Datasets:
Tasks:
Text Retrieval
Modalities:
Text
Formats:
parquet
Sub-tasks:
document-retrieval
Languages:
Chinese
Size:
1M - 10M
License:
metadata
dataset_info:
features:
- name: docid
dtype: string
- name: title
dtype: string
- name: text
dtype: string
- name: emb
sequence: float32
splits:
- name: train
num_bytes: 16896544391
num_examples: 4934368
download_size: 19326395123
dataset_size: 16896544391