mpi-less baselines (#689)

* make baselines run without mpi wip

* squash-merged latest master

* further removing MPI references where unnecessary

* more MPI removal

* syntax and flake8

* MpiAdam becomes regular Adam if Mpi not present

* autopep8

* add assertion to test in mpi_adam; fix trpo_mpi failure without MPI on cartpole

* mpiless ddpg

baselines/ddpg/ddpg.py

@@ -12,8 +12,11 @@ import baselines.common.tf_util as U
 
 from baselines import logger
 import numpy as np
-from mpi4py import MPI
 
+try:
+    from mpi4py import MPI
+except ImportError:
+    MPI = None
 
 def learn(network, env,
           seed=None,
@@ -49,7 +52,11 @@ def learn(network, env,
     else:
         nb_epochs = 500
 
-    rank = MPI.COMM_WORLD.Get_rank()
+    if MPI is not None:
+        rank = MPI.COMM_WORLD.Get_rank()
+    else:
+        rank = 0
+
     nb_actions = env.action_space.shape[-1]
     assert (np.abs(env.action_space.low) == env.action_space.high).all()  # we assume symmetric actions.
 
@@ -199,7 +206,11 @@ def learn(network, env,
                             eval_episode_rewards_history.append(eval_episode_reward[d])
                             eval_episode_reward[d] = 0.0
 
-        mpi_size = MPI.COMM_WORLD.Get_size()
+        if MPI is not None:
+            mpi_size = MPI.COMM_WORLD.Get_size()
+        else:
+            mpi_size = 1
+
         # Log stats.
         # XXX shouldn't call np.mean on variable length lists
         duration = time.time() - start_time
@@ -233,7 +244,10 @@ def learn(network, env,
             else:
                 raise ValueError('expected scalar, got %s'%x)
 
-        combined_stats_sums = MPI.COMM_WORLD.allreduce(np.array([ np.array(x).flatten()[0] for x in combined_stats.values()]))
+        combined_stats_sums = np.array([ np.array(x).flatten()[0] for x in combined_stats.values()])
+        if MPI is not None:
+            combined_stats_sums = MPI.COMM_WORLD.allreduce(combined_stats_sums)
+
         combined_stats = {k : v / mpi_size for (k,v) in zip(combined_stats.keys(), combined_stats_sums)}
 
         # Total statistics.

baselines/ddpg/ddpg_learner.py

@@ -9,7 +9,10 @@ from baselines import logger
 from baselines.common.mpi_adam import MpiAdam
 import baselines.common.tf_util as U
 from baselines.common.mpi_running_mean_std import RunningMeanStd
-from mpi4py import MPI
+try:
+    from mpi4py import MPI
+except ImportError:
+    MPI = None
 
 def normalize(x, stats):
     if stats is None:
@@ -358,6 +361,11 @@ class DDPG(object):
         return stats
 
     def adapt_param_noise(self):
+        try:
+            from mpi4py import MPI
+        except ImportError:
+            MPI = None
+
         if self.param_noise is None:
             return 0.
 
@@ -371,7 +379,16 @@ class DDPG(object):
             self.param_noise_stddev: self.param_noise.current_stddev,
         })
 
-        mean_distance = MPI.COMM_WORLD.allreduce(distance, op=MPI.SUM) / MPI.COMM_WORLD.Get_size()
+        if MPI is not None:
+            mean_distance = MPI.COMM_WORLD.allreduce(distance, op=MPI.SUM) / MPI.COMM_WORLD.Get_size()
+        else:
+            mean_distance = distance
+
+        if MPI is not None:
+            mean_distance = MPI.COMM_WORLD.allreduce(distance, op=MPI.SUM) / MPI.COMM_WORLD.Get_size()
+        else:
+            mean_distance = distance
+
         self.param_noise.adapt(mean_distance)
         return mean_distance