onnx_graphsurgeon: toposort failed on a graph without cycle

Description

Run toposort on this simple onnx got an error:
OnnxGraphSurgeonException: Cycle detected in graph! Are there tensors with duplicate names in the graph?

Environment

onnx==1.16.2
onnx_graphsurgeon==0.5.2

Operating System:
Ubuntu 20.04

Python Version (if applicable):
python==3.8.10

Relevant Files

Model link:
toposort.onnx.zip

Steps To Reproduce

Commands or scripts:

import onnx
import onnx_graphsurgeon as gs

graph = gs.import_onnx(onnx.load('./toposort.onnx'))
graph.toposort()
# `OnnxGraphSurgeonException: Cycle detected in graph! Are there tensors with duplicate names in the graph?`

Have you tried the latest release?:
Yes

Can this model run on other frameworks? For example run ONNX model with ONNXRuntime (polygraphy run <model.onnx> --onnxrt):
Yes

import onnx
import onnxruntime as ort

model = onnx.load('./toposort.onnx')
sess = ort.InferenceSession(model.SerializeToString())
sess.run(['mul_0_output'], {'input': np.array(3., dtype=np.float32)})
# Out: [array(-9., dtype=float32)]

Reproduce the failure with the provided toposort.onnx model using the shown Python script, starting at the graph.toposort() entry point. Compare the reported cycle with the model's actual graph structure; done means this acyclic model sorts successfully without the cycle exception, with coverage for the regression if the project has an appropriate test location.