How to optimize LSTM with Multi objective dragon fly algorithm?

Hi every one.

I have a question about optimization. I made a prediction with LSTM that does not have good accuracy and I want to increase the prediction accuracy by optimizing the weight and bias of LSTM with the multi-objective dragonfly algorithm. But I do not know how to do this. Do I have to give the trained network weights and bias as input to the MODA after training the network?If I have to do this, in which parameter should the weights and bias be placed in the MODA, and after optimization, how should I use weights and bias? I downloaded the dragonfly code from the "MODA: Multi-Objective Dragonfly Algorithm - File Exchange - MATLAB Central (mathworks.com)" but I do not know how to use this code to optimize the forecasting. please guide me. Thankful.