I’m trying to move from torch.onnx.export to torch.onnx.dynamo_export
By using torch.onnx.export I can determine the input and output names, activate constant folding on the model and decide to activate the verbose mode during the export operation.
Now, when I’m trying to use torch.onnx.dynamo_export, I don’t find such way to do the same things. These things are pretty basic so I think that maybe I miss something.
My second question is if there is a way to iterate the operators in the nn.Module and predict if the operators are supported by torch.onnx.dynamo_export. Is it possible to do so? I want to print all the unsupported operators in the nn.Module