metadata
dataset_info:
features:
- name: path
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: condition_on_prev
sequence: int64
- name: whisper_transcript
dtype: string
splits:
- name: train
num_bytes: 631685248
num_examples: 714
- name: validation
num_bytes: 359421903
num_examples: 406
- name: test
num_bytes: 502350075
num_examples: 576
download_size: 1398110261
dataset_size: 1493457226
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*