I’m a bit confused by some “hanging nodes” I am seeing in the Nuetron onnx visualizer. They seem to have no input, only output. Can someone explain to me how these integrate with the rest of the tree? Here is a screen shot from the Python Onnx Export example here:
https://pytorch.org/tutorials//beginner/onnx/export_simple_model_to_onnx_tutorial.html
The page also has another screen shot which doesn’t seem related to the example code shown, but displays the same artifact(s). In this case they are transpose operations with no inputs, only outputs.
Thanks!