| 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). | |