standalone val (#56)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

ultralytics/yolo/v8/classify/__init__.py

@@ -1,4 +1,4 @@
 from ultralytics.yolo.v8.classify.train import ClassificationTrainer, train
-from ultralytics.yolo.v8.classify.val import ClassificationValidator
+from ultralytics.yolo.v8.classify.val import ClassificationValidator, val
 
 __all__ = ["train"]

ultralytics/yolo/v8/classify/train.py

@@ -19,6 +19,13 @@ class ClassificationTrainer(BaseTrainer):
         else:
             model = ClassificationModel(model_cfg, weights, data["nc"])
         ClassificationModel.reshape_outputs(model, data["nc"])
+        for m in model.modules():
+            if not weights and hasattr(m, 'reset_parameters'):
+                m.reset_parameters()
+            if isinstance(m, torch.nn.Dropout) and self.args.dropout is not None:
+                m.p = self.args.dropout  # set dropout
+        for p in model.parameters():
+            p.requires_grad = True  # for training
         return model
 
     def get_dataloader(self, dataset_path, batch_size, rank=0, mode="train"):

ultralytics/yolo/v8/classify/val.py

@@ -1,5 +1,8 @@
+import hydra
 import torch
 
+from ultralytics.yolo.data import build_classification_dataloader
+from ultralytics.yolo.engine.trainer import DEFAULT_CONFIG
 from ultralytics.yolo.engine.validator import BaseValidator
 
 
@@ -24,6 +27,21 @@ class ClassificationValidator(BaseValidator):
         top1, top5 = acc.mean(0).tolist()
         return {"top1": top1, "top5": top5, "fitness": top5}
 
+    def get_dataloader(self, dataset_path, batch_size):
+        return build_classification_dataloader(path=dataset_path, imgsz=self.args.img_size, batch_size=batch_size)
+
     @property
     def metric_keys(self):
         return ["top1", "top5"]
+
+
+@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
+def val(cfg):
+    cfg.data = cfg.data or "imagenette160"
+    cfg.model = cfg.model or "resnet18"
+    validator = ClassificationValidator(args=cfg)
+    validator(model=cfg.model)
+
+
+if __name__ == "__main__":
+    val()