Skip to content

HassanTaleb90/OpenAIKit

 
 

Repository files navigation

OpenAIKit

Swift Platforms Swift Package Manager Swift

OpenAIKit is a community-maintained API for the OpenAI REST endpoint used to get output data from Dall-E 2 and GPT-3 models.

Features

  • Generate new, edited, and variations of images using Dall-E 2.
  • Generate edits and completions using GPT-3.
  • List avaiable models for use with GPT-3.
  • Retrieve embeddings for GPT-3 prompts.
  • Generate Chat responses using ChatGPT.
  • Upload and view training set files for Fine-tuning.
  • Generate and use Fine-tuned models.
  • View whether a prompt is flagged by the Moderations endpoint or not.
  • Comprehensive Unit and Integration Test coverage.
  • Swift Concurrency Support back to iOS 13, macOS 10.15, tvOS 13, and watchOS 6.
  • Complete documentation of OpenAIKit.

ToDo

  • Implement data streaming for Fine-Tune and GPT-3 event streams.

Requirements

Platform Minimum Swift Version Installation Status
iOS 13.0+ / macOS 10.15+ / tvOS 13.0+ / watchOS 6.0+ 5.5 Swift Package Manager Fully Tested

Installation

Swift Package Manager

The Swift Package Manager allows for developers to easily integrate packages into their Xcode projects and packages; and is also fully integrated into the swift compiler.

SPM Through XCode Project

  • File > Swift Packages > Add Package Dependency
  • Add https://github.com/marcodotio/OpenAIKit.git
  • Select "Up to next Major" with "1.2"

SPM Through Xcode Package

Once you have your Swift package set up, add the Git link within the dependencies value of your Package.swift file.

dependencies: [
    .package(url: "https://github.com/marcodotio/OpenAIKit.git", .upToNextMajor(from: "1.2"))
]

Using OpenAIKit

OpenAIKit is designed to be very easy to integrate into your own projects. The main method of utilizing OpenAIKit is to set a OpenAI class object:

import OpenAIKit

// An API key and Organization ID is required to use the API library.
// Note: It's recommended to load the API Key through a .plist dictionary, rather than hard coding it in a String.
let config = Configuration(organizationId: "INSERT-ORGANIZATION-ID", apiKey: "INSERT-API-KEY")

// Create an `OpenAI` object using the Configuration object.
let openAI = OpenAI(config)

From there, it's as easy as calling one of the provided function members. The code below demonstrates how you can generate an image using createImage():

do {
    let imageParam = ImageParameters(
      prompt: "a red apple", 
      resolution: .small, 
      responseFormat: .base64Json
    )
    let result = try await openAi.createImage(
      parameters: imageParam
    )
    let b64Image = result.data[0].image
    let image = try openAi.decodeBase64Image(b64Image)
} catch {
    // Insert your own error handling method here.
}

As well, you are able to generate completions using GPT-3:

do {
    let completionParameter = CompletionParameters(
      model: "text-davinci-001", 
      prompt: ["Say this is a test ->"], 
      maxTokens: 4, 
      temperature: 0.98
    )
    let completionResponse = try await openAI.generateCompletion(
      parameters: completionParameter
    )
    let responseText = completionResponse.choices[0].text
} catch {
    // Insert your own error handling method here.
}

Development And Testing

I welcome anyone to contribute to the project through posting issues, if they encounter any bugs / glitches while using OpenAIKit; and as well with creating pull issues that add any additional features to OpenAIKit.

Next Steps

  • In the near future, there will be full documentation outlining how a user can fully utilize OpenAIKit.
  • As well, more features listed in ToDo will be fully implemented.
  • More examples, from other platforms, will be uploaded for developers to be able to focus more on implementing the end user experience, and less time figuring out their project's architecture.

Credits

I would like to personally thank the OpenAI Team for implementing the REST endpoint and implementing the models themselves, as without them, this project wouldn't have been possible.

As well, I would like to personally thank YufeiG for providing troubleshooting help on sending Image data for the Image Edit and Variations endpoints.

License

OpenAIKit is released under the MIT license, and any use of OpenAI's REST endpoint will be under the Usage policies set by them. See LICENSE for details.

About

Swift Package for OpenAI's API

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Swift 100.0%