Bugfix in vision transformer example - change lr datatype to float (pytorch#1161)

change lr datatype to float

Author: siddharth9820

vision_transformer/main.py

@@ -229,7 +229,7 @@ def main():
  help='number of classes in dataset (default : 10 for CIFAR10)')
  parser.add_argument('--epochs', type=int, default=10,
  help='number of epochs (default : 10)')
- parser.add_argument('--lr', type=int, default=1e-2,
+ parser.add_argument('--lr', type=float, default=1e-2,
  help='base learning rate (default : 0.01)')
  parser.add_argument('--weight-decay', type=int, default=3e-2,
  help='weight decay value (default : 0.03)')