Is it possible to export only one onnx file of DeepFilterNet？

[hqynew]

Hi,
We want to export and get the onnx of whole model.
When run the following export script with exprot-full mode:
def export(
model,
export_dir: str,
df_state: DF,
check: bool = True,
simplify: bool = True,
opset=14,
export_full: bool = True,
print_graph: bool = False,
):
model = deepcopy(model).to("cpu")
model.eval()
p = ModelParams()
audio = torch.randn((1, 1 * p.sr))
spec, feat_erb, feat_spec = df_features(audio, df_state, p.nb_df, device="cpu")

# Export full model
if export_full:
    path = os.path.join(export_dir, "deepfilternet3.onnx")
    input_names = ["spec", "feat_erb", "feat_spec"]
    dynamic_axes = {
        "spec": {2: "S"},
        "feat_erb": {2: "S"},
        "feat_spec": {2: "S"},
        "enh": {2: "S"},
        "m": {2: "S"},
        "lsnr": {1: "S"},
    }
    inputs = (spec, feat_erb, feat_spec)
    output_names = ["enh", "m", "lsnr", "coefs"]
    export_impl(
        path,
        model,
        inputs=inputs,
        input_names=input_names,
        output_names=output_names,
        dynamic_axes=dynamic_axes,
        jit=False,
        check=check,
        simplify=simplify,
        opset_version=opset,
        print_graph=print_graph,
    )

it outputs the error:
torch.onnx.errors.UnsupportedOperatorError: Exporting the operator 'aten::view_as_complex' to ONNX opset version 17 is not supported. Please feel free to request support or submit a pull request on PyTorch GitHub: https://github.com/pytorch/pytorch/issues.

opset = 14 also has same error.

Can anyone help?

Thanks!