v3.0.1

• Fix model auto selection logic

v3.0.0

• Add support for sherpa-onnx and Nvidia's parakeet model
• Add support for GigaAM for Russian via onnx-asr
• Add --stt-library to select speech-to-text library (deprecate --use-transformers)
• Default --model to "auto" (prefer parakeet)
• Add Docker build here
• Default --language to "auto"
• Add --cpu-threads for faster-whisper (@Zerwin)

v2.5.0

• Add support for HuggingFace transformers Whisper models (--use-transformers)

v2.4.0

2.4.0

• Add "auto" for model and beam size (0) to select values based on CPU

v2.3.0

• Bump faster-whisper package to 1.1.0
• Supports model turbo for faster processing

v2.2.0

2.2.0

• Bump faster-whisper package to 1.0.3

v2.0.0

2.0.0

• Use faster-whisper PyPI package
• --model can now be a HuggingFace model ID like Systran/faster-distil-whisper-small.en as well as one of:
  • tiny-int8
  • tiny.en
  • tiny
  • base-int8
  • base.en
  • base
  • small-int8
  • small.en
  • small
  • medium-int8
  • medium.en
  • medium
  • large-v1
  • large-v2
  • large-v3
  • large
  • distil-large-v2
  • distil-medium.en
  • distil-small.en
• --model may also be a path to a custom model directory