README.md

metadata

license: other
license_name: mixed-see-licensing-information
task_categories:
  - visual-question-answering
  - image-text-to-text
language:
  - en
  - zh
tags:
  - multimodal
  - vision-language
  - instruction-tuning
  - chain-of-thought
  - sft
  - ocr
  - chart
  - document
  - grounding
  - gui
size_categories:
  - 10M<n<100M
pretty_name: Honey-Data-V2
dataset_info:
  - config_name: ST-VQA
    features: &ref_0
      - name: id
        dtype: string
      - name: images
        list: image
      - name: n_images
        dtype: int32
      - name: image_sizes
        list:
          list: int64
      - name: image_phashes
        list: string
      - name: category
        dtype: string
      - name: source_subset
        dtype: string
      - name: categories
        list: string
      - name: sources
        list: string
      - name: source_subsets
        list: string
      - name: data
        list:
          - name: id
            dtype: string
          - name: ori_id
            dtype: string
          - name: subset
            dtype: string
          - name: messages
            list:
              - name: from
                dtype: string
              - name: value
                dtype: string
          - name: n_turns
            dtype: int32
          - name: cot_level
            dtype: string
          - name: original_answers
            list: string
          - name: provenance
            struct:
              - name: question_image_consist
                dtype: string
              - name: answer_rewrite
                dtype: string
              - name: reasoning_rewrite
                dtype: string
              - name: answer_consist
                dtype: string
          - name: source
            dtype: string
          - name: source_subset
            dtype: string
          - name: category
            dtype: string
    splits:
      - name: train
        num_bytes: 2988595335
        num_examples: 11605
    download_size: 2967260810
    dataset_size: 2988595335
  - config_name: Sherlock
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3418323607
        num_examples: 4062
    download_size: 3400791046
    dataset_size: 3418323607
  - config_name: WIT
    features: *ref_0
    splits:
      - name: train
        num_bytes: 648331736071
        num_examples: 290499
    download_size: 645799935461
    dataset_size: 648331736071
  - config_name: coyo
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2833147
        num_examples: 42
    download_size: 2648600
    dataset_size: 2833147
  - config_name: densefusion_1m
    features: *ref_0
    splits:
      - name: train
        num_bytes: 466175738828
        num_examples: 1054732
    download_size: 465086950073
    dataset_size: 466175738828
  - config_name: laion_gpt4v_train_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15178934
        num_examples: 65
    download_size: 14922856
    dataset_size: 15178934
  - config_name: Caption-other
    features: *ref_0
    splits:
      - name: train
        num_bytes: 318243622
        num_examples: 879
    download_size: 312477439
    dataset_size: 318243622
  - config_name: pixmo-cap
    features: *ref_0
    splits:
      - name: train
        num_bytes: 10994682
        num_examples: 27
    download_size: 10532165
    dataset_size: 10994682
  - config_name: rlaifv_chosen_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1913905
        num_examples: 30
    download_size: 1863846
    dataset_size: 1913905
  - config_name: sharegpt4o_longcap_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 58639
        num_examples: 1
    download_size: 52654
    dataset_size: 58639
  - config_name: sherlock_317k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5369331694
        num_examples: 6627
    download_size: 5345947049
    dataset_size: 5369331694
  - config_name: Caption-st_vqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 122603900
        num_examples: 878
    download_size: 120526942
    dataset_size: 122603900
  - config_name: wit
    features: *ref_0
    splits:
      - name: train
        num_bytes: 89312784709
        num_examples: 43675
    download_size: 89097809580
    dataset_size: 89312784709
  - config_name: ArxivQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2571361179
        num_examples: 15773
    download_size: 1881120324
    dataset_size: 2571361179
  - config_name: Chart2Text
    features: *ref_0
    splits:
      - name: train
        num_bytes: 980546836
        num_examples: 28235
    download_size: 581405308
    dataset_size: 980546836
  - config_name: ChartQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 477996271
        num_examples: 17299
    download_size: 297474118
    dataset_size: 477996271
  - config_name: ChartVerse-SFT-1.8M
    features: *ref_0
    splits:
      - name: train
        num_bytes: 732362911848
        num_examples: 1642665
    download_size: 725773977292
    dataset_size: 732362911848
  - config_name: Chart-CoSyn
    features: *ref_0
    splits:
      - name: train
        num_bytes: 19507659457
        num_examples: 209229
    download_size: 16991341679
    dataset_size: 19507659457
  - config_name: Chart-CoSyn_400k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1839779062
        num_examples: 10716
    download_size: 1768790846
    dataset_size: 1839779062
  - config_name: DVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8497478379
        num_examples: 318906
    download_size: 3602198010
    dataset_size: 8497478379
  - config_name: FigureQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3428269166
        num_examples: 103837
    download_size: 1439736438
    dataset_size: 3428269166
  - config_name: FigureQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 57653734
        num_examples: 2730
    download_size: 31812371
    dataset_size: 57653734
  - config_name: HiTab
    features: *ref_0
    splits:
      - name: train
        num_bytes: 70532165
        num_examples: 2406
    download_size: 42449621
    dataset_size: 70532165
  - config_name: Infographic
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1299713162
        num_examples: 1308
    download_size: 1265718261
    dataset_size: 1299713162
  - config_name: LRV_Chart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 125867479
        num_examples: 2353
    download_size: 75314045
    dataset_size: 125867479
  - config_name: MAVIS_Function
    features: *ref_0
    splits:
      - name: train
        num_bytes: 342364093
        num_examples: 9840
    download_size: 263985517
    dataset_size: 342364093
  - config_name: MMC_Instruction
    features: *ref_0
    splits:
      - name: train
        num_bytes: 211132559
        num_examples: 10587
    download_size: 134494475
    dataset_size: 211132559
  - config_name: MMTab
    features: *ref_0
    splits:
      - name: train
        num_bytes: 14926048872
        num_examples: 56553
    download_size: 8549148816
    dataset_size: 14926048872
  - config_name: Chart-MapQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 10164433
        num_examples: 333
    download_size: 8745110
    dataset_size: 10164433
  - config_name: MultiHiertt
    features: *ref_0
    splits:
      - name: train
        num_bytes: 13259096
        num_examples: 857
    download_size: 10888966
    dataset_size: 13259096
  - config_name: PlotQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3853360301
        num_examples: 69960
    download_size: 1328807522
    dataset_size: 3853360301
  - config_name: RobuT
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4191125066
        num_examples: 118573
    download_size: 2590185116
    dataset_size: 4191125066
  - config_name: SciTSR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 94380491
        num_examples: 8193
    download_size: 70773417
    dataset_size: 94380491
  - config_name: SimChart9K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 72681884
        num_examples: 2978
    download_size: 62992064
    dataset_size: 72681884
  - config_name: Chart-SynMultiImageQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 40976373523
        num_examples: 120250
    download_size: 39697634878
    dataset_size: 40976373523
  - config_name: SynthChartNet
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6776887848
        num_examples: 499550
    download_size: 6623860168
    dataset_size: 6776887848
  - config_name: TabMWP
    features: *ref_0
    splits:
      - name: train
        num_bytes: 384941507
        num_examples: 41931
    download_size: 227576747
    dataset_size: 384941507
  - config_name: TabMWP_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9376373
        num_examples: 682
    download_size: 5765092
    dataset_size: 9376373
  - config_name: TinyChart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 22468212094
        num_examples: 615483
    download_size: 11177981691
    dataset_size: 22468212094
  - config_name: UReader
    features: *ref_0
    splits:
      - name: train
        num_bytes: 23984919538
        num_examples: 78895
    download_size: 21578853538
    dataset_size: 23984919538
  - config_name: UniChart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7679791335
        num_examples: 309800
    download_size: 4489975566
    dataset_size: 7679791335
  - config_name: Unichart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 727918738
        num_examples: 31888
    download_size: 398529698
    dataset_size: 727918738
  - config_name: VisText
    features: *ref_0
    splits:
      - name: train
        num_bytes: 337220799
        num_examples: 13228
    download_size: 211121969
    dataset_size: 337220799
  - config_name: arxiv_qa_chart_gpt4o
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3766222255
        num_examples: 109209
    download_size: 3163019710
    dataset_size: 3766222255
  - config_name: arxivqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4024034095
        num_examples: 7639
    download_size: 3996804367
    dataset_size: 4024034095
  - config_name: arxivqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7706551561
        num_examples: 39965
    download_size: 5952708492
    dataset_size: 7706551561
  - config_name: arxivqa_total
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1404954173
        num_examples: 8217
    download_size: 1070857427
    dataset_size: 1404954173
  - config_name: blockdiagramhandwritten
    features: *ref_0
    splits:
      - name: train
        num_bytes: 56432547
        num_examples: 1014
    download_size: 54357413
    dataset_size: 56432547
  - config_name: chart2text_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 62015486
        num_examples: 3475
    download_size: 48018218
    dataset_size: 62015486
  - config_name: chartqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12003823
        num_examples: 618
    download_size: 9472480
    dataset_size: 12003823
  - config_name: cosyn_chart_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 157657056
        num_examples: 1429
    download_size: 148611133
    dataset_size: 157657056
  - config_name: cosyn_diagram_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 47249000
        num_examples: 715
    download_size: 44785837
    dataset_size: 47249000
  - config_name: cosyn_graphic_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 33085663
        num_examples: 3016
    download_size: 18951302
    dataset_size: 33085663
  - config_name: cosyn_table_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 16351202
        num_examples: 335
    download_size: 14849425
    dataset_size: 16351202
  - config_name: diagram_image_to_text_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 30458
        num_examples: 2
    download_size: 44551
    dataset_size: 30458
  - config_name: dvqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 118741
        num_examples: 8
    download_size: 86307
    dataset_size: 118741
  - config_name: dvqa_train_200k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 54704092
        num_examples: 2507
    download_size: 28382703
    dataset_size: 54704092
  - config_name: figureqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 73643418
        num_examples: 2634
    download_size: 33100251
    dataset_size: 73643418
  - config_name: figureqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 11692198
        num_examples: 778
    download_size: 5953395
    dataset_size: 11692198
  - config_name: finqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 48609848
        num_examples: 4023
    download_size: 30093937
    dataset_size: 48609848
  - config_name: hitab_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3772926
        num_examples: 156
    download_size: 2713411
    dataset_size: 3772926
  - config_name: Chart-infographic_gpt4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5959775
        num_examples: 10
    download_size: 5872929
    dataset_size: 5959775
  - config_name: Chart-infographic_vqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2328873
        num_examples: 4
    download_size: 2321122
    dataset_size: 2328873
  - config_name: Chart-infographic_vqa_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 14480411
        num_examples: 17
    download_size: 14339226
    dataset_size: 14480411
  - config_name: lrv_chart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4262356
        num_examples: 87
    download_size: 2862417
    dataset_size: 4262356
  - config_name: lrv_chart_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 61178
        num_examples: 4
    download_size: 73961
    dataset_size: 61178
  - config_name: mavis_math_function_caption_to_question_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1202241
        num_examples: 32
    download_size: 1088007
    dataset_size: 1202241
  - config_name: mmc_inst_qa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 45384826
        num_examples: 2917
    download_size: 34755354
    dataset_size: 45384826
  - config_name: mmtab_instruct_pretrain_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 129290045
        num_examples: 1060
    download_size: 83835774
    dataset_size: 129290045
  - config_name: mmtab_instruct_sft_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1188303363
        num_examples: 1703
    download_size: 632537362
    dataset_size: 1188303363
  - config_name: multihiertt
    features: *ref_0
    splits:
      - name: train
        num_bytes: 47484639
        num_examples: 1053
    download_size: 45226762
    dataset_size: 47484639
  - config_name: multihiertt_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 40994801
        num_examples: 3281
    download_size: 32484518
    dataset_size: 40994801
  - config_name: multimodal_arxivqa52k_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 357226882
        num_examples: 2052
    download_size: 270376564
    dataset_size: 357226882
  - config_name: Chart-other
    features: *ref_0
    splits:
      - name: train
        num_bytes: 32469978
        num_examples: 1201
    download_size: 21035774
    dataset_size: 32469978
  - config_name: plotqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 706579204
        num_examples: 20028
    download_size: 247795858
    dataset_size: 706579204
  - config_name: plotqa_v1_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 99384
        num_examples: 6
    download_size: 76815
    dataset_size: 99384
  - config_name: robut
    features: *ref_0
    splits:
      - name: train
        num_bytes: 403742777
        num_examples: 15829
    download_size: 383534788
    dataset_size: 403742777
  - config_name: robut_sqa_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6214535
        num_examples: 231
    download_size: 4443617
    dataset_size: 6214535
  - config_name: robut_wikisql_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 164619242
        num_examples: 6626
    download_size: 127922021
    dataset_size: 164619242
  - config_name: robut_wtq_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 81111066
        num_examples: 2476
    download_size: 65370695
    dataset_size: 81111066
  - config_name: scitsr_markdown_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9990169
        num_examples: 605
    download_size: 7551900
    dataset_size: 9990169
  - config_name: simchart9k_csv_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 23931352
        num_examples: 1065
    download_size: 21600363
    dataset_size: 23931352
  - config_name: tabmwp
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8436635
        num_examples: 1366
    download_size: 7450225
    dataset_size: 8436635
  - config_name: tabmwp_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 190006
        num_examples: 22
    download_size: 175482
    dataset_size: 190006
  - config_name: tat_qa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 21349761
        num_examples: 1560
    download_size: 14501041
    dataset_size: 21349761
  - config_name: tinychart_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 929160398
        num_examples: 42570
    download_size: 640316732
    dataset_size: 929160398
  - config_name: unichart_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 405304590
        num_examples: 19416
    download_size: 320569565
    dataset_size: 405304590
  - config_name: Chart-ureader_kg
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1653527951
        num_examples: 7329
    download_size: 1599519790
    dataset_size: 1653527951
  - config_name: ureader_qa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 278192229
        num_examples: 1370
    download_size: 270393591
    dataset_size: 278192229
  - config_name: vistext_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12913331
        num_examples: 665
    download_size: 10522073
    dataset_size: 12913331
  - config_name: Document-CoSyn
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7685772910
        num_examples: 77715
    download_size: 7038568420
    dataset_size: 7685772910
  - config_name: CoSyn_400k_document
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1651082603
        num_examples: 6225
    download_size: 1623563789
    dataset_size: 1651082603
  - config_name: DocReason
    features: *ref_0
    splits:
      - name: train
        num_bytes: 286370273
        num_examples: 1730
    download_size: 277686048
    dataset_size: 286370273
  - config_name: DocVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 295126072
        num_examples: 443
    download_size: 294671132
    dataset_size: 295126072
  - config_name: DoclingMatix
    features: *ref_0
    splits:
      - name: train
        num_bytes: 343983649075
        num_examples: 1216923
    download_size: 336083027373
    dataset_size: 343983649075
  - config_name: Docmatix
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3184759361
        num_examples: 18765
    download_size: 2978226422
    dataset_size: 3184759361
  - config_name: EATEN
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1946865640
        num_examples: 29991
    download_size: 1885671058
    dataset_size: 1946865640
  - config_name: FinTabNet
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1634710686
        num_examples: 37527
    download_size: 1298844274
    dataset_size: 1634710686
  - config_name: InfoVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 26655077
        num_examples: 63
    download_size: 21753770
    dataset_size: 26655077
  - config_name: InfographicVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1579332440
        num_examples: 1648
    download_size: 1539348666
    dataset_size: 1579332440
  - config_name: LLaVAR_GPT4
    features: *ref_0
    splits:
      - name: train
        num_bytes: 842635067
        num_examples: 20012
    download_size: 723689503
    dataset_size: 842635067
  - config_name: OCR-VQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3278615150
        num_examples: 71258
    download_size: 2699561645
    dataset_size: 3278615150
  - config_name: OCR-VQA-LLaVA-NeXT
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2584978
        num_examples: 74
    download_size: 2291914
    dataset_size: 2584978
  - config_name: POIE
    features: *ref_0
    splits:
      - name: train
        num_bytes: 388208102
        num_examples: 2249
    download_size: 384634911
    dataset_size: 388208102
  - config_name: ScreenQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 21590040996
        num_examples: 60478
    download_size: 21506417926
    dataset_size: 21590040996
  - config_name: SynMultiImageQA_Doc
    features: *ref_0
    splits:
      - name: train
        num_bytes: 91163411939
        num_examples: 57241
    download_size: 90703877561
    dataset_size: 91163411939
  - config_name: TextVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 16981331225
        num_examples: 24883
    download_size: 16942250104
    dataset_size: 16981331225
  - config_name: UreaderOCR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 596108180
        num_examples: 3830
    download_size: 578038426
    dataset_size: 596108180
  - config_name: Ureader_Chart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 94356331896
        num_examples: 202908
    download_size: 92229575892
    dataset_size: 94356331896
  - config_name: VisualMRC
    features: *ref_0
    splits:
      - name: train
        num_bytes: 818632110
        num_examples: 2184
    download_size: 805748398
    dataset_size: 818632110
  - config_name: cosyn_document_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4274020
        num_examples: 58
    download_size: 4139429
    dataset_size: 4274020
  - config_name: cosyn_nutrition_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 153609
        num_examples: 2
    download_size: 160649
    dataset_size: 153609
  - config_name: docmatix_single_20k_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 61979102
        num_examples: 304
    download_size: 60751981
    dataset_size: 61979102
  - config_name: docreason_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 25586066
        num_examples: 19
    download_size: 25558912
    dataset_size: 25586066
  - config_name: docvqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 133535
        num_examples: 2
    download_size: 147428
    dataset_size: 133535
  - config_name: eaten_passport_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 448846
        num_examples: 4
    download_size: 462139
    dataset_size: 448846
  - config_name: fintabnet_markdown_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 454314323
        num_examples: 11217
    download_size: 408153460
    dataset_size: 454314323
  - config_name: handwriting_forms
    features: *ref_0
    splits:
      - name: train
        num_bytes: 65376540
        num_examples: 1302
    download_size: 64554826
    dataset_size: 65376540
  - config_name: hw_squad
    features: *ref_0
    splits:
      - name: train
        num_bytes: 11946273885
        num_examples: 20429
    download_size: 11900434245
    dataset_size: 11946273885
  - config_name: Document-infographic_gpt4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 292080150
        num_examples: 573
    download_size: 248625737
    dataset_size: 292080150
  - config_name: Document-infographic_vqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 384244242
        num_examples: 1219
    download_size: 325317510
    dataset_size: 384244242
  - config_name: Document-infographic_vqa_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 318758651
        num_examples: 517
    download_size: 317633180
    dataset_size: 318758651
  - config_name: infovqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 539122
        num_examples: 1
    download_size: 512516
    dataset_size: 539122
  - config_name: llavar_gpt4_20k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12912555
        num_examples: 334
    download_size: 11819053
    dataset_size: 12912555
  - config_name: olmOCR-mix-0225-books
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2895726903
        num_examples: 11270
    download_size: 2848559709
    dataset_size: 2895726903
  - config_name: olmOCR-mix-0225-documents
    features: *ref_0
    splits:
      - name: train
        num_bytes: 38413314839
        num_examples: 227958
    download_size: 37951071790
    dataset_size: 38413314839
  - config_name: Document-other
    features: *ref_0
    splits:
      - name: train
        num_bytes: 52872677
        num_examples: 157
    download_size: 46179316
    dataset_size: 52872677
  - config_name: pdfvqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 698643140
        num_examples: 8562
    download_size: 634300951
    dataset_size: 698643140
  - config_name: poie_multi_turn_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 72305
        num_examples: 1
    download_size: 85920
    dataset_size: 72305
  - config_name: screen-qa-cleaned
    features: *ref_0
    splits:
      - name: train
        num_bytes: 665774214
        num_examples: 1775
    download_size: 663485102
    dataset_size: 665774214
  - config_name: slidevqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1770250489
        num_examples: 1910
    download_size: 1728746455
    dataset_size: 1770250489
  - config_name: sujet_finance
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3556914909
        num_examples: 9785
    download_size: 3443715961
    dataset_size: 3556914909
  - config_name: tat_dqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12282874
        num_examples: 942
    download_size: 9629394
    dataset_size: 12282874
  - config_name: ureader
    features: *ref_0
    splits:
      - name: train
        num_bytes: 10615847415
        num_examples: 39568
    download_size: 10585001392
    dataset_size: 10615847415
  - config_name: ureader_chart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 18788411390
        num_examples: 56079
    download_size: 18335126578
    dataset_size: 18788411390
  - config_name: Document-ureader_kg
    features: *ref_0
    splits:
      - name: train
        num_bytes: 263653214
        num_examples: 788
    download_size: 253674045
    dataset_size: 263653214
  - config_name: ureader_ocr
    features: *ref_0
    splits:
      - name: train
        num_bytes: 101963017
        num_examples: 502
    download_size: 100108086
    dataset_size: 101963017
  - config_name: visualmrc_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 27926067
        num_examples: 40
    download_size: 27868895
    dataset_size: 27926067
  - config_name: vqaonbd
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4449802554
        num_examples: 39780
    download_size: 3631951798
    dataset_size: 4449802554
  - config_name: aria_ui_desktop
    features: *ref_0
    splits:
      - name: train
        num_bytes: 384113750
        num_examples: 3458
    download_size: 283756743
    dataset_size: 384113750
  - config_name: aria_ui_mobile
    features: *ref_0
    splits:
      - name: train
        num_bytes: 35213632674
        num_examples: 70028
    download_size: 34587123131
    dataset_size: 35213632674
  - config_name: aria_ui_web
    features: *ref_0
    splits:
      - name: train
        num_bytes: 82929947194
        num_examples: 414005
    download_size: 81938891167
    dataset_size: 82929947194
  - config_name: groundui
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3829069643
        num_examples: 13425
    download_size: 3824253461
    dataset_size: 3829069643
  - config_name: A-OKVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5077229588
        num_examples: 18292
    download_size: 5023747048
    dataset_size: 5077229588
  - config_name: ALFWorld
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1476177193
        num_examples: 21222
    download_size: 1243344377
    dataset_size: 1476177193
  - config_name: ALLaVA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 188064516930
        num_examples: 587072
    download_size: 178999350532
    dataset_size: 188064516930
  - config_name: ALLaVA-Instruct-LAION-4V_Chinese
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5563294986
        num_examples: 9645
    download_size: 5533340625
    dataset_size: 5563294986
  - config_name: ALLaVA-Instruct-VFLAN-4V_Chinese
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2895574860
        num_examples: 7417
    download_size: 2878008027
    dataset_size: 2895574860
  - config_name: ART500K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 621238281
        num_examples: 8425
    download_size: 583064828
    dataset_size: 621238281
  - config_name: Birds-to-Words
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1395170064
        num_examples: 9448
    download_size: 1366971260
    dataset_size: 1395170064
  - config_name: Cambrian
    features: *ref_0
    splits:
      - name: train
        num_bytes: 575987195
        num_examples: 2840
    download_size: 512678064
    dataset_size: 575987195
  - config_name: Cambrian_Filter
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4312715057
        num_examples: 23339
    download_size: 3726052164
    dataset_size: 4312715057
  - config_name: Co-Instruct
    features: *ref_0
    splits:
      - name: train
        num_bytes: 16398867371
        num_examples: 27125
    download_size: 16300826504
    dataset_size: 16398867371
  - config_name: DreamSim
    features: *ref_0
    splits:
      - name: train
        num_bytes: 24503814625
        num_examples: 10710
    download_size: 24496861291
    dataset_size: 24503814625
  - config_name: EST-VQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3741195579
        num_examples: 13086
    download_size: 3674931905
    dataset_size: 3741195579
  - config_name: GQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4238702807
        num_examples: 26174
    download_size: 3802845698
    dataset_size: 4238702807
  - config_name: GQA-LLaVA-NeXT
    features: *ref_0
    splits:
      - name: train
        num_bytes: 34406121
        num_examples: 222
    download_size: 33098294
    dataset_size: 34406121
  - config_name: Hateful_Memes
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1389751071
        num_examples: 5348
    download_size: 1362999753
    dataset_size: 1389751071
  - config_name: IDK
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1660533881
        num_examples: 9355
    download_size: 1502836019
    dataset_size: 1660533881
  - config_name: General-IconQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12998
        num_examples: 2
    download_size: 27383
    dataset_size: 12998
  - config_name: InternVL-SA-1B-Caption
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4943619472
        num_examples: 4972
    download_size: 4916460389
    dataset_size: 4943619472
  - config_name: KVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 17827052535
        num_examples: 17212
    download_size: 17628434132
    dataset_size: 17827052535
  - config_name: KonIQ-10k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 888425787
        num_examples: 9102
    download_size: 856894377
    dataset_size: 888425787
  - config_name: LLaVA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 137295323560
        num_examples: 634600
    download_size: 129258337018
    dataset_size: 137295323560
  - config_name: LLaVA-NeXT-Data
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4198272732
        num_examples: 89068
    download_size: 3835688103
    dataset_size: 4198272732
  - config_name: LRV_Normal
    features: *ref_0
    splits:
      - name: train
        num_bytes: 801419868
        num_examples: 10456
    download_size: 518142426
    dataset_size: 801419868
  - config_name: LVIS-InstructV4
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12208490041
        num_examples: 65502
    download_size: 10994721682
    dataset_size: 12208490041
  - config_name: MMChat-Twitter-Post
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2270647482
        num_examples: 1410
    download_size: 2253717667
    dataset_size: 2270647482
  - config_name: NewYorker_Caption
    features: *ref_0
    splits:
      - name: train
        num_bytes: 69405822
        num_examples: 2252
    download_size: 59102218
    dataset_size: 69405822
  - config_name: Objects365
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1053976275
        num_examples: 5122
    download_size: 1025108932
    dataset_size: 1053976275
  - config_name: PixMo
    features: *ref_0
    splits:
      - name: train
        num_bytes: 131499807387
        num_examples: 292991
    download_size: 129239670511
    dataset_size: 131499807387
  - config_name: SVIT
    features: *ref_0
    splits:
      - name: train
        num_bytes: 204758295899
        num_examples: 602448
    download_size: 195252423968
    dataset_size: 204758295899
  - config_name: SVIT_core_150K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 842180196
        num_examples: 2869
    download_size: 829989122
    dataset_size: 842180196
  - config_name: SVIT_mix_665K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8648472405
        num_examples: 25962
    download_size: 8531085938
    dataset_size: 8648472405
  - config_name: ScanQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2922293401
        num_examples: 3649
    download_size: 2879154629
    dataset_size: 2922293401
  - config_name: ShareGPT4V
    features: *ref_0
    splits:
      - name: train
        num_bytes: 17296086471
        num_examples: 56571
    download_size: 16830383251
    dataset_size: 17296086471
  - config_name: ShareGPT4o
    features: *ref_0
    splits:
      - name: train
        num_bytes: 23459512303
        num_examples: 41025
    download_size: 22567782735
    dataset_size: 23459512303
  - config_name: VIST
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4880051011
        num_examples: 17466
    download_size: 4807243558
    dataset_size: 4880051011
  - config_name: VSR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 548704002
        num_examples: 2062
    download_size: 542437608
    dataset_size: 548704002
  - config_name: ViQuAE
    features: *ref_0
    splits:
      - name: train
        num_bytes: 437884698
        num_examples: 1755
    download_size: 432348525
    dataset_size: 437884698
  - config_name: Vision_FLAN
    features: *ref_0
    splits:
      - name: train
        num_bytes: 64784189373
        num_examples: 333793
    download_size: 63858374144
    dataset_size: 64784189373
  - config_name: Visual7W
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5226359664
        num_examples: 24365
    download_size: 5036058291
    dataset_size: 5226359664
  - config_name: General-VizWiz_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 97744433
        num_examples: 700
    download_size: 96283050
    dataset_size: 97744433
  - config_name: WildVision
    features: *ref_0
    splits:
      - name: train
        num_bytes: 322271123
        num_examples: 1619
    download_size: 294297550
    dataset_size: 322271123
  - config_name: alfworldgpt
    features: *ref_0
    splits:
      - name: train
        num_bytes: 85454420
        num_examples: 1835
    download_size: 81312740
    dataset_size: 85454420
  - config_name: alfworldgpt_45k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 50778218
        num_examples: 854
    download_size: 46461174
    dataset_size: 50778218
  - config_name: allava
    features: *ref_0
    splits:
      - name: train
        num_bytes: 28038133067
        num_examples: 59026
    download_size: 27833641884
    dataset_size: 28038133067
  - config_name: allava-laion-500k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1669374259
        num_examples: 8436
    download_size: 1590632418
    dataset_size: 1669374259
  - config_name: allava-vflan-200k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1414341188
        num_examples: 7981
    download_size: 1379061525
    dataset_size: 1414341188
  - config_name: allava_instruct_gpt4v_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 297762797
        num_examples: 1427
    download_size: 282413186
    dataset_size: 297762797
  - config_name: allava_instruct_gpt4v_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 91199890
        num_examples: 342
    download_size: 88260334
    dataset_size: 91199890
  - config_name: allava_instruct_laion4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 249009815
        num_examples: 2217
    download_size: 228465335
    dataset_size: 249009815
  - config_name: allava_instruct_vflan4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 94048666
        num_examples: 1528
    download_size: 85401085
    dataset_size: 94048666
  - config_name: allava_writing_gpt4o_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1043236
        num_examples: 5
    download_size: 955700
    dataset_size: 1043236
  - config_name: aokvqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 923678559
        num_examples: 3431
    download_size: 915978459
    dataset_size: 923678559
  - config_name: art500k_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12628420
        num_examples: 183
    download_size: 12218203
    dataset_size: 12628420
  - config_name: cambrain_multiturn_gpt4o_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 668865
        num_examples: 9
    download_size: 517675
    dataset_size: 668865
  - config_name: cambrian_filtered
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15778762
        num_examples: 342
    download_size: 13716032
    dataset_size: 15778762
  - config_name: chinesememe
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6335129834
        num_examples: 36883
    download_size: 6265331439
    dataset_size: 6335129834
  - config_name: coco_colors
    features: *ref_0
    splits:
      - name: train
        num_bytes: 789901147
        num_examples: 2495
    download_size: 786269108
    dataset_size: 789901147
  - config_name: coco_negative_gpt4o_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 117206
        num_examples: 1
    download_size: 132244
    dataset_size: 117206
  - config_name: drivelm
    features: *ref_0
    splits:
      - name: train
        num_bytes: 20352764762
        num_examples: 3700
    download_size: 20027130586
    dataset_size: 20352764762
  - config_name: estvqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 86633173
        num_examples: 321
    download_size: 85032040
    dataset_size: 86633173
  - config_name: google_landmarks
    features: *ref_0
    splits:
      - name: train
        num_bytes: 422190686152
        num_examples: 190899
    download_size: 421458036789
    dataset_size: 422190686152
  - config_name: gpt-4o
    features: *ref_0
    splits:
      - name: train
        num_bytes: 173499178
        num_examples: 489
    download_size: 168575201
    dataset_size: 173499178
  - config_name: gpt4gen_rd_boxcot_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1303848
        num_examples: 10
    download_size: 1289232
    dataset_size: 1303848
  - config_name: gpt77k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 13642554
        num_examples: 102
    download_size: 12845087
    dataset_size: 13642554
  - config_name: gqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1352786
        num_examples: 9
    download_size: 1347485
    dataset_size: 1352786
  - config_name: hateful_memes_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 359602121
        num_examples: 1320
    download_size: 355024905
    dataset_size: 359602121
  - config_name: idefics375k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 17155956029
        num_examples: 226591
    download_size: 10856280781
    dataset_size: 17155956029
  - config_name: idk_11k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 33423051
        num_examples: 208
    download_size: 32582218
    dataset_size: 33423051
  - config_name: image_textualization(filtered)
    features: *ref_0
    splits:
      - name: train
        num_bytes: 290008206
        num_examples: 1137
    download_size: 287626503
    dataset_size: 290008206
  - config_name: inat_train2018_merge_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 345592
        num_examples: 1
    download_size: 364524
    dataset_size: 345592
  - config_name: indoor_qa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 490556436
        num_examples: 3006
    download_size: 487564845
    dataset_size: 490556436
  - config_name: internvl_sa1b_caption_gpt4o_review_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5875685817
        num_examples: 5896
    download_size: 5837387686
    dataset_size: 5875685817
  - config_name: koniq10k_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 26108778
        num_examples: 224
    download_size: 25395741
    dataset_size: 26108778
  - config_name: kvqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 238173294
        num_examples: 390
    download_size: 234562915
    dataset_size: 238173294
  - config_name: laion_gpt4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 610611393
        num_examples: 3497
    download_size: 602854181
    dataset_size: 610611393
  - config_name: llava_instruct_300k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 538370121
        num_examples: 1441
    download_size: 533867730
    dataset_size: 538370121
  - config_name: lnqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 171239341199
        num_examples: 295106
    download_size: 170199636172
    dataset_size: 171239341199
  - config_name: localized_narratives
    features: *ref_0
    splits:
      - name: train
        num_bytes: 107737701435
        num_examples: 185122
    download_size: 107552261219
    dataset_size: 107737701435
  - config_name: lrv_normal_filtered
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12086273
        num_examples: 132
    download_size: 7390997
    dataset_size: 12086273
  - config_name: lvis_instruct4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1262030811
        num_examples: 3498
    download_size: 1242832270
    dataset_size: 1262030811
  - config_name: lvis_instruct4v_220k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 952511366
        num_examples: 5689
    download_size: 923728872
    dataset_size: 952511366
  - config_name: memotion
    features: *ref_0
    splits:
      - name: train
        num_bytes: 439743958
        num_examples: 1870
    download_size: 434077935
    dataset_size: 439743958
  - config_name: mmevol
    features: *ref_0
    splits:
      - name: train
        num_bytes: 29333111565
        num_examples: 22222
    download_size: 29164763349
    dataset_size: 29333111565
  - config_name: mmsoc_memotion
    features: *ref_0
    splits:
      - name: train
        num_bytes: 476773974
        num_examples: 2040
    download_size: 470322306
    dataset_size: 476773974
  - config_name: newyorker_caption_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5880933
        num_examples: 288
    download_size: 5107725
    dataset_size: 5880933
  - config_name: nlvr2
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6292246807
        num_examples: 6776
    download_size: 6277412509
    dataset_size: 6292246807
  - config_name: objects365_qa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8302792
        num_examples: 42
    download_size: 7999882
    dataset_size: 8302792
  - config_name: okvqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2269313
        num_examples: 14
    download_size: 2253531
    dataset_size: 2269313
  - config_name: pixmo-cap-qa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 11449754684
        num_examples: 31108
    download_size: 11061434367
    dataset_size: 11449754684
  - config_name: pixmo_ask_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6077773605
        num_examples: 8704
    download_size: 6050228846
    dataset_size: 6077773605
  - config_name: pixmo_point_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1481482058
        num_examples: 3428
    download_size: 1474481595
    dataset_size: 1481482058
  - config_name: scienceqa_nona_context
    features: *ref_0
    splits:
      - name: train
        num_bytes: 10264711
        num_examples: 244
    download_size: 7995506
    dataset_size: 10264711
  - config_name: screen2words
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2652009106
        num_examples: 8043
    download_size: 2637692698
    dataset_size: 2652009106
  - config_name: screenqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2603671588
        num_examples: 6995
    download_size: 2596865103
    dataset_size: 2603671588
  - config_name: sharegpt4o
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3609196037
        num_examples: 7835
    download_size: 3578653073
    dataset_size: 3609196037
  - config_name: sharegpt4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6371997405
        num_examples: 9140
    download_size: 6346635492
    dataset_size: 6371997405
  - config_name: sharegpt4v_coco
    features: *ref_0
    splits:
      - name: train
        num_bytes: 369366
        num_examples: 1
    download_size: 380648
    dataset_size: 369366
  - config_name: sharegpt4v_instruct_gpt4-vision_cap100k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15180829
        num_examples: 12
    download_size: 15168513
    dataset_size: 15180829
  - config_name: sharegpt4v_knowledge
    features: *ref_0
    splits:
      - name: train
        num_bytes: 34531051
        num_examples: 203
    download_size: 33334756
    dataset_size: 34531051
  - config_name: sharegpt4v_sam
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3406267555
        num_examples: 8603
    download_size: 3382405739
    dataset_size: 3406267555
  - config_name: sketchyvqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 108258773
        num_examples: 5923
    download_size: 104813447
    dataset_size: 108258773
  - config_name: spark
    features: *ref_0
    splits:
      - name: train
        num_bytes: 311666711
        num_examples: 2508
    download_size: 308073221
    dataset_size: 311666711
  - config_name: spatialsense
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2364261557
        num_examples: 8819
    download_size: 2354459038
    dataset_size: 2364261557
  - config_name: spot_the_diff
    features: *ref_0
    splits:
      - name: train
        num_bytes: 674391020
        num_examples: 5107
    download_size: 668220691
    dataset_size: 674391020
  - config_name: General-st_vqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 948851
        num_examples: 8
    download_size: 952265
    dataset_size: 948851
  - config_name: stvqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 71615323
        num_examples: 475
    download_size: 70432280
    dataset_size: 71615323
  - config_name: textvqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 30277214
        num_examples: 106
    download_size: 30081916
    dataset_size: 30277214
  - config_name: tqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 242795985
        num_examples: 1332
    download_size: 240075036
    dataset_size: 242795985
  - config_name: unknown
    features: *ref_0
    splits:
      - name: train
        num_bytes: 67655738
        num_examples: 487
    download_size: 64759362
    dataset_size: 67655738
  - config_name: vflan_cleaned
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3672652581
        num_examples: 16357
    download_size: 3644039792
    dataset_size: 3672652581
  - config_name: viquae
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9123696
        num_examples: 43
    download_size: 9056267
    dataset_size: 9123696
  - config_name: viquae_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5117862
        num_examples: 65
    download_size: 5071145
    dataset_size: 5117862
  - config_name: vision_flan(filtered)
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2101477782
        num_examples: 11829
    download_size: 2086694908
    dataset_size: 2101477782
  - config_name: vision_flan_filtered
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8109565086
        num_examples: 40520
    download_size: 7933716377
    dataset_size: 8109565086
  - config_name: visual7w_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 16923829
        num_examples: 87
    download_size: 16621934
    dataset_size: 16923829
  - config_name: visual7w_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 230652
        num_examples: 6
    download_size: 240066
    dataset_size: 230652
  - config_name: General-vqarad_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8874671
        num_examples: 52
    download_size: 8726583
    dataset_size: 8874671
  - config_name: vsr_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 57934955
        num_examples: 248
    download_size: 57396509
    dataset_size: 57934955
  - config_name: vsr_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 336164
        num_examples: 3
    download_size: 348333
    dataset_size: 336164
  - config_name: wildvision_gpt4o_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 42455474
        num_examples: 196
    download_size: 38368859
    dataset_size: 42455474
  - config_name: wildvision_gpt4v_to_gpt4o_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8433178
        num_examples: 41
    download_size: 8131095
    dataset_size: 8433178
  - config_name: yesbut
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1591031968
        num_examples: 3846
    download_size: 1584663872
    dataset_size: 1591031968
  - config_name: CLEVR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15420475854
        num_examples: 127798
    download_size: 13115306354
    dataset_size: 15420475854
  - config_name: CLEVR-Math_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 31691723
        num_examples: 1067
    download_size: 28768640
    dataset_size: 31691723
  - config_name: CLEVR_Math
    features: *ref_0
    splits:
      - name: train
        num_bytes: 104934996
        num_examples: 3467
    download_size: 92551983
    dataset_size: 104934996
  - config_name: IconQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 274816772
        num_examples: 28534
    download_size: 182109118
    dataset_size: 274816772
  - config_name: Grounding-IconQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 11768509
        num_examples: 1013
    download_size: 6561048
    dataset_size: 11768509
  - config_name: MathV360K_VQA-AS
    features: *ref_0
    splits:
      - name: train
        num_bytes: 81444561
        num_examples: 657
    download_size: 75797243
    dataset_size: 81444561
  - config_name: MovieNet
    features: *ref_0
    splits:
      - name: train
        num_bytes: 121741378
        num_examples: 6158
    download_size: 102377508
    dataset_size: 121741378
  - config_name: Super-CLEVR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 262176833
        num_examples: 3210
    download_size: 250256603
    dataset_size: 262176833
  - config_name: Super-CLEVR_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 101895625
        num_examples: 1166
    download_size: 96975487
    dataset_size: 101895625
  - config_name: TQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 157635944
        num_examples: 16695
    download_size: 124564387
    dataset_size: 157635944
  - config_name: TallyQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 16202436246
        num_examples: 69863
    download_size: 16001032047
    dataset_size: 16202436246
  - config_name: VisualGenome-LLaVA-NeXT
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9985372446
        num_examples: 68188
    download_size: 9688981552
    dataset_size: 9985372446
  - config_name: clevr_700k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 55422397
        num_examples: 531
    download_size: 52747062
    dataset_size: 55422397
  - config_name: coco_rem_en_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 340685
        num_examples: 2
    download_size: 354650
    dataset_size: 340685
  - config_name: ctw
    features: *ref_0
    splits:
      - name: train
        num_bytes: 28504814379
        num_examples: 9960
    download_size: 28482854244
    dataset_size: 28504814379
  - config_name: fsc147_train_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6040442
        num_examples: 146
    download_size: 4601957
    dataset_size: 6040442
  - config_name: iconqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 10883577
        num_examples: 1939
    download_size: 8826475
    dataset_size: 10883577
  - config_name: iconqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15675710
        num_examples: 2202
    download_size: 11792933
    dataset_size: 15675710
  - config_name: movienet_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4396879
        num_examples: 230
    download_size: 3831859
    dataset_size: 4396879
  - config_name: objects365_qa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 681596874070
        num_examples: 1620920
    download_size: 678037560766
    dataset_size: 681596874070
  - config_name: oodvqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2111349363
        num_examples: 5689
    download_size: 2106843567
    dataset_size: 2111349363
  - config_name: Grounding-other
    features: *ref_0
    splits:
      - name: train
        num_bytes: 77225346
        num_examples: 800
    download_size: 74465208
    dataset_size: 77225346
  - config_name: tallyqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 224819928
        num_examples: 918
    download_size: 224182918
    dataset_size: 224819928
  - config_name: tallyqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1443235617
        num_examples: 5766
    download_size: 1409055763
    dataset_size: 1443235617
  - config_name: tallyqa_coco_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5959952
        num_examples: 32
    download_size: 5136303
    dataset_size: 5959952
  - config_name: tallyqa_vg_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4551915
        num_examples: 30
    download_size: 4114150
    dataset_size: 4551915
  - config_name: tqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 18278207
        num_examples: 2215
    download_size: 14343745
    dataset_size: 18278207
  - config_name: ArXiv_OCR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 88014661665
        num_examples: 159299
    download_size: 86019288203
    dataset_size: 88014661665
  - config_name: CTW
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9229760060
        num_examples: 9314
    download_size: 9199261285
    dataset_size: 9229760060
  - config_name: Chrome-Writing
    features: *ref_0
    splits:
      - name: train
        num_bytes: 54802331
        num_examples: 9247
    download_size: 32323077
    dataset_size: 54802331
  - config_name: HME
    features: *ref_0
    splits:
      - name: train
        num_bytes: 489939109
        num_examples: 93639
    download_size: 362707932
    dataset_size: 489939109
  - config_name: IAM
    features: *ref_0
    splits:
      - name: train
        num_bytes: 539850765
        num_examples: 5658
    download_size: 534457783
    dataset_size: 539850765
  - config_name: ICDAR-LSVT-zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6781990158
        num_examples: 22483
    download_size: 6621783850
    dataset_size: 6781990158
  - config_name: ICDAR2017
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6130470023
        num_examples: 8251
    download_size: 5963468425
    dataset_size: 6130470023
  - config_name: ICDAR2019
    features: *ref_0
    splits:
      - name: train
        num_bytes: 773127162
        num_examples: 2614
    download_size: 760600721
    dataset_size: 773127162
  - config_name: IIIT_5K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6219271
        num_examples: 1976
    download_size: 4768769
    dataset_size: 6219271
  - config_name: K12_Printing
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1115846174
        num_examples: 211104
    download_size: 923171470
    dataset_size: 1115846174
  - config_name: MTWI_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1310124735
        num_examples: 7338
    download_size: 1252679985
    dataset_size: 1310124735
  - config_name: Orand-Car-A
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7405377
        num_examples: 1999
    download_size: 4864871
    dataset_size: 7405377
  - config_name: ReCTs
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8980635524
        num_examples: 17000
    download_size: 8964616072
    dataset_size: 8980635524
  - config_name: Rendered_Text
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7682659886
        num_examples: 9995
    download_size: 7651724807
    dataset_size: 7682659886
  - config_name: Sroie
    features: *ref_0
    splits:
      - name: train
        num_bytes: 476820673
        num_examples: 31832
    download_size: 376310130
    dataset_size: 476820673
  - config_name: TextOCR
    features: *ref_0
    splits:
      - name: train
        num_bytes: 18104595556
        num_examples: 47656
    download_size: 17899806143
    dataset_size: 18104595556
  - config_name: VCR-Wiki
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2320870582
        num_examples: 73235
    download_size: 2147179859
    dataset_size: 2320870582
  - config_name: arxiv_ocr_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1389256
        num_examples: 1
    download_size: 1407332
    dataset_size: 1389256
  - config_name: bentham
    features: *ref_0
    splits:
      - name: train
        num_bytes: 823638472
        num_examples: 10843
    download_size: 820359974
    dataset_size: 823638472
  - config_name: captcha
    features: *ref_0
    splits:
      - name: train
        num_bytes: 913167643
        num_examples: 113062
    download_size: 902213368
    dataset_size: 913167643
  - config_name: ctw_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2069318885
        num_examples: 2079
    download_size: 2064047118
    dataset_size: 2069318885
  - config_name: hme100k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 26070804
        num_examples: 3714
    download_size: 25646283
    dataset_size: 26070804
  - config_name: icdar2017_rctw17_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 833052946
        num_examples: 742
    download_size: 813979175
    dataset_size: 833052946
  - config_name: icdar2017_rects_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 22280337
        num_examples: 298
    download_size: 21466490
    dataset_size: 22280337
  - config_name: icdar2019_art_task1_3_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 250242172
        num_examples: 547
    download_size: 247655411
    dataset_size: 250242172
  - config_name: icdar_lsvt_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 959923524
        num_examples: 3174
    download_size: 937568325
    dataset_size: 959923524
  - config_name: imgur5k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8482063359
        num_examples: 5928
    download_size: 8462954064
    dataset_size: 8482063359
  - config_name: invoices_receipts
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1033470861
        num_examples: 2691
    download_size: 1031617080
    dataset_size: 1033470861
  - config_name: k12_printing
    features: *ref_0
    splits:
      - name: train
        num_bytes: 748590824
        num_examples: 106920
    download_size: 736022396
    dataset_size: 748590824
  - config_name: latex_handwritten
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3864608013
        num_examples: 19992
    download_size: 3860142491
    dataset_size: 3864608013
  - config_name: latexformulas
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2290798161
        num_examples: 550760
    download_size: 2184501437
    dataset_size: 2290798161
  - config_name: mathwriting-google
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4395886699
        num_examples: 299977
    download_size: 4358303360
    dataset_size: 4395886699
  - config_name: mtwi_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 106127323
        num_examples: 654
    download_size: 101101257
    dataset_size: 106127323
  - config_name: ocrvqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 18331187909
        num_examples: 102929
    download_size: 18033831475
    dataset_size: 18331187909
  - config_name: sroie
    features: *ref_0
    splits:
      - name: train
        num_bytes: 13142832
        num_examples: 5132
    download_size: 12740775
    dataset_size: 13142832
  - config_name: svrd
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2738681138
        num_examples: 4396
    download_size: 2734179148
    dataset_size: 2738681138
  - config_name: synth_ocr_fanti_regular_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 351371
        num_examples: 1
    download_size: 354002
    dataset_size: 351371
  - config_name: synthdog
    features: *ref_0
    splits:
      - name: train
        num_bytes: 172797839264
        num_examples: 460352
    download_size: 172676638757
    dataset_size: 172797839264
  - config_name: tal_ocr_eng
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1152076582
        num_examples: 148658
    download_size: 1134636326
    dataset_size: 1152076582
  - config_name: textvqa_train_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1447216148
        num_examples: 2094
    download_size: 1443738910
    dataset_size: 1447216148
  - config_name: vcr_wiki_en_easy
    features: *ref_0
    splits:
      - name: train
        num_bytes: 26778803
        num_examples: 693
    download_size: 23572642
    dataset_size: 26778803
  - config_name: vcr_wiki_en_hard
    features: *ref_0
    splits:
      - name: train
        num_bytes: 22798150
        num_examples: 616
    download_size: 20251361
    dataset_size: 22798150
  - config_name: vcr_wiki_zh_easy
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9782853
        num_examples: 309
    download_size: 8895013
    dataset_size: 9782853
  - config_name: vcr_wiki_zh_hard
    features: *ref_0
    splits:
      - name: train
        num_bytes: 25613396
        num_examples: 787
    download_size: 22242400
    dataset_size: 25613396
  - config_name: wordart
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1993527064
        num_examples: 4804
    download_size: 1992907632
    dataset_size: 1993527064
  - config_name: AI2D
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1166637298
        num_examples: 12450
    download_size: 1022982283
    dataset_size: 1166637298
  - config_name: CMM-Math
    features: *ref_0
    splits:
      - name: train
        num_bytes: 28348857
        num_examples: 1341
    download_size: 15975037
    dataset_size: 28348857
  - config_name: STEM-CoSyn
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1822279360
        num_examples: 60794
    download_size: 1263785893
    dataset_size: 1822279360
  - config_name: STEM-CoSyn_400k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 498318799
        num_examples: 18013
    download_size: 427413878
    dataset_size: 498318799
  - config_name: GEOS_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 368011
        num_examples: 27
    download_size: 162948
    dataset_size: 368011
  - config_name: Geo170K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 760627410
        num_examples: 13284
    download_size: 189508390
    dataset_size: 760627410
  - config_name: GeoQA+
    features: *ref_0
    splits:
      - name: train
        num_bytes: 142465758
        num_examples: 10639
    download_size: 50566703
    dataset_size: 142465758
  - config_name: GeoQA+_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 9688869
        num_examples: 864
    download_size: 3862441
    dataset_size: 9688869
  - config_name: GeomVerse
    features: *ref_0
    splits:
      - name: train
        num_bytes: 338892645
        num_examples: 3243
    download_size: 315158242
    dataset_size: 338892645
  - config_name: Geometry3K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 139404282
        num_examples: 7786
    download_size: 65691076
    dataset_size: 139404282
  - config_name: Geometry3K_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15731065
        num_examples: 932
    download_size: 7953717
    dataset_size: 15731065
  - config_name: STEM-IconQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 149928
        num_examples: 19
    download_size: 128863
    dataset_size: 149928
  - config_name: InterGPS
    features: *ref_0
    splits:
      - name: train
        num_bytes: 197610
        num_examples: 4
    download_size: 86438
    dataset_size: 197610
  - config_name: MAVIS-Metagen
    features: *ref_0
    splits:
      - name: train
        num_bytes: 782381529
        num_examples: 46440
    download_size: 259899882
    dataset_size: 782381529
  - config_name: MMChem
    features: *ref_0
    splits:
      - name: train
        num_bytes: 455785868
        num_examples: 8728
    download_size: 348165929
    dataset_size: 455785868
  - config_name: MapQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2690326077
        num_examples: 37626
    download_size: 1796385878
    dataset_size: 2690326077
  - config_name: STEM-MapQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7538535
        num_examples: 267
    download_size: 5566565
    dataset_size: 7538535
  - config_name: MathV360K_TQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 230707416
        num_examples: 1134
    download_size: 208808646
    dataset_size: 230707416
  - config_name: MathV360K_VQA-RAD
    features: *ref_0
    splits:
      - name: train
        num_bytes: 77005
        num_examples: 1
    download_size: 91650
    dataset_size: 77005
  - config_name: MetaMathQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2397674215
        num_examples: 95255
    download_size: 1683356353
    dataset_size: 2397674215
  - config_name: PMC-VQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 360698757
        num_examples: 18754
    download_size: 239667526
    dataset_size: 360698757
  - config_name: PMC-VQA_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 79796933
        num_examples: 3874
    download_size: 64582442
    dataset_size: 79796933
  - config_name: PathVQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 623218748
        num_examples: 8152
    download_size: 563932984
    dataset_size: 623218748
  - config_name: RAVEN
    features: *ref_0
    splits:
      - name: train
        num_bytes: 227159396
        num_examples: 9340
    download_size: 128799790
    dataset_size: 227159396
  - config_name: ScienceQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 329761370
        num_examples: 9434
    download_size: 262081043
    dataset_size: 329761370
  - config_name: STEM-SynMultiImageQA
    features: *ref_0
    splits:
      - name: train
        num_bytes: 897232377
        num_examples: 13095
    download_size: 739229122
    dataset_size: 897232377
  - config_name: SynthCodeNet
    features: *ref_0
    splits:
      - name: train
        num_bytes: 22529635590
        num_examples: 498730
    download_size: 21935989459
    dataset_size: 22529635590
  - config_name: SynthFormulaNet
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1238424256
        num_examples: 497492
    download_size: 1092826586
    dataset_size: 1238424256
  - config_name: UniGeo
    features: *ref_0
    splits:
      - name: train
        num_bytes: 65793305
        num_examples: 5561
    download_size: 23715934
    dataset_size: 65793305
  - config_name: UniGeo_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6258412
        num_examples: 564
    download_size: 2471011
    dataset_size: 6258412
  - config_name: VQA-RAD
    features: *ref_0
    splits:
      - name: train
        num_bytes: 20016600
        num_examples: 241
    download_size: 13824583
    dataset_size: 20016600
  - config_name: VisualWebInstruct
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2827433232
        num_examples: 57445
    download_size: 2323727118
    dataset_size: 2827433232
  - config_name: VisualWebInstruct_filtered
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7075784864
        num_examples: 167283
    download_size: 5517703006
    dataset_size: 7075784864
  - config_name: VizWiz
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2971775838
        num_examples: 3101
    download_size: 2957605855
    dataset_size: 2971775838
  - config_name: STEM-VizWiz_MathV360K
    features: *ref_0
    splits:
      - name: train
        num_bytes: 538519547
        num_examples: 559
    download_size: 535665763
    dataset_size: 538519547
  - config_name: WebSight
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4619266618
        num_examples: 7574
    download_size: 4547122608
    dataset_size: 4619266618
  - config_name: ai2d_cap_gpt4o_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8560794
        num_examples: 70
    download_size: 8032611
    dataset_size: 8560794
  - config_name: ai2d_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2277751
        num_examples: 18
    download_size: 2253555
    dataset_size: 2277751
  - config_name: ai2d_cot_gpt4o_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 215829
        num_examples: 4
    download_size: 226417
    dataset_size: 215829
  - config_name: ai2d_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 35293
        num_examples: 3
    download_size: 46102
    dataset_size: 35293
  - config_name: ai2d_gpt4v
    features: *ref_0
    splits:
      - name: train
        num_bytes: 6196004
        num_examples: 185
    download_size: 5006657
    dataset_size: 6196004
  - config_name: ai2d_internvl
    features: *ref_0
    splits:
      - name: train
        num_bytes: 15212196
        num_examples: 155
    download_size: 14804770
    dataset_size: 15212196
  - config_name: cmm_math_cot_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2883765
        num_examples: 128
    download_size: 1897675
    dataset_size: 2883765
  - config_name: cosyn_chemical_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2377235
        num_examples: 113
    download_size: 1789952
    dataset_size: 2377235
  - config_name: cosyn_circuit_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 265106
        num_examples: 14
    download_size: 206468
    dataset_size: 265106
  - config_name: cosyn_math_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 158689459
        num_examples: 4501
    download_size: 129199735
    dataset_size: 158689459
  - config_name: cosyn_music_train
    features: *ref_0
    splits:
      - name: train
        num_bytes: 23423
        num_examples: 1
    download_size: 35930
    dataset_size: 23423
  - config_name: datik
    features: *ref_0
    splits:
      - name: train
        num_bytes: 959345896
        num_examples: 73990
    download_size: 558415206
    dataset_size: 959345896
  - config_name: datikz
    features: *ref_0
    splits:
      - name: train
        num_bytes: 173698098
        num_examples: 27158
    download_size: 136704674
    dataset_size: 173698098
  - config_name: geo170k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 4443558
        num_examples: 323
    download_size: 1554550
    dataset_size: 4443558
  - config_name: geo170k_align
    features: *ref_0
    splits:
      - name: train
        num_bytes: 27263393
        num_examples: 1416
    download_size: 8981889
    dataset_size: 27263393
  - config_name: geo170k_qa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3973766
        num_examples: 153
    download_size: 2048689
    dataset_size: 3973766
  - config_name: geo3k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 525230
        num_examples: 28
    download_size: 401037
    dataset_size: 525230
  - config_name: geomverse
    features: *ref_0
    splits:
      - name: train
        num_bytes: 207427367
        num_examples: 1720
    download_size: 189944421
    dataset_size: 207427367
  - config_name: geomverse_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 328893422
        num_examples: 2157
    download_size: 316933097
    dataset_size: 328893422
  - config_name: geoqa_plus_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 36772
        num_examples: 3
    download_size: 44975
    dataset_size: 36772
  - config_name: intergps_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 14167424
        num_examples: 909
    download_size: 10248002
    dataset_size: 14167424
  - config_name: mapqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 11504590
        num_examples: 262
    download_size: 10223487
    dataset_size: 11504590
  - config_name: mathvision_3k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 8392377
        num_examples: 178
    download_size: 4185633
    dataset_size: 8392377
  - config_name: mavis_math
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1286808609
        num_examples: 24249
    download_size: 1099625818
    dataset_size: 1286808609
  - config_name: mavis_math_geometry_caption_to_question_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 801941113
        num_examples: 7970
    download_size: 272651564
    dataset_size: 801941113
  - config_name: mavis_math_metagen
    features: *ref_0
    splits:
      - name: train
        num_bytes: 239524666
        num_examples: 10141
    download_size: 140188545
    dataset_size: 239524666
  - config_name: mavis_math_rule_geo
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2361780765
        num_examples: 17663
    download_size: 852603586
    dataset_size: 2361780765
  - config_name: metamathqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 28784900
        num_examples: 995
    download_size: 18920110
    dataset_size: 28784900
  - config_name: metamathqa_ocr_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 36966712
        num_examples: 1273
    download_size: 25223177
    dataset_size: 36966712
  - config_name: mm_chem_exam_single_image_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 54768944
        num_examples: 1096
    download_size: 44609561
    dataset_size: 54768944
  - config_name: mm_chem_ocr_smiles_gpt4o_zh
    features: *ref_0
    splits:
      - name: train
        num_bytes: 1979221
        num_examples: 186
    download_size: 1058169
    dataset_size: 1979221
  - config_name: mmc_instruct
    features: *ref_0
    splits:
      - name: train
        num_bytes: 7034960487
        num_examples: 149978
    download_size: 6266213254
    dataset_size: 7034960487
  - config_name: STEM-other
    features: *ref_0
    splits:
      - name: train
        num_bytes: 121826117
        num_examples: 1754
    download_size: 97535114
    dataset_size: 121826117
  - config_name: pathvqa
    features: *ref_0
    splits:
      - name: train
        num_bytes: 149751890
        num_examples: 787
    download_size: 148414193
    dataset_size: 149751890
  - config_name: pathvqa_32k
    features: *ref_0
    splits:
      - name: train
        num_bytes: 514496048
        num_examples: 6717
    download_size: 481300395
    dataset_size: 514496048
  - config_name: pmc_vqa(mathv360k)
    features: *ref_0
    splits:
      - name: train
        num_bytes: 12228682
        num_examples: 848
    download_size: 10713551
    dataset_size: 12228682
  - config_name: pmc_vqa_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 424567
        num_examples: 6
    download_size: 425464
    dataset_size: 424567
  - config_name: raven
    features: *ref_0
    splits:
      - name: train
        num_bytes: 116068422
        num_examples: 5653
    download_size: 76627458
    dataset_size: 116068422
  - config_name: raven_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 45401743
        num_examples: 2079
    download_size: 28580293
    dataset_size: 45401743
  - config_name: scienceqa_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 3633222
        num_examples: 176
    download_size: 3305300
    dataset_size: 3633222
  - config_name: scienceqa_choice_aug_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 68800
        num_examples: 2
    download_size: 73894
    dataset_size: 68800
  - config_name: study_com_en
    features: *ref_0
    splits:
      - name: train
        num_bytes: 36017
        num_examples: 3
    download_size: 37002
    dataset_size: 36017
  - config_name: visualwebinstruct(filtered)
    features: *ref_0
    splits:
      - name: train
        num_bytes: 5090456105
        num_examples: 43717
    download_size: 4987862836
    dataset_size: 5090456105
  - config_name: visualwebinstruct_si
    features: *ref_0
    splits:
      - name: train
        num_bytes: 313939304
        num_examples: 5013
    download_size: 280122069
    dataset_size: 313939304
  - config_name: STEM-vqarad_cauldron_llava_format
    features: *ref_0
    splits:
      - name: train
        num_bytes: 2498002
        num_examples: 15
    download_size: 2460972
    dataset_size: 2498002
  - config_name: websight
    features: *ref_0
    splits:
      - name: train
        num_bytes: 487832744
        num_examples: 1229
    download_size: 482049554
    dataset_size: 487832744
  - config_name: websight_cauldron
    features: *ref_0
    splits:
      - name: train
        num_bytes: 10459216
        num_examples: 30
    download_size: 10294797
    dataset_size: 10459216
configs:
  - config_name: ST-VQA
    data_files:
      - split: train
        path: Caption/ST-VQA/train-*
  - config_name: Sherlock
    data_files:
      - split: train
        path: Caption/Sherlock/train-*
  - config_name: WIT
    data_files:
      - split: train
        path: Caption/WIT/train-*
  - config_name: coyo
    data_files:
      - split: train
        path: Caption/coyo/train-*
  - config_name: densefusion_1m
    data_files:
      - split: train
        path: Caption/densefusion_1m/train-*
  - config_name: laion_gpt4v_train_format
    data_files:
      - split: train
        path: Caption/laion_gpt4v_train_format/train-*
  - config_name: Caption-other
    data_files:
      - split: train
        path: Caption/other/train-*
  - config_name: pixmo-cap
    data_files:
      - split: train
        path: Caption/pixmo-cap/train-*
  - config_name: rlaifv_chosen_en
    data_files:
      - split: train
        path: Caption/rlaifv_chosen_en/train-*
  - config_name: sharegpt4o_longcap_en
    data_files:
      - split: train
        path: Caption/sharegpt4o_longcap_en/train-*
  - config_name: sherlock_317k
    data_files:
      - split: train
        path: Caption/sherlock_317k/train-*
  - config_name: Caption-st_vqa_cauldron_llava_format
    data_files:
      - split: train
        path: Caption/st_vqa_cauldron_llava_format/train-*
  - config_name: wit
    data_files:
      - split: train
        path: Caption/wit/train-*
  - config_name: ArxivQA
    data_files:
      - split: train
        path: Chart/ArxivQA/train-*
  - config_name: Chart2Text
    data_files:
      - split: train
        path: Chart/Chart2Text/train-*
  - config_name: ChartQA
    data_files:
      - split: train
        path: Chart/ChartQA/train-*
  - config_name: ChartVerse-SFT-1.8M
    data_files:
      - split: train
        path: Chart/ChartVerse-SFT-1.8M/train-*
  - config_name: Chart-CoSyn
    data_files:
      - split: train
        path: Chart/CoSyn/train-*
  - config_name: Chart-CoSyn_400k
    data_files:
      - split: train
        path: Chart/CoSyn_400k/train-*
  - config_name: DVQA
    data_files:
      - split: train
        path: Chart/DVQA/train-*
  - config_name: FigureQA
    data_files:
      - split: train
        path: Chart/FigureQA/train-*
  - config_name: FigureQA_MathV360K
    data_files:
      - split: train
        path: Chart/FigureQA_MathV360K/train-*
  - config_name: HiTab
    data_files:
      - split: train
        path: Chart/HiTab/train-*
  - config_name: Infographic
    data_files:
      - split: train
        path: Chart/Infographic/train-*
  - config_name: LRV_Chart
    data_files:
      - split: train
        path: Chart/LRV_Chart/train-*
  - config_name: MAVIS_Function
    data_files:
      - split: train
        path: Chart/MAVIS_Function/train-*
  - config_name: MMC_Instruction
    data_files:
      - split: train
        path: Chart/MMC_Instruction/train-*
  - config_name: MMTab
    data_files:
      - split: train
        path: Chart/MMTab/train-*
  - config_name: Chart-MapQA_MathV360K
    data_files:
      - split: train
        path: Chart/MapQA_MathV360K/train-*
  - config_name: MultiHiertt
    data_files:
      - split: train
        path: Chart/MultiHiertt/train-*
  - config_name: PlotQA
    data_files:
      - split: train
        path: Chart/PlotQA/train-*
  - config_name: RobuT
    data_files:
      - split: train
        path: Chart/RobuT/train-*
  - config_name: SciTSR
    data_files:
      - split: train
        path: Chart/SciTSR/train-*
  - config_name: SimChart9K
    data_files:
      - split: train
        path: Chart/SimChart9K/train-*
  - config_name: Chart-SynMultiImageQA
    data_files:
      - split: train
        path: Chart/SynMultiImageQA/train-*
  - config_name: SynthChartNet
    data_files:
      - split: train
        path: Chart/SynthChartNet/train-*
  - config_name: TabMWP
    data_files:
      - split: train
        path: Chart/TabMWP/train-*
  - config_name: TabMWP_MathV360K
    data_files:
      - split: train
        path: Chart/TabMWP_MathV360K/train-*
  - config_name: TinyChart
    data_files:
      - split: train
        path: Chart/TinyChart/train-*
  - config_name: UReader
    data_files:
      - split: train
        path: Chart/UReader/train-*
  - config_name: UniChart
    data_files:
      - split: train
        path: Chart/UniChart/train-*
  - config_name: Unichart
    data_files:
      - split: train
        path: Chart/Unichart/train-*
  - config_name: VisText
    data_files:
      - split: train
        path: Chart/VisText/train-*
  - config_name: arxiv_qa_chart_gpt4o
    data_files:
      - split: train
        path: Chart/arxiv_qa_chart_gpt4o/train-*
  - config_name: arxivqa
    data_files:
      - split: train
        path: Chart/arxivqa/train-*
  - config_name: arxivqa_en
    data_files:
      - split: train
        path: Chart/arxivqa_en/train-*
  - config_name: arxivqa_total
    data_files:
      - split: train
        path: Chart/arxivqa_total/train-*
  - config_name: blockdiagramhandwritten
    data_files:
      - split: train
        path: Chart/blockdiagramhandwritten/train-*
  - config_name: chart2text_cauldron
    data_files:
      - split: train
        path: Chart/chart2text_cauldron/train-*
  - config_name: chartqa_cauldron_llava_format
    data_files:
      - split: train
        path: Chart/chartqa_cauldron_llava_format/train-*
  - config_name: cosyn_chart_train
    data_files:
      - split: train
        path: Chart/cosyn_chart_train/train-*
  - config_name: cosyn_diagram_train
    data_files:
      - split: train
        path: Chart/cosyn_diagram_train/train-*
  - config_name: cosyn_graphic_train
    data_files:
      - split: train
        path: Chart/cosyn_graphic_train/train-*
  - config_name: cosyn_table_train
    data_files:
      - split: train
        path: Chart/cosyn_table_train/train-*
  - config_name: diagram_image_to_text_cauldron
    data_files:
      - split: train
        path: Chart/diagram_image_to_text_cauldron/train-*
  - config_name: dvqa_cauldron_llava_format
    data_files:
      - split: train
        path: Chart/dvqa_cauldron_llava_format/train-*
  - config_name: dvqa_train_200k
    data_files:
      - split: train
        path: Chart/dvqa_train_200k/train-*
  - config_name: figureqa
    data_files:
      - split: train
        path: Chart/figureqa/train-*
  - config_name: figureqa_cauldron_llava_format
    data_files:
      - split: train
        path: Chart/figureqa_cauldron_llava_format/train-*
  - config_name: finqa
    data_files:
      - split: train
        path: Chart/finqa/train-*
  - config_name: hitab_cauldron_llava_format
    data_files:
      - split: train
        path: Chart/hitab_cauldron_llava_format/train-*
  - config_name: Chart-infographic_gpt4v
    data_files:
      - split: train
        path: Chart/infographic_gpt4v/train-*
  - config_name: Chart-infographic_vqa
    data_files:
      - split: train
        path: Chart/infographic_vqa/train-*
  - config_name: Chart-infographic_vqa_llava_format
    data_files:
      - split: train
        path: Chart/infographic_vqa_llava_format/train-*
  - config_name: lrv_chart
    data_files:
      - split: train
        path: Chart/lrv_chart/train-*
  - config_name: lrv_chart_en
    data_files:
      - split: train
        path: Chart/lrv_chart_en/train-*
  - config_name: mavis_math_function_caption_to_question_en
    data_files:
      - split: train
        path: Chart/mavis_math_function_caption_to_question_en/train-*
  - config_name: mmc_inst_qa_en
    data_files:
      - split: train
        path: Chart/mmc_inst_qa_en/train-*
  - config_name: mmtab_instruct_pretrain_en
    data_files:
      - split: train
        path: Chart/mmtab_instruct_pretrain_en/train-*
  - config_name: mmtab_instruct_sft_en
    data_files:
      - split: train
        path: Chart/mmtab_instruct_sft_en/train-*
  - config_name: multihiertt
    data_files:
      - split: train
        path: Chart/multihiertt/train-*
  - config_name: multihiertt_cauldron
    data_files:
      - split: train
        path: Chart/multihiertt_cauldron/train-*
  - config_name: multimodal_arxivqa52k_en
    data_files:
      - split: train
        path: Chart/multimodal_arxivqa52k_en/train-*
  - config_name: Chart-other
    data_files:
      - split: train
        path: Chart/other/train-*
  - config_name: plotqa
    data_files:
      - split: train
        path: Chart/plotqa/train-*
  - config_name: plotqa_v1_en
    data_files:
      - split: train
        path: Chart/plotqa_v1_en/train-*
  - config_name: robut
    data_files:
      - split: train
        path: Chart/robut/train-*
  - config_name: robut_sqa_cauldron
    data_files:
      - split: train
        path: Chart/robut_sqa_cauldron/train-*
  - config_name: robut_wikisql_cauldron
    data_files:
      - split: train
        path: Chart/robut_wikisql_cauldron/train-*
  - config_name: robut_wtq_cauldron_llava_format
    data_files:
      - split: train
        path: Chart/robut_wtq_cauldron_llava_format/train-*
  - config_name: scitsr_markdown_zh
    data_files:
      - split: train
        path: Chart/scitsr_markdown_zh/train-*
  - config_name: simchart9k_csv_en
    data_files:
      - split: train
        path: Chart/simchart9k_csv_en/train-*
  - config_name: tabmwp
    data_files:
      - split: train
        path: Chart/tabmwp/train-*
  - config_name: tabmwp_cauldron
    data_files:
      - split: train
        path: Chart/tabmwp_cauldron/train-*
  - config_name: tat_qa
    data_files:
      - split: train
        path: Chart/tat_qa/train-*
  - config_name: tinychart_train
    data_files:
      - split: train
        path: Chart/tinychart_train/train-*
  - config_name: unichart_en
    data_files:
      - split: train
        path: Chart/unichart_en/train-*
  - config_name: Chart-ureader_kg
    data_files:
      - split: train
        path: Chart/ureader_kg/train-*
  - config_name: ureader_qa
    data_files:
      - split: train
        path: Chart/ureader_qa/train-*
  - config_name: vistext_cauldron
    data_files:
      - split: train
        path: Chart/vistext_cauldron/train-*
  - config_name: Document-CoSyn
    data_files:
      - split: train
        path: Document/CoSyn/train-*
  - config_name: CoSyn_400k_document
    data_files:
      - split: train
        path: Document/CoSyn_400k_document/train-*
  - config_name: DocReason
    data_files:
      - split: train
        path: Document/DocReason/train-*
  - config_name: DocVQA
    data_files:
      - split: train
        path: Document/DocVQA/train-*
  - config_name: DoclingMatix
    data_files:
      - split: train
        path: Document/DoclingMatix/train-*
  - config_name: Docmatix
    data_files:
      - split: train
        path: Document/Docmatix/train-*
  - config_name: EATEN
    data_files:
      - split: train
        path: Document/EATEN/train-*
  - config_name: FinTabNet
    data_files:
      - split: train
        path: Document/FinTabNet/train-*
  - config_name: InfoVQA
    data_files:
      - split: train
        path: Document/InfoVQA/train-*
  - config_name: InfographicVQA
    data_files:
      - split: train
        path: Document/InfographicVQA/train-*
  - config_name: LLaVAR_GPT4
    data_files:
      - split: train
        path: Document/LLaVAR_GPT4/train-*
  - config_name: OCR-VQA
    data_files:
      - split: train
        path: Document/OCR-VQA/train-*
  - config_name: OCR-VQA-LLaVA-NeXT
    data_files:
      - split: train
        path: Document/OCR-VQA-LLaVA-NeXT/train-*
  - config_name: POIE
    data_files:
      - split: train
        path: Document/POIE/train-*
  - config_name: ScreenQA
    data_files:
      - split: train
        path: Document/ScreenQA/train-*
  - config_name: SynMultiImageQA_Doc
    data_files:
      - split: train
        path: Document/SynMultiImageQA_Doc/train-*
  - config_name: TextVQA
    data_files:
      - split: train
        path: Document/TextVQA/train-*
  - config_name: UreaderOCR
    data_files:
      - split: train
        path: Document/UreaderOCR/train-*
  - config_name: Ureader_Chart
    data_files:
      - split: train
        path: Document/Ureader_Chart/train-*
  - config_name: VisualMRC
    data_files:
      - split: train
        path: Document/VisualMRC/train-*
  - config_name: cosyn_document_train
    data_files:
      - split: train
        path: Document/cosyn_document_train/train-*
  - config_name: cosyn_nutrition_train
    data_files:
      - split: train
        path: Document/cosyn_nutrition_train/train-*
  - config_name: docmatix_single_20k_en
    data_files:
      - split: train
        path: Document/docmatix_single_20k_en/train-*
  - config_name: docreason_en
    data_files:
      - split: train
        path: Document/docreason_en/train-*
  - config_name: docvqa_en
    data_files:
      - split: train
        path: Document/docvqa_en/train-*
  - config_name: eaten_passport_zh
    data_files:
      - split: train
        path: Document/eaten_passport_zh/train-*
  - config_name: fintabnet_markdown_zh
    data_files:
      - split: train
        path: Document/fintabnet_markdown_zh/train-*
  - config_name: handwriting_forms
    data_files:
      - split: train
        path: Document/handwriting_forms/train-*
  - config_name: hw_squad
    data_files:
      - split: train
        path: Document/hw_squad/train-*
  - config_name: Document-infographic_gpt4v
    data_files:
      - split: train
        path: Document/infographic_gpt4v/train-*
  - config_name: Document-infographic_vqa
    data_files:
      - split: train
        path: Document/infographic_vqa/train-*
  - config_name: Document-infographic_vqa_llava_format
    data_files:
      - split: train
        path: Document/infographic_vqa_llava_format/train-*
  - config_name: infovqa_en
    data_files:
      - split: train
        path: Document/infovqa_en/train-*
  - config_name: llavar_gpt4_20k
    data_files:
      - split: train
        path: Document/llavar_gpt4_20k/train-*
  - config_name: olmOCR-mix-0225-books
    data_files:
      - split: train
        path: Document/olmOCR-mix-0225-books/train-*
  - config_name: olmOCR-mix-0225-documents
    data_files:
      - split: train
        path: Document/olmOCR-mix-0225-documents/train-*
  - config_name: Document-other
    data_files:
      - split: train
        path: Document/other/train-*
  - config_name: pdfvqa
    data_files:
      - split: train
        path: Document/pdfvqa/train-*
  - config_name: poie_multi_turn_en
    data_files:
      - split: train
        path: Document/poie_multi_turn_en/train-*
  - config_name: screen-qa-cleaned
    data_files:
      - split: train
        path: Document/screen-qa-cleaned/train-*
  - config_name: slidevqa
    data_files:
      - split: train
        path: Document/slidevqa/train-*
  - config_name: sujet_finance
    data_files:
      - split: train
        path: Document/sujet_finance/train-*
  - config_name: tat_dqa
    data_files:
      - split: train
        path: Document/tat_dqa/train-*
  - config_name: ureader
    data_files:
      - split: train
        path: Document/ureader/train-*
  - config_name: ureader_chart
    data_files:
      - split: train
        path: Document/ureader_chart/train-*
  - config_name: Document-ureader_kg
    data_files:
      - split: train
        path: Document/ureader_kg/train-*
  - config_name: ureader_ocr
    data_files:
      - split: train
        path: Document/ureader_ocr/train-*
  - config_name: visualmrc_cauldron
    data_files:
      - split: train
        path: Document/visualmrc_cauldron/train-*
  - config_name: vqaonbd
    data_files:
      - split: train
        path: Document/vqaonbd/train-*
  - config_name: aria_ui_desktop
    data_files:
      - split: train
        path: GUI-Grounding/aria_ui_desktop/train-*
  - config_name: aria_ui_mobile
    data_files:
      - split: train
        path: GUI-Grounding/aria_ui_mobile/train-*
  - config_name: aria_ui_web
    data_files:
      - split: train
        path: GUI-Grounding/aria_ui_web/train-*
  - config_name: groundui
    data_files:
      - split: train
        path: GUI-Grounding/groundui/train-*
  - config_name: A-OKVQA
    data_files:
      - split: train
        path: General/A-OKVQA/train-*
  - config_name: ALFWorld
    data_files:
      - split: train
        path: General/ALFWorld/train-*
  - config_name: ALLaVA
    data_files:
      - split: train
        path: General/ALLaVA/train-*
  - config_name: ALLaVA-Instruct-LAION-4V_Chinese
    data_files:
      - split: train
        path: General/ALLaVA-Instruct-LAION-4V_Chinese/train-*
  - config_name: ALLaVA-Instruct-VFLAN-4V_Chinese
    data_files:
      - split: train
        path: General/ALLaVA-Instruct-VFLAN-4V_Chinese/train-*
  - config_name: ART500K
    data_files:
      - split: train
        path: General/ART500K/train-*
  - config_name: Birds-to-Words
    data_files:
      - split: train
        path: General/Birds-to-Words/train-*
  - config_name: Cambrian
    data_files:
      - split: train
        path: General/Cambrian/train-*
  - config_name: Cambrian_Filter
    data_files:
      - split: train
        path: General/Cambrian_Filter/train-*
  - config_name: Co-Instruct
    data_files:
      - split: train
        path: General/Co-Instruct/train-*
  - config_name: DreamSim
    data_files:
      - split: train
        path: General/DreamSim/train-*
  - config_name: EST-VQA
    data_files:
      - split: train
        path: General/EST-VQA/train-*
  - config_name: GQA
    data_files:
      - split: train
        path: General/GQA/train-*
  - config_name: GQA-LLaVA-NeXT
    data_files:
      - split: train
        path: General/GQA-LLaVA-NeXT/train-*
  - config_name: Hateful_Memes
    data_files:
      - split: train
        path: General/Hateful_Memes/train-*
  - config_name: IDK
    data_files:
      - split: train
        path: General/IDK/train-*
  - config_name: General-IconQA_MathV360K
    data_files:
      - split: train
        path: General/IconQA_MathV360K/train-*
  - config_name: InternVL-SA-1B-Caption
    data_files:
      - split: train
        path: General/InternVL-SA-1B-Caption/train-*
  - config_name: KVQA
    data_files:
      - split: train
        path: General/KVQA/train-*
  - config_name: KonIQ-10k
    data_files:
      - split: train
        path: General/KonIQ-10k/train-*
  - config_name: LLaVA
    data_files:
      - split: train
        path: General/LLaVA/train-*
  - config_name: LLaVA-NeXT-Data
    data_files:
      - split: train
        path: General/LLaVA-NeXT-Data/train-*
  - config_name: LRV_Normal
    data_files:
      - split: train
        path: General/LRV_Normal/train-*
  - config_name: LVIS-InstructV4
    data_files:
      - split: train
        path: General/LVIS-InstructV4/train-*
  - config_name: MMChat-Twitter-Post
    data_files:
      - split: train
        path: General/MMChat-Twitter-Post/train-*
  - config_name: NewYorker_Caption
    data_files:
      - split: train
        path: General/NewYorker_Caption/train-*
  - config_name: Objects365
    data_files:
      - split: train
        path: General/Objects365/train-*
  - config_name: PixMo
    data_files:
      - split: train
        path: General/PixMo/train-*
  - config_name: SVIT
    data_files:
      - split: train
        path: General/SVIT/train-*
  - config_name: SVIT_core_150K
    data_files:
      - split: train
        path: General/SVIT_core_150K/train-*
  - config_name: SVIT_mix_665K
    data_files:
      - split: train
        path: General/SVIT_mix_665K/train-*
  - config_name: ScanQA
    data_files:
      - split: train
        path: General/ScanQA/train-*
  - config_name: ShareGPT4V
    data_files:
      - split: train
        path: General/ShareGPT4V/train-*
  - config_name: ShareGPT4o
    data_files:
      - split: train
        path: General/ShareGPT4o/train-*
  - config_name: VIST
    data_files:
      - split: train
        path: General/VIST/train-*
  - config_name: VSR
    data_files:
      - split: train
        path: General/VSR/train-*
  - config_name: ViQuAE
    data_files:
      - split: train
        path: General/ViQuAE/train-*
  - config_name: Vision_FLAN
    data_files:
      - split: train
        path: General/Vision_FLAN/train-*
  - config_name: Visual7W
    data_files:
      - split: train
        path: General/Visual7W/train-*
  - config_name: General-VizWiz_MathV360K
    data_files:
      - split: train
        path: General/VizWiz_MathV360K/train-*
  - config_name: WildVision
    data_files:
      - split: train
        path: General/WildVision/train-*
  - config_name: alfworldgpt
    data_files:
      - split: train
        path: General/alfworldgpt/train-*
  - config_name: alfworldgpt_45k
    data_files:
      - split: train
        path: General/alfworldgpt_45k/train-*
  - config_name: allava
    data_files:
      - split: train
        path: General/allava/train-*
  - config_name: allava-laion-500k
    data_files:
      - split: train
        path: General/allava-laion-500k/train-*
  - config_name: allava-vflan-200k
    data_files:
      - split: train
        path: General/allava-vflan-200k/train-*
  - config_name: allava_instruct_gpt4v_en
    data_files:
      - split: train
        path: General/allava_instruct_gpt4v_en/train-*
  - config_name: allava_instruct_gpt4v_zh
    data_files:
      - split: train
        path: General/allava_instruct_gpt4v_zh/train-*
  - config_name: allava_instruct_laion4v
    data_files:
      - split: train
        path: General/allava_instruct_laion4v/train-*
  - config_name: allava_instruct_vflan4v
    data_files:
      - split: train
        path: General/allava_instruct_vflan4v/train-*
  - config_name: allava_writing_gpt4o_zh
    data_files:
      - split: train
        path: General/allava_writing_gpt4o_zh/train-*
  - config_name: aokvqa_cauldron_llava_format
    data_files:
      - split: train
        path: General/aokvqa_cauldron_llava_format/train-*
  - config_name: art500k_en
    data_files:
      - split: train
        path: General/art500k_en/train-*
  - config_name: cambrain_multiturn_gpt4o_en
    data_files:
      - split: train
        path: General/cambrain_multiturn_gpt4o_en/train-*
  - config_name: cambrian_filtered
    data_files:
      - split: train
        path: General/cambrian_filtered/train-*
  - config_name: chinesememe
    data_files:
      - split: train
        path: General/chinesememe/train-*
  - config_name: coco_colors
    data_files:
      - split: train
        path: General/coco_colors/train-*
  - config_name: coco_negative_gpt4o_en
    data_files:
      - split: train
        path: General/coco_negative_gpt4o_en/train-*
  - config_name: drivelm
    data_files:
      - split: train
        path: General/drivelm/train-*
  - config_name: estvqa_en
    data_files:
      - split: train
        path: General/estvqa_en/train-*
  - config_name: google_landmarks
    data_files:
      - split: train
        path: General/google_landmarks/train-*
  - config_name: gpt-4o
    data_files:
      - split: train
        path: General/gpt-4o/train-*
  - config_name: gpt4gen_rd_boxcot_en
    data_files:
      - split: train
        path: General/gpt4gen_rd_boxcot_en/train-*
  - config_name: gpt77k
    data_files:
      - split: train
        path: General/gpt77k/train-*
  - config_name: gqa_en
    data_files:
      - split: train
        path: General/gqa_en/train-*
  - config_name: hateful_memes_cauldron_llava_format
    data_files:
      - split: train
        path: General/hateful_memes_cauldron_llava_format/train-*
  - config_name: idefics375k
    data_files:
      - split: train
        path: General/idefics375k/train-*
  - config_name: idk_11k
    data_files:
      - split: train
        path: General/idk_11k/train-*
  - config_name: image_textualization(filtered)
    data_files:
      - split: train
        path: General/image_textualization(filtered)/train-*
  - config_name: inat_train2018_merge_en
    data_files:
      - split: train
        path: General/inat_train2018_merge_en/train-*
  - config_name: indoor_qa
    data_files:
      - split: train
        path: General/indoor_qa/train-*
  - config_name: internvl_sa1b_caption_gpt4o_review_en
    data_files:
      - split: train
        path: General/internvl_sa1b_caption_gpt4o_review_en/train-*
  - config_name: koniq10k_en
    data_files:
      - split: train
        path: General/koniq10k_en/train-*
  - config_name: kvqa_en
    data_files:
      - split: train
        path: General/kvqa_en/train-*
  - config_name: laion_gpt4v
    data_files:
      - split: train
        path: General/laion_gpt4v/train-*
  - config_name: llava_instruct_300k
    data_files:
      - split: train
        path: General/llava_instruct_300k/train-*
  - config_name: lnqa
    data_files:
      - split: train
        path: General/lnqa/train-*
  - config_name: localized_narratives
    data_files:
      - split: train
        path: General/localized_narratives/train-*
  - config_name: lrv_normal_filtered
    data_files:
      - split: train
        path: General/lrv_normal_filtered/train-*
  - config_name: lvis_instruct4v
    data_files:
      - split: train
        path: General/lvis_instruct4v/train-*
  - config_name: lvis_instruct4v_220k
    data_files:
      - split: train
        path: General/lvis_instruct4v_220k/train-*
  - config_name: memotion
    data_files:
      - split: train
        path: General/memotion/train-*
  - config_name: mmevol
    data_files:
      - split: train
        path: General/mmevol/train-*
  - config_name: mmsoc_memotion
    data_files:
      - split: train
        path: General/mmsoc_memotion/train-*
  - config_name: newyorker_caption_en
    data_files:
      - split: train
        path: General/newyorker_caption_en/train-*
  - config_name: nlvr2
    data_files:
      - split: train
        path: General/nlvr2/train-*
  - config_name: objects365_qa_en
    data_files:
      - split: train
        path: General/objects365_qa_en/train-*
  - config_name: okvqa_en
    data_files:
      - split: train
        path: General/okvqa_en/train-*
  - config_name: pixmo-cap-qa
    data_files:
      - split: train
        path: General/pixmo-cap-qa/train-*
  - config_name: pixmo_ask_train
    data_files:
      - split: train
        path: General/pixmo_ask_train/train-*
  - config_name: pixmo_point_train
    data_files:
      - split: train
        path: General/pixmo_point_train/train-*
  - config_name: scienceqa_nona_context
    data_files:
      - split: train
        path: General/scienceqa_nona_context/train-*
  - config_name: screen2words
    data_files:
      - split: train
        path: General/screen2words/train-*
  - config_name: screenqa
    data_files:
      - split: train
        path: General/screenqa/train-*
  - config_name: sharegpt4o
    data_files:
      - split: train
        path: General/sharegpt4o/train-*
  - config_name: sharegpt4v
    data_files:
      - split: train
        path: General/sharegpt4v/train-*
  - config_name: sharegpt4v_coco
    data_files:
      - split: train
        path: General/sharegpt4v_coco/train-*
  - config_name: sharegpt4v_instruct_gpt4-vision_cap100k
    data_files:
      - split: train
        path: General/sharegpt4v_instruct_gpt4-vision_cap100k/train-*
  - config_name: sharegpt4v_knowledge
    data_files:
      - split: train
        path: General/sharegpt4v_knowledge/train-*
  - config_name: sharegpt4v_sam
    data_files:
      - split: train
        path: General/sharegpt4v_sam/train-*
  - config_name: sketchyvqa
    data_files:
      - split: train
        path: General/sketchyvqa/train-*
  - config_name: spark
    data_files:
      - split: train
        path: General/spark/train-*
  - config_name: spatialsense
    data_files:
      - split: train
        path: General/spatialsense/train-*
  - config_name: spot_the_diff
    data_files:
      - split: train
        path: General/spot_the_diff/train-*
  - config_name: General-st_vqa_cauldron_llava_format
    data_files:
      - split: train
        path: General/st_vqa_cauldron_llava_format/train-*
  - config_name: stvqa_en
    data_files:
      - split: train
        path: General/stvqa_en/train-*
  - config_name: textvqa_en
    data_files:
      - split: train
        path: General/textvqa_en/train-*
  - config_name: tqa_en
    data_files:
      - split: train
        path: General/tqa_en/train-*
  - config_name: unknown
    data_files:
      - split: train
        path: General/unknown/train-*
  - config_name: vflan_cleaned
    data_files:
      - split: train
        path: General/vflan_cleaned/train-*
  - config_name: viquae
    data_files:
      - split: train
        path: General/viquae/train-*
  - config_name: viquae_en
    data_files:
      - split: train
        path: General/viquae_en/train-*
  - config_name: vision_flan(filtered)
    data_files:
      - split: train
        path: General/vision_flan(filtered)/train-*
  - config_name: vision_flan_filtered
    data_files:
      - split: train
        path: General/vision_flan_filtered/train-*
  - config_name: visual7w_cauldron_llava_format
    data_files:
      - split: train
        path: General/visual7w_cauldron_llava_format/train-*
  - config_name: visual7w_en
    data_files:
      - split: train
        path: General/visual7w_en/train-*
  - config_name: General-vqarad_cauldron_llava_format
    data_files:
      - split: train
        path: General/vqarad_cauldron_llava_format/train-*
  - config_name: vsr_cauldron_llava_format
    data_files:
      - split: train
        path: General/vsr_cauldron_llava_format/train-*
  - config_name: vsr_en
    data_files:
      - split: train
        path: General/vsr_en/train-*
  - config_name: wildvision_gpt4o_en
    data_files:
      - split: train
        path: General/wildvision_gpt4o_en/train-*
  - config_name: wildvision_gpt4v_to_gpt4o_en
    data_files:
      - split: train
        path: General/wildvision_gpt4v_to_gpt4o_en/train-*
  - config_name: yesbut
    data_files:
      - split: train
        path: General/yesbut/train-*
  - config_name: CLEVR
    data_files:
      - split: train
        path: Grounding/CLEVR/train-*
  - config_name: CLEVR-Math_MathV360K
    data_files:
      - split: train
        path: Grounding/CLEVR-Math_MathV360K/train-*
  - config_name: CLEVR_Math
    data_files:
      - split: train
        path: Grounding/CLEVR_Math/train-*
  - config_name: IconQA
    data_files:
      - split: train
        path: Grounding/IconQA/train-*
  - config_name: Grounding-IconQA_MathV360K
    data_files:
      - split: train
        path: Grounding/IconQA_MathV360K/train-*
  - config_name: MathV360K_VQA-AS
    data_files:
      - split: train
        path: Grounding/MathV360K_VQA-AS/train-*
  - config_name: MovieNet
    data_files:
      - split: train
        path: Grounding/MovieNet/train-*
  - config_name: Super-CLEVR
    data_files:
      - split: train
        path: Grounding/Super-CLEVR/train-*
  - config_name: Super-CLEVR_MathV360K
    data_files:
      - split: train
        path: Grounding/Super-CLEVR_MathV360K/train-*
  - config_name: TQA
    data_files:
      - split: train
        path: Grounding/TQA/train-*
  - config_name: TallyQA
    data_files:
      - split: train
        path: Grounding/TallyQA/train-*
  - config_name: VisualGenome-LLaVA-NeXT
    data_files:
      - split: train
        path: Grounding/VisualGenome-LLaVA-NeXT/train-*
  - config_name: clevr_700k
    data_files:
      - split: train
        path: Grounding/clevr_700k/train-*
  - config_name: coco_rem_en_zh
    data_files:
      - split: train
        path: Grounding/coco_rem_en_zh/train-*
  - config_name: ctw
    data_files:
      - split: train
        path: Grounding/ctw/train-*
  - config_name: fsc147_train_en
    data_files:
      - split: train
        path: Grounding/fsc147_train_en/train-*
  - config_name: iconqa
    data_files:
      - split: train
        path: Grounding/iconqa/train-*
  - config_name: iconqa_cauldron_llava_format
    data_files:
      - split: train
        path: Grounding/iconqa_cauldron_llava_format/train-*
  - config_name: movienet_en
    data_files:
      - split: train
        path: Grounding/movienet_en/train-*
  - config_name: objects365_qa
    data_files:
      - split: train
        path: Grounding/objects365_qa/train-*
  - config_name: oodvqa
    data_files:
      - split: train
        path: Grounding/oodvqa/train-*
  - config_name: Grounding-other
    data_files:
      - split: train
        path: Grounding/other/train-*
  - config_name: tallyqa
    data_files:
      - split: train
        path: Grounding/tallyqa/train-*
  - config_name: tallyqa_cauldron_llava_format
    data_files:
      - split: train
        path: Grounding/tallyqa_cauldron_llava_format/train-*
  - config_name: tallyqa_coco_en
    data_files:
      - split: train
        path: Grounding/tallyqa_coco_en/train-*
  - config_name: tallyqa_vg_en
    data_files:
      - split: train
        path: Grounding/tallyqa_vg_en/train-*
  - config_name: tqa_cauldron_llava_format
    data_files:
      - split: train
        path: Grounding/tqa_cauldron_llava_format/train-*
  - config_name: ArXiv_OCR
    data_files:
      - split: train
        path: OCR/ArXiv_OCR/train-*
  - config_name: CTW
    data_files:
      - split: train
        path: OCR/CTW/train-*
  - config_name: Chrome-Writing
    data_files:
      - split: train
        path: OCR/Chrome-Writing/train-*
  - config_name: HME
    data_files:
      - split: train
        path: OCR/HME/train-*
  - config_name: IAM
    data_files:
      - split: train
        path: OCR/IAM/train-*
  - config_name: ICDAR-LSVT-zh
    data_files:
      - split: train
        path: OCR/ICDAR-LSVT-zh/train-*
  - config_name: ICDAR2017
    data_files:
      - split: train
        path: OCR/ICDAR2017/train-*
  - config_name: ICDAR2019
    data_files:
      - split: train
        path: OCR/ICDAR2019/train-*
  - config_name: IIIT_5K
    data_files:
      - split: train
        path: OCR/IIIT_5K/train-*
  - config_name: K12_Printing
    data_files:
      - split: train
        path: OCR/K12_Printing/train-*
  - config_name: MTWI_zh
    data_files:
      - split: train
        path: OCR/MTWI_zh/train-*
  - config_name: Orand-Car-A
    data_files:
      - split: train
        path: OCR/Orand-Car-A/train-*
  - config_name: ReCTs
    data_files:
      - split: train
        path: OCR/ReCTs/train-*
  - config_name: Rendered_Text
    data_files:
      - split: train
        path: OCR/Rendered_Text/train-*
  - config_name: Sroie
    data_files:
      - split: train
        path: OCR/Sroie/train-*
  - config_name: TextOCR
    data_files:
      - split: train
        path: OCR/TextOCR/train-*
  - config_name: VCR-Wiki
    data_files:
      - split: train
        path: OCR/VCR-Wiki/train-*
  - config_name: arxiv_ocr_train
    data_files:
      - split: train
        path: OCR/arxiv_ocr_train/train-*
  - config_name: bentham
    data_files:
      - split: train
        path: OCR/bentham/train-*
  - config_name: captcha
    data_files:
      - split: train
        path: OCR/captcha/train-*
  - config_name: ctw_zh
    data_files:
      - split: train
        path: OCR/ctw_zh/train-*
  - config_name: hme100k
    data_files:
      - split: train
        path: OCR/hme100k/train-*
  - config_name: icdar2017_rctw17_zh
    data_files:
      - split: train
        path: OCR/icdar2017_rctw17_zh/train-*
  - config_name: icdar2017_rects_zh
    data_files:
      - split: train
        path: OCR/icdar2017_rects_zh/train-*
  - config_name: icdar2019_art_task1_3_zh
    data_files:
      - split: train
        path: OCR/icdar2019_art_task1_3_zh/train-*
  - config_name: icdar_lsvt_zh
    data_files:
      - split: train
        path: OCR/icdar_lsvt_zh/train-*
  - config_name: imgur5k
    data_files:
      - split: train
        path: OCR/imgur5k/train-*
  - config_name: invoices_receipts
    data_files:
      - split: train
        path: OCR/invoices_receipts/train-*
  - config_name: k12_printing
    data_files:
      - split: train
        path: OCR/k12_printing/train-*
  - config_name: latex_handwritten
    data_files:
      - split: train
        path: OCR/latex_handwritten/train-*
  - config_name: latexformulas
    data_files:
      - split: train
        path: OCR/latexformulas/train-*
  - config_name: mathwriting-google
    data_files:
      - split: train
        path: OCR/mathwriting-google/train-*
  - config_name: mtwi_zh
    data_files:
      - split: train
        path: OCR/mtwi_zh/train-*
  - config_name: ocrvqa
    data_files:
      - split: train
        path: OCR/ocrvqa/train-*
  - config_name: sroie
    data_files:
      - split: train
        path: OCR/sroie/train-*
  - config_name: svrd
    data_files:
      - split: train
        path: OCR/svrd/train-*
  - config_name: synth_ocr_fanti_regular_zh
    data_files:
      - split: train
        path: OCR/synth_ocr_fanti_regular_zh/train-*
  - config_name: synthdog
    data_files:
      - split: train
        path: OCR/synthdog/train-*
  - config_name: tal_ocr_eng
    data_files:
      - split: train
        path: OCR/tal_ocr_eng/train-*
  - config_name: textvqa_train_format
    data_files:
      - split: train
        path: OCR/textvqa_train_format/train-*
  - config_name: vcr_wiki_en_easy
    data_files:
      - split: train
        path: OCR/vcr_wiki_en_easy/train-*
  - config_name: vcr_wiki_en_hard
    data_files:
      - split: train
        path: OCR/vcr_wiki_en_hard/train-*
  - config_name: vcr_wiki_zh_easy
    data_files:
      - split: train
        path: OCR/vcr_wiki_zh_easy/train-*
  - config_name: vcr_wiki_zh_hard
    data_files:
      - split: train
        path: OCR/vcr_wiki_zh_hard/train-*
  - config_name: wordart
    data_files:
      - split: train
        path: OCR/wordart/train-*
  - config_name: AI2D
    data_files:
      - split: train
        path: STEM/AI2D/train-*
  - config_name: CMM-Math
    data_files:
      - split: train
        path: STEM/CMM-Math/train-*
  - config_name: STEM-CoSyn
    data_files:
      - split: train
        path: STEM/CoSyn/train-*
  - config_name: STEM-CoSyn_400k
    data_files:
      - split: train
        path: STEM/CoSyn_400k/train-*
  - config_name: GEOS_MathV360K
    data_files:
      - split: train
        path: STEM/GEOS_MathV360K/train-*
  - config_name: Geo170K
    data_files:
      - split: train
        path: STEM/Geo170K/train-*
  - config_name: GeoQA+
    data_files:
      - split: train
        path: STEM/GeoQA+/train-*
  - config_name: GeoQA+_MathV360K
    data_files:
      - split: train
        path: STEM/GeoQA+_MathV360K/train-*
  - config_name: GeomVerse
    data_files:
      - split: train
        path: STEM/GeomVerse/train-*
  - config_name: Geometry3K
    data_files:
      - split: train
        path: STEM/Geometry3K/train-*
  - config_name: Geometry3K_MathV360K
    data_files:
      - split: train
        path: STEM/Geometry3K_MathV360K/train-*
  - config_name: STEM-IconQA_MathV360K
    data_files:
      - split: train
        path: STEM/IconQA_MathV360K/train-*
  - config_name: InterGPS
    data_files:
      - split: train
        path: STEM/InterGPS/train-*
  - config_name: MAVIS-Metagen
    data_files:
      - split: train
        path: STEM/MAVIS-Metagen/train-*
  - config_name: MMChem
    data_files:
      - split: train
        path: STEM/MMChem/train-*
  - config_name: MapQA
    data_files:
      - split: train
        path: STEM/MapQA/train-*
  - config_name: STEM-MapQA_MathV360K
    data_files:
      - split: train
        path: STEM/MapQA_MathV360K/train-*
  - config_name: MathV360K_TQA
    data_files:
      - split: train
        path: STEM/MathV360K_TQA/train-*
  - config_name: MathV360K_VQA-RAD
    data_files:
      - split: train
        path: STEM/MathV360K_VQA-RAD/train-*
  - config_name: MetaMathQA
    data_files:
      - split: train
        path: STEM/MetaMathQA/train-*
  - config_name: PMC-VQA
    data_files:
      - split: train
        path: STEM/PMC-VQA/train-*
  - config_name: PMC-VQA_MathV360K
    data_files:
      - split: train
        path: STEM/PMC-VQA_MathV360K/train-*
  - config_name: PathVQA
    data_files:
      - split: train
        path: STEM/PathVQA/train-*
  - config_name: RAVEN
    data_files:
      - split: train
        path: STEM/RAVEN/train-*
  - config_name: ScienceQA
    data_files:
      - split: train
        path: STEM/ScienceQA/train-*
  - config_name: STEM-SynMultiImageQA
    data_files:
      - split: train
        path: STEM/SynMultiImageQA/train-*
  - config_name: SynthCodeNet
    data_files:
      - split: train
        path: STEM/SynthCodeNet/train-*
  - config_name: SynthFormulaNet
    data_files:
      - split: train
        path: STEM/SynthFormulaNet/train-*
  - config_name: UniGeo
    data_files:
      - split: train
        path: STEM/UniGeo/train-*
  - config_name: UniGeo_MathV360K
    data_files:
      - split: train
        path: STEM/UniGeo_MathV360K/train-*
  - config_name: VQA-RAD
    data_files:
      - split: train
        path: STEM/VQA-RAD/train-*
  - config_name: VisualWebInstruct
    data_files:
      - split: train
        path: STEM/VisualWebInstruct/train-*
  - config_name: VisualWebInstruct_filtered
    data_files:
      - split: train
        path: STEM/VisualWebInstruct_filtered/train-*
  - config_name: VizWiz
    data_files:
      - split: train
        path: STEM/VizWiz/train-*
  - config_name: STEM-VizWiz_MathV360K
    data_files:
      - split: train
        path: STEM/VizWiz_MathV360K/train-*
  - config_name: WebSight
    data_files:
      - split: train
        path: STEM/WebSight/train-*
  - config_name: ai2d_cap_gpt4o_en
    data_files:
      - split: train
        path: STEM/ai2d_cap_gpt4o_en/train-*
  - config_name: ai2d_cauldron_llava_format
    data_files:
      - split: train
        path: STEM/ai2d_cauldron_llava_format/train-*
  - config_name: ai2d_cot_gpt4o_en
    data_files:
      - split: train
        path: STEM/ai2d_cot_gpt4o_en/train-*
  - config_name: ai2d_en
    data_files:
      - split: train
        path: STEM/ai2d_en/train-*
  - config_name: ai2d_gpt4v
    data_files:
      - split: train
        path: STEM/ai2d_gpt4v/train-*
  - config_name: ai2d_internvl
    data_files:
      - split: train
        path: STEM/ai2d_internvl/train-*
  - config_name: cmm_math_cot_zh
    data_files:
      - split: train
        path: STEM/cmm_math_cot_zh/train-*
  - config_name: cosyn_chemical_train
    data_files:
      - split: train
        path: STEM/cosyn_chemical_train/train-*
  - config_name: cosyn_circuit_train
    data_files:
      - split: train
        path: STEM/cosyn_circuit_train/train-*
  - config_name: cosyn_math_train
    data_files:
      - split: train
        path: STEM/cosyn_math_train/train-*
  - config_name: cosyn_music_train
    data_files:
      - split: train
        path: STEM/cosyn_music_train/train-*
  - config_name: datik
    data_files:
      - split: train
        path: STEM/datik/train-*
  - config_name: datikz
    data_files:
      - split: train
        path: STEM/datikz/train-*
  - config_name: geo170k
    data_files:
      - split: train
        path: STEM/geo170k/train-*
  - config_name: geo170k_align
    data_files:
      - split: train
        path: STEM/geo170k_align/train-*
  - config_name: geo170k_qa
    data_files:
      - split: train
        path: STEM/geo170k_qa/train-*
  - config_name: geo3k
    data_files:
      - split: train
        path: STEM/geo3k/train-*
  - config_name: geomverse
    data_files:
      - split: train
        path: STEM/geomverse/train-*
  - config_name: geomverse_cauldron
    data_files:
      - split: train
        path: STEM/geomverse_cauldron/train-*
  - config_name: geoqa_plus_en
    data_files:
      - split: train
        path: STEM/geoqa_plus_en/train-*
  - config_name: intergps_cauldron_llava_format
    data_files:
      - split: train
        path: STEM/intergps_cauldron_llava_format/train-*
  - config_name: mapqa_cauldron_llava_format
    data_files:
      - split: train
        path: STEM/mapqa_cauldron_llava_format/train-*
  - config_name: mathvision_3k
    data_files:
      - split: train
        path: STEM/mathvision_3k/train-*
  - config_name: mavis_math
    data_files:
      - split: train
        path: STEM/mavis_math/train-*
  - config_name: mavis_math_geometry_caption_to_question_en
    data_files:
      - split: train
        path: STEM/mavis_math_geometry_caption_to_question_en/train-*
  - config_name: mavis_math_metagen
    data_files:
      - split: train
        path: STEM/mavis_math_metagen/train-*
  - config_name: mavis_math_rule_geo
    data_files:
      - split: train
        path: STEM/mavis_math_rule_geo/train-*
  - config_name: metamathqa_en
    data_files:
      - split: train
        path: STEM/metamathqa_en/train-*
  - config_name: metamathqa_ocr_en
    data_files:
      - split: train
        path: STEM/metamathqa_ocr_en/train-*
  - config_name: mm_chem_exam_single_image_zh
    data_files:
      - split: train
        path: STEM/mm_chem_exam_single_image_zh/train-*
  - config_name: mm_chem_ocr_smiles_gpt4o_zh
    data_files:
      - split: train
        path: STEM/mm_chem_ocr_smiles_gpt4o_zh/train-*
  - config_name: mmc_instruct
    data_files:
      - split: train
        path: STEM/mmc_instruct/train-*
  - config_name: STEM-other
    data_files:
      - split: train
        path: STEM/other/train-*
  - config_name: pathvqa
    data_files:
      - split: train
        path: STEM/pathvqa/train-*
  - config_name: pathvqa_32k
    data_files:
      - split: train
        path: STEM/pathvqa_32k/train-*
  - config_name: pmc_vqa(mathv360k)
    data_files:
      - split: train
        path: STEM/pmc_vqa(mathv360k)/train-*
  - config_name: pmc_vqa_en
    data_files:
      - split: train
        path: STEM/pmc_vqa_en/train-*
  - config_name: raven
    data_files:
      - split: train
        path: STEM/raven/train-*
  - config_name: raven_cauldron
    data_files:
      - split: train
        path: STEM/raven_cauldron/train-*
  - config_name: scienceqa_cauldron_llava_format
    data_files:
      - split: train
        path: STEM/scienceqa_cauldron_llava_format/train-*
  - config_name: scienceqa_choice_aug_en
    data_files:
      - split: train
        path: STEM/scienceqa_choice_aug_en/train-*
  - config_name: study_com_en
    data_files:
      - split: train
        path: STEM/study_com_en/train-*
  - config_name: visualwebinstruct(filtered)
    data_files:
      - split: train
        path: STEM/visualwebinstruct(filtered)/train-*
  - config_name: visualwebinstruct_si
    data_files:
      - split: train
        path: STEM/visualwebinstruct_si/train-*
  - config_name: STEM-vqarad_cauldron_llava_format
    data_files:
      - split: train
        path: STEM/vqarad_cauldron_llava_format/train-*
  - config_name: websight
    data_files:
      - split: train
        path: STEM/websight/train-*
  - config_name: websight_cauldron
    data_files:
      - split: train
        path: STEM/websight_cauldron/train-*

Honey-Data-V2

[🏠 Homepage] [📖 arXiv (conference version)] [🤗 Bee models & datasets] [🍯 Honey-Data-15M] [💻 Code]

Honey-Data-V2 is the corpus of the journal version of Bee, Bee: A High-Quality Corpus and Full-Stack Suite for End-to-End Training of Fully Open MLLMs, accepted to Computational Visual Media. The conference version (ICLR 2026) released Honey-Data-15M and the Bee-8B models trained on it. The journal version scales the corpus up by reapplying the same curation pipeline, HoneyPipe, at a substantially larger scale.

Honey-Data-V2 is a multimodal supervised fine-tuning corpus of 44.3M samples comprising 100.3M QA pairs. It is stored as 19,707,852 image groups carrying 44,295,078 conversations, spread over 8 task categories and 420 subsets (5.81 TB). Long chain-of-thought covers 17.9% of the samples (7.9M, against 2.9M in Honey-Data-15M).

The released Bee-8B checkpoints, and all results reported in the paper, were trained on Honey-Data-15M. Evaluating models trained on Honey-Data-V2 is left for future work.

What is new compared with Honey-Data-15M

Honey-Data-V2 is not built from scratch. It extends Honey-Data-15M in three ways, all executed with the existing operators of DataStudio:

• Re-curation of the existing pool. Acting on community defect reports, samples whose final turn carries no answer were removed, and 492.6K grounding instructions were rewritten to state their coordinate convention up front. 1.8M machine-recaptioning samples were dropped, because they dominated by volume while adding little instruction-following signal. Image placeholders were canonicalized, and degenerate generations (unpaired reasoning tags, repetition collapse, placeholder leakage) were rejected.
• Multi-model rollout. On top of the original annotation pass, two further independent rollouts answered the same instructions with an upgraded model stack. Every verified answer is kept, so an instruction carries one, two or three independently produced answers. Over the overlapping subset, two answers to the same instruction agree verbatim in only 0.00%–0.89% of cases.
• New source incorporation. The FineVision datasets absent from Honey-Data-15M, together with Molmo2-SynMultiImageQA, ChartVerse and Aria-UI, were passed through the same pipeline rather than concatenated as-is. Sources whose answers are already determinate, such as OCR transcription and box-level grounding, were filtered but not rewritten.

lineage	samples	long CoT
re-curated Honey-Data-15M pool (rollout passes 1 / 2 / 3: 9.6M / 12.8M / 7.0M)	29.4M	15.5%
newly incorporated sources	14.9M	22.6%
Honey-Data-V2 total	44.3M	17.9%

role	Honey-Data-15M	Honey-Data-V2
short CoT	Qwen2.5-VL-72B/32B	Qwen3-VL-235B-A22B-Instruct, Kimi-K2.5
long CoT	proprietary MLLMs	Qwen3-VL-235B-A22B-Thinking, Doubao-Thinking, Kimi-K2.5
verification	Qwen2.5-VL-72B	Qwen3-235B-A22B-Instruct
GUI grounding	–	MAI-UI-8B

One record is one image group

Unlike the usual one-row-per-conversation layout, a record here collects every conversation that refers to the same image. The image bytes are therefore stored once, no matter how many dialogues discuss it. Across the corpus this is the difference between 19,708,047 stored images and the 44,295,078 conversations that use them.

from datasets import load_dataset, get_dataset_config_names

subsets = get_dataset_config_names("HoneyDataV2/Honey-Data-V2")
ds = load_dataset("HoneyDataV2/Honey-Data-V2", subsets[0], split="train")

rec = ds[0]
rec["images"][0]                    # a PIL image
len(rec["data"])                    # how many dialogues share it
rec["data"][0]["messages"]          # [{'from': 'human', 'value': ...}, {'from': 'gpt', 'value': ...}]
rec["data"][0]["provenance"]        # which model wrote and which verified this answer

Image encoding

Images are stored as lossless WebP. Where the source was already JPEG, or where WebP came out larger than the original, the original bytes are kept untouched. WebP holds the same pixels as the PNGs it replaces in roughly 60% of the bytes. Decoding is transparent: datasets returns PIL images either way.

For RGBA images the alpha channel is preserved exactly and RGB is preserved wherever alpha > 0. PNG stores arbitrary colour under fully transparent pixels and WebP does not carry it over; no reader can observe those values.

Record fields

column	type	meaning
id	string	identifier of the image group
images	list of image	the images, decoded to PIL by datasets
n_images	int32	how many images the group holds
image_sizes	list of [width, height]	pixel dimensions, one pair per image
image_phashes	list of string	perceptual hash, one per image
category	string	task category; matches the directory
source_subset	string	subset; matches the directory
categories	list of string	every category label the upstream corpora gave this group
sources	list of string	every upstream corpus that contributed a dialogue
source_subsets	list of string	every upstream subset name
data	list of struct	the dialogues, below

Each element of data:

field	meaning
id, ori_id	identifiers of this dialogue and of the record it came from
messages	the dialogue as from / value pairs, human and gpt
n_turns	number of turns
cot_level	short or long chain-of-thought
original_answers	the answers before rewriting
provenance	which model performed each step
source, subset, source_subset, category	where this dialogue came from

provenance names the model used at each stage, so a subset of the corpus can be selected by annotator: question_image_consist (the image-question consistency check), answer_rewrite, reasoning_rewrite, and answer_consist (verification of the rewritten answer). A value of null means that stage did not apply — determinate tasks such as OCR transcription and box-level grounding are filtered but never rewritten.

Composition

category	subsets	image groups	conversations	size
Caption	13	1,413,122	2,011,253	1.21 TB
Chart	77	4,665,363	11,036,064	0.87 TB
Document	49	2,271,245	4,766,060	0.67 TB
GUI-Grounding	4	500,916	2,803,848	0.12 TB
General	122	4,298,102	12,952,088	1.75 TB
Grounding	27	1,979,196	3,435,355	0.75 TB
OCR	45	2,564,793	3,625,483	0.37 TB
STEM	83	2,015,115	3,664,927	0.07 TB
total	420	19,707,852	44,295,078	5.81 TB

A subset name that appears in more than one category is qualified with its category, for example Chart-CoSyn and Document-CoSyn.

Provenance of the answers

stage	models
answer rewriting	Qwen3-VL-235B-A22B-Instruct, Kimi-K2.5, MAI-UI-8B
reasoning rewriting	Qwen3-VL-235B-A22B-Thinking, Kimi-K2.5, Doubao-Thinking
image-question consistency	Qwen3-VL-30B-A3B-Instruct
answer verification	Qwen3-235B-A22B-Instruct

sources records which annotation pass a dialogue belongs to: Honey-Data-15M-Rollout1 is the re-curated original pool, Rollout2 and Rollout3 are the two further independent passes over the same instructions. Dialogues taken from newly incorporated community corpora carry that corpus's name instead.

Licensing information

Honey-Data-V2 aggregates many publicly available datasets, each governed by its own licence. Anyone using this corpus must observe the licence attached to each constituent dataset. To the extent we hold any rights in the curation and the generated answers, those are released under CC-BY-4.0.

Citation

If you use Honey-Data-V2, please cite the journal version:

@article{zhang2026beecvm,
  title   = {Bee: A High-Quality Corpus and Full-Stack Suite for End-to-End Training of Fully Open MLLMs},
  author  = {Yi Zhang and Bolin Ni and Xin-Sheng Chen and Heng-Rui Zhang and Yongming Rao and
             Houwen Peng and Qinglin Lu and Han Hu and Meng-Hao Guo and Shi-Min Hu},
  journal = {Computational Visual Media},
  year    = {2026},
  note    = {to appear},
}

The conference version, which introduced Honey-Data-15M and Bee-8B:

@inproceedings{zhang2026bee,
  title     = {Bee: A High-Quality Corpus and Full-Stack Suite to Unlock Advanced Fully Open MLLMs},
  author    = {Yi Zhang and Bolin Ni and Xin-Sheng Chen and Heng-Rui Zhang and Yongming Rao and
               Houwen Peng and Qinglin Lu and Han Hu and Meng-Hao Guo and Shi-Min Hu},
  booktitle = {The Fourteenth International Conference on Learning Representations},
  year      = {2026},
}