From 668f555755602fc42a114b0469c956b33b8cb1e1 Mon Sep 17 00:00:00 2001
From: Eric Liang <ekhliang@gmail.com>
Date: Thu, 6 Aug 2020 18:29:04 -0700
Subject: [PATCH] [rllib] Clean up outdated docs #9915

---
 doc/source/rllib-toc.rst | 2 --
 doc/source/rllib.rst     | 4 ++--
 rllib/policy/policy.py   | 5 ++++-
 3 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/doc/source/rllib-toc.rst b/doc/source/rllib-toc.rst
index 4210d9a50..0dcbea76d 100644
--- a/doc/source/rllib-toc.rst
+++ b/doc/source/rllib-toc.rst
@@ -195,8 +195,6 @@ If you encounter errors like
 try setting ``OMP_NUM_THREADS=1``. Similarly, check configured system limits with
 `ulimit -a` for other resource limit errors.
 
-If you encounter out-of-memory errors, consider setting ``redis_max_memory`` and ``object_store_memory`` in ``ray.init()`` to reduce memory usage.
-
 For debugging unexpected hangs or performance problems, you can run ``ray stack`` to dump
 the stack traces of all Ray workers on the current node, ``ray timeline`` to dump
 a timeline visualization of tasks to a file, and ``ray memory`` to list all object
diff --git a/doc/source/rllib.rst b/doc/source/rllib.rst
index eb60390bc..0f7fd7cd1 100644
--- a/doc/source/rllib.rst
+++ b/doc/source/rllib.rst
@@ -38,7 +38,7 @@ Then, you can try out training in the following equivalent ways:
   from ray import tune
   from ray.rllib.agents.ppo import PPOTrainer
   tune.run(PPOTrainer, config={"env": "CartPole-v0"})  # "log_level": "INFO" for verbose,
-                                                       # "framework": "tfe" for tf-eager,
+                                                       # "framework": "tfe"/"tf2" for eager,
                                                        # "framework": "torch" for PyTorch
 
 Next, we'll cover three key concepts in RLlib: Policies, Samples, and Trainers.
@@ -92,7 +92,7 @@ Policies each define a ``learn_on_batch()`` method that improves the policy give
 
 - Simple `policy gradient loss <https://github.com/ray-project/ray/blob/master/rllib/agents/pg/pg_tf_policy.py>`__
 - Simple `Q-function loss <https://github.com/ray-project/ray/blob/a1d2e1762325cd34e14dc411666d63bb15d6eaf0/rllib/agents/dqn/simple_q_policy.py#L136>`__
-- Importance-weighted `APPO surrogate loss <https://github.com/ray-project/ray/blob/master/rllib/agents/ppo/appo_policy.py>`__
+- Importance-weighted `APPO surrogate loss <https://github.com/ray-project/ray/blob/master/rllib/agents/ppo/appo_torch_policy.py>`__
 
 RLlib `Trainer classes <rllib-concepts.html#trainers>`__ coordinate the distributed workflow of running rollouts and optimizing policies. They do this by leveraging Ray `parallel iterators <iter.html>`__ to implement the desired computation pattern. The following figure shows *synchronous sampling*, the simplest of `these patterns <rllib-algorithms.html>`__:
 
diff --git a/rllib/policy/policy.py b/rllib/policy/policy.py
index 580f0bf41..15380e2d4 100644
--- a/rllib/policy/policy.py
+++ b/rllib/policy/policy.py
@@ -36,7 +36,10 @@ class Policy(metaclass=ABCMeta):
     graphs and multi-GPU support.
 
     Attributes:
-        observation_space (gym.Space): Observation space of the policy.
+        observation_space (gym.Space): Observation space of the policy. For
+            complex spaces (e.g., Dict), this will be flattened version of the
+            space, and you can access the original space via
+            ``observation_space.original_space``.
         action_space (gym.Space): Action space of the policy.
         exploration (Exploration): The exploration object to use for
             computing actions, or None.