Skip to content

semipy.methods.utils.get_metrics

Warning

This section is in construction.

    semipy.methods.utils.get_metrics(config,
                                     num_validation_loaders=1,
                                     num_test_loaders=1,
                                     device='cpu',
                                     **kwargs)

This function loads and reads a custom metrics YAML file. It returns a nn.ModuleDict of MetricsCollection.

Parameters

  • config - Either a dict, AttributeDict or a path to a config file.
  • num_validation_loaders (int) - Number of validation dataloaders in case using multiple validation sets. Default: 1
  • num_test_loaders (int) - Number of test dataloaders in case using multiple test sets. Default: 1
  • device (str) - The device used during training. Default: 'cpu'