Back to search
ipython notebooks do some sample experiments , make some idea
Stars
24
Forks
8
Watchers
24
Open Issues
0
Overall repository health assessment
No package.json found
This might not be a Node.js project
adaptive RAG 主要是用LLM作为路由,判别器(好、坏), 本质上是使用llm作为分类任务在处理逻辑分支,利用搜索(深度搜索),让其自适应(self-reflection)获取结果。 但是处理时间会长些,如果遇到badcase, 就会一直自适应的判定循环,直到判定次数上限退出。
25aa114View on GitHub377
commits