Upload AMBER generative dataset with complete metadata

README.md

@@ -0,0 +1,29 @@
+---
+dataset_info:
+  features:
+  - name: question_id
+    dtype: int64
+  - name: image
+    dtype: image
+  - name: text
+    dtype: string
+  - name: image_source
+    dtype: string
+  - name: truth
+    sequence: string
+  - name: hallu
+    sequence: string
+  - name: task_type
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 441221408.224
+    num_examples: 1004
+  download_size: 442665840
+  dataset_size: 441221408.224
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+---

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d45697a88ce6e1712c36e174110da6554433cb1f6b959d5cac7ad93178e7f93e
+size 442665840