agentlans commited on
Commit
7c30a4b
·
verified ·
1 Parent(s): f6a3c6b

Upload 3 files

Browse files
Files changed (3) hide show
  1. README.md +52 -3
  2. test.txt.gz +3 -0
  3. train.txt.gz +3 -0
README.md CHANGED
@@ -1,3 +1,52 @@
1
- ---
2
- license: odc-by
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ## High-Quality English Sentences
2
+
3
+ ### Dataset Description
4
+
5
+ This dataset contains a collection of high-quality English sentences sourced from [C4](https://huggingface.co/datasets/allenai/c4) and [FineWeb](https://huggingface.co/datasets/HuggingFaceFW/fineweb) (*not* [FineWeb-Edu](https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu)). The sentences have been carefully filtered and processed to ensure quality and uniqueness.
6
+
7
+ ### Source Data
8
+
9
+ Before filtering:
10
+ - **C4:** 1 million sentences
11
+ - **FineWeb:** 1 million sentences
12
+
13
+ ### Data Processing
14
+
15
+ 1. **Initial Filtering:**
16
+ - Quality score > 0.5 with [agentlans/deberta-v3-xsmall-quality](https://huggingface.co/agentlans/deberta-v3-xsmall-quality) classifier
17
+ - Sentence length >= 20 characters
18
+
19
+ 2. **Additional Filtering:**
20
+ - Removed sentences not starting with a capital letter
21
+ - Removed sentences with unmatched parentheses or brackets
22
+
23
+ 3. **Deduplication:**
24
+ - Combined sentences from both sources
25
+ - Filtered for unique sentences
26
+
27
+ 4. **Train-Test Split:**
28
+ - Train set: 90% of data
29
+ - Test set: 10% of data
30
+
31
+ ### Data Fields
32
+
33
+ No data field. Just one sentence per line as a string.
34
+
35
+ ### Data Splits
36
+
37
+ | Split | Number of Sentences |
38
+ |---------|---------------------|
39
+ | Train | 1 534 699 |
40
+ | Test | 170 522 |
41
+ | Total | 1 705 221 |
42
+
43
+ ### Dataset Creation
44
+
45
+ - **Curation Rationale:** To create a high-quality dataset of diverse English sentences for various NLP tasks.
46
+ - **Source Contributions:** Equal initial contributions from C4 and FineWeb, with subsequent filtering and processing.
47
+
48
+ ### Considerations for Using the Data
49
+
50
+ - The dataset deliberately uses FineWeb and C4 instead of FineWeb-Edu in order to broaden beyond educational and academic content. However, it's still biased towards those genres.
51
+ - Titles and citations are considered "sentences".
52
+ - The filtering process may have introduced some biases, such as favouring more formal writing styles and omitting lists due to the capitalization requirement.
test.txt.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70f1b4e0ea23de2d3e2060372587512906c37b120747f31a2d4373d868850923
3
+ size 9488687
train.txt.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edbcece9d0236c149f872b1098a1cb952e43a61bfc6cde0c42b44adbeb594406
3
+ size 85467066