Datasets:
metadata
task_categories:
- translation
multilinguality:
- translation
task_ids: []
language:
- af
- ar
- be
- bp
- bs
- cs
- da
- de
- el
- es
- et
- fa
- fi
- fr
- he
- hu
- id
- it
- lt
- lv
- 'no'
- pl
- pt
- ro
- ru
- sq
- sr
- sv
- tr
- uk
size_categories:
- 1K<n<10K
configs:
- config_name: en-cs
data_files:
- split: test
path: data/en-cs.csv
- config_name: en-et
data_files:
- split: test
path: data/en-et.csv
- config_name: en-hu
data_files:
- split: test
path: data/en-hu.csv
- config_name: en-no
data_files:
- split: test
path: data/en-no.csv
- config_name: en-ru
data_files:
- split: test
path: data/en-ru.csv
- config_name: en-uk
data_files:
- split: test
path: data/en-uk.csv
- config_name: en-af
data_files:
- split: test
path: data/en-af.csv
- config_name: en-da
data_files:
- split: test
path: data/en-da.csv
- config_name: en-fa
data_files:
- split: test
path: data/en-fa.csv
- config_name: en-id
data_files:
- split: test
path: data/en-id.csv
- config_name: en-pl
data_files:
- split: test
path: data/en-pl.csv
- config_name: en-sq
data_files:
- split: test
path: data/en-sq.csv
- config_name: en-ar
data_files:
- split: test
path: data/en-ar.csv
- config_name: en-de
data_files:
- split: test
path: data/en-de.csv
- config_name: en-fi
data_files:
- split: test
path: data/en-fi.csv
- config_name: en-it
data_files:
- split: test
path: data/en-it.csv
- config_name: en-pt_br
data_files:
- split: test
path: data/en-pt_br.csv
- config_name: en-sr
data_files:
- split: test
path: data/en-sr.csv
- config_name: en-be
data_files:
- split: test
path: data/en-be.csv
- config_name: en-el
data_files:
- split: test
path: data/en-el.csv
- config_name: en-fr
data_files:
- split: test
path: data/en-fr.csv
- config_name: en-lt
data_files:
- split: test
path: data/en-lt.csv
- config_name: en-pt
data_files:
- split: test
path: data/en-pt.csv
- config_name: en-sv
data_files:
- split: test
path: data/en-sv.csv
- config_name: en-bs
data_files:
- split: test
path: data/en-bs.csv
- config_name: en-es
data_files:
- split: test
path: data/en-es.csv
- config_name: en-he
data_files:
- split: test
path: data/en-he.csv
- config_name: en-lv
data_files:
- split: test
path: data/en-lv.csv
- config_name: en-ro
data_files:
- split: test
path: data/en-ro.csv
- config_name: en-tr
data_files:
- split: test
path: data/en-tr.csv
- config_name: en-uk
data_files:
- split: test
path: data/en-uk.csv
license: cc-by-nc-sa-4.0
LTR
LTR -- Lingvanex Test References for MT Evaluation from English into a total of 30 target languages for a big variety of cases.
TEST CASES
Parameter | Description |
---|---|
Length | Sentences from 1 to 100 words. |
Domain | Medicine (12%), Automobile (11%), Finance (8%) |
Tokenizer | Jupiter is 1.000.000 km far. Ask Mr. Johnson for training |
Tags | I want to eat and swim |
Capitalisation (Case) | HELLO my Dear frIEND |
Different languages in one text (Up to 3 languages) | I see "Купалинка" performance near the theater. |
Styling | Hello Dude! |
Errors (Grammar, OCR) | I neet (need) to buy a kat (cat) |
Abbreviations | The model was named 15.BVcX-10 |
Named Entities | Let’s go to New York city |
Idioms | A piece of cake. Once in a blue moon’ |
Formulas (Math, Physics, Chemistry) | Cr2(SO4)3 + CO2 + H2O + K2SO4 + KNO3 |
Romanian numbers | It was in MCMXVII year |
Unicode Special Characters | №%&*/# |
Updates
Data package has been released on 11/18/2023. Enjoy!
License
The LTR data set is released under the CC BY-SA 4.0 license.
How to Cite
@inproceedings{
title = "{LTR} Lingvanex Test References for {MT} Evaluation",
author = "Aliaksei Rudak",
url = "https://lingvanex.com",
Year = "2023"
}