Add reasoning VQA dataset

README.md

@@ -40,6 +40,26 @@ dataset_info:
     num_examples: 400
   download_size: 45849706
   dataset_size: 48951712.0
+- config_name: reasoning_vqa
+  features:
+  - name: image
+    dtype: image
+  - name: messages
+    dtype: string
+  - name: answer
+    dtype: string
+  - name: meta_info
+    struct:
+    - name: id
+      dtype: string
+    - name: question_type
+      dtype: string
+  splits:
+  - name: train
+    num_bytes: 51553044.0
+    num_examples: 400
+  download_size: 48717300
+  dataset_size: 51553044.0
 configs:
 - config_name: cognition_vqa
   data_files:
@@ -49,4 +69,8 @@ configs:
   data_files:
   - split: train
     path: document_classification/train-*
+- config_name: reasoning_vqa
+  data_files:
+  - split: train
+    path: reasoning_vqa/train-*
 ---

reasoning_vqa/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c87069987bf77dd6b6d7ba6813f83de4eee2bc3e937e85b71add93084629280
+size 48717300