Image Editing Service API

Create a Python service, API endpoint, view, and serializer that implement a new image editing function capable of processing multiple image inputs. This function must:

- Accept multiple image files opened in binary mode (rb) as a list and pass them to the OpenAI image editing client.
- Support both "gpt-image-1" model, which accepts multiple images, and "dall-e-2" model, which accepts only one image.
- Use the OpenAI Python client for image editing as shown in the provided sample code.
- Handle exceptions robustly, including but not limited to file handling errors, API errors, and validation errors.
- Implement structured logging using `structlog` to log key events and errors clearly.
- Collect and store OpenAI token usage details (such as total tokens, input and output tokens, and token breakdowns) after each image editing request.
- Return or store the edited image(s) in base64 format after decoding, similar to the sample.

Requirements:

- The API should accept image uploads and a prompt.
- Validate the number of images based on the model selected.
- Ensure clean and maintainable code with appropriate use of serializers for input validation.
- The service layer should encapsulate the interaction with the OpenAI API.


# Steps

1. Design a serializer class to validate the input images and prompt text.
2. Create a service function that:
   - Opens images in 'rb' mode,
   - Calls the appropriate OpenAI image edit method,
   - Handles different model capabilities for multiple images,
   - Logs request and error details with structlog,
   - Extracts and stores token usage.
3. Develop an API view (using Django Rest Framework or similar framework) to:
   - Accept multi-part/form-data including images and prompt,
   - Select the model and enforce image count restrictions,
   - Invoke the service function,
   - Handle and return proper HTTP responses.
4. Implement error handling across all layers with meaningful log entries.

# Output Format

Provide complete, runnable Python code snippets for:
- Serializer class(es).
- Image editing service function/class.
- API view handling the request.

Include inline comments explaining key parts of the code. Use idiomatic and clear Python 3.x syntax.

# Notes

- Assume the OpenAI Python client and structlog are installed and configured.
- You may use Django Rest Framework or a similar web framework for API view and serializer implementations.
- Focus on clarity, error handling, and logging completeness.

# Examples

Use the provided sample code as a reference for calling the OpenAI image editing endpoint with multiple images.