v0.13.0

What's Changed

• Upgrade dependencies via pip-compile --upgrade by @acmiyaguchi in #41
• Fix #40 - Implement undersampling of classes and improved randomization/augmentation in classify-nn model by @acmiyaguchi in #42
• Revised Clustering Scripts by @ddudley18 in #35
• Add larger model and prediction command by @acmiyaguchi in #43

Full Changelog: v0.12.0...v0.13.0

v0.12.0

What's Changed

• fix #10 - updated classify dataloader by @JiangyueYu in #37
• Fix broken test in model classifier datasets by @acmiyaguchi in #38
• Update ci configuration - caching, main branch ci, and diff on pre-commit failures by @acmiyaguchi in #39
• Add neural net based classifier with multi-label prediction by @acmiyaguchi in #36

New Contributors

• @JiangyueYu made their first contribution in #37

Full Changelog: v0.11.0...v0.12.0

v0.11.0

What's Changed

• Resolve IndexErrors with get_motif_pairs in DataLoader by @bran22 in #32
• Change default parameters for tilenet model and add training augmentation by @acmiyaguchi in #30

Full Changelog: v0.10.0...v0.11.0

v0.10.0: Merge pull request #23 from acmiyaguchi/iterable-dataloader

Implement iterable dataloader for triplets

v0.9.0: Merge pull request #8 from acmiyaguchi/motif-features

Add classifier model with motif features [behind feature flag]

v0.8.0: Merge pull request #7 from acmiyaguchi/motif-features

Add classification model using motif features

v0.7.0: Merge pull request #6 from acmiyaguchi/motif-features

Use multioutput for classifier

v0.6.0: Merge pull request #5 from acmiyaguchi/submission

Add workflow for performing main classification task

v0.5.0: Merge pull request #4 from acmiyaguchi/labelstudio

Add initial work for label studio and fix bugs in loading audio tracks

v0.4.0: Merge pull request #3 from acmiyaguchi/gbdt

Add call/no-call training with lightgbm