Merge pull request MacPaw#160 from kalafus/MacPaw.completions_legacy_removed

remove legacy completions endpoint

Author: nezhyborets

README.md

@@ -15,8 +15,6 @@ This repository contains Swift community-maintained implementation over [OpenAI]
 - [Installation](#installation)
 - [Usage](#usage)
     - [Initialization](#initialization)
-    - [Completions](#completions)
-        - [Completions Streaming](#completions-streaming)
     - [Chats](#chats)
         - [Chats Streaming](#chats-streaming) 
     - [Images](#images)
@@ -84,115 +82,6 @@ let openAI = OpenAI(configuration: configuration)
 
 Once token you posses the token, and the instance is initialized you are ready to make requests.
 
-### Completions
-
-Given a prompt, the model will return one or more predicted completions, and can also return the probabilities of alternative tokens at each position.
-
-**Request**
-
-```swift
-struct CompletionsQuery: Codable {
-    /// ID of the model to use.
-    public let model: Model
-    /// The prompt(s) to generate completions for, encoded as a string, array of strings, array of tokens, or array of token arrays.
-    public let prompt: String
-    /// What sampling temperature to use. Higher values means the model will take more risks. Try 0.9 for more creative applications, and 0 (argmax sampling) for ones with a well-defined answer.
-    public let temperature: Double?
-    /// The maximum number of tokens to generate in the completion.
-    public let maxTokens: Int?
-    /// An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
-    public let topP: Double?
-    /// Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
-    public let frequencyPenalty: Double?
-    /// Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
-    public let presencePenalty: Double?
-    /// Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.
-    public let stop: [String]?
-    /// A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.
-    public let user: String?
-}
-```
-
-**Response**
-
-```swift
-struct CompletionsResult: Codable, Equatable {
-    public struct Choice: Codable, Equatable {
-        public let text: String
-        public let index: Int
-    }
-
-    public let id: String
-    public let object: String
-    public let created: TimeInterval
-    public let model: Model
-    public let choices: [Choice]
-    public let usage: Usage
-}
-```
-**Example**
-
-```swift
-let query = CompletionsQuery(model: .textDavinci_003, prompt: "What is 42?", temperature: 0, maxTokens: 100, topP: 1, frequencyPenalty: 0, presencePenalty: 0, stop: ["\\n"])
-openAI.completions(query: query) { result in
-  //Handle result here
-}
-//or
-let result = try await openAI.completions(query: query)
-```
-
-```
-(lldb) po result
-▿ CompletionsResult
-  - id : "cmpl-6P9be2p2fQlwB7zTOl0NxCOetGmX3"
-  - object : "text_completion"
-  - created : 1671453146.0
-  - model : OpenAI.Model.textDavinci_003
-  ▿ choices : 1 element
-    ▿ 0 : Choice
-      - text : "\n\n42 is the answer to the ultimate question of life, the universe, and everything, according to the book The Hitchhiker\'s Guide to the Galaxy."
-      - index : 0
-```
-
-#### Completions Streaming
-
-Completions streaming is available by using `completionsStream` function. Tokens will be sent one-by-one.
-
-**Closures**
-```swift
-openAI.completionsStream(query: query) { partialResult in
-    switch partialResult {
-    case .success(let result):
-        print(result.choices)
-    case .failure(let error):
-        //Handle chunk error here
-    }
-} completion: { error in
-    //Handle streaming error here
-}
-```
-
-**Combine**
-
-```swift
-openAI
-    .completionsStream(query: query)
-    .sink { completion in
-        //Handle completion result here
-    } receiveValue: { result in
-        //Handle chunk here
-    }.store(in: &cancellables)
-```
-
-**Structured concurrency**
-```swift
-for try await result in openAI.completionsStream(query: query) {
-   //Handle result here
-}
-```
-
-Review [Completions Documentation](https://platform.openai.com/docs/api-reference/completions) for more info.
-
 ### Chats
 
 Using the OpenAI Chat API, you can build your own applications with `gpt-3.5-turbo` to do things like:
@@ -933,7 +822,6 @@ Read more about Cosine Similarity [here](https://en.wikipedia.org/wiki/Cosine_si
 The library contains built-in [Combine](https://developer.apple.com/documentation/combine) extensions.
 
 ```swift
-func completions(query: CompletionsQuery) -> AnyPublisher<CompletionsResult, Error>
 func images(query: ImagesQuery) -> AnyPublisher<ImagesResult, Error>
 func embeddings(query: EmbeddingsQuery) -> AnyPublisher<EmbeddingsResult, Error>
 func chats(query: ChatQuery) -> AnyPublisher<ChatResult, Error>

Sources/OpenAI/OpenAI.swift

@@ -65,14 +65,6 @@ final public class OpenAI: OpenAIProtocol {
         self.init(configuration: configuration, session: session as URLSessionProtocol)
     }
 
-    public func completions(query: CompletionsQuery, completion: @escaping (Result<CompletionsResult, Error>) -> Void) {
-        performRequest(request: JSONRequest<CompletionsResult>(body: query, url: buildURL(path: .completions)), completion: completion)
-    }
-    
-    public func completionsStream(query: CompletionsQuery, onResult: @escaping (Result<CompletionsResult, Error>) -> Void, completion: ((Error?) -> Void)?) {
-        performStreamingRequest(request: JSONRequest<CompletionsResult>(body: query.makeStreamable(), url: buildURL(path: .completions)), onResult: onResult, completion: completion)
-    }
-    
     public func images(query: ImagesQuery, completion: @escaping (Result<ImagesResult, Error>) -> Void) {
         performRequest(request: JSONRequest<ImagesResult>(body: query, url: buildURL(path: .images)), completion: completion)
     }
@@ -225,7 +217,6 @@ extension OpenAI {
 typealias APIPath = String
 extension APIPath {
 
-    static let completions = "/v1/completions"
     static let embeddings = "/v1/embeddings"
     static let chats = "/v1/chat/completions"
     static let models = "/v1/models"

Sources/OpenAI/Public/Models/CompletionsQuery.swift

@@ -80,19 +80,6 @@ public extension Model {
 
     /// Snapshot of `gpt-3.5-turbo-16k` from June 13th 2023. Unlike `gpt-3.5-turbo-16k`, this model will not receive updates, and will be deprecated 3 months after a new version is released.
     static let gpt3_5Turbo_16k_0613 = "gpt-3.5-turbo-16k-0613"
-
-    // Completions
-    
-    /// Can do any language task with better quality, longer output, and consistent instruction-following than the curie, babbage, or ada models. Also supports inserting completions within text.
-    static let textDavinci_003 = "text-davinci-003"
-    /// Similar capabilities to text-davinci-003 but trained with supervised fine-tuning instead of reinforcement learning.
-    static let textDavinci_002 = "text-davinci-002"
-    /// Very capable, faster and lower cost than Davinci.
-    static let textCurie = "text-curie-001"
-    /// Capable of straightforward tasks, very fast, and lower cost.
-    static let textBabbage = "text-babbage-001"
-    /// Capable of very simple tasks, usually the fastest model in the GPT-3 series, and lowest cost.
-    static let textAda = "text-ada-001"
 
     // Speech
 

Sources/OpenAI/Public/Models/CompletionsResult.swift

@@ -12,32 +12,6 @@ import Foundation
 @available(tvOS 13.0, *)
 @available(watchOS 6.0, *)
 public extension OpenAIProtocol {
-    func completions(
-        query: CompletionsQuery
-    ) async throws -> CompletionsResult {
-        try await withCheckedThrowingContinuation { continuation in
-            completions(query: query) { result in
-                switch result {
-                case let .success(success):
-                    return continuation.resume(returning: success)
-                case let .failure(failure):
-                    return continuation.resume(throwing: failure)
-                }
-            }
-        }
-    }
-    
-    func completionsStream(
-        query: CompletionsQuery
-    ) -> AsyncThrowingStream<CompletionsResult, Error> {
-        return AsyncThrowingStream { continuation in
-            return completionsStream(query: query) { result in
-                continuation.yield(with: result)
-            } completion: { error in
-                continuation.finish(throwing: error)
-            }
-        }
-    }
 
     func images(
         query: ImagesQuery

Sources/OpenAI/Public/Models/Models/Models.swift

@@ -15,27 +15,6 @@ import Combine
 @available(watchOS 6.0, *)
 public extension OpenAIProtocol {
 
-    func completions(query: CompletionsQuery) -> AnyPublisher<CompletionsResult, Error> {
-        Future<CompletionsResult, Error> {
-            completions(query: query, completion: $0)
-        }
-        .eraseToAnyPublisher()
-    }
-    
-    func completionsStream(query: CompletionsQuery) -> AnyPublisher<Result<CompletionsResult, Error>, Error> {
-        let progress = PassthroughSubject<Result<CompletionsResult, Error>, Error>()
-        completionsStream(query: query) { result in
-            progress.send(result)
-        } completion: { error in
-            if let error {
-                progress.send(completion: .failure(error))
-            } else {
-                progress.send(completion: .finished)
-            }
-        }
-        return progress.eraseToAnyPublisher()
-    }
-
     func images(query: ImagesQuery) -> AnyPublisher<ImagesResult, Error> {
         Future<ImagesResult, Error> {
             images(query: query, completion: $0)

Sources/OpenAI/Public/Protocols/OpenAIProtocol+Async.swift

@@ -9,41 +9,6 @@ import Foundation
 
 public protocol OpenAIProtocol {
 
-    /**
-     This function sends a completions query to the OpenAI API and retrieves generated completions in response. The Completions API enables you to build applications using OpenAI's language models, like the powerful GPT-3.
-
-     Example:
-     ```
-     let query = CompletionsQuery(model: .textDavinci_003, prompt: "What is 42?")
-     openAI.completions(query: query) { result in
-       //Handle result here
-     }
-     ```
-     
-     - Parameters:
-       - query: A `CompletionsQuery` object containing the input parameters for the API request. This includes the prompt, model, temperature, max tokens, and other settings.
-       - completion: A closure which receives the result when the API request finishes. The closure's parameter, `Result<CompletionsResult, Error>`, will contain either the `CompletionsResult` object with the generated completions, or an error if the request failed.
-    **/
-    func completions(query: CompletionsQuery, completion: @escaping (Result<CompletionsResult, Error>) -> Void)
-
-    /**
-     This function sends a completions query to the OpenAI API and retrieves generated completions in response. The Completions API enables you to build applications using OpenAI's language models, like the powerful GPT-3. The result is returned by chunks.
-
-     Example:
-     ```
-     let query = CompletionsQuery(model: .textDavinci_003, prompt: "What is 42?")
-     openAI.completions(query: query) { result in
-       //Handle result here
-     }
-     ```
-     
-     - Parameters:
-       - query: A `CompletionsQuery` object containing the input parameters for the API request. This includes the prompt, model, temperature, max tokens, and other settings.
-       - onResult: A closure which receives the result when the API request finishes. The closure's parameter, `Result<CompletionsResult, Error>`, will contain either the `CompletionsResult` object with the generated completions, or an error if the request failed.
-       - completion: A closure that is being called when all chunks are delivered or uncrecoverable error occured
-    **/
-    func completionsStream(query: CompletionsQuery, onResult: @escaping (Result<CompletionsResult, Error>) -> Void, completion: ((Error?) -> Void)?)
-    
     /**
      This function sends an images query to the OpenAI API and retrieves generated images in response. The Images Generation API enables you to create various images or graphics using OpenAI's powerful deep learning models.