Remove default model i will add back later, add back train.py, fix some rewarding

train.py

@@ -0,0 +1,30 @@
+from stable_baselines3 import PPO
+from utils.rl import SpaceInvadersEnv
+from stable_baselines3.common.vec_env import DummyVecEnv
+
+def make_env(rank: int, seed: int = 0):
+    def _init():
+        env = SpaceInvadersEnv()
+        return env
+    return _init
+
+env = SpaceInvadersEnv()
+
+n_envs = 128
+
+env = DummyVecEnv([make_env(i) for i in range(n_envs)])
+model = PPO(
+    "MlpPolicy", 
+    env, 
+    n_steps=8192,
+    batch_size=256,
+    n_epochs=7,
+    learning_rate=0.001,
+    verbose=1, 
+    device="cpu", 
+    gamma=0.985, 
+    ent_coef=0.015,
+    clip_range=0.2,
+)
+model.learn(75_000_000)
+model.save("invader_agent")