Merge pull request #202 from lcskrishna/cl/enable-fp16-nn

enabled fp16 tests that work with compiler fixes in test_nn

Author: iotamudelta

test/test_nn.py

@@ -6556,7 +6556,6 @@ def add(test_name, fn):
         add(cuda_test_name + '_double', lambda self,
             test=test, kwargs=kwargs: test.test_cuda(self, dtype=torch.double, **kwargs))
 
-        @skipIfRocm
         def test_half(self, test=test, kwargs=kwargs):
             test.test_cuda(self, dtype=torch.half, **kwargs)
         if getattr(test, 'check_half', True):