Release version 2.0

This MR aims at finalizing the package to proceed with its public release, as a clean 2.0 version.

Release To-Do List

Must have

Polish up the code:

• add ClassVar type annotations to class attributes
• add NoReturn type annotations to exception-raising functions (if any)
• clean up __init_subclass__ implementations see docs
• verify module docstrings;
• verify __all__ statements
• garbage-collect unused statements in abstract methods
• garbage-collect deprecated Model.compute_loss method
• implement pending hotfixes (notably to Readme and Vector classes)

Add required metadata:

• add LICENSE file
• add license-related header in all code files
• add AUTHORS file
• update package version (use a VERSION file?)

Test the package extensively:

• run tox with python [3.8, 3.9, 3.10, 3.11]
• drop support for python 3.7*

*Support for python 3.7 was dropped due to functions such as typing.get_args and typing.get_origin having been introduced with Python 3.8.

Could have

Improve dependency specifications:

• update dependencies' versions (possibly running tests, e.g. for TF)
• update black version (~=23.0)
• revise version requirements for pylint and mypy (current are OK and CI/CD uses the latest anyway)

Add more unit tests:

• units tests for Vector (see MR !22 (merged))

Add coverage stats:

• compute (and report in some sort) tests' coverage

Unplanned additions

• Modularize tests execution to be resilient to tensorflow and/or torch not being available.

declearn/aggregator/__init__.py

 # coding: utf-8
 
-"""Framework-agnostic Vector aggregation API and tools."""
+# Copyright 2023 Inria (Institut National de Recherche en Informatique
+# et Automatique)
+#
+# 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.
+
+"""Model updates aggregating API and implementations. An Aggregator
+is typically meant to be used on a round-wise basis by the orchestrating
+server of a centralized federated learning process, to aggregate the
+client-wise model updated into a Vector that may then be used as "gradients"
+by the server's Optimizer to update the global model.
+
+This declearn submodule provides with:
+* Aggregator : abstract class defining an API for Vector aggregation
+* AveragingAggregator : average-based-aggregation Aggregator subclass
+* GradientMaskedAveraging : gradient Masked Averaging Aggregator subclass
+"""
 
 from ._api import Aggregator
 from ._base import AveragingAggregator