Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -9,9 +9,9 @@ Based on the [Kuaishou](https://www.kuaishou.com/) short video data, we construc
|
|
| 9 |
| Task | Description |
|
| 10 |
| -------------- | --------------------------------------------------------------------------- |
|
| 11 |
| CPV | The task of predicting product attributes in e-commerce. |
|
| 12 |
-
|
|
| 13 |
-
|
|
| 14 |
-
|
|
| 15 |
| High_Like | A binary classification task to determine the rate of likes of a short video. |
|
| 16 |
| SPU | The task of determining whether two items are the same product in e-commerce. |
|
| 17 |
|
|
@@ -19,9 +19,9 @@ Based on the [Kuaishou](https://www.kuaishou.com/) short video data, we construc
|
|
| 19 |
| Task | Qwen2.5-VL-3B | Qwen2.5-VL-7B | InternVL-3-8B | MiMo-VL-7B | Kwai Keye-VL-8B |
|
| 20 |
| -------------- | ------------- | ------------- | ------------- | ------- | ---- |
|
| 21 |
| CPV | 12.39 | 20.08 | 14.95 | 16.66 | 55.13 |
|
| 22 |
-
|
|
| 23 |
-
|
|
| 24 |
-
|
|
| 25 |
| High_Like | 48.85 | 47.94 | 47.03 | 51.14 | 55.25 |
|
| 26 |
| SPU | 74.09 | 81.34 | 75.64 | 81.86 | 87.05 |
|
| 27 |
|
|
@@ -37,17 +37,17 @@ Here is an example of an evaluation using VLMs on our datasets. The following co
|
|
| 37 |
"class": "KwaiVQADataset",
|
| 38 |
"dataset": "CPV"
|
| 39 |
},
|
| 40 |
-
"
|
| 41 |
"class": "KwaiVQADataset",
|
| 42 |
-
"dataset": "
|
| 43 |
},
|
| 44 |
-
"
|
| 45 |
"class": "KwaiVQADataset",
|
| 46 |
-
"dataset": "
|
| 47 |
},
|
| 48 |
-
"
|
| 49 |
"class": "KwaiYORNDataset",
|
| 50 |
-
"dataset": "
|
| 51 |
},
|
| 52 |
"High_like":{
|
| 53 |
"class":"KwaiYORNDataset",
|
|
|
|
| 9 |
| Task | Description |
|
| 10 |
| -------------- | --------------------------------------------------------------------------- |
|
| 11 |
| CPV | The task of predicting product attributes in e-commerce. |
|
| 12 |
+
| Hot_Videos_Aggregation | The task of determining whether multiple videos belong to the same topic. |
|
| 13 |
+
| Collection_Order | The task of determining the logical order between multiple videos with the same topic. |
|
| 14 |
+
| Pornographic_Comment | The task of whether short video comments contain pornographic content. |
|
| 15 |
| High_Like | A binary classification task to determine the rate of likes of a short video. |
|
| 16 |
| SPU | The task of determining whether two items are the same product in e-commerce. |
|
| 17 |
|
|
|
|
| 19 |
| Task | Qwen2.5-VL-3B | Qwen2.5-VL-7B | InternVL-3-8B | MiMo-VL-7B | Kwai Keye-VL-8B |
|
| 20 |
| -------------- | ------------- | ------------- | ------------- | ------- | ---- |
|
| 21 |
| CPV | 12.39 | 20.08 | 14.95 | 16.66 | 55.13 |
|
| 22 |
+
| Hot_Videos_Aggregation | 42.38 | 46.35 | 52.31 | 49.00 | 54.30 |
|
| 23 |
+
| Collection_Order | 36.88 | 59.83 | 64.75 | 78.68 | 84.43 |
|
| 24 |
+
| Pornographic_Comment | 56.61 | 56.08 | 57.14 | 68.25 | 71.96 |
|
| 25 |
| High_Like | 48.85 | 47.94 | 47.03 | 51.14 | 55.25 |
|
| 26 |
| SPU | 74.09 | 81.34 | 75.64 | 81.86 | 87.05 |
|
| 27 |
|
|
|
|
| 37 |
"class": "KwaiVQADataset",
|
| 38 |
"dataset": "CPV"
|
| 39 |
},
|
| 40 |
+
"Hot_Videos_Aggregation": {
|
| 41 |
"class": "KwaiVQADataset",
|
| 42 |
+
"dataset": "Hot_Videos_Aggregation"
|
| 43 |
},
|
| 44 |
+
"Collection_Order": {
|
| 45 |
"class": "KwaiVQADataset",
|
| 46 |
+
"dataset": "Collection_Order"
|
| 47 |
},
|
| 48 |
+
"Pornographic_Comment": {
|
| 49 |
"class": "KwaiYORNDataset",
|
| 50 |
+
"dataset": "Pornographic_Comment"
|
| 51 |
},
|
| 52 |
"High_like":{
|
| 53 |
"class":"KwaiYORNDataset",
|