ui export

main.py

@@ -64,7 +64,7 @@ env_config = {
     'allow_observation_scaling': True,
     'dense_log_frequency': 100, 
     'world_dense_log_frequency':1,
-    'energy_cost':0,
+    'energy_cost':0.2,
     'energy_warmup_method': "auto",
     'energy_warmup_constant': 4000,
     
@@ -120,7 +120,7 @@ eval_env_config = {
     'isoelastic_eta':0.001,
     'dense_log_frequency': 1, 
     'world_dense_log_frequency':1, 
-    'energy_cost':0,
+    'energy_cost':0.2,
     'energy_warmup_method': "auto",
     'energy_warmup_constant': 4000,
     
@@ -138,6 +138,7 @@ eval_env_config = {
 }
 
 num_frames=1
+CONNECTION_STRING = "mongodb+srv://root:econ@localhost:/econ"
 
 class TensorboardCallback(BaseCallback):
     """
@@ -292,8 +293,8 @@ runname="run_{}".format(run_number)
 model_db=[None,None] # object for storing model
 
 
-model = MaskablePPO("MlpPolicy",n_steps=int(env_config['episode_length']*2),ent_coef=0.001, vf_coef=0.5 ,gamma=0.99, learning_rate=learning_rate_adj_basic,env=stackenv_basic, seed=445,verbose=1,device="cuda",tensorboard_log="./log")
-model_trade=MaskablePPO("MlpPolicy",n_steps=int(env_config['episode_length']*2),ent_coef=0.001, vf_coef=0.5 ,gamma=0.99, learning_rate=learning_rate_adj_trade,env=stackenv_traid, seed=445,verbose=1,device="cuda",tensorboard_log="./log")
+model = MaskablePPO("MlpPolicy",n_steps=int(env_config['episode_length']*2),ent_coef=0.001, vf_coef=0.5 ,gamma=0.99, learning_rate=4e-4,env=stackenv_basic, seed=500,verbose=1,device="cuda",tensorboard_log="./log")
+model_trade=MaskablePPO("MlpPolicy",n_steps=int(env_config['episode_length']*2),ent_coef=0.001, vf_coef=0.5 ,gamma=0.99, learning_rate=4e-4,env=stackenv_traid, seed=500,verbose=1,device="cuda",tensorboard_log="./log")
 
 n_agents=econ.n_agents
 
@@ -308,9 +309,9 @@ while True:
     #Train
     runname="run_{}_{}".format(run_number,"basic")
 
-    thread_model=Thread(target=train,args=(model,total_required_for_episode_basic*150,econ,True,runname,model_db,0))
+    thread_model=Thread(target=train,args=(model,total_required_for_episode_basic*50,econ,True,runname,model_db,0))
     runname="run_{}_{}".format(run_number,"trader")
-    thread_model_traid=Thread(target=train,args=(model_trade,total_required_for_episode_traid*150,econ,False,runname,model_db,1))
+    thread_model_traid=Thread(target=train,args=(model_trade,total_required_for_episode_traid*50,econ,False,runname,model_db,1))
     
     thread_model.start()
     thread_model_traid.start()