HELLASWAG_DE

NAME = HELLASWAG_DE
DATASET_PATH = LeoLM/HellaSwag_de
SAMPLE_SPLIT = validation
FEWSHOT_SPLIT = train
RESPONSE_TYPE = LOGLIKELIHOODS
METRICS = [AccuracyLoglikelihood, AccuracyNormLoglikelihood]
SUBJECTS = ['no_subject']
LANGUAGE = <Language.DEU: 'German'>

• Module: eval_framework.tasks.benchmarks.hellaswag_de

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

• Link to dataset: https://huggingface.co/datasets/LeoLM/HellaSwag_de

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 "HELLASWAG_DE".