upload subset_20k/test_100 's json, videos and frames

data/subset_20k/test_100/data_inc.json

@@ -0,0 +1,1568 @@
+[
+    {
+        "id": "2002126",
+        "images": [
+            "2002126_00.jpg",
+            "2002126_01.jpg",
+            "2002126_02.jpg",
+            "2002126_03.jpg",
+            "2002126_04.jpg",
+            "2002126_05.jpg",
+            "2002126_06.jpg",
+            "2002126_07.jpg",
+            "2002126_08.jpg",
+            "2002126_09.jpg",
+            "2002126_10.jpg",
+            "2002126_11.jpg",
+            "2002126_12.jpg",
+            "2002126_13.jpg",
+            "2002126_14.jpg",
+            "2002126_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a smiling and winged ham, a sausage with sunglasses and a creamy chop on top and a ribs with smoke coming out moving  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 2\n factual consistency: 1\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1001545",
+        "images": [
+            "1001545_00.jpg",
+            "1001545_01.jpg",
+            "1001545_02.jpg",
+            "1001545_03.jpg",
+            "1001545_04.jpg",
+            "1001545_05.jpg",
+            "1001545_06.jpg",
+            "1001545_07.jpg",
+            "1001545_08.jpg",
+            "1001545_09.jpg",
+            "1001545_10.jpg",
+            "1001545_11.jpg",
+            "1001545_12.jpg",
+            "1001545_13.jpg",
+            "1001545_14.jpg",
+            "1001545_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a yellow bird singing a son  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 2\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "0002572",
+        "images": [
+            "0002572_00.jpg",
+            "0002572_01.jpg",
+            "0002572_02.jpg",
+            "0002572_03.jpg",
+            "0002572_04.jpg",
+            "0002572_05.jpg",
+            "0002572_06.jpg",
+            "0002572_07.jpg",
+            "0002572_08.jpg",
+            "0002572_09.jpg",
+            "0002572_10.jpg",
+            "0002572_11.jpg",
+            "0002572_12.jpg",
+            "0002572_13.jpg",
+            "0002572_14.jpg",
+            "0002572_15.jpg",
+            "0002572_16.jpg",
+            "0002572_17.jpg",
+            "0002572_18.jpg",
+            "0002572_19.jpg",
+            "0002572_20.jpg",
+            "0002572_21.jpg",
+            "0002572_22.jpg",
+            "0002572_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a warrior in a dark night holding weapons ready to fight a wolf  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "2000773",
+        "images": [
+            "2000773_00.jpg",
+            "2000773_01.jpg",
+            "2000773_02.jpg",
+            "2000773_03.jpg",
+            "2000773_04.jpg",
+            "2000773_05.jpg",
+            "2000773_06.jpg",
+            "2000773_07.jpg",
+            "2000773_08.jpg",
+            "2000773_09.jpg",
+            "2000773_10.jpg",
+            "2000773_11.jpg",
+            "2000773_12.jpg",
+            "2000773_13.jpg",
+            "2000773_14.jpg",
+            "2000773_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a supermodel walks the fashion runway on the surface of Mars  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "2001062",
+        "images": [
+            "2001062_00.jpg",
+            "2001062_01.jpg",
+            "2001062_02.jpg",
+            "2001062_03.jpg",
+            "2001062_04.jpg",
+            "2001062_05.jpg",
+            "2001062_06.jpg",
+            "2001062_07.jpg",
+            "2001062_08.jpg",
+            "2001062_09.jpg",
+            "2001062_10.jpg",
+            "2001062_11.jpg",
+            "2001062_12.jpg",
+            "2001062_13.jpg",
+            "2001062_14.jpg",
+            "2001062_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is hyperrealistic picture of ancient time family, background from the past  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 1\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "7000811",
+        "images": [
+            "7000811_00.jpg",
+            "7000811_01.jpg",
+            "7000811_02.jpg",
+            "7000811_03.jpg",
+            "7000811_04.jpg",
+            "7000811_05.jpg",
+            "7000811_06.jpg",
+            "7000811_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is create an images avaatar of buiness man  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 2\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "0000654",
+        "images": [
+            "0000654_00.jpg",
+            "0000654_01.jpg",
+            "0000654_02.jpg",
+            "0000654_03.jpg",
+            "0000654_04.jpg",
+            "0000654_05.jpg",
+            "0000654_06.jpg",
+            "0000654_07.jpg",
+            "0000654_08.jpg",
+            "0000654_09.jpg",
+            "0000654_10.jpg",
+            "0000654_11.jpg",
+            "0000654_12.jpg",
+            "0000654_13.jpg",
+            "0000654_14.jpg",
+            "0000654_15.jpg",
+            "0000654_16.jpg",
+            "0000654_17.jpg",
+            "0000654_18.jpg",
+            "0000654_19.jpg",
+            "0000654_20.jpg",
+            "0000654_21.jpg",
+            "0000654_22.jpg",
+            "0000654_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is People from diverse backgrounds joining an online discussion  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 2\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "1000534",
+        "images": [
+            "1000534_00.jpg",
+            "1000534_01.jpg",
+            "1000534_02.jpg",
+            "1000534_03.jpg",
+            "1000534_04.jpg",
+            "1000534_05.jpg",
+            "1000534_06.jpg",
+            "1000534_07.jpg",
+            "1000534_08.jpg",
+            "1000534_09.jpg",
+            "1000534_10.jpg",
+            "1000534_11.jpg",
+            "1000534_12.jpg",
+            "1000534_13.jpg",
+            "1000534_14.jpg",
+            "1000534_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is christmas doggo relieves himself under the christmas tree  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 2\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "0002560",
+        "images": [
+            "0002560_00.jpg",
+            "0002560_01.jpg",
+            "0002560_02.jpg",
+            "0002560_03.jpg",
+            "0002560_04.jpg",
+            "0002560_05.jpg",
+            "0002560_06.jpg",
+            "0002560_07.jpg",
+            "0002560_08.jpg",
+            "0002560_09.jpg",
+            "0002560_10.jpg",
+            "0002560_11.jpg",
+            "0002560_12.jpg",
+            "0002560_13.jpg",
+            "0002560_14.jpg",
+            "0002560_15.jpg",
+            "0002560_16.jpg",
+            "0002560_17.jpg",
+            "0002560_18.jpg",
+            "0002560_19.jpg",
+            "0002560_20.jpg",
+            "0002560_21.jpg",
+            "0002560_22.jpg",
+            "0002560_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is diplomats negotiating peace terms over a map of europe  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "7000581",
+        "images": [
+            "7000581_00.jpg",
+            "7000581_01.jpg",
+            "7000581_02.jpg",
+            "7000581_03.jpg",
+            "7000581_04.jpg",
+            "7000581_05.jpg",
+            "7000581_06.jpg",
+            "7000581_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is It generates an image of a young man dressed as a mercenary astronaut, without a helmet, with a rough, serious face, futuristic desert clothing, dark colors, riding on a machine, a robot shaped like a metal horse, on a planet of exotic vegetation. in the background, five big stars.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "5000037",
+        "images": [
+            "5000037_00.jpg",
+            "5000037_01.jpg",
+            "5000037_02.jpg",
+            "5000037_03.jpg",
+            "5000037_04.jpg",
+            "5000037_05.jpg",
+            "5000037_06.jpg",
+            "5000037_07.jpg",
+            "5000037_08.jpg",
+            "5000037_09.jpg",
+            "5000037_10.jpg",
+            "5000037_11.jpg",
+            "5000037_12.jpg",
+            "5000037_13.jpg",
+            "5000037_14.jpg",
+            "5000037_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is an aerial view of an old wooden house found in a far village, a long alley of trees leading to the house  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1001576",
+        "images": [
+            "1001576_00.jpg",
+            "1001576_01.jpg",
+            "1001576_02.jpg",
+            "1001576_03.jpg",
+            "1001576_04.jpg",
+            "1001576_05.jpg",
+            "1001576_06.jpg",
+            "1001576_07.jpg",
+            "1001576_08.jpg",
+            "1001576_09.jpg",
+            "1001576_10.jpg",
+            "1001576_11.jpg",
+            "1001576_12.jpg",
+            "1001576_13.jpg",
+            "1001576_14.jpg",
+            "1001576_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is gaming monitor with ocean shore water flowing through monitor screen onto desk  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 2\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "1002026",
+        "images": [
+            "1002026_00.jpg",
+            "1002026_01.jpg",
+            "1002026_02.jpg",
+            "1002026_03.jpg",
+            "1002026_04.jpg",
+            "1002026_05.jpg",
+            "1002026_06.jpg",
+            "1002026_07.jpg",
+            "1002026_08.jpg",
+            "1002026_09.jpg",
+            "1002026_10.jpg",
+            "1002026_11.jpg",
+            "1002026_12.jpg",
+            "1002026_13.jpg",
+            "1002026_14.jpg",
+            "1002026_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is He immediately rushed to his side, offering first aid for his wounds and taking him to a nearby inn for further care.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 1\n factual consistency: 3\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "0001559",
+        "images": [
+            "0001559_00.jpg",
+            "0001559_01.jpg",
+            "0001559_02.jpg",
+            "0001559_03.jpg",
+            "0001559_04.jpg",
+            "0001559_05.jpg",
+            "0001559_06.jpg",
+            "0001559_07.jpg",
+            "0001559_08.jpg",
+            "0001559_09.jpg",
+            "0001559_10.jpg",
+            "0001559_11.jpg",
+            "0001559_12.jpg",
+            "0001559_13.jpg",
+            "0001559_14.jpg",
+            "0001559_15.jpg",
+            "0001559_16.jpg",
+            "0001559_17.jpg",
+            "0001559_18.jpg",
+            "0001559_19.jpg",
+            "0001559_20.jpg",
+            "0001559_21.jpg",
+            "0001559_22.jpg",
+            "0001559_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A group of teenagers accidentally enter the mysterious area of the old house  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "2000771",
+        "images": [
+            "2000771_00.jpg",
+            "2000771_01.jpg",
+            "2000771_02.jpg",
+            "2000771_03.jpg",
+            "2000771_04.jpg",
+            "2000771_05.jpg",
+            "2000771_06.jpg",
+            "2000771_07.jpg",
+            "2000771_08.jpg",
+            "2000771_09.jpg",
+            "2000771_10.jpg",
+            "2000771_11.jpg",
+            "2000771_12.jpg",
+            "2000771_13.jpg",
+            "2000771_14.jpg",
+            "2000771_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Sikh asternout with turban in space, close shot,behind spaceship with sikh logo in it  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1001670",
+        "images": [
+            "1001670_00.jpg",
+            "1001670_01.jpg",
+            "1001670_02.jpg",
+            "1001670_03.jpg",
+            "1001670_04.jpg",
+            "1001670_05.jpg",
+            "1001670_06.jpg",
+            "1001670_07.jpg",
+            "1001670_08.jpg",
+            "1001670_09.jpg",
+            "1001670_10.jpg",
+            "1001670_11.jpg",
+            "1001670_12.jpg",
+            "1001670_13.jpg",
+            "1001670_14.jpg",
+            "1001670_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A dinosaur skeleton and a military skeleton go trick or treating.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "1001702",
+        "images": [
+            "1001702_00.jpg",
+            "1001702_01.jpg",
+            "1001702_02.jpg",
+            "1001702_03.jpg",
+            "1001702_04.jpg",
+            "1001702_05.jpg",
+            "1001702_06.jpg",
+            "1001702_07.jpg",
+            "1001702_08.jpg",
+            "1001702_09.jpg",
+            "1001702_10.jpg",
+            "1001702_11.jpg",
+            "1001702_12.jpg",
+            "1001702_13.jpg",
+            "1001702_14.jpg",
+            "1001702_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A monkey and a duck are skiing in Antarctica, and the video should be very realistic. The video should be one minute long  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 2\n factual consistency: 2\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1000312",
+        "images": [
+            "1000312_00.jpg",
+            "1000312_01.jpg",
+            "1000312_02.jpg",
+            "1000312_03.jpg",
+            "1000312_04.jpg",
+            "1000312_05.jpg",
+            "1000312_06.jpg",
+            "1000312_07.jpg",
+            "1000312_08.jpg",
+            "1000312_09.jpg",
+            "1000312_10.jpg",
+            "1000312_11.jpg",
+            "1000312_12.jpg",
+            "1000312_13.jpg",
+            "1000312_14.jpg",
+            "1000312_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is blonde long hair woman sitting under the blossom tree and playing on flute  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 3\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "2000769",
+        "images": [
+            "2000769_00.jpg",
+            "2000769_01.jpg",
+            "2000769_02.jpg",
+            "2000769_03.jpg",
+            "2000769_04.jpg",
+            "2000769_05.jpg",
+            "2000769_06.jpg",
+            "2000769_07.jpg",
+            "2000769_08.jpg",
+            "2000769_09.jpg",
+            "2000769_10.jpg",
+            "2000769_11.jpg",
+            "2000769_12.jpg",
+            "2000769_13.jpg",
+            "2000769_14.jpg",
+            "2000769_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Consequences on Collective Memory and Identity  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "2000778",
+        "images": [
+            "2000778_00.jpg",
+            "2000778_01.jpg",
+            "2000778_02.jpg",
+            "2000778_03.jpg",
+            "2000778_04.jpg",
+            "2000778_05.jpg",
+            "2000778_06.jpg",
+            "2000778_07.jpg",
+            "2000778_08.jpg",
+            "2000778_09.jpg",
+            "2000778_10.jpg",
+            "2000778_11.jpg",
+            "2000778_12.jpg",
+            "2000778_13.jpg",
+            "2000778_14.jpg",
+            "2000778_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Kaal confronts the evil forces within the mansion, facing his fears and using his wits and courage to overcome them. (Style: Anime, dynamic and climactic)  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 2\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "7000582",
+        "images": [
+            "7000582_00.jpg",
+            "7000582_01.jpg",
+            "7000582_02.jpg",
+            "7000582_03.jpg",
+            "7000582_04.jpg",
+            "7000582_05.jpg",
+            "7000582_06.jpg",
+            "7000582_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is young jungle animals cartoon characters pixar style  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 2\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "1001532",
+        "images": [
+            "1001532_00.jpg",
+            "1001532_01.jpg",
+            "1001532_02.jpg",
+            "1001532_03.jpg",
+            "1001532_04.jpg",
+            "1001532_05.jpg",
+            "1001532_06.jpg",
+            "1001532_07.jpg",
+            "1001532_08.jpg",
+            "1001532_09.jpg",
+            "1001532_10.jpg",
+            "1001532_11.jpg",
+            "1001532_12.jpg",
+            "1001532_13.jpg",
+            "1001532_14.jpg",
+            "1001532_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A spinning top whirls very quickly in the middle of a table ar 4  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "1000052",
+        "images": [
+            "1000052_00.jpg",
+            "1000052_01.jpg",
+            "1000052_02.jpg",
+            "1000052_03.jpg",
+            "1000052_04.jpg",
+            "1000052_05.jpg",
+            "1000052_06.jpg",
+            "1000052_07.jpg",
+            "1000052_08.jpg",
+            "1000052_09.jpg",
+            "1000052_10.jpg",
+            "1000052_11.jpg",
+            "1000052_12.jpg",
+            "1000052_13.jpg",
+            "1000052_14.jpg",
+            "1000052_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Hello and welcome to our history story channel.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "2002116",
+        "images": [
+            "2002116_00.jpg",
+            "2002116_01.jpg",
+            "2002116_02.jpg",
+            "2002116_03.jpg",
+            "2002116_04.jpg",
+            "2002116_05.jpg",
+            "2002116_06.jpg",
+            "2002116_07.jpg",
+            "2002116_08.jpg",
+            "2002116_09.jpg",
+            "2002116_10.jpg",
+            "2002116_11.jpg",
+            "2002116_12.jpg",
+            "2002116_13.jpg",
+            "2002116_14.jpg",
+            "2002116_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A handsome young Asian man in a metallic silver sport coat walking out of a jail cell.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 2\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "0002570",
+        "images": [
+            "0002570_00.jpg",
+            "0002570_01.jpg",
+            "0002570_02.jpg",
+            "0002570_03.jpg",
+            "0002570_04.jpg",
+            "0002570_05.jpg",
+            "0002570_06.jpg",
+            "0002570_07.jpg",
+            "0002570_08.jpg",
+            "0002570_09.jpg",
+            "0002570_10.jpg",
+            "0002570_11.jpg",
+            "0002570_12.jpg",
+            "0002570_13.jpg",
+            "0002570_14.jpg",
+            "0002570_15.jpg",
+            "0002570_16.jpg",
+            "0002570_17.jpg",
+            "0002570_18.jpg",
+            "0002570_19.jpg",
+            "0002570_20.jpg",
+            "0002570_21.jpg",
+            "0002570_22.jpg",
+            "0002570_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is make a video there man coding text fast  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 2\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "2000770",
+        "images": [
+            "2000770_00.jpg",
+            "2000770_01.jpg",
+            "2000770_02.jpg",
+            "2000770_03.jpg",
+            "2000770_04.jpg",
+            "2000770_05.jpg",
+            "2000770_06.jpg",
+            "2000770_07.jpg",
+            "2000770_08.jpg",
+            "2000770_09.jpg",
+            "2000770_10.jpg",
+            "2000770_11.jpg",
+            "2000770_12.jpg",
+            "2000770_13.jpg",
+            "2000770_14.jpg",
+            "2000770_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A Space Shuttle adorned with vibrant African colors soaring through the depths of space  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 2\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "0002563",
+        "images": [
+            "0002563_00.jpg",
+            "0002563_01.jpg",
+            "0002563_02.jpg",
+            "0002563_03.jpg",
+            "0002563_04.jpg",
+            "0002563_05.jpg",
+            "0002563_06.jpg",
+            "0002563_07.jpg",
+            "0002563_08.jpg",
+            "0002563_09.jpg",
+            "0002563_10.jpg",
+            "0002563_11.jpg",
+            "0002563_12.jpg",
+            "0002563_13.jpg",
+            "0002563_14.jpg",
+            "0002563_15.jpg",
+            "0002563_16.jpg",
+            "0002563_17.jpg",
+            "0002563_18.jpg",
+            "0002563_19.jpg",
+            "0002563_20.jpg",
+            "0002563_21.jpg",
+            "0002563_22.jpg",
+            "0002563_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A man running from a cave, there are  monsters standing in front of the man in a winter theme, first person view  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "1001807",
+        "images": [
+            "1001807_00.jpg",
+            "1001807_01.jpg",
+            "1001807_02.jpg",
+            "1001807_03.jpg",
+            "1001807_04.jpg",
+            "1001807_05.jpg",
+            "1001807_06.jpg",
+            "1001807_07.jpg",
+            "1001807_08.jpg",
+            "1001807_09.jpg",
+            "1001807_10.jpg",
+            "1001807_11.jpg",
+            "1001807_12.jpg",
+            "1001807_13.jpg",
+            "1001807_14.jpg",
+            "1001807_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a women who is surfing with smiley face  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1000300",
+        "images": [
+            "1000300_00.jpg",
+            "1000300_01.jpg",
+            "1000300_02.jpg",
+            "1000300_03.jpg",
+            "1000300_04.jpg",
+            "1000300_05.jpg",
+            "1000300_06.jpg",
+            "1000300_07.jpg",
+            "1000300_08.jpg",
+            "1000300_09.jpg",
+            "1000300_10.jpg",
+            "1000300_11.jpg",
+            "1000300_12.jpg",
+            "1000300_13.jpg",
+            "1000300_14.jpg",
+            "1000300_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is german looking person looking at the cell phone in a group  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 2\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "1000497",
+        "images": [
+            "1000497_00.jpg",
+            "1000497_01.jpg",
+            "1000497_02.jpg",
+            "1000497_03.jpg",
+            "1000497_04.jpg",
+            "1000497_05.jpg",
+            "1000497_06.jpg",
+            "1000497_07.jpg",
+            "1000497_08.jpg",
+            "1000497_09.jpg",
+            "1000497_10.jpg",
+            "1000497_11.jpg",
+            "1000497_12.jpg",
+            "1000497_13.jpg",
+            "1000497_14.jpg",
+            "1000497_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a dog rolling over in the grass,4k  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1000663",
+        "images": [
+            "1000663_00.jpg",
+            "1000663_01.jpg",
+            "1000663_02.jpg",
+            "1000663_03.jpg",
+            "1000663_04.jpg",
+            "1000663_05.jpg",
+            "1000663_06.jpg",
+            "1000663_07.jpg",
+            "1000663_08.jpg",
+            "1000663_09.jpg",
+            "1000663_10.jpg",
+            "1000663_11.jpg",
+            "1000663_12.jpg",
+            "1000663_13.jpg",
+            "1000663_14.jpg",
+            "1000663_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is beautiful pear shaped body women anime  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 1\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1001613",
+        "images": [
+            "1001613_00.jpg",
+            "1001613_01.jpg",
+            "1001613_02.jpg",
+            "1001613_03.jpg",
+            "1001613_04.jpg",
+            "1001613_05.jpg",
+            "1001613_06.jpg",
+            "1001613_07.jpg",
+            "1001613_08.jpg",
+            "1001613_09.jpg",
+            "1001613_10.jpg",
+            "1001613_11.jpg",
+            "1001613_12.jpg",
+            "1001613_13.jpg",
+            "1001613_14.jpg",
+            "1001613_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Preparing for the Storm:  Sarah building a makeshift raft from debris. Sealing windows and doors to protect her home.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "7001216",
+        "images": [
+            "7001216_00.jpg",
+            "7001216_01.jpg",
+            "7001216_02.jpg",
+            "7001216_03.jpg",
+            "7001216_04.jpg",
+            "7001216_05.jpg",
+            "7001216_06.jpg",
+            "7001216_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is realistic samurai entry with smoke in background 4k hd  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "8000778",
+        "images": [
+            "8000778_00.jpg",
+            "8000778_01.jpg",
+            "8000778_02.jpg",
+            "8000778_03.jpg",
+            "8000778_04.jpg",
+            "8000778_05.jpg",
+            "8000778_06.jpg",
+            "8000778_07.jpg",
+            "8000778_08.jpg",
+            "8000778_09.jpg",
+            "8000778_10.jpg",
+            "8000778_11.jpg",
+            "8000778_12.jpg",
+            "8000778_13.jpg",
+            "8000778_14.jpg",
+            "8000778_15.jpg",
+            "8000778_16.jpg",
+            "8000778_17.jpg",
+            "8000778_18.jpg",
+            "8000778_19.jpg",
+            "8000778_20.jpg",
+            "8000778_21.jpg",
+            "8000778_22.jpg",
+            "8000778_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Cyborg robot with lots of wires, booting up and placing hands on screen like its a pane of glass, horrified facial expression  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 1\n factual consistency: 2\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "0000658",
+        "images": [
+            "0000658_00.jpg",
+            "0000658_01.jpg",
+            "0000658_02.jpg",
+            "0000658_03.jpg",
+            "0000658_04.jpg",
+            "0000658_05.jpg",
+            "0000658_06.jpg",
+            "0000658_07.jpg",
+            "0000658_08.jpg",
+            "0000658_09.jpg",
+            "0000658_10.jpg",
+            "0000658_11.jpg",
+            "0000658_12.jpg",
+            "0000658_13.jpg",
+            "0000658_14.jpg",
+            "0000658_15.jpg",
+            "0000658_16.jpg",
+            "0000658_17.jpg",
+            "0000658_18.jpg",
+            "0000658_19.jpg",
+            "0000658_20.jpg",
+            "0000658_21.jpg",
+            "0000658_22.jpg",
+            "0000658_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Create enchanting images of elegant and alluring women, each framed in exquisite poses and adorned with captivating backgrounds that amplify their beauty. Explore a diverse range of settings, from picturesque landscapes to sophisticated urban scenes, to bring forth the essence of glamour and allure in the generated images.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "7001213",
+        "images": [
+            "7001213_00.jpg",
+            "7001213_01.jpg",
+            "7001213_02.jpg",
+            "7001213_03.jpg",
+            "7001213_04.jpg",
+            "7001213_05.jpg",
+            "7001213_06.jpg",
+            "7001213_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is one group in underground basement doing. Plan for robari  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "7001217",
+        "images": [
+            "7001217_00.jpg",
+            "7001217_01.jpg",
+            "7001217_02.jpg",
+            "7001217_03.jpg",
+            "7001217_04.jpg",
+            "7001217_05.jpg",
+            "7001217_06.jpg",
+            "7001217_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Once upon a time, in a dense forest, there were two friends named John and Michael. They were inseparable and did everything together. One sunny morning, as they were strolling through the woods, they suddenly came face to face with a menacing bear.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 1\n factual consistency: 2\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "7000583",
+        "images": [
+            "7000583_00.jpg",
+            "7000583_01.jpg",
+            "7000583_02.jpg",
+            "7000583_03.jpg",
+            "7000583_04.jpg",
+            "7000583_05.jpg",
+            "7000583_06.jpg",
+            "7000583_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is News Anchorzooms  in at the news desk,camerass swing around with a view from the director behind thecameramann pointing at the news anchor at the news desk, studio light,s and microphones everywhere. digital 4k realistic mood lighting  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "1001870",
+        "images": [
+            "1001870_00.jpg",
+            "1001870_01.jpg",
+            "1001870_02.jpg",
+            "1001870_03.jpg",
+            "1001870_04.jpg",
+            "1001870_05.jpg",
+            "1001870_06.jpg",
+            "1001870_07.jpg",
+            "1001870_08.jpg",
+            "1001870_09.jpg",
+            "1001870_10.jpg",
+            "1001870_11.jpg",
+            "1001870_12.jpg",
+            "1001870_13.jpg",
+            "1001870_14.jpg",
+            "1001870_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is zoom in animation of a beach sunset view  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1001924",
+        "images": [
+            "1001924_00.jpg",
+            "1001924_01.jpg",
+            "1001924_02.jpg",
+            "1001924_03.jpg",
+            "1001924_04.jpg",
+            "1001924_05.jpg",
+            "1001924_06.jpg",
+            "1001924_07.jpg",
+            "1001924_08.jpg",
+            "1001924_09.jpg",
+            "1001924_10.jpg",
+            "1001924_11.jpg",
+            "1001924_12.jpg",
+            "1001924_13.jpg",
+            "1001924_14.jpg",
+            "1001924_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A Chinese man is practicing martial arts  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 3\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "8000589",
+        "images": [
+            "8000589_00.jpg",
+            "8000589_01.jpg",
+            "8000589_02.jpg",
+            "8000589_03.jpg",
+            "8000589_04.jpg",
+            "8000589_05.jpg",
+            "8000589_06.jpg",
+            "8000589_07.jpg",
+            "8000589_08.jpg",
+            "8000589_09.jpg",
+            "8000589_10.jpg",
+            "8000589_11.jpg",
+            "8000589_12.jpg",
+            "8000589_13.jpg",
+            "8000589_14.jpg",
+            "8000589_15.jpg",
+            "8000589_16.jpg",
+            "8000589_17.jpg",
+            "8000589_18.jpg",
+            "8000589_19.jpg",
+            "8000589_20.jpg",
+            "8000589_21.jpg",
+            "8000589_22.jpg",
+            "8000589_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A man, head slowly nodding, standing under street light, light breeze blowing trees and clouds, bright moon, dog walking in background,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 3\n text-to-video alignment: 1\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1000618",
+        "images": [
+            "1000618_00.jpg",
+            "1000618_01.jpg",
+            "1000618_02.jpg",
+            "1000618_03.jpg",
+            "1000618_04.jpg",
+            "1000618_05.jpg",
+            "1000618_06.jpg",
+            "1000618_07.jpg",
+            "1000618_08.jpg",
+            "1000618_09.jpg",
+            "1000618_10.jpg",
+            "1000618_11.jpg",
+            "1000618_12.jpg",
+            "1000618_13.jpg",
+            "1000618_14.jpg",
+            "1000618_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is photorealistic trippy violin turning into a piano turning into a saxophone.  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 1\n text-to-video alignment: 2\n factual consistency: 1\n overall score: 1\n"
+            }
+        ]
+    },
+    {
+        "id": "1001925",
+        "images": [
+            "1001925_00.jpg",
+            "1001925_01.jpg",
+            "1001925_02.jpg",
+            "1001925_03.jpg",
+            "1001925_04.jpg",
+            "1001925_05.jpg",
+            "1001925_06.jpg",
+            "1001925_07.jpg",
+            "1001925_08.jpg",
+            "1001925_09.jpg",
+            "1001925_10.jpg",
+            "1001925_11.jpg",
+            "1001925_12.jpg",
+            "1001925_13.jpg",
+            "1001925_14.jpg",
+            "1001925_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is happy birthday balloons man and woman walking hand in hand on the beach as the sun sets over the ocean  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 2\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "7001211",
+        "images": [
+            "7001211_00.jpg",
+            "7001211_01.jpg",
+            "7001211_02.jpg",
+            "7001211_03.jpg",
+            "7001211_04.jpg",
+            "7001211_05.jpg",
+            "7001211_06.jpg",
+            "7001211_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is High resolution, a female mechanic works on her 60s muscle car, all  that can be see are her beautiful legs and lower half of her torso, the other half is under the car working and is not seen, She is wearing ultra High Heeled shoes and short shorts, bright outdoor lighting  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 1\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "8000870",
+        "images": [
+            "8000870_00.jpg",
+            "8000870_01.jpg",
+            "8000870_02.jpg",
+            "8000870_03.jpg",
+            "8000870_04.jpg",
+            "8000870_05.jpg",
+            "8000870_06.jpg",
+            "8000870_07.jpg",
+            "8000870_08.jpg",
+            "8000870_09.jpg",
+            "8000870_10.jpg",
+            "8000870_11.jpg",
+            "8000870_12.jpg",
+            "8000870_13.jpg",
+            "8000870_14.jpg",
+            "8000870_15.jpg",
+            "8000870_16.jpg",
+            "8000870_17.jpg",
+            "8000870_18.jpg",
+            "8000870_19.jpg",
+            "8000870_20.jpg",
+            "8000870_21.jpg",
+            "8000870_22.jpg",
+            "8000870_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Chinese Song Dynasty farmers discovered fungi on trees in the forest  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 3\n dynamic degree: 2\n motion smoothness: 3\n text-to-video alignment: 1\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1000269",
+        "images": [
+            "1000269_00.jpg",
+            "1000269_01.jpg",
+            "1000269_02.jpg",
+            "1000269_03.jpg",
+            "1000269_04.jpg",
+            "1000269_05.jpg",
+            "1000269_06.jpg",
+            "1000269_07.jpg",
+            "1000269_08.jpg",
+            "1000269_09.jpg",
+            "1000269_10.jpg",
+            "1000269_11.jpg",
+            "1000269_12.jpg",
+            "1000269_13.jpg",
+            "1000269_14.jpg",
+            "1000269_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is A perfect whey protein powder that promotes muscle production and immunity  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 2\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 1\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "2002122",
+        "images": [
+            "2002122_00.jpg",
+            "2002122_01.jpg",
+            "2002122_02.jpg",
+            "2002122_03.jpg",
+            "2002122_04.jpg",
+            "2002122_05.jpg",
+            "2002122_06.jpg",
+            "2002122_07.jpg",
+            "2002122_08.jpg",
+            "2002122_09.jpg",
+            "2002122_10.jpg",
+            "2002122_11.jpg",
+            "2002122_12.jpg",
+            "2002122_13.jpg",
+            "2002122_14.jpg",
+            "2002122_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is nintendo 8 bit relaxing rain video  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 2\n motion smoothness: 3\n text-to-video alignment: 3\n factual consistency: 1\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "7000580",
+        "images": [
+            "7000580_00.jpg",
+            "7000580_01.jpg",
+            "7000580_02.jpg",
+            "7000580_03.jpg",
+            "7000580_04.jpg",
+            "7000580_05.jpg",
+            "7000580_06.jpg",
+            "7000580_07.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is Capture the final scene of Evergreen, with the trio and the townspeople celebrating under the moonlit snowfall  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 3\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 2\n"
+            }
+        ]
+    },
+    {
+        "id": "1001527",
+        "images": [
+            "1001527_00.jpg",
+            "1001527_01.jpg",
+            "1001527_02.jpg",
+            "1001527_03.jpg",
+            "1001527_04.jpg",
+            "1001527_05.jpg",
+            "1001527_06.jpg",
+            "1001527_07.jpg",
+            "1001527_08.jpg",
+            "1001527_09.jpg",
+            "1001527_10.jpg",
+            "1001527_11.jpg",
+            "1001527_12.jpg",
+            "1001527_13.jpg",
+            "1001527_14.jpg",
+            "1001527_15.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is a yolk in sea going through waves  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 3\n object consistency: 2\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 3\n factual consistency: 3\n overall score: 3\n"
+            }
+        ]
+    },
+    {
+        "id": "8000780",
+        "images": [
+            "8000780_00.jpg",
+            "8000780_01.jpg",
+            "8000780_02.jpg",
+            "8000780_03.jpg",
+            "8000780_04.jpg",
+            "8000780_05.jpg",
+            "8000780_06.jpg",
+            "8000780_07.jpg",
+            "8000780_08.jpg",
+            "8000780_09.jpg",
+            "8000780_10.jpg",
+            "8000780_11.jpg",
+            "8000780_12.jpg",
+            "8000780_13.jpg",
+            "8000780_14.jpg",
+            "8000780_15.jpg",
+            "8000780_16.jpg",
+            "8000780_17.jpg",
+            "8000780_18.jpg",
+            "8000780_19.jpg",
+            "8000780_20.jpg",
+            "8000780_21.jpg",
+            "8000780_22.jpg",
+            "8000780_23.jpg"
+        ],
+        "conversations": [
+            {
+                "from": "human",
+                "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos, \nplease watch the following frames of a given video and see the text prompt for generating the video, \nthen give scores from 7 different dimensions:\n(1) visual quality, \n(2) object consistency,\n(3) dynamic degree,\n(4) motion smoothness,\n(5) text-to-video alignment,\n(6) factual consistency, \n(7) overall score\nfor each dimension, output a number from [1,2,3], in which '1' stands for 'Bad', '2' stands for 'Average', '3' stands for 'Good'.\nHere is an output example: \nvisual quality: 3\nobject consistency: 2 \ndynamic degree: 2\nmotion smoothness: 1\ntext-to-video alignment: 1\nfactual consistency: 2\noverall score: 1\n\nFor this item, the text prompt is two people symbolizing  oposites, two different sites of power  ,\nall the frames of video are as follows: \n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> \n"
+            },
+            {
+                "from": "gpt",
+                "value": "visual quality: 1\n object consistency: 1\n dynamic degree: 3\n motion smoothness: 2\n text-to-video alignment: 1\n factual consistency: 2\n overall score: 2\n"
+            }
+        ]
+    }
+]