metadata
dataset_info:
features:
- name: idx
dtype: int64
- name: image
dtype: image
- name: masks
sequence: image
- name: points
sequence: string
- name: label
dtype: string
- name: question
dtype: string
- name: hash_match
dtype: bool
splits:
- name: train
num_bytes: 136900004
num_examples: 330
download_size: 136386638
dataset_size: 136900004
configs:
- config_name: default
data_files:
- split: train
path: data/train-*