openai

package

v0.0.0-...-80055c2 Latest Latest Go to latest Published: May 15, 2023 License: MIT Imports: 6 Imported by: 0

Details

Valid go.mod file
Redistributable license
Tagged version
Stable version
Learn more about best practices

Repository

github.com/philiplinell/commit-msg

Links

Open Source Insights

Documentation ¶

Index ¶

Constants
type ChatCompletionResponse
type Client
- func NewClient(httpClient Doer, apiKey string) *Client
- func (c *Client) ChatCompletionRequest(ctx context.Context, messages []Message, model aiModel, temperature float32) (ChatCompletionResponse, error)
type Coster
type Doer
type Message

Constants ¶

View Source

const (
	// The user messages help instruct the assistant. They can be generated by
	// the end users of an application, or set by a developer as an
	// instruction.
	UserRole aiRole = "user"

	// The system message helps set the behavior of the assistant. E.g. the
	// assistant can be instructed with "You are a helpful assistant".
	SystemRole aiRole = "system"

	// The assistant messages help store prior responses. They can also be
	// written by a developer to help give examples of desired behavior.
	AssistantRole aiRole = "assistant"
)

View Source

const (
	// GPT3_5Turbo - The most capable GPT-3.5 model and optimized for chat at
	// 1/10th the cost of text-davinci-003. Will be updated with the latest
	// model iteration.
	// gpt-3.5-turbo is recomennded over the other GPT-3.5 model due to its
	// (lowest) cost.
	GPT3_5Turbo aiModel = "gpt-3.5-turbo"
)

Variables ¶

This section is empty.

Functions ¶

This section is empty.

Types ¶

type ChatCompletionResponse ¶

type ChatCompletionResponse struct {
	Created time.Time
	Model   aiModel

	// Cost is the cost for the request in dollars.
	Cost float64

	Messages []string
}

func NewClient ¶

func NewClient(httpClient Doer, apiKey string) *Client

NewClient creates a new OpenAI API client.

func (*Client) ChatCompletionRequest ¶

func (c *Client) ChatCompletionRequest(ctx context.Context, messages []Message, model aiModel, temperature float32) (ChatCompletionResponse, error)

ChatCompletionRequest does a request to the openai chat completion API.

temperature decides how deterministic the model is in generating a response. It must be a value between 0 and 1 (inclusive). A lower temperature means that completions will be more accurate and deterministic. A higher temperature value means that the completions will be more diverse. See more about temperature here: https://platform.openai.com/docs/quickstart/adjust-your-settings

type Coster ¶

type Coster interface {
	// Cost returns the cost in dollars of the request based on the total
	// tokens used.
	Cost(totalTokens int) float64
}

Coster is an interface that models can implement to calculate the cost of request based on the total tokens used.

type Doer ¶

type Doer interface {
	Do(req *http.Request) (*http.Response, error)
}

type Message ¶

type Message struct {
	// Role is the role of the message.
	Role aiRole `json:"role"`
	// Content is the message to send to the OpenAI model.
	Content string `json:"content"`
}

Message is used to interact with the OpenAI model.

Source Files ¶

View all Source files

client.go

?	: This menu
/	: Search site
f or F	: Jump to
y or Y	: Canonical URL