metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcript1
dtype: string
- name: transcript2
dtype: string
- name: subcorpus
dtype: string
- name: id
dtype: string
- name: max_variant_counter
dtype: int64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1730909970.666
num_examples: 8558
download_size: 1731514886
dataset_size: 1730909970.666
configs:
- config_name: default
data_files:
- split: train
path: data/train-*