We will transfer the funds to verified humanitarian aid foundations to Ukraine 🇺🇦
For the last couple of years, there were countless case studies of how client-side applications take over more and more tasks traditionally handled by the backend. How does it apply to ML? How can we run models on our client’s devices and why can it be beneficial? The goal of this workshop is to answer these questions in the context of the iOS platform. You will gain a hands-on experience with CoreML, how it is used to train models from scratch in Xcode, convert existing models implemented on other ML frameworks, and deploy them to the iOS environment.
The number of participants: up to 30 participants
Approximate duration: 2 hours. Starts on May 18, at 3:00 p.m. till 5 p.m. (GMT+3)
The event will be held in English.
This workshop is aimed at software engineers who are interested in iOS development and ML. Basic Swift and ML-related Python knowledge will be a plus.
The workshop is run fully online, using Zoom and various online collaboration tools.
Attendees will be writing and running Xcode apps during the workshop, so will need to join the workshop on a computer running macOS with Xcode installed and configured. Also, make sure your computer has Python 3 and you are able to run scripts and install packages via pip or conda. Some of the tasks will involve working with peripherals so you’ll need a physical Apple device to run such iOS apps (emulator won’t work in such cases) - iPhone, iPad, or Mac with a functioning front camera.
Structure and Contents:
1. CoreML and CreateML. How and when to use each of them. Latest features
2. Exercise: coremltools - convert PyTorch models to CoreML format in Python
3. Exercise: train CoreML text classifier in CreateML
4. Exercise: app with CoreML image classifier
5. Exercise: app which utilizes camera with a background removal model
6. Exercise: app featuring on-device training of the recommender model
There is limited space available and we’d love to see you there, so please buy a ticket to secure your spot.