Using the Workspace in Grace for AI Model Development

Overview

The Workspace in Grace provides an integrated environment for coding, developing AI models, and performing data science tasks. Workspaces can be customized with resources, IDE preferences, and development images, giving users flexibility while staying within project quotas.

User Guide

Step 1: Activate the Workspace

1. Navigate to your Project in Grace.
2. Locate the Workspace section.
3. Check the toggle:
   • Gray = Inactive
   • Orange = Active
4. If inactive, click the toggle and select Activate.

Step 2: Configure Resources

1. Once the workspace is active, you can adjust the resources:
2. Click Edit Resource Setting.
3. Choose:
   • CPU allocation
   • Memory (GB)
4. Limits are defined by the Project Administrator’s quota.
5. Example: 1 CPU and 5 GB memory.

Step 3: Select IDE

1. You can choose how to code within Grace:
2. Grace UI IDE
   • Default option is JupyterLab, launched directly in the platform.
3. External IDE
   • Optionally, connect an external IDE (covered separately).

Step 4: Choose Development Image

1. Select the desired runtime environment.
2. Example: Python version (Python 3.11).
3. This ensures compatibility with your project’s dependencies.

Step 5: Start Coding

1. Click Workspace to launch your IDE.
2. By default, JupyterLab opens inside Grace:
   • Provides a native development environment.
   • Side ribbon can be minimized for more space.
   • Use the maximize button (top-right) for full-screen mode, replicating the experience of running JupyterLab locally.

Summary

Step	Description
Activate Workspace	Toggle on to enable development
Configure Resources	Adjust CPU and memory within project quota
Select IDE	Choose JupyterLab (default) or connect an external IDE
Choose Image	Pick runtime environment (e.g., Python 3.11)
Start Coding	Launch IDE and begin development in JupyterLab

By following these steps, you can activate and configure your workspace in Grace, providing a fully functional environment for AI model development and data science projects.

GRACE_MLOPs_workspace_v.1.57.html Displaying GRACE_MLOPs_workspace_v.1.57.html.