Discrepancy between torch AUPRC and sklearn area under precision_recall_curve
I am writing a script that extends torch metrics to gives me some additional ease in using AUPRC and and AUROC in torch. Essentially I want them to be wrapped with a task parameter that allows me to select the task on init and an average parameter that allows me to select the micro average as in sklearns implementation (https://scikit-learn.org/stable/modules/generated/sklearn.metrics.roc_auc_score.html).