My problem is:
I have a pdf file. while storing this in pinecone DB, it divides my pdf into chunks and stores as multiple records. Now my whole pdf have single metadata key-value pair with 17 fields. How do i group this multiple records of pdf to this single meta-data?
Hi @abhinav1 and welcome to the Pinecone community forums!
Thank you for your question.
If I’m understanding your question correctly, then the solution is to attach the same metadata key when upserting each of the 17 chunks - for example:
{"source":"my-file-1.pdf"}
This way, all your chunks will be associated via the source key in your metadata.
Hope this helps!
Best,
Zack
Hi @ZacharyProser,
Thank you so much for the reply.
Understanding your response. Lets say i have pdf “my-file-1.pdf”, While storing this in pinecone it created 50 records. Now, Some important information i have extracted from pdf and want to store it as metadata. So for this 50 records you are asking me to attach same metadata. Is that correct?
So By attaching same metadata key everytime for the multiple records of pdf wouldnt create redundancy issues? because i will be updating the data in pdf say weekly once, at that time metadata also changes.
Will i be able to update the metadata without any redundancy issues.
1 Like