Skip to content

zyz954489346/opanai-php

BranchesTags
 
 

Repository files navigation

OpenAI GPT-3 Api Client in PHP


Fork from orhanerday/open-ai
Add http exception catching Add curl prox support

Requires PHP 7.4+

Update Record

date features
2023-04-12 - add support to Azure OpenAI API

Endpoint Support

Installation

You can install the package via composer:

composer require ze/openai-php

Quick Start

Before you get starting, you should set OPENAI_API_KEY as ENV key, and set OpenAI key as env value with the following commands;

Powershell

$Env:OPENAI_API_KEY = "sk-gjtv....."

Cmd

set OPENAI_API_KEY=sk-gjtv.....

Linux or macOS

export OPENAI_API_KEY=sk-gjtv.....

Getting issues while setting up env? Please read the article or you can check my StackOverflow answer for the Windows® ENV setup.

Create your index.php file and paste the following code part into the file.

<?php

require __DIR__ . '/vendor/autoload.php'; // remove this line if you use a PHP Framework.

use Ze\OpenAi\OpenAi;

$open_ai_key = getenv('OPENAI_API_KEY');
$open_ai = new OpenAi($open_ai_key);

$complete = $open_ai->chat([
   'model' => 'gpt-3.5-turbo',
   'messages' => [
       [
           "role" => "system",
           "content" => "You are a helpful assistant."
       ],
       [
           "role" => "user",
           "content" => "Who won the world series in 2020?"
       ],
       [
           "role" => "assistant",
           "content" => "The Los Angeles Dodgers won the World Series in 2020."
       ],
       [
           "role" => "user",
           "content" => "Where was it played?"
       ],
   ],
   'temperature' => 1.0,
   'max_tokens' => 4000,
   'frequency_penalty' => 0,
   'presence_penalty' => 0,
]);

var_dump($complete);

Run the server with the following command

php -S localhost:8000 -t .

Usage

Load your key from an environment variable.

According to the following code $open_ai is the base variable for all open-ai operations.

use Ze\OpenAi\OpenAi;

$open_ai = new OpenAi(env('OPEN_AI_API_KEY'));

Requesting organization

For users who belong to multiple organizations, you can pass a header to specify which organization is used for an API request. Usage from these API requests will count against the specified organization's subscription quota.

$open_ai_key = getenv('OPENAI_API_KEY');
$open_ai = new OpenAi($open_ai_key, "org-IKN2E1nI3kFYU8ywaqgFRKqi");

Custom URL

You can specify Origin URL with the third parameter of the OpenAI constructor method;

$open_ai_key = getenv('OPENAI_API_KEY');
$organization = ""; // the empty string means there is no organization
$originURL = "https://ai.example.com/"; // the empty string mean the origin URL is 'https://api.openai.com'
$open_ai = new OpenAi($open_ai_key, $organization, $originURL);

Chat (as known as ChatGPT API)

Given a chat conversation, the model will return a chat completion response.

$complete = $open_ai->chat([
   'model' => 'gpt-3.5-turbo',
   'messages' => [
       [
           "role" => "system",
           "content" => "You are a helpful assistant."
       ],
       [
           "role" => "user",
           "content" => "Who won the world series in 2020?"
       ],
       [
           "role" => "assistant",
           "content" => "The Los Angeles Dodgers won the World Series in 2020."
       ],
       [
           "role" => "user",
           "content" => "Where was it played?"
       ],
   ],
   'temperature' => 1.0,
   'max_tokens' => 4000,
   'frequency_penalty' => 0,
   'presence_penalty' => 0,
]);

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.

$complete = $open_ai->completion([
   'model' => 'text-davinci-002',
   'prompt' => 'Hello',
   'temperature' => 0.9,
   'max_tokens' => 150,
   'frequency_penalty' => 0,
   'presence_penalty' => 0.6,
]);

Stream Example

This feature might sound familiar from ChatGPT.

Whether to stream back partial progress. If set, tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message.

$open_ai = new OpenAi(env('OPEN_AI_API_KEY'));

$opts = [
   'prompt' => "Hello",
   'temperature' => 0.9,
   "max_tokens" => 150,
   "frequency_penalty" => 0,
   "presence_penalty" => 0.6,
   "stream" => true,
];

header('Content-type: text/event-stream');
header('Cache-Control: no-cache');

$open_ai->completion($opts, function ($curl_info, $data) {
   echo $data . "<br><br>";
   echo PHP_EOL;
   ob_flush();
   flush();
   return strlen($data);
});

Add this part inside <body> of the HTML

<div id="divID">Hello</div>
<script>
   var eventSource = new EventSource("/");
   var div = document.getElementById('divID');


   eventSource.onmessage = function (e) {
      if(e.data == "[DONE]")
      {
          div.innerHTML += "<br><br>Hello";
      }
       div.innerHTML += JSON.parse(e.data).choices[0].text;
   };
   eventSource.onerror = function (e) {
       console.log(e);
   };
</script>

You should see a response like the in video;

stream-event.mp4

Edits

Creates a new edit for the provided input, instruction, and parameters

   $result = $open_ai->createEdit([
       "model" => "text-davinci-edit-001",
       "input" => "What day of the wek is it?",
       "instruction" => "Fix the spelling mistakes",
   ]);

Images (DALL·E)

All DALL·E Examples available in this repo.

Given a prompt, the model will return one or more generated images as urls or base64 encoded.

Create image

Creates an image given a prompt.

$complete = $open_ai->image([
   "prompt" => "A cat drinking milk",
   "n" => 1,
   "size" => "256x256",
   "response_format" => "url",
]);

Create image edit

Creates an edited or extended image given an original image and a prompt.

You need HTML upload for image edit or variation? Please check DALL·E Examples

$otter = curl_file_create(__DIR__ . './files/otter.png');
$mask = curl_file_create(__DIR__ . './files/mask.jpg');

$result = $open_ai->imageEdit([
    "image" => $otter,
    "mask" => $mask,
    "prompt" => "A cute baby sea otter wearing a beret",
    "n" => 2,
    "size" => "1024x1024",
]);

Create image variation

Creates a variation of a given image.

$otter = curl_file_create(__DIR__ . './files/otter.png');

$result = $open_ai->createImageVariation([
    "image" => $otter,
    "n" => 2,
    "size" => "256x256",
]);

Searches

(Deprecated)

This endpoint is deprecated and will be removed on December 3rd, 2022 OpenAI developed new methods with better performance. Learn more.

Given a query and a set of documents or labels, the model ranks each document based on its semantic similarity to the provided query.

$search = $open_ai->search([
    'engine' => 'ada',
    'documents' => ['White House', 'hospital', 'school'],
    'query' => 'the president',
]);

Embeddings

Get a vector representation of a given input that can be easily consumed by machine learning models and algorithms.

Related guide: Embeddings

Create embeddings

$result = $open_ai->embeddings([
    "model" => "text-similarity-babbage-001",
    "input" => "The food was delicious and the waiter..."
]);

Answers

(Deprecated)

This endpoint is deprecated and will be removed on December 3rd, 2022 We’ve developed new methods with better performance. Learn more.

Given a question, a set of documents, and some examples, the API generates an answer to the question based on the information in the set of documents. This is useful for question-answering applications on sources of truth, like company documentation or a knowledge base.

$answer = $open_ai->answer([
  'documents' => ['Puppy A is happy.', 'Puppy B is sad.'],
  'question' => 'which puppy is happy?',
  'search_model' => 'ada',
  'model' => 'curie',
  'examples_context' => 'In 2017, U.S. life expectancy was 78.6 years.',
  'examples' => [['What is human life expectancy in the United States?', '78 years.']],
  'max_tokens' => 5,
  'stop' => ["\n", '<|endoftext|>'],
]);

Classifications

(Deprecated)

This endpoint is deprecated and will be removed on December 3rd, 2022 OpenAI developed new methods with better performance. Learn more.

Given a query and a set of labeled examples, the model will predict the most likely label for the query. Useful as a drop-in replacement for any ML classification or text-to-label task.

$classification = $open_ai->classification([
   'examples' => [
       ['A happy moment', 'Positive'],
       ['I am sad.', 'Negative'],
       ['I am feeling awesome', 'Positive'],
   ],
   'labels' => ['Positive', 'Negative', 'Neutral'],
   'query' => 'It is a raining day =>(',
   'search_model' => 'ada',
   'model' => 'curie',
]);

Content Moderations

Given a input text, outputs if the model classifies it as violating OpenAI's content policy.

$flags = $open_ai->moderation([
    'input' => 'I want to kill them.'
]);

Know more about Content Moderations here: OpenAI Moderations

List engines

(Deprecated)

The Engines endpoints are deprecated. Please use their replacement, Models, instead. Learn more.

Lists the currently available engines, and provides basic information about each one such as the owner and availability.

$engines = $open_ai->engines();

Files

Files are used to upload documents that can be used across features like Answers, Search, and Classifications

List files

Returns a list of files that belong to the user's organization.

$files = $open_ai->listFiles();

Upload file

Upload a file that contains document(s) to be used across various endpoints/features. Currently, the size of all the files uploaded by one organization can be up to 1 GB. Please contact OpenAI if you need to increase the storage limit.

$c_file = curl_file_create(__DIR__ . 'files/sample_file_1.jsonl');
$result = $open_ai->uploadFile([
            "purpose" => "answers",
            "file" => $c_file,
]);

Upload file with HTML Form

<form action="index.php" method="post" enctype="multipart/form-data">
    Select file to upload:
    <input type="file" name="fileToUpload" id="fileToUpload">
    <input type="submit" value="Upload File" name="submit">
</form>
<?php
require __DIR__ . '/vendor/autoload.php';

use Ze\OpenAi\OpenAi;

if ($_SERVER['REQUEST_METHOD'] == 'POST') {
    ob_clean();
    $open_ai = new OpenAi(env('OPEN_AI_API_KEY'));
    $tmp_file = $_FILES['fileToUpload']['tmp_name'];
    $file_name = basename($_FILES['fileToUpload']['name']);
    $c_file = curl_file_create($tmp_file, $_FILES['fileToUpload']['type'], $file_name);

    echo "[";
    echo $open_ai->uploadFile(
        [
            "purpose" => "answers",
            "file" => $c_file,
        ]
    );
    echo ",";
    echo $open_ai->listFiles();
    echo "]";

}

Delete file

$result = $open_ai->deleteFile('file-xxxxxxxx');

Retrieve file

$file = $open_ai->retrieveFile('file-xxxxxxxx');

Retrieve file content

$file = $open_ai->retrieveFileContent('file-xxxxxxxx');

Fine-tunes

Manage fine-tuning jobs to tailor a model to your specific training data.

Create fine-tune

$result = $open_ai->createFineTune([
       "training_file" => "file-U3KoAAtGsjUKSPXwEUDdtw86",
]);

List fine-tune

$fine_tunes = $open_ai->listFineTunes();

Retrieve fine-tune

$fine_tune = $open_ai->retrieveFineTune('ft-AF1WoRqd3aJAHsqc9NY7iL8F');

Cancel fine-tune

$result = $open_ai->cancelFineTune('ft-AF1WoRqd3aJAHsqc9NY7iL8F');

List fine-tune events

$fine_tune_events = $open_ai->listFineTuneEvents('ft-AF1WoRqd3aJAHsqc9NY7iL8F');

Delete fine-tune model

$result = $open_ai->deleteFineTune('curie:ft-acmeco-2021-03-03-21-44-20');

Retrieve engine

(Deprecated)

Retrieves an engine instance, providing basic information about the engine such as the owner and availability.

$engine = $open_ai->engine('davinci');

Models

List and describe the various models available in the API.

List models

Lists the currently available models, and provides basic information about each one such as the owner and availability.

$result = $open_ai->listModels();

Retrieve model

Retrieves a model instance, providing basic information about the model such as the owner and permissioning.

$result = $open_ai->retrieveModel("text-ada-001");

Printing results i.e. $search

echo $search;

Testing

To run all tests:

composer test

To run only those tests that work for most user (exclude those that require a missing folder or that hit deprecated endpoints no longer available to most users):

./vendor/bin/pest --group=working

License

The MIT License (MIT). Please see License File for more information.

About

OpenAI PHP SDK

packagist.org/packages/orhanerday/open-ai

Resources

Readme

License

MIT license

Security policy

Security policy
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases 2

v1.1 Latest
Mar 7, 2023
+ 1 release

Packages

No packages published

Languages

  • PHP 100.0%