# file: /home/runner/work/aelfrice/aelfrice/benchmarks/amabench_adapter.py
# hypothesis_version: 6.152.4

[0.0, 34.36, 44.8, 46.06, 57.22, 72.26, 1000, 2000, '--budget', '--domain', '--max-episodes', '--output', '--qa-type', '--retrieve-only', 'A', 'AMA-Agent', 'AMA-bench/AMA-bench', 'B', 'BM25 (Qwen3-32B)', 'C', 'Causal Inference', 'D', 'GPT 5.2 long-context', 'HippoRAG2', 'MemoRAG', 'PATH', 'State Abstraction', 'State Updating', '__main__', '_gt', 'accuracy_by_domain', 'accuracy_by_type', 'action', 'amabench', 'amabench-benchmark', 'amabench_', 'answer', 'context', 'count', 'domain', 'domain_counts', 'episode_id', 'exact_match', 'f1', 'num_turns', 'observation', 'per_question', 'qa_pairs', 'qa_type', 'qa_type_name', 'question', 'question_uuid', 'success', 'task', 'task_type', 'test', 'total_episodes', 'total_qa', 'total_tokens', 'trajectory', 'turn_idx', 'type', 'type_counts', 'unknown', 'utf-8', 'w']