v0.0.12

What's Changed

• Reduce Memory usage by clearing cache in whisper
• Add lyrics to midi file
• Split word by note changes
• Upgrade UltraStar Format to Version 1.2.0 (Use of VIDEOURL)
• Use yt with cookies
• Some docker container improvements
• Fix keep-cache option
• Fix numbers in lyrics and transcribtion
• Fix model path option
• Fix error in PDF sheet generation
• Fix hypen language download
• Fix install scripts
• Some bug fixes and improved error handling and logs

PR

• Cherry pick from evaluation by @rakuri255 in #159
• Refactor/class refactorings by @rakuri255 in #162
• add cookiefile option and pass it to yt-dlp by @BWagener in #184
• keep_cache parameter did not work by @agwosdz in #189
• New feature: Multilingual num2words by @agwosdz in #188
• "/" in whisper align model string get interpreted as path by @agwosdz in #190
• Added error handling to PDF sheet generation by @agwosdz in #192
• optimize container related configuration and documentation by @BWagener in #186
• Added VideoURL and bump format version to 1.2.0 by @rakuri255 in #193

New Contributors

• @agwosdz made their first contribution in #189

Full Changelog: v0.0.11...v0.0.12

v0.0.11

What's Changed

• Better linebreak calculation
• Remove cache folder when finished
• Remove audio from yt video
• Added install and start scripts
• Added Docker support
• Added sheet music generation

PRs

• Added sheet music generation by @rakuri255 in #158
• remove audio from yt video by @rakuri255 in #144
• Docker Support with GPU by @binarygeek119 in #148
• Docker ci by @rakuri255 in #154
• added docker-compose.yml file and fix commands by @binarygeek119 in #153
• Feature/pitch clustering refactoring by @BWagener in #149
• Feature/pitch clustering by @BWagener in #145

New Contributors

• @binarygeek119 made their first contribution in #148

Full Changelog: v0.0.10...v0.0.11

v0.0.10

Fix:

• remove whitespace from the beginning and end of title and artist
• image conversion to jpeg for transparent or RGBA
• index out of range error in when list is empty from musicbrainz

Version: 0.0.9

This is the initial "source code" release.
I have simply added all release notes below to make it easier to track what has changed.

• Fix:
  • Re-Pitch mode now re-pitch the audio again
  • Re-Pitch mode now show the text and lines in plot

Version: 0.0.8

Date: 2024.01.03

• Changes:
  • Plot words
• Fix:
  • Missing word lines in plot

Version: 0.0.7

Date: 2023.12.29

• Changes:
  • Added format version support for 0.3.0, 1.0.0 and 1.1.0

Version: 0.0.6

Date: 2023.12.28

• Changes:
  • Optimized the removing of silence in transcription data
  • Mute the processing audio in parts where no singing is detected
  • Plot muted audio

Version: 0.0.5

Date: 2023.12.23

• Changes:
  • Format GENRE string
  • Extract year from date

Version: 0.0.4

Date: 2023.12.16

• Changes:
  • Optimized the conversion to Mono
  • Removed limitation to mp3 and wav audio formats
  • Added option float32 to demucs

New Contributors

• @rakuri255 made their first contribution in #1
• @bohning made their first contribution in #41
• @BWagener made their first contribution in #89
• @taynaron made their first contribution in #99