pytorch - 💡(How to fix) Fix DISABLED test_jacfwd (__main__.FuncTorchHigherOrderOpTestsWithCompiledAutograd) [2 comments, 1 participants]

Traceback (most recent call last): File "/var/lib/jenkins/workspace/test/dynamo/test_higher_order_ops.py", line 5214, in test_jacfwd wrapped_gm = self._compile_check(wrapper_fn, (x,)) File "/var/lib/jenkins/workspace/test/dynamo/test_higher_order_ops.py", line 3789, in _compile_check expected = torch.compile(fn, backend=backend, fullgraph=fullgraph)(*inputs) File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 1051, in compile_wrapper raise e.with_traceback(None) from e.cause # User compiler error torch._dynamo.exc.Unsupported: Attempted to call a super() attribute that is not a function or method Explanation: Dynamo does not know how to trace the call super().unflatten() because super().unflatten is not a function or method attribute. Hint: Ensure the attribute accessed via super() is a standard method or function.

Developer debug context: call_method SuperVariable() unflatten

For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0006.html

from user code: File "/var/lib/jenkins/workspace/test/dynamo/test_higher_order_ops.py", line 5211, in wrapper_fn return torch.func.jacfwd(torch.sin)(x) File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_functorch/eager_transforms.py", line 1401, in wrapper_fn jac_outs_ins = tuple( File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_functorch/eager_transforms.py", line 1402, in <genexpr> tuple( File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_functorch/eager_transforms.py", line 1403, in <genexpr> safe_unflatten(jac_out_in, -1, primal.shape) File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_functorch/eager_transforms.py", line 1247, in safe_unflatten return tensor.unflatten(dim, shape) File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/utils/_device.py", line 122, in torch_function return func(*args, **kwargs) File "/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/_tensor.py", line 1456, in unflatten return super().unflatten(dim, sizes)

Set TORCHDYNAMO_VERBOSE=1 for the internal stack trace (please do this especially if you're reporting a bug to PyTorch). For even more developer context, set TORCH_LOGS="+dynamo"

To execute this test, run the following from the base repo dir: python test/dynamo/test_higher_order_ops.py FuncTorchHigherOrderOpTestsWithCompiledAutograd.test_jacfwd

This message can be suppressed by setting PYTORCH_PRINT_REPRO_ON_FAILURE=0

TL;DR

The test failure is likely due to an issue with Dynamo's ability to trace the call super().unflatten() because super().unflatten is not a function or method attribute, and setting TORCHDYNAMO_VERBOSE=1 may provide more insight into the problem.

Guidance

• Review the error message and the provided debugging instructions to understand the issue better.
• Set TORCHDYNAMO_VERBOSE=1 to get the internal stack trace, which may help identify the root cause of the problem.
• Check the compile-graph-break-site for more information about the graph break.
• Run the test with the provided command python test/dynamo/test_higher_order_ops.py FuncTorchHigherOrderOpTestsWithCompiledAutograd.test_jacfwd to reproduce the issue.

Example

No code example is provided as the issue is related to a specific test failure and requires more context to provide a code snippet.

Notes

The issue seems to be related to the Dynamo compiler and its ability to trace certain function calls. The provided error message and debugging instructions should help identify the root cause of the problem.

Recommendation

Apply workaround: Set TORCHDYNAMO_VERBOSE=1 to get more information about the issue and review the provided debugging instructions to understand the problem better. This may help identify a solution or provide more insight into the issue.