README.md

metadata

dataset_info:
  - config_name: CoSyn_400k_chart
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 25619852113.664
        num_examples: 116814
    download_size: 25239736178
    dataset_size: 25619852113.664
  - config_name: CoSyn_400k_chemical
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 284197936.992
        num_examples: 8942
    download_size: 273097193
    dataset_size: 284197936.992
  - config_name: CoSyn_400k_circuit
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 395788840.72
        num_examples: 10470
    download_size: 381928378
    dataset_size: 395788840.72
  - config_name: CoSyn_400k_diagram
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 7305286810.288
        num_examples: 34963
    download_size: 7234372451
    dataset_size: 7305286810.288
  - config_name: CoSyn_400k_document
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 24180793229.832
        num_examples: 71282
    download_size: 24015209877
    dataset_size: 24180793229.832
  - config_name: CoSyn_400k_graphic
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 335694057.168
        num_examples: 26968
    download_size: 313408282
    dataset_size: 335694057.168
  - config_name: CoSyn_400k_math
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 6107895469.064
        num_examples: 66714
    download_size: 6057279470
    dataset_size: 6107895469.064
  - config_name: CoSyn_400k_music
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 405064954.03
        num_examples: 11969
    download_size: 377684174
    dataset_size: 405064954.03
  - config_name: CoSyn_400k_nutrition
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1508994014.353
        num_examples: 6931
    download_size: 1485416825
    dataset_size: 1508994014.353
  - config_name: CoSyn_400k_table
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 7684052957.968
        num_examples: 46518
    download_size: 7560488723
    dataset_size: 7684052957.968
  - config_name: DoclingMatix
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 960982374796.28
        num_examples: 1270910
    download_size: 950807646187
    dataset_size: 960982374796.28
  - config_name: LLaVA_Instruct_150K
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 76726976312.25
        num_examples: 157710
    download_size: 76639461610
    dataset_size: 76726976312.25
  - config_name: SynthChartNet
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 17908210462
        num_examples: 500000
    download_size: 17714786653
    dataset_size: 17908210462
  - config_name: SynthCodeNet
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 61998944812.125
        num_examples: 499983
    download_size: 61472605472
    dataset_size: 61998944812.125
  - config_name: SynthFormulaNet
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2640399650.375
        num_examples: 499997
    download_size: 2534243196
    dataset_size: 2640399650.375
  - config_name: Unichart
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 18177703609.375
        num_examples: 611925
    download_size: 16923868243
    dataset_size: 18177703609.375
  - config_name: a_okvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 22759813096.74
        num_examples: 54602
    download_size: 22740515076
    dataset_size: 22759813096.74
  - config_name: aguvis-stage-1
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 234712403450.264
        num_examples: 458957
    download_size: 227839724491
    dataset_size: 234712403450.264
  - config_name: ai2d_merged
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 867183847.75
        num_examples: 4858
    download_size: 860582630
    dataset_size: 867183847.75
  - config_name: alfworldgpt
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3890916009.875
        num_examples: 45073
    download_size: 2887255617
    dataset_size: 3890916009.875
  - config_name: allava_laion
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 366924129181.264
        num_examples: 468664
    download_size: 366513300480
    dataset_size: 366924129181.264
  - config_name: allava_vflan
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 92699946037.528
        num_examples: 177078
    download_size: 92524279162
    dataset_size: 92699946037.528
  - config_name: aokvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 896746993.93
        num_examples: 16539
    download_size: 893471601
    dataset_size: 896746993.93
  - config_name: art
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 5141087027.04
        num_examples: 5492
    download_size: 5140689948
    dataset_size: 5141087027.04
  - config_name: arxivqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 81923895225
        num_examples: 100000
    download_size: 81811622698
    dataset_size: 81923895225
  - config_name: bentham
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1450159090.168
        num_examples: 10843
    download_size: 1449119505
    dataset_size: 1450159090.168
  - config_name: blockdiagramcomputerized
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 28792412
        num_examples: 502
    download_size: 28553870
    dataset_size: 28792412
  - config_name: blockdiagramhandwritten
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 146651431.23
        num_examples: 1029
    download_size: 146191172
    dataset_size: 146651431.23
  - config_name: cambrian(filtered)_processed
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 37993953612.448
        num_examples: 83123
    download_size: 37702528512
    dataset_size: 37993953612.448
  - config_name: captcha
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1108385677.25
        num_examples: 113062
    download_size: 1093568723
    dataset_size: 1108385677.25
  - config_name: chart2text
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1129182569.736
        num_examples: 26961
    download_size: 1108115443
    dataset_size: 1129182569.736
  - config_name: chartqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 815177635.55
        num_examples: 18265
    download_size: 803910718
    dataset_size: 815177635.55
  - config_name: chinesememe
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 14244173434.512
        num_examples: 54212
    download_size: 14222203753
    dataset_size: 14244173434.512
  - config_name: chrome_writting
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 80739342.2
        num_examples: 8825
    download_size: 79343529
    dataset_size: 80739342.2
  - config_name: clevr
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 10557164224
        num_examples: 70000
    download_size: 10465001066
    dataset_size: 10557164224
  - config_name: clevr_math
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 9394753178
        num_examples: 70000
    download_size: 9344480504
    dataset_size: 9394753178
  - config_name: clevr_math(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 708071620.6
        num_examples: 5280
    download_size: 697997425
    dataset_size: 708071620.6
  - config_name: coco_colors
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 55374513710.125
        num_examples: 118287
    download_size: 55344845137
    dataset_size: 55374513710.125
  - config_name: cocoqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2402176655.69
        num_examples: 46287
    download_size: 2394615512
    dataset_size: 2402176655.69
  - config_name: cocotext
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 7930321103.875
        num_examples: 16169
    download_size: 7928989554
    dataset_size: 7930321103.875
  - config_name: ctw
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 109319015738.75
        num_examples: 24290
    download_size: 109306604047
    dataset_size: 109319015738.75
  - config_name: datik
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3646550812.875
        num_examples: 220537
    download_size: 3482030545
    dataset_size: 3646550812.875
  - config_name: datikz
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 642401206.67
        num_examples: 47441
    download_size: 591381396
    dataset_size: 642401206.67
  - config_name: densefusion_1m
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 146400262606.897
        num_examples: 1058751
    download_size: 144502318476
    dataset_size: 146400262606.897
  - config_name: diagram_image_to_text
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 18704652
        num_examples: 300
    download_size: 18534456
    dataset_size: 18704652
  - config_name: docvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 12018085125.664
        num_examples: 10189
    download_size: 12007345171
    dataset_size: 12018085125.664
  - config_name: drivelm
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 37226102202.192
        num_examples: 4072
    download_size: 34029716036
    dataset_size: 37226102202.192
  - config_name: dvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4581122677
        num_examples: 200000
    download_size: 4302544626
    dataset_size: 4581122677
  - config_name: est_vqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 18902348521.25
        num_examples: 19358
    download_size: 18901853752
    dataset_size: 18902348521.25
  - config_name: face_emotion
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 15127430
        num_examples: 797
    download_size: 14983116
    dataset_size: 15127430
  - config_name: figureqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2346521984
        num_examples: 100000
    download_size: 2222862886
    dataset_size: 2346521984
  - config_name: figureqa(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 419233306.69
        num_examples: 17587
    download_size: 414996519
    dataset_size: 419233306.69
  - config_name: finqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 138086601.5
        num_examples: 5276
    download_size: 123625992
    dataset_size: 138086601.5
  - config_name: funsd
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 35306471
        num_examples: 194
    download_size: 35104326
    dataset_size: 35306471
  - config_name: geo170k(align)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 197478416.875
        num_examples: 35297
    download_size: 161890724
    dataset_size: 197478416.875
  - config_name: geo170k(qa)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 87625237.87
        num_examples: 12101
    download_size: 52163819
    dataset_size: 87625237.87
  - config_name: geo3k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 38756856.17
        num_examples: 2091
    download_size: 37400382
    dataset_size: 38756856.17
  - config_name: geometry3k(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 197867040.88
        num_examples: 9724
    download_size: 184961625
    dataset_size: 197867040.88
  - config_name: geomverse
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1183897660.128
        num_examples: 9303
    download_size: 1062185395
    dataset_size: 1183897660.128
  - config_name: geoqa+(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 94213384.94
        num_examples: 17162
    download_size: 90953636
    dataset_size: 94213384.94
  - config_name: geos(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3936551
        num_examples: 498
    download_size: 1613784
    dataset_size: 3936551
  - config_name: google_landmarks
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 189460937100.184
        num_examples: 299993
    download_size: 189343235587
    dataset_size: 189460937100.184
  - config_name: groundui
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 6044682942.056
        num_examples: 13531
    download_size: 6027988163
    dataset_size: 6044682942.056
  - config_name: handwriting_forms
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 168001610
        num_examples: 1400
    download_size: 164655119
    dataset_size: 168001610
  - config_name: hateful_memes
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3059106937.5
        num_examples: 8500
    download_size: 3058138125
    dataset_size: 3059106937.5
  - config_name: hitab
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 163934179
        num_examples: 2500
    download_size: 160422628
    dataset_size: 163934179
  - config_name: hme100k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1547322234.04
        num_examples: 74492
    download_size: 1538339958
    dataset_size: 1547322234.04
  - config_name: hw_squad
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 21637654637.632
        num_examples: 20457
    download_size: 21633468499
    dataset_size: 21637654637.632
  - config_name: iam
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1138239910.217
        num_examples: 5663
    download_size: 1134974960
    dataset_size: 1138239910.217
  - config_name: iconqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 330918682.09
        num_examples: 27307
    download_size: 326819099
    dataset_size: 330918682.09
  - config_name: iconqa(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 209363820.43
        num_examples: 22589
    download_size: 204676537
    dataset_size: 209363820.43
  - config_name: idk
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 5197458667.01
        num_examples: 11123
    download_size: 5194521196
    dataset_size: 5197458667.01
  - config_name: iiit5k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 21788858.44
        num_examples: 1990
    download_size: 21513252
    dataset_size: 21788858.44
  - config_name: image_textualization(filtered)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 39882386250.375
        num_examples: 99573
    download_size: 39829746385
    dataset_size: 39882386250.375
  - config_name: imgur5k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 12591193342.434
        num_examples: 5934
    download_size: 12590763433
    dataset_size: 12591193342.434
  - config_name: indoor_qa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 797864863.25
        num_examples: 3350
    download_size: 797431780
    dataset_size: 797864863.25
  - config_name: infographic(gpt4v)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2014741558.032
        num_examples: 1982
    download_size: 2011159744
    dataset_size: 2014741558.032
  - config_name: infographic_vqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4467479648.894
        num_examples: 4394
    download_size: 4465512444
    dataset_size: 4467479648.894
  - config_name: infographic_vqa_llava_format
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1765450951.75
        num_examples: 2113
    download_size: 1764585485
    dataset_size: 1765450951.75
  - config_name: intergps
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 25159455
        num_examples: 1280
    download_size: 24899065
    dataset_size: 25159455
  - config_name: invoices_receipts
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1925658845.375
        num_examples: 3013
    download_size: 1923244863
    dataset_size: 1925658845.375
  - config_name: k12_printing
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4587776492.32
        num_examples: 256636
    download_size: 4546453791
    dataset_size: 4587776492.32
  - config_name: laion_gpt4v
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3021991360.375
        num_examples: 9301
    download_size: 3017230039
    dataset_size: 3021991360.375
  - config_name: latex_handwritten
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 12665387206.408
        num_examples: 39583
    download_size: 12655091924
    dataset_size: 12665387206.408
  - config_name: latexformulas
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 5604066568.5
        num_examples: 552340
    download_size: 5525103231
    dataset_size: 5604066568.5
  - config_name: llavar_gpt4_20k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4235159184.04
        num_examples: 19790
    download_size: 4229077598
    dataset_size: 4235159184.04
  - config_name: lnqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 266234680687.28
        num_examples: 302780
    download_size: 266088073857
    dataset_size: 266234680687.28
  - config_name: localized_narratives
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 21346019807.448
        num_examples: 199998
    download_size: 21291848742
    dataset_size: 21346019807.448
  - config_name: lrv_chart
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 86444276.008
        num_examples: 1776
    download_size: 85369432
    dataset_size: 86444276.008
  - config_name: lrv_normal(filtered)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2985153010.43
        num_examples: 10489
    download_size: 2967270530
    dataset_size: 2985153010.43
  - config_name: lvis_instruct4v
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 107372123408.125
        num_examples: 222711
    download_size: 107199700503
    dataset_size: 107372123408.125
  - config_name: mapqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3371567797.875
        num_examples: 37417
    download_size: 3308958271
    dataset_size: 3371567797.875
  - config_name: mapqa(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 351524458.75
        num_examples: 5225
    download_size: 345656246
    dataset_size: 351524458.75
  - config_name: maptext
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1504185688
        num_examples: 200
    download_size: 1504165598
    dataset_size: 1504185688
  - config_name: mathwriting-google
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 12299849132
        num_examples: 300000
    download_size: 12219456415
    dataset_size: 12299849132
  - config_name: mavis_math_metagen
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3975734405.048
        num_examples: 87348
    download_size: 3266775104
    dataset_size: 3975734405.048
  - config_name: mavis_math_rule_geo
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 20031463665.136
        num_examples: 99986
    download_size: 19769419782
    dataset_size: 20031463665.136
  - config_name: memotion
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2530734349.206
        num_examples: 6991
    download_size: 2528737208
    dataset_size: 2530734349.206
  - config_name: mimic_cgd
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 13184046225.25
        num_examples: 70939
    download_size: 13149862823
    dataset_size: 13184046225.25
  - config_name: mmc_instruct
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 16504670029.128
        num_examples: 168178
    download_size: 16230725185
    dataset_size: 16504670029.128
  - config_name: mmevol
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 25742246427.065
        num_examples: 160215
    download_size: 25480716864
    dataset_size: 25742246427.065
  - config_name: mmra
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1289479228.248
        num_examples: 1024
    download_size: 1249496994
    dataset_size: 1289479228.248
  - config_name: mmsoc_memotion
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2531497426.206
        num_examples: 6991
    download_size: 2529088456
    dataset_size: 2531497426.206
  - config_name: multihiertt
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1378944031.237
        num_examples: 7619
    download_size: 1362595573
    dataset_size: 1378944031.237
  - config_name: nlvr2
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 23552929006.152
        num_examples: 50426
    download_size: 23481437598
    dataset_size: 23552929006.152
  - config_name: objects365_qa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 202609182505.89
        num_examples: 1665847
    download_size: 199554203410
    dataset_size: 202609182505.89
  - config_name: ocrvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 6148678275.896
        num_examples: 165746
    download_size: 6057032047
    dataset_size: 6148678275.896
  - config_name: olmOCR-mix-0225-books
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 6633513593.78
        num_examples: 15194
    download_size: 6618802397
    dataset_size: 6633513593.78
  - config_name: olmOCR-mix-0225-documents
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 97945411922.46
        num_examples: 228858
    download_size: 97308921712
    dataset_size: 97945411922.46
  - config_name: oodvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4694217369.688
        num_examples: 8488
    download_size: 4653237790
    dataset_size: 4694217369.688
  - config_name: orand_car_a
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 23695905.13
        num_examples: 1999
    download_size: 23351148
    dataset_size: 23695905.13
  - config_name: pathvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 18649772240
        num_examples: 32632
    download_size: 18155570098
    dataset_size: 18649772240
  - config_name: pdfvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1663168578.91
        num_examples: 8593
    download_size: 1645451234
    dataset_size: 1663168578.91
  - config_name: plotqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 8939469643.25
        num_examples: 157070
    download_size: 5345605223
    dataset_size: 8939469643.25
  - config_name: pmc_vqa(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3445692373.648
        num_examples: 35948
    download_size: 3437305247
    dataset_size: 3445692373.648
  - config_name: raven
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1734017137
        num_examples: 42000
    download_size: 1721095694
    dataset_size: 1734017137
  - config_name: rendered_text
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 11087697572
        num_examples: 10000
    download_size: 11087197372
    dataset_size: 11087697572
  - config_name: robut_sqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 691188383.864
        num_examples: 8514
    download_size: 684223334
    dataset_size: 691188383.864
  - config_name: robut_wikisql
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 6319659576.464
        num_examples: 74989
    download_size: 6292705239
    dataset_size: 6319659576.464
  - config_name: robut_wtq
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4150727243.896
        num_examples: 38246
    download_size: 4125713020
    dataset_size: 4150727243.896
  - config_name: scienceqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 287033977.12
        num_examples: 4976
    download_size: 283309644
    dataset_size: 287033977.12
  - config_name: scienceqa(nona_context)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2014832143.96
        num_examples: 19208
    download_size: 1968554064
    dataset_size: 2014832143.96
  - config_name: screen2words
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1693100006.1
        num_examples: 15730
    download_size: 1345772929
    dataset_size: 1693100006.1
  - config_name: screenqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 44877746311.875
        num_examples: 80761
    download_size: 44817901938
    dataset_size: 44877746311.875
  - config_name: sharegpt4o
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 39874535436.384
        num_examples: 57284
    download_size: 39791929512
    dataset_size: 39874535436.384
  - config_name: sharegpt4v(coco)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 20028717347.875
        num_examples: 50017
    download_size: 20005211134
    dataset_size: 20028717347.875
  - config_name: sharegpt4v(knowledge)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2405546343.928
        num_examples: 1988
    download_size: 2404816763
    dataset_size: 2405546343.928
  - config_name: sharegpt4v(llava)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 5644424724.75
        num_examples: 29986
    download_size: 5627968195
    dataset_size: 5644424724.75
  - config_name: sharegpt4v(sam)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 31591491436.24
        num_examples: 8990
    download_size: 31588799545
    dataset_size: 31591491436.24
  - config_name: sketchyvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 462161568
        num_examples: 8000
    download_size: 454872096
    dataset_size: 462161568
  - config_name: slidevqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4252726221.347
        num_examples: 1919
    download_size: 2508659044
    dataset_size: 4252726221.347
  - config_name: spark
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1062710726.48
        num_examples: 3904
    download_size: 1061887418
    dataset_size: 1062710726.48
  - config_name: spatialsense
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3539733377.8
        num_examples: 10440
    download_size: 3537019555
    dataset_size: 3539733377.8
  - config_name: spot_the_diff
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1656404738.5
        num_examples: 8566
    download_size: 1590994273
    dataset_size: 1656404738.5
  - config_name: sroie
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 382776987.92
        num_examples: 33616
    download_size: 377976339
    dataset_size: 382776987.92
  - config_name: st_vqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 878506863.672
        num_examples: 17247
    download_size: 876025784
    dataset_size: 878506863.672
  - config_name: sujet_finance
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4870775458.875
        num_examples: 9801
    download_size: 4859136094
    dataset_size: 4870775458.875
  - config_name: super_clevr(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2494651975.75
        num_examples: 8642
    download_size: 2481704024
    dataset_size: 2494651975.75
  - config_name: svrd
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4770520445.396
        num_examples: 4396
    download_size: 4294464627
    dataset_size: 4770520445.396
  - config_name: synthdog
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 312184613844
        num_examples: 500000
    download_size: 312088721977
    dataset_size: 312184613844
  - config_name: tabmwp
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 311512614.14
        num_examples: 22722
    download_size: 306381970
    dataset_size: 311512614.14
  - config_name: tabmwp(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 308654364.24
        num_examples: 22452
    download_size: 304900175
    dataset_size: 308654364.24
  - config_name: tal_ocr_eng
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4597914661.25
        num_examples: 256646
    download_size: 4570495461
    dataset_size: 4597914661.25
  - config_name: tallyqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4683118092
        num_examples: 98680
    download_size: 4663708984
    dataset_size: 4683118092
  - config_name: tat_dqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 56046237.351
        num_examples: 2207
    download_size: 51835448
    dataset_size: 56046237.351
  - config_name: tat_qa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 74054155.13
        num_examples: 2199
    download_size: 70887428
    dataset_size: 74054155.13
  - config_name: text_OpenMathInstruct-2
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1397157730
        num_examples: 1000000
    download_size: 716116310
    dataset_size: 1397157730
  - config_name: text_code_feedback
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 395467554
        num_examples: 66383
    download_size: 177807405
    dataset_size: 395467554
  - config_name: text_codefeedback_filtered_instruction
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 366717666
        num_examples: 156525
    download_size: 189995909
    dataset_size: 366717666
  - config_name: text_infinitymath
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 36914765
        num_examples: 101380
    download_size: 19626819
    dataset_size: 36914765
  - config_name: text_mathinstruct
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 199703065
        num_examples: 262039
    download_size: 111115007
    dataset_size: 199703065
  - config_name: text_mathqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 309234514
        num_examples: 394996
    download_size: 161782592
    dataset_size: 309234514
  - config_name: text_mathstepdpo10k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 5594152
        num_examples: 10795
    download_size: 3115607
    dataset_size: 5594152
  - config_name: text_numinamath_cot
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1292196153
        num_examples: 859494
    download_size: 708138975
    dataset_size: 1292196153
  - config_name: text_openhermes_2_5
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1573997409
        num_examples: 1001551
    download_size: 893716086
    dataset_size: 1573997409
  - config_name: text_openorca
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 6918913888
        num_examples: 4233853
    download_size: 4333157269
    dataset_size: 6918913888
  - config_name: text_orcamath
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 238325136
        num_examples: 200035
    download_size: 96194490
    dataset_size: 238325136
  - config_name: text_pythoncode25k
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 26382364
        num_examples: 49626
    download_size: 14144270
    dataset_size: 26382364
  - config_name: text_pythoncodealpaca
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 12659736
        num_examples: 18612
    download_size: 6520090
    dataset_size: 12659736
  - config_name: text_ruozhiba
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 624832
        num_examples: 1496
    download_size: 425925
    dataset_size: 624832
  - config_name: text_theoremqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 221207
        num_examples: 800
    download_size: 137531
    dataset_size: 221207
  - config_name: text_wizardlm_evol
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 135800299
        num_examples: 69999
    download_size: 73655559
    dataset_size: 135800299
  - config_name: textcaps
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 18032419585.056
        num_examples: 21906
    download_size: 18030277823
    dataset_size: 18032419585.056
  - config_name: textocr(gpt4v)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 20681382782.56
        num_examples: 25060
    download_size: 20676591460
    dataset_size: 20681382782.56
  - config_name: textvqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2145630068.41
        num_examples: 21943
    download_size: 2140127123
    dataset_size: 2145630068.41
  - config_name: tqa
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 659074476.63
        num_examples: 2749
    download_size: 656833279
    dataset_size: 659074476.63
  - config_name: unigeo(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 68744643.63
        num_examples: 11949
    download_size: 66357327
    dataset_size: 68744643.63
  - config_name: ureader_cap
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 75068370896.84
        num_examples: 91215
    download_size: 75059667838
    dataset_size: 75068370896.84
  - config_name: ureader_ie
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 9192939954.32
        num_examples: 17320
    download_size: 9130561778
    dataset_size: 9192939954.32
  - config_name: ureader_kg_processed
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 19867189334.8
        num_examples: 37550
    download_size: 19845137580
    dataset_size: 19867189334.8
  - config_name: ureader_qa_processed
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 150865741789.528
        num_examples: 252953
    download_size: 149919280977
    dataset_size: 150865741789.528
  - config_name: vision_flan(filtered)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 82861363780.064
        num_examples: 175964
    download_size: 82808510061
    dataset_size: 82861363780.064
  - config_name: vistext
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 551951787.944
        num_examples: 9969
    download_size: 544522262
    dataset_size: 551951787.944
  - config_name: visual7w
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4452152281.25
        num_examples: 14366
    download_size: 4443416633
    dataset_size: 4452152281.25
  - config_name: visualmrc
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 1765924752.032
        num_examples: 3027
    download_size: 1762499163
    dataset_size: 1765924752.032
  - config_name: visualwebinstruct(filtered)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 36324448983.375
        num_examples: 263581
    download_size: 36238426746
    dataset_size: 36324448983.375
  - config_name: vizwiz(mathv360k)
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 10145613386.2
        num_examples: 6604
    download_size: 9953393596
    dataset_size: 10145613386.2
  - config_name: vqaonbd
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 7937425917.75
        num_examples: 39986
    download_size: 7749504823
    dataset_size: 7937425917.75
  - config_name: vqarad
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 17413769
        num_examples: 313
    download_size: 16998366
    dataset_size: 17413769
  - config_name: vqav2
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 4295913716.5
        num_examples: 82772
    download_size: 4262624701
    dataset_size: 4295913716.5
  - config_name: vsr
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 108075930.59
        num_examples: 2157
    download_size: 107612332
    dataset_size: 108075930.59
  - config_name: websight
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 8465472626
        num_examples: 10000
    download_size: 8422476950
    dataset_size: 8465472626
  - config_name: wildvision
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 365457306
        num_examples: 333
    download_size: 253665007
    dataset_size: 365457306
  - config_name: wordart
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 2837619342.804
        num_examples: 4804
    download_size: 2837305200
    dataset_size: 2837619342.804
  - config_name: yesbut
    features:
      - name: images
        list: image
      - name: texts
        list:
          - name: user
            dtype: string
          - name: assistant
            dtype: string
      - name: source
        dtype: string
      - name: image_correspondence_ratings
        list: int64
      - name: image_correspondence_min
        dtype: int64
      - name: formatting_ratings
        list: int64
      - name: formatting_min
        dtype: int64
      - name: visual_dependency_ratings
        list: int64
      - name: visual_dependency_min
        dtype: int64
      - name: relevance_ratings
        list: int64
      - name: relevance_min
        dtype: int64
    splits:
      - name: train
        num_bytes: 3172970131.082
        num_examples: 4318
    download_size: 3171611057
    dataset_size: 3172970131.082
configs:
  - config_name: CoSyn_400k_chart
    data_files:
      - split: train
        path: CoSyn_400k_chart/train-*
  - config_name: CoSyn_400k_chemical
    data_files:
      - split: train
        path: CoSyn_400k_chemical/train-*
  - config_name: CoSyn_400k_circuit
    data_files:
      - split: train
        path: CoSyn_400k_circuit/train-*
  - config_name: CoSyn_400k_diagram
    data_files:
      - split: train
        path: CoSyn_400k_diagram/train-*
  - config_name: CoSyn_400k_document
    data_files:
      - split: train
        path: CoSyn_400k_document/train-*
  - config_name: CoSyn_400k_graphic
    data_files:
      - split: train
        path: CoSyn_400k_graphic/train-*
  - config_name: CoSyn_400k_math
    data_files:
      - split: train
        path: CoSyn_400k_math/train-*
  - config_name: CoSyn_400k_music
    data_files:
      - split: train
        path: CoSyn_400k_music/train-*
  - config_name: CoSyn_400k_nutrition
    data_files:
      - split: train
        path: CoSyn_400k_nutrition/train-*
  - config_name: CoSyn_400k_table
    data_files:
      - split: train
        path: CoSyn_400k_table/train-*
  - config_name: DoclingMatix
    data_files:
      - split: train
        path: DoclingMatix/train-*
  - config_name: LLaVA_Instruct_150K
    data_files:
      - split: train
        path: LLaVA_Instruct_150K/train-*
  - config_name: SynthChartNet
    data_files:
      - split: train
        path: SynthChartNet/train-*
  - config_name: SynthCodeNet
    data_files:
      - split: train
        path: SynthCodeNet/train-*
  - config_name: SynthFormulaNet
    data_files:
      - split: train
        path: SynthFormulaNet/train-*
  - config_name: Unichart
    data_files:
      - split: train
        path: Unichart/train-*
  - config_name: a_okvqa
    data_files:
      - split: train
        path: a_okvqa/train-*
  - config_name: aguvis-stage-1
    data_files:
      - split: train
        path: aguvis-stage-1/train-*
  - config_name: ai2d_merged
    data_files:
      - split: train
        path: ai2d_merged/train-*
  - config_name: alfworldgpt
    data_files:
      - split: train
        path: alfworldgpt/train-*
  - config_name: allava_laion
    data_files:
      - split: train
        path: allava_laion/train-*
  - config_name: allava_vflan
    data_files:
      - split: train
        path: allava_vflan/train-*
  - config_name: aokvqa
    data_files:
      - split: train
        path: aokvqa/train-*
  - config_name: art
    data_files:
      - split: train
        path: art/train-*
  - config_name: arxivqa
    data_files:
      - split: train
        path: arxivqa/train-*
  - config_name: bentham
    data_files:
      - split: train
        path: bentham/train-*
  - config_name: blockdiagramcomputerized
    data_files:
      - split: train
        path: blockdiagramcomputerized/train-*
  - config_name: blockdiagramhandwritten
    data_files:
      - split: train
        path: blockdiagramhandwritten/train-*
  - config_name: cambrian(filtered)_processed
    data_files:
      - split: train
        path: cambrian(filtered)_processed/train-*
  - config_name: captcha
    data_files:
      - split: train
        path: captcha/train-*
  - config_name: chart2text
    data_files:
      - split: train
        path: chart2text/train-*
  - config_name: chartqa
    data_files:
      - split: train
        path: chartqa/train-*
  - config_name: chinesememe
    data_files:
      - split: train
        path: chinesememe/train-*
  - config_name: chrome_writting
    data_files:
      - split: train
        path: chrome_writting/train-*
  - config_name: clevr
    data_files:
      - split: train
        path: clevr/train-*
  - config_name: clevr_math
    data_files:
      - split: train
        path: clevr_math/train-*
  - config_name: clevr_math(mathv360k)
    data_files:
      - split: train
        path: clevr_math(mathv360k)/train-*
  - config_name: coco_colors
    data_files:
      - split: train
        path: coco_colors/train-*
  - config_name: cocoqa
    data_files:
      - split: train
        path: cocoqa/train-*
  - config_name: cocotext
    data_files:
      - split: train
        path: cocotext/train-*
  - config_name: ctw
    data_files:
      - split: train
        path: ctw/train-*
  - config_name: datik
    data_files:
      - split: train
        path: datik/train-*
  - config_name: datikz
    data_files:
      - split: train
        path: datikz/train-*
  - config_name: densefusion_1m
    data_files:
      - split: train
        path: densefusion_1m/train-*
  - config_name: diagram_image_to_text
    data_files:
      - split: train
        path: diagram_image_to_text/train-*
  - config_name: docvqa
    data_files:
      - split: train
        path: docvqa/train-*
  - config_name: drivelm
    data_files:
      - split: train
        path: drivelm/train-*
  - config_name: dvqa
    data_files:
      - split: train
        path: dvqa/train-*
  - config_name: est_vqa
    data_files:
      - split: train
        path: est_vqa/train-*
  - config_name: face_emotion
    data_files:
      - split: train
        path: face_emotion/train-*
  - config_name: figureqa
    data_files:
      - split: train
        path: figureqa/train-*
  - config_name: figureqa(mathv360k)
    data_files:
      - split: train
        path: figureqa(mathv360k)/train-*
  - config_name: finqa
    data_files:
      - split: train
        path: finqa/train-*
  - config_name: funsd
    data_files:
      - split: train
        path: funsd/train-*
  - config_name: geo170k(align)
    data_files:
      - split: train
        path: geo170k(align)/train-*
  - config_name: geo170k(qa)
    data_files:
      - split: train
        path: geo170k(qa)/train-*
  - config_name: geo3k
    data_files:
      - split: train
        path: geo3k/train-*
  - config_name: geometry3k(mathv360k)
    data_files:
      - split: train
        path: geometry3k(mathv360k)/train-*
  - config_name: geomverse
    data_files:
      - split: train
        path: geomverse/train-*
  - config_name: geoqa+(mathv360k)
    data_files:
      - split: train
        path: geoqa+(mathv360k)/train-*
  - config_name: geos(mathv360k)
    data_files:
      - split: train
        path: geos(mathv360k)/train-*
  - config_name: google_landmarks
    data_files:
      - split: train
        path: google_landmarks/train-*
  - config_name: groundui
    data_files:
      - split: train
        path: groundui/train-*
  - config_name: handwriting_forms
    data_files:
      - split: train
        path: handwriting_forms/train-*
  - config_name: hateful_memes
    data_files:
      - split: train
        path: hateful_memes/train-*
  - config_name: hitab
    data_files:
      - split: train
        path: hitab/train-*
  - config_name: hme100k
    data_files:
      - split: train
        path: hme100k/train-*
  - config_name: hw_squad
    data_files:
      - split: train
        path: hw_squad/train-*
  - config_name: iam
    data_files:
      - split: train
        path: iam/train-*
  - config_name: iconqa
    data_files:
      - split: train
        path: iconqa/train-*
  - config_name: iconqa(mathv360k)
    data_files:
      - split: train
        path: iconqa(mathv360k)/train-*
  - config_name: idk
    data_files:
      - split: train
        path: idk/train-*
  - config_name: iiit5k
    data_files:
      - split: train
        path: iiit5k/train-*
  - config_name: image_textualization(filtered)
    data_files:
      - split: train
        path: image_textualization(filtered)/train-*
  - config_name: imgur5k
    data_files:
      - split: train
        path: imgur5k/train-*
  - config_name: indoor_qa
    data_files:
      - split: train
        path: indoor_qa/train-*
  - config_name: infographic(gpt4v)
    data_files:
      - split: train
        path: infographic(gpt4v)/train-*
  - config_name: infographic_vqa
    data_files:
      - split: train
        path: infographic_vqa/train-*
  - config_name: infographic_vqa_llava_format
    data_files:
      - split: train
        path: infographic_vqa_llava_format/train-*
  - config_name: intergps
    data_files:
      - split: train
        path: intergps/train-*
  - config_name: invoices_receipts
    data_files:
      - split: train
        path: invoices_receipts/train-*
  - config_name: k12_printing
    data_files:
      - split: train
        path: k12_printing/train-*
  - config_name: laion_gpt4v
    data_files:
      - split: train
        path: laion_gpt4v/train-*
  - config_name: latex_handwritten
    data_files:
      - split: train
        path: latex_handwritten/train-*
  - config_name: latexformulas
    data_files:
      - split: train
        path: latexformulas/train-*
  - config_name: llavar_gpt4_20k
    data_files:
      - split: train
        path: llavar_gpt4_20k/train-*
  - config_name: lnqa
    data_files:
      - split: train
        path: lnqa/train-*
  - config_name: localized_narratives
    data_files:
      - split: train
        path: localized_narratives/train-*
  - config_name: lrv_chart
    data_files:
      - split: train
        path: lrv_chart/train-*
  - config_name: lrv_normal(filtered)
    data_files:
      - split: train
        path: lrv_normal(filtered)/train-*
  - config_name: lvis_instruct4v
    data_files:
      - split: train
        path: lvis_instruct4v/train-*
  - config_name: mapqa
    data_files:
      - split: train
        path: mapqa/train-*
  - config_name: mapqa(mathv360k)
    data_files:
      - split: train
        path: mapqa(mathv360k)/train-*
  - config_name: maptext
    data_files:
      - split: train
        path: maptext/train-*
  - config_name: mathwriting-google
    data_files:
      - split: train
        path: mathwriting-google/train-*
  - config_name: mavis_math_metagen
    data_files:
      - split: train
        path: mavis_math_metagen/train-*
  - config_name: mavis_math_rule_geo
    data_files:
      - split: train
        path: mavis_math_rule_geo/train-*
  - config_name: memotion
    data_files:
      - split: train
        path: memotion/train-*
  - config_name: mimic_cgd
    data_files:
      - split: train
        path: mimic_cgd/train-*
  - config_name: mmc_instruct
    data_files:
      - split: train
        path: mmc_instruct/train-*
  - config_name: mmevol
    data_files:
      - split: train
        path: mmevol/train-*
  - config_name: mmra
    data_files:
      - split: train
        path: mmra/train-*
  - config_name: mmsoc_memotion
    data_files:
      - split: train
        path: mmsoc_memotion/train-*
  - config_name: multihiertt
    data_files:
      - split: train
        path: multihiertt/train-*
  - config_name: nlvr2
    data_files:
      - split: train
        path: nlvr2/train-*
  - config_name: objects365_qa
    data_files:
      - split: train
        path: objects365_qa/train-*
  - config_name: ocrvqa
    data_files:
      - split: train
        path: ocrvqa/train-*
  - config_name: olmOCR-mix-0225-books
    data_files:
      - split: train
        path: olmOCR-mix-0225-books/train-*
  - config_name: olmOCR-mix-0225-documents
    data_files:
      - split: train
        path: olmOCR-mix-0225-documents/train-*
  - config_name: oodvqa
    data_files:
      - split: train
        path: oodvqa/train-*
  - config_name: orand_car_a
    data_files:
      - split: train
        path: orand_car_a/train-*
  - config_name: pathvqa
    data_files:
      - split: train
        path: pathvqa/train-*
  - config_name: pdfvqa
    data_files:
      - split: train
        path: pdfvqa/train-*
  - config_name: plotqa
    data_files:
      - split: train
        path: plotqa/train-*
  - config_name: pmc_vqa(mathv360k)
    data_files:
      - split: train
        path: pmc_vqa(mathv360k)/train-*
  - config_name: raven
    data_files:
      - split: train
        path: raven/train-*
  - config_name: rendered_text
    data_files:
      - split: train
        path: rendered_text/train-*
  - config_name: robut_sqa
    data_files:
      - split: train
        path: robut_sqa/train-*
  - config_name: robut_wikisql
    data_files:
      - split: train
        path: robut_wikisql/train-*
  - config_name: robut_wtq
    data_files:
      - split: train
        path: robut_wtq/train-*
  - config_name: scienceqa
    data_files:
      - split: train
        path: scienceqa/train-*
  - config_name: scienceqa(nona_context)
    data_files:
      - split: train
        path: scienceqa(nona_context)/train-*
  - config_name: screen2words
    data_files:
      - split: train
        path: screen2words/train-*
  - config_name: screenqa
    data_files:
      - split: train
        path: screenqa/train-*
  - config_name: sharegpt4o
    data_files:
      - split: train
        path: sharegpt4o/train-*
  - config_name: sharegpt4v(coco)
    data_files:
      - split: train
        path: sharegpt4v(coco)/train-*
  - config_name: sharegpt4v(knowledge)
    data_files:
      - split: train
        path: sharegpt4v(knowledge)/train-*
  - config_name: sharegpt4v(llava)
    data_files:
      - split: train
        path: sharegpt4v(llava)/train-*
  - config_name: sharegpt4v(sam)
    data_files:
      - split: train
        path: sharegpt4v(sam)/train-*
  - config_name: sketchyvqa
    data_files:
      - split: train
        path: sketchyvqa/train-*
  - config_name: slidevqa
    data_files:
      - split: train
        path: slidevqa/train-*
  - config_name: spark
    data_files:
      - split: train
        path: spark/train-*
  - config_name: spatialsense
    data_files:
      - split: train
        path: spatialsense/train-*
  - config_name: spot_the_diff
    data_files:
      - split: train
        path: spot_the_diff/train-*
  - config_name: sroie
    data_files:
      - split: train
        path: sroie/train-*
  - config_name: st_vqa
    data_files:
      - split: train
        path: st_vqa/train-*
  - config_name: sujet_finance
    data_files:
      - split: train
        path: sujet_finance/train-*
  - config_name: super_clevr(mathv360k)
    data_files:
      - split: train
        path: super_clevr(mathv360k)/train-*
  - config_name: svrd
    data_files:
      - split: train
        path: svrd/train-*
  - config_name: synthdog
    data_files:
      - split: train
        path: synthdog/train-*
  - config_name: tabmwp
    data_files:
      - split: train
        path: tabmwp/train-*
  - config_name: tabmwp(mathv360k)
    data_files:
      - split: train
        path: tabmwp(mathv360k)/train-*
  - config_name: tal_ocr_eng
    data_files:
      - split: train
        path: tal_ocr_eng/train-*
  - config_name: tallyqa
    data_files:
      - split: train
        path: tallyqa/train-*
  - config_name: tat_dqa
    data_files:
      - split: train
        path: tat_dqa/train-*
  - config_name: tat_qa
    data_files:
      - split: train
        path: tat_qa/train-*
  - config_name: text_OpenMathInstruct-2
    data_files:
      - split: train
        path: text_OpenMathInstruct-2/train-*
  - config_name: text_code_feedback
    data_files:
      - split: train
        path: text_code_feedback/train-*
  - config_name: text_codefeedback_filtered_instruction
    data_files:
      - split: train
        path: text_codefeedback_filtered_instruction/train-*
  - config_name: text_infinitymath
    data_files:
      - split: train
        path: text_infinitymath/train-*
  - config_name: text_mathinstruct
    data_files:
      - split: train
        path: text_mathinstruct/train-*
  - config_name: text_mathqa
    data_files:
      - split: train
        path: text_mathqa/train-*
  - config_name: text_mathstepdpo10k
    data_files:
      - split: train
        path: text_mathstepdpo10k/train-*
  - config_name: text_numinamath_cot
    data_files:
      - split: train
        path: text_numinamath_cot/train-*
  - config_name: text_openhermes_2_5
    data_files:
      - split: train
        path: text_openhermes_2_5/train-*
  - config_name: text_openorca
    data_files:
      - split: train
        path: text_openorca/train-*
  - config_name: text_orcamath
    data_files:
      - split: train
        path: text_orcamath/train-*
  - config_name: text_pythoncode25k
    data_files:
      - split: train
        path: text_pythoncode25k/train-*
  - config_name: text_pythoncodealpaca
    data_files:
      - split: train
        path: text_pythoncodealpaca/train-*
  - config_name: text_ruozhiba
    data_files:
      - split: train
        path: text_ruozhiba/train-*
  - config_name: text_theoremqa
    data_files:
      - split: train
        path: text_theoremqa/train-*
  - config_name: text_wizardlm_evol
    data_files:
      - split: train
        path: text_wizardlm_evol/train-*
  - config_name: textcaps
    data_files:
      - split: train
        path: textcaps/train-*
  - config_name: textocr(gpt4v)
    data_files:
      - split: train
        path: textocr(gpt4v)/train-*
  - config_name: textvqa
    data_files:
      - split: train
        path: textvqa/train-*
  - config_name: tqa
    data_files:
      - split: train
        path: tqa/train-*
  - config_name: unigeo(mathv360k)
    data_files:
      - split: train
        path: unigeo(mathv360k)/train-*
  - config_name: ureader_cap
    data_files:
      - split: train
        path: ureader_cap/train-*
  - config_name: ureader_ie
    data_files:
      - split: train
        path: ureader_ie/train-*
  - config_name: ureader_kg_processed
    data_files:
      - split: train
        path: ureader_kg_processed/train-*
  - config_name: ureader_qa_processed
    data_files:
      - split: train
        path: ureader_qa_processed/train-*
  - config_name: vision_flan(filtered)
    data_files:
      - split: train
        path: vision_flan(filtered)/train-*
  - config_name: vistext
    data_files:
      - split: train
        path: vistext/train-*
  - config_name: visual7w
    data_files:
      - split: train
        path: visual7w/train-*
  - config_name: visualmrc
    data_files:
      - split: train
        path: visualmrc/train-*
  - config_name: visualwebinstruct(filtered)
    data_files:
      - split: train
        path: visualwebinstruct(filtered)/train-*
  - config_name: vizwiz(mathv360k)
    data_files:
      - split: train
        path: vizwiz(mathv360k)/train-*
  - config_name: vqaonbd
    data_files:
      - split: train
        path: vqaonbd/train-*
  - config_name: vqarad
    data_files:
      - split: train
        path: vqarad/train-*
  - config_name: vqav2
    data_files:
      - split: train
        path: vqav2/train-*
  - config_name: vsr
    data_files:
      - split: train
        path: vsr/train-*
  - config_name: websight
    data_files:
      - split: train
        path: websight/train-*
  - config_name: wildvision
    data_files:
      - split: train
        path: wildvision/train-*
  - config_name: wordart
    data_files:
      - split: train
        path: wordart/train-*
  - config_name: yesbut
    data_files:
      - split: train
        path: yesbut/train-*
size_categories:
  - 10M<n<100M

Fine Vision

FineVision is a massive collection of datasets with 17.3M images, 24.3M samples, 88.9M turns, and 9.5B answer tokens, designed for training state-of-the-art open Vision-Language-Models.

More detail can be found in the blog post: https://huggingface.co/spaces/HuggingFaceM4/FineVision

Load the data

  from datasets import load_dataset, get_dataset_config_names

  # Get all subset names and load the first one
  available_subsets = get_dataset_config_names('HuggingFaceM4/FineVision')
  ds = load_dataset('HuggingFaceM4/FineVision', name=available_subsets[0], split='train', streaming=True)
  
  # Inspect the first sample
  ds[0]

Structure

{
  'images': [<PIL.PngImagePlugin.PngImageFile image mode=RGB size=387x194 at 0x7F8F0B308200>],
  'texts': [{'user': 'Question: What is between the reticulum and the abomasum?\nChoices:\nA. Intestine\nB. Omasum\nC. Stomach\nD. Rumen\nAnswer with the letter.', 'assistant': 'Answer: B'},
            {'user': 'Here is a diagram figure extracted from some Grade 1 - 6 science books.\nPlease first describe the content of this figure in detail, including how the knowledge visually displayed in the diagram.\nThen start with a section title "related knowledge:", briefly and concisely highlight the related domain knowledge and theories that underly this diagram. Note that you do not need to provide much detail. Simply cover the most important concepts.', 'assistant': "The figure is a simple diagram of the four compartments of a ruminant animal's stomach, which are the rumen, reticulum, omasum, and abomasum. The diagram shows the relative size and position of each compartment within the stomach, with arrows indicating the direction of food flow from one compartment to the next. \n\nRelated Knowledge:\n- Ruminant Digestion: Ruminants such as cows, sheep, and goats have a unique digestive system that allows them to break down fibrous plant material, like grass, that other animals cannot digest.\n- Four Stomach Compartments: The rumen is the largest compartment and serves as a fermentation vat where microbes break down fibrous material. The reticulum traps foreign objects and also helps in fermentation. The omasum absorbs water and nutrients, and the abomasum is the true stomach where digestion occurs similarly to monogastric animals.\n- Microbial Fermentation: The microbes in the rumen produce volatile fatty acids which are the primary energy source for ruminants. They also produce gases like methane, which are eructated (belched) out.\n- Ruminant Nutrition: Ruminants rely on a high-fiber diet and have to consume large quantities of forage to meet their nutritional needs."}],
  'source': 'original',
  'image_correspondence_ratings': [4, 3],
  'image_correspondence_min': 3,
  'visual_dependency_ratings': [4, 5],
  'visual_dependency_min': 4,
  'formatting_ratings': [4, 4],
  'formatting_min': 4,
  'relevance_ratings': [5, 5],
  'relevance_min': 5
}

Categories

Licensing Information

Each of the publicly available sub-datasets present in FineVision are governed by specific licensing conditions. Therefore, when making use of them you must take into consideration each of the licenses governing each dataset. To the extent we have any rights in the prompts, these are licensed under CC-BY-4.0.

Citation

If you find this dataset useful, please cite:

@misc{wiedmann2025finevisionopendataneed,
      title={FineVision: Open Data Is All You Need}, 
      author={Luis Wiedmann and Orr Zohar and Amir Mahla and Xiaohan Wang and Rui Li and Thibaud Frere and Leandro von Werra and Aritra Roy Gosthipaty and Andrés Marafioti},
      year={2025},
      eprint={2510.17269},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2510.17269}, 
}