task: mbpp_instruct dataset_path: google-research-datasets/mbpp dataset_name: full unsafe_code: true output_type: generate_until test_split: test doc_to_text: !function utils.doc_to_text doc_to_target: "{{test_list[1]}}\n{{test_list[2]}}" gen_prefix: "\n```python\n" target_delimiter: "" metric_list: - metric: !function utils.pass_at_10 aggregation: mean higher_is_better: true k: [ 10 ] filter_list: - name: "extract_code" filter: - function: "custom" filter_fn: !function utils.build_predictions repeats: 20 generation_kwargs: max_gen_toks: 256 until: [ "\nclass", "\nassert", '\n"""', "\nprint", "\nif", "\n```", "\n#", "\n<|/", "<|eot_id|>", ] do_sample: true temperature: 0.8 top_p: 0.95 num_fewshot: 0 fewshot_config: sampler: first_n samples: !function utils.list_fewshot_samples metadata: version: 1.0