azamsharp/trySwiftCreateMLWorkshop

trySwift Create ML Workshop

Prerequisites

• Knowledge of Swift language is recommended
• Knowledge of SwiftUI framework is recommended
• Xcode 14.X is recommended
• Physical device is recommended. Make sure you can run iOS apps on your physical device through Xcode.
• Please make sure you have a Kaggle account. We wil use Kaggle to download datasets for our workshop.

Projects

Image Classfication (Existing Model)

• Download the MobileNetV2 Image Classification Model
• Image Classification Starter Project
• Download Final Project

Image Classification (Cat and Dogs)

• Kaggle: Download Cats and Dogs Image Classification Dataset
• Image Classification Starter Project (Same as above)
• Image Classification Dogs and Cats Final
• Activity: Fruits Classification Dataset

Object Detection

• Article - Object Detection with Create ML
• Visio ML Annotating Images
• Object Detection Final

Sentiment Analysis

• Downloading the Financial News Dataset
• Sentiment Analysis Swift Playgrounds
• OPTIONAL: Python Script
• OPTIONAL: Swift Script to Create Folders
• Sentiment Analysis Starter Project
• Sentiment Analysis Final Project
• Activity: Movies Review Dataset

NOTE: Dummy sentiment analysis data can be generated by ChatGPT. Dummy data should not be used in actual apps.

Tabular Data

• Carvana Project Starter
• Carvana Dataset from Kaggle
• Carvana Project Final

Activity (Coffee & Sleep)

• DataSet as CSV File

The following files contains the final code. But I encourage you to train the Coffee & Sleep model yourself and then create a SwiftUI/UIKit app to consume the model.

• Xcode Playgrounds File - Final
• SwiftUI Project Final