How to set up inputTensor when deploying multi-input model using c++？

[Shyuna]

I am trying to deploy the text2vec model on Windows using the onnxruntime C++ API. The onnx model used was downloaded from https://hf-mirror.com/shibing624/text2vec-base-chinese/tree/main/onnx
I used the following code to get the number of inputs and outputs for the model, which shows three inputs: input_ids, attention_mask, and token_type_ids. How do I set the parameters of session.Run, especially inputTensor?

void printModelInfo(Ort::Session& session, Ort::AllocatorWithDefaultOptions& allocator)
{
	size_t num_input_nodes = session.GetInputCount();
	size_t num_output_nodes = session.GetOutputCount();
	cout << "Number of input node is:" << num_input_nodes << endl;
	cout << "Number of output node is:" << num_output_nodes << endl;

	for (auto i = 0; i < num_input_nodes; i++)
	{
		std::vector<int64_t> input_dims = session.GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape();
		cout << endl << "input " << i << " dim is: ";
		for (auto j = 0; j < input_dims.size(); j++)
			cout << input_dims[j] << " ";
	}
	for (auto i = 0; i < num_output_nodes; i++)
	{
		std::vector<int64_t> output_dims = session.GetOutputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape();
		cout << endl << "output " << i << " dim is: ";
		for (auto j = 0; j < output_dims.size(); j++)
			cout << output_dims[j] << " ";
	}

	cout << endl;
	for (auto i = 0; i < num_input_nodes; i++) 
        {
		Ort::AllocatedStringPtr input_name_Ptr = session.GetInputNameAllocated(i, allocator);
		cout << "The input op-name " << i << " is:" << input_name_Ptr.get() << endl;
	}

	for (auto i = 0; i < num_output_nodes; i++)
	{
		Ort::AllocatedStringPtr output_name_Ptr = session.GetOutputNameAllocated(i, allocator);
		cout << "The output op-name " << i << " is:" << output_name_Ptr.get() << endl;
	}
}

Ort::Env env(ORT_LOGGING_LEVEL_WARNING, "Txt2vecModel");
Ort::SessionOptions sessionOptions;

std::wstring modelPath = L"D:\\txt2vec\\m3e-base-onnx\\model.onnx";

Ort::Session session(env, modelPath.c_str(), sessionOptions);
Ort::AllocatorWithDefaultOptions allocator;

printModelInfo(session, allocator);

Putting a statement through the tokenizer of text2vec gets

input_ids: [[ 101 2644 1962  117 6468 6468 2644 8013  102]]
token_type_ids: [[0 0 0 0 0 0 0 0 0]]
attention_mask: [[1 1 1 1 1 1 1 1 1]]

I tried the following code to set up inputTensor and it compiled, but the runtime threw an error. I would like to know how to solve this problem, thank you very much！

std::vector<int64_t> inputShape = { 3, 9 };
Ort::MemoryInfo memoryInfo = Ort::MemoryInfo::CreateCpu(OrtAllocatorType::OrtArenaAllocator, OrtMemType::OrtMemTypeDefault);
std::vector<int64_t>  inputIds = { 101, 2644, 1962,  117, 6468, 6468, 2644, 8013,  102 , 0, 0, 0, 0, 0, 0, 0, 0, 0 , 1, 1, 1, 1, 1, 1, 1, 1, 1  };
Ort::Value inputTensor = Ort::Value::CreateTensor<int64_t>(memoryInfo, inputIds.data(), inputIds.size(), inputShape.data(), inputShape.size());

std::vector<const char*> inputNodeNames;
for (size_t i = 0; i < session.GetInputCount(); ++i) 
{
	Ort::AllocatorWithDefaultOptions allocator;
	Ort::AllocatedStringPtr input_name_Ptr = session.GetInputNameAllocated(i, allocator);
	inputNodeNames.push_back(input_name_Ptr.get());
}

std::vector<Ort::Value> outputTensors = session.Run(Ort::RunOptions{nullptr},
	inputNodeNames.data(),
	&inputTensor,
	3, 
	nullptr, 
	1); 

[Shyuna]

std::array<int64_t, 2> input_shape_{1, static_cast<int64_t>(input_ids.size())};

std::vector<Ort::Value> input_tensors;
input_tensors.push_back(Ort::Value::CreateTensor<int64_t>(
    memoryInfo, input_ids.data(), input_ids.size(), input_shape_.data(), input_shape_.size()));
input_tensors.push_back(
    Ort::Value::CreateTensor<int64_t>(memoryInfo, attention_mask.data(), attention_mask.size(),
                                      input_shape_.data(), input_shape_.size()));
input_tensors.push_back(
    Ort::Value::CreateTensor<int64_t>(memoryInfo, token_type_ids.data(), token_type_ids.size(),
                                      input_shape_.data(), input_shape_.size()));