arabic_quranic_asr / README.md
Sadique5's picture
Upload dataset
b0e37cb verified
|
raw
history blame
1.07 kB
metadata
language:
  - ar
license: mit
size_categories:
  - 10K<n<100K
task_categories:
  - automatic-speech-recognition
  - text-to-speech
dataset_info:
  features:
    - name: text
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: input_features
      sequence:
        sequence: float32
    - name: labels
      sequence: int64
  splits:
    - name: train
      num_bytes: 17975826804.530933
      num_examples: 15027
    - name: test
      num_bytes: 1997712834.4690664
      num_examples: 1670
  download_size: 7908179259
  dataset_size: 19973539639
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*

Dataset details

This dataset contains quran recitations of every ayats or verses. Also contains 10k unique words from quran.

Dataset Purpose

This dataset can be used to train ASR models that can be used for teaching beginners to recite quran. It can also be used for training TTS models that produces quran recitations in a way so that beginners can easily learn.