Update README.md
Browse files
README.md
CHANGED
|
@@ -21,4 +21,21 @@ configs:
|
|
| 21 |
path: data/train-*
|
| 22 |
- split: test
|
| 23 |
path: data/test-*
|
|
|
|
|
|
|
| 24 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
path: data/train-*
|
| 22 |
- split: test
|
| 23 |
path: data/test-*
|
| 24 |
+
language:
|
| 25 |
+
- ur
|
| 26 |
---
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
# Urdu Error Correction Dataset
|
| 30 |
+
|
| 31 |
+
This dataset contains paired Urdu text for error correction tasks.
|
| 32 |
+
|
| 33 |
+
## Dataset Details
|
| 34 |
+
|
| 35 |
+
- **Splits:**
|
| 36 |
+
- **Train:** 540,000 examples
|
| 37 |
+
- **Test:** 60,000 examples
|
| 38 |
+
|
| 39 |
+
- **Columns:**
|
| 40 |
+
- `input_text`: The text containing simulated errors.
|
| 41 |
+
- `correct_text`: The corresponding correct text.
|