metadata
dataset_info:
features:
- name: text
dtype: string
- name: ipa
dtype: string
- name: text_confidence
dtype: float64
- name: ipa_confidence
dtype: float64
- name: text_word_count
dtype: int64
- name: ipa_word_count
dtype: int64
- name: ipa_ref
dtype: string
- name: ipa_cer
dtype: float64
- name: abs_ipa_text_word_count_diff
dtype: int64
- name: start
dtype: float64
- name: duration
dtype: float64
- name: audio_url
dtype: string
splits:
- name: train
num_bytes: 62944
num_examples: 78
download_size: 42136
dataset_size: 62944
configs:
- config_name: default
data_files:
- split: train
path: data/train-*