metadata
dataset_info:
features:
- name: duration
dtype: float64
- name: transcription
dtype: string
- name: french
dtype: string
- name: language
dtype: string
- name: input_features
sequence:
sequence: float32
- name: input_length
dtype: int64
- name: labels
sequence: int64
- name: text_length
dtype: int64
splits:
- name: train
num_bytes: 139298795190
num_examples: 144995
download_size: 20894722994
dataset_size: 139298795190
configs:
- config_name: default
data_files:
- split: train
path: data/train-*