---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: option_a
    dtype: string
  - name: option_b
    dtype: string
  - name: option_c
    dtype: string
  - name: option_d
    dtype: string
  - name: answer
    dtype: string
  - name: id
    dtype: int64
  splits:
  - name: train
    num_bytes: 295127
    num_examples: 1588
  download_size: 124202
  dataset_size: 295127
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

QA dataset for testing LLM's ability to understand the meaning of Danish idioms. The dataset is based on [Juunge/danske-talemaader](https://huggingface.co/datasets/Juunge/danske-talemaader).