---
layout: single
title:  "RAG 2.0"
date:   2025-03-22 10:00:00 +0800
categories: [AI 与大模型, DevOps]
tags: [RAG, LLM]
---

![](/images/2025/RAG/RAG.001.jpeg)
![](/images/2025/RAG/RAG.002.jpeg)
![](/images/2025/RAG/RAG.003.jpeg)
![](/images/2025/RAG/RAG.004.jpeg)
![](/images/2025/RAG/RAG.005.jpeg)
![](/images/2025/RAG/RAG.006.jpeg)
![](/images/2025/RAG/RAG.007.jpeg)
![](/images/2025/RAG/RAG.008.jpeg)
![](/images/2025/RAG/RAG.009.jpeg)
![](/images/2025/RAG/RAG.010.jpeg)
![](/images/2025/RAG/RAG.011.jpeg)
![](/images/2025/RAG/RAG.012.jpeg)
![](/images/2025/RAG/RAG.013.jpeg)
![](/images/2025/RAG/RAG.014.jpeg)
![](/images/2025/RAG/RAG.015.jpeg)
![](/images/2025/RAG/RAG.016.jpeg)
![](/images/2025/RAG/RAG.017.jpeg)
![](/images/2025/RAG/RAG.018.jpeg)
![](/images/2025/RAG/RAG.019.jpeg)
![](/images/2025/RAG/RAG.020.jpeg)
![](/images/2025/RAG/RAG.021.jpeg)
![](/images/2025/RAG/RAG.022.jpeg)
![](/images/2025/RAG/RAG.023.jpeg)
![](/images/2025/RAG/RAG.024.jpeg)
![](/images/2025/RAG/RAG.025.jpeg)
![](/images/2025/RAG/RAG.026.jpeg)
![](/images/2025/RAG/RAG.027.jpeg)
![](/images/2025/RAG/RAG.028.jpeg)
![](/images/2025/RAG/RAG.029.jpeg)
![](/images/2025/RAG/RAG.030.jpeg)
![](/images/2025/RAG/RAG.031.jpeg)
![](/images/2025/RAG/RAG.032.jpeg)
![](/images/2025/RAG/RAG.033.jpeg)
![](/images/2025/RAG/RAG.034.jpeg)
![](/images/2025/RAG/RAG.035.jpeg)
![](/images/2025/RAG/RAG.036.jpeg)
![](/images/2025/RAG/RAG.037.jpeg)
![](/images/2025/RAG/RAG.038.jpeg)
![](/images/2025/RAG/RAG.039.jpeg)
![](/images/2025/RAG/RAG.040.jpeg)
![](/images/2025/RAG/RAG.041.jpeg)
![](/images/2025/RAG/RAG.042.jpeg)
![](/images/2025/RAG/RAG.043.jpeg)
![](/images/2025/RAG/RAG.044.jpeg)
![](/images/2025/RAG/RAG.045.jpeg)
![](/images/2025/RAG/RAG.046.jpeg)
![](/images/2025/RAG/RAG.047.jpeg)
![](/images/2025/RAG/RAG.048.jpeg)


## 参考资料
- [2024 年 RAG 的崛起与演变年度回顾](https://ragflow.io/blog/the-rise-and-evolution-of-rag-in-2024-a-year-in-review)
- [所见即所得：多模态RAG正在向我们走来](https://news.qq.com/rain/a/20241022A04GE100)
- [What is Retrieval Augmented Generation (RAG)?](https://www.trantorinc.com/blog/what-is-rag-retrieval-augmented-generation)
- [Build with Claude - Prompt caching](https://docs.anthropic.com/en/docs/build-with-claude/prompt-caching)
- [Introducing Contextual Retrieval](https://www.anthropic.com/news/contextual-retrieval)
- [Cookbook - Retrieval Augmented Generation with Contextual Embeddings](https://github.com/anthropics/anthropic-cookbook/tree/main/skills/contextual-embeddings)
- [Infinity](https://infiniflow.org/)
- [Dense vector + Sparse vector + Full text search + Tensor reranker = Best retrieval for RAG?](https://infiniflow.org/blog/best-hybrid-search-solution)
- [Sparse embedding or BM25?](https://infiniflow.org/blog/sparse-embedding-bm25)
- [ColPali: Efficient Document Retrieval with Vision Language Models](https://arxiv.org/abs/2407.01449)
- [PaliGemma 正式发布 — Google 最新发布的前沿开放视觉语言模型](https://huggingface.co/blog/zh/paligemma)
- [SPLADE v2: Sparse Lexical and Expansion Model for Information Retrieval](https://arxiv.org/abs/2109.10086)
- [RAG的技巧：TF-IDF與BM25的介紹與使用](https://chtseng.wordpress.com/2024/11/18/rag%E7%9A%84%E6%8A%80%E5%B7%A7%EF%BC%9Atf-idf%E8%88%87bm25%E7%9A%84%E4%BB%8B%E7%B4%B9%E8%88%87%E4%BD%BF%E7%94%A8/)
- [BM25算法---计算文档和query相关性](https://www.cnblogs.com/Lee-yl/p/11149879.html)
- [Agentic RAG: Definition and Low-code Implementation](https://medium.com/@infiniflowai/agentic-rag-definition-and-low-code-implementation-d0744815029c)
- [Adaptive RAG](https://x.com/Pavan_Belagatti/status/1804919895580135886)
- [MTEB Leaderboard](https://huggingface.co/spaces/mteb/leaderboard)
- [What Infrastructure Capabilities does RAG Need beyond Hybrid Search](https://ragflow.io/blog/what-infrastructure-capabilities-does-rag-need-beyond-hybrid-search)