How the rewards are incorporated into the learning process?

I have an environment in which I am trying to learn the system dynamics using baseline3 library. Since my example is quite big, I am using the following example to clarify the problem I have.

The problems

1. Where actually environment reword incorporate into the learning process?

2. Is there any upper and lower bound for the rewards?

My code:

from stable_baselines.common.policies import MlpPolicy
from stable_baselines.common.vec_env import DummyVecEnv
from stable_baselines import PPO2

env = gym.make('CartPole-v1')
   

model = PPO2(MlpPolicy, env, verbose=1)
model.learn(total_timesteps=10000)

obs = env.reset()
for i in range(1000):
    action, _states = model.predict(obs)
    obs, rewards, dones, info = env.step(action)
    env.render()

I could not find any information regarding this yet.