Merge pull request #8 from junweima/fix

fix 0 percent training accuracy

Author: ynd

train.py

@@ -158,8 +158,8 @@ def train(epoch):
         train_loss += loss.data[0]
         _, predicted = torch.max(outputs.data, 1)
         total += targets.size(0)
-        correct += (lam * predicted.eq(targets_a.data).cpu().sum()
-                    + (1 - lam) * predicted.eq(targets_b.data).cpu().sum())
+        correct += (lam * predicted.eq(targets_a.data).cpu().sum().float()
+                    + (1 - lam) * predicted.eq(targets_b.data).cpu().sum().float())
 
         optimizer.zero_grad()
         loss.backward()