Complements requirements.txt and add target-net to DQN

[GiriYomi]

1. pip environment freeze txt file is now complement with 99% of packs we are using.
   Some pack, for example: pickle, will run into error on python 3.10 on windows since it already exist with python from python 3.8
   After this project is complete we can add a venv folder and bash/bat file for 1 key run our project demo. But thing have to be done after wrapping up the project to prevent frequently update venv.

2. DQN will have both policy net and target net after this update which make DQN algorithm structure correct. However on our maze demo it still won't converge. Further in-variable value debugging and baseline (casual way of RL) will be needed.