ivrit-ai/ivrit.ai

Issues

• result is worse than regular whisper-large-v3. did I do anything wrong?

  #54 opened 3 months ago
  1

• Issue downloading Knesset Plenum data

  #53 opened 3 months ago
  0

• Extract and Organize Metadata for Audio Files

  #48 opened 4 months ago
  0

• Build a New Tagging App for Collecting Volunteer Readings

  #47 opened 4 months ago
  0

• Design and Characterize Search Interface for Corpus

  #46 opened 4 months ago
  0

• UI: Allow correcting the former clip

  #42 opened 4 months ago
  1

• Train and Evaluate Models on Noise-Filtered Data

  #37 opened 5 months ago
  1

• Automatically Generate Faster-Whisper and GGML Models from HuggingFace Whisper Model

  #36 opened 5 months ago
  1

• Use speaker diarization

  #35 opened 4 months ago
  1

• Better evaluation

  #33 opened 8 months ago
  1

• תצוגת נייד

  #26 opened 10 months ago
  0

• Bug: Google Blocks Access to Interface from Social Media Apps

  #23 opened 10 months ago
  1

• ftc.py: support production WSGI server

  #19 opened a year ago
  0

• Transcription web-page: explicit slow-speed audo setting

  #18 opened a year ago
  1

• Transcription webpage: add statistics

  #16 opened a year ago
  0

• Feature Request: Prevent Submission Without Playing Audio Once to Avoid Mistakes

  #15 opened a year ago
  0

• Transcription web-page: more statistics

  #14 opened a year ago
  0

• Transcription webpage: prefetch data

  #12 opened a year ago
  0

• Transcription webpage: layout modifications

  #11 opened a year ago
  0

• Add per-user settings

  #10 opened a year ago
  1

• Pass audio, transcript directories as arguments to ftc.py

  #9 opened a year ago
  0

• Provide ability to select specific content source for transcribing

  #8 opened 7 months ago
  2

• Add instructions/help screen

  #7 opened a year ago
  1

• Maintain minutes-transcribed counter across session

  #6 opened a year ago
  0

• Provide uniquely-identifying ID (using segment index or otherwise) in transcription screen

  #4 opened a year ago
  3

• Transcribed minutes appears with fractional part, instead of mm:ss

  #3 opened a year ago
  1