dataset_info: | |
features: | |
- name: audio_token | |
sequence: | |
sequence: | |
sequence: int64 | |
- name: audio_duration | |
dtype: float64 | |
- name: nemo | |
dtype: string | |
- name: cer | |
dtype: float64 | |
- name: text | |
dtype: string | |
splits: | |
- name: train | |
num_bytes: 23162502823 | |
num_examples: 8282987 | |
download_size: 5158446109 | |
dataset_size: 23162502823 | |
configs: | |
- config_name: default | |
data_files: | |
- split: train | |
path: data/train-* | |