Interface Overview: Foundation Models in BaseModel

The Foundation Models interface in BaseModel provides users with an organized view of all foundational models, their statuses, and relevant actions. This guide breaks down the components of the interface to help users navigate and manage their models effectively.

For more information about building Foundation Models in BaseModel.Ai, please refer to the documentation here.

1. Navigation Panel (Left Sidebar)

The left sidebar provides filters to help users sort and manage their Foundation Models effectively:

• All: Displays all models regardless of their status.
• Draft: Models that are still being configured and not yet finalized.
• Not Trained: Models that have been created but haven't undergone any training.
• Operational: Models that are fully trained and currently in use.

2. Foundation Models List (Main Content Area)

The main panel presents a list of all foundation models and their metadata:

• Name: The model's title. Models without a custom name appear as Untitled model.
• Status: Indicates the current state, such as:
  • DRAFT — Not yet trained
  • OPERATIONAL — Ready and usable
• Last Run: Timestamp of the most recent model training run (if applicable).
• Next Run: When the model is scheduled to run next (if scheduled).
• Edited: Shows how recently the model configuration was updated.

Additional Features:

• Total Results Count: A label at the top (e.g., “54 Results”) displays how many models are currently shown.
• Three-Dot Menu: Each row includes a context menu offering additional actions like editing or deleting the model.

3. Top Bar Actions

• New Foundation Model: A blue button in the top-right corner used to start configuring a new Foundation Model.
• Backup: Opens options for backing up model configurations or metadata.
• Learn More: A link to official documentation and resources about Foundation Models.

This view is typically the starting point in a machine learning workflow within BaseModel, where foundational data logic and schema are established before proceeding to fine-tuning and inference.