Skip to main content
Version: 1.0.0

Hyper Parameter Tuning

Hyper parameter helps you to more tune your mode, which can be called by pipeline, it is having a UI where you can see the model learning rate, momentum is giving the better accuracy, there are multiple runs internally which happens in Kubernetes

Components of the Hyperparameter Tuning#

  • HP Submit: here you submit the model which you want to tune for hyper parameter tuning.

  • HP Monitor: here you can see the different hyper parameter for the model in the UI with different loss, learning rate and momentum, also see the graph of how loss is changing.

  • NAS Submit:

  • NAS Monitor: