Add 10 million timestep model, improve README, add diffculty/mode selector, score, make model training have instant graphs and also multiple envs for faster training, better plotting, improve RL model by including multiple players, better reard system, use EnemyFormation instead of single Enemy-es

README.md

@@ -2,7 +2,9 @@ Fleet Commander is like Space Invaders but you are the enemy instead of the play
 
 It uses AI (Reinforcement Learning) for the Player, and you, the Enemy has to defeat it.
 
-You can train yourself, or use the default model which comes with the game.
+I know the game is too easy and is too simple, but please understand that doing RL isnt the easiest thing ever. I also did this so late so yeah.
+
+You can train yourself, or use the default model(10 million timesteps) which comes with the game.
 
 # Install steps:
 
@@ -18,4 +20,7 @@ You can train yourself, or use the default model which comes with the game.
 - `pip3 install -r requirements.txt`
 - `pip3 install torch --index-url https://download.pytorch.org/whl/cpu`
 - `pip3 install stable_baselines3`
-- `python3 run.py`
+- `python3 run.py`
+
+# Disclaimer
+AI assistance was used in this project, since i never did any RL work before. But every instance of AI code was heavily modified by me.