[tune] Support user-defined trainable functions / classes / envs with a shared object registry (#1226)

doc/source/conf.py

@@ -19,10 +19,19 @@ import shlex
 # These lines added to enable Sphinx to work without installing Ray.
 import mock
 MOCK_MODULES = ["gym",
+                "gym.spaces",
+                "scipy",
+                "scipy.signal",
                 "tensorflow",
                 "tensorflow.contrib",
+                "tensorflow.contrib.layers",
                 "tensorflow.contrib.slim",
                 "tensorflow.contrib.rnn",
+                "tensorflow.core",
+                "tensorflow.core.util",
+                "tensorflow.python",
+                "tensorflow.python.client",
+                "tensorflow.python.util",
                 "pyarrow",
                 "pyarrow.plasma",
                 "smart_open",

doc/source/example-a3c.rst

@@ -25,7 +25,7 @@ You can run the code with
 
 .. code-block:: bash
 
-  python/ray/rllib/train.py --env=Pong-ram-v4 --alg=A3C --config='{"num_workers": N}'
+  python/ray/rllib/train.py --env=Pong-ram-v4 --run=A3C --config='{"num_workers": N}'
 
 Reinforcement Learning
 ----------------------

doc/source/example-evolution-strategies.rst

@@ -18,7 +18,7 @@ on the ``Humanoid-v1`` gym environment.
 
 .. code-block:: bash
 
-  python/ray/rllib/train.py --env=Humanoid-v1 --alg=ES
+  python/ray/rllib/train.py --env=Humanoid-v1 --run=ES
 
 To train a policy on a cluster (e.g., using 900 workers), run the following.
 
@@ -26,7 +26,7 @@ To train a policy on a cluster (e.g., using 900 workers), run the following.
 
   python ray/python/ray/rllib/train.py \
       --env=Humanoid-v1 \
-      --alg=ES \
+      --run=ES \
       --redis-address=<redis-address> \
       --config='{"num_workers": 900, "episodes_per_batch": 10000, "timesteps_per_batch": 100000}'
 

doc/source/example-policy-gradient.rst

@@ -16,7 +16,7 @@ Then you can run the example as follows.
 
 .. code-block:: bash
 
-  python/ray/rllib/train.py --env=Pong-ram-v4 --alg=PPO
+  python/ray/rllib/train.py --env=Pong-ram-v4 --run=PPO
 
 This will train an agent on the ``Pong-ram-v4`` Atari environment. You can also
 try passing in the ``Pong-v0`` environment or the ``CartPole-v0`` environment.

doc/source/rllib.rst

@@ -30,7 +30,7 @@ You can run training with
 
 ::
 
-    python ray/python/ray/rllib/train.py --env CartPole-v0 --alg PPO --config '{"timesteps_per_batch": 10000}'
+    python ray/python/ray/rllib/train.py --env CartPole-v0 --run PPO --config '{"timesteps_per_batch": 10000}'
 
 By default, the results will be logged to a subdirectory of ``/tmp/ray``.
 This subdirectory will contain a file ``config.json`` which contains the
@@ -51,7 +51,7 @@ The ``train.py`` script has a number of options you can show by running
 
 The most important options are for choosing the environment
 with ``--env`` (any OpenAI gym environment including ones registered by the user
-can be used) and for choosing the algorithm with ``--alg``
+can be used) and for choosing the algorithm with ``-run``
 (available options are ``PPO``, ``A3C``, ``ES`` and ``DQN``). Each algorithm
 has specific hyperparameters that can be set with ``--config``, see the
 ``DEFAULT_CONFIG`` variable in