Hello,
I’m having trouble translating annotations from v2 wizard responses when using the file_search tool, something like this answer
"partial response【18:0†source】【18:1†source】【18:2†source】.\n\partial response【18:4†source】【18:12†source】."
Once I have replaced the indexes of the references, I am trying to make the reference itself using the annotation.
When I try to create my references/legend from these annotations I ran into 2 crashes.
The first is that the file having purpose assistant is not allowed to be downloaded, resulting in a 400 http.
. The second is choosing to store the file in parallel in another binary data store. But, when I try to replace the start_index and end_index positions on the content of the file, it doesn’t quite add up. They are fragments cut without much sense. I have verified that they are not file encoding issues.
So I don’t really understand how I have to proceed once I have 【x:y†source】
to be able to relate it directly to the specific text fragment in the file.
Thank you very much in advance.