README.md

metadata

readme: '::readme::'
library_name: dataset
tags:
  - audio
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: sentence
      dtype: string
    - name: path
      dtype: string
    - name: duration
      dtype: float32
  splits:
    - name: train
      num_examples: 10115

Masri Audio Raw (Pre-SAM)

This dataset contains 10115 raw audio files packaged for SAM cleaning. Structure: 100% Train (No splits yet). Audio: 16kHz Embedded Parquet.