metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: audio_id
dtype: int64
- name: transcription
dtype: string
- name: tokens
dtype: string
- name: word_timestamps
dtype: string
- name: ner_tags
dtype: string
- name: ner_timestamps
dtype: string
- name: ner_two_classes
dtype: string
- name: ner_four_classes
dtype: string
splits:
- name: train
num_bytes: 2112999522.625
num_examples: 3627
download_size: 2069590011
dataset_size: 2112999522.625
configs:
- config_name: default
data_files:
- split: train
path: data/train-*