[upstream_ut] float64 got "Exception: Jacobian computed with forward mode mismatch for output 0 with respect to input 0" #2360
Description
🐛 Describe the bug
Cases:
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD___rmatmul___xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD_addmm_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD_addmv_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD_baddbmm_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD_matmul_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD_mv_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_inplace_forward_mode_AD_addmv_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_forward_mode_AD_addbmm_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_inplace_forward_mode_AD_addbmm_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_inplace_forward_mode_AD_addmm_decomposed_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_inplace_forward_mode_AD_addmm_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_inplace_forward_mode_AD_baddbmm_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_fn_fwgrad_bwgrad_matmul_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_fn_fwgrad_bwgrad_mv_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_fn_fwgrad_bwgrad_addmv_xpu_float64
op_ut,test.test_ops_fwd_gradients.TestFwdGradientsXPU,test_fn_fwgrad_bwgrad_addr_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad___rmatmul___xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_addmm_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_addmv_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_baddbmm_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_cdist_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_matmul_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_mv_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_nn_functional_multi_head_attention_forward_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_grad_addmv_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_grad_addbmm_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_gradgrad___rmatmul___xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_grad_addbmm_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_grad_addmm_decomposed_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_grad_addmm_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_grad_baddbmm_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_gradgrad_matmul_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_gradgrad_mv_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_gradgrad_addmv_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_fn_gradgrad_addr_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_gradgrad_addmv_xpu_float64
op_ut,test.test_ops_gradients.TestBwdGradientsXPU,test_inplace_gradgrad_addr_xpu_float64
pytest_command:
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD___rmatmul___xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_addmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_baddbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_matmul_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_mv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_inplace_forward_mode_AD_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_addbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_inplace_forward_mode_AD_addbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_inplace_forward_mode_AD_addmm_decomposed_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_inplace_forward_mode_AD_addmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_inplace_forward_mode_AD_baddbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_matmul_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_mv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_addr_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad___rmatmul___xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_addmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_baddbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_cdist_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_matmul_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_mv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_nn_functional_multi_head_attention_forward_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_grad_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_grad_addbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_gradgrad___rmatmul___xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_grad_addbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_grad_addmm_decomposed_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_grad_addmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_grad_baddbmm_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_gradgrad_matmul_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_gradgrad_mv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_gradgrad_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_fn_gradgrad_addr_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_gradgrad_addmv_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_gradgrad_addr_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_nn_functional_conv_transpose2d_xpu_complex128
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_nn_functional_conv_transpose2d_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_nn_functional_conv_transpose3d_xpu_complex128
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_fn_fwgrad_bwgrad_nn_functional_conv_transpose3d_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_nn_functional_conv_transpose2d_xpu_complex128
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_nn_functional_conv_transpose2d_xpu_float64
PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_forward_mode_AD_nn_functional_conv_transpose3d_xpu_complex128
Trace Example:
Command: PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_fwd_gradients.py -k test_inplace_forward_mode_AD_addmv_xpu_float64
Traceback (most recent call last):
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 1151, in test_wrapper
return test(*args, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 1762, in wrapper
fn(*args, **kwargs)
File "/home/daisyden/upstream/gradient_test/test/test_ops_fwd_gradients.py", line 100, in test_inplace_forward_mode_AD
self._forward_grad_helper(
File "/home/daisyden/upstream/gradient_test/test/test_ops_fwd_gradients.py", line 72, in _forward_grad_helper
call_grad_test_helper()
File "/home/daisyden/upstream/gradient_test/test/test_ops_fwd_gradients.py", line 60, in call_grad_test_helper
self._grad_test_helper(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 5628, in _grad_test_helper
return self._check_helper(device, dtype, op, variant, 'gradcheck', check_forward_ad=check_forward_ad,
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 5597, in _check_helper
self.assertTrue(gradcheck(fn, gradcheck_args,
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 5187, in gradcheck
return torch.autograd.gradcheck(fn, inputs, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 2086, in gradcheck
return _gradcheck_helper(**args)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 2115, in _gradcheck_helper
_gradcheck_real_imag(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 1585, in _gradcheck_real_imag
gradcheck_fn(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 1956, in _fast_gradcheck
_check_analytical_numerical_equal(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 1884, in _check_analytical_numerical_equal
raise GradcheckError(
torch.autograd.gradcheck.GradcheckError: Jacobian computed with forward mode mismatch for output 0 with respect to input 0,
numerical:tensor([0., 0., 0., 0., 0.], device='xpu:0', dtype=torch.float64)
analytical:tensor([0.2034, 0.6639, 0.2551, 0.3067, 0.5989], device='xpu:0',
dtype=torch.float64)
The above quantities relating the numerical and analytical jacobians are computed
in fast mode. See: https://github.com/pytorch/pytorch/issues/53876 for more background
about fast mode. Below, we recompute numerical and analytical jacobians in slow mode:
Numerical:
tensor([[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.]], device='xpu:0', dtype=torch.float64)
Analytical:
tensor([[1., 0., 0., 0., 0.],
[0., 1., 0., 0., 0.],
[0., 0., 1., 0., 0.],
[0., 0., 0., 1., 0.],
[0., 0., 0., 0., 1.]], device='xpu:0', dtype=torch.float64,
grad_fn=<CopySlices>)
The max per-element difference (slow mode) is: 1.0.
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 3334, in wrapper
method(*args, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 428, in instantiated_test
result = test(self, **param_kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 1741, in wrapper
fn(*args, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 1163, in test_wrapper
raise e_tracked from e
Exception: Jacobian computed with forward mode mismatch for output 0 with respect to input 0,
numerical:tensor([0., 0., 0., 0., 0.], device='xpu:0', dtype=torch.float64)
analytical:tensor([0.2034, 0.6639, 0.2551, 0.3067, 0.5989], device='xpu:0',
dtype=torch.float64)
The above quantities relating the numerical and analytical jacobians are computed
in fast mode. See: https://github.com/pytorch/pytorch/issues/53876 for more background
about fast mode. Below, we recompute numerical and analytical jacobians in slow mode:
Numerical:
tensor([[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.]], device='xpu:0', dtype=torch.float64)
Analytical:
tensor([[1., 0., 0., 0., 0.],
[0., 1., 0., 0., 0.],
[0., 0., 1., 0., 0.],
[0., 0., 0., 1., 0.],
[0., 0., 0., 0., 1.]], device='xpu:0', dtype=torch.float64,
grad_fn=<CopySlices>)
The max per-element difference (slow mode) is: 1.0.
Caused by sample input at index 0: SampleInput(input=Tensor[size=(5,), device="xpu:0", dtype=torch.float64], args=TensorList[Tensor[size=(5, 10), device="xpu:0", dtype=torch.float64], Tensor[size=(10,), device="xpu:0", dtype=torch.float64]], kwargs={'beta': '1', 'alpha': '1'}, broadcasts_input=False, name='')
To execute this test, run the following from the base repo dir:
PYTORCH_OPINFO_SAMPLE_INPUT_INDEX=0 python test/test_ops_fwd_gradients.py TestFwdGradientsXPU.test_inplace_forward_mode_AD_addmv_xpu_float64
This message can be suppressed by setting PYTORCH_PRINT_REPRO_ON_FAILURE=0
Command: PYTORCH_TEST_WITH_SLOW=1 pytest -v test/test_ops_gradients.py -k test_inplace_grad_baddbmm_xpu_float64
Traceback (most recent call last):
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 1151, in test_wrapper
return test(*args, **kwargs)
File "/home/daisyden/upstream/gradient_test/test/test_ops_gradients.py", line 62, in test_inplace_grad
self._grad_test_helper(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 5628, in _grad_test_helper
return self._check_helper(device, dtype, op, variant, 'gradcheck', check_forward_ad=check_forward_ad,
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 5597, in _check_helper
self.assertTrue(gradcheck(fn, gradcheck_args,
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 5187, in gradcheck
return torch.autograd.gradcheck(fn, inputs, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 2086, in gradcheck
return _gradcheck_helper(**args)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 2115, in _gradcheck_helper
_gradcheck_real_imag(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 1518, in _gradcheck_real_imag
gradcheck_fn(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 1956, in _fast_gradcheck
_check_analytical_numerical_equal(
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/autograd/gradcheck.py", line 1884, in _check_analytical_numerical_equal
raise GradcheckError(
torch.autograd.gradcheck.GradcheckError: Jacobian mismatch for output 0 with respect to input 1,
numerical:-2.1440760583563283
analytical:-1.4775956362133393
The above quantities relating the numerical and analytical jacobians are computed
in fast mode. See: https://github.com/pytorch/pytorch/issues/53876 for more background
about fast mode. Below, we recompute numerical and analytical jacobians in slow mode:
Numerical:
tensor([[-5.8413, 8.8215, -5.7220, ..., 0.0000, 0.0000, 0.0000],
[-4.4107, -5.2452, 9.5367, ..., 0.0000, 0.0000, 0.0000],
[-5.0068, -6.9141, 1.9073, ..., 0.0000, 0.0000, 0.0000],
...,
[ 0.0000, 0.0000, 0.0000, ..., 0.0000, 5.7220, 7.6294],
[ 0.0000, 0.0000, 0.0000, ..., -7.6294, -5.7220, -1.9073],
[ 0.0000, 0.0000, 0.0000, ..., -7.6294, -1.9073, 5.7220]],
device='xpu:0', dtype=torch.float64)
Analytical:
tensor([[-5.8068, 8.7546, -5.8279, ..., 0.0000, 0.0000, 0.0000],
[-4.4340, -5.2358, 7.9194, ..., 0.0000, 0.0000, 0.0000],
[-4.9885, -7.0509, 0.8998, ..., 0.0000, 0.0000, 0.0000],
...,
[ 0.0000, 0.0000, 0.0000, ..., -0.7064, 4.3908, 7.2704],
[ 0.0000, 0.0000, 0.0000, ..., -5.4474, -4.6293, -1.6703],
[ 0.0000, 0.0000, 0.0000, ..., -4.4873, -0.2133, 4.4432]],
device='xpu:0', dtype=torch.float64)
The max per-element difference (slow mode) is: 7.513290386896653.
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 3334, in wrapper
method(*args, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 428, in instantiated_test
result = test(self, **param_kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 1741, in wrapper
fn(*args, **kwargs)
File "/home/daisyden/miniforge3/envs/gradient_test/lib/python3.10/site-packages/torch/testing/_internal/common_device_type.py", line 1163, in test_wrapper
raise e_tracked from e
Exception: Jacobian mismatch for output 0 with respect to input 1,
numerical:-2.1440760583563283
analytical:-1.4775956362133393
The above quantities relating the numerical and analytical jacobians are computed
in fast mode. See: https://github.com/pytorch/pytorch/issues/53876 for more background
about fast mode. Below, we recompute numerical and analytical jacobians in slow mode:
Numerical:
tensor([[-5.8413, 8.8215, -5.7220, ..., 0.0000, 0.0000, 0.0000],
[-4.4107, -5.2452, 9.5367, ..., 0.0000, 0.0000, 0.0000],
[-5.0068, -6.9141, 1.9073, ..., 0.0000, 0.0000, 0.0000],
...,
[ 0.0000, 0.0000, 0.0000, ..., 0.0000, 5.7220, 7.6294],
[ 0.0000, 0.0000, 0.0000, ..., -7.6294, -5.7220, -1.9073],
[ 0.0000, 0.0000, 0.0000, ..., -7.6294, -1.9073, 5.7220]],
device='xpu:0', dtype=torch.float64)
Analytical:
tensor([[-5.8068, 8.7546, -5.8279, ..., 0.0000, 0.0000, 0.0000],
[-4.4340, -5.2358, 7.9194, ..., 0.0000, 0.0000, 0.0000],
[-4.9885, -7.0509, 0.8998, ..., 0.0000, 0.0000, 0.0000],
...,
[ 0.0000, 0.0000, 0.0000, ..., -0.7064, 4.3908, 7.2704],
[ 0.0000, 0.0000, 0.0000, ..., -5.4474, -4.6293, -1.6703],
[ 0.0000, 0.0000, 0.0000, ..., -4.4873, -0.2133, 4.4432]],
device='xpu:0', dtype=torch.float64)
The max per-element difference (slow mode) is: 7.513290386896653.
Caused by sample input at index 0: SampleInput(input=Tensor[size=(5, 5, 10), device="xpu:0", dtype=torch.float64], args=TensorList[Tensor[size=(5, 5, 5), device="xpu:0", dtype=torch.float64], Tensor[size=(5, 5, 10), device="xpu:0", dtype=torch.float64]], kwargs={'beta': '1', 'alpha': '1'}, broadcasts_input=False, name='')
To execute this test, run the following from the base repo dir:
PYTORCH_OPINFO_SAMPLE_INPUT_INDEX=0 PYTORCH_TEST_WITH_SLOW=1 python test/test_ops_gradients.py TestBwdGradientsXPU.test_inplace_grad_baddbmm_xpu_float64
This message can be suppressed by setting PYTORCH_PRINT_REPRO_ON_FAILURE=0
Versions
Collecting environment information...
PyTorch version: N/A
Is debug build: N/A
CUDA used to build PyTorch: N/A
ROCM used to build PyTorch: N/A
OS: Ubuntu 22.04.5 LTS (x86_64)
GCC version: (Ubuntu 13.1.0-8ubuntu1~22.04) 13.1.0
Clang version: Could not collect
CMake version: version 3.31.6
Libc version: glibc-2.35
Python version: 3.10.19 | packaged by conda-forge | (main, Oct 22 2025, 22:29:10) [GCC 14.3.0] (64-bit runtime)
Python platform: Linux-5.15.0-73-generic-x86_64-with-glibc2.35
Is CUDA available: N/A
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: N/A
GPU models and configuration: Could not collect
Nvidia driver version: Could not collect
cuDNN version: Could not collect
Is XPU available: N/A
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: N/A
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 46 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 48
On-line CPU(s) list: 0-47
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Gold 6342 CPU @ 2.80GHz
CPU family: 6
Model: 106
Thread(s) per core: 2
Core(s) per socket: 24
Socket(s): 1
Stepping: 6
CPU max MHz: 3500.0000
CPU min MHz: 800.0000
BogoMIPS: 5600.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 invpcid_single intel_ppin ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid fsrm md_clear pconfig flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 1.1 MiB (24 instances)
L1i cache: 768 KiB (24 instances)
L2 cache: 30 MiB (24 instances)
L3 cache: 36 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-47
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT vulnerable
Vulnerability Retbleed: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Versions of relevant libraries:
[pip3] dpcpp-cpp-rt==2025.2.1
[pip3] impi-rt==2021.16.1
[pip3] intel-cmplr-lib-rt==2025.2.1
[pip3] intel-cmplr-lib-ur==2025.2.1
[pip3] intel-cmplr-lic-rt==2025.2.1
[pip3] intel-opencl-rt==2025.2.1
[pip3] intel-openmp==2025.2.1
[pip3] intel-pti==0.13.1
[pip3] intel-sycl-rt==2025.2.1
[pip3] mkl==2025.2.0
[pip3] mypy==1.16.0
[pip3] mypy_extensions==1.1.0
[pip3] numpy==1.22.4
[pip3] oneccl==2021.16.1
[pip3] oneccl-devel==2021.16.1
[pip3] onemkl-sycl-blas==2025.2.0
[pip3] onemkl-sycl-dft==2025.2.0
[pip3] onemkl-sycl-lapack==2025.2.0
[pip3] onemkl-sycl-rng==2025.2.0
[pip3] onemkl-sycl-sparse==2025.2.0
[pip3] onnx==1.19.1
[pip3] onnx-ir==0.1.12
[pip3] onnxscript==0.5.4
[pip3] optree==0.13.0
[pip3] pytorch-triton-xpu==3.5.0+git1b0418a9
[pip3] tbb==2022.2.0
[pip3] tcmlib==1.4.0
[pip3] torch==2.10.0.dev20251110+xpu
[pip3] torchaudio==2.10.0.dev20251111+xpu
[pip3] torchvision==0.25.0.dev20251111+xpu
[pip3] umf==0.11.0
[conda] dpcpp-cpp-rt 2025.2.1 pypi_0 pypi
[conda] impi-rt 2021.16.1 pypi_0 pypi
[conda] intel-cmplr-lib-rt 2025.2.1 pypi_0 pypi
[conda] intel-cmplr-lib-ur 2025.2.1 pypi_0 pypi
[conda] intel-cmplr-lic-rt 2025.2.1 pypi_0 pypi
[conda] intel-opencl-rt 2025.2.1 pypi_0 pypi
[conda] intel-openmp 2025.2.1 pypi_0 pypi
[conda] intel-pti 0.13.1 pypi_0 pypi
[conda] intel-sycl-rt 2025.2.1 pypi_0 pypi
[conda] mkl 2025.2.0 pypi_0 pypi
[conda] numpy 1.22.4 pypi_0 pypi
[conda] oneccl 2021.16.1 pypi_0 pypi
[conda] oneccl-devel 2021.16.1 pypi_0 pypi
[conda] onemkl-sycl-blas 2025.2.0 pypi_0 pypi
[conda] onemkl-sycl-dft 2025.2.0 pypi_0 pypi
[conda] onemkl-sycl-lapack 2025.2.0 pypi_0 pypi
[conda] onemkl-sycl-rng 2025.2.0 pypi_0 pypi
[conda] onemkl-sycl-sparse 2025.2.0 pypi_0 pypi
[conda] optree 0.13.0 pypi_0 pypi
[conda] pytorch-triton-xpu 3.5.0+git1b0418a9 pypi_0 pypi
[conda] tbb 2022.2.0 pypi_0 pypi
[conda] tcmlib 1.4.0 pypi_0 pypi
[conda] torch 2.10.0.dev20251110+xpu pypi_0 pypi
[conda] torchaudio 2.10.0.dev20251111+xpu pypi_0 pypi
[conda] torchvision 0.25.0.dev20251111+xpu pypi_0 pypi
[conda] umf 0.11.0 pypi_0 pypi