Convert dataset to Parquet

Convert dataset to Parquet.

This dataset uses `tasks`, which are deprecated and will raise an error after the next release of `datasets`. See: https://github.com/huggingface/datasets/pull/6999

README.md

@@ -48,10 +48,15 @@ dataset_info:
     sequence: int32
   splits:
   - name: train
-    num_bytes: 1423090
+    num_bytes: 1423066
     num_examples: 1458
-  download_size: 118173
-  dataset_size: 1423090
+  download_size: 218736
+  dataset_size: 1423066
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
 ---
 
 # Dataset Card for OneStopQA

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe10e4a2bd7488614df3d2084436b2050e6b8bed8ff23aed53da1da785baa48e
+size 218736