mmcls.core¶

This package includes some runtime components. These components are useful in classification tasks but not supported by MMCV yet.

Note

Some components may be moved to MMCV in the future.

mmcls.core

Evaluation¶

Evaluation metrics calculation functions

 precision Calculate precision according to the prediction and target. recall Calculate recall according to the prediction and target. f1_score Calculate F1 score according to the prediction and target. precision_recall_f1 Calculate precision, recall and f1 score according to the prediction and target. average_precision Calculate the average precision for a single class. mAP Calculate the mean average precision with respect of classes. support Calculate the total number of occurrences of each label according to the prediction and target. average_performance Calculate CP, CR, CF1, OP, OR, OF1, where C stands for per-class average, O stands for overall average, P stands for precision, R stands for recall and F1 stands for F1-score. calculate_confusion_matrix Calculate confusion matrix according to the prediction and target.

Hook¶

 ClassNumCheckHook PreciseBNHook Precise BN hook. CosineAnnealingCooldownLrUpdaterHook Cosine annealing learning rate scheduler with cooldown. MMClsWandbHook Enhanced Wandb logger hook for classification.

Optimizers¶

 Lamb A pure pytorch variant of FuseLAMB (NvLamb variant) optimizer.