Update README.md
Browse files
README.md
CHANGED
@@ -2,6 +2,7 @@
|
|
2 |
license: mit
|
3 |
datasets:
|
4 |
- mlburnham/PoliStance_Affect
|
|
|
5 |
pipeline_tag: zero-shot-classification
|
6 |
language:
|
7 |
- en
|
@@ -16,9 +17,11 @@ It is currently trained for zero-shot classification of stances towards politica
|
|
16 |
Further capabilities will be added and benchmarked as more training data is developed.
|
17 |
|
18 |
# Training Data
|
19 |
-
The model was trained using the [PoliStance Affect](https://huggingface.co/datasets/mlburnham/PoliStance_Affect)
|
20 |
-
|
21 |
-
|
|
|
|
|
22 |
|
23 |
# Evaluation
|
24 |
Results below are performance on the PoliStance Affect test set.
|
|
|
2 |
license: mit
|
3 |
datasets:
|
4 |
- mlburnham/PoliStance_Affect
|
5 |
+
- mlburnham/PoliStance_Affect_QT
|
6 |
pipeline_tag: zero-shot-classification
|
7 |
language:
|
8 |
- en
|
|
|
17 |
Further capabilities will be added and benchmarked as more training data is developed.
|
18 |
|
19 |
# Training Data
|
20 |
+
The model was trained using the [PoliStance Affect](https://huggingface.co/datasets/mlburnham/PoliStance_Affect) and [PoliStance Affect_QT](https://huggingface.co/datasets/mlburnham/PoliStance_Affect_QT) datasets.
|
21 |
+
- Polistance Affect: ~27,000 political texts about U.S. politicians and political groups that have been triple coded for stance.
|
22 |
+
- Polistance Affect QT: A set of quote tweets about U.S. politicians that pose a particularly challenging classification task.
|
23 |
+
|
24 |
+
The test set for both datasets contains documents about six politicians that were not included in the training set in order to evaluate zero-shot classification performance.
|
25 |
|
26 |
# Evaluation
|
27 |
Results below are performance on the PoliStance Affect test set.
|