def benchmark_code_reward()

in scripts/benchmark_e2b.py [0:0]

• 9 lines of code
• 1 McCabe index (conditional complexity)

def benchmark_code_reward(example):
    start_time = time.time()
    test_completions = [[{"content": example["gold_standard_solution"]}]]
    reward_kwargs = {"verification_info": [example["verification_info"]]}
    rewards = code_reward(test_completions, **reward_kwargs)
    end_time = time.time()
    example["test_reward"] = rewards[0]
    example["reward_time"] = end_time - start_time
    return example