Chinese version-----
我想要从V1升级到V2,以下是我的升级更改,请帮忙指正,非常感谢:
1、我需要替换请求头的Bate版本
2、修改原来的助手工具retrieval
为新工具file_search
3、对于初始化助手工具的文件的情况,需要修改文件到tool_resources
4、对于原来的retrieval
工具,新的file_search
需要使用向量库?(这里我比较迷惑,可以帮忙拟一个流程吗)
5、在线程内的消息,以前生成文件是file_ids
现在为attachments
。这个位置我想问下我生成的文件是会被助手添加到向量库,并索引到助手吗?
6、我不需要访问v1的旧文件数据,不需要关注跨版本访问的问题。
7、我没有使用python
使用的curl
直接请求。
English version-----
I want to upgrade from V1 to V2. Below are the changes I plan to make for the upgrade. Please correct me if there’s anything wrong. Thank you very much:
- I need to replace the Beta version in the request header.
- Change the original helper tool
retrieval
to the new toolfile_search
. - For the initialization of the helper tool’s files, I need to move files to
tool_resources
. - For the original
retrieval
tool, does the newfile_search
need to use a vector database? (I’m a bit confused here; can you help draft a process?) - Within thread messages, files were previously identified as
file_ids
but now asattachments
. At this point, I want to ask if the files I generate will be added to the vector database by the helper and indexed? - I don’t need access to old file data from V1 and do not need to worry about cross-version access issues.
- I haven’t used
python
; I made direct requests withcurl
.