peacock-data-public-datasets-idc-llm_eval
/
lm-evaluation
/build
/lib
/lm_eval
/tasks
/hellaswag
/hellaswag.yaml
| group: | |
| - multiple_choice | |
| task: hellaswag | |
| dataset_path: hellaswag | |
| dataset_name: null | |
| output_type: multiple_choice | |
| training_split: train | |
| validation_split: validation | |
| test_split: null | |
| process_docs: !function utils.process_docs | |
| doc_to_text: "{{query}}" | |
| doc_to_target: "{{label}}" | |
| doc_to_choice: "choices" | |
| metric_list: | |
| - metric: acc | |
| aggregation: mean | |
| higher_is_better: true | |
| - metric: acc_norm | |
| aggregation: mean | |
| higher_is_better: true | |
| metadata: | |
| version: 1.0 | |