EDGEwww25/EDGE-Model
Updated • 3 • 2
id stringclasses 26
values | messages listlengths 2 36 | images sequencelengths 1 1 |
|---|---|---|
icon_mixed_cropped;icon_grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_icon_mixed/images/20901.png</img>\nFor the icons I describe in this page capture, can you predict their positions (with point (x, y))?\nicon representing PPTX file type",
"role": "user"
},
{
"content": "(0.6808, 0.8724)",
"role": "assista... | [
"../../datasets/anno/cropped_icon_mixed/images/20901.png"
] |
accessibility;image_alt | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/205264_btm.png</img>\nGiven this screenshot of a webpage, I specify locations of some images; please describe these images there (with bbox (x1, x2, y1, y2)).\n(0.027, 0.0, 0.3664, 0.103)",
"role": "user"
},
{
"content": "Ne... | [
"../../datasets/anno/annotated/fineweb-edu/raw/205264_btm.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000507580.jpg</img>\nWhat safety concerns might arise in this situation?",
"role": "user"
},
{
"content": "In the image, two horses are standing in the road, and two peo... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000507580.jpg"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/gqa/images/2360267.jpg</img>\nWhich kind of food is not green, the beans or the vegetables?",
"role": "user"
},
{
"content": "vegetables",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/gqa/images/2360267.jpg"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/5855.jpg</img>\nIn this image, to open item below bt remote pc icon, where should I click on (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.2236, 0.2012, 0.4313, 0.3566)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/5855.jpg"
] |
som_cropped;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_general/93827.png</img>\nWhat is being framed in the screen? What's the content of it? What's the position of it (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "BLOODY SCOTLAND (0.3824, 0.3069, 0.5381, 0.3413)",
... | [
"../../datasets/anno/cropped_som/images_general/93827.png"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/DUE_Benchmark/DocVQA/pngs/npjh0227_4.png</img>\nWhich generation of his family does Mr. Bunker belong to in the sugar industry",
"role": "user"
},
{
"content": "fourth",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/DUE_Benchmark/DocVQA/pngs/npjh0227_4.png"
] |
rico_tasks;widget-grounding | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/5643.jpg</img>\nOn this screen, where can I find the feature that allows me to \"play bottom video\" (with point (x, y))?",
"role": "user"
},
{
"content": "(0.5, 0.8762)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/5643.jpg"
] |
basic_cropped;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/006009_mid.png</img>\nFrom this screenshot, I'll give coordinates; can you describe the text of the elements at these points (with point (x, y))?\n(0.14, 0.055)",
"role": "user"
},
{
"content": "Lascelve Graham | Sport... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/006009_mid.png"
] |
icon_mixed;icon_grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/icon_mixed/images/25127.png</img>\nFrom the given webpage screenshot, I need you to identify the locations of described float icons (with point (x, y)).\nicon of a desk with a monitor",
"role": "user"
},
{
"content": "(0.3789, 0.2018)",
"role": "... | [
"../../datasets/anno/icon_mixed/images/25127.png"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000034938.jpg</img>\nCan you provide a detailed description of the picture?",
"role": "user"
},
{
"content": "A suitcase sits in a bedroom in front of a dresser.",
"... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000034938.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/23865_btm.png</img>\nPlease determine the location of elements I describe in this webpage screenshot (with point (x, y)).\nPrivacy Notices",
"role": "user"
},
{
"content": "(0.7503, 0.9269)",
"role": "assistant"
},
{... | [
"../../datasets/anno/annotated/fineweb-edu/raw/23865_btm.png"
] |
accessibility;general_acb | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/121976_top.png</img>\nIn this web page capture, please predict the positions of elements I describe (with bbox (x1, x2, y1, y2)).\n(0.68, 0.2567, 0.8, 0.29)",
"role": "user"
},
{
"content": "logo - opens in a new window",
... | [
"../../datasets/anno/annotated/fineweb-edu/raw/121976_top.png"
] |
som_cropped;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_general/78538.png</img>\nWhere is the element marked by the striking box and what's the content of it (with point (x, y))?",
"role": "user"
},
{
"content": "Share on Facebook (0.1494, 0.2679)",
"role": "assistant"
}
] | [
"../../datasets/anno/cropped_som/images_general/78538.png"
] |
basic_cropped;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/023362_btm.png</img>\nUsing this screenshot of a webpage, I'll indicate coordinates; can you predict the text of the elements (with bbox (x1, x2, y1, y2))?\n(0.0731, 0.7789, 0.1225, 0.804)",
"role": "user"
},
{
"conten... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/023362_btm.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/206804_top.png</img>\nIn the screenshot of this web page, please give me the coordinates of the element I want to click on according to my instructions (with point (x, y)).\nInstructions",
"role": "user"
},
{
"content": "(0.... | [
"../../datasets/anno/annotated/fineweb-edu/raw/206804_top.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/203221_btm.png</img>\nUsing this screenshot of a webpage, I'll indicate coordinates; can you predict the text of the elements (with bbox (x1, x2, y1, y2))?\n(0.144, 0.431, 0.466, 0.4533)",
"role": "user"
},
{
"content": "In ... | [
"../../datasets/anno/annotated/fineweb-edu/raw/203221_btm.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000414849.jpg</img>\nHow many glazed donuts are on the plate in the image?",
"role": "user"
},
{
"content": "There are two glazed donuts on the plate in the image.",
... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000414849.jpg"
] |
som;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/som/images_general/29113.png</img>\nWhere is the element marked by the striking box and what's the content of it (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "refresh\nReset all (0.1487, 0.0289, 0.215, 0.0689)",
"role": "assista... | [
"../../datasets/anno/som/images_general/29113.png"
] |
icon_mixed_cropped;icon_grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_icon_mixed/images/21761.png</img>\nI'll give descriptions of certain webpage float icons; please identify where they are in this screenshot (with point (x, y)).\nTreehouse logo",
"role": "user"
},
{
"content": "(0.3307, 0.4893)",
"role": ... | [
"../../datasets/anno/cropped_icon_mixed/images/21761.png"
] |
som_cropped;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_general/76978.png</img>\nLocate the rectangular bounding box with grounding and describe its content briefly (with point (x, y)).",
"role": "user"
},
{
"content": "Listen (0.4219, 0.1853)",
"role": "assistant"
}
] | [
"../../datasets/anno/cropped_som/images_general/76978.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/15193.jpg</img>\nOn this interface, to flip to the home icon, where should I click (with point (x, y))?",
"role": "user"
},
{
"content": "(0.1264, 0.8996)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/15193.jpg"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/216093_btm.png</img>\nUsing this screenshot of a webpage, I'll indicate coordinates; can you predict the text of the elements (with bbox (x1, x2, y1, y2))?\n(0.474, 0.743, 0.489, 0.77)",
"role": "user"
},
{
"content": "Faceb... | [
"../../datasets/anno/annotated/fineweb-edu/raw/216093_btm.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/42929.jpg</img>\nOn this screen, where can I find the feature that allows me to \"scroll until the choose a network\" (with point (x, y))?",
"role": "user"
},
{
"content": "(0.5, 0.0984)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/42929.jpg"
] |
basic_cropped;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/001609_top.png</img>\nI'll describe certain elements on this webpage image; your task is to find their locations (with point (x, y)).\nCinema",
"role": "user"
},
{
"content": "(0.608, 0.2174)",
"role": "assistant"
... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/001609_top.png"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/densecap_data/GCC_train_000050388.jpg</img>\nCan you provide a detailed description of the picture?",
"role": "user"
},
{
"content": "In the image, a man dressed as Santa Claus is seen spreading festive che... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/densecap_data/GCC_train_000050388.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/99952_top.png</img>\nGiven this screenshot of a webpage, I'll describe some elements; locate them for me (with point (x, y)).\nOrder Now",
"role": "user"
},
{
"content": "(0.8687, 0.0326)",
"role": "assistant"
},
{
... | [
"../../datasets/anno/annotated/fineweb-edu/raw/99952_top.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000010318.jpg</img>\nWhat type of bus is depicted in the image?",
"role": "user"
},
{
"content": "The image features a double-decker bus, likely a tour bus, driving near... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000010318.jpg"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/19208.jpg</img>\nIn the current UI, I want to open item next to i want to item, where should I click (with point (x, y))?",
"role": "user"
},
{
"content": "(0.0681, 0.0711)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/19208.jpg"
] |
som_cropped;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_general/92749.png</img>\nWhat is being framed in the screen? What's the content of it? What's the position of it (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "Powered By (0.7294, 0.7331, 0.7768, 0.7556)",
"rol... | [
"../../datasets/anno/cropped_som/images_general/92749.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/39312.jpg</img>\nIn this digital interface, to initiate \"press the icon to the right of the call received activate\", where is my point of interest (with point (x, y))?",
"role": "user"
},
{
"content": "(0.9028, 0.6629)"... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/39312.jpg"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/014667_btm.png</img>\nOn this page, I will give you a series of coordinates and ask you to predict the text of the clickable element that corresponds to these coordinates (with bbox (x1, x2, y1, y2)).\n(0.0067, 0.876, 0.122, 0.909... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/014667_btm.png"
] |
rico_tasks;widget-grounding | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/66703.jpg</img>\nFacing this interface, where do I access to \"enter wages option bar\" (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.0632, 0.4488, 0.9368, 0.5027)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/66703.jpg"
] |
accessibility;general_acb | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/113959_top.png</img>\nI will describe elements from a webpage screenshot; your role is to locate them (with point (x, y)).\n(0.8289, 0.175)",
"role": "user"
},
{
"content": "Open Search Bar",
"role": "assistant"
}
] | [
"../../datasets/anno/annotated/fineweb-edu/raw/113959_top.png"
] |
accessibility;general_acb | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/222822_btm.png</img>\nI'll give descriptions of certain webpage elements; please identify where they are in this screenshot (with bbox (x1, x2, y1, y2)).\n(0.68, 0.087, 1.0, 0.8972)",
"role": "user"
},
{
"content": "With-An-... | [
"../../datasets/anno/annotated/fineweb-edu/raw/222822_btm.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/57448.jpg</img>\nIn this interface snapshot, to begin \"click the icon at the bottom left corner\", what is the clicking point (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.0, 0.8414, 0.0778, 0.8852)"... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/57448.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/106785_mid.png</img>\nI will describe features on this webpage screenshot; please predict their positions (with bbox (x1, x2, y1, y2)).\nAthletics",
"role": "user"
},
{
"content": "(0.4392, 0.0234, 0.5051, 0.1042)",
"rol... | [
"../../datasets/anno/annotated/fineweb-edu/raw/106785_mid.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/68895_btm.png</img>\nBased on this screenshot, I provide a series of coordinates; describe the text of the elements there (with bbox (x1, x2, y1, y2)).\n(0.721, 0.0, 0.7775, 0.056)",
"role": "user"
},
{
"content": "Culture",... | [
"../../datasets/anno/annotated/fineweb-edu/raw/68895_btm.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/022190_mid.png</img>\nBased on the screenshot of the page, I give a text description and you give its corresponding location (with bbox (x1, x2, y1, y2)).\nBuilding community networks to strengthen learning",
"role": "user"
... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/022190_mid.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/16795.jpg</img>\nIn this UI design, to process the instruction \"turn on hunter weapons addon icon\", where should I activate (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.0319, 0.7289, 0.7535, 0.7621... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/16795.jpg"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/122025_mid.png</img>\nIn this screenshot, I'll give coordinates and ask you to describe the text of the elements there (with bbox (x1, x2, y1, y2)).\n(0.081, 0.281, 0.6965, 0.3269)",
"role": "user"
},
{
"content": "You shoul... | [
"../../datasets/anno/annotated/fineweb-edu/raw/122025_mid.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000232401.jpg</img>\nWhat specialized adaptation allows the giraffe to reach its food source?",
"role": "user"
},
{
"content": "The giraffe has a specialized adaptation ... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000232401.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/149986_btm.png</img>\nOn this webpage screenshot, I will point out some elements; please predict their exact coordinates (with bbox (x1, x2, y1, y2)).\nHow to Spray a Cherry Tree",
"role": "user"
},
{
"content": "(0.775, 0.6... | [
"../../datasets/anno/annotated/fineweb-edu/raw/149986_btm.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000373758.jpg</img>\nWhere is the cow standing in the image?",
"role": "user"
},
{
"content": "The cow is standing on a pile of garbage located by a busy or crowded stre... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000373758.jpg"
] |
som;som_icon | [
{
"content": "Picture 1: <img>../../datasets/anno/som/images_icon_mixed/02286.png</img>\nObserve the region enclosed by the rectangular box. Generate a brief caption of it and point out its position with grounding (with point (x, y)).",
"role": "user"
},
{
"content": "representation of currency with... | [
"../../datasets/anno/som/images_icon_mixed/02286.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/006097_top.png</img>\nUsing the image of this webpage, I'll specify locations; you need to predict the text of the elements (with bbox (x1, x2, y1, y2)).\n(0.01, 0.204, 0.6549, 0.26)",
"role": "user"
},
{
"content": "F... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/006097_top.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/125732_mid.png</img>\nUsing the screenshot of this webpage, identify the coordinates of elements I describe (with bbox (x1, x2, y1, y2)).\nusing",
"role": "user"
},
{
"content": "(0.1078, 0.3208, 0.1391, 0.3528)",
"role"... | [
"../../datasets/anno/annotated/fineweb-edu/raw/125732_mid.png"
] |
icon_mixed;icon_referring | [
{
"content": "Picture 1: <img>../../datasets/anno/icon_mixed/images/38074.png</img>\nFrom the given webpage image, I'll provide coordinates of some icons; can you describe their main content or role in the webpage at these points (with bbox (x1, x2, y1, y2))?\n(0.35, 0.39, 0.424, 0.509)",
"role": "user"
}... | [
"../../datasets/anno/icon_mixed/images/38074.png"
] |
rico_tasks;widget-grounding | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/14550.jpg</img>\nIn the UI, where should I click if I want to complete instruction \"view this user 's profile\" (with point (x, y))?",
"role": "user"
},
{
"content": "(0.0792, 0.6719)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/14550.jpg"
] |
basic_cropped;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/013564_mid.png</img>\nWith this webpage capture, I provide a series of coordinates; please identify the text content of each element (with point (x, y)).\n(0.6424, 0.902)",
"role": "user"
},
{
"content": "Menu",
"r... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/013564_mid.png"
] |
som;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/som/images_general/94538.png</img>\nLocate the rectangular bounding box with grounding and describe its content briefly (with point (x, y)).",
"role": "user"
},
{
"content": "LOGIN (0.7144, 0.0745)",
"role": "assistant"
}
] | [
"../../datasets/anno/som/images_general/94538.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/40644_mid.png</img>\nOn this captured webpage, I provide specific coordinates; you need to predict the text of the elements there (with point (x, y)).\n(0.191, 0.57)",
"role": "user"
},
{
"content": "Email: Neurodiversity0-1... | [
"../../datasets/anno/annotated/fineweb-edu/raw/40644_mid.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/18651_top.png</img>\nBased on this web page screenshot, I provide coordinates; please predict the textual content at these spots (with bbox (x1, x2, y1, y2)).\n(0.76, 0.1389, 0.833, 0.19)",
"role": "user"
},
{
"content": "TE... | [
"../../datasets/anno/annotated/fineweb-edu/raw/18651_top.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/223553_btm.png</img>\nWith this image of the web page, I give specific coordinates; your task is to describe the text of the elements at these points (with point (x, y)).\n(0.4591, 0.55)",
"role": "user"
},
{
"content": "Sen... | [
"../../datasets/anno/annotated/fineweb-edu/raw/223553_btm.png"
] |
som;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/som/images_general/61915.png</img>\nAn element is marked with a bounding box. What's the text and the position of it (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "SEARCH (0.1672, 0.7788, 0.2539, 0.82)",
"role": "assistant"
}
] | [
"../../datasets/anno/som/images_general/61915.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/70555_top.png</img>\nBased on the screenshot of the web page, I give you the location to click on and you predict the text content of the corresponding element (with point (x, y)).\n(0.53, 0.0415)",
"role": "user"
},
{
"cont... | [
"../../datasets/anno/annotated/fineweb-edu/raw/70555_top.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/147641_mid.png</img>\nBased on the screenshot of the page, I give a text description and you give its corresponding location (with point (x, y)).\nASK A QUESTION",
"role": "user"
},
{
"content": "(0.5, 0.8883)",
"role": ... | [
"../../datasets/anno/annotated/fineweb-edu/raw/147641_mid.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000012930.jpg</img>\nWhat makes this image ideal for promoting a relaxing breakfast or brunch experience?",
"role": "user"
},
{
"content": "This image effectively convey... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000012930.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/ranking/raw/0015625_top.png</img>\nPlease determine the location of elements I describe in this webpage screenshot (with bbox (x1, x2, y1, y2)).\nBien. Mejor. SiteGround. Hosting web perfeccionado",
"role": "user"
},
{
"content": "(0.0333, ... | [
"../../datasets/anno/annotated/ranking/raw/0015625_top.png"
] |
som_cropped;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_general/42072.png</img>\nAn element is marked with a bounding box. What's the text and the position of it (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "Sample Business Plans (0.3454, 0.5951, 0.4966, 0.616)",
"... | [
"../../datasets/anno/cropped_som/images_general/42072.png"
] |
icon_mixed_cropped;icon_referring | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_icon_mixed/images/11523.png</img>\nUsing this web page screenshot, I'll indicate some float icons' locations; please summarize these icons in a few words (with point (x, y)).\n(0.7495, 0.3925)",
"role": "user"
},
{
"content": "icon of vertica... | [
"../../datasets/anno/cropped_icon_mixed/images/11523.png"
] |
accessibility;general_acb | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/117803_top.png</img>\nOn this webpage capture, I'll point out specific elements for you to locate (with point (x, y)).\n(0.15, 0.106)",
"role": "user"
},
{
"content": "About Us",
"role": "assistant"
},
{
"content... | [
"../../datasets/anno/annotated/fineweb-edu/raw/117803_top.png"
] |
icon_mixed;icon_referring | [
{
"content": "Picture 1: <img>../../datasets/anno/icon_mixed/images/35629.png</img>\nUsing this web page screenshot, I'll indicate some float icons' locations; please summarize these icons in a few words (with bbox (x1, x2, y1, y2)).\n(0.5638, 0.58, 0.5957, 0.63)",
"role": "user"
},
{
"content": "ic... | [
"../../datasets/anno/icon_mixed/images/35629.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/5940_top.png</img>\nFor the elements I describe in this page capture, can you predict their positions (with point (x, y))?\nABOUT",
"role": "user"
},
{
"content": "(0.7038, 0.1806)",
"role": "assistant"
},
{
"con... | [
"../../datasets/anno/annotated/fineweb-edu/raw/5940_top.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/64537.jpg</img>\nWithin this graphical user interface, to \"press the item below the structure 6 public\", which icon should I be looking for (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.4812, 0.7617... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/64537.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/67015_btm.png</img>\nUsing the screenshot of this webpage, identify the coordinates of elements I describe (with bbox (x1, x2, y1, y2)).\nConclusion",
"role": "user"
},
{
"content": "(0.0933, 0.11, 0.1758, 0.14)",
"role"... | [
"../../datasets/anno/annotated/fineweb-edu/raw/67015_btm.png"
] |
accessibility;general_acb | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/9026_btm.png</img>\nIn the image above, I will give a series of descriptions of the elements to be clicked. Please predict where you want to click (with bbox (x1, x2, y1, y2)).\n(0.2278, 0.505, 0.28, 0.5344)",
"role": "user"
},
... | [
"../../datasets/anno/annotated/fineweb-edu/raw/9026_btm.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000363370.jpg</img>\nHow many people are sitting at the table in the image?",
"role": "user"
},
{
"content": "There are two people sitting at the table in the image.",
... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000363370.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/183824_btm.png</img>\nUsing this webpage screenshot, I'll describe some elements. Please locate them (with bbox (x1, x2, y1, y2)).\nReject all",
"role": "user"
},
{
"content": "(0.4229, 0.5958, 0.4938, 0.6417)",
"role": ... | [
"../../datasets/anno/annotated/fineweb-edu/raw/183824_btm.png"
] |
som;som_icon | [
{
"content": "Picture 1: <img>../../datasets/anno/som/images_icon_mixed/40854.png</img>\nWhat is being framed in the screen? What's the content of it? What's the position of it (with point (x, y))?",
"role": "user"
},
{
"content": "logo of Stack Overflow (0.3409, 0.4572)",
"role": "assistant"
... | [
"../../datasets/anno/som/images_icon_mixed/40854.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/97771_btm.png</img>\nUsing this webpage screenshot, I'll describe some elements. Please locate them (with point (x, y)).\nTerms of Service",
"role": "user"
},
{
"content": "(0.6078, 0.9139)",
"role": "assistant"
},
{... | [
"../../datasets/anno/annotated/fineweb-edu/raw/97771_btm.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/32458.jpg</img>\nIn the UI, where should I click if I want to complete instruction \"choose the icon next to the fake gps search icon\" (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.5826, 0.7051, 0.64... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/32458.jpg"
] |
icon_mixed;icon_grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/icon_mixed/images/44425.png</img>\nPlease use the provided webpage screenshot to locate the icons I describe (with point (x, y)).\nminus button in a filled square",
"role": "user"
},
{
"content": "(0.9368, 0.7759)",
"role": "assistant"
},
{
... | [
"../../datasets/anno/icon_mixed/images/44425.png"
] |
basic_cropped;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/025481.png</img>\nOn this page screenshot, I give coordinates; please predict the text content of the corresponding elements (with bbox (x1, x2, y1, y2)).\n(0.0, 0.02, 0.13, 0.05)",
"role": "user"
},
{
"content": "Astr... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/025481.png"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/densecap_data/GCC_train_001728671.jpg</img>\nCan you provide a detailed description of the picture?",
"role": "user"
},
{
"content": "In the image, we see a man standing in the opening of a door, with light... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/densecap_data/GCC_train_001728671.jpg"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/TextCaps/train_images/1b2195d9cd1b2b6f.jpg</img>\nGenerate a detailed caption of this image.",
"role": "user"
},
{
"content": "A street that people are walking on that has a thrift store next to tall buildi... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/TextCaps/train_images/1b2195d9cd1b2b6f.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/181092_top.png</img>\nPlease identify the coordinates of the webpage elements I describe based on the provided screenshot (with bbox (x1, x2, y1, y2)).\n+",
"role": "user"
},
{
"content": "(0.8375, 0.4444, 0.8639, 0.4796)",
... | [
"../../datasets/anno/annotated/fineweb-edu/raw/181092_top.png"
] |
rico_tasks;widget-grounding | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/66762.jpg</img>\nIn the current UI, I want to forwad, where should I click (with point (x, y))?",
"role": "user"
},
{
"content": "(0.958, 0.2734)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/66762.jpg"
] |
icon_mixed;icon_grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/icon_mixed/images/26647.png</img>\nIn this web page image, please locate the float icons as I describe them (with bbox (x1, x2, y1, y2)).\nicon depicting a scanning interface",
"role": "user"
},
{
"content": "(0.466, 0.0396, 0.4931, 0.0802)",
"ro... | [
"../../datasets/anno/icon_mixed/images/26647.png"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/209224_btm.png</img>\nFrom this webpage image, I will describe specific elements. Please predict their exact locations (with bbox (x1, x2, y1, y2)).\nCar Accidents",
"role": "user"
},
{
"content": "(0.5089, 0.4248, 0.5667, 0... | [
"../../datasets/anno/annotated/fineweb-edu/raw/209224_btm.png"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000563342.jpg</img>\nWhat is the color of the cat in the image?",
"role": "user"
},
{
"content": "The cat in the image is black.",
"role": "assistant"
},
{
"... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000563342.jpg"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/738.jpg</img>\nIn this interface snapshot, to begin \"flip until strictly friends item\", what is the clicking point (with point (x, y))?",
"role": "user"
},
{
"content": "(0.5, 0.3203)",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/738.jpg"
] |
monkey_training;monkey_training | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/densecap_data/GCC_train_001132699.jpg</img>\nCan you provide a detailed description of the picture?",
"role": "user"
},
{
"content": "In this image, we see a protest march in Yangon against the government o... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/densecap_data/GCC_train_001132699.jpg"
] |
captioning;description | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/ranking/raw/0007132_mid.png</img>\nGenerate a brief introduction or description of this webpage for future appearance in the summary of the search result below the webpage title.",
"role": "user"
},
{
"content": "A fast VPN service that enc... | [
"../../datasets/anno/annotated/ranking/raw/0007132_mid.png"
] |
rico_tasks;widget-caption | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/70181.jpg</img>\nCan you explain what the user interface element at (0.702, 0.884) does?",
"role": "user"
},
{
"content": "pull up keypad",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/70181.jpg"
] |
som_cropped;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_general/42451.png</img>\nObserve the region enclosed by the rectangular box. Generate a brief caption of it and point out its position with grounding (with point (x, y)).",
"role": "user"
},
{
"content": "Dissertation Writing Servi... | [
"../../datasets/anno/cropped_som/images_general/42451.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/36366.jpg</img>\nWhen interacting with this UI, for the operation labeled \"choose the let the journey\", what is my target (with point (x, y))?",
"role": "user"
},
{
"content": "(0.5, 0.8172)",
"role": "assistant"
... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/36366.jpg"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/1080.jpg</img>\nOn this screen, where can I find the feature that allows me to \"select icon to the right of the pre-register\" (with point (x, y))?",
"role": "user"
},
{
"content": "(0.892, 0.0957)",
"role": "assista... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/1080.jpg"
] |
llava_instruct;llava_instruct | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000433310.jpg</img>\nWhat type of animal is dressed up in the image?",
"role": "user"
},
{
"content": "The animal dressed up in the image is a cow.",
"role": "assist... | [
"/data1/cxt/projects/mmllm/datasets/MonkeyData/Monkey_Train_data/COCO2014/train2014/COCO_train2014_000000433310.jpg"
] |
som_cropped;som_icon | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_icon_mixed/86484.png</img>\nAn element is marked with a bounding box. What's the text and the position of it (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "icon for dislike action (0.8082, 0.8617, 0.848, 0.9141)",
... | [
"../../datasets/anno/cropped_som/images_icon_mixed/86484.png"
] |
rico_tasks;widget-caption | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/11906.jpg</img>\nWhat is the significance of the element located at (0.5431, 0.18, 1.0, 0.2437) in the application?",
"role": "user"
},
{
"content": "open the history screen",
"role": "assistant"
}
] | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/11906.jpg"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/16698.jpg</img>\nIn the UI here, I need to open the update stories even icon, what is the coordinates of the element is related to this (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.0389, 0.3602, 0.64... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/16698.jpg"
] |
basic;grounding | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/38180_btm.png</img>\nFrom this webpage image, I will describe specific elements. Please predict their exact locations (with point (x, y)).\nAll Talents",
"role": "user"
},
{
"content": "(0.8017, 0.4911)",
"role": "assist... | [
"../../datasets/anno/annotated/fineweb-edu/raw/38180_btm.png"
] |
accessibility;general_acb | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/fineweb-edu/raw/90099_top.png</img>\nI'll provide textual descriptions of elements in this webpage screenshot. Can you find their coordinates (with bbox (x1, x2, y1, y2))?\n(0.05, 0.06, 0.11, 0.17)",
"role": "user"
},
{
"content": "Skegby J... | [
"../../datasets/anno/annotated/fineweb-edu/raw/90099_top.png"
] |
som;som_general | [
{
"content": "Picture 1: <img>../../datasets/anno/som/images_general/10237.png</img>\nLocate the rectangular bounding box with grounding and describe its content briefly (with bbox (x1, x2, y1, y2)).",
"role": "user"
},
{
"content": "A typical vehicle with a standard internal combustion engine (ICE)... | [
"../../datasets/anno/som/images_general/10237.png"
] |
basic;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/049386_top.png</img>\nWith this webpage image, I provide coordinates; can you tell me the text of the elements at these points (with point (x, y))?\n(0.4372, 0.5195)",
"role": "user"
},
{
"content": "NICK GRANT",
"... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/049386_top.png"
] |
rico_tasks;ricosca | [
{
"content": "Picture 1: <img>/data1/cxt/projects/mmllm/datasets/Rico/combined/69817.jpg</img>\nWithin this graphical user interface, to \"press the marquis home item\", which icon should I be looking for (with bbox (x1, x2, y1, y2))?",
"role": "user"
},
{
"content": "(0.3847, 0.1012, 0.7597, 0.1492... | [
"/data1/cxt/projects/mmllm/datasets/Rico/combined/69817.jpg"
] |
basic_cropped;ocr | [
{
"content": "Picture 1: <img>../../datasets/anno/annotated/autoAnnosLargeNew/raw/047810_btm.png</img>\nBased on this web page screenshot, I provide coordinates; please predict the textual content at these spots (with bbox (x1, x2, y1, y2)).\n(0.0156, 0.144, 0.04, 0.16)",
"role": "user"
},
{
"conten... | [
"../../datasets/anno/annotated/autoAnnosLargeNew/raw/047810_btm.png"
] |
icon_mixed;icon_referring | [
{
"content": "Picture 1: <img>../../datasets/anno/icon_mixed/images/07052.png</img>\nUsing this web page screenshot, I'll indicate some float icons' locations; please summarize these icons in a few words (with point (x, y)).\n(0.3489, 0.89)",
"role": "user"
},
{
"content": "icon of a filled calendar... | [
"../../datasets/anno/icon_mixed/images/07052.png"
] |
som_cropped;som_icon | [
{
"content": "Picture 1: <img>../../datasets/anno/cropped_som/images_icon_mixed/11352.png</img>\nWhat's the text and coordinates of the marked element (with point (x, y))?",
"role": "user"
},
{
"content": "wallet icon from Ant Design (0.7197, 0.7546)",
"role": "assistant"
}
] | [
"../../datasets/anno/cropped_som/images_icon_mixed/11352.png"
] |
End of preview. Expand in Data Studio
This is the dataset repository of paper EDGE: Enhanced Grounded GUI Understanding with Enriched Multi-Granularity Synthetic Data.
Considering the huge number of images, the all_items.jsonl provided here contains the final QA pairs for training, but does not contain images for the time being. We will release all images as soon as possible.
You can also follow the mark_webpages and dataset.py scripts provided in the code repository to generate your own webpage image-question-answering dataset.
The model parameters is published in repository EDGE-Model.
The model training and inference scripts are published in anonymous repository EDGE.
- Downloads last month
- 10
Size of downloaded dataset files:
1.39 GB
Size of the auto-converted Parquet files:
478 MB
Number of rows:
1,660,709
Models trained or fine-tuned on EDGEwww25/EDGE-Dataset