Updated sweep

reinforcement_learning/multi_agent_training.py

@@ -488,7 +488,6 @@ if __name__ == "__main__":
         "observation_max_path_depth": 30
     }
 
-
     def check_env_config(id):
         if id >= len(env_params) or id < 0:
             print("\n🛑 Invalid environment configuration, only Test_0 to Test_{} are supported.".format(len(env_params) - 1))

sweep.yaml

@@ -5,16 +5,15 @@ metric:
     goal: maximize
 parameters:
     n_episodes:
-        values: [1500]
-    learning_rate:
-        # default: 0.00005 (rllib: same)
-        values: [0.000040, 0.000045, 0.000050, 0.000055, 0.000060]
+        values: [2000]
     hidden_size:
-        # default: 128
+        # default: 256
         values: [128, 256, 512]
     buffer_size:
-        # default: 100000
+        # default: 50000
         values: [50000, 100000, 500000, 1000000]
     batch_size:
-        # default: 512 (rllib: 32)
-        values: [32, 64, 128, 256]
\ No newline at end of file
+        # default: 32
+        values: [16, 32, 64, 128]
+    training_env_config:
+        values: [0, 1, 2]
\ No newline at end of file