Changelog

Generated from merged pull requests with build_tools/changelog.py; see that file’s docstring for how classification works.

0.1.0

First release on PyPI. PyQit puts a Trainer, a DataModule and model classes on top of PennyLane QNodes, so training a variational circuit is a fit call. PyTorch and PyTorch Lightning are optional: pyqit.set_backend("torch") moves the same code onto Lightning.

What is added:

• Models. VQCClassifier, VQCRegressor, DataReuploadingClassifier and the hybrid DressedQuantumClassifier. QuantumLayer, DenseLayer and DenseClassifier are the stages hybrids are built from.

• Circuits. Embeddings AngleEmbedding, HadamardAngleEmbedding, AmplitudeEmbedding, IQPEmbedding and ZZFeatureMap; ansatzes SELAnsatz, RealAmplitudesAnsatz, EfficientSU2Ansatz, SimplifiedTwoDesignAnsatz, BasicEntanglerAnsatz and CNOTLadderAnsatz; measurements measure_probs, measure_expval_z, measure_expval_x and measure_parity_z.

• Training. Trainer with per-weight-group learning rates, diff_method selection, and a check_bp=True barren-plateau pre-flight. EarlyStopping, ModelCheckpoint and HistoryCallback work on both backends; checkpoints hold weights, optimizer state and history, and resume from any of them.

• Data. DataModule splits, normalizes on the train split only, and prescales inputs for the model’s embedding. DataModule.save/load and for_prediction reuse a fitted DataModule on new rows.

• Pipelines. QuantumPipeline chains stages sequentially, as an ensemble, or trains them jointly as one hybrid network.

• Losses. MSELoss, HingeLoss, CrossEntropyLoss, or any callable.

• Devices. Any PennyLane device name, lightning.qubit and the PennyLane-Qiskit plugin included, via the qiskit extra.

Bug fixes

• Bugfixes (#35) by @phoeenniixx

• bugfixes and removal of dead code (#20) by @phoeenniixx

Enhancements

• add diff_method, run the nb and parameter shift debug (#46) by @phoeenniixx

• update tutorials and add tags collection (#43) by @phoeenniixx

• Add dm checkpointing, update model checkpoint, chack_bp debug for joint pipeline (#42) by @phoeenniixx

• Add pipeline support for hybrid models (#40) by @phoeenniixx

• Add new Models (#38) by @phoeenniixx

• add new anastz (#37) by @phoeenniixx

• refactor vqc and add ZZFeatureMap (#36) by @phoeenniixx

• Improve API (#34) by @phoeenniixx

• add test and validate API end points to trainer and pipeline (#33) by @phoeenniixx

• debugging and add tests for datamodule and pipeline (#31) by @phoeenniixx

• add readthedocs (#28) by @phoeenniixx

• add pennylane tensor as return_format option (#24) by @phoeenniixx

• Refactor Trainer (#21) by @phoeenniixx

• Major - 2 (#17) by @phoeenniixx

• Major updates - 2 (#3) by @phoeenniixx

• add vectorization (#1) by @phoeenniixx

Documentation

• update docs to use tags (#44) by @phoeenniixx

• update docs (#41) by @phoeenniixx

• update docs (#39) by @phoeenniixx

• add new pyqit icon (#29) by @phoeenniixx

• Add docs, community md etc (#25) by @phoeenniixx

• Add a new callbacks nb and fix a cross-entropy bug found due to new pennylane ver (#23) by @phoeenniixx

• update nbs, Readme and add workflows to run nbs (#22) by @phoeenniixx

Maintenance

• [Dependabot](deps): Update scikit-base requirement from <1.1.0 to <1.2.0 (#18) by @app/dependabot

• [Dependabot](deps): Bump actions/setup-python from 6 to 7 (#16) by @app/dependabot

• [Dependabot](deps): Bump actions/checkout from 6 to 7 (#15) by @app/dependabot

• [Dependabot](deps): Update scikit-base requirement from <0.14.0 to <1.1.0 (#12) by @app/dependabot

• [Dependabot](deps-dev): Update setuptools requirement from >=70.0.0 to >=82.0.1 (#11) by @app/dependabot

• [Dependabot](deps-dev): Update torch requirement from !=2.0.1,<3.0.0,>=2.0.0 to !=2.0.1,>=2.11.0,<3.0.0 (#10) by @app/dependabot

• [Dependabot](deps): Update scikit-learn requirement from <2.0,>=1.2 to >=1.7.2,<2.0 (#9) by @app/dependabot

• [Dependabot](deps): Update pandas requirement from <3.1.0,>=1.3.0 to >=2.3.3,<3.1.0 (#8) by @app/dependabot

• [Dependabot](deps-dev): Update ipywidgets requirement from <9.0.0,>=8.0.1 to >=8.1.8,<9.0.0 (#7) by @app/dependabot

Needs a label

• Minor update - 1 (#13) by @phoeenniixx

Contributors

@phoeenniixx