metadata
language:
- ka
license: gpl-3.0
size_categories:
- n<1K
task_categories:
- translation
- text2text-generation
- sentence-similarity
pretty_name: Svan-Georgian parallel corpora
dataset_info:
features:
- name: verse
dtype: string
- name: sva
dtype: string
- name: kat
dtype: string
- name: line_id
dtype: int64
- name: text_id
dtype: int64
- name: text_source
dtype: string
splits:
- name: train
num_bytes: 1297709
num_examples: 6270
download_size: 391580
dataset_size: 1297709
configs:
- config_name: default
data_files:
- split: train
path: data/train-*