Update README.md
Browse files
README.md
CHANGED
@@ -8,7 +8,7 @@ language:
|
|
8 |
|
9 |
### Dataset Summary
|
10 |
|
11 |
-
This dataset is a DeepL -based machine translation of the English SQuAD2.0 dataset which combines the 100,000 questions in
|
12 |
SQuAD1.1 with over 50,000 unanswerable questions written adversarially by crowdworkers to look similar to answerable ones.
|
13 |
To do well on SQuAD2.0, systems must not only answer questions when possible, but also determine when no answer is supported
|
14 |
by the paragraph and abstain from answering.
|
|
|
8 |
|
9 |
### Dataset Summary
|
10 |
|
11 |
+
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
|
12 |
SQuAD1.1 with over 50,000 unanswerable questions written adversarially by crowdworkers to look similar to answerable ones.
|
13 |
To do well on SQuAD2.0, systems must not only answer questions when possible, but also determine when no answer is supported
|
14 |
by the paragraph and abstain from answering.
|