coze-studio-0.3.0版本知识库上传文件报错code=105000004

分别在windows和linux上部署coze-studio-0.3.0，知识库上传文件均报错，错误不一致。

**windows下**
文档分段超过10段时，就会报错
错误：code=105000004 message=SearchStore operation failed: store search store failed, err: [slices2Columns] embed failed, error, status code: 400, status: 400 Bad Request, message: <400> InternalError.Algo.InvalidParameter: Value error, batch size is invalid, it should not be larger than 10.: input.contents
如图：

.env相关配置如下图：

**linux下**
错误：code=105000004 message=SearchStore operation failed: store searchstore failed, err:[slices2Columns] embed failed,json: cannot unmarshal array into Go struct field Embedding.data.embedding of type float32
如图：

.env相关配置如下图：

Start by reproducing knowledge-base file uploads with coze-studio 0.3.0 on Windows and Linux, using the environment configurations shown in the issue. Inspect the embedding and search-store error paths, including the batch-size limit and the Go embedding response parsing. Done means uploads succeed for documents over 10 segments on both platforms.

Written by the indexing model from the issue text.