Model Validation

Pro supports a visual model validation framework to compare multiple trained models on the same dataset. It allows users to evaluate model performance metric-wise— all in a single view. This helps streamline decision-making by avoiding manual checks and enabling faster deployment choices based on visual insights

Object Detection

• Click on the top right button , and you will see sort: option

  • your ID will be displayed—each ID represents a sample trial.

    For example, if the number of samples is set to 5, there will be 5 trial IDs, each corresponding to a separate experiment run with different parameter combinations.

  • You can also view the Training Loss and Validation Loss for each trial to compare performance and determine which configuration performed best.

Segmentation:

• Click on the top right button , for sorting
  • You can also view the Training Loss and Validation Loss for each trial to compare performance and determine which configuration performed best.