Restructure dataset for HF viewer: split JSONL, flatten metadata, add feature config

- Split qa.json into data/test.jsonl and data/val.jsonl for proper split detection
- Flatten metadata fields (category, domain, source, source_id, raw_folder) into top-level columns
- Keep generation_info as JSON string for heterogeneous source-specific data
- Add configs and dataset_info to README YAML for correct feature types (Image for image columns)
- Remove qa.json (replaced by per-split JSONL files)

README.md

@@ -14,6 +14,41 @@ tags:
   - multi-image
 size_categories:
   - 10K<n<100K
+configs:
+  - config_name: default
+    data_files:
+      - split: test
+        path: data/test.jsonl
+      - split: val
+        path: data/val.jsonl
+dataset_info:
+  features:
+    - name: image_1
+      dtype: image
+    - name: image_2
+      dtype: image
+    - name: question
+      dtype: string
+    - name: answer
+      dtype: string
+    - name: distractors
+      sequence: string
+    - name: has_caption
+      dtype: bool
+    - name: caption
+      dtype: string
+    - name: category
+      dtype: string
+    - name: domain
+      dtype: string
+    - name: source
+      dtype: string
+    - name: source_id
+      dtype: string
+    - name: raw_folder
+      dtype: string
+    - name: generation_info
+      dtype: string
 ---
 
 # VLM-SubtleBench