Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -139,9 +139,24 @@ You can also access the data processed for reranking tasks. To do that:
|
|
| 139 |
```python
|
| 140 |
from datasets import load_dataset
|
| 141 |
|
| 142 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 143 |
|
| 144 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 145 |
```
|
| 146 |
|
| 147 |
## License and Terms of Use
|
|
|
|
| 139 |
```python
|
| 140 |
from datasets import load_dataset
|
| 141 |
|
| 142 |
+
name = 'xhluca/WebLINX-testing'
|
| 143 |
+
# validation split:
|
| 144 |
+
valid = load_dataset(name, 'reranking', split='validation')
|
| 145 |
+
# test-iid split
|
| 146 |
+
test_iid = load_dataset(name, 'reranking', split='test_iid')
|
| 147 |
+
# other options: test_cat, test_geo, test_vis, test_web
|
| 148 |
|
| 149 |
+
print("Query:")
|
| 150 |
+
print(valid[0]['query'])
|
| 151 |
+
|
| 152 |
+
print("\nPositive:")
|
| 153 |
+
print(valid[0]['positives'][0])
|
| 154 |
+
|
| 155 |
+
print("\nNegative #1:")
|
| 156 |
+
print(valid[0]['negatives'][0])
|
| 157 |
+
|
| 158 |
+
print("\nNegative #2:")
|
| 159 |
+
print(valid[0]['negatives'][1])
|
| 160 |
```
|
| 161 |
|
| 162 |
## License and Terms of Use
|