Update README.md
Browse files
README.md
CHANGED
|
@@ -55,6 +55,8 @@ The [SQuAD-NL v2.0](https://github.com/wietsedv/NLP-NL/tree/squad-nl-v1.0?tab=re
|
|
| 55 |
We added an extra column `score` to the original dataset.
|
| 56 |
The value of `score` is `1.0` if the question has an answer in the context (no matter where), and `0.0` if there are no answers in the context.
|
| 57 |
The allows the evaluation of embedding models that aim to pair queries and document fragments.
|
|
|
|
|
|
|
| 58 |
|
| 59 |
## Translations
|
| 60 |
SQuAD-NL is translated from the original [SQuAD](https://rajpurkar.github.io/SQuAD-explorer/) and [XQuAD](https://github.com/google-deepmind/xquad) English-language datasets.
|
|
|
|
| 55 |
We added an extra column `score` to the original dataset.
|
| 56 |
The value of `score` is `1.0` if the question has an answer in the context (no matter where), and `0.0` if there are no answers in the context.
|
| 57 |
The allows the evaluation of embedding models that aim to pair queries and document fragments.
|
| 58 |
+
Please note that is a very hard task for embedding models, because SQuAD v2.0 was specifically designed to contain unanswerable questions adversarially written to look similar to answerable ones.
|
| 59 |
+
Expect your models to perform poorly.
|
| 60 |
|
| 61 |
## Translations
|
| 62 |
SQuAD-NL is translated from the original [SQuAD](https://rajpurkar.github.io/SQuAD-explorer/) and [XQuAD](https://github.com/google-deepmind/xquad) English-language datasets.
|