I’m wanted to know if we can upload large files. I know that there is a 1 GB limit to the size of the file, but I tried uploading a database schema via the api but found out that there is a size/token limit for each line (JSON dictionary) of the JSONL file. Is there a way around it?
I researched a little on the forum and found out that chunking is a good option, but automating that process is a little tricky as I couldn’t find an exact relation between characters and tokens (how many characters is equal to one token).
Furthermore, does anyone know if it’s possible to integrate the codex engine with the answers api and if we can fine-tune it for a particular usecase?
Yes it is very easy to use codex with answers api, set the
search_model to codex series engines you want to use, add relevant
example_context, pass your
documents for factual answers and the
question you want to answer.
Here’s API reference for /answers
Thank you! I’ll try it out😁
Thanks! I’ll take a look at embeddings and see if I can use it for my usecase!