README.md

metadata

language:
  - ru
license: mit
dataset_info:
  features:
    - name: word
      dtype: string
    - name: definition
      dtype: string
    - name: type
      dtype: string
  splits:
    - name: train
      num_bytes: 73977
      num_examples: 540
  download_size: 34766
  dataset_size: 73977
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

This is a small testing dataset for Russian-language reverse dictionaries. It consists of two parts:

• there are 360 word-definition pairs taken from a Russian explanatory dictionary - Efremova's New Explanatory-Morphological Dictionary (2000) in this case;
• there are also 180 words with descriptions written and validated by native Russian speakers, written in a more natural way than lexicographical definitions.

The training dataset is available here for now: https://huggingface.co/datasets/snagbreac/russian-reverse-dictionary-train-data .