onnx Infer problem?

[montensorrt]

Is your feature request related to a problem? Please describe.
when infer the onnx model,don‘t find meta_data.json file?
if i want infer onnx,must be using openvino libs?
maybe provide some scripts for onnx model infer ,such as using onnxruntime infer? or other?

[samet-akcay]

Hi @montensorrt, PR #636 addresses this issue. We hope to be able to merge it by the end of today. Hope that answers your question.

[montensorrt]

it's solved，but have a lit problem,why must to export_mode == "openvino",the meta_data.json file would be created,when I only export_mode == "onnx“，the file wondn't created,maybe ,should adjust the order,
Such as:
torch.onnx.export(
model.model,
torch.zeros((1, 3, height, width)).to(model.device),
onnx_path,
opset_version=12,
input_names=["input"],
output_names=["output"],
)

with open(Path(export_path) / "meta_data.json", "w", encoding="utf-8") as metadata_file:
    meta_data = get_model_metadata(model)
    # Convert metadata from torch
    for key, value in meta_data.items():
        if isinstance(value, Tensor):
            meta_data[key] = value.numpy().tolist()
    json.dump(meta_data, metadata_file, ensure_ascii=False, indent=4)

if export_mode == "openvino":