new hparams

environments/factory/base_factory.py

@@ -265,7 +265,6 @@ class BaseFactory(gym.Env):
         # d = {key: val._asdict() if hasattr(val, '_asdict') else val for key, val in self.__dict__.items()
         d = {key: val for key, val in self.__dict__.items() if not key.startswith('_') and not key.startswith('__')}
         filepath.parent.mkdir(parents=True, exist_ok=True)
-        super(BaseFactory, self).save_params()
         with filepath.open('w') as f:
             yaml.dump(d, f)
             # pickle.dump(d, f, protocol=pickle.HIGHEST_PROTOCOL)

environments/factory/simple_factory.py

@@ -192,11 +192,6 @@ if __name__ == '__main__':
     factory = SimpleFactory(movement_properties=move_props, dirt_properties=dirt_props, n_agents=10,
                             combin_agent_slices_in_obs=True, omit_agent_slice_in_obs=False, level_name='rooms')
 
-    # dirt_props = DirtProperties()
-    # move_props = MovementProperties(allow_diagonal_movement=False, allow_no_op=False)
-    # factory = SimpleFactory(n_agents=2, dirt_properties=dirt_props, movement_properties=move_props, level='rooms',
-    #                         pomdp_radius=2)
-
     n_actions = factory.action_space.n - 1
 
     for epoch in range(100):

environments/logging/plotting.py

@@ -41,6 +41,6 @@ def prepare_plot(filepath, results_df, ext='png', hue='Measurement', style=None)
         plt.close('all')
         sns.set(rc={'text.usetex': False}, style='whitegrid')
         lineplot = sns.lineplot(data=df, x='Episode', y='Score', hue=hue, style=style,
-                                ci=95, palette=PALETTE, hue_order=hue_order)
+                                     ci=95, palette=PALETTE, hue_order=hue_order)
         lineplot.set_title(f'{sorted(list(df["Measurement"].unique()))}')
         plot(filepath, ext=ext)

main.py

@@ -101,14 +101,20 @@ if __name__ == '__main__':
 
     out_path = None
 
-    for modeL_type in [PPO, A2C]:  # , RegDQN, DQN]:
+    for modeL_type in [A2C, PPO, RegDQN, DQN]:  # , QRDQN]:
         for seed in range(3):
 
             with SimpleFactory(n_agents=1, dirt_properties=dirt_props, pomdp_radius=2, max_steps=400,
                                movement_properties=move_props, level_name='rooms', frames_to_stack=4,
                                omit_agent_slice_in_obs=False, combin_agent_slices_in_obs=True) as env:
 
-                kwargs = dict(ent_coef=0.01) if isinstance(modeL_type, (PPO, A2C)) else {}
+                if modeL_type.__name__ in ["PPO", "A2C"]:
+                    kwargs = dict(ent_coef=0.01)
+                elif modeL_type.__name__ in ["RegDQN", "DQN", "QRDQN"]:
+                    kwargs = dict(target_update_interval=500, buffer_size=30000, learning_starts=5000,
+                                  exploration_final_eps=0.01, batch_size=96)
+                else:
+                    raise NameError(f'The model "{model.__name__}" has the wrong name.')
                 model = modeL_type("MlpPolicy", env, verbose=1, seed=seed, device='cpu', **kwargs)
 
                 out_path = Path('debug_out') / f'{model.__class__.__name__}_{time_stamp}'