diff --git a/README.md b/README.md
index f1e9cf3..4f20b6e 100644
--- a/README.md
+++ b/README.md
@@ -7,13 +7,12 @@ Implemented algorithms:
 * Deep Q-Learning (DQN)
 * Double DQN
 * Dueling DQN
-* Async Advantage Actor Critic (A3C)
+* (Async) Advantage Actor Critic (A3C / A2C)
 * Async One-Step Q-Learning
 * Async One-Step Sarsa 
 * Async N-Step Q-Learning
 * Continuous A3C
 * Distributed Deep Deterministic Policy Gradient (Distributed DDPG, aka D3PG)
-* Hybrid Reward Architecture (HRA)
 * Parallelized Proximal Policy Optimization (P3O, similar to DPPO)
 * Action Conditional Video Prediction
 
diff --git a/agent/A2C_agent.py b/agent/A2C_agent.py
index 1d7509f..4b7aa6d 100644
--- a/agent/A2C_agent.py
+++ b/agent/A2C_agent.py
@@ -100,7 +100,7 @@ class A2CAgent:
         value_loss = 0.5 * (Variable(returns) - value).pow(2)
 
         self.optimizer.zero_grad()
-        (policy_loss + value_loss).mean().backward()
+        (policy_loss + value_loss).sum().backward()
         nn.utils.clip_grad_norm(self.network.parameters(), config.gradient_clip)
         self.optimizer.step()
 
diff --git a/main.py b/main.py
index e15ac5b..8aa9b89 100644
--- a/main.py
+++ b/main.py
@@ -308,13 +308,14 @@ if __name__ == '__main__':
     mkdir('data/video')
     mkdir('log')
     os.system('export OMP_NUM_THREADS=1')
+    os.system('export CUDA_VISIBLE_DEVICES=0')
     # logger.setLevel(logging.DEBUG)
     logger.setLevel(logging.INFO)
 
-    dqn_cart_pole()
+    # dqn_cart_pole()
     # async_cart_pole()
     # a3c_cart_pole()
-    # a2c_cart_pole()
+    a2c_cart_pole()
     # a3c_continuous()
     # p3o_continuous()
     # d3pg_continuous()