Datasets:
saattrupdan
commited on
Commit
•
b58d856
1
Parent(s):
30f6ac6
docs: Add links to DeepL and gTranslate, and fix a typo
Browse files
README.md
CHANGED
@@ -49,14 +49,16 @@ answers with plausible "answer candidates", being slight perturbations or transl
|
|
49 |
of the answer.
|
50 |
|
51 |
With the contexts extracted, we translated these to Danish, Swedish and Norwegian using
|
52 |
-
the DeepL translation service for
|
53 |
-
|
54 |
-
|
|
|
|
|
55 |
|
56 |
As we are filtering the MKQA samples at both the "merging stage" and the "translation
|
57 |
stage", we are not able to fully convert the 10,000 samples to the Scandinavian
|
58 |
languages, and instead get roughly 8,000 samples per language. These have further been
|
59 |
-
split into a training, validation and test split, with the
|
60 |
roughly 750 samples. The splits have been created in such a way that the proportion of
|
61 |
samples without an answer is roughly the same in each split.
|
62 |
|
|
|
49 |
of the answer.
|
50 |
|
51 |
With the contexts extracted, we translated these to Danish, Swedish and Norwegian using
|
52 |
+
the [DeepL translation service](https://www.deepl.com/pro-api?cta=header-pro-api) for
|
53 |
+
Danish and Swedish, and the [Google Translation
|
54 |
+
service](https://cloud.google.com/translate/docs/reference/rest/) for Norwegian. After
|
55 |
+
translation we ensured that the Scandinavian answers do indeed occur in the translated
|
56 |
+
contexts.
|
57 |
|
58 |
As we are filtering the MKQA samples at both the "merging stage" and the "translation
|
59 |
stage", we are not able to fully convert the 10,000 samples to the Scandinavian
|
60 |
languages, and instead get roughly 8,000 samples per language. These have further been
|
61 |
+
split into a training, validation and test split, with the latter two containing
|
62 |
roughly 750 samples. The splits have been created in such a way that the proportion of
|
63 |
samples without an answer is roughly the same in each split.
|
64 |
|