docs/major update to docs, sphinx, and docs ci script #36
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add wrapper load_config and wrapper utility functions
Update install instructions to first favor installing in a conda env and installing the dependencies and then installing with pip from the GH repo instead of having to clone the repo.
Move x86 off of pre torch 12 b/c nightly fixes the mkl issue Also will add issue to move off of nightly once that is released on stable
…-install-w-pip-from-gh Update install instructions to include pip from repo
… additional class ModularMetric use to use a MetricToEval class for json serialization, which wasn't necessary at all. Removed that class, and moved the functionality onto ModularMetric. This reduced clutter and confusion. Now when you look at a metric, metric.f/metric.metric_to_eval is the function, instead of having todo metric.f.func/metric.metric_to_eval.func Which feels much easier and cleaner. This also made it easier to make all the built--in metrics subclasses, which allowed easier documentation with sphinx rendering. Added a test that saves and loads a scheduler and resumes an experiment to check that the scheduler saving works. Added another test to save an arbitrary callable that is passed in. it doesn't work right now, marked as skip and made an issue to get that functionality for scheduler saving and loading with arbitrary callables working at some point.
Modified code reference api section to not use the template as much, and allowing each module to be generated when it is needed in the api section. This is a tradeoff of convencience for more customizability and a better doc design Also added more module docstrings and improved the ones that were already there
Add docutils func that adds a reference to the relative __init__.py of all files __doc__ in each submodule change sphinx configuration to be nitpicky which revealed a bunch of broken sphinx links which were then fixed (except ax and botorch links which can't be fixed right now because they don't build sphinx in a normal way. see facebook/Ax#1227 and see pytorch/botorch#1464 for more info on when ax and botroch can be linked). Add more docstrings
Add an example notebook that runs a toy example in python and saves the output and another one that loads that output from the scheduler.json file and then displays the plots
madeline-scyphers
changed the title
jemissik
reviewed
Nov 5, 2022
jemissik
reviewed
Nov 5, 2022
jemissik
reviewed
Nov 5, 2022
- Minor edits to docs - Fixed refs/linking in notebooks
jemissik
requested changes
Nov 5, 2022
PR gave feedback on missing info on passthrough metric, on how to use the experiment_dir in the config file, and changed the contribution guidelines to say fork the repo instead of clone it. Also changed the the x86mac install back to how it was before before we switched it to nightly pytorch since nightly pytorch is broken right now.
jemissik
reviewed
Nov 5, 2022
jemissik
approved these changes
Nov 6, 2022
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Update ModularMetric to directly save metric_to_eval instead of using… …
… additional class
ModularMetric use to use a MetricToEval class for json serialization, which wasn't necessary
at all. Removed that class, and moved the functionality onto ModularMetric. This reduced
clutter and confusion. Now when you look at a metric, metric.f/metric.metric_to_eval is
the function, instead of having todo metric.f.func/metric.metric_to_eval.func Which feels
much easier and cleaner. This also made it easier to make all the built--in metrics
subclasses, which allowed easier documentation with sphinx rendering.
Added a test that saves and loads a scheduler and resumes an experiment to check that
the scheduler saving works. Added another test to save an arbitrary callable that is
passed in. it doesn't work right now, marked as skip and made an issue to get that
functionality for scheduler saving and loading with arbitrary callables working at
Modify api template …
Modified code reference api section to not use the
template as much, and allowing each module to be generated
when it is needed in the api section. This is a tradeoff of
convencience for more customizability and a better doc design
More docstring updates, fix broken rst links …
Add docutils func that adds a reference to the relative
init.py of all files doc in each submodule
change sphinx configuration to be nitpicky which revealed a bunch
of broken sphinx links which were then fixed (except ax and
botorch links which can't be fixed right now because
they don't build sphinx in a normal way.
see intersphinx mapping facebook/Ax#1227
and see Intersphinx Mapping Request pytorch/botorch#1464
for more info on when ax and botroch can be linked).
Add examples, boa cli info, config info
Add example cli boa script with output and basic instructions on how to do boa cli
Add example notebooks and add docs build to CI
Add an example notebook that runs a toy example in python and
saves the output and another one that loads that output
from the scheduler.json file and then displays the plots