Haystack 发布 v3.3.0-rc1：BM25 检索与 CJK 分词修复

AI 摘要

Haystack 发布 v3.3.0-rc1 候选版本，主要更新包括：InMemoryDocumentStore 的 BM25L/BM25Plus 检索现在只返回包含至少一个查询词的文档，并修复了中日韩文本的 BM25 分词问题；SentenceWindowRetriever 改为每次调用只查询一次文档存储以降低 OpenSearch/Elasticsearch 负载；同时因 CVE-2026-63374 要求 anyio>=4.14.2，并修复了 ChatPromptBuilder、DOCXToDocument、MetaFieldRanker 等多个组件的缺陷。

正文节选

# Release Notes ## v3.3.0-rc1 ### Upgrade Notes - `InMemoryDocumentStore.bm25_retrieval` and `InMemoryBM25Retriever` with `BM25L` (the default) or `BM25Plus` now return only documents that contain at least one query term, with or without `scale_score`. Previously, documents without any query term were returned with a positive score and filled up `top_k`. As a result, retrieval can now return fewer than `top_k` documents, or none at all. The `delta` lower bound now applies only to query ter

发布时间：2026-09-30 21:33

抓取时间：2026-09-30 21:51

来源机构：deepset