1938a5af6d1ea3a2d8dccbb831f82990800fa3e1,allennlp/training/metrics/boolean_accuracy.py,BooleanAccuracy,__call__,#BooleanAccuracy#Any#Any#Any#,23

Before Change

        // The .prod() here is functioning as a logical and.
        correct = predictions.eq(gold_labels).prod(dim=1).float()
        count = torch.ones(gold_labels.size(0))
        self._correct_count += correct.sum()
        self._total_count += count.sum()

    def get_metric(self, reset: bool = False):

After Change

            // We want to skip predictions that are completely masked;
            // so we"ll keep predictions that aren"t.
            keep = mask.view(batch_size, -1).max(dim=1)[0].float()
        else:
            keep = torch.ones(batch_size).float()

        predictions = predictions.view(batch_size, -1)

Instances