Convert dataset to Parquet

Convert dataset to Parquet.

README.md

@@ -31,16 +31,25 @@ dataset_info:
     dtype: int32
   splits:
   - name: train
-    num_bytes: 17089298
+    num_bytes: 16992007
     num_examples: 29808
   - name: test
-    num_bytes: 1532223
+    num_bytes: 1522546
     num_examples: 3003
   - name: validation
-    num_bytes: 3779584
+    num_bytes: 3757111
     num_examples: 6894
-  download_size: 5247733
-  dataset_size: 22401105
+  download_size: 4248819
+  dataset_size: 22271664
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: test
+    path: data/test-*
+  - split: validation
+    path: data/validation-*
 ---
 
 # Dataset Card for "wiqa"

data/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0af5d229a14369004f1edde21b97586307d85b9cdcd1d38e3de6440bc882e907
+size 193432

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed1ba14025996701297e83880caa824d3d4d579546728eeac8bdb891a2d902c9
+size 3532349

data/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2dc03842e8a60767a6bc630b98219881cc04755b0498ee9dc1cea157b90e8bc
+size 523038