distribution, batches are handled with `jax.vmap` rather than by passing in arrays
+with a leading batch axis.
+
+Three flow types are implemented so far: coupling, masked autoregressive, and planar.
+
+## Example
+
+```python
+import jax.numpy as jnp
+import jax.random as jr
+
+import fleqx
+
+key = jr.key(0)
+flow = fleqx.flows.coupling_flow(key, dim=2)
+
+sample = flow.sample(jr.key(1))
+log_p = flow.log_prob(sample)
+
+data = jr.normal(jr.key(2), (1000, 2))
+flow, losses = fleqx.train.fit(jr.key(3), flow, data)
+```
+
+## Installation
+
+```
+pip install fleqx
+```
+
+See the [documentation](https://gvcallen.github.io/fleqx) for the full API.
+
+## Acknowledgements
+
+Built on [distreqx](https://github.com/lockwo/distreqx) (Owen Lockwood). The
+bijectors were ported from [flowjax](https://github.com/danielward27/flowjax)
+(Daniel Ward), a more complete flows library that's worth using directly if you don't
+specifically need the distreqx API.
+
+## Development
+
+```
+pip install -e ".[test]"
+pytest
+```
+
+---
+
+This library was written by Claude, porting the bijectors directly from flowjax.
+Behaviour should be nearly identical to flowjax's, though minor differences may
+remain, and the code hasn't yet had a full human review.

fleqx-0.0.2/README.md

@@ -0,0 +1,57 @@
+# fleqx
+
+Normalizing flows for JAX, built directly on [distreqx](https://github.com/lockwo/distreqx).
+
+A flow from `fleqx` is a plain `distreqx.distributions.Transformed` — there's no
+flow-specific wrapper, so `log_prob`, `sample`, `optax` training, etc. all work
+exactly as they would for any other distreqx distribution. As with any distreqx
+distribution, batches are handled with `jax.vmap` rather than by passing in arrays
+with a leading batch axis.
+
+Three flow types are implemented so far: coupling, masked autoregressive, and planar.
+
+## Example
+
+```python
+import jax.numpy as jnp
+import jax.random as jr
+
+import fleqx
+
+key = jr.key(0)
+flow = fleqx.flows.coupling_flow(key, dim=2)
+
+sample = flow.sample(jr.key(1))
+log_p = flow.log_prob(sample)
+
+data = jr.normal(jr.key(2), (1000, 2))
+flow, losses = fleqx.train.fit(jr.key(3), flow, data)
+```
+
+## Installation
+
+```
+pip install fleqx
+```
+
+See the [documentation](https://gvcallen.github.io/fleqx) for the full API.
+
+## Acknowledgements
+
+Built on [distreqx](https://github.com/lockwo/distreqx) (Owen Lockwood). The
+bijectors were ported from [flowjax](https://github.com/danielward27/flowjax)
+(Daniel Ward), a more complete flows library that's worth using directly if you don't
+specifically need the distreqx API.
+
+## Development
+
+```
+pip install -e ".[test]"
+pytest
+```
+
+---
+
+This library was written by Claude, porting the bijectors directly from flowjax.
+Behaviour should be nearly identical to flowjax's, though minor differences may
+remain, and the code hasn't yet had a full human review.

fleqx-0.0.2/fleqx/__init__.py

@@ -0,0 +1,12 @@
+"""fleqx: normalizing flows for JAX, with a distreqx-native API."""
+
+from . import bijectors as bijectors
+from . import flows as flows
+from . import train as train
+
+try:  # pragma: no cover
+    import importlib.metadata
+
+    __version__ = importlib.metadata.version("fleqx")
+except importlib.metadata.PackageNotFoundError:  # pragma: no cover
+    __version__ = "0.0.1"

fleqx-0.0.2/fleqx/bijectors/__init__.py

@@ -0,0 +1,13 @@
+"""Bijectors used to build normalizing flows.
+
+Every class here is a genuine `distreqx.bijectors.AbstractBijector`, implemented
+directly against the distreqx API with no third-party flow library involved. Most
+users should build flows with a constructor from [`fleqx.flows`][] instead; this
+module is for composing layers by hand.
+"""
+
+from ._coupling import Coupling as Coupling
+from ._invert import Invert as Invert
+from ._masked_autoregressive import MaskedAutoregressive as MaskedAutoregressive
+from ._permute import Permute as Permute
+from ._planar import Planar as Planar

fleqx-0.0.2/fleqx/bijectors/_affine_common.py

@@ -0,0 +1,31 @@
+"""Shared helper for bijectors parameterising an elementwise affine transform."""
+
+from __future__ import annotations
+
+import jax.nn as jnn
+import jax.numpy as jnp
+from jaxtyping import Array
+
+
+def positive_scale(raw_scale: Array, min_scale: float) -> Array:
+    """Maps an unconstrained array to strictly positive values via `softplus`.
+
+    `min_scale` is added as a floor, both for numerical stability (avoiding
+    `log(scale)` blowing up as `scale -> 0`) and to guarantee positivity outright.
+    """
+    return jnn.softplus(raw_scale) + min_scale
+
+
+def identity_init_offset(min_scale: float) -> Array:
+    """Offset added to a conditioner's raw scale output before `positive_scale`.
+
+    A freshly initialised conditioner network outputs values close to zero, and
+    `positive_scale(0, min_scale)` is around 0.7, not 1 -- so without this offset, a
+    layer starts noticeably *shrinking* its input rather than close to the identity.
+    That compounds across layers (e.g. 0.7^8 ~ 0.06), making a deep flow start far
+    from the base distribution's scale and harder to train. Adding this offset before
+    `positive_scale` makes `positive_scale(0 + offset, min_scale) == 1` exactly, so a
+    freshly initialised layer starts at (approximately) the identity.
+    """
+    target = 1.0 - min_scale
+    return jnp.log(-jnp.expm1(-target)) + target

fleqx-0.0.2/fleqx/bijectors/_coupling.py

@@ -0,0 +1,104 @@
+"""Coupling bijector."""
+
+from __future__ import annotations
+
+from collections.abc import Callable
+
+import equinox as eqx
+import jax
+import jax.numpy as jnp
+from distreqx.bijectors import (
+    AbstractBijector,
+    AbstractForwardInverseBijector,
+    AbstractFwdLogDetJacBijector,
+    AbstractInvLogDetJacBijector,
+)
+from jaxtyping import Array, PRNGKeyArray
+
+from ._affine_common import identity_init_offset, positive_scale
+
+
+class Coupling(
+    AbstractForwardInverseBijector,
+    AbstractFwdLogDetJacBijector,
+    AbstractInvLogDetJacBijector,
+):
+    """Coupling layer with an elementwise affine transform.
+
+    See [Dinh et al., 2016](https://arxiv.org/abs/1605.08803). Splits the input into
+    two parts. The first `untransformed_dim` elements are left untouched and fed
+    through an MLP conditioner to produce a per-element shift and (positive) scale,
+    which affinely transform the remaining elements.
+    """
+
+    untransformed_dim: int
+    dim: int
+    min_scale: float
+    conditioner: eqx.nn.MLP
+    _is_constant_jacobian: bool
+    _is_constant_log_det: bool
+
+    def __init__(
+        self,
+        key: PRNGKeyArray,
+        *,
+        untransformed_dim: int,
+        dim: int,
+        nn_width: int,
+        nn_depth: int,
+        nn_activation: Callable = jax.nn.relu,
+        min_scale: float = 1e-2,
+    ):
+        """Initializes a `Coupling` bijector.
+
+        **Arguments:**
+
+        - `key`: JAX random key used to initialise the conditioner MLP.
+        - `untransformed_dim`: Number of leading elements left untouched.
+        - `dim`: Total dimension.
+        - `nn_width`: Conditioner hidden layer width.
+        - `nn_depth`: Conditioner hidden layer depth.
+        - `nn_activation`: Conditioner activation function. Defaults to
+            `jax.nn.relu`.
+        - `min_scale`: Lower bound added to the transform's scale, for numerical
+            stability (also keeps the scale strictly positive). Defaults to 0.01.
+        """
+        self.untransformed_dim = untransformed_dim
+        self.dim = dim
+        self.min_scale = min_scale
+        transform_dim = dim - untransformed_dim
+        self.conditioner = eqx.nn.MLP(
+            in_size=untransformed_dim,
+            out_size=2 * transform_dim,
+            width_size=nn_width,
+            depth=nn_depth,
+            activation=nn_activation,
+            key=key,
+        )
+        self._is_constant_jacobian = False
+        self._is_constant_log_det = False
+
+    def _shift_and_scale(self, x_cond: Array) -> tuple[Array, Array]:
+        # Each dimension's (shift, raw_scale) pair is adjacent in the conditioner's
+        # output, not grouped by kind -- i.e. [shift_0, scale_0, shift_1, scale_1, ...].
+        shift, raw_scale = jnp.reshape(self.conditioner(x_cond), (-1, 2)).T
+        offset = identity_init_offset(self.min_scale)
+        return shift, positive_scale(raw_scale + offset, self.min_scale)
+
+    def forward_and_log_det(self, x: Array) -> tuple[Array, Array]:
+        """Computes y = f(x) and log|det J(f)(x)|."""
+        x_cond, x_trans = x[: self.untransformed_dim], x[self.untransformed_dim :]
+        shift, scale = self._shift_and_scale(x_cond)
+        y = jnp.concatenate([x_cond, x_trans * scale + shift])
+        return y, jnp.sum(jnp.log(scale))
+
+    def inverse_and_log_det(self, y: Array) -> tuple[Array, Array]:
+        """Computes x = f^{-1}(y) and log|det J(f^{-1})(y)|."""
+        y_cond, y_trans = y[: self.untransformed_dim], y[self.untransformed_dim :]
+        shift, scale = self._shift_and_scale(y_cond)
+        x = jnp.concatenate([y_cond, (y_trans - shift) / scale])
+        return x, -jnp.sum(jnp.log(scale))
+
+    def same_as(self, other: AbstractBijector) -> bool:
+        """Returns True if this bijector is guaranteed to be the same as `other`."""
+        return self is other

fleqx-0.0.2/fleqx/bijectors/_invert.py

@@ -0,0 +1,62 @@
+"""Bijector inversion."""
+
+from __future__ import annotations
+
+from distreqx.bijectors import AbstractBijector
+from jaxtyping import Array
+
+
+class Invert(AbstractBijector):
+    """Inverts a bijector, swapping its forward and inverse directions.
+
+    Bijectors are usually implemented with their cheaper direction as "forward". For
+    a coupling flow, the inverse transform is used by `log_prob` and the forward
+    transform by `sample`, so wrapping the flow's bijector in `Invert` swaps which of
+    the two is fast -- useful for prioritising fast maximum-likelihood training,
+    which repeatedly evaluates `log_prob`.
+    """
+
+    bijector: AbstractBijector
+    _is_constant_jacobian: bool
+    _is_constant_log_det: bool
+
+    def __init__(self, bijector: AbstractBijector):
+        """Initializes an `Invert` bijector.
+
+        **Arguments:**
+
+        - `bijector`: The bijector to invert.
+        """
+        self.bijector = bijector
+        self._is_constant_jacobian = bijector.is_constant_jacobian
+        self._is_constant_log_det = bijector.is_constant_log_det
+
+    def forward(self, x: Array) -> Array:
+        """Computes y = f(x)."""
+        return self.bijector.inverse(x)
+
+    def inverse(self, y: Array) -> Array:
+        """Computes x = f^{-1}(y)."""
+        return self.bijector.forward(y)
+
+    def forward_log_det_jacobian(self, x: Array) -> Array:
+        """Computes log|det J(f)(x)|."""
+        return self.bijector.inverse_log_det_jacobian(x)
+
+    def inverse_log_det_jacobian(self, y: Array) -> Array:
+        """Computes log|det J(f^{-1})(y)|."""
+        return self.bijector.forward_log_det_jacobian(y)
+
+    def forward_and_log_det(self, x: Array) -> tuple[Array, Array]:
+        """Computes y = f(x) and log|det J(f)(x)|."""
+        return self.bijector.inverse_and_log_det(x)
+
+    def inverse_and_log_det(self, y: Array) -> tuple[Array, Array]:
+        """Computes x = f^{-1}(y) and log|det J(f^{-1})(y)|."""
+        return self.bijector.forward_and_log_det(y)
+
+    def same_as(self, other: AbstractBijector) -> bool:
+        """Returns True if this bijector is guaranteed to be the same as `other`."""
+        if type(other) is Invert:
+            return self.bijector.same_as(other.bijector)
+        return False