MMCR / mmcr.jsonld
zhangluning's picture
Upload mmcr.jsonld
d8aa1b1 verified
{
"@context": "https://huggingface.co/datasets/croissant/context.jsonld",
"@type": "Dataset",
"name": "MMCR",
"description": "A benchmark dataset for evaluating multimodal compositional reasoning.",
"license": "https://creativecommons.org/licenses/by/4.0/",
"creator": {
"@type": "Person",
"name": "Luning Zhang"
},
"distribution": {
"@type": "DataDownload",
"contentUrl": "https://www.kaggle.com/datasets/zhangluning1217/mmcr-benchmark/download",
"encodingFormat": "application/zip"
},
"keywords": ["multimodal", "reasoning", "benchmark", "image-text", "compositionality"],
"variableMeasured": [
{
"@type": "PropertyValue",
"name": "image",
"description": "The image input involved in the reasoning process.",
"valueType": "string"
},
{
"@type": "PropertyValue",
"name": "question",
"description": "A natural language question about the image.",
"valueType": "string"
},
{
"@type": "PropertyValue",
"name": "answer",
"description": "The correct answer based on the image and question.",
"valueType": "string"
}
]
}