A Python library for amortized Bayesian workflows using generative neural networks.
Full Changelog: https://github.com/stefanradev93/BayesFlow/compare/v1.1.5...v1.1.6
Full Changelog: https://github.com/stefanradev93/BayesFlow/compare/v1.1.4...v1.1.5
State of software at JOSS publication.
SimulationMemory
affecting the use of empty folders for initializing a Trainer
;Trainer.train_from_presimulation()
for model comparison tasks;c2st
in computational_utilities
.SetTransformer
with induced pointsEnable PyPI integration through GitHub workflows.
Following multiple improvements and being actively used in multiple projects, the BayesFlow library is ready to move beyond the beta phase!
Features:
permutation='learnable'
when creating an InvertibleNetwork
coupling_design in ["affine", "spline", "interleaved"]
when creating an InvertibleNetwork
inference_network = InvertibleNetwork(num_params=20, coupling_net_settings={'mc_dropout': True})
to get a Bayesian neural network.PMPNetwork
has been added for model comparison according to findings in https://arxiv.org/abs/2301.11873
diagnostics.py
and is accessible as plot_calibration_curves()
experimental
has been added currently containing rectifiers.py
.posterior_calibration_error()
General Improvements:
diagnostics.py
have been improved and prettifiedsensitivity.py
for testing the sensitivity of neural approximators to model misspecificationThe project now also features automatic PyPI publishing. :)
Welcome to the Future!