Commit
·
adc1b80
1
Parent(s):
a24c517
Initial upload of Wikimedia paragraphs dataset
Browse files- README.md +14 -0
- wikimedia_paragraphs.parquet +3 -0
README.md
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# My Wikimedia Dataset
|
| 2 |
+
|
| 3 |
+
This dataset contains paragraphs extracted from Wikimedia Wikipedia texts. Each entry includes:
|
| 4 |
+
|
| 5 |
+
- `paragraph_text`: Text of the paragraph
|
| 6 |
+
- `original_url`: URL of the source article
|
| 7 |
+
- `original_id`: ID of the source article
|
| 8 |
+
- `paragraph_number`: Paragraph number within the original text
|
| 9 |
+
- `combined_id`: Unique hash ID for the paragraph
|
| 10 |
+
- `length`: Paragraph length in characters
|
| 11 |
+
- `length_tokens`: Paragraph length in tokens (Gemma tokenizer)
|
| 12 |
+
|
| 13 |
+
Format: Parquet (zstd compression).
|
| 14 |
+
|
wikimedia_paragraphs.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dcb17b3739b6df6ce8f25cf1841733b4af7615cd9d93fcd756519d8577c03db3
|
| 3 |
+
size 14729671727
|