Convert dataset to Parquet

Convert dataset to Parquet.

README.md

@@ -46,13 +46,13 @@ dataset_info:
     dtype: string
   splits:
   - name: train
-    num_bytes: 3943484
+    num_bytes: 3943436
     num_examples: 8782
   - name: validation
-    num_bytes: 472121
+    num_bytes: 472073
     num_examples: 980
-  download_size: 7352932
-  dataset_size: 4415605
+  download_size: 1931441
+  dataset_size: 4415509
 - config_name: proto_qa_cs
   features:
   - name: normalized-question
@@ -93,6 +93,14 @@ dataset_info:
     num_examples: 52
   download_size: 24755
   dataset_size: 12473
+configs:
+- config_name: proto_qa
+  data_files:
+  - split: train
+    path: proto_qa/train-*
+  - split: validation
+    path: proto_qa/validation-*
+  default: true
 ---
 
 # Dataset Card for [Dataset Name]

proto_qa/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89d160832f41afe5bea913dca67222629e427b52115c88db9386970e9a42c48b
+size 1733846

proto_qa/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f89cd5033fd72776bbaa8a49a45517c3bb8ac54e9a10c2df8403360cc2ba2ab2
+size 197595