Update README.md

README.md

@@ -8,7 +8,7 @@ language:
 
 ### Dataset Summary
 
-This dataset is a DeepL -based machine translation of the English SQuAD2.0 dataset which combines the 100,000 questions in 
+This is a Finnish SQuAD question answering dataset. It is a DeepL -based machine translation of the English SQuAD2.0 dataset which combines the 100,000 questions in 
 SQuAD1.1 with over 50,000 unanswerable questions written adversarially by crowdworkers to look similar to answerable ones. 
 To do well on SQuAD2.0, systems must not only answer questions when possible, but also determine when no answer is supported 
 by the paragraph and abstain from answering.