bugfix for envs with max_action < 1

Author: sfujim

main.py

@@ -67,7 +67,7 @@ def evaluate_policy(policy, eval_episodes=10):
 
 	state_dim = env.observation_space.shape[0]
 	action_dim = env.action_space.shape[0] 
-	max_action = int(env.action_space.high[0])
+	max_action = float(env.action_space.high[0])
 
 	# Initialize policy
 	if args.policy_name == "TD3": policy = TD3.TD3(state_dim, action_dim, max_action)