env_id:  "overcooked"
policy:  "MlpPolicy"
model_name: "PPO"
model_type:
  _partial_: true
  _target_: stable_baselines3.PPO
model_type_inference:
  _partial_: true
  _target_: stable_baselines3.PPO.load
total_timesteps:  3_000_000 # hendric sagt eher so 300_000_000 schritte
number_envs_parallel:  16
learning_rate: 0.0003
n_steps: 2048
batch_size: 16
n_epochs: 10
gamma: 0.99
gae_lambda: 0.95
clip_range: 0.2
clip_range_vf:  None
normalize_advantage: True
ent_coef: 0.0
vf_coef: 0.5
max_grad_norm: 0.5
use_sde: False
sde_sample_freq: -1
rollout_buffer_class: None
rollout_buffer_kwargs: None
target_kl: None
stats_window_size: 100
tensorboard_log: f"logs/reinforcement_learning/runs/{0}"
policy_kwargs: None
verbose: 0
seed: None
device: "auto"
_init_setup_model: True