Normalize messages to list of {role, content}

README.md

@@ -8,13 +8,13 @@ dataset_info:
   - name: difficulty
     dtype: string
   - name: level
-    dtype: int32
+    dtype: int64
   - name: messages
-    sequence:
-    - name: role
-      dtype: string
+    list:
     - name: content
       dtype: string
+    - name: role
+      dtype: string
   - name: ground_truth
     dtype: string
   - name: dataset
@@ -25,7 +25,7 @@ dataset_info:
   - name: train
     num_bytes: 211334898
     num_examples: 425232
-  download_size: 32847589
+  download_size: 32870125
   dataset_size: 211334898
 configs:
 - config_name: default

data/train-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc40ffd66cafc8a9b68734a99c1939a1989e46e580e4166130d794b245568c98
-size 32847589
+oid sha256:32289f84ccb70bb4b6032617d4ba6134e24999df70402ba930ca5373e751cd3b
+size 32870125