SDMGRModuleLoss

class mmocr.models.kie.SDMGRModuleLoss(weight_node=1.0, weight_edge=1.0, ignore_idx=- 100)

The implementation the loss of key information extraction proposed in the paper: Spatial Dual-Modality Graph Reasoning for Key Information Extraction.

Parameters

• weight_node (float) – Weight of node loss. Defaults to 1.0.

• weight_edge (float) – Weight of edge loss. Defaults to 1.0.

• ignore_idx (int) – Node label to ignore. Defaults to -100.

Return type

None

forward(preds, data_samples)

Forward function.

Parameters

• preds (tuple(Tensor, Tensor)) –

• data_samples (list[KIEDataSample]) – A list of datasamples containing gt_instances.labels and gt_instances.edge_labels.

Returns

Loss dict, containing loss_node, loss_edge, acc_node and acc_edge.

Return type

dict(str, Tensor)