Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -38,7 +38,13 @@ Note that the 2000 examples included in MNLI (and machine translated in French h
|
|
38 |
|
39 |
### Data Fields
|
40 |
|
41 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
|
43 |
### Data Splits
|
44 |
|
|
|
38 |
|
39 |
### Data Fields
|
40 |
|
41 |
+
- `premise`: The machine translated premise in the target language.
|
42 |
+
- `hypothesis`: The machine translated premise in the target language.
|
43 |
+
- `label`: The classification label, with possible values `entailment` (0), `neutral` (1), `contradiction` (2).
|
44 |
+
- `pairID`: Unique identifier for pair.
|
45 |
+
- `promptID`: Unique identifier for prompt.
|
46 |
+
- `premise_original`: The original premise from the English source dataset.
|
47 |
+
- `hypothesis_original`: The original hypothesis from the English source dataset.
|
48 |
|
49 |
### Data Splits
|
50 |
|