All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Add Linear2DBasisQFE
- Move type aliases to a separate module
- Update Makefile
- Add github workflows
- Add TwoQubitMPS and EmbedU layers
- Hat basis negative close to 0 values, sqrt in HatBasisQFE returned nan
- MPS contraction
- Hat basis modules
- MPS modules
- Hat basis QFE to qlayer
- Output shape for batched inputs consistent with PyTorch
- Added variable exponent base to data embeddings
- Added multiple observables option to Measurement Layer
- Added quantum kernel models and ridge regression trainers
- Changed model builders and configs in examples
- Refactored and simplified trainer, better logging
- Naive (non-efficient) estimation of Fisher matrices
- Estimation of effective dimension (see arXiv:2112.04807)
- Hybrid model functionality.
- Capacity metrics estimation routines.
- Synthetic data generation methods.
- Trainers.
- Basic unit tests for all the features.
- Initial documentation for the API.