metadata
dataset_info:
features:
- name: cluster_id
dtype: string
- name: doc_id
dtype: string
- name: main_text
dtype: string
- name: image
dtype: image
- name: video
dtype: 'null'
- name: audio
dtype: 'null'
- name: kg_embedding
sequence:
sequence: float32
splits:
- name: train
num_bytes: 197952
num_examples: 1
download_size: 0
dataset_size: 197952
configs:
- config_name: default
data_files:
- split: train
path: data/train-*