README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: channel
      dtype: string
    - name: transcript_whisper
      dtype: string
    - name: title
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: transcript_sensevoice
      dtype: string
    - name: emotion_sensevoice
      sequence: string
    - name: event_sensevoice
      sequence: string
    - name: c50
      dtype: string
    - name: snr
      dtype: string
    - name: speech_duration
      dtype: string
    - name: emotion_emotion2vec
      dtype: string
  splits:
    - name: train
      num_bytes: 544878197109.913
      num_examples: 1478337
  download_size: 527078747956
  dataset_size: 544878197109.913
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

An earlier version of the dataset contains some duplicated data, a fix is on the way. Sorry for the inconvenience

Cantonese Youtube Pseudo-Transcription Dataset

• Contains approximately 10k hours of audio sourced from YouTube
  • Videos are chosen at random, and scraped on a channel basis
  • Includes news, vlogs, entertainment, stories, health
• Columns
  • transcript_whisper: Transcribed using Scrya/whisper-large-v2-cantonese with alvanlii/whisper-small-cantonese for speculative decoding
  • transcript_sensevoice: Transcribed using FunAudioLLM/SenseVoiceSmall
    • used OpenCC to convert to traditional chinese
    • isolated event tags to event_sensevoice
    • isolated emotion tags to emotion_sensevoice
  • snr: Signal-to-noise ratio, extracted from ylacombe/brouhaha-best
  • c50: Speech clarity, extracted from ylacombe/brouhaha-best
  • emotion: Emotion, extracted from emotion2vec/emotion2vec_plus_large
• Processing
  • The full audio is split using WhisperX, using Scrya/whisper-large-v2-cantonese
  • Preliminary filtering includes filtering out phrases like:
    • liking/subscribing to YouTube channel
    • subtitles by [xxxx]
    • Additional filtering is recommended for your own use