Add the `subject` field
Browse files- README.md +12 -3
- train.jsonl +0 -0
README.md
CHANGED
|
@@ -4,8 +4,8 @@ license: mit
|
|
| 4 |
|
| 5 |
# ZNO dataset
|
| 6 |
|
| 7 |
-
This dataset contains
|
| 8 |
-
|
| 9 |
|
| 10 |
Question subjects are:
|
| 11 |
|
|
@@ -30,7 +30,8 @@ Every line in a .jsonl file contains a structure like this:
|
|
| 30 |
{ "marker": "Г", "text": "загадка" },
|
| 31 |
{ "marker": "Д", "text": "русло" }
|
| 32 |
],
|
| 33 |
-
"correct_answers": ["Д"]
|
|
|
|
| 34 |
}
|
| 35 |
|
| 36 |
```
|
|
@@ -39,6 +40,14 @@ Currently, all questions have exactly one correct answer, stored in
|
|
| 39 |
`correct_answers[0]`.
|
| 40 |
|
| 41 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
## Pre/postprocessing
|
| 43 |
|
| 44 |
* Question text is converted to Markdown.
|
|
|
|
| 4 |
|
| 5 |
# ZNO dataset
|
| 6 |
|
| 7 |
+
This dataset contains machine-readable questions and answers from Ukrainian
|
| 8 |
+
External independent testing (called _ЗНО_/_ZNO_ in Ukrainian).
|
| 9 |
|
| 10 |
Question subjects are:
|
| 11 |
|
|
|
|
| 30 |
{ "marker": "Г", "text": "загадка" },
|
| 31 |
{ "marker": "Д", "text": "русло" }
|
| 32 |
],
|
| 33 |
+
"correct_answers": ["Д"],
|
| 34 |
+
"subject": "ukrainian-language-and-literature"
|
| 35 |
}
|
| 36 |
|
| 37 |
```
|
|
|
|
| 40 |
`correct_answers[0]`.
|
| 41 |
|
| 42 |
|
| 43 |
+
## Dataset structure
|
| 44 |
+
|
| 45 |
+
| Subject | Subset | Size |
|
| 46 |
+
|-------------------------------------|--------|----------------|
|
| 47 |
+
|ukrainian-language-and-literature | train | 1925 questions |
|
| 48 |
+
|history-of-ukraine | train | 1138 questions |
|
| 49 |
+
|
| 50 |
+
|
| 51 |
## Pre/postprocessing
|
| 52 |
|
| 53 |
* Question text is converted to Markdown.
|
train.jsonl
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|