Merge pull request #330 from mblondel:release_v0.5.5

PiperOrigin-RevId: 482426902

Author: JAXopt authors

docs/changelog.rst

@@ -1,13 +1,28 @@
 Changelog
 =========
 
-Main branch
------------
+Version 0.5.5
+-------------
 
 New features
 ~~~~~~~~~~~~
 
-- Added example :doc:`/notebooks/deep_learning/maml` by Fabian Pedregosa based on initial code by Paul Vicol and Eric Jiang.
+- Added MAML example by Fabian Pedregosa based on initial code by Paul Vicol and Eric Jiang.
+- Added the possibility to stop LBFGS after a line search failure, by Zaccharie Ramzi.
+- Added gamma to LBFGS state, by Zaccharie Ramzi.
+- Added :class:`jaxopt.BFGS`, by Mathieu Blondel.
+- Added value_and_grad option to all gradient-based solvers, by Mathieu Blondel.
+- Added Fenchel-Young loss, by Quentin Berthet.
+- Added :func:`projection_sparse_simplex <jaxopt.projection.projection_sparse_simplex>`, by Tianlin Liu.
+
+Bug fixes and enhancements
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+- Fixed missing args,kwargs in resnet example, by Louis Béthune.
+- Corrected the implicit diff examples, by Zaccharie Ramzi.
+- Small optimization in l2-regularized semi-dual OT, by Mathieu Blondel.
+- Numerical stability improvements in :class:`jaxopt.LevenbergMarquardt`, by Amir Saadat.
+- Dtype consistency in LBFGS, by Alex Botev.
 
 Deprecations
 ~~~~~~~~~~~~
@@ -16,6 +31,11 @@ Deprecations
   :class:`jaxopt.CvxpyQP`, :class:`jaxopt.OSQP`, :class:`jaxopt.BoxOSQP` and
   :class:`jaxopt.EqualityConstrainedQP` instead.
 
+Contributors
+~~~~~~~~~~~~
+
+Alex Botev, Amir Saadat, Fabian Pedregosa, Louis Béthune, Mathieu Blondel, Quentin Berthet, Tianlin Liu, Zaccharie Ramzi.
+
 Version 0.5
 -----------
 

docs/constrained.rst

@@ -92,6 +92,7 @@ The following operators are available.
     jaxopt.projection.projection_non_negative
     jaxopt.projection.projection_box
     jaxopt.projection.projection_simplex
+    jaxopt.projection.projection_sparse_simplex
     jaxopt.projection.projection_l1_sphere
     jaxopt.projection.projection_l1_ball
     jaxopt.projection.projection_l2_sphere

docs/index.rst

@@ -46,7 +46,6 @@ Alternatively, it can be be installed from sources with the following command::
    implicit_diff
    objective_and_loss
    line_search
-   developer
 
 .. toctree::
    :maxdepth: 1
@@ -69,6 +68,7 @@ Alternatively, it can be be installed from sources with the following command::
    changelog
    Source code <https://github.com/google/jaxopt>
    Issue tracker <https://github.com/google/jaxopt/issues>
+   developer
 
 Support
 -------

docs/notebooks/index.rst

@@ -28,7 +28,7 @@ Deep learning
 
 .. raw:: html
 
-    <div class="sphx-glr-thumbcontainer" tooltip="Adversarial training with jaxopt.">
+    <div class="sphx-glr-thumbcontainer" tooltip="Adversarial training with JAXopt.">
 
 .. only:: html
 
@@ -79,4 +79,4 @@ Implicit Differentiation
 
 .. raw:: html
 
-    </div>
+    </div>

jaxopt/version.py

@@ -14,4 +14,4 @@
 
 """JAXopt version."""
 
-__version__ = "0.5"
+__version__ = "0.5.5"