model: text-davinci-003

prompt: Extract keywords from ...

temperature: 0.5

max_token: 60

frequency_penalty: 0.8

fetching from: 

https://api.openai.com/v1/completions