remove categories, fix logging

adversarial_attack/fgsm.py

@@ -37,7 +37,6 @@ def standard_attack(
     model: torch.nn.Module,
     tensor: torch.Tensor,
     truth: torch.Tensor,
-    categories: list[str],
     epsilon: float = 1e-3,
     max_iter: int = 50,
 ) -> ty.Optional[tuple[torch.Tensor, int, int]]:
@@ -48,7 +47,6 @@ def standard_attack(
         model (torch.Model): PyTorch model to attack.
         tensor (torch.Tensor): Tensor to attack.
         truth (torch.Tensor): Tensor representing true category.
-        categories (list[str]): List of categories for the model.
         epsilon (float): Maximum perturbation allowed.
         max_iter (int): Maximum number of iterations to perform.
 
@@ -80,7 +78,9 @@ def standard_attack(
 
     for i in range(max_iter):
         model.zero_grad()
-        grad = compute_gradient(model=model, input=adv_tensor, target=torch.tensor([orig_pred_idx]))
+        grad = compute_gradient(
+            model=model, input=adv_tensor, target=torch.tensor([orig_pred_idx])
+        )
         adv_tensor = torch.clamp(adv_tensor + epsilon * grad.sign(), -2, 2)
         new_pred_idx = model(adv_tensor).argmax()
         if orig_pred_idx != new_pred_idx:
@@ -97,7 +97,6 @@ def targeted_attack(
     tensor: torch.Tensor,
     truth: torch.Tensor,
     target: torch.Tensor,
-    categories: list[str],
     epsilon: float = 1e-3,
     max_iter: int = 50,
 ) -> ty.Optional[tuple[torch.Tensor, int, int]]:
@@ -109,7 +108,6 @@ def targeted_attack(
         tensor (torch.Tensor): Tensor to attack.
         truth (torch.Tensor): Tensor representing true category.
         target (torch.Tensor): Tensor representing targeted category.
-        categories (list[str]): List of categories for the model.
         epsilon (float): Maximum perturbation allowed.
         max_iter (int): Maximum number of iterations to perform.
 
@@ -128,9 +126,11 @@ def targeted_attack(
     truth_idx: int = truth.item()
 
     if orig_pred_idx != truth_idx:
-        raise ValueError(
-            f"Model prediction {orig_pred_idx} does not match true class {truth_idx}.",
-            f"It is therefore pointless to perform an attack.",
+        logger.warning(
+            (
+                f"Model prediction {orig_pred_idx} does not match true class {truth_idx}."
+                f"It is therefore pointless to perform an attack.",
+            )
         )
         return None
 
@@ -170,7 +170,6 @@ def execute():
                 model,
                 tensor=tensor,
                 truth=truth,
-                categories=categories,
                 epsilon=epsilon,
                 max_iter=max_iter,
             )
@@ -185,7 +184,6 @@ def execute():
                 tensor=tensor,
                 truth=truth,
                 target=target,
-                categories=categories,
                 epsilon=epsilon,
                 max_iter=max_iter,
             )

tests/unit/test_fgsm.py

@@ -42,15 +42,13 @@ def test_compute_gradient(model, inputs_and_targets):
 def test_standard_attack_success(model):
     tensor = torch.tensor([[1.0, 0.0, 0.5]])
     truth = torch.tensor([0])
-    categories = ["cat1", "cat2", "cat3"]
     epsilon = 0.1
     max_iter = 50
 
     result = standard_attack(
         model=model,
         tensor=tensor,
         truth=truth,
-        categories=categories,
         epsilon=epsilon,
         max_iter=max_iter,
     )
@@ -64,7 +62,6 @@ def test_targeted_attack_success(model):
     tensor = torch.tensor([[1.0, 0.0, 0.5]])
     truth = torch.tensor([0])
     target = torch.tensor([2])
-    categories = ["cat1", "cat2", "cat3"]
     epsilon = 0.1
     max_iter = 50
 
@@ -73,7 +70,6 @@ def test_targeted_attack_success(model):
         tensor=tensor,
         truth=truth,
         target=target,
-        categories=categories,
         epsilon=epsilon,
         max_iter=max_iter,
     )
@@ -86,22 +82,20 @@ def test_targeted_attack_success(model):
 def test_standard_attack_failure(model):
     tensor = torch.tensor([[1.0, 0.0, 0.5]])
     truth = torch.tensor([1])  # Intentionally mismatched target
-    categories = ["cat1", "cat2", "cat3"]
     epsilon = 0.1
     max_iter = 50
 
-    assert standard_attack(model=model, tensor=tensor, truth=truth, categories=categories, epsilon=epsilon, max_iter=max_iter) is None
+    assert standard_attack(model=model, tensor=tensor, truth=truth, epsilon=epsilon, max_iter=max_iter) is None
 
 
 def test_targeted_attack_failure(model):
     tensor = torch.tensor([[1.0, 0.0, 0.5]])
     truth = torch.tensor([1])  # Intentionally mismatched target
     target = torch.tensor([2])
-    categories = ["cat1", "cat2", "cat3"]
     epsilon = 0.1
     max_iter = 50
 
-    assert targeted_attack(model=model, tensor=tensor, truth=truth, target=target, categories=categories, epsilon=epsilon, max_iter=max_iter) is None
+    assert targeted_attack(model=model, tensor=tensor, truth=truth, target=target, epsilon=epsilon, max_iter=max_iter) is None
 
 def test_standard_attack_no_change(model):
     tensor = torch.tensor([[1.0, 0.0, 0.5]])
@@ -114,7 +108,6 @@ def test_standard_attack_no_change(model):
         model=model,
         tensor=tensor,
         truth=truth,
-        categories=categories,
         epsilon=epsilon,
         max_iter=max_iter,
     )
@@ -138,7 +131,6 @@ def test_targeted_attack_no_change(model):
         tensor=tensor,
         truth=truth,
         target=target,
-        categories=categories,
         epsilon=epsilon,
         max_iter=max_iter,
     )