README.md

metadata

dataset_info:
  features:
    - name: sq
      dtype: string
    - name: en
      dtype: string
    - name: dataset
      dtype: string
    - name: subset
      dtype: string
  splits:
    - name: train
      num_bytes: 521626304
      num_examples: 3192496
  download_size: 242504356
  dataset_size: 521626304
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: gpl-3.0
task_categories:
  - translation
  - fill-mask
language:
  - sq
  - en
size_categories:
  - 1M<n<10M

Albanian-English bundled dataset

This dataset is created by normalizing and bundling the following datasets:

Dataset	Subset
aiana94/polynews-parallel	eng_Latn-sqi_Latn
ayymen/Pontoon-Translations	en-sq
ayymen/Weblate-Translations	en_devel-sq_al
ayymen/Weblate-Translations	en_devel-sq
ayymen/Weblate-Translations	en_GB-sq
ayymen/Weblate-Translations	en-sq_AL
ayymen/Weblate-Translations	en-sq-rAL
ayymen/Weblate-Translations	en-sq
ayymen/Weblate-Translations	en_US-sq
davidstap/ted_talks	sq_en
Helsinki-NLP/bible_para	lang1="en", lang2="sq"
Helsinki-NLP/open_subtitles	lang1="en", lang2="sq"
Helsinki-NLP/qed_amara	lang1="en", lang2="sq"
Helsinki-NLP/tanzil	lang1="en", lang2="sq"
Helsinki-NLP/tatoeba	lang1="en", lang2="sq"
Helsinki-NLP/tilde_model	lang1="en", lang2="sq"
Helsinki-NLP/opus-100	en-sq
ted_talks_iwslt	language_pair=("sq", "en") year="2014"
ted_talks_iwslt	language_pair=("sq", "en") year="2015"
ted_talks_iwslt	language_pair=("sq", "en") year="2016"