Add chunk 21 (50,000 samples)

README.md

@@ -25,10 +25,10 @@ dataset_info:
         dtype: float32
   splits:
   - name: train
-    num_bytes: 989327422
+    num_bytes: 867846529
     num_examples: 50000
-  download_size: 987392012
-  dataset_size: 989327422
+  download_size: 865907305
+  dataset_size: 867846529
 configs:
 - config_name: default
   data_files:

data/train-00000-of-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:75b53bd8ba9eff37f032712dde13be477e36aba806bc4bd4d0f199186cba7b57
-size 502206082
+oid sha256:2b7784ac8f1a7260dbd148de59bde09da832537bf164db00d5a53df3d3a00aff
+size 434487649

data/train-00001-of-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c9cb83467f78cc3eda43c85ce76ff50804fb2d4399b007957593e48d28cf56b0
-size 485185930
+oid sha256:42c6193862a0459d6286b3b34ebae22869bbc1a1a531f38a464af7d847d3e983
+size 431419656