Datasets:

Graimond
/

ECBERT-idioms-dataset

dataset_info:
  features:
    - name: idiom
      dtype: string
    - name: mask
      dtype: string
  splits:
    - name: train
      num_bytes: 3961
      num_examples: 100
  download_size: 3637
  dataset_size: 3961
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Dataset description

This is the digitalized version of "Appendix 1: Masked idiom dataset" from Leonardo Gambacorta, Byeungchun Kwon, Taejin Park, Pietro Patelli, and Sonya Zhu. CB-LMs: Language Models for Central Banking. Bank for International Settlements, BIS Working Papers No. 1215, October 2024. Available at: https://www.bis.org/publ/work1215.pdf.