Loading a Model

Requirements

• A .rime model package (see What is CMF?)
• The session must have the signal channels the model requires

Load a model

Models → Load Model (or the model loader button in the toolbar)

1. Browse to the .rime package folder or zip file
2. RIME reads the model's config.json and validates that the required signals are present in the session
3. The model appears in the Model Runner panel

Run inference

In the Model Runner panel:

• Click Run to run the model on the current session
• Inference progress is shown in the panel
• Output is overlaid on the timeline in the configured lane

Adjust model parameters

Click Settings in the Model Runner panel to adjust:

• Detection threshold
• Window size / stride
• Any model-specific parameters declared in config.json

Included models

The RIME repository includes several example .rime packages in models/:

Package	Description
freeze-index.rime	Freeze Index from accelerometer signal
step-detector.rime	Step detection from foot accelerometer
walking-classifier.rime	Walking vs. non-walking classifier
movement-video.rime	Movement detection from video