metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: speaker_id
dtype: string
- name: transcript
dtype: string
- name: accent
dtype: string
- name: duration
dtype: float64
- name: age
dtype: string
- name: gender
dtype: string
- name: whisper-tiny_transcript
dtype: string
- name: whisper-base_transcript
dtype: string
- name: whisper-medium_transcript
dtype: string
- name: whisper-large-v2_transcript
dtype: string
- name: wav2vec2-large-960h-lv60-self_transcript
dtype: string
- name: hubert-large-ls960-ft_transcript
dtype: string
splits:
- name: test
num_bytes: 453619676
num_examples: 2999
- name: val
num_bytes: 289484499
num_examples: 1879
- name: train
num_bytes: 7499881659
num_examples: 48662
download_size: 7892235103
dataset_size: 8242985834
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: val
path: data/val-*
- split: train
path: data/train-*