#45 - Validating using RxSwift, CoreML, Jira CLI

Apple has definitely invested a lot in machine learning in the past years. Both internally to power tech like Siri, and externally with CoreML for developers.

It’s very exciting to see how the open source community has jumped on it. For the 3rd week we have interesting CoreML projects to share 🤖🎉.


Graph illustrating how to use a .mlmodel with MLModelCamera: just drag-and-drop it

MLModelCamera is an open source iOS app built by Shuichi Tsutsumi to quickly test trained CoreML models. Simply put the .mlmodel file or files in the /models folder and build on your device.


Reza Shirazian with a very deep dive into the new quantization feature introduced in WWDC 2018.

Model size is one of the most common reason for skipping using a local model and opting for an online cloud solution. Fully trained models can go into 100s of MBs and can easily deter potential users from downloading our app. However if you followed WWDC’s What’s new in Core ML session we got a taste of quantization. An approach that can possibly cut down the size of a fully trained model by two third without losing much in accuracy or performance.

An example of quantization applied to a pizza image 🍕


A validation library from Geum Sang Yoo built on to for RxSwift providing a easy to read API, and aiming to be flexible and extensible.

Validate.to("There's nothing like spaghetti bolognese 🍝")
    .validate(StringIsNotOverflowThen(maxLength: 140))


Taiki Suzuki share a CLI client for world famous Atlassian’s project manager Jira.

Even if you’re not a Jira user you might find this project interesting as an example of how to build a CLI using Swift and Swift Package Manager.

Other interesting projects