Update README.md
Browse files
README.md
CHANGED
|
@@ -37,3 +37,24 @@ configs:
|
|
| 37 |
- split: train
|
| 38 |
path: data/train-*
|
| 39 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
- split: train
|
| 38 |
path: data/train-*
|
| 39 |
---
|
| 40 |
+
|
| 41 |
+
## Dataset description
|
| 42 |
+
|
| 43 |
+
This is a small sample of []() for testing purposes. It was obtained as:
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
```python
|
| 47 |
+
import duckdb
|
| 48 |
+
from datasets import Dataset
|
| 49 |
+
|
| 50 |
+
con = duckdb.connect()
|
| 51 |
+
con.execute("INSTALL httpfs;")
|
| 52 |
+
con.execute("LOAD httpfs;")
|
| 53 |
+
|
| 54 |
+
url = r"https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus/resolve/main/fineweb-edu-dedup/train-00000-of-00234.parquet"
|
| 55 |
+
result = con.sql(f"SELECT * FROM '{url}' LIMIT(20)")
|
| 56 |
+
|
| 57 |
+
ds = Dataset.from_pandas(result.df(), preserve_index=False)
|
| 58 |
+
ds.push_to_hub("plaguss/fineweb-edu-dedup-testing-sample")
|
| 59 |
+
```
|
| 60 |
+
|