change 2 dqn

chris-chris · Nov 3, 2017 · ef948f9 · ef948f9
1 parent 90f5151
commit ef948f9
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 8 deletions.
diff --git a/deepq_mineral_shards.py b/deepq_mineral_shards.py
@@ -257,7 +257,7 @@ def make_obs_ph(name):
 
   player_relative = obs[0].observation["screen"][_PLAYER_RELATIVE]
 
-  screen = player_relative #+ path_memory
+  screen = (player_relative == _PLAYER_NEUTRAL).astype(int) #+ path_memory
 
   player_y, player_x = (player_relative == _PLAYER_FRIENDLY).nonzero()
   player = [int(player_x.mean()), int(player_y.mean())]
@@ -314,7 +314,7 @@ def make_obs_ph(name):
       obs = env.step(actions=new_action)
 
       player_relative = obs[0].observation["screen"][_PLAYER_RELATIVE]
-      new_screen = player_relative #+ path_memory
+      new_screen = (player_relative == _PLAYER_NEUTRAL).astype(int)
 
       player_y, player_x = (player_relative == _PLAYER_FRIENDLY).nonzero()
       player = [int(player_x.mean()), int(player_y.mean())]
@@ -335,8 +335,7 @@ def make_obs_ph(name):
       if done:
         obs = env.reset()
         player_relative = obs[0].observation["screen"][_PLAYER_RELATIVE]
-
-        screen = player_relative #+ path_memory
+        screent = (player_relative == _PLAYER_NEUTRAL).astype(int)
 
         player_y, player_x = (player_relative == _PLAYER_FRIENDLY).nonzero()
         player = [int(player_x.mean()), int(player_y.mean())]

diff --git a/train_mineral_shards.py b/train_mineral_shards.py
@@ -119,11 +119,11 @@ def main():
       act = deepq_mineral_shards.learn(
         env,
         q_func=model,
-        num_actions=64,
-        lr=1e-3,
-        max_timesteps=20000000,
+        num_actions=32,
+        lr=FLAGS.lr,
+        max_timesteps=FLAGS.timesteps,
         buffer_size=10000,
-        exploration_fraction=0.5,
+        exploration_fraction=FLAGS.exploration_fraction,
         exploration_final_eps=0.01,
         train_freq=4,
         learning_starts=10000,