RuntimeError: PyTorch convert function for op 'pythonop' not implemented.

Hey there,

I am getting the following error when converting a model using coreml.
: RuntimeError: PyTorch convert function for op 'pythonop' not implemented.
"PyTorch convert function for op '{}' not implemented.".format(node.kind)

I am using the strict flag before in the trace function. (since the input size can vary).
traced_model_detector = torch.jit.trace(...., strict=False)

Did anybody encounter this issue before?

Thank you!

Start by reproducing conversion of the traced_model_detector created with torch.jit.trace(..., strict=False), then inspect why the graph contains the unsupported 'pythonop'. The issue does not name a source file or test; done would require identifying the cause and establishing a supported conversion path or a documented limitation.

Written by the indexing model from the issue text.