metadata

dataset_info:
  features:
    - name: index
      dtype: string
    - name: original_word
      dtype: string
    - name: correct_word
      dtype: string
  splits:
    - name: train
      num_bytes: 10270649
      num_examples: 152631
  download_size: 9173907
  dataset_size: 10270649
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - fill-mask
language:
  - az
tags:
  - spelling
  - localdoc
pretty_name: Spelling corrected words in Azerbaijani
size_categories:
  - 100K<n<1M

Spelling Corrected Words in Azerbaijani

Dataset Overview

This dataset, "Spelling Corrected Words in Azerbaijani," is designed for the task of correcting spelling errors in Azerbaijani texts. It contains pairs of words where each pair consists of an original word and its corrected version. The dataset is intended to be used for training and evaluating models that perform the task of filling in masked words correctly.

Dataset Structure

Columns

index: A unique identifier for each row.
original_word: The original word, which may contain spelling errors.
correct_word: The correct version of the word.

License

This dataset licensed under the CC BY-NC-ND 4.0 license. What does this license allow?

Attribution: You must give appropriate credit, provide a link to the license, and indicate if changes were made.
Non-Commercial: You may not use the material for commercial purposes.
No Derivatives: If you remix, transform, or build upon the material, you may not distribute the modified material.

For more information, please refer to the CC BY-NC-ND 4.0 license.

Contact

For more information, questions, or issues, please contact LocalDoc at [[email protected]].