def get_reward_model_scores()

in cached_classes.py [0:0]

• 11 lines of code
• 4 McCabe index (conditional complexity)

    def get_reward_model_scores(self, prompt: List[Dict[str, str]], completions: List[List[Dict[str, str]]]) -> List[float]:
        completion_rewards = []
        for idx_c, completion in enumerate(completions):
            lookup_hash = (to_string(prompt), to_string(completion))
            if lookup_hash in self.train_data:
                completion_rewards.append(self.train_data[lookup_hash])
            elif lookup_hash in self.test_data:
                completion_rewards.append(self.test_data[lookup_hash])
            else:
                raise ValueError(f"No cached reward model score found for prompt: {prompt}, completion: {idx_c}, {completion}")

        return completion_rewards