coreml-web-api

Deploy CoreML models on the server with Vapor

This is a prototype project to demonstrate one way to deploy CoreML models on the server with Vapor. It is not intended for production use.

Overview

This project demonstrates one way how to deploy a CoreML model on the server with Vapor. The project includes a simple Vapor server that accepts images and returns a prediction using a CoreML model. The model is a pre-trained Resnet50 model. The server uses this to classify the primary subject of images uploaded to the server.

In the Box

• Server - Vapor server with CoreML model, and Dockerfile
• SwiftUI iOS client

Prerequisites

• Xcode 15 (client)
• Swift 5.5

Getting started

1. Clone repo
2. Open terminal and navigate to the project directory
3. Open the project with Xcode. xed .
4. cd server && swift run App serve
5. Back in Xcode, set the development team in Signing & Capabilities to your own team/profile
6. Select the Run scheme of your choice and run the project in Xcode
7. On simulator, select Photo Library then one of the default images, then Upload Image
8. Get results!

NOTE: additional steps may (most likely will) be required to run the project on a physical device. see The Serve Command about binding to host and port.

---

Source code for the article Deploy CoreML Models on the Server with Vapor