All pull requests

Pull requests

• Support loading model files from file-like objects in memory

  #1669·Purfview opened Oct 12, 2026Collaborator·

  ·

• Fix speech timestamp restoration at floating-point boundaries

  #1667·ktz03 opened Oct 9, 2026·

  ·

• Raise an error for illegal values in clip_timestamps arg

  #1654·Purfview opened Oct 5, 2026Collaborator·

  ·
  2

• Raise an error for languages unsupported by a model

  #1564·Purfview opened Oct 1, 2026Collaborator·

  ·
  1

• feat: support os.PathLike and pathlib.Path across audio inputs and model paths

  #1489·hsynylmz-umhg opened Sep 14, 2026·

  ·

• Load the transcription stack on first use

  #1480·Kayvan-Zahiri opened Aug 26, 2026·

  ·
  1

• Reset the prompt when multilingual detection changes language

  #1477·Kayvan-Zahiri opened Aug 16, 2026·

  ·
  1

• Add optional language-aware VAD segmentation for multilingual inference

  #1468·Frenzie opened Jul 13, 2026·

  ·

• Add language output to each segment

  #1466·Frenzie opened Jul 9, 2026·

  ·

• Add a simple batch transcribe - only <30s audio files are supported

  #1453·yoadsn opened Jun 5, 2026·

  ·

• Fix merge_punctuations for multi-char words with apostrophes

  #1423·Br1an67 opened Mar 1, 2026·

  ·
  1

• Bugfix: Fix inconsistent VAD defaults in BatchedInferencePipeline

  #1409·Purfview opened Dec 31, 2025Collaborator·

  ·
  2

• Adds optional batched VAD feature

  #1408·Purfview opened Dec 30, 2025Collaborator·

  ·
  9

• VAD Fix: explicit safe context, reduced RAM usage

  #1406·Purfview opened Dec 25, 2025Collaborator·

  ·
  3

• Add language output to each segment

  #1274·erichaydel opened Mar 28, 2025·

  ·
  1

• Update download_dir to specific directory

  #1247·tnt305 opened Feb 17, 2025·

  ·

• Feat: correct type hints

  #1150·Equipo45 opened Nov 17, 2024·

  ·

• Ignore repeated prompt

  #163·heimoshuiyu opened Apr 19, 2023Contributor·

  ·
  4