API nodejs

[celest stag]

Has anyone used the API and figured out how to get the entire text instead of just the first line in the response

[burnt tusk]

You can get the entire text response in response.data.choices[0].text

[celest stag]

that is what im doing rn actually

i used the gpt-3 template given

  const configuration = new Configuration({
    
    apiKey: process.env.OPENAI_API_KEY,
  });
  
  const openai = new OpenAIApi(configuration);

  
    const completion = await openai.createCompletion({
      model: "text-davinci-002",
      prompt: generatePrompt(req.body.promptText),
      temperature: 0.6,
    });
 

 
   res.status(200).json({ result: completion.data.choices[0].text });
});```

this is the endpoint

i only ever get 1 phrase

    event.preventDefault();

    const response = await fetch("http://localhost:3001/api/generate", {
        method: "POST",
        headers: {
            "Content-Type": "application/json",
        },
        body: JSON.stringify({ promptText }),
    });
    const data = await response.json();

    setResponseText(data.result);
  

};```
this is my react side

[burnt tusk]

In your api, if you console.log completion.data.choices[0].text, did you het the whole text

[celest stag]

nope just one line

i noticed its a readable stream is it sent in chunks or something maybe thats why? @burnt tusk

@burnt tusk

[sinful sparrow]

It's sends one line because your code is not complete

[celest stag]

@sinful sparrow wdym what am I missing

[sinful sparrow]

           model: "text-davinci-003", 
           max_tokens: 2000, 
           temperature: 0.6, 
           frequency_penalty: 0, 
           presence_penalty: 0.6,            prompt: generatePrompt(req.body.promptText), 
         }); ```

Test it now

[celest stag]

Oo

works now

[celest stag]

is max tokens the number of tokens in the response?

and what is the actual max we can use

[sinful sparrow]

Yes

[sinful sparrow]

I think davinci is 2000 token

[celest stag]

Thanks brotha