File size: 1,919 Bytes
efaaf8c
44b3828
 
efaaf8c
 
85ce6b8
efaaf8c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c2d0432
 
 
 
 
 
 
efaaf8c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
---
language:
- en
license: cc-by-4.0
---
# LLM Benchmark for English-Cebuano Translation

This dataset contains parallel sentences of English and Cebuano extracted from the Bible corpus available at [https://github.com/christos-c/bible-corpus](https://github.com/christos-c/bible-corpus). The dataset is formatted for use in training machine translation models, particularly with the Transformers library from Hugging Face.

## Dataset Description

* **Source:** Bible corpus (https://github.com/christos-c/bible-corpus)
* **Languages:** English (en) and Cebuano (ceb)
* **Format:** CSV files with two columns: `english` and `cebuano`.
* **Extraction:** Sentences were extracted from the XML files inside the `<seg>` tags, and matched by their `id` attribute.
* **Purpose:** To create a benchmark for English-Cebuano machine translation in Large Language Models.

## Usage

This dataset can be used to evaluate the performance of Large Language Models for English-Cebuano machine translation using libraries like Transformers.

**Example (Loading with Pandas):**

```python
import pandas as pd

train_df = pd.read_csv("english-ceb-bible-prompt.csv")
print(train_df.head())

```

**Example (Loading with datsets):**
```python
from datasets import load_dataset

dataset = load_dataset("eemberda/english-ceb-bible-prompt")
```

## Notes
The quality of the sentence alignment may vary.
The dataset is intended for research and educational purposes.
Please cite the original Bible corpus if you use this dataset in your work.

## Citation
If you use this dataset in your research, please include the original Bible corpus in your citations:

```
@misc{bible-corpus,
  author = {Christos Christodoulopoulos},
  title = {Bible Corpus},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{[https://github.com/christos-c/bible-corpus](https://github.com/christos-c/bible-corpus)}}
}