Shortcuts

# Functional Metrics¶

## Aggregation Metrics¶

 auc Computes Area Under the Curve (AUC) using the trapezoidal rule. mean Compute weighted mean. sum Compute weighted sum. throughput Calculate the throughput value which is the number of elements processed per second.

## Classification Metrics¶

 binary_accuracy Compute binary accuracy score, which is the frequency of input matching target. binary_auprc Compute AUPRC, also called Average Precision, which is the area under the Precision-Recall Curve, for binary classification. binary_auroc Compute AUROC, which is the area under the ROC Curve, for binary classification. binary_binned_auroc Compute AUROC, which is the area under the ROC Curve, for binary classification. binary_binned_precision_recall_curve Compute precision recall curve with given thresholds. binary_confusion_matrix Compute binary confusion matrix, a 2 by 2 tensor with counts ( (true positive, false negative) , (false positive, true negative) ) binary_f1_score Compute binary f1 score, the harmonic mean of precision and recall. binary_normalized_entropy Compute the normalized binary cross entropy between predicted input and ground-truth binary target. binary_precision Compute precision score for binary classification class, which is calculated as the ratio between the number of true positives (TP) and the total number of predicted positives (TP + FP). binary_precision_recall_curve Returns precision-recall pairs and their corresponding thresholds for binary classification tasks. binary_recall Compute recall score for binary classification class, which is calculated as the ratio between the number of true positives (TP) and the total number of actual positives (TP + FN). binary_recall_at_fixed_precision Returns the highest possible recall value given the minimum precision for binary classification tasks. multiclass_accuracy Compute accuracy score, which is the frequency of input matching target. multiclass_auprc Compute AUPRC, also called Average Precision, which is the area under the Precision-Recall Curve, for multiclass classification. multiclass_auroc Compute AUROC, which is the area under the ROC Curve, for multiclass classification. multiclass_binned_auroc Compute AUROC, which is the area under the ROC Curve, for multiclass classification. multiclass_binned_precision_recall_curve Compute precision recall curve with given thresholds. multiclass_confusion_matrix Compute multi-class confusion matrix, a matrix of dimension num_classes x num_classes where each element at position (i,j) is the number of examples with true class i that were predicted to be class j. multiclass_f1_score Compute f1 score, which is defined as the harmonic mean of precision and recall. multiclass_precision Compute precision score, which is the ratio of the true positives (TP) and the total number of points classified as positives (TP + FP). multiclass_precision_recall_curve Returns precision-recall pairs and their corresponding thresholds for multi-class classification tasks. multiclass_recall Compute recall score, which is calculated as the ratio between the number of true positives (TP) and the total number of actual positives (TP + FN). multilabel_accuracy Compute multilabel accuracy score, which is the frequency of input matching target. multilabel_auprc Compute AUPRC, also called Average Precision, which is the area under the Precision-Recall Curve, for multilabel classification. multilabel_precision_recall_curve Returns precision-recall pairs and their corresponding thresholds for multi-label classification tasks. multilabel_recall_at_fixed_precision Returns the highest possible recall value give the minimum precision for each label and their corresponding thresholds for multi-label classification tasks. topk_multilabel_accuracy Compute multilabel accuracy score, which is the frequency of the top k label predicted matching target.

## Ranking Metrics¶

 click_through_rate Compute the click through rate given a click events. frequency_at_k Calculate the frequency given a list of frequencies and threshold k. hit_rate Compute the hit rate of the correct class among the top predicted classes. num_collisions Compute the number of collisions given a list of input(ids). reciprocal_rank Compute the reciprocal rank of the correct class among the top predicted classes. weighted_calibration Compute weighted calibration metric.

## Regression Metrics¶

 mean_squared_error Compute Mean Squared Error, which is the mean of squared error of input and target Its class version is torcheval.metrics.MeanSquaredError. r2_score Compute R-squared score, which is the proportion of variance in the dependent variable that can be explained by the independent variable.

## Text Metrics¶

 bleu_score Compute BLEU score given translations and references for each translation. perplexity Perplexity measures how well a model predicts sample data. word_error_rate Compute the word error rate of the predicted word sequence(s) with the reference word sequence(s). word_information_preserved Compute the word information preserved score of the predicted word sequence(s) against the reference word sequence(s). word_information_lost Word Information Lost rate is a metric of the performance of an automatic speech recognition system.

## Docs

Access comprehensive developer documentation for PyTorch

View Docs

## Tutorials

Get in-depth tutorials for beginners and advanced developers

View Tutorials