hL*SSKJr SSKJr SSKJr SSKrSSKJr SSK J r J r SSK Jr SSKJrJr SSKJrJr SS KJr "S S \5rg) ) annotations)Path)AnyN)ClassificationDatasetbuild_dataloader) BaseValidator)LOGGERRANK)ClassifyMetricsConfusionMatrix) plot_imagesc^\rSrSrSrSSU4SjjjrSSjrSSjrSSjrSSjr SSjr SS jr SS jr SS jr SS jrSS jrSSjrSSjrSSjrSrU=r$)ClassificationValidatorau A class extending the BaseValidator class for validation based on a classification model. This validator handles the validation process for classification models, including metrics calculation, confusion matrix generation, and visualization of results. Attributes: targets (list[torch.Tensor]): Ground truth class labels. pred (list[torch.Tensor]): Model predictions. metrics (ClassifyMetrics): Object to calculate and store classification metrics. names (dict): Mapping of class indices to class names. nc (int): Number of classes. confusion_matrix (ConfusionMatrix): Matrix to evaluate model performance across classes. Methods: get_desc: Return a formatted string summarizing classification metrics. init_metrics: Initialize confusion matrix, class names, and tracking containers. preprocess: Preprocess input batch by moving data to device. update_metrics: Update running metrics with model predictions and batch targets. finalize_metrics: Finalize metrics including confusion matrix and processing speed. postprocess: Extract the primary prediction from model output. get_stats: Calculate and return a dictionary of metrics. build_dataset: Create a ClassificationDataset instance for validation. get_dataloader: Build and return a data loader for classification validation. print_results: Print evaluation metrics for the classification model. plot_val_samples: Plot validation image samples with their ground truth labels. plot_predictions: Plot images with their predicted class labels. Examples: >>> from ultralytics.models.yolo.classify import ClassificationValidator >>> args = dict(model="yolo11n-cls.pt", data="imagenet10") >>> validator = ClassificationValidator(args=args) >>> validator() Notes: Torchvision classification models can also be passed to the 'model' argument, i.e. model='resnet18'. c>[TU]XX45 SUlSUlSURl[ 5Ulg)a Initialize ClassificationValidator with dataloader, save directory, and other parameters. Args: dataloader (torch.utils.data.DataLoader, optional): Dataloader to use for validation. save_dir (str | Path, optional): Directory to save results. args (dict, optional): Arguments containing model and validation configuration. _callbacks (list, optional): List of callback functions to be called during validation. Examples: >>> from ultralytics.models.yolo.classify import ClassificationValidator >>> args = dict(model="yolo11n-cls.pt", data="imagenet10") >>> validator = ClassificationValidator(args=args) >>> validator() Nclassify)super__init__targetspredargstaskr metrics)self dataloadersave_dirr _callbacks __class__s ^/home/james-whalen/.local/lib/python3.13/site-packages/ultralytics/models/yolo/classify/val.pyr ClassificationValidator.__init__9s7 t@  # &( c SS-$)z=Return a formatted string summarizing classification metrics.z %22s%11s%11s)classestop1_acctop5_accrs rget_desc ClassificationValidator.get_descOs#'JJJr!cURUl[UR5Ul/Ul/Ul[ URS9Ulg)z^Initialize confusion matrix, class names, and tracking containers for predictions and targets.)namesN)r+lenncrrr confusion_matrix)rmodels r init_metrics$ClassificationValidator.init_metricsSs<[[ ekk"  /ekk Br!cjUSRURURRS:HS9US'URR(aUSR 5OUSR 5US'USRURURRS:HS9US'U$)zTPreprocess input batch by moving data to device and converting to appropriate dtype.imgcuda) non_blockingcls)todevicetyperhalffloat)rbatchs r preprocess"ClassificationValidator.preprocess[sU|t{{AQAQU[A[\e .2iinnuU|((*%,BTBTBVe U|t{{AQAQU[A[\e  r!c[[UR5S5nURR UR SSS9SS2SU24R [R5R55 URR USR [R5R55 g)a Update running metrics with model predictions and batch targets. Args: preds (torch.Tensor): Model predictions, typically logits or probabilities for each class. batch (dict): Batch data containing images and class labels. Notes: This method appends the top-N predictions (sorted by confidence in descending order) to the prediction list for later evaluation. N is limited to the minimum of 5 and the number of classes. T) descendingNr6) minr,r+rappendargsortr9torchint32cpur)rpredsr<n5s rupdate_metrics&ClassificationValidator.update_metricsbsTZZ! $ qT:1crc6BGG TXXZ[ E%L--ekk:>>@Ar!cURRURUR5 URR (a7SH1nURR URXRS9 M3 URURl URURlURURlg)aM Finalize metrics including confusion matrix and processing speed. Notes: This method processes the accumulated predictions and targets to generate the confusion matrix, optionally plots it, and updates the metrics object with speed information. Examples: >>> validator = ClassificationValidator() >>> validator.pred = [torch.tensor([[0, 1, 2]])] # Top-3 predictions for one sample >>> validator.targets = [torch.tensor([0])] # Ground truth class >>> validator.finalize_metrics() >>> print(validator.metrics.confusion_matrix) # Access the confusion matrix )TF)r normalizeon_plotN) r.process_cls_predsrrrplotsplotrrOspeedr)rrNs rfinalize_metrics(ClassificationValidator.finalize_metricsrs // 4<<H 99??( %%**DMMY`l`l*m)!ZZ  $  (,(=(= %r!cF[U[[45(aUS$U$)zSExtract the primary prediction from model output if it's in a list or tuple format.r) isinstancelisttuple)rrIs r postprocess#ClassificationValidator.postprocesss!%edE];;uQxFFr!cURRURUR5 URR$)zSCalculate and return a dictionary of metrics by processing targets and predictions.)rprocessrr results_dictr's r get_stats!ClassificationValidator.get_statss. T\\4995||(((r!c4[S:XaS/[R"5-nS/[R"5-n[R"URUSS9 [R"UR USS9 UVVs/sH o3HoDPM M snnUlUVVs/sH o3HoUPM M snnUlg[S:aA[R"URSSS9 [R"UR SSS9 ggs snnfs snnf)zGather stats from all GPUs.rN)dst)r distget_world_size gather_objectrr)rgathered_predsgathered_targetsrankrrs r gather_stats$ClassificationValidator.gather_statss 19"Vd&9&9&;;N $v(;(;(==    tyy.a @   t||-=1 E*8J.$TTT.JDI0@U0@PTWGPTG0@UDL AX   tyy$A 6   t||Tq 9KUs D$DcT[XRSURRS9$)z7Create a ClassificationDataset instance for validation.F)rootraugmentprefix)rrsplit)rimg_paths r build_dataset%ClassificationValidator.build_datasets"$(EZ^ZcZcZiZijjr!cbURU5n[X2URRSS9$)a! Build and return a data loader for classification validation. Args: dataset_path (str | Path): Path to the dataset directory. batch_size (int): Number of samples per batch. Returns: (torch.utils.data.DataLoader): DataLoader object for the classification validation dataset. )rh)rqrrworkers)r dataset_path batch_sizedatasets rget_dataloader&ClassificationValidator.get_dataloaders-$$\2TYY5F5FRPPr!cSS[URR5--n[R"USURR URR 4-5 g)z6Print evaluation metrics for the classification model.z%22sz%11.3gallN)r,rkeysr infotop1top5)rpfs r print_results%ClassificationValidator.print_resultssL hT\\%6%6!77 7 B%!2!2DLL4E4EFFGr!c[R"USRS5US'[UURSUS3- UR UR S9 g)a Plot validation image samples with their ground truth labels. Args: batch (dict[str, Any]): Dictionary containing batch data with 'img' (images) and 'cls' (class labels). ni (int): Batch index used for naming the output file. Examples: >>> validator = ClassificationValidator() >>> batch = {"img": torch.rand(16, 3, 224, 224), "cls": torch.randint(0, 10, (16,))} >>> validator.plot_val_samples(batch, 0) r3r batch_idx val_batchz _labels.jpg)labelsfnamer+rON)rFarangeshaper rr+rO)rr<nis rplot_val_samples(ClassificationValidator.plot_val_samplessT#\\%,*<*>> validator = ClassificationValidator() >>> batch = {"img": torch.rand(16, 3, 224, 224)} >>> preds = torch.rand(16, 10) # 16 images, 10 classes >>> validator.plot_predictions(batch, preds, 0) r3rrA)dim)r3rr6confrz _pred.jpg)rr+rON) dictrFrrargmaxamaxr rr+rO)rr<rIr batched_predss rplot_predictions(ClassificationValidator.plot_predictionss|e ll5<#5#5a#89 U*Eq)   --IbT";;**LL  r!)r.rr+r-rr)NNNN)returnNone)rstr)r/ztorch.nn.Modulerr)r<dict[str, Any]rr)rI torch.Tensorr<rrr)rIz7torch.Tensor | list[torch.Tensor] | tuple[torch.Tensor]rr)rzdict[str, float])rprrr)rvz Path | strrwintrztorch.utils.data.DataLoader)r<rrrrr)r<rrIrrrrr)__name__ __module__ __qualname____firstlineno____doc__rr(r0r=rKrTrZr_rirqryrrr__static_attributes__ __classcell__)rs@rrrs`$L)),KCB >.G) :k QH  *  r!r) __future__rpathlibrtypingrrFtorch.distributed distributedrcultralytics.datarrultralytics.engine.validatorrultralytics.utilsr r ultralytics.utils.metricsr r ultralytics.utils.plottingr rr&r!rrs3# D6*F2S mS r!