NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

Meta: registered at /dev/null:241 [kernel]
BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

Meta: registered at /dev/null:241 [kernel]
BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

Meta: registered at /dev/null:241 [kernel]
BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
onnxscript\tests\function_libs\torch_lib\ops_test.py:209: in run_test_output_match
    torch_output = op(*inputs, **cpu_sample.kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\testing\_internal\opinfo\core.py:1112: in __call__
    return self.op(*args, **kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\_ops.py:825: in __call__
    return self_._op(*args, **(kwargs or {}))
E   NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].
E   
E   Meta: registered at /dev/null:241 [kernel]
E   BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
E   Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
E   FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
E   Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
E   Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
E   Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
E   Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
E   ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
E   ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
E   AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
E   AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
E   AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
E   FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
E   BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
E   FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
E   Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
E   VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
E   FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
E   PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
E   FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
E   PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
E   PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
onnxscript\tests\function_libs\torch_lib\ops_test.py:209: in run_test_output_match
    torch_output = op(*inputs, **cpu_sample.kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\testing\_internal\opinfo\core.py:1112: in __call__
    return self.op(*args, **kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\_ops.py:825: in __call__
    return self_._op(*args, **(kwargs or {}))
E   NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].
E   
E   Meta: registered at /dev/null:241 [kernel]
E   BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
E   Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
E   FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
E   Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
E   Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
E   Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
E   Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
E   ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
E   ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
E   AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
E   AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
E   AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
E   FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
E   BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
E   FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
E   Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
E   VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
E   FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
E   PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
E   FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
E   PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
E   PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
onnxscript\tests\function_libs\torch_lib\ops_test.py:209: in run_test_output_match
    torch_output = op(*inputs, **cpu_sample.kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\testing\_internal\opinfo\core.py:1112: in __call__
    return self.op(*args, **kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\_ops.py:825: in __call__
    return self_._op(*args, **(kwargs or {}))
E   NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].
E   
E   Meta: registered at /dev/null:241 [kernel]
E   BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
E   Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
E   FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
E   Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
E   Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
E   Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
E   Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
E   ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
E   ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
E   AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
E   AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
E   AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
E   FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
E   BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
E   FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
E   Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
E   VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
E   FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
E   PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
E   FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
E   PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
E   PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]

TypeError: 'NoneType' object is not subscriptable
TypeError: 'NoneType' object is not subscriptable
onnxscript/tests/function_libs/torch_lib/ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript/tests/function_libs/torch_lib/ops_test_common.py:590: in executor
    return function(*args, **kwargs)
onnxscript/values.py:577: in __call__
    return self.func(*args, **kwargs)
onnxscript/function_libs/torch_lib/ops/nn.py:2322: in aten_upsample_bilinear2d_vec
    self, output_size, scale_factors[0], scale_factors[1], align_corners
E   TypeError: 'NoneType' object is not subscriptable
onnxscript/tests/function_libs/torch_lib/ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript/tests/function_libs/torch_lib/ops_test_common.py:590: in executor
    return function(*args, **kwargs)
onnxscript/values.py:577: in __call__
    return self.func(*args, **kwargs)
onnxscript/function_libs/torch_lib/ops/nn.py:2322: in aten_upsample_bilinear2d_vec
    self, output_size, scale_factors[0], scale_factors[1], align_corners
E   TypeError: 'NoneType' object is not subscriptable

EOFError
onnxscript/tests/function_libs/torch_lib/ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript/tests/function_libs/torch_lib/ops_test_common.py:540: in _capture_graph_and_evaluate_torch_script_evaluator
    return _safe_ort_session_run(onnx_model.SerializeToString(), ort_inputs)
onnxscript/tests/function_libs/torch_lib/ops_test_common.py:345: in _safe_ort_session_run
    return_dict = manager.dict()
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/managers.py:723: in temp
    token, exp = self._create(typeid, *args, **kwds)
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/managers.py:606: in _create
    conn = self._Client(self._address, authkey=self._authkey)
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/connection.py:508: in Client
    answer_challenge(c, authkey)
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/connection.py:752: in answer_challenge
    message = connection.recv_bytes(256)         # reject large message
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/connection.py:216: in recv_bytes
    buf = self._recv_bytes(maxlength)
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/connection.py:414: in _recv_bytes
    buf = self._recv(4)
/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/multiprocessing/connection.py:383: in _recv
    raise EOFError
E   EOFError

TypeError: 'NoneType' object is not subscriptable
TypeError: 'NoneType' object is not subscriptable
onnxscript/tests/function_libs/torch_lib/ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript/tests/function_libs/torch_lib/ops_test_common.py:504: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript/values.py:577: in __call__
    return self.func(*args, **kwargs)
onnxscript/function_libs/torch_lib/ops/nn.py:2322: in aten_upsample_bilinear2d_vec
    self, output_size, scale_factors[0], scale_factors[1], align_corners
E   TypeError: 'NoneType' object is not subscriptable
onnxscript/tests/function_libs/torch_lib/ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript/tests/function_libs/torch_lib/ops_test_common.py:504: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript/values.py:577: in __call__
    return self.func(*args, **kwargs)
onnxscript/function_libs/torch_lib/ops/nn.py:2322: in aten_upsample_bilinear2d_vec
    self, output_size, scale_factors[0], scale_factors[1], align_corners
E   TypeError: 'NoneType' object is not subscriptable

NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

Meta: registered at /dev/null:241 [kernel]
BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

Meta: registered at /dev/null:241 [kernel]
BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

Meta: registered at /dev/null:241 [kernel]
BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
onnxscript\tests\function_libs\torch_lib\ops_test.py:209: in run_test_output_match
    torch_output = op(*inputs, **cpu_sample.kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\testing\_internal\opinfo\core.py:1112: in __call__
    return self.op(*args, **kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\_ops.py:825: in __call__
    return self_._op(*args, **(kwargs or {}))
E   NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].
E   
E   Meta: registered at /dev/null:241 [kernel]
E   BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
E   Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
E   FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
E   Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
E   Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
E   Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
E   Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
E   ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
E   ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
E   AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
E   AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
E   AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
E   FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
E   BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
E   FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
E   Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
E   VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
E   FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
E   PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
E   FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
E   PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
E   PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
onnxscript\tests\function_libs\torch_lib\ops_test.py:209: in run_test_output_match
    torch_output = op(*inputs, **cpu_sample.kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\testing\_internal\opinfo\core.py:1112: in __call__
    return self.op(*args, **kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\_ops.py:825: in __call__
    return self_._op(*args, **(kwargs or {}))
E   NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].
E   
E   Meta: registered at /dev/null:241 [kernel]
E   BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
E   Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
E   FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
E   Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
E   Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
E   Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
E   Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
E   ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
E   ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
E   AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
E   AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
E   AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
E   FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
E   BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
E   FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
E   Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
E   VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
E   FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
E   PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
E   FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
E   PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
E   PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]
onnxscript\tests\function_libs\torch_lib\ops_test.py:209: in run_test_output_match
    torch_output = op(*inputs, **cpu_sample.kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\testing\_internal\opinfo\core.py:1112: in __call__
    return self.op(*args, **kwargs)
.nox\test_torch_nightly\lib\site-packages\torch\_ops.py:825: in __call__
    return self_._op(*args, **(kwargs or {}))
E   NotImplementedError: Could not run 'aten::_scaled_dot_product_flash_attention' with arguments from the 'CPU' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::_scaled_dot_product_flash_attention' is only available for these backends: [Meta, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradHIP, AutogradXLA, AutogradMPS, AutogradIPU, AutogradXPU, AutogradHPU, AutogradVE, AutogradLazy, AutogradMTIA, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, AutogradMeta, AutogradNestedTensor, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].
E   
E   Meta: registered at /dev/null:241 [kernel]
E   BackendSelect: fallthrough registered at ..\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
E   Python: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:154 [backend fallback]
E   FuncTorchDynamicLayerBackMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:498 [backend fallback]
E   Functionalize: registered at ..\aten\src\ATen\FunctionalizeFallbackKernel.cpp:324 [backend fallback]
E   Named: registered at ..\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
E   Conjugate: registered at ..\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
E   Negative: registered at ..\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
E   ZeroTensor: registered at ..\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
E   ADInplaceOrView: fallthrough registered at ..\aten\src\ATen\core\VariableFallbackKernel.cpp:86 [backend fallback]
E   AutogradOther: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradCUDA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHIP: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXLA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMPS: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradIPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradXPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradHPU: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradVE: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradLazy: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMTIA: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse1: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse2: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradPrivateUse3: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradMeta: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   AutogradNestedTensor: registered at ..\torch\csrc\autograd\generated\VariableType_1.cpp:16340 [autograd kernel]
E   Tracer: registered at ..\torch\csrc\autograd\generated\TraceType_1.cpp:16033 [kernel]
E   AutocastCPU: fallthrough registered at ..\aten\src\ATen\autocast_mode.cpp:378 [backend fallback]
E   AutocastCUDA: registered at ..\aten\src\ATen\autocast_mode.cpp:248 [kernel]
E   FuncTorchBatched: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:720 [backend fallback]
E   BatchedNestedTensor: registered at ..\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:746 [backend fallback]
E   FuncTorchVmapMode: fallthrough registered at ..\aten\src\ATen\functorch\VmapModeRegistrations.cpp:28 [backend fallback]
E   Batched: registered at ..\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
E   VmapMode: fallthrough registered at ..\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
E   FuncTorchGradWrapper: registered at ..\aten\src\ATen\functorch\TensorWrapper.cpp:203 [backend fallback]
E   PythonTLSSnapshot: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:162 [backend fallback]
E   FuncTorchDynamicLayerFrontMode: registered at ..\aten\src\ATen\functorch\DynamicLayer.cpp:494 [backend fallback]
E   PreDispatch: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:166 [backend fallback]
E   PythonDispatcher: registered at ..\aten\src\ATen\core\PythonFallbackKernel.cpp:158 [backend fallback]

AssertionError: Unable to import 'onnxscript.tests.onnx_backend_test_code.test_size' (file: WindowsPath('D:/a/onnxscript/onnxscript/onnxscript/tests/onnx_backend_test_code/test_size.py'))
----
import numpy
from onnx import TensorProto
from onnx.helper import make_tensor
from onnxscript import script, external_tensor
from onnxscript.values import Opset
from onnxscript.onnx_types import FLOAT, INT64
from onnxscript.onnx_opset import opset19

@script()
def bck_test_size(x: FLOAT[3,4,5]) -> (INT64):
    y = opset19.Size(x)
    return y
onnxscript\backend\onnx_export_test.py:116: in extract_functions
    mod = importlib.import_module(import_name)
C:\hostedtoolcache\windows\Python\3.8.10\x64\lib\importlib\__init__.py:127: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'onnxscript.tests.onnx_backend_test_code.test_size'

The above exception was the direct cause of the following exception:
.nox\test\lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
onnxscript\backend\onnx_export_test.py:247: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
onnxscript\backend\onnx_export_test.py:118: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'onnxscript.tests.onnx_backend_test_code.test_size' (file: WindowsPath('D:/a/onnxscript/onnxscript/onnxscript/tests/onnx_backend_test_code/test_size.py'))
E   ----
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import FLOAT, INT64
E   from onnxscript.onnx_opset import opset19
E   
E   @script()
E   def bck_test_size(x: FLOAT[3,4,5]) -> (INT64):
E       y = opset19.Size(x)
E       return y

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2, float16[1,2,3] input_3) => (float16[1,2,3] _val_4, float16[1,1,1] _val_5, float16[1,1,1] _val_6) 
   <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2, float16[1,2,3] input_3, float16[1,2,3] _val_4, float16[1,1,1] _val_5, float16[1,1,1] _val_6>
{
   _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, input_2, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_3) => (float16[1,2,3] _val_7, float16[1,1,1] _val_8, float16[1,1,1] _val_9) 
   <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_3, float16[1,2,3] _val_7, float16[1,1,1] _val_8, float16[1,1,1] _val_9, float[1] _val_3, int64[3] _val_4, float[1,2,3] _val_5, float16[1,2,3] _val_6>
{
   _val_3 = Constant <value_floats: floats = [1]> ()
   _val_4 = Shape <start: int = -3> (input_0)
   _val_5 = Expand (_val_3, _val_4)
   _val_6 = CastLike (_val_5, input_0)
   _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, _val_6, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2) => (float16[1,2,3] _val_3, float16[1,1,1] _val_4, float16[1,1,1] _val_5) 
   <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2, float16[1,2,3] _val_3, float16[1,1,1] _val_4, float16[1,1,1] _val_5>
{
   _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, input_2)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2,3] input_0, int64[3] input_1) => (float16[1,2,3] _val_6, float16[1,1,1] _val_7, float16[1,1,1] _val_8) 
   <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] _val_6, float16[1,1,1] _val_7, float16[1,1,1] _val_8, float[1] _val_2, int64[3] _val_3, float[1,2,3] _val_4, float16[1,2,3] _val_5>
{
   _val_2 = Constant <value_floats: floats = [1]> ()
   _val_3 = Shape <start: int = -3> (input_0)
   _val_4 = Expand (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
   _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_2, float16[2,3] input_3) => (float16[2,2,3] _val_4, float16[2,1,1] _val_5, float16[2,1,1] _val_6) 
   <float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_2, float16[2,3] input_3, float16[2,2,3] _val_4, float16[2,1,1] _val_5, float16[2,1,1] _val_6>
{
   _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, input_2, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_3) => (float16[2,2,3] _val_7, float16[2,1,1] _val_8, float16[2,1,1] _val_9) 
   <float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_3, float16[2,2,3] _val_7, float16[2,1,1] _val_8, float16[2,1,1] _val_9, float[1] _val_3, int64[2] _val_4, float[2,3] _val_5, float16[2,3] _val_6>
{
   _val_3 = Constant <value_floats: floats = [1]> ()
   _val_4 = Shape <start: int = -2> (input_0)
   _val_5 = Expand (_val_3, _val_4)
   _val_6 = CastLike (_val_5, input_0)
   _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, _val_6, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_2) => (float16[2,2,3] _val_3, float16[2,1,1] _val_4, float16[2,1,1] _val_5) 
   <float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_2, float16[2,2,3] _val_3, float16[2,1,1] _val_4, float16[2,1,1] _val_5>
{
   _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, input_2)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[2,2,3] input_0, int64[2] input_1) => (float16[2,2,3] _val_6, float16[2,1,1] _val_7, float16[2,1,1] _val_8) 
   <float16[2,2,3] input_0, int64[2] input_1, float16[2,2,3] _val_6, float16[2,1,1] _val_7, float16[2,1,1] _val_8, float[1] _val_2, int64[2] _val_3, float[2,3] _val_4, float16[2,3] _val_5>
{
   _val_2 = Constant <value_floats: floats = [1]> ()
   _val_3 = Shape <start: int = -2> (input_0)
   _val_4 = Expand (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
   _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3) => (float16[1] _val_4, float16[1] _val_5, float16[1] _val_6) 
   <float16[1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3, float16[1] _val_4, float16[1] _val_5, float16[1] _val_6>
{
   _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1] input_0, int64[1] input_1, float16[1] input_3) => (float16[1] _val_7, float16[1] _val_8, float16[1] _val_9) 
   <float16[1] input_0, int64[1] input_1, float16[1] input_3, float16[1] _val_7, float16[1] _val_8, float16[1] _val_9, float[1] _val_3, int64[1] _val_4, float[1] _val_5, float16[1] _val_6>
{
   _val_3 = Constant <value_floats: floats = [1]> ()
   _val_4 = Shape <start: int = -1> (input_0)
   _val_5 = Expand (_val_3, _val_4)
   _val_6 = CastLike (_val_5, input_0)
   _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_6, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1] input_0, int64[1] input_1, float16[1] input_2) => (float16[1] _val_3, float16[1] _val_4, float16[1] _val_5) 
   <float16[1] input_0, int64[1] input_1, float16[1] input_2, float16[1] _val_3, float16[1] _val_4, float16[1] _val_5>
{
   _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1] input_0, int64[1] input_1) => (float16[1] _val_6, float16[1] _val_7, float16[1] _val_8) 
   <float16[1] input_0, int64[1] input_1, float16[1] _val_6, float16[1] _val_7, float16[1] _val_8, float[1] _val_2, int64[1] _val_3, float[1] _val_4, float16[1] _val_5>
{
   _val_2 = Constant <value_floats: floats = [1]> ()
   _val_3 = Shape <start: int = -1> (input_0)
   _val_4 = Expand (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
   _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2] input_0, int64[1] input_1, float16[2] input_2, float16[2] input_3) => (float16[1,2] _val_4, float16[1,1] _val_5, float16[1,1] _val_6) 
   <float16[1,2] input_0, int64[1] input_1, float16[2] input_2, float16[2] input_3, float16[1,2] _val_4, float16[1,1] _val_5, float16[1,1] _val_6>
{
   _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2] input_0, int64[1] input_1, float16[2] input_3) => (float16[1,2] _val_7, float16[1,1] _val_8, float16[1,1] _val_9) 
   <float16[1,2] input_0, int64[1] input_1, float16[2] input_3, float16[1,2] _val_7, float16[1,1] _val_8, float16[1,1] _val_9, float[1] _val_3, int64[1] _val_4, float[2] _val_5, float16[2] _val_6>
{
   _val_3 = Constant <value_floats: floats = [1]> ()
   _val_4 = Shape <start: int = -1> (input_0)
   _val_5 = Expand (_val_3, _val_4)
   _val_6 = CastLike (_val_5, input_0)
   _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_6, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2] input_0, int64[1] input_1, float16[2] input_2) => (float16[1,2] _val_3, float16[1,1] _val_4, float16[1,1] _val_5) 
   <float16[1,2] input_0, int64[1] input_1, float16[2] input_2, float16[1,2] _val_3, float16[1,1] _val_4, float16[1,1] _val_5>
{
   _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[1,2] input_0, int64[1] input_1) => (float16[1,2] _val_6, float16[1,1] _val_7, float16[1,1] _val_8) 
   <float16[1,2] input_0, int64[1] input_1, float16[1,2] _val_6, float16[1,1] _val_7, float16[1,1] _val_8, float[1] _val_2, int64[1] _val_3, float[2] _val_4, float16[2] _val_5>
{
   _val_2 = Constant <value_floats: floats = [1]> ()
   _val_3 = Shape <start: int = -1> (input_0)
   _val_4 = Expand (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
   _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[0,1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3) => (float16[0,1] _val_4, float16[0,1] _val_5, float16[0,1] _val_6) 
   <float16[0,1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3, float16[0,1] _val_4, float16[0,1] _val_5, float16[0,1] _val_6>
{
   _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[0,1] input_0, int64[1] input_1, float16[1] input_3) => (float16[0,1] _val_7, float16[0,1] _val_8, float16[0,1] _val_9) 
   <float16[0,1] input_0, int64[1] input_1, float16[1] input_3, float16[0,1] _val_7, float16[0,1] _val_8, float16[0,1] _val_9, float[1] _val_3, int64[1] _val_4, float[1] _val_5, float16[1] _val_6>
{
   _val_3 = Constant <value_floats: floats = [1]> ()
   _val_4 = Shape <start: int = -1> (input_0)
   _val_5 = Expand (_val_3, _val_4)
   _val_6 = CastLike (_val_5, input_0)
   _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_6, input_3)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[0,1] input_0, int64[1] input_1, float16[1] input_2) => (float16[0,1] _val_3, float16[0,1] _val_4, float16[0,1] _val_5) 
   <float16[0,1] input_0, int64[1] input_1, float16[1] input_2, float16[0,1] _val_3, float16[0,1] _val_4, float16[0,1] _val_5>
{
   _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.3.0"
>
main_graph (float16[0,1] input_0, int64[1] input_1) => (float16[0,1] _val_6, float16[0,1] _val_7, float16[0,1] _val_8) 
   <float16[0,1] input_0, int64[1] input_1, float16[0,1] _val_6, float16[0,1] _val_7, float16[0,1] _val_8, float[1] _val_2, int64[1] _val_3, float[1] _val_4, float16[1] _val_5>
{
   _val_2 = Constant <value_floats: floats = [1]> ()
   _val_3 = Shape <start: int = -1> (input_0)
   _val_4 = Expand (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
   _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2, float16[1,2,3] input_3) => (float16[1,2,3] _val_4, float16[1,1,1] _val_5, float16[1,1,1] _val_6) 
E      <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2, float16[1,2,3] input_3, float16[1,2,3] _val_4, float16[1,1,1] _val_5, float16[1,1,1] _val_6>
E   {
E      _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, input_2, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_3) => (float16[1,2,3] _val_7, float16[1,1,1] _val_8, float16[1,1,1] _val_9) 
E      <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_3, float16[1,2,3] _val_7, float16[1,1,1] _val_8, float16[1,1,1] _val_9, float[1] _val_3, int64[3] _val_4, float[1,2,3] _val_5, float16[1,2,3] _val_6>
E   {
E      _val_3 = Constant <value_floats: floats = [1]> ()
E      _val_4 = Shape <start: int = -3> (input_0)
E      _val_5 = Expand (_val_3, _val_4)
E      _val_6 = CastLike (_val_5, input_0)
E      _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, _val_6, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2) => (float16[1,2,3] _val_3, float16[1,1,1] _val_4, float16[1,1,1] _val_5) 
E      <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] input_2, float16[1,2,3] _val_3, float16[1,1,1] _val_4, float16[1,1,1] _val_5>
E   {
E      _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, input_2)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2,3] input_0, int64[3] input_1) => (float16[1,2,3] _val_6, float16[1,1,1] _val_7, float16[1,1,1] _val_8) 
E      <float16[1,2,3] input_0, int64[3] input_1, float16[1,2,3] _val_6, float16[1,1,1] _val_7, float16[1,1,1] _val_8, float[1] _val_2, int64[3] _val_3, float[1,2,3] _val_4, float16[1,2,3] _val_5>
E   {
E      _val_2 = Constant <value_floats: floats = [1]> ()
E      _val_3 = Shape <start: int = -3> (input_0)
E      _val_4 = Expand (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E      _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -3, epsilon: float = 0.5, stash_type: int = 1> (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
on…0, int64[2] input_1, float16[2,3] input_3) => (float16[2,2,3] _val_7, float16[2,1,1] _val_8, float16[2,1,1] _val_9) 
E      <float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_3, float16[2,2,3] _val_7, float16[2,1,1] _val_8, float16[2,1,1] _val_9, float[1] _val_3, int64[2] _val_4, float[2,3] _val_5, float16[2,3] _val_6>
E   {
E      _val_3 = Constant <value_floats: floats = [1]> ()
E      _val_4 = Shape <start: int = -2> (input_0)
E      _val_5 = Expand (_val_3, _val_4)
E      _val_6 = CastLike (_val_5, input_0)
E      _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, _val_6, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_2) => (float16[2,2,3] _val_3, float16[2,1,1] _val_4, float16[2,1,1] _val_5) 
E      <float16[2,2,3] input_0, int64[2] input_1, float16[2,3] input_2, float16[2,2,3] _val_3, float16[2,1,1] _val_4, float16[2,1,1] _val_5>
E   {
E      _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, input_2)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[2,2,3] input_0, int64[2] input_1) => (float16[2,2,3] _val_6, float16[2,1,1] _val_7, float16[2,1,1] _val_8) 
E      <float16[2,2,3] input_0, int64[2] input_1, float16[2,2,3] _val_6, float16[2,1,1] _val_7, float16[2,1,1] _val_8, float[1] _val_2, int64[2] _val_3, float[2,3] _val_4, float16[2,3] _val_5>
E   {
E      _val_2 = Constant <value_floats: floats = [1]> ()
E      _val_3 = Shape <start: int = -2> (input_0)
E      _val_4 = Expand (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E      _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -2, epsilon: float = -0.5, stash_type: int = 1> (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3) => (float16[1] _val_4, float16[1] _val_5, float16[1] _val_6) 
E      <float16[1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3, float16[1] _val_4, float16[1] _val_5, float16[1] _val_6>
E   {
E      _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1] input_0, int64[1] input_1, float16[1] input_3) => (float16[1] _val_7, float16[1] _val_8, float16[1] _val_9) 
E      <float16[1] input_0, int64[1] input_1, float16[1] input_3, float16[1] _val_7, float16[1] _val_8, float16[1] _val_9, float[1] _val_3, int64[1] _val_4, float[1] _val_5, float16[1] _val_6>
E   {
E      _val_3 = Constant <value_floats: floats = [1]> ()
E      _val_4 = Shape <start: int = -1> (input_0)
E      _val_5 = Expand (_val_3, _val_4)
E      _val_6 = CastLike (_val_5, input_0)
E      _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_6, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1] input_0, int64[1] input_1, float16[1] input_2) => (float16[1] _val_3, float16[1] _val_4, float16[1] _val_5) 
E      <float16[1] input_0, int64[1] input_1, float16[1] input_2, float16[1] _val_3, float16[1] _val_4, float16[1] _val_5>
E   {
E      _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1] input_0, int64[1] input_1) => (float16[1] _val_6, float16[1] _val_7, float16[1] _val_8) 
E      <float16[1] input_0, int64[1] input_1, float16[1] _val_6, float16[1] _val_7, float16[1] _val_8, float[1] _val_2, int64[1] _val_3, float[1] _val_4, float16[1] _val_5>
E   {
E      _val_2 = Constant <value_floats: floats = [1]> ()
E      _val_3 = Shape <start: int = -1> (input_0)
E      _val_4 = Expand (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E      _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2] input_0, int64[1] input_1, float16[2] input_2, float16[2] input_3) => (float16[1,2] _val_4, float16[1,1] _val_5, float16[1,1] _val_6) 
E      <float16[1,2] input_0, int64[1] input_1, float16[2] input_2, float16[2] input_3, float16[1,2] _val_4, float16[1,1] _val_5, float16[1,1] _val_6>
E   {
E      _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2] input_0, int64[1] input_1, float16[2] input_3) => (float16[1,2] _val_7, float16[1,1] _val_8, float16[1,1] _val_9) 
E      <float16[1,2] input_0, int64[1] input_1, float16[2] input_3, float16[1,2] _val_7, float16[1,1] _val_8, float16[1,1] _val_9, float[1] _val_3, int64[1] _val_4, float[2] _val_5, float16[2] _val_6>
E   {
E      _val_3 = Constant <value_floats: floats = [1]> ()
E      _val_4 = Shape <start: int = -1> (input_0)
E      _val_5 = Expand (_val_3, _val_4)
E      _val_6 = CastLike (_val_5, input_0)
E      _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_6, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2] input_0, int64[1] input_1, float16[2] input_2) => (float16[1,2] _val_3, float16[1,1] _val_4, float16[1,1] _val_5) 
E      <float16[1,2] input_0, int64[1] input_1, float16[2] input_2, float16[1,2] _val_3, float16[1,1] _val_4, float16[1,1] _val_5>
E   {
E      _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[1,2] input_0, int64[1] input_1) => (float16[1,2] _val_6, float16[1,1] _val_7, float16[1,1] _val_8) 
E      <float16[1,2] input_0, int64[1] input_1, float16[1,2] _val_6, float16[1,1] _val_7, float16[1,1] _val_8, float[1] _val_2, int64[1] _val_3, float[2] _val_4, float16[2] _val_5>
E   {
E      _val_2 = Constant <value_floats: floats = [1]> ()
E      _val_3 = Shape <start: int = -1> (input_0)
E      _val_4 = Expand (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E      _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[0,1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3) => (float16[0,1] _val_4, float16[0,1] _val_5, float16[0,1] _val_6) 
E      <float16[0,1] input_0, int64[1] input_1, float16[1] input_2, float16[1] input_3, float16[0,1] _val_4, float16[0,1] _val_5, float16[0,1] _val_6>
E   {
E      _val_4, _val_5, _val_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[0,1] input_0, int64[1] input_1, float16[1] input_3) => (float16[0,1] _val_7, float16[0,1] _val_8, float16[0,1] _val_9) 
E      <float16[0,1] input_0, int64[1] input_1, float16[1] input_3, float16[0,1] _val_7, float16[0,1] _val_8, float16[0,1] _val_9, float[1] _val_3, int64[1] _val_4, float[1] _val_5, float16[1] _val_6>
E   {
E      _val_3 = Constant <value_floats: floats = [1]> ()
E      _val_4 = Shape <start: int = -1> (input_0)
E      _val_5 = Expand (_val_3, _val_4)
E      _val_6 = CastLike (_val_5, input_0)
E      _val_7, _val_8, _val_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_6, input_3)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_0): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[0,1] input_0, int64[1] input_1, float16[1] input_2) => (float16[0,1] _val_3, float16[0,1] _val_4, float16[0,1] _val_5) 
E      <float16[0,1] input_0, int64[1] input_1, float16[1] input_2, float16[0,1] _val_3, float16[0,1] _val_4, float16[0,1] _val_5>
E   {
E      _val_3, _val_4, _val_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, input_2)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:528: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox\test_torch_nightly\lib\site-packages\onnx\checker.py:171: in check_model
    C.check_model(protobuf_string, full_check, skip_opset_compatibility_check)
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:LayerNormalization, node name: LayerNormalization_4): [TypeInferenceError] Inferred elem type differs from existing elem type: (1) vs (10)

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:229: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
onnxscript\tests\function_libs\torch_lib\ops_test_common.py:530: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.3.0"
E   >
E   main_graph (float16[0,1] input_0, int64[1] input_1) => (float16[0,1] _val_6, float16[0,1] _val_7, float16[0,1] _val_8) 
E      <float16[0,1] input_0, int64[1] input_1, float16[0,1] _val_6, float16[0,1] _val_7, float16[0,1] _val_8, float[1] _val_2, int64[1] _val_3, float[1] _val_4, float16[1] _val_5>
E   {
E      _val_2 = Constant <value_floats: floats = [1]> ()
E      _val_3 = Shape <start: int = -1> (input_0)
E      _val_4 = Expand (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E      _val_6, _val_7, _val_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-05, stash_type: int = 1> (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
onnxscript\tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: The values for attribute 'dtype' do not match: torch.float32 != torch.float16.

The above exception was the direct cause of the following exception:
onnxscript\tests\function_libs\torch_lib\ops_test.py:280: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch