Johnu George
Technical Lead
Cisco Systems
location_on India
Member since 3 years
Share
Johnu George
Specialises In
Johnu George is a technical Lead in Cloud CTO at Cisco Systems. His research interests are in areas of distributed systems and scalable infrastructure for big data applications. He is an active open source contributor and currently a PMC member of Apache Mnemonic. He is currently involved in developing Kubeflow stack and one of the key maintainers of Katib and Tensorflow/PyTorch operators in Kubeflow. He has been a speaker at various conferences including Apache Big Data conference.
-
keyboard_arrow_down
A Scalable Hyperparameter Optimization framework for ML workloads
Johnu GeorgeTechnical LeadCisco Systems
Ramdoot Kumar PTECHNICAL LEADCisco Systemsschedule 3 years ago
20 Mins
Demonstration
Intermediate
In machine learning, hyperparameters are parameters that governs the training process itself. For example, learning rate, number of hidden layers, number of nodes per layer are typical hyperparameters for neural networks. Hyperparameter Tuning is the process of searching the best hyper parameters to initialize the learning algorithm, thus improving training performance.
We present Katib, a scalable and general hyper parameter tuning framework based on Kubernetes which is ML framework agnostic (Tensorflow, Pytorch, MXNet, XGboost etc). You will learn about Katib in Kubeflow, an open source ML toolkit for Kubernetes, as we demonstrate the advantages of hyperparameter optimization by running a sample classification problem. In addition, as we dive into the implementation details, you will learn how to contribute as we expand this platform to include autoML tools.
-
-
No more submissions exist.
