Fix bug in LSUN dataset for test split. (#435)

Both conditional branches in `__init__` were executed, since `classes` for test
becomes a list and `if` was used instead of `elif`.

Author: iacolippo

torchvision/datasets/lsun.py

@@ -85,7 +85,7 @@ def __init__(self, root, classes='train',
                 classes = [classes]
             else:
                 classes = [c + '_' + classes for c in categories]
-        if type(classes) == list:
+        elif type(classes) == list:
             for c in classes:
                 c_short = c.split('_')
                 c_short.pop(len(c_short) - 1)