metadata
dataset_info:
features:
- name: image
dtype: image
- name: prompt
dtype: string
- name: answer
dtype: string
- name: x
dtype: int32
- name: 'y'
dtype: int32
- name: frame_index
dtype: int32
splits:
- name: train
num_bytes: 461033
num_examples: 100
- name: validation
num_bytes: 922391
num_examples: 200
download_size: 1378611
dataset_size: 1383424
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*