Datasets:
metadata
language:
- en
- vi
size_categories:
- 1M<n<10M
task_categories:
- translation
dataset_info:
features:
- name: vi
dtype: string
- name: en
dtype: string
splits:
- name: train
num_bytes: 1169167942.9334285
num_examples: 3750006
- name: test
num_bytes: 129907757.06657156
num_examples: 416668
download_size: 834752619
dataset_size: 1299075700
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
This dataset is cloned from another dataset.
We saw that the samples in this dataset is augmented from ~4.2M to ~8.3M samples to allow both En-Vi and Vi-En translation.
However, we only want En-Vi translation, so we remove duplication in this dataset.
The original version belongs to VietAI but it is unable to accept now. You can read the paper about the dataset here
for more information.