Use Libtorch and C++ to add features to the existing

My coworker and I are developing a nonlinear regression model (Mixed frequency sampling if you want to search for more detail) in C++. He used the Libtorch to optimize model parameters but he can't proceed the work because of family reasons. Right now we have finished the data preparation, parameter optimization but we still need need to use the optimized model parameter to calculate regression coefficient, standard errors, t-stats, pvalue, aic/bic, MSE and check the model forecast on a test dataset. I have written the calculation formulas in Python+Pytorch. I need someone to first read the existing C++ code structure and then add more functions to calculate the regression statistics. Also output all the results. I can provide the existing C++/Libtorch code and my Python/pytorch code per requested.