metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: filename
dtype: string
- name: text
dtype: string
- name: duration
dtype: float64
- name: spk
dtype: string
- name: id
dtype: int64
- name: caption
dtype: string
- name: key
dtype: string
- name: seq_len
dtype: int64
- name: asmr
dtype: string
- name: input_ids
list: int32
- name: attention_mask
list: int8
- name: codes
dtype: string
- name: __index_level_0__
dtype: int64
- name: nano_codes
list:
list: int64
splits:
- name: train
num_bytes: 85837213132
num_examples: 6793807
download_size: 12684964885
dataset_size: 85837213132