Datasets:
metadata
language:
- ar
- en
task_categories:
- translation
dataset_info:
features:
- name: en
dtype: string
- name: ar
dtype: string
splits:
- name: test
num_bytes: 1936892.4714943743
num_examples: 5298
- name: train
num_bytes: 9683000
num_examples: 26486
download_size: 16774039
dataset_size: 11619892.471494375
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
Source Data
- ImruQays: Quran Parallel Texts
- Shamela Books: Shamela Open Library
Limitations
- Translation Quality: The translation of Shamela Books was done using Google Translate, which showed some inaccuracies with complex classical Arabic terms.