metadata
dataset_info:
features:
- name: spk
dtype: int64
- name: age
dtype: int64
- name: gender
dtype: string
- name: utt_name
dtype: int64
- name: utt_text
dtype: string
- name: utt_accuracy
dtype: int64
- name: utt_completeness
dtype: float64
- name: utt_fluency
dtype: int64
- name: utt_prosodic
dtype: int64
- name: utt_total
dtype: int64
- name: words
dtype: string
- name: words_accuracy
dtype: string
- name: words_stress
dtype: string
- name: words_total
dtype: string
- name: phones
dtype: string
- name: phones_godness
dtype: string
- name: duration
dtype: float64
- name: file_name
dtype: string
- name: Transcript
dtype: string
- name: CanonicalOld
dtype: string
- name: Canonical
dtype: string
- name: Error
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 333634943
num_examples: 2500
- name: test
num_bytes: 290174138
num_examples: 2500
download_size: 612746792
dataset_size: 623809081
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*