metadata
dataset_info:
features:
- name: id
dtype: int64
- name: image_id
dtype: int64
- name: audio
dtype:
audio:
decode: false
- name: image
dtype:
image:
decode: false
- name: text
dtype: string
- name: duration
dtype: float32
- name: timecode
dtype: string
- name: speaker
dtype: string
- name: gender
dtype: string
- name: nationality
dtype: string
- name: speed
dtype: float32
- name: disfluency_pos
dtype: string
- name: disfluency_val
dtype: string
splits:
- name: validation
num_bytes: 55043673499.742
num_examples: 202654
download_size: 51627673529
dataset_size: 55043673499.742
configs:
- config_name: default
data_files:
- split: validation
path: data/validation-*