上述代码要改成c++

把上述Python代码翻译成C++的代码：

#include<iostream>
#include<fstream>
#include<vector>
#include<string>
#include<torch/torch.h>
#include<torch/script.h>

int main() {
    std::string checkpoint_path = "chinese_gpt.pth";
    std::string vocab_path = "vocab.txt";
    std::string sentence = "我喜欢";
    int max_len = 20;

    // 读取词表
    std::vector<std::string> vocab;
    std::ifstream fin(vocab_path);
    std::string line;
    while (std::getline(fin, line)) {
        vocab.push_back(line);
    }

    // 加载模型
    torch::jit::script::Module module = torch::jit::load(checkpoint_path);

    // 输入处理
    std::vector<int64_t> input_ids;
    for (char c : sentence) {
        auto it = std::find(vocab.begin(), vocab.end(), std::string(1, c));
        int64_t id = 0;
        if (it != vocab.end()) {
            id = std::distance(vocab.begin(), it);
        }
        input_ids.push_back(id);
    }

    // 生成文本
    std::vector<int64_t> output_ids;
    output_ids.reserve(max_len);
    auto input_tensor = torch::from_blob(input_ids.data(), {1, input_ids.size()}, torch::kInt64).clone();
    for (int i = 0; i < max_len; ++i) {
        std::vector<torch::jit::IValue> inputs;
        inputs.push_back(input_tensor);
        auto outputs = module.forward(inputs).toTensor();
        int64_t predicted_id = outputs[0][input_ids.size()-1].argmax().item<int64_t>();
        if (predicted_id == 2) {  // <EOS>
            break;
        }
        output_ids.push_back(predicted_id);
        input_ids.push_back(predicted_id);
        input_tensor = torch::from_blob(input_ids.data(), {1, input_ids.size()}, torch::kInt64).clone();
    }

    // 输出结果
    std::string output_text;
    for (int64_t id : output_ids) {
        output_text += vocab[id];
    }
    std::cout << output_text << std::endl;

    return 0;
}

注意：上述代码使用了PyTorch C++接口，请确保已经正确安装了PyTorch库。