haripritam
commited on
Commit
•
73c3769
1
Parent(s):
531cbed
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,39 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
---
|
4 |
+
# Telugu GK Questions Dataset
|
5 |
+
|
6 |
+
## Overview
|
7 |
+
|
8 |
+
This dataset consists of General Knowledge (GK) questions scraped from the Telugu Tech Badi website.
|
9 |
+
A separate data cleaning script refines the extracted questions for better readability and analysis.
|
10 |
+
|
11 |
+
## Tasks
|
12 |
+
|
13 |
+
### Task
|
14 |
+
- **Objective:** Extract GK questions from a list of URLs.
|
15 |
+
- **Challenges:** Some of the URLs follow a different format than others, so modify the code for specific URLs.
|
16 |
+
- **Colab Notebook:** Modifying the .jsonl file is not in the Colab as there may be other ways. [Link](https://colab.research.google.com/drive/1m1I4kV_UFccm76fE2MWB0rLJ9IzoQZfX?usp=sharing)
|
17 |
+
|
18 |
+
## Data Cleaning
|
19 |
+
|
20 |
+
### Script
|
21 |
+
- **Objective:** Refine the extracted questions for better readability and analysis. This includes removing any indicators such as "Question No," etc.
|
22 |
+
- Read the JSON Lines files from the scraping tasks and create a new file with cleaned data.
|
23 |
+
- Verify data for any errors or null values.
|
24 |
+
|
25 |
+
## Data Format
|
26 |
+
|
27 |
+
The data is saved in JSON Lines format, where each line represents a JSON object with the following structure:
|
28 |
+
|
29 |
+
```json
|
30 |
+
{
|
31 |
+
"input": "Question text in Telugu",
|
32 |
+
"instruction": "Instructions in Telugu",
|
33 |
+
"output": "Answer text in Telugu"
|
34 |
+
}
|
35 |
+
```
|
36 |
+
|
37 |
+
## Source
|
38 |
+
|
39 |
+
GK Questions from - [telugutechbadi](https://www.telugutechbadi.com/)
|