Commit
•
1a16188
1
Parent(s):
d4b220c
Update README.md (#1)
Browse files- Update README.md (f8b29cb3cf0ca38f656517ce8c50560902afa609)
Co-authored-by: Gokcimen <[email protected]>
README.md
CHANGED
|
@@ -21,3 +21,33 @@ configs:
|
|
| 21 |
- split: train
|
| 22 |
path: data/train-*
|
| 23 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
- split: train
|
| 22 |
path: data/train-*
|
| 23 |
---
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
# Spellchecker Dataset
|
| 27 |
+
|
| 28 |
+
This dataset contains correct and incorrect spellings of various words in Turkish. It is intended for use in training and evaluating spellchecking models.
|
| 29 |
+
|
| 30 |
+
## Dataset Structure
|
| 31 |
+
|
| 32 |
+
The dataset is structured in the following format:
|
| 33 |
+
- `correct`: The correct spelling of the word.
|
| 34 |
+
- `incorrect1`: A slightly misspelled version of the word.
|
| 35 |
+
- `incorrect2`: A moderately misspelled version of the word.
|
| 36 |
+
- `incorrect3`: A severely misspelled version of the word.
|
| 37 |
+
|
| 38 |
+
### Example
|
| 39 |
+
|
| 40 |
+
| correct | incorrect1 | incorrect2 | incorrect3 |
|
| 41 |
+
|----------|------------|------------|------------|
|
| 42 |
+
| bernard | bernard | bernrd | rnard |
|
| 43 |
+
| yemekten | yemekten | yemektn | yemekn |
|
| 44 |
+
|
| 45 |
+
## Usage
|
| 46 |
+
|
| 47 |
+
You can load this dataset using the `datasets` library from Hugging Face:
|
| 48 |
+
|
| 49 |
+
```python
|
| 50 |
+
from datasets import load_dataset
|
| 51 |
+
|
| 52 |
+
dataset = load_dataset("pnr-svc/spellchecker-dataset")
|
| 53 |
+
|