pygip.models.defense.SurviveWM¶

Module Attributes

class pygip.models.defense.SurviveWM.SurviveWM(dataset, attack_node_fraction, model_path=None)[source]¶

Bases: BaseDefense

_abc_impl = <_abc_data object>¶

_load_model()[source]¶: Load a pre-trained model.

_to_cpu(tensor)[source]¶: Safely move tensor to CPU for NumPy operations

combine_with_trigger(base_graph, base_features, base_labels, trigger_data)[source]¶

compute_metrics(y_true, y_pred, y_score=None)[source]¶

defend()[source]¶: Execute the defense mechanism.

generate_key_graph(num_nodes=10, edge_prob=0.3)[source]¶

snn_loss(x, y, T=0.5)[source]¶

supported_api_types = {'dgl'}¶

train_with_snnl(model, graph, features, labels, train_mask, optimizer, T=0.5, alpha=0.1)[source]¶

verify_watermark(model, trigger_graph, trigger_labels)¶