dgs.models.metric.metric.compute_accuracy

dgs.models.metric.metric.compute_accuracy(prediction: torch.Tensor, target: torch.Tensor, topk: list[int] | None = None) → dict[int, float]

Compute the accuracies of a predictor over a tuple of k-top predictions. Will use the k-biggest values in prediction.

Parameters:

• prediction – prediction matrix with shape [B x num_classes].

• target – ground truth labels with shape [B].

• topk – A list containing the number of values to check for the top-k accuracies. Default [1].

Returns:

The accuracies for each of the k-top predictions.