Datasets:
metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 1165116125.04
num_examples: 6216
download_size: 1309837395
dataset_size: 1165116125.04
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: mit
task_categories:
- automatic-speech-recognition
language:
- ar
- en
Please see paper & code for more information: