Update README.md
Browse files
README.md
CHANGED
@@ -32,3 +32,21 @@ configs:
|
|
32 |
- split: validation
|
33 |
path: data/validation-*
|
34 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
- split: validation
|
33 |
path: data/validation-*
|
34 |
---
|
35 |
+
### Dataset Summary
|
36 |
+
|
37 |
+
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
|
38 |
+
SQuAD1.1 with over 50,000 unanswerable questions written adversarially by crowdworkers to look similar to answerable ones.
|
39 |
+
To do well on SQuAD2.0, systems must not only answer questions when possible, but also determine when no answer is supported
|
40 |
+
by the paragraph and abstain from answering.
|
41 |
+
|
42 |
+
### Considerations for Using the Data
|
43 |
+
|
44 |
+
Due to DeepL terms and conditions, this dataset **must not be used for any machine translation work**, namely machine translation
|
45 |
+
system development and evaluation of any kind. In general, we wish you do not pair the original English data with the translations
|
46 |
+
except when working on research unrelated to machine translation, so as not to infringe on the terms and conditions.
|
47 |
+
|
48 |
+
### Licensing Information
|
49 |
+
|
50 |
+
Contents of this repository are distributed under the
|
51 |
+
[Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0)](https://creativecommons.org/licenses/by-sa/4.0/).
|
52 |
+
Copyright of the dataset contents belongs to the original copyright holders.
|