LlamaIndex使用HyDE的demo HyDE[1] 是先用query生成伪答案再用伪答案检索。 importosos.environ["OPENAI_API_KEY"]="sk-..."os.environ["http_proxy"]="http://127.0.0.1:7890"os.environ["https_proxy"]="http://127.0.0.1:7890"fromllama_index.coreimportVectorStoreIndex,SimpleDirectoryReaderfromllama_i...
之后,我们使用这个假设性文档来进行嵌入式查找,而不是直接使用原始查询。 # 加载文档,构建索引documents=SimpleDirectoryReader("../paul_graham_essay/data").load_data()index=VectorStoreIndex(documents)# 使用HyDE查询转换运行查询query_str="what did paul graham do after going to RISD"hyde=HyDEQueryTransform...
HyDE Query Transform Multi-Step Query Engine Sub Question Query Engine (Intro) Build your own OpenAI Agent OpenAI Agent with Query Engine Tools Retrieval-Augmented OpenAI Agent OpenAI Agent + Query Engine Experimental Cookbook OpenAI Agent Query Planning Context-Augmented OpenAI Agent 分类...
ImportError: cannot import name 'HyDEQueryTransform' from 'llama_index.indices.query.query_transform' (/usr/local/lib/python3.9/dist-packages/llama_index/indices/query/query_transform/__init__.py) Import like this work for me: from llama_index.indices.query.query_transform.base import (HyDEQu...
Hypothetical document embeddings (HyDE) Typically, when we ask a question about an external document, what we normally do is that we use text embeddings to create vector representations for both the question and the document. Then we use semantic search to find the text chunks that are the m...
4 changes: 2 additions & 2 deletions 4 docs/docs/examples/agent/agent_runner/agent_around_query_pipeline_with_HyDE_for_PDFs.ipynb Original file line numberDiff line numberDiff line change @@ -50,8 +50,8 @@ "outputs": [], "source": [ "!mkdir -p 'data/10k/'\n", "!wget 'http...
LlamaIndex 的优势在于易用性、灵活的索引方法和高效的检索能力，但它在定制化检索和生成策略方面的灵活性...
Added a QueryTransform class allowing us to transform queries within our data structures - first implementation is with HyDE Some examples are given here: https://github.com/jerryjliu/gpt_index/blob/main/examples/query_transformations/HyDEQueryTransformDemo.ipynb相关...
HyDE Query Transform Multi-Step Query Engine Sub Question Query Engine (Intro) Build your own OpenAI Agent OpenAI Agent with Query Engine Tools Retrieval-Augmented OpenAI Agent OpenAI Agent + Query Engine Experimental Cookbook OpenAI Agent Query Planning Context-Augmented OpenAI Agent 分类...
HyDE(Hypothetical Document Embeddings) 提供另外一种思路：如果用户输入的问题过于简短，那么可以基于LLM的...