metadata
dataset_info:
features:
- name: index
dtype: int64
- name: filename
dtype: string
- name: content
dtype: string
- name: duration
dtype: float64
- name: codes
dtype: string
- name: spk
dtype: string
- name: id
dtype: int64
- name: output
dtype: string
- name: input_ids
list: int32
- name: attention_mask
list: int8
- name: labels
dtype: float64
splits:
- name: train
num_bytes: 7982177350
num_examples: 2506291
- name: test
num_bytes: 7956041
num_examples: 2509
download_size: 660908155
dataset_size: 7990133391
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*