MBA
MBA copied to clipboard
FUTUREEEEEE
COLING 2025: MBA-RAG: a Bandit Approach for Adaptive Retrieval-Augmented Generation through Question Complexity
开源代码里的 reward： # Environment._get_reward() if method in answer: if method == 'zero': return reward_zero # 1 elif method == 'one': return reward_one # 3 elif method == 'multiple': return reward_multiple...
Hello author, when I was reproducing your MBA-RAG, in your **MBA-RAG/classifier/postprocess/predict_complexity_on_classification_results.py**, there is a **stepNum_result_file** that records the steps. How was this file obtained? Can you provide it within the...
Metadata
Owner
Metadata
COLING 2025: MBA-RAG: a Bandit Approach for Adaptive Retrieval-Augmented Generation through Question Complexity