Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -11,4 +11,22 @@ task_categories:
|
|
| 11 |
- conditional-text-generation
|
| 12 |
task_ids:
|
| 13 |
- summarization
|
| 14 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
- conditional-text-generation
|
| 12 |
task_ids:
|
| 13 |
- summarization
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
|
| 17 |
+
# Dataset Card for Swedish pubmed Dataset
|
| 18 |
+
The Swedish pubmed dataset has only been machine-translated to improve downstream fine-tuning on Swedish summarization tasks.
|
| 19 |
+
|
| 20 |
+
## Dataset Summary
|
| 21 |
+
Read about the full details at original English version: https://huggingface.co/datasets/pubmed
|
| 22 |
+
|
| 23 |
+
### Data Fields
|
| 24 |
+
- `document`: a string containing the body of the paper
|
| 25 |
+
- `summary`: a string containing the abstract of the paper
|
| 26 |
+
|
| 27 |
+
### Data Splits
|
| 28 |
+
The Swedish pubmed dataset follows the same splits as the original English version and has 1 splits: _train_.
|
| 29 |
+
|
| 30 |
+
| Dataset Split | Number of Instances in Split |
|
| 31 |
+
| ------------- | ------------------------------------------- |
|
| 32 |
+
| Train | 90,000 |
|