1.0.3

What's Changed

• BUG: avoid KeyError in train.frame_classification, fix #779 by @NickleDave in #781

Full Changelog: 1.0.2...1.0.3

1.0.2

What's Changed

• DOC: Update autoannoate tutorial, fix #768 by @NickleDave in #769
• docs: add meriablue as a contributor for doc by @allcontributors in #770
• ENH: fix frame classification model to work with BioSoundSegBench by @NickleDave in #774

Full Changelog: 1.0.1...1.0.2

1.0.1

What's Changed

• DEV: Require Python>=3.10, update other dependencies by @NickleDave in #765
• BUG: minor fixes to run with biosoundsegbench dataset by @NickleDave in #766

Full Changelog: 1.0.0...1.0.1

1.0.0

What's Changed

• Version 1.0 by @NickleDave in #639
• DOC: Add version 1.0 announcement [skip ci] by @NickleDave in #640
• DOC: Add version 1.0 announcement to README [skip ci] by @NickleDave in #642
• DOC: Add API reference, fix #441 by @NickleDave in #644
• DOC: Revise reference/about.md [skip ci] by @NickleDave in #646
• CLN: Refactor window dataset, add tests by @NickleDave in #654
• ENH: Use crowsetta 5.0, fixes #526 by @NickleDave in #657
• Prep datasets as directories, fixes #649 #650 #651 by @NickleDave in #658
• BUG: Fix usage of crowsetta 5.0 by @NickleDave in #659
• BUG/CLN: Fixup prepare dataset as directory by @NickleDave in #660
• replace union to future syntax by @Ja-sonYun in #661
• docs: add Ja-sonYun as a contributor for code by @allcontributors in #662
• BUG: Change labelmap for validation step of FrameWindowClassificationModel only, fix #664 by @NickleDave in #665
• Refactor api, fixes #663 by @NickleDave in #666
• Refactor frame classification, add audio datasets + DAS model, fixes #630 #652 #667 by @NickleDave in #670
• Remove initial implementation of DAS model by @NickleDave in #675
• Add decorators to register models and model families, fix #623 by @NickleDave in #676
• ENH: Add ED-TCN model by @NickleDave in #677
• ENH: Determine network kwargs dynamically in vak.models.get by @NickleDave in #680
• BUG/CLN: Add/fix log statements in FrameClassificationModel by @NickleDave in #681
• Add parametric UMAP model family by @NickleDave in #688
• Fix and run linting by @NickleDave in #690
• TST/CLN: Fix unit tests by @NickleDave in #693
• docs: add JacquelineGoe as a contributor for bug by @allcontributors in #695
• BUG: Make distance metrics return tensors, fix #700 #701 by @NickleDave in #702
• docs: add VenetianRed as a contributor for bug by @allcontributors in #703
• docs: add zhileiz1992 as a contributor for bug by @allcontributors in #705
• DOC, Make minor doc fixes, fixes #704 by @NickleDave in #706
• docs: add zhileiz1992 as a contributor for code by @allcontributors in #711
• docs: add marisbasha as a contributor for ideas by @allcontributors in #712
• docs: add marisbasha as a contributor for code by @allcontributors in #713
• DOC: Add eval to autoannotate, fix #460 by @NickleDave in #715
• docs: add vivinastase as a contributor for ideas by @allcontributors in #716
• ENH: Minimize frame classification dataset size, fix #717 by @NickleDave in #718
• BUG: Fix models to log train loss on step, fixes #720 by @NickleDave in #722
• CLN: Rename "segment error rate" to "character error rate", fix #721 by @NickleDave in #723
• docs: add danielmk as a contributor for doc by @allcontributors in #732
• DOC: fix autoannotate tutorial configs, fixes #734 by @NickleDave in #735
• TST: Fix test in common.labels by @NickleDave in #747
• ENH: Switch to version 1.0 of config file format, fix #685 #345 #748 by @NickleDave in #750
• ENH: Add lightning.Trainer config, fix #691 #687 #742 #745 by @NickleDave in #752
• CLN: Refactor model abstraction, fix #737 #726 by @NickleDave in #753
• CLN/ENH: Rename and refactor datapipes, add datasets; fix #574 #724 #754 by @NickleDave in #755

New Contributors

• @Ja-sonYun made their first contribution in #661

Full Changelog: 0.8.2...1.0.0

0.8.2

What's Changed

• BUG: fix default for post_tfm_kwargs, fixes Inconsistent syllable error by @zhileiz1992 in #710

New Contributors

• @zhileiz1992 made their first contribution in #710

Full Changelog: 0.8.1...0.8.2

1.0.0a3

What's Changed

• BUG: Make distance metrics return tensors, fix #700 #701 by @NickleDave in #702
• docs: add VenetianRed as a contributor for bug by @allcontributors in #703

Full Changelog: 1.0.0a2...1.0.0a3

1.0.0a2

What's Changed

• BUG: Have to_segments return all Nones for no segments, fix #634 by @NickleDave in #636
• docs: add nhoglen as a contributor for bug by @allcontributors in #637
• Version 1.0 by @NickleDave in #639
• DOC: Add version 1.0 announcement [skip ci] by @NickleDave in #640
• DOC: Add version 1.0 announcement to README [skip ci] by @NickleDave in #642
• DOC: Add API reference, fix #441 by @NickleDave in #644
• DOC: Revise reference/about.md [skip ci] by @NickleDave in #646
• CLN: Refactor window dataset, add tests by @NickleDave in #654
• ENH: Use crowsetta 5.0, fixes #526 by @NickleDave in #657
• Prep datasets as directories, fixes #649 #650 #651 by @NickleDave in #658
• BUG: Fix usage of crowsetta 5.0 by @NickleDave in #659
• BUG/CLN: Fixup prepare dataset as directory by @NickleDave in #660
• replace union to future syntax by @Ja-sonYun in #661
• docs: add Ja-sonYun as a contributor for code by @allcontributors in #662
• BUG: Change labelmap for validation step of FrameWindowClassificationModel only, fix #664 by @NickleDave in #665
• Refactor api, fixes #663 by @NickleDave in #666
• Refactor frame classification, add audio datasets + DAS model, fixes #630 #652 #667 by @NickleDave in #670
• Remove initial implementation of DAS model by @NickleDave in #675
• Add decorators to register models and model families, fix #623 by @NickleDave in #676
• ENH: Add ED-TCN model by @NickleDave in #677
• ENH: Determine network kwargs dynamically in vak.models.get by @NickleDave in #680
• BUG/CLN: Add/fix log statements in FrameClassificationModel by @NickleDave in #681
• Add parametric UMAP model family by @NickleDave in #688
• Fix and run linting by @NickleDave in #690
• TST/CLN: Fix unit tests by @NickleDave in #693
• docs: add JacquelineGoe as a contributor for bug by @allcontributors in #695

New Contributors

• @Ja-sonYun made their first contribution in #661

Full Changelog: 0.8.0...1.0.0a2

1.0.0a1

What's Changed

• BUG: Have to_segments return all Nones for no segments, fix #634 by @NickleDave in #636
• docs: add nhoglen as a contributor for bug by @allcontributors in #637
• Version 1.0 by @NickleDave in #639

Full Changelog: 0.8.0...1.0.0a1

0.8.1

What's Changed

• BUG: Have to_segments return all Nones for no segments, fix #634 by @NickleDave in #636
• docs: add nhoglen as a contributor for bug by @allcontributors in #637

Full Changelog: 0.8.0...0.8.1

0.8.0

0.8.0 release notes

2023-02-09

Added

• Add options for how audio.to_spect calls dask.bag, to help with memory issues when processing large files
  #611. Fixes #580.
• Add ability to run evaluation of models with and without post-processing transforms. This is done by specifying an option post_tfm_kwargs in the [EVAL] or [LEARNCURVE] sections of a .toml configuration file. If the option is not specified, then models are evaluated as they were previously, by converting the predicted label for each time bin to a label for each continuous segment, represented as a string. If the option is specified, then the post-processing is applied to the model predictions before converting to strings. Metrics are computed for outputs with and without post-processing, to be able to compare the two. #621. Fixes #472.
• vak.core.eval now logs computed evaluation metrics so they can be quickly inspected in the terminal or log files before full analysis
  #621. Fixes #471.

Changed

• Rewrite post-processing transforms applied to network outputs as transforms, with functional and class implementations,
  to make it possible to compose these transforms, and more easily evaluate model performance with and without them
  #621. Fixes #537.