[Tune] [PBT] [Doc] Fix and clean up PBT examples (ray-project#29060)

jkpjkpjkp · Oct 27, 2022 · 8c4e6dc · 8c4e6dc
1 parent cd031a0
commit 8c4e6dc
Show file tree

Hide file tree

Showing 20 changed files with 973 additions and 326 deletions.
diff --git a/doc/source/_toc.yml b/doc/source/_toc.yml
@@ -101,7 +101,7 @@ parts:
               - file: tune/tutorials/tune-metrics
                 title: "Using Callbacks and Metrics"
               - file: tune/examples/tune_analyze_results
-              - file: tune/tutorials/tune-advanced-tutorial
+              - file: tune/examples/pbt_guide
                 sections:
                   - file: tune/examples/pbt_visualization/pbt_visualization
                     title: "Visualizing and Understanding PBT"

diff --git a/doc/source/images/tune_advanced_dcgan_Gloss.png b/doc/source/images/tune_advanced_dcgan_Gloss.png
diff --git a/doc/source/images/tune_advanced_dcgan_inscore.png b/doc/source/images/tune_advanced_dcgan_inscore.png
diff --git a/doc/source/tune/api_docs/schedulers.rst b/doc/source/tune/api_docs/schedulers.rst
@@ -147,7 +147,7 @@ the trials to explore the new hyperparameter configuration starting from a parti
 trained model (e.g. by copying model weights from one of the top-performing trials).
 
 Take a look at :doc:`/tune/examples/pbt_visualization/pbt_visualization` to get an idea
-of how PBT operates. :doc:`/tune/tutorials/tune-advanced-tutorial` gives more examples
+of how PBT operates. :doc:`/tune/examples/pbt_guide` gives more examples
 of PBT usage.
 
 .. autoclass:: ray.tune.schedulers.PopulationBasedTraining

diff --git a/doc/source/tune/examples/includes/images/pbt_function_results.png b/doc/source/tune/examples/includes/images/pbt_function_results.png
diff --git a/doc/source/tune/examples/includes/pbt_function.rst b/doc/source/tune/examples/includes/pbt_function.rst
@@ -3,4 +3,9 @@
 PBT Function Example
 ~~~~~~~~~~~~~~~~~~~~
 
+The following script produces the following results. For a population of 8 trials,
+the PBT learning rate schedule roughly matches the optimal learning rate schedule.
+
+.. image:: images/pbt_function_results.png
+
 .. literalinclude:: /../../python/ray/tune/examples/pbt_function.py
diff --git a/doc/source/tune/examples/pbt_guide.ipynb b/doc/source/tune/examples/pbt_guide.ipynb
diff --git a/doc/source/tune/tutorials/overview.rst b/doc/source/tune/tutorials/overview.rst
@@ -87,9 +87,9 @@ Tune Feature Guides
     ---
     :img-top: /images/tune.png
 
-    .. link-button:: tune-advanced-tutorial
+    .. link-button:: ../examples/pbt_guide
         :type: ref
-        :text: Using Population-Based Training
+        :text: A Guide to Population-Based Training
         :classes: btn-link btn-block stretched-link
 
     ---

diff --git a/doc/source/tune/tutorials/tune-advanced-tutorial.rst b/doc/source/tune/tutorials/tune-advanced-tutorial.rst