The directory contains an information about a parallel corpus for English-Azerbaijani and Azerbaijani-English translation tasks. The main goal of introducing the data is to help engineers and developers to create a universal language model (fine-tuned models) for Azerbaijani language.
Read this in other languages: Azərbaycan dili
The corpus can be obtained by writing to rasul.karimov@skoltech.ru
or mammad.hajili@epfl.ch
. Just state in your letter how the obtained information will be used. Whether it is a private or public research. Also, it will be really helpful to give more information about the organization you are working in.
In case of some problems with availability of the dataset, twitter: @rkarrimov, @mammadhajili
File | #sentences | #words | #vocabulary |
---|---|---|---|
train.en | 68,201 | 1,205,183 | 42,315 |
train.az | 68,201 | 928,474 | 88,002 |
dev.en | 500 | 9,015 | 2,867 |
dev.az | 500 | 6,852 | 3,832 |
test.en | 500 | 9,032 | 2,941 |
test.az | 500 | 6,905 | 3,815 |