Upload dataset

README.md

@@ -1,5 +1,40 @@
 ---
 license: mit
+dataset_info:
+  features:
+  - name: img
+    dtype: image
+  - name: question
+    dtype: string
+  - name: choices
+    dtype: string
+  - name: answer_0
+    dtype: string
+  - name: answer_1
+    dtype: string
+  - name: answer_2
+    dtype: string
+  - name: answer_3
+    dtype: string
+  - name: answer_4
+    dtype: string
+  - name: gt
+    dtype: string
+  - name: avg_rating
+    dtype: float64
+  - name: category
+    dtype: int64
+  splits:
+  - name: non_bounding_box
+    num_bytes: 4641231.0
+    num_examples: 36
+  download_size: 2685946
+  dataset_size: 4641231.0
+configs:
+- config_name: default
+  data_files:
+  - split: non_bounding_box
+    path: data/non_bounding_box-*
 ---
 
 This is the visual human preferences dataset with both the bounding-box and non-bounding-box variants (bb_responses.csv and reg_responses.csv respectively) from the paper "A Dataset for Dynamic Human Preferences for Vision Language Models".

data/non_bounding_box-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be17457bb960a20d5d4dd9cad9dfdee14ecfcebc2f64d20995cd111011065156
+size 2685946