Extracting text fro PDF into OpenAI

I have connected cloudconvert to doccrafter to get the text from a pdf of a webpage i need however the text wont go through to OpenAI to work with, why?


1 Like

Hi @Sam_Van_Leeuwen

Thanks for reaching out to the community. Please try inputting 0 in the field value of “Max Tokens”.

If 0, the limit of the model will be used. If the issue persists, ensure that you have added billing here or you may try with a lower model.

Regards,
Msquare Automation - Gold Partner of Make

Book a Free Consultation | Connect Live

Explore our YouTube Channel for valuable insights and updates!

same problem with max tokens at 0

@Sam_Van_Leeuwen ,

This issue is related to the character length of the input that you have given to the GPT. To avoid this, you have two options:

1:- You can split the input into two parts, get a summary of each, and then use another GPT to aggregate the two summaries and get the final output.

The entire flow will be as shown below:

  1. Split the input into two parts.
  2. Get a summary for each part.
  3. Aggregate the two summaries.
  4. Use another GPT to generate the final output from the aggregated summaries.

2 :-

The new GPT has introduced a method to send almost 50,000 API calls at the same time. You can split your input content, create a Jsonl file, and make the API calls work.
For more information please check the below documentation. https://platform.openai.com/docs/guides/batch