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
doc_to_text: "{% set text = activity_label ~ ': ' ~ ctx_a ~ ' ' ~ ctx_b.capitalize() %}{{text|trim|replace(' [title]', '. ')|regex_replace('\\[.*?\\]', '')|replace('  ', ' ')}}"
doc_to_target: "{{label}}"
doc_to_choice: "{{endings|map('trim')|map('replace', ' [title]', '. ')|map('regex_replace', '\\[.*?\\]', '')|map('replace', '  ', ' ')|list}}"
metric_list:
  - metric: acc
    aggregation: mean
    higher_is_better: true
  - metric: acc_norm
    aggregation: mean
    higher_is_better: true