hot update A2C

projects/codes/DQN/main.py

@@ -24,6 +24,7 @@ def get_args():
     parser.add_argument('--env_name',default='CartPole-v0',type=str,help="name of environment")
     parser.add_argument('--train_eps',default=200,type=int,help="episodes of training")
     parser.add_argument('--test_eps',default=20,type=int,help="episodes of testing")
+    parser.add_argument('--ep_max_steps',default = 100000,type=int,help="steps per episode, much larger value can simulate infinite steps")
     parser.add_argument('--gamma',default=0.95,type=float,help="discounted factor")
     parser.add_argument('--epsilon_start',default=0.95,type=float,help="initial value of epsilon")
     parser.add_argument('--epsilon_end',default=0.01,type=float,help="final value of epsilon")
@@ -72,7 +73,7 @@ def train(cfg, env, agent):
         ep_reward = 0  # reward per episode
         ep_step = 0
         state = env.reset()  # reset and obtain initial state
-        while True:
+        for _ in range(cfg['ep_max_steps']):
             ep_step += 1
             action = agent.sample_action(state)  # sample action
             next_state, reward, done, _ = env.step(action)  # update env and return transitions
@@ -91,7 +92,7 @@ def train(cfg, env, agent):
             print(f'Episode: {i_ep+1}/{cfg["train_eps"]}, Reward: {ep_reward:.2f}: Epislon: {agent.epsilon:.3f}')
     print("Finish training!")
     env.close()
-    res_dic = {'episodes':range(len(rewards)),'rewards':rewards}
+    res_dic = {'episodes':range(len(rewards)),'rewards':rewards,'steps':steps}
     return res_dic
 
 def test(cfg, env, agent):
@@ -103,7 +104,7 @@ def test(cfg, env, agent):
         ep_reward = 0  # reward per episode
         ep_step = 0
         state = env.reset()  # reset and obtain initial state
-        while True:
+        for _ in range(cfg['ep_max_steps']):
             ep_step+=1
             action = agent.predict_action(state)  # predict action
             next_state, reward, done, _ = env.step(action)  
@@ -116,7 +117,7 @@ def test(cfg, env, agent):
         print(f"Episode: {i_ep+1}/{cfg['test_eps']}，Reward: {ep_reward:.2f}")
     print("Finish testing!")
     env.close()
-    return {'episodes':range(len(rewards)),'rewards':rewards}
+    return {'episodes':range(len(rewards)),'rewards':rewards,'steps':steps}
 
 
 if __name__ == "__main__":

projects/codes/DQN/outputs/CartPole-v1/20220828-214702/results/params.json

@@ -0,0 +1 @@
+{"algo_name": "DQN", "env_name": "CartPole-v1", "train_eps": 2000, "test_eps": 20, "ep_max_steps": 100000, "gamma": 0.99, "epsilon_start": 0.95, "epsilon_end": 0.01, "epsilon_decay": 6000, "lr": 1e-05, "memory_capacity": 200000, "batch_size": 64, "target_update": 4, "hidden_dim": 256, "device": "cuda", "seed": 10, "show_fig": false, "save_fig": true, "result_path": "C:\\Users\\24438\\Desktop\\rl-tutorials\\codes\\DQN/outputs/CartPole-v1/20220828-214702/results", "model_path": "C:\\Users\\24438\\Desktop\\rl-tutorials\\codes\\DQN/outputs/CartPole-v1/20220828-214702/models", "n_states": 4, "n_actions": 2}

projects/codes/DQN/outputs/CartPole-v1/20220828-214702/results/testing_results.csv

@@ -0,0 +1,21 @@
+episodes,rewards,steps
+0,371.0,371
+1,446.0,446
+2,300.0,300
+3,500.0,500
+4,313.0,313
+5,500.0,500
+6,341.0,341
+7,489.0,489
+8,304.0,304
+9,358.0,358
+10,278.0,278
+11,500.0,500
+12,500.0,500
+13,500.0,500
+14,500.0,500
+15,476.0,476
+16,308.0,308
+17,394.0,394
+18,500.0,500
+19,500.0,500