offer hard distillation as well

README.md

@@ -93,7 +93,8 @@ distiller = DistillWrapper(
     student = v,
     teacher = teacher,
     temperature = 3,           # temperature of distillation
-    alpha = 0.5                # trade between main loss and distillation loss
+    alpha = 0.5,               # trade between main loss and distillation loss
+    hard = False               # whether to use soft or hard distillation
 )
 
 img = torch.randn(2, 3, 256, 256)

setup.py

@@ -3,7 +3,7 @@
 setup(
   name = 'vit-pytorch',
   packages = find_packages(exclude=['examples']),
-  version = '0.14.2',
+  version = '0.14.4',
   license='MIT',
   description = 'Vision Transformer (ViT) - Pytorch',
   author = 'Phil Wang',

vit_pytorch/distill.py

@@ -104,7 +104,8 @@ def __init__(
         teacher,
         student,
         temperature = 1.,
-        alpha = 0.5
+        alpha = 0.5,
+        hard = False
     ):
         super().__init__()
         assert (isinstance(student, (DistillableViT, DistillableT2TViT, DistillableEfficientViT))) , 'student must be a vision transformer'
@@ -116,6 +117,7 @@ def __init__(
         num_classes = student.num_classes
         self.temperature = temperature
         self.alpha = alpha
+        self.hard = hard
 
         self.distillation_token = nn.Parameter(torch.randn(1, 1, dim))
 
@@ -137,11 +139,15 @@ def forward(self, img, labels, temperature = None, alpha = None, **kwargs):
 
         loss = F.cross_entropy(student_logits, labels)
 
-        distill_loss = F.kl_div(
-            F.log_softmax(distill_logits / T, dim = -1),
-            F.softmax(teacher_logits / T, dim = -1).detach(),
-        reduction = 'batchmean')
+        if not self.hard:
+            distill_loss = F.kl_div(
+                F.log_softmax(distill_logits / T, dim = -1),
+                F.softmax(teacher_logits / T, dim = -1).detach(),
+            reduction = 'batchmean')
+            distill_loss *= T ** 2
 
-        distill_loss *= T ** 2
+        else:
+            teacher_labels = teacher_logits.argmax(dim = -1)
+            distill_loss = F.cross_entropy(student_logits, teacher_labels)
 
         return loss * alpha + distill_loss * (1 - alpha)