Release 0.2.4 - QOL improvements and bug fixes

• only append align_json if defined
• use code beautifier black and enforce isort ordering
• refactor exemplar class
• enforce segment validation
• expose minimum sizes in prepare_audio_corpora

Release 0.2.3

bug fixes

• preserves noise tags with brackets
• docker image functional again
  new features
• corpus splitting can occur based on number of words

0.1.20

Merge pull request #75 from finos/matthew/FixAudioSplitter

fix audio splitter

0.1.19

Safety rails put in place around most scripts
Pep8 style in place now

0.1.18

bump version

0.1.17

Release 0.1.17

0.1.16

Recent code changes include extending the nonsilence noise list and refactoring the corpus object for stability

0.1.13

Relaxing requirements for pandas since functionality is uncommonly used.

0.1.12: Merge pull request #30 from finos-voice/matthew/tweakCleanUp

improve clean_formatting routines

0.1.11

Bugfix release stopping regex explosion for select test cases.