Hi all. I want to take multiple inputs that a user is contributing to, feed it through OpenAI, and have it output something based on everything they entered in the inputs. Below is an example for more context of what i would like to happen.
Essentially, the box in the second screenshot will populate a list of things based on what the user put in all of the inputs. Also each input has a separate field under one data type. In the past, I have accessed a single input to output system using OpenAI with the basic API call shown below and arbitrary text to feed it the prompt.
I have been trying to use the same call to access multiple inputs but I’m pretty sure that is not going to work but I really don’t know how to proceed as setting up the first call was challenging enough (for my beginner self) . If anyone has any idea regarding how to attempt this, that would be great! Also: if extra screenshots of my data types or workflows are needed please let me know and I will attach them right away. Thanks in advance.
Each Input is separate and unique. Each input is not in a repeating group.
So, where you have content in your API call as a dynamic expression, that would need to change. The issue is not getting multiple inputs to become one CONTENT, instead you need to restructure your API call to have Parameters for each Input value type.
OR, you need to ensure you can take each inputs value and form a full coherent single Content. I can see you have TONE, and BACKGROUND…so you could use ARBITRARY TEXT operator to format your CONTENT, by inside of ARBITRARY TEXT write something like ‘Multiline Input Background written in tone of Dropdown Tone value’…
So the ultimate goal is to figure out a Templated version of all the different inputs values so as to swap them out easily and have the overall Content of the call still be understood by the AI, and you can do that using ARBITRARY TEXT operator.
Okay cool I think I’m understanding. So just like the screenshot of my other page below, where I have the prompt and then the multiline input I’m referencing for a singular input to output function…
What you’re saying is that I can take each input’s value and plug in play throughout the prompt, create a completely NEW data type (name it something like results) and then set that equal to the result of step 1 (chatgpt API call plugin) and then I should be golden without even having to modify my API call.
Let me know if I misinterpreted you, but I’ll give that a try and let you know how it goes.
Solved! Thank you so much for your help, saved me hours of wasted time that I was about to get myself into with messing around with the API’s. Appreciate it!
