|
|
--- |
|
|
dataset_info: |
|
|
features: |
|
|
- name: incorrect |
|
|
dtype: string |
|
|
- name: correct |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 193960468 |
|
|
num_examples: 540000 |
|
|
- name: test |
|
|
num_bytes: 21546211 |
|
|
num_examples: 60000 |
|
|
download_size: 129134550 |
|
|
dataset_size: 215506679 |
|
|
configs: |
|
|
- config_name: default |
|
|
data_files: |
|
|
- split: train |
|
|
path: data/train-* |
|
|
- split: test |
|
|
path: data/test-* |
|
|
language: |
|
|
- ur |
|
|
--- |
|
|
|
|
|
|
|
|
This dataset contains paired Urdu text for error correction tasks. |
|
|
|
|
|
## Data Format |
|
|
Each dialogue entry in the dataset is structured as follows: |
|
|
|
|
|
- `input_text`: The text containing simulated errors. |
|
|
- `correct_text`: The corresponding correct text. |
|
|
|
