Datasets:
metadata
language:
- ar
- en
task_categories:
- translation
dataset_info:
features:
- name: source
dtype: string
- name: target
dtype: string
splits:
- name: train
num_bytes: 35638442.55816875
num_examples: 91084
download_size: 20087251
dataset_size: 35638442.55816875
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
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.