__________________________ test_index_sort[device0] ___________________________

device = device(type='cpu')

    @pytest.mark.parametrize('device', DEVICES)
    def test_index_sort(device):
        inputs = torch.randperm(100_000, device=device)
        ref_sorted_input, ref_indices = torch.sort(inputs, stable=True)
        sorted_input, indices = pyg_lib.ops.index_sort(inputs)
>       assert torch.all(ref_sorted_input == sorted_input)
E       assert tensor(False)
E        +  where tensor(False) = <built-in method all of type object at 0x00007FFFB4FFEE80>(tensor([    0, 11966,  7382,  ..., 24002,  2670, 16435]) == tensor([    0,     1,     2,  ..., 99997, 99998, 99999])
E        +    where <built-in method all of type object at 0x00007FFFB4FFEE80> = torch.all
E           
E           Full diff:
E           - tensor([    0,     1,     2,  ..., 99997, 99998, 99999])
E           + tensor([    0, 11966,  7382,  ..., 24002,  2670, 16435]))

test\ops\test_index_sort.py:16: AssertionError
 ________________________ test_softmax_csr_autograd[0] _________________________

dim = 0

    @pytest.mark.parametrize('dim', list(range(3)))
    def test_softmax_csr_autograd(dim):
        sizes = (16, 32, 64)
        src1 = torch.rand(sizes, requires_grad=True)
        src2 = src1.detach().clone()
        src2.requires_grad = True
        dim_size = sizes[dim]
        ptr = torch.tensor([0, 1, 4, 5, dim_size - 1, dim_size])
        out_grad = torch.randn(sizes)
    
        expected_out = softmax_reference(src1, ptr, dim)
        out = pyg_lib.ops.softmax_csr(src2, ptr, dim)
>       assert torch.allclose(expected_out, out, atol=1e-6)
E       assert False
E        +  where False = <built-in method allclose of type object at 0x00007FFFB4FFEE80>(tensor([[[0.1250, 0.1250, 0.1250,  ..., 0.1250, 0.1250, 0.1250],\n         [0.1250, 0.1250, 0.1250,  ..., 0.1250, 0.1250, 0.1250],\n         [0.1250, 0.1250, 0.1250,  ..., 0.1250, 0.1250, 0.1250],\n         ...,\n         [0.1250, 0.1250, 0.1250,  ..., 0.1250, 0.1250, 0.1250],\n         [0.1250, 0.1250, 0.1250,  ..., 0.1250, 0.1250, 0.1250],\n         [0.1250, 0.1250, 0.1250,  ..., 0.1250, 0.1250, 0.1250]],\n\n        [[   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         ...,\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf]],\n\n        [[   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         ...,\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf],\n         [   inf,    inf,    inf,  ...,    inf,    inf,    inf]],\n\n        ...,\n\n        [[0.1112, 0.1154, 0.1628,  ..., 0.2388, 0.1397, 0.1385],\n         [0.2727, 0.1811, 0.1359,  ..., 0.1575, 0.1239, 0.1146],\n         [0.1527, 0.2101, 0.1511,  ..., 0.2917, 0.2265, 0.1673],\n         ...,\n         [0.1504, 0.0841, 0.2485,  ..., 0.1421, 0.1068, 0.1552],\n         [0.1738, 0.2317, 0.2033,  ..., 0.1912, 0.1105, 0.1442],\n         [0.1625, 0.2572, 0.1764,  ..., 0.1542, 0.1133, 0.1440]],\n\n        [[0.2061, 0.2132, 0.2313,  ..., 0.1056, 0.2529, 0.1111],\n         [0.2058, 0.1829, 0.2990,  ..., 0.1247, 0.1509, 0.2046],\n         [0.2379, 0.2679, 0.1090,  ..., 0.2855, 0.1633, 0.1737],\n         ...,\n         [0.1331, 0.1018, 0.1329,  ..., 0.1130, 0.1155, 0.1681],\n         [0.1935, 0.1238, 0.2346,  ..., 0.1846, 0.1407, 0.1730],\n         [0.2061, 0.1988, 0.2449,  ..., 0.1516, 0.1032, 0.1285]],\n\n        [[1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         ...,\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000]]],\n       grad_fn=<DivBackward0>), tensor([[[1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         ...,\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000]],\n\n        [[0.2417, 0.2579, 0.4961,  ..., 0.3641, 0.2242, 0.4179],\n         [0.3829, 0.3277, 0.3676,  ..., 0.3801, 0.2794, 0.2856],\n         [0.3129, 0.2644, 0.5144,  ..., 0.3335, 0.2922, 0.4699],\n         ...,\n         [0.3454, 0.3659, 0.3029,  ..., 0.4260, 0.2055, 0.5255],\n         [0.2670, 0.2713, 0.3273,  ..., 0.4415, 0.3480, 0.4439],\n         [0.3507, 0.2773, 0.3717,  ..., 0.4621, 0.3282, 0.4235]],\n\n        [[0.4173, 0.3118, 0.2460,  ..., 0.3144, 0.3545, 0.2162],\n         [0.4176, 0.2632, 0.2321,  ..., 0.2850, 0.3394, 0.4088],\n         [0.2204, 0.2341, 0.2112,  ..., 0.4103, 0.4245, 0.3126],\n         ...,\n         [0.4126, 0.1978, 0.2525,  ..., 0.3116, 0.4507, 0.2546],\n         [0.3021, 0.3805, 0.3337,  ..., 0.2540, 0.2477, 0.2672],\n         [0.3875, 0.2745, 0.2679,  ..., 0.2447, 0.3154, 0.2707]],\n\n        ...,\n\n        [[0.0621, 0.0659, 0.0977,  ..., 0.1387, 0.0852, 0.0911],\n         [0.1352, 0.1012, 0.0754,  ..., 0.0990, 0.0800, 0.0710],\n         [0.0897, 0.1091, 0.0916,  ..., 0.1439, 0.1281, 0.1022],\n         ...,\n         [0.0973, 0.0588, 0.1368,  ..., 0.0836, 0.0691, 0.1009],\n         [0.1036, 0.1466, 0.1165,  ..., 0.1123, 0.0659, 0.0844],\n         [0.0937, 0.1372, 0.0980,  ..., 0.0893, 0.0755, 0.0829]],\n\n        [[0.1151, 0.1217, 0.1388,  ..., 0.0613, 0.1542, 0.0731],\n         [0.1021, 0.1022, 0.1658,  ..., 0.0784, 0.0975, 0.1266],\n         [0.1398, 0.1391, 0.0661,  ..., 0.1408, 0.0923, 0.1061],\n         ...,\n         [0.0861, 0.0712, 0.0732,  ..., 0.0665, 0.0748, 0.1093],\n         [0.1153, 0.0783, 0.1344,  ..., 0.1085, 0.0838, 0.1013],\n         [0.1187, 0.1060, 0.1361,  ..., 0.0879, 0.0688, 0.0740]],\n\n        [[1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         ...,\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000],\n         [1.0000, 1.0000, 1.0000,  ..., 1.0000, 1.0000, 1.0000]]],\n       grad_fn=<SoftmaxCSR>>), atol=1e-06)
E        +    where <built-in method allclose of type object at 0x00007FFFB4FFEE80> = torch.allclose

test\ops\test_softmax.py:47: AssertionError