Exception: Tensor-likes are not close! in test/functorch/test_ops.py #2238
Description
🐛 Describe the bug
Cases:
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vjp_nn_functional_conv3d_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vjp_nn_functional_scaled_dot_product_attention_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vmap_autograd_grad_nn_functional_group_norm_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vmap_autograd_grad_nn_functional_group_norm_xpu_float64
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_grad_nn_functional_conv3d_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_jvpvjp_nn_functional_conv3d_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vjpvjp_index_reduce_prod_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vmapvjp_nn_functional_embedding_functorch_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vmapvjpvjp_nn_functional_conv2d_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_grad_nn_functional_scaled_dot_product_attention_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_grad_to_sparse_xpu_float32
op_ut,third_party.torch-xpu-ops.test.xpu.functorch.test_ops_functorch_xpu.TestOperatorsXPU,test_vmapvjp_nn_functional_ctc_loss_xpu_float32
__________ TestOperatorsXPU.test_vjp_nn_functional_conv3d_xpu_float32 __________
[gw2] linux -- Python 3.10.19 /tmp/xpu-tool/Python/3.10.19/x64/bin/python
Traceback (most recent call last):
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 1151, in test_wrapper
return test(*args, **kwargs)
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_cuda.py", line 291, in wrapped
return f(*args, **kwargs)
File "/__w/torch-xpu-ops/torch-xpu-ops/pytorch/third_party/torch-xpu-ops/test/xpu/functorch/test_ops_functorch_xpu.py", line 850, in test_vjp
_test(op)
File "/__w/torch-xpu-ops/torch-xpu-ops/pytorch/third_party/torch-xpu-ops/test/xpu/functorch/test_ops_functorch_xpu.py", line 836, in _test
self.assertEqual(result_vjps, expected_vjps)
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 4255, in assertEqual
raise error_metas.pop()[0].to_error( # type: ignore[index]
AssertionError: Tensor-likes are not close!
Mismatched elements: 1 / 10 (10.0%)
Greatest absolute difference: 2.09808349609375e-05 at index (0, 0, 0, 1, 4) (up to 1e-05 allowed)
Greatest relative difference: 2.726984348555561e-06 at index (0, 0, 0, 1, 4) (up to 1.3e-06 allowed)
The failure occurred for item [1]
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/unittest/case.py", line 59, in testPartExecutor
yield
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/unittest/case.py", line 591, in run
self._callTestMethod(testMethod)
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/unittest/case.py", line 549, in _callTestMethod
method()
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 3300, in wrapper
method(*args, **kwargs)
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 428, in instantiated_test
result = test(self, **param_kwargs)
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 1707, in wrapper
fn(*args, **kwargs)
File "/tmp/xpu-tool/Python/3.10.19/x64/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 1163, in test_wrapper
raise e_tracked from e
Exception: Tensor-likes are not close!
Mismatched elements: 1 / 10 (10.0%)
Greatest absolute difference: 2.09808349609375e-05 at index (0, 0, 0, 1, 4) (up to 1e-05 allowed)
Greatest relative difference: 2.726984348555561e-06 at index (0, 0, 0, 1, 4) (up to 1.3e-06 allowed)
The failure occurred for item [1]
Caused by sample input at index 8: SampleInput(input=Tensor[size=(1, 1, 10, 11, 12), device="xpu:0", dtype=torch.float32], args=(Tensor[size=(1, 1, 1, 2, 5), device="xpu:0", dtype=torch.float32],None), kwargs={'padding': "'same'"}, broadcasts_input=False, name='')
To execute this test, run the following from the base repo dir:
PYTORCH_OPINFO_SAMPLE_INPUT_INDEX=8 PYTORCH_TEST_WITH_SLOW=1 python test/xpu/functorch/test_ops_functorch_xpu.py TestOperatorsXPU.test_vjp_nn_functional_conv3d_xpu_float32
This message can be suppressed by setting PYTORCH_PRINT_REPRO_ON_FAILURE=0