d9media/openai-text-rewriter repository overview

⁠OpenAI Rewriter Docker Service

This project provides a simple Docker-based service for rewriting content using the OpenAI API. It is designed to be multi-architecture compatible (x86_64, arm64, etc.), allowing it to run on various platforms, including Macs with M1/M2 chips. The service accepts content via a POST request and returns rewritten content.

⁠Features

• Accepts OpenAI prompts and content via POST requests.
• Requires authentication via a custom server API key.
• Supports configurable OpenAI parameters (model, max_completion_tokens, max_tokens, temperature).
• Compatible with both x86_64 (amd64) and arm64 architectures.
• Responses are returned as base64-encoded HTML, maintaining original formatting.

⁠Environment Variables

• OPENAI_API_KEY: Your OpenAI API key (required).
• SERVER_API_KEY: Custom API key for authenticating server requests (required) - this can be any text string you define. Recommended length: 128.
• OPENAI_MODEL: Default OpenAI model to use when a request does not include model (default: gpt-4o-mini).
• SERVER_PORT: The port the server listens on (default: 5000).

⁠Usage

1. Build the Docker Image

   docker buildx build --platform linux/amd64,linux/arm64 -t your_dockerhub_username/openai-rewriter-enhanced --push .
   

   Copy

2. Run the Docker Container

   Replace your_openai_api_key and your_server_api_key with your actual keys:

   docker run -d -p 5000:5000 \
   -e OPENAI_API_KEY=your_openai_api_key \
   -e SERVER_API_KEY=your_server_api_key \
   -e OPENAI_MODEL=gpt-4o-mini \
   -e SERVER_PORT=5000 \
   your_dockerhub_username/openai-rewriter-enhanced
   

   Copy

3. API Request

   The service listens on the root endpoint (/) and accepts POST requests with the following parameters:

   ⁠Request Parameters

   • api_key: The server's custom API key to authenticate requests. (Not your OpenAI API Key. OpenAI is set in your docker container.)
   • task_prompt: The task description for the OpenAI model (e.g., "Rewrite the following article").
   • content: Base64-encoded HTML content to be rewritten.
   • model (optional): Chat Completions-compatible OpenAI model ID to use for this request. Falls back to OPENAI_MODEL.
   • max_completion_tokens (optional): Maximum number of tokens to generate (default: 500).
   • max_tokens (optional): Backward-compatible alias for max_completion_tokens.
   • temperature (optional): The creativity level of the response. Omit it to use the selected model's default.

   ⁠Example PHP Usage

   You can use the following PHP code to make a request to the service and retrieve the rewritten content:

   function rewrite_content($api_key, $task_prompt, $content) {
       $url = 'http://localhost:5000';
       
       // Base64 encode the content
       $encoded_content = base64_encode($content);
   
       // Prepare the data for the POST request
       $data = json_encode([
           'api_key' => $api_key,
           'task_prompt' => $task_prompt,
           'content' => $encoded_content,
           'model' => 'gpt-4o-mini', // optional
           'max_completion_tokens' => 600, // optional
           'temperature' => 0.5  // optional
       ]);
   
       // Use cURL to make the request
       $ch = curl_init($url);
       curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
       curl_setopt($ch, CURLOPT_HTTPHEADER, [
           'Content-Type: application/json'
       ]);
       curl_setopt($ch, CURLOPT_POST, true);
       curl_setopt($ch, CURLOPT_POSTFIELDS, $data);
   
       // Execute the request
       $response = curl_exec($ch);
       curl_close($ch);
   
       // Parse the JSON response
       $responseData = json_decode($response, true);
   
       // Decode the base64 content and return it
       if (isset($responseData['rewritten_content'])) {
           return base64_decode($responseData['rewritten_content']);
       } else {
           return 'Error rewriting content';
       }
   }
   

   Copy

⁠Example Request

The following example shows the structure of a request to the service:

{
    "api_key": "your_server_api_key",
    "task_prompt": "Rewrite the following article:",
    "content": "base64_encoded_html_content",
    "model": "gpt-4o-mini",
    "max_completion_tokens": 600,
    "temperature": 0.5
}

⁠Source code

https://github.com/d9media/openai-text-rewriter⁠

⁠Pull

docker pull d9media/openai-text-rewriter:latest

⁠License

This project is licensed under the MIT License.