HowardChan
HowardChan
Upgrade it to the latest version solved it for me.
You're right. Based on #1085, the chunk_ids of LIGHTRAG_VDB_ENTITY and LIGHTRAG_VDB_RELATION should be an array of VARCHAR. I believe it's also related to #1091.
> error:column "chunk_ids" of relation "lightrag_vdb_entity" does not exist Failed to extract entities and relationships > > reporting the same issue.. Hey, Please pull the latest code and rebuild your...
> See comments Hi! I'm glad to see your comment, I've resolved it.PTAL.
> @Howe829 Hello, please resolve deps issue here. Hi there! I've resolved the deps issues.The Pytest failed due to the api key provided in test_pinecone.py is not real.
> > > @Howe829 Hello, please resolve deps issue here. > > > > > > Hi there! I've resolved the deps issues.The Pytest failed due to the api key...
> Seems there more conflicts here. Resolved.PTAL @crazywoola
> > > 因为用的还是朴素rag的那一套做的，想要效果好还得结合高级rag和搜索引擎 > > > > > > 高级的rag是什么样子的？graph rag？advanced rag？总感觉是模型上对中文支持的问题 > > 举个例子，在做文本切分的时候还是用的chunk_size，不能按句子、段落或语义切分 > >  你贴的这个图，就是按句子切分的SentenceWindow技术额