metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
sequence: float32
- name: samplerate
dtype: int64
- name: secs
dtype: float64
- name: speaker_id
dtype: int64
- name: _speaker_id
dtype: 'null'
- name: flag
dtype: 'null'
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 624115466
num_examples: 2905
download_size: 626541257
dataset_size: 624115466
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
language:
- ar
pretty_name: wasmdashai/db-arabic-f1-nn