update mujoco and tensorflow 1.14/2.0 api calls in efficient-hrl (#6525)

* updated mujoco api calls to data, and tensorflow call to CriticalSection (moved from contrib.framework)

* changed efficient-hrl envs so self.physics points to correct PyMj object based on mujoco_py version

* corrected mujoco version number checkm in AntEnv.physics and PointEnv.physics properties

* fix AntEnv.set_xy by reverting to using self.physics

research/efficient-hrl/agents/circular_buffer.py

@@ -47,7 +47,7 @@ class CircularBuffer(object):
     self._tensors = collections.OrderedDict()
     with tf.variable_scope(self._scope):
       self._num_adds = tf.Variable(0, dtype=tf.int64, name='num_adds')
-    self._num_adds_cs = tf.contrib.framework.CriticalSection(name='num_adds')
+    self._num_adds_cs = tf.CriticalSection(name='num_adds')
 
   @property
   def buffer_size(self):

research/efficient-hrl/environments/ant.py

@@ -17,6 +17,7 @@
 
 import math
 import numpy as np
+import mujoco_py
 from gym import utils
 from gym.envs.mujoco import mujoco_env
 
@@ -50,7 +51,13 @@ class AntEnv(mujoco_env.MujocoEnv, utils.EzPickle):
 
   @property
   def physics(self):
-    return self.model
+    # check mujoco version is greater than version 1.50 to call correct physics
+    # model containing PyMjData object for getting and setting position/velocity
+    # check https://github.com/openai/mujoco-py/issues/80 for updates to api
+    if mujoco_py.get_version() >= '1.50':
+      return self.sim
+    else:
+      return self.model
 
   def _step(self, a):
     return self.step(a)
@@ -117,7 +124,7 @@ class AntEnv(mujoco_env.MujocoEnv, utils.EzPickle):
 
   def get_ori(self):
     ori = [0, 1, 0, 0]
-    rot = self.model.data.qpos[self.__class__.ORI_IND:self.__class__.ORI_IND + 4]  # take the quaternion
+    rot = self.physics.data.qpos[self.__class__.ORI_IND:self.__class__.ORI_IND + 4]  # take the quaternion
     ori = q_mult(q_mult(rot, ori), q_inv(rot))[1:3]  # project onto x-y plane
     ori = math.atan2(ori[1], ori[0])
     return ori

research/efficient-hrl/environments/maze_env.py

@@ -228,7 +228,7 @@ class MazeEnv(gym.Env):
         raise Exception("Every geom of the torso must have a name "
                         "defined")
 
-    _, file_path = tempfile.mkstemp(text=True)
+    _, file_path = tempfile.mkstemp(text=True, suffix='.xml')
     tree.write(file_path)
 
     self.wrapped_env = model_cls(*args, file_path=file_path, **kwargs)

research/efficient-hrl/environments/point.py

@@ -17,6 +17,7 @@
 
 import math
 import numpy as np
+import mujoco_py
 from gym import utils
 from gym.envs.mujoco import mujoco_env
 
@@ -33,7 +34,13 @@ class PointEnv(mujoco_env.MujocoEnv, utils.EzPickle):
 
   @property
   def physics(self):
-    return self.model
+    # check mujoco version is greater than version 1.50 to call correct physics
+    # model containing PyMjData object for getting and setting position/velocity
+    # check https://github.com/openai/mujoco-py/issues/80 for updates to api
+    if mujoco_py.get_version() >= '1.50':  
+      return self.sim
+    else:
+      return self.model
 
   def _step(self, a):
     return self.step(a)
@@ -80,7 +87,7 @@ class PointEnv(mujoco_env.MujocoEnv, utils.EzPickle):
     return self._get_obs()
 
   def get_ori(self):
-    return self.model.data.qpos[self.__class__.ORI_IND]
+    return self.physics.data.qpos[self.__class__.ORI_IND]
 
   def set_xy(self, xy):
     qpos = np.copy(self.physics.data.qpos)

research/efficient-hrl/scripts/local_train.py

@@ -28,7 +28,7 @@ CONFIGS_PATH = './configs'
 CONTEXT_CONFIGS_PATH = './context/configs'
 
 def main():
-  bb = './'
+  bb = '.'
   base_num_args = 6
   if len(sys.argv) < base_num_args:
     print(