Frequent Exceptions¶
Unsupported functions or classes¶
If the converter to onnx fails, function bypass_export_some_errors
may help solving some of them. The ocumentation of this function
gives the list of issues it can bypass.
from experimental_experiment.torch_interpreter.onnx_export_errors import bypass_export_some_errors
with bypass_export_some_errors():
onx = to_onnx(...)
torch._dynamo.exc.Unsupported¶
torch._dynamo.exc.Unsupported: call_function BuiltinVariable(NotImplementedError) [ConstantVariable()] {}
This exception started to show up with transformers==4.38.2 but it does not seem related to it. Wrapping the code with the following fixes it.
with torch.no_grad():
# ...
RuntimeError¶
RuntimeError: Encountered autograd state manager op <built-in function _set_grad_enabled> trying to change global autograd state while exporting.
Wrapping the code around probably solves this issue.
with torch.no_grad():
# ...