single-shot RAG에서 agentic RAG로 — 내 블로그 챗봇을 다시 짰다한 번 검색해서 끝내던 RAG 챗봇을 도구 8개랑 max 5 iter 루프 가진 agentic으로 갈아엎었다. 옮기다 만난 함정 두 개가 사실 본편보다 재밌었음.https://zerry.co.kr/blog/from-rag-to-agentic-rag#agenticrag #rag #llm #groq #functioncalling #chatbot
Related
2/2...meanwhile back at the ranch the next gen work in progress, along with the highly dependable unreasonable effective...
2/2...meanwhile back at the ranch the next gen work in progress, along with the highly dependable unreasonable effectiveness of mathematics, is humming.Admittedly I'm still a bit c...
Vision-Language-Action (VLA) models process long multimodal token sequences, making inference expensive in both memory a...
Vision-Language-Action (VLA) models process long multimodal token sequences, making inference expensive in both memory and computation. Role-Conditioned Sub-Token Routing (RoleSub)...
#ai #usa2aTwo things fell hard in the last couple of years.- The Turing Test- The US 2nd AmmendmentDead and done.
#ai #usa2aTwo things fell hard in the last couple of years.- The Turing Test- The US 2nd AmmendmentDead and done.