2021-11-14, 11:00–12:00, Room 3
--Interested in one of the hottest trends in ML for production environent? 🔥
--Curious about how to deploy your ML model effortless? 🚀
This workshop is designed for anyone wants to take his/her ML models a step further 🦸
MLOps are becoming an essential component in order to automate your ML project lifecycle.
As machine learning models become part of real-world applications, is vital for engineers to shift from the research oriented approach to business and product needs.
The purpose of the workshop is to demonstrate MLFlow capabilities to machine learning and open source community. MLFlow is a powerful tool, that can integrate with most of the modern ML frameworks, while is adopted by many famous organization, to leverage a machine learning lifecycle e.g. keep track of ML projects, log different models with numerous of parameters, register and deploy models to production.
The workshop will demonstrate MLFlow API covering the following topics:
- install MLFlow and walkthrough in MLFlow server
- develop a ml pipeline
- train/evaluate models
- use MLFlow to track parameters and log trained models and datasets
- deploy models and serve them with the built-in MLFlow API
- consume the deployed models through its MLFlow built-in endpoint
Prerequisites: Knowledge in Python, general ML concepts, Familiarity w/ Linux
Stavros Niafas has received his diploma in Computer Engineering from TEI of Central Greece. He also holds an MSc in Image Synthesis & Multimedia and an Msc in Data Science from NCSR Demokritos and the University of Peloponnese.
He works as Machine Learning engineer in Digital Market Intelligence while his research interests expand in domains of Machine/Deep learning, Computer Vision and data-centric AI. He is also actively engaged in systems engineering, FLOSS & photography.