metadata
dataset_info:
config_name: fa
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: 3544929223.413
num_examples: 3981
- name: validation
num_bytes: 1440270416.228
num_examples: 1636
- name: test
num_bytes: 1679224745.288
num_examples: 1928
download_size: 5848777724
dataset_size: 6664424384.929
configs:
- config_name: fa
data_files:
- split: train
path: fa/train-*
- split: validation
path: fa/validation-*
- split: test
path: fa/test-*