A few days back, I stumbled across the new Feature launched by OPENAI and it opens new doors for a developer like me and it can help business to minimize the customer need to look for each feature in the app and use it.
For example,
I am Fabric.js(it is a canvas library) Developer and I wanted to build a customize prompt for users where they can enter their query and based on that query I can perform functions on Canvas and this way users can have the option to customize their design templates very easily.
Practical Use Case:
I have an object which has many key value pairs and I want to get the key and value from users by providing an interactive prompt so if user types "Change the text to Hello World" then I can make a request like that and get results accordingly in a JSON formatted way.
curl https://api.openai.com/v1/chat/completions -u :$OPENAI_KEY -H 'Content-Type: application/json' -d '{
"model": "gpt-3.5-turbo-0613",
"messages": [
{"role": "user", "content": "Change the text to Hello World"}
],
"functions": [
{
"name": "change_value",
"description": "Update value of object key",
"parameters": {
"type": "object",
"properties": {
"key": {
"type": "string",
"description": "Key that needs to be changed"
},
"value": {
"type": "string",
"description": "value that needs to be updated"
}
},
"required": ["key","value"]
}
}
]
}'
This is a curl command making a POST request to OpenAI's GPT-3.5-turbo model. It will return this JSON response
{
"key": "text",
"value": "Hello World"
}
You can play around with it here.
http://langtale.ai/playground/p/fD10_ciAcM
If you enjoyed the article and have questions, make sure to:
🔔 Connect with me on mudassirali.com
Top comments (0)