[rllib] MAML Agent (#8862)

* Halfway done with transferring MAML to new Ray

* MAML Beta Out

* Debugging MAML atm

* Distributed Execution

* Pendulum Mass Working

* All experiments complete

* Cleaned up codebase

* Travis CI

* Travis CI

* Tests

* Merged conflicts

* Fixed variance bug conflict

* Comment resolved

* Apply suggestions from code review

fixed test_maml

* Update rllib/agents/maml/tests/test_maml.py

* asdf

* Fix testing

Co-authored-by: Sven Mika <sven@anyscale.io>

rllib/examples/env/pendulum_mass.py

@@ -0,0 +1,28 @@
+import numpy as np
+import gym
+from gym.envs.classic_control.pendulum import PendulumEnv
+
+
+class PendulumMassEnv(PendulumEnv, gym.utils.EzPickle):
+    """PendulumMassEnv varies the weight of the pendulum
+
+    Tasks are defined to be weight uniformly sampled between [0.5,2]
+    """
+
+    def sample_tasks(self, n_tasks):
+        # Mass is a random float between 0.5 and 2
+        return np.random.uniform(low=0.5, high=2.0, size=(n_tasks, ))
+
+    def set_task(self, task):
+        """
+        Args:
+            task: task of the meta-learning environment
+        """
+        self.m = task
+
+    def get_task(self):
+        """
+        Returns:
+            task: task of the meta-learning environment
+        """
+        return self.m