Embedding维度不匹配如何解决

> 萌新小白本地部署coze求救 ToT

embedding一个.txt纯文档
[Story0_Snow_White.txt](https://github.com/user-attachments/files/24165664/Story0_Snow_White.txt)
出现报错：
code=105000004 message=SearchStore operation failed: store search store failed, err: [Store] upsert failed, the length(66048) of float data should divide the dim(1024)

embedding model: ollama nomic-embed-text
embed dim = 1024

问题出在哪里？

Start by reproducing the reported embedding upload with Ollama's nomic-embed-text model and the supplied Story0_Snow_White.txt document. Trace the embedding dimension used by the search store against the returned float-data length, then determine whether the mismatch is caused by configuration or dimension handling. Done means the cause and a reproducible correction are documented or implemented.

Written by the indexing model from the issue text.