💾GPU & Storage
This section explains how to monitor GPU usage and manage persistent storage for your containers in AI-Stack.
GPU Usage
The GPU Usage page helps you monitor how much GPU power is being used by each container created by your project members.
On this page, you can:
View GPU usage hours per container
Track total container run time
Check GPU utilization rates
See data grouped by container name
🧠 Tip: Use this info to optimize resource usage and track which containers consume the most GPU power.
Storage Management
Containers in AI-Stack use ephemeral file systems—meaning any data saved inside the container will be lost if it is stopped, restarted, or deleted.
⚠️ Important: Always use a volume to store important files that need to survive beyond the container’s lifecycle.
How to Use Persistent Storage:
Create a volume in your storage settings.
Assign a mount path when launching a container.
The volume acts as external storage, keeping your data safe and accessible across container sessions.
💡 Example Use Case: When training models, use a mounted volume to store datasets, checkpoints, and logs.
Last updated