metadata
dataset_info:
features:
- name: file
dtype: string
- name: index
dtype: string
- name: txt_embed
sequence:
sequence: float32
- name: txt_raw_embed
sequence:
sequence: float32
- name: txt_raw
dtype: string
- name: txt
dtype: string
- name: flac
dtype:
audio:
decode: false
splits:
- name: train
num_bytes: 2424498690.875
num_examples: 50001
- name: test
num_bytes: 53002206
num_examples: 999
download_size: 2559177723
dataset_size: 2477500896.875
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*