SavviHub overview

Conceptual overview

‚ÄčSavviHub is the platform for machine learning researchers and data scientists to run an experiment on the Kubernetes cluster easily. SavviHub helps to import source code from the git provider and to retrieve the dataset from local disk or cloud vendors. The experiment runs on various types of Kubernetes clusters, and the experiment results are handled (view/export/service) on SavviHub.

SavviHub resources

SavviHub workflow

- Sign up and create a SavviHub organization

Sign up and create an organization on SavviHub.

- Create a project and import source code

Create a new project and import the source code from the git provider on SavviHub.

- Retrieve dataset from external sources

Retrieve the dataset from the local disk or cloud vendors on SavviHub.

- Run an experiment

Run a machine learning experiment on the Kubernetes cluster without configuration on SavviHub.

- Examine results

Examine the training results of experiments on SavviHub. Go to the experiment results page to learn how to analyze the results.

- Manage SavviHub resources with CLI

Manage your SavviHub resources with SavviHub Command-Line Interface(CLI), SavviHub Client. Go to the SavviHub Client page to learn how to use it.