metadata
dataset_info:
features:
- name: audio
struct:
- name: bytes
dtype: binary
- name: path
dtype: string
- name: gender
dtype: string
- name: accent
dtype: string
- name: speaker_id
dtype: string
- name: text
dtype: string
- name: speaker_name
dtype: string
- name: normalized_text
dtype: string
- name: absolute_audio_url
dtype: string
- name: transcription_length
dtype: int64
- name: standardize_text
dtype: string
- name: audio duration
dtype: float64
- name: word_count
dtype: int64
- name: cwr
dtype: float64
- name: s2n
sequence: float32
splits:
- name: train
num_bytes: 479800652
num_examples: 14739
download_size: 464343562
dataset_size: 479800652
configs:
- config_name: default
data_files:
- split: train
path: data/train-*