🧪 fix(tests): CLI unit test (#3161)

* Use right parent class for registering the usage

Signed-off-by: Ashwin Vaidya <[email protected]>

* Add weights_only kwarg

Signed-off-by: Ashwin Vaidya <[email protected]>

* Add missing weights_only kwarg

Signed-off-by: Ashwin Vaidya <[email protected]>

---------

Signed-off-by: Ashwin Vaidya <[email protected]>

Author: ashwinvaidya17

src/anomalib/cli/cli.py

@@ -99,9 +99,9 @@ def init_parser(**kwargs) -> ArgumentParser:
     def subcommands() -> dict[str, set[str]]:
         """Skip predict subcommand as it is added later."""
         return {
-            "fit": {"model", "train_dataloaders", "val_dataloaders", "datamodule"},
-            "validate": {"model", "dataloaders", "datamodule"},
-            "test": {"model", "dataloaders", "datamodule"},
+            "fit": {"model", "train_dataloaders", "val_dataloaders", "datamodule", "weights_only"},
+            "validate": {"model", "dataloaders", "datamodule", "weights_only"},
+            "test": {"model", "dataloaders", "datamodule", "weights_only"},
         }
 
     @staticmethod

src/anomalib/cli/utils/help_formatter.py

@@ -327,7 +327,7 @@ def add_usage(self, usage: str | None, actions: list, *args, **kwargs) -> None:
             elif self.verbosity_level == 1:
                 actions = [action for action in actions if action.dest in REQUIRED_ARGUMENTS[self.subcommand]]
 
-        super().add_usage(usage, actions, *args, **kwargs)
+        super(RichHelpFormatter, self).add_usage(usage, actions, *args, **kwargs)
 
     def add_argument(self, action: argparse.Action) -> None:
         """Add an argument to the help formatter.

src/anomalib/engine/engine.py

@@ -407,9 +407,15 @@ def fit(
         self._setup_trainer(model)
         if model.learning_type in {LearningType.ZERO_SHOT, LearningType.FEW_SHOT}:
             # if the model is zero-shot or few-shot, we only need to run validate for normalization and thresholding
-            self.trainer.validate(model, val_dataloaders, datamodule=datamodule, ckpt_path=ckpt_path)
+            self.trainer.validate(
+                model,
+                val_dataloaders,
+                datamodule=datamodule,
+                ckpt_path=ckpt_path,
+                weights_only=False,
+            )
         else:
-            self.trainer.fit(model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)
+            self.trainer.fit(model, train_dataloaders, val_dataloaders, datamodule, ckpt_path, weights_only=False)
 
     def validate(
         self,
@@ -457,7 +463,7 @@ def validate(
             ckpt_path = Path(ckpt_path).resolve()
         if model:
             self._setup_trainer(model)
-        return self.trainer.validate(model, dataloaders, ckpt_path, verbose, datamodule)
+        return self.trainer.validate(model, dataloaders, ckpt_path, verbose, datamodule, weights_only=False)
 
     def test(
         self,
@@ -551,7 +557,7 @@ def test(
         if self._should_run_validation(model or self.model, ckpt_path):
             logger.info("Running validation before testing to collect normalization metrics and/or thresholds.")
             self.trainer.validate(model, dataloaders, None, verbose=False, datamodule=datamodule)
-        return self.trainer.test(model, dataloaders, ckpt_path, verbose, datamodule)
+        return self.trainer.test(model, dataloaders, ckpt_path, verbose, datamodule, weights_only=False)
 
     def predict(
         self,
@@ -658,9 +664,10 @@ def predict(
                 ckpt_path=None,
                 verbose=False,
                 datamodule=datamodule,
+                weights_only=False,
             )
 
-        return self.trainer.predict(model, dataloaders, datamodule, return_predictions, ckpt_path)
+        return self.trainer.predict(model, dataloaders, datamodule, return_predictions, ckpt_path, weights_only=False)
 
     def train(
         self,
@@ -716,8 +723,14 @@ def train(
             # if the model is zero-shot or few-shot, we only need to run validate for normalization and thresholding
             self.trainer.validate(model, val_dataloaders, None, verbose=False, datamodule=datamodule)
         else:
-            self.trainer.fit(model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)
-        return self.trainer.test(model, test_dataloaders, ckpt_path=ckpt_path, datamodule=datamodule)
+            self.trainer.fit(model, train_dataloaders, val_dataloaders, datamodule, ckpt_path, weights_only=False)
+        return self.trainer.test(
+            model,
+            test_dataloaders,
+            ckpt_path=ckpt_path,
+            datamodule=datamodule,
+            weights_only=False,
+        )
 
     def export(
         self,
@@ -816,7 +829,7 @@ def export(
         self._setup_trainer(model)
         if ckpt_path:
             ckpt_path = Path(ckpt_path).resolve()
-            model = model.__class__.load_from_checkpoint(ckpt_path)
+            model = model.__class__.load_from_checkpoint(ckpt_path, weights_only=False)
 
         if export_root is None:
             export_root = Path(self.trainer.default_root_dir)