Update jparacrawl.py

jparacrawl.py

@@ -10,6 +10,7 @@ import unicodedata
 
 import packaging.version
 import pandas as pd
+import pyarrow  # use pyarrow engine for pd.read_csv() 
 
 import datasets
 from datasets.features import Translation, Value
@@ -154,6 +155,7 @@ class JParaCrawl(datasets.GeneratorBasedBuilder):
                     index_col=None,
                     sep="\t\t",  # dummy delimiter
                     encoding="utf8",
+                    engine='pyarrow',
                 )
                 break
         assert df is not None, extracted_path