Fabio/voice_audio_multi
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1
voice_audio_multi/tests/parakeet_tdt/assets
History
Fabio c36510c21e
Some checks failed
Linux Build / Linux cpu (push) Has been cancelled
Details
Linux Build / Linux cpu-clang (push) Has been cancelled
Details
Linux Build / Linux vulkan (push) Has been cancelled
Details
mac-build / macOS CPU (push) Has been cancelled
Details
Nix Build / Nix (cpu) (push) Has been cancelled
Details
Nix Build / Nix (vulkan) (push) Has been cancelled
Details
windows-build / Windows CPU (push) Has been cancelled
Details
Initial import: audio.cpp with Orange Pi documentation
2026-11-04 19:05:22 +01:00
..
2086-149220-0033.wav Initial import: audio.cpp with Orange Pi documentation 2026-11-04 19:05:22 +01:00
README.md Initial import: audio.cpp with Orange Pi documentation 2026-11-04 19:05:22 +01:00

README.md

Parakeet TDT test fixtures

  • 2086-149220-0033.wav
    • 16 kHz mono, ~7.44 s speech clip, extracted from the LibriSpeech test-clean corpus, utterance 2086-149220-0033 ("Well, I don't wish to see it any more, observed Phoebe, turning away her eyes. It is certainly very like the old portrait.").
    • LibriSpeech (Panayotov et al., 2015) is distributed under CC BY 4.0: https://www.openslr.org/12
    • Used by test_golden_transcription.cpp and test_streaming_transcription.cpp as a fixed-input regression check for offline full-context, bounded-window long-form, word timestamps, buffered streaming, finalize, and reset. Both tests accept the installed safetensors directory or a standalone GGUF through --model. The expected transcription was verified against the actual NeMo reference model (nvidia/parakeet-tdt-0.6b-v3 via nemo.collections.asr.models.ASRModel.from_pretrained) for this exact clip, not just eyeballed.