类class
agentscope.ragVectorRecord
class VectorRecord(BaseModel)源码Source
agentscope/rag/_vdb/_vector_store.py:28–50向量记录，把私有文档转化为可检索的知识。位于 agentscope.rag 包。
Pairs a :class:Chunk (the business payload — content, source,
structural metadata) with its dense embedding vector. Chunk
is intentionally not extended with an embedding field so its
semantics stay stable across the indexing pipeline; instead the
vector lives in this wrapper whose only purpose is "I am about
to be inserted into a vector database."
属性摘要Attribute Summary
| 属性Attribute | 类型Type | 说明Description | 值Value |
|---|---|---|---|
vector | list[float] | 向量字段。 | — |
document_id | str | 文档标识字段。 | — |
chunk | Chunk | 分块字段。 | — |
属性详情Attribute Detail
vector
list[float]向量字段。声明为 list[float]。
The vector field.
源码Source
agentscope/rag/_vdb/_vector_store.py:39document_id
str文档标识字段。声明为 str。
The document id field.
源码Source
agentscope/rag/_vdb/_vector_store.py:42chunk
Chunk分块字段。声明为 Chunk。
The chunk field.
源码Source
agentscope/rag/_vdb/_vector_store.py:49