Hi everyone,
I’ve a project that under development and i need an openai model for a process.
Process is a mapping process. I have a class and class has 3 props is string 2 is date. I am going to send a request to openai api with a text. I need a response that just consist of a json object. And this json object must be map according to text i sent. For this process which model I should use? and api can you show how api request should be.
1 Like
gpt-3.5-turbo is the best model to start with for just about any task. From there, try the others, it’s real easy to switch between them. Relevant documentation here and here
1 Like
Do you know what kind of differences it has compared to 4.0
Hi, you can do this with different models.
What you have to do is send a system message, saying that it has to return a json, send a simulated user message and a simulated assistant response with the response in JSON format.
It may also be a good option to reiterate at the end of the user message that you want the output in JSON format