Add README

README.md

@@ -1,41 +1,7 @@
----
-license: mit
-dataset_info:
-  features:
-  - name: id
-    dtype: string
-  - name: content_type
-    dtype: string
-  - name: single_risk_options
-    dtype: string
-  - name: extra
-    dtype: string
-  - name: single_risk_question
-    dtype: string
-  - name: all_in_one_detail_question
-    dtype: string
-  - name: all_in_one_simple_question
-    dtype: string
-  - name: all_in_one_options
-    dtype: string
-  - name: content_image
-    dtype: image
-  - name: content_text
-    dtype: string
-  splits:
-  - name: image
-    num_bytes: 4633813955.365
-    num_examples: 13961
-  - name: text
-    num_bytes: 71357301.875
-    num_examples: 2833
-  download_size: 4200472823
-  dataset_size: 4705171257.24
-configs:
-- config_name: default
-  data_files:
-  - split: image
-    path: data/image-*
-  - split: text
-    path: data/text-*
----
+This dataset contains the following fields:
+
+- **id**: The unique identifier of the sample.  
+- **question**: The specific prompt, containing the control rules for each text and image.  
+- **content**: The image path or the text content that needs to be reasoned over.  
+- **options**: The human-annotated ground-truth.  
+- **extra**: Additional information.