[Doc] Clean up remaining RLlib broken links (#46102)

## Why are these changes needed?

This PR cleans up the last remaining broken links in the RLlib
documentation.

## Related issue number

Partially completes #39658.

## Checks

- [x] I've signed off every commit(by using the -s flag, i.e., `git
commit -s`) in this PR.
- [x] I've run `scripts/format.sh` to lint the changes in this PR.
- [x] I've included any doc changes needed for
https://docs.ray.io/en/master/.
- [x] I've added any new APIs to the API Reference. For example, if I
added a
method in Tune, I've added it in `doc/source/tune/api/` under the
           corresponding `.rst` file.
- [x] I've made sure the tests are passing. Note that there might be a
few flaky tests, see the recent failures at https://flakey-tests.ray.io/
- Testing Strategy
   - [x] Unit tests
   - [ ] Release tests
   - [ ] This PR is not tested :(

Signed-off-by: pdmurray <peynmurray@gmail.com>

doc/source/rllib/package_ref/rl_modules.rst

@@ -82,6 +82,9 @@ Forward methods
     ~RLModule.forward_train
     ~RLModule.forward_exploration
     ~RLModule.forward_inference
+    ~RLModule._forward_train
+    ~RLModule._forward_exploration
+    ~RLModule._forward_inference
 
 IO specifications
 +++++++++++++++++

doc/source/rllib/rllib-advanced-api.rst

@@ -42,7 +42,7 @@ are represented by slightly different maps that the agent has to navigate.
    :end-before: __END_curriculum_learning_example_env_options__
 
 Then, define the central piece controlling the curriculum, which is a custom callbacks class
-overriding the :py:meth:`~ray.rllib.algorithms.callbacks.Callbacks.on_train_result`.
+overriding the :py:meth:`~ray.rllib.algorithms.callbacks.DefaultCallbacks.on_train_result`.
 
 
 .. TODO move to doc_code and make it use algo configs.

doc/source/rllib/rllib-examples.rst

@@ -64,9 +64,8 @@ Algorithms
 ----------
 
 - |new_stack| `How to write a custom Algorith.training_step() method combining on- and off-policy training <https://github.com/ray-project/ray/blob/master/rllib/examples/algorithms/custom_training_step_on_and_off_policy_combined.py>`__:
-   Example of how to override the :py:meth:`~ray.rllib.algorithms.algorithm.training_step` method of the
-   :py:class:`~ray.rllib.algorithms.algorithm.Algorithm` class to train two different policies in parallel
-   (also using multi-agent API).
+   Example of how to override the :py:meth:`~ray.rllib.algorithms.algorithm.Algorithm.training_step` method
+   to train two different policies in parallel (also using multi-agent API).
 
 Checkpoints
 -----------