Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -76,6 +76,18 @@ All data at [Fastcode](https://fastcode.im/Templates?TemplatesOnly=True) are onl
|
|
| 76 |
|
| 77 |
You can find in folder `scripts_parsing`!
|
| 78 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 79 |
## Citation
|
| 80 |
|
| 81 |
Please cite this dataset if you use it in your work:
|
|
|
|
| 76 |
|
| 77 |
You can find in folder `scripts_parsing`!
|
| 78 |
|
| 79 |
+
### ❗IMPORTANT❗: Escaping special characters
|
| 80 |
+
|
| 81 |
+
Values in columns `prompt`, `think_process`, and `solution` have escaped special characters like `\n`, `\t`, and `\\`!
|
| 82 |
+
|
| 83 |
+
The reason for this behavior: the Python's `csv` module that I'm using in the parser to save rows saves `\n` as an actual newline character,
|
| 84 |
+
which makes reading the prepared file with Pandas impossible.
|
| 85 |
+
|
| 86 |
+
To fix it use simple lambda code:
|
| 87 |
+
```python
|
| 88 |
+
return_back_special_symbols = lambda x: x.replace("\\n", "\n").replace("\\t", "\t").replace("\\\\", "\\")
|
| 89 |
+
```
|
| 90 |
+
|
| 91 |
## Citation
|
| 92 |
|
| 93 |
Please cite this dataset if you use it in your work:
|