---
dataset_info:
  features:
  - name: pairID
    dtype: string
  - name: image
    dtype: image
  - name: hypothesis
    dtype: string
  - name: label
    dtype: string
  splits:
  - name: train
    num_bytes: 57828653285.576
    num_examples: 401717
  - name: validation
    num_bytes: 1997015375.232
    num_examples: 14339
  - name: test
    num_bytes: 2021864147.84
    num_examples: 14740
  download_size: 8313812469
  dataset_size: 61847532808.647995
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---