ZERO_SCROLLS_GOV_REPORT

NAME = ZERO_SCROLLS_GOV_REPORT
DATASET_PATH = tau/zero_scrolls
SAMPLE_SPLIT = validation
FEWSHOT_SPLIT = validation
RESPONSE_TYPE = COMPLETION
METRICS = [ROUGE_GEOMETRIC_MEAN]
SUBJECTS = ['gov_report']

• Module: eval_framework.tasks.benchmarks.zero_scrolls

• File: src/eval_framework/tasks/benchmarks/zero_scrolls.py | View on GitHub

• Link to dataset: https://huggingface.co/datasets/tau/zero_scrolls

More detailed documentation, with prompt examples and ground truth completions, can be generated with uv run -m eval_framework.utils.generate_task_docs --add-prompt-examples --only-tasks "ZERO_SCROLLS_GOV_REPORT".