Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -16,4 +16,30 @@ configs:
|
|
16 |
data_files:
|
17 |
- split: train
|
18 |
path: data/train-*
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
data_files:
|
17 |
- split: train
|
18 |
path: data/train-*
|
19 |
+
language:
|
20 |
+
- da
|
21 |
+
task_categories:
|
22 |
+
- feature-extraction
|
23 |
+
- question-answering
|
24 |
+
license: cc-by-sa-4.0
|
25 |
---
|
26 |
+
|
27 |
+
# SQuAD question-answer pairs in Danish
|
28 |
+
|
29 |
+
## About
|
30 |
+
|
31 |
+
This dataset is a version of the [SQuAD question-answer pairs dataset](https://huggingface.co/datasets/sentence-transformers/embedding-training-data) machine-translated from English to Danish ([link to original dataset](https://huggingface.co/datasets/squad)).
|
32 |
+
|
33 |
+
Machine translation is performed using the Helsinki NLP [English-to-Danish OPUS-MT model](https://huggingface.co/Helsinki-NLP/opus-mt-en-da).
|
34 |
+
|
35 |
+
The dataset contains ~87k question-answer pairs and can be used to train embedding and question-answer models. Each pair consists of one question ('query') and one passage containing the answer ('passage').
|
36 |
+
|
37 |
+
## Usage
|
38 |
+
|
39 |
+
Using the HuggingFace datasets library:
|
40 |
+
|
41 |
+
```python
|
42 |
+
from datasets import load_dataset
|
43 |
+
|
44 |
+
dataset = load_dataset("KennethTM/squad_pairs_danish")
|
45 |
+
```
|