metadata
dataset_info:
features:
- name: corpus
dtype: string
- name: speaker
dtype: string
- name: sample_rate
dtype: int64
- name: gender
dtype: string
- name: text
dtype: string
- name: l2_level
dtype: string
- name: variety
dtype: string
- name: locale
dtype: string
- name: audio
sequence: float32
- name: age_str
dtype: string
splits:
- name: train
num_bytes: 5686131898
num_examples: 12815
download_size: 4638674329
dataset_size: 5686131898
configs:
- config_name: default
data_files:
- split: train
path: data/train-*