metadata
dataset_info:
features:
- name: fluency
dtype: int64
- name: text
dtype: string
- name: speaker
dtype: string
- name: audio
dtype: audio
- name: label_fluency
dtype: string
- name: audio_duration
dtype: float64
- name: speech_rate
dtype: float64
- name: 1gram_repeat
dtype: int64
- name: 2gram_repeat
dtype: int64
- name: 3gram_repeat
dtype: int64
- name: 4gram_repeat
dtype: int64
- name: 5gram_repeat
dtype: int64
- name: input_values
sequence: float32
- name: attention_mask
sequence: int32
- name: labels
dtype: int64
splits:
- name: train
num_bytes: 2183559779
num_examples: 2500
- name: test
num_bytes: 2009219809
num_examples: 2500
download_size: 1569823139
dataset_size: 4192779588
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*