Dialog/widgets for showing metrics on trained models.

class sleap.gui.dialogs.metrics.DetailedMetricsDialog(cfg_info: sleap.gui.learning.configs.ConfigFileInfo)[source]

Dialog to show detailed metrics for a trained model.


cfg_info – The ConfigFileInfo object (from TrainingConfigsGetter) for the model we want to show.

class sleap.gui.dialogs.metrics.MetricsTableDialog(labels_filename: str = '')[source]

Dialog for showing table with multiple models.

The dialog is can show multiple models, including those which don’t already have metrics (ideally you’d be able to generate evals but this isn’t currently supported).

You can then view details on the models (hyperparameters or more detailed metrics).

The typical use-case is to init dialog with path to labels file, and it will then show all trained models found within subdirectories.

class sleap.gui.dialogs.metrics.MetricsTableModel(items: Optional[list] = None, properties: Optional[List[str]] = None, context: Optional[sleap.gui.commands.CommandContext] = None)[source]

Model (i.e. Qt model/view) for table in MetricsTableDialog.