The chat completions endpoint

The chat completions endpoint can be used to perform many tasks on a text, including classification, generation, transformation, completion of incomplete text, factual responses, and others. The endpoint takes an input message from the user and the role assigned to it and returns a JSON object.

Fields	Format	Type	Description
`messages`	Array	Required	This is a list of messages comprising the conversation as of yet.
`model`	String	Required	This is the ID of the model that the chat completions endpoint will use.
`max_tokens`	Integer	Optional	This is the maximum number of tokens to generate in the chat completion.
`temperature`	Float	Optional	Which sampling temperature should be employed, ranging from `0` to `2`? The default temperature is set at `1`. Opting for higher values, such as `0.8`, will increase randomness in the output, whereas lower values, like `0.2`, will enhance focus and determinism.
`top_p`	Float	Optional	Nucleus sampling is an alternative to temperature sampling in which the model evaluates the outcomes of tokens with top p probability mass. So, 0.1 indicates that only the top 10% probability mass tokens will be evaluated. Default value: `1`
`response_format`	object	Optional	This is an object that specifies the format of the output. This parameter is compatible with the newer models. Setting to `{ "type" : "json_object" }` will guarantee a valid JSON object.
`n`	Integer	Optional	It indicates the number of chat completions to generate. Default value: `1`
`logprobs`	Integer	Optional	This value decides whether to return the log probabilities of output tokens. This option is not available on the `gpt-4-vision-preview` model. Default value: `false`
`stop`	String/array	Optional	This allows you to provide four sequences where the API will stop generating further tokens. Default value: `null`
`presence_penalty`	Float	Optional	A value range between -2.0 to 2.0. The positive value penalizes the new tokens if they exist in the text, increasing the possibility of generating new things. Default value: `0`
`frequency_penalty`	Float	Optional	A value range between -2.0 to 2.0. The positive value penalizes the new tokens concerning their existing frequency in the text, decreasing the possibility of repeating the same words. Default value: `0`

Fields	Format	Description
`id`	String	This is a unique ID for the chat completion.
`choices`	Array	This is an array of objects. Every object contains valuable information about the response. The size of the array will be equal to the n parameter that we provided in the request parameters.
`model`	String	This is the Unix timestamp (in seconds) of when the chat completions was created.
`usage`	Object	These usage statistics specify the `prompt_tokens`, `completion_tokens`, and `total_tokens` for the chat completion requests.

Fields	Format	Description
`choices[i].finish_reason`	String	This provides the reason the model stopped generating tokens. Here are the reasons sent: `stop`: This is a natural stop point. `length`: The number of tokens specified in the request has been reached. `content_filter`: Content is omitted due to a flag generated by a filter. `tool_calls`: The model calls a tool.
`choices[i].index`	Integer	This gives the index of the choice in a list of choices.
`choices[i].message`	Object	This is the chat completion response message generated by the model.
`choices[i].logprobs`	Object	This is the log probability of the response choice.

Basics of prompt design

OpenAI can perform many simple and complex tasks in text analysis. To get the most out of it, we have to be very specific about our prompt. We must take particular care about the following points:

Proofread the prompt for spelling and grammatical mistakes because that can affect the output.
Check for setting parameters such as temperature and top_p.
Ensure clarity about the intended generation of the text.
Provide one or more examples. (We’ll see this prompt design in the next lessons.)

Let’s look at a few examples by executing various prompts.

Using the chat completions endpoint

Let’s use the chat completions endpoint to write something about artificial intelligence. In the code widget below, we use the GPT-3.5-Turbo model for the chat completions task. The temperature value used is 0.9. Make sure you’ve added your SECRET_KEY. Press the "Run" button to see the response for our prompt.

The following prompt exhibits the generation use case by generating some text corresponding to the provided prompt:

Javascript (babel-node)

// Define endpoint URL here
const endpointUrl = "https://api.openai.com/v1/chat/completions";
// Define Header Parameters here
const headerParameters = {
  "Authorization": "Bearer {{SECRET_KEY}}",
  "Content-Type": "application/json"
};
// Body Parameters
const bodyParameters = JSON.stringify({
  model: "gpt-3.5-turbo",
  messages: [{role:"system",content:"You are a helpful assistant"},
             {role:"user",content:"Write a tagline about artificial intelligence."}],
  temperature: 0.9
});
// Setting API call options
const options = {
  method: "POST",
  headers: headerParameters,
  body: bodyParameters
};
// Function to make API call
async function createCompletion() {
  try {
    const response = await fetch(`${endpointUrl}`, options);
    // Printing response
    printResponse(response);
  } catch (error) {
    // Printing error message
    printError(error);
  }
}
// Calling function to make API call
createCompletion();

We have imported the node-fetch library to make an API call in the above code. Let’s see some code details:

Line 2: We define the endpoint URL.
Line 5–8: We define the header, which includes the authorization token and content type.
Line 11–15: We define the request parameters required to make the API call.
Line 19–23: We set options by specifying the header, body parameters, and the request method as POST.
Line 26–35: We create a function createCompletion to make an API call using fetch and handle any exception if it occurs. The printResponse and printError are the custom functions to print the respective objects.
Line 37: We invoke the createCompletion function.

Try out different prompts and examine the varying responses. Here are some examples:

1.Introduction

2.The Completions Endpoint

3.Miscellaneous Endpoints

4.APIs Integration in React

5.Conclusion

6.Appendix

Chat Completions

The chat completions endpoint

Understanding the chat completions endpoint

Request parameters

Response fields

Prompt design

Basics of prompt design

Using the chat completions endpoint