nebula.core.utils.helper

nebula.core.utils.helper#

Functions#

cosine_metric2(model1, model2[, similarity])

cosine_metric(model1, model2[, similarity])

euclidean_metric(model1, model2[, standardized, ...])

minkowski_metric(model1, model2, p[, similarity])

manhattan_metric(model1, model2[, similarity])

pearson_correlation_metric(model1, model2[, similarity])

jaccard_metric(model1, model2[, similarity])

normalise_layers(untrusted_params, trusted_params)

Module Contents#

nebula.core.utils.helper.cosine_metric2(model1, model2, similarity=False)#
Parameters:

  • model1 (OrderedDict[str, torch.Tensor])

  • model2 (OrderedDict[str, torch.Tensor])

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.cosine_metric(model1, model2, similarity=False)#
Parameters:

  • model1 (OrderedDict)

  • model2 (OrderedDict)

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.euclidean_metric(model1, model2, standardized=False, similarity=False)#
Parameters:

  • model1 (OrderedDict[str, torch.Tensor])

  • model2 (OrderedDict[str, torch.Tensor])

  • standardized (bool)

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.minkowski_metric(model1, model2, p, similarity=False)#
Parameters:

  • model1 (OrderedDict[str, torch.Tensor])

  • model2 (OrderedDict[str, torch.Tensor])

  • p (int)

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.manhattan_metric(model1, model2, similarity=False)#
Parameters:

  • model1 (OrderedDict[str, torch.Tensor])

  • model2 (OrderedDict[str, torch.Tensor])

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.pearson_correlation_metric(model1, model2, similarity=False)#
Parameters:

  • model1 (OrderedDict[str, torch.Tensor])

  • model2 (OrderedDict[str, torch.Tensor])

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.jaccard_metric(model1, model2, similarity=False)#
Parameters:

  • model1 (OrderedDict[str, torch.Tensor])

  • model2 (OrderedDict[str, torch.Tensor])

  • similarity (bool)

Return type:

Optional[float]

nebula.core.utils.helper.normalise_layers(untrusted_params, trusted_params)#
Contents