metadata
dataset_info:
features:
- name: idiom
dtype: string
- name: meaning
dtype: string
splits:
- name: train
num_bytes: 83549
num_examples: 1624
download_size: 52424
dataset_size: 83549
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
This dataset is scraped from "Annes liste over talemåder" (https://www.overlevelsesguiden.dk/files/annes_liste_over_talemaader.pdf) Code to generate dataset can be found here: https://github.com/kasperjunge/danish-evals/blob/main/talemaader.py