Create README.md
Browse files
README.md
CHANGED
@@ -1,20 +1,18 @@
|
|
1 |
-
|
2 |
-
dataset_info:
|
3 |
-
features:
|
4 |
-
- name: text
|
5 |
-
dtype: string
|
6 |
-
splits:
|
7 |
-
- name: train
|
8 |
-
num_bytes: 7136.413530148701
|
9 |
-
num_examples: 24
|
10 |
-
download_size: 1864
|
11 |
-
dataset_size: 7136.413530148701
|
12 |
-
configs:
|
13 |
-
- config_name: default
|
14 |
-
data_files:
|
15 |
-
- split: train
|
16 |
-
path: data/train-*
|
17 |
-
---
|
18 |
-
# Dataset Card for "wikitext-tiny"
|
19 |
|
20 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
This dataset is sampled from `wikitext/wikitext-2-v1/train`.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
+
Codes to generate this dataset:
|
4 |
+
|
5 |
+
```python
|
6 |
+
import datasets
|
7 |
+
dataset = datasets.load_dataset('wikitext', 'wikitext-2-v1')
|
8 |
+
|
9 |
+
selected = []
|
10 |
+
i = -1
|
11 |
+
while len(selected) < 24:
|
12 |
+
i += 1
|
13 |
+
text = dataset['train'][i]['text']
|
14 |
+
if 8 < len(text.split(' ')) <= 16 and '=' not in text:
|
15 |
+
selected.append(i)
|
16 |
+
|
17 |
+
tiny_dataset = dataset['train'].select(selected)
|
18 |
+
```
|