metadata
dataset_info:
features:
- name: split
dtype: string
- name: idx
dtype: int64
- name: text_tokens
sequence: int64
- name: audio_tokens
sequence: int64
- name: text
dtype: string
- name: whisper_transcript
dtype: string
- name: b_words
sequence: string
splits:
- name: train.clean.100
num_bytes: 58202606
num_examples: 9515
- name: train.clean.360
num_bytes: 215671964
num_examples: 35598
- name: train.other.500
num_bytes: 330697199
num_examples: 56343
- name: validation.clean
num_bytes: 2511334
num_examples: 586
- name: validation.other
num_bytes: 2984808
num_examples: 787
- name: test.clean
num_bytes: 2473488
num_examples: 538
- name: test.other
num_bytes: 3352951
num_examples: 881
download_size: 120427336
dataset_size: 615894350
configs:
- config_name: default
data_files:
- split: train.clean.100
path: data/train.clean.100-*
- split: train.clean.360
path: data/train.clean.360-*
- split: train.other.500
path: data/train.other.500-*
- split: validation.clean
path: data/validation.clean-*
- split: validation.other
path: data/validation.other-*
- split: test.clean
path: data/test.clean-*
- split: test.other
path: data/test.other-*