Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,17 @@
|
|
1 |
-
---
|
2 |
-
license: odc-by
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: odc-by
|
3 |
+
---
|
4 |
+
# Fineweb-edu Sentences
|
5 |
+
|
6 |
+
**Description:**
|
7 |
+
A dataset of sentences collected from the web.
|
8 |
+
The dataset was created by splitting the text into individual sentences using the [spaCy package](https://spacy.io/),
|
9 |
+
then removing duplicates and filtering for complete sentences in a semi-automated process.
|
10 |
+
|
11 |
+
**Source:** [HuggingFaceFW/fineweb-edu](https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu)
|
12 |
+
|
13 |
+
**Size:** About 700,000 English language sentences. Each sentence is 512 tokens long or less as assessed using the BERT tokenizer.
|
14 |
+
|
15 |
+
**Annotations:** The `source` field contains the URL of each sentence from FineWeb-Edu.
|
16 |
+
|
17 |
+
**License:** Open Data Commons License Attribution (same as HuggingFaceFW/fineweb-edu)
|