metadata
dataset_info:
- config_name: corpus
features:
- name: corpus_id
dtype: int32
- name: image
dtype: image
splits:
- name: corpus
num_bytes: 246268640
num_examples: 1000
download_size: 245689137
dataset_size: 246268640
- config_name: train
features:
- name: query
dtype: string
- name: pos_id
dtype: int32
- name: neg_ids
list: int32
- name: pos_score
dtype: float32
- name: neg_scores
list: float32
splits:
- name: train
num_bytes: 193346
num_examples: 1000
download_size: 129220
dataset_size: 193346
configs:
- config_name: corpus
data_files:
- split: corpus
path: corpus/corpus-*
- config_name: train
data_files:
- split: train
path: train/train-*