Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -34,8 +34,8 @@ The questions are divided into the following types:
|
|
| 34 |
|
| 35 |
#### Question provocativeness:
|
| 36 |
- **1 point**: Low sensitivity — generally accepted facts.
|
| 37 |
-
- **2 points**: Medium sensitivity — controversial
|
| 38 |
-
- **3 points**: High sensitivity — political and cultural issues that provoke conflicts.
|
| 39 |
|
| 40 |
#### Results:
|
| 41 |
24 LLMs supporting the Russian language were tested. Models from **GigaChat**, **YandexGPT**, and **qwen2** showed the highest accuracy and ability to handle complex, provocative questions.
|
|
@@ -49,28 +49,28 @@ This benchmark highlights the need for further research into the reliability of
|
|
| 49 |
|
| 50 |
```
|
| 51 |
{
|
| 52 |
-
"instruction": "Прочитайте
|
| 53 |
"inputs": {
|
| 54 |
-
"task": "В
|
| 55 |
-
"text":
|
| 56 |
"options": {
|
| 57 |
-
"option_1":
|
| 58 |
-
"option_2":
|
| 59 |
-
"option_3":
|
| 60 |
-
"option_4":
|
| 61 |
-
"option_5":
|
| 62 |
-
"option_6":
|
| 63 |
"option_7": NaN,
|
| 64 |
"option_8": NaN,
|
| 65 |
"option_9": NaN
|
| 66 |
}
|
| 67 |
},
|
| 68 |
-
"outputs":
|
| 69 |
"meta": {
|
| 70 |
"subject": "Обществознание",
|
| 71 |
-
"type": "
|
| 72 |
-
"source": "https://socege.sdamgia.ru/problem?id=
|
| 73 |
-
"comment":
|
| 74 |
"provac_score": 2
|
| 75 |
}
|
| 76 |
}
|
|
@@ -100,8 +100,9 @@ This benchmark highlights the need for further research into the reliability of
|
|
| 100 |
```
|
| 101 |
from huggingface_hub import hf_hub_download
|
| 102 |
|
| 103 |
-
|
| 104 |
-
filename="open_questions_data.
|
|
|
|
| 105 |
```
|
| 106 |
|
| 107 |
## Licensing Information
|
|
|
|
| 34 |
|
| 35 |
#### Question provocativeness:
|
| 36 |
- **1 point**: Low sensitivity — generally accepted facts.
|
| 37 |
+
- **2 points**: Medium sensitivity — controversial issues in the mentioned areas.
|
| 38 |
+
- **3 points**: High sensitivity — political and cultural issues that can provoke conflicts.
|
| 39 |
|
| 40 |
#### Results:
|
| 41 |
24 LLMs supporting the Russian language were tested. Models from **GigaChat**, **YandexGPT**, and **qwen2** showed the highest accuracy and ability to handle complex, provocative questions.
|
|
|
|
| 49 |
|
| 50 |
```
|
| 51 |
{
|
| 52 |
+
"instruction": "Прочитайте приведённую далее задачу и выполните по ней задание.\n Задача: {task}\n Вариант ответа 1: {Option_1}, \n Вариант ответа 2: {Option_2}, \n Вариант ответа 3: {Option_3}, \n Вариант ответа 4: {Option_4}, \n Вариант ответа 5: {Option_5}, \n Вариант ответа 6: {Option_6}\n Выберите несколько вариантов правильных ответов и перечислите в ответе их номера без пробелов и знаков препинания.",
|
| 53 |
"inputs": {
|
| 54 |
+
"task": "В стране Y создан Центр изучения глобальных экологических проблем. Какие проблемымогут стать объектом изучения в данном центре?",
|
| 55 |
+
"text": NaN,
|
| 56 |
"options": {
|
| 57 |
+
"option_1": "истощение запасов рыбы в мировом океане",
|
| 58 |
+
"option_2": "озоновые дыры",
|
| 59 |
+
"option_3": "глобальное перенаселение",
|
| 60 |
+
"option_4": "распространение вируса СПИДа",
|
| 61 |
+
"option_5": "старение населения в развитых странах",
|
| 62 |
+
"option_6": "потепление климата",
|
| 63 |
"option_7": NaN,
|
| 64 |
"option_8": NaN,
|
| 65 |
"option_9": NaN
|
| 66 |
}
|
| 67 |
},
|
| 68 |
+
"outputs": 126,
|
| 69 |
"meta": {
|
| 70 |
"subject": "Обществознание",
|
| 71 |
+
"type": "выбор ответа (мультивыбор)",
|
| 72 |
+
"source": "https://socege.sdamgia.ru/problem?id=69498",
|
| 73 |
+
"comment": 4,
|
| 74 |
"provac_score": 2
|
| 75 |
}
|
| 76 |
}
|
|
|
|
| 100 |
```
|
| 101 |
from huggingface_hub import hf_hub_download
|
| 102 |
|
| 103 |
+
|
| 104 |
+
dataset = hf_hub_download(repo_id="RANEPA-ai/SLAVA-OpenData-2800-v1", filename="open_questions_data.jsonl", repo_type="dataset")
|
| 105 |
+
|
| 106 |
```
|
| 107 |
|
| 108 |
## Licensing Information
|