README.md

metadata

language:
  - eve
task_categories:
  - automatic-speech-recognition
size_categories:
  - 1K<n<10K

This dataset consists of audiofiles with a speech in Even language.

The correspondence between text and audio is in the table metadata.csv.

The data was collected from religious texts written down by Institute for Bible Translation

1. Һөвки Дукундукун укчэнэкэл. Institute for Bible Translation, Moscow, 2018.
2. Притчал. Institute for Bible Translation, Moscow, 2019.

Sourse	Dialect	Total length (min)
Religious texts	Lamunkhin	67.96

Another dataset of Even speech:

• field records of HSE University expedition: https://huggingface.co/datasets/tbkazakova/even_speech_hse

There is also unified data from the project (Aralova et al. 2007-2023)[1], but for copyright reasons, it can only be made available by personal agreement with all copyright holders.

[1]: Natalia Aralova, Brigitte Pakendorf, Alexandra Lavrillier, Dejan Matić, Katharina Gernet, Tat'jana Vasil'evna Zakharova, Raisa Petrovna Kuzmina, and Luise Zippel (2007 - 2023). Collection "Even". The Language Archive. https://hdl.handle.net/1839/07210104-91d6-4133-b067-b21eadc35f9a