reduce session length

tasks/rl_hopper.py

@@ -258,7 +258,7 @@ def prepare_data_tuples(states, actions, rewards, num_layers, skip_steps):
                 actions = []
                 rewards = []
 
-                for j in range(1000):
+                for j in range(200):
                     if random.random() <= epsilon or course == 0:
                         selected_action = env.action_space.sample()
                         # quantize to -1 0 1