GSM8KEvalHarness

NAME = GSM8KEvalHarness
DATASET_PATH = gsm8k
SAMPLE_SPLIT = test
FEWSHOT_SPLIT = train
RESPONSE_TYPE = COMPLETION
METRICS = [AccuracyCompletion]
SUBJECTS = ['main']
LANGUAGE = <Language.ENG: 'English'>

• Module: eval_framework.tasks.benchmarks.gsm8k

• File: src/eval_framework/tasks/benchmarks/gsm8k.py | View on GitHub

• Link to dataset: https://huggingface.co/datasets/gsm8k

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 "GSM8KEvalHarness".