What does the files status "error" mean

Prompt Assistance
#1

I’ve uploaded a couple documents to the /files endpoint with:

openai.File.create(file=open(“upload-files.JSONL”, encoding=‘utf-8’), purpose=“search”)

I get no error here but when looking at the files with:

curl https://api.openai.com/v1/files
-H ‘Authorization: Bearer YOUR_API_KEY’

I get the following response:
“id”: “file-xxxxxxxxxxxxxxxxxxxxxxxxxx”,
“object”: “file”,
“bytes”: 36611,
“created_at”: 1618160623,
“filename”: “upload-files.JSONL”,
“purpose”: “search”,
"status": "error"

Does anybody what kind of error this may be or how I could get more information on it?

btw:
I tried searching through the documents with the exact text from one of the documents. This failed with the error “No similar documents were found in file with ID ‘file-xxxxxxxxxxxxxxx’.Please upload more documents or adjust your query.”

#2

Hey Kris, with an error like this, it may that the document isn’t in the proper JSONL format. Would it be possible to share a snippet of your JSONL file (text anonymised if needed)?

Since there was an error in creating the file, you wouldn’t be able to search through the documents.

#3

Yes, the first line looks like this. I replaced the text for readability. The longest text in whole file has 350 tokens.

{“text”: “…text…” , “metadata”: “00010 Name”}

When writing this I did notice the extra space (the italic one :wink: ) after the comma. I reuploaded the file without but got the same (?) error.

#4

RESOLVED:
Thanks to @joey to for pointing me in the right direction. The script that creates the upload file just needed to pass the data as a dict to the jsonlines writer (not as a string).

Great service, would error again :+1:

5 Likes