Rename flag

tests/conftest.py

@@ -9,11 +9,11 @@
 ModelAndData = Tuple[fbn.Model, ep.Tensor, ep.Tensor]
 CallableModelAndDescription = NamedTuple(
     "CallableModelAndDescription",
-    [("model_fn", Callable[..., ModelAndData]), ("real", bool), ("small", bool)],
+    [("model_fn", Callable[..., ModelAndData]), ("real", bool), ("small_input", bool)],
 )
 ModelDescriptionAndData = NamedTuple(
     "ModelDescriptionAndData",
-    [("model_and_data", ModelAndData), ("real", bool), ("small", bool)],
+    [("model_and_data", ModelAndData), ("real", bool), ("small_input", bool)],
 )
 
 models: Dict[str, CallableModelAndDescription] = {}
@@ -35,7 +35,7 @@ def dummy(request: Any) -> ep.Tensor:
 
 
 def register(
-    backend: str, *, real: bool = False, small: bool = False, attack: bool = True
+    backend: str, *, real: bool = False, small_input: bool = False, attack: bool = True
 ) -> Callable[[Callable], Callable]:
     def decorator(f: Callable[[Any], ModelAndData]) -> Callable[[Any], ModelAndData]:
         @functools.wraps(f)
@@ -48,7 +48,7 @@ def model(request: Any) -> ModelAndData:
         global real_models
 
         models[model.__name__] = CallableModelAndDescription(
-            model_fn=model, real=real, small=small
+            model_fn=model, real=real, small_input=small_input
         )
         if attack:
             models_for_attacks.append(model.__name__)
@@ -118,7 +118,7 @@ def pytorch_simple_model_object(request: Any) -> ModelAndData:
     return pytorch_simple_model(torch.device("cpu"))
 
 
-@register("pytorch", real=True, small=True)
+@register("pytorch", real=True, small_input=True)
 def pytorch_mnist(request: Any) -> ModelAndData:
     fmodel = fbn.zoo.ModelLoader.get().load(
         "examples/zoo/mnist/", module_name="foolbox_model"
@@ -165,12 +165,12 @@ def tensorflow_simple_sequential(
     return fmodel, x, y
 
 
-@register("tensorflow", small=True)
+@register("tensorflow", small_input=True)
 def tensorflow_simple_sequential_cpu(request: Any) -> ModelAndData:
     return tensorflow_simple_sequential("cpu", None)
 
 
-@register("tensorflow", small=True)
+@register("tensorflow", small_input=True)
 def tensorflow_simple_sequential_native_tensors(request: Any) -> ModelAndData:
     import tensorflow as tf
 
@@ -179,14 +179,14 @@ def tensorflow_simple_sequential_native_tensors(request: Any) -> ModelAndData:
     return tensorflow_simple_sequential("cpu", dict(mean=mean, std=std))
 
 
-@register("tensorflow", small=True)
+@register("tensorflow", small_input=True)
 def tensorflow_simple_sequential_eagerpy_tensors(request: Any) -> ModelAndData:
     mean = ep.tensorflow.zeros(1)
     std = ep.tensorflow.ones(1) * 255.0
     return tensorflow_simple_sequential("cpu", dict(mean=mean, std=std))
 
 
-@register("tensorflow", small=True)
+@register("tensorflow", small_input=True)
 def tensorflow_simple_subclassing(request: Any) -> ModelAndData:
     import tensorflow as tf
 
@@ -267,7 +267,7 @@ def tensorflow_resnet50(request: Any) -> ModelAndData:
     return fmodel, x, y
 
 
-@register("jax", small=True)
+@register("jax", small_input=True)
 def jax_simple_model(request: Any) -> ModelAndData:
     import jax
 

tests/test_attacks.py

@@ -22,7 +22,7 @@ class AttackTestTarget(NamedTuple):
     epsilon: Optional[float] = None
     uses_grad: Optional[bool] = False
     requires_real_model: Optional[bool] = False
-    requires_small_model: Optional[bool] = False
+    requires_small_input: Optional[bool] = False
 
 
 def get_attack_id(x: AttackTestTarget) -> str:
@@ -127,10 +127,10 @@ def test_untargeted_attacks(
     attack_test_target: AttackTestTarget,
 ) -> None:
 
-    (fmodel, x, y), real, small = fmodel_and_data_ext_for_attacks
+    (fmodel, x, y), real, small_input = fmodel_and_data_ext_for_attacks
     if attack_test_target.requires_real_model and not real:
         pytest.skip()
-    if attack_test_target.requires_small_model and not small:
+    if attack_test_target.requires_small_input and not small_input:
         pytest.skip()
     if isinstance(x, ep.NumPyTensor) and attack_test_target.uses_grad:
         pytest.skip()
@@ -186,10 +186,10 @@ def test_targeted_attacks(
     attack_test_target: AttackTestTarget,
 ) -> None:
 
-    (fmodel, x, y), real, small = fmodel_and_data_ext_for_attacks
+    (fmodel, x, y), real, small_input = fmodel_and_data_ext_for_attacks
     if attack_test_target.requires_real_model and not real:
         pytest.skip()
-    if attack_test_target.requires_small_model and not small:
+    if attack_test_target.requires_small_input and not small_input:
         pytest.skip()
 
     if isinstance(x, ep.NumPyTensor) and attack_test_target.uses_grad: