SambaStudio introduction

The left panel contains a selectable list of navigation menu items, including:

The SambaStudio release version number is located at the bottom of the left panel.

The top menu bar provides information items, including:

The Dashboard displays platform relative information via the Jobs and Live endpoints tables. Each table is sortable by column and includes robust filtering options.

See Dashboard for more information.

A project is logical grouping of workloads and artifacts within a tenant. Projects are used to group jobs for training and batch inference as well as endpoints for deployment. Creating a project is typically the first step of a workflow and is done by clicking New project. See Projects for more information.

Within the SambaStudio platform, jobs run training and batch inference while endpoints generate predictions from models. Both are assigned to projects. Jobs and endpoints are created from a project window, which assigns the job or endpoint to that project. You can fine-tune existing models by creating a new training job. You can generate predictions from your models on your data by deploying it to an endpoint. Or, you can generate predictions on bulk data by creating a batch inference job. You can select either a platform provided dataset or your own dataset.

The Model Hub provides a two panel interface for viewing the SambaStudio model cards. The right panel displays a grid layout of model cards. Each model card describes one of the available models. The left panel hosts a robust set of filter options that refine the display of the model cards panel. See Model Hub for more information.

The Dataset Hub provides an interface for managing datasets. The top of the window displays the number of SambaNova and user datasets available in the platform. A list of datasets is displayed in a multi-column table that can be filtered via the checkboxes and filter drop-down options. See Datasets for more information.

The Playground provides an in-platform experience for generating predictions using deployed generative tuning endpoints. You can choose between a chat mode and completion mode experience. Chat mode provides a word-by-word response to your prompt with follow-on prompts kept within the context of your conversation. Completion mode provides complete statement responses to your prompt. A user preset option is available to populate the editor and quickly experience generative tuning. Alternatively, you can input text directly into the editor, without selecting a user preset. See Playground for more information.

SambaStudio provides three predefined roles to be assigned to users within the platform: organization administrator (OrgAdmin), tenant administrator (TenantAdmin), and user. A user can be part of and assigned to multiple tenants. User management and role assignment is performed by an organization administrator (OrgAdmin) or tenant administrator (TenantAdmin) from the User management window. See Users and management for more information.

A tenant is an allotment of users and resources within an organization. Organizations can leverage tenants for distinct purposes, such as development or production, and then share common resources across tenants. Each tenant has its own resources, users, and workloads. SambaStudio allows organizations to allocate resources across multiple tenants. Tenant management is performed by an organization administrator (OrgAdmin). See Tenants and management for more information.

The Notifications panel displays platform specific messages in a scrollable list.

The help drop-down is accessed by clicking the help icon in the top menu bar. It provides links to the SambaNova support page and the SambaNova resources.

A help system is available within the snapi CLI by typing --help after a command. The command below displays the options available for the snapi --help command.