def test_batch_norm(conf): x = torch.randn(100, 16) norm = BatchNorm(16, affine=conf, track_running_stats=conf) assert norm.__repr__() == 'BatchNorm(16)' torch.jit.script(norm) out = norm(x) assert out.size() == (100, 16)
def test_batch_norm(): norm = BatchNorm(16) assert norm.__repr__() == ( 'BatchNorm(16, eps=1e-05, momentum=0.1, affine=True, ' 'track_running_stats=True)') out = norm(torch.randn(100, 16)) assert out.size() == (100, 16) norm = BatchNorm(16, affine=False, track_running_stats=False) out = norm(torch.randn(100, 16)) assert out.size() == (100, 16)