certified

Tune.py

@@ -41,6 +41,7 @@ EPOCHS = 10
 # Inception = 299
 RESOLUTION = 224
 
+FINETUNE = False # whether to fintune or just load finetuned model from dir
 # model architecture options are 
 #  - RESNET18
 #  - RESNET50
@@ -112,19 +113,24 @@ device = SetUp.get_device()
 for i in range(0,1):#CLASS_SIZE):
     FORGET_CLASS_IDX = i
     # Create model using Factory
-    model = Model.create(
-        arch = arch, 
-        device = device, 
-        size = CLASS_SIZE)
 
-    # we may need to load existing model or finetune
-    #model.train(
-    #        epochs = EPOCHS, 
-    #        loader = train_loader, 
-    #        rate = LR_RATE)
+    model = None
 
-        # save.
-    #model.save(filename=arch.name.lower())
+    if FINETUNE:
+        model = Model.create(
+            arch = arch, 
+            device = device, 
+            size = CLASS_SIZE)
+
+        # we may need to load existing model or finetune
+        model.train(
+                epochs = EPOCHS, 
+                loader = train_loader, 
+                rate = LR_RATE)
+
+            # save.
+        file_name = f"{arch.name.lower}_{dataset_name.name.lower()}"
+        model.save(filename=arch.name.lower())
 
 
     # done tuning
@@ -147,18 +153,21 @@ for i in range(0,1):#CLASS_SIZE):
 
     # Evaluate
     current_mode = "Finetuned"
-    #accuracy, report_dict = model.evaluate(
-    #    loader = test_loader,
-    #    mode=current_mode
-    #)
+    if FINETUNE:
 
-    Util._log_to_csv(
-        arch=model.__class__.__name__, 
-        mode = current_mode, 
-        accuracy=accuracy, 
-        report_dict=report_dict, 
-        strategy="base"
-    )
+        #current_mode = "Finetuned"
+        accuracy, report_dict = model.evaluate(
+            loader = test_loader,
+            mode=current_mode
+        )
+
+        Util._log_to_csv(
+            arch=model.__class__.__name__, 
+            mode = current_mode, 
+            accuracy=accuracy, 
+            report_dict=report_dict, 
+            strategy="base"
+        )
 
     # unlearning algorithms
     #linear_filtration = LinearFiltration(target_class_index=FORGET_CLASS_IDX)
@@ -167,7 +176,14 @@ for i in range(0,1):#CLASS_SIZE):
     #weight_filtration = WeightFiltration(num_classes = CLASS_SIZE,target_class_idx=FORGET_CLASS_IDX)
     #weight_filtration.apply(reloaded.model)
 
-    certified_removal = CertifiedRemoval(target_class_index=FORGET_CLASS_IDX,removal_bound=0.05, epsilon=0.5, l2_reg=15)
+    certified_removal = CertifiedRemoval(
+        target_class_index=FORGET_CLASS_IDX,
+        s1=2,
+        s2=500,
+        unlearn_bs=2,
+        scale=100.0,    # Drop scale to match lower s2 depth
+        std=0.00001)
+    #,removal_bound=0.05, epsilon=0.5, l2_reg=15)
     #certified_removal.apply(reloaded.model)
 
     # to be unlearned
@@ -200,6 +216,12 @@ for i in range(0,1):#CLASS_SIZE):
         #   loader = test_loader
         #)
 
+        if not FINETUNE:
+            reloaded.evaluate(
+            loader = test_loader,
+            mode=current_mode
+        )
+
         # Unlearning
         # train loaders passed here
         strategy.apply(reloaded.model, forget_train_loader, retain_train_loader)