Fix seed typing as int

Make sure the seed input to be interpreted as `int` type. Thanks wang6213@purdue.edu for pointing out.

configs.py

@@ -202,6 +202,7 @@ def dataloader(d, bs=256, shuffle=True, workers=-1, drop_last=True):
 
 def seeding(seed):
     if seed != -1:
+        seed = int(seed)           # make sure seed in `int` type
         torch.manual_seed(seed)
         np.random.seed(seed)
         random.seed(seed)