Skip to content

[tune] Update Trainable doc to expose interface #2272

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
richardliaw merged 2 commits into from
Jun 20, 2018
Merged

[tune] Update Trainable doc to expose interface #2272

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
0f88bd6
more clarity for trainable
richardliaw Jun 19, 2018
517c60f
run_experiment docs
richardliaw Jun 19, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 6 additions & 3 deletions doc/source/tune.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ Quick Start
ray.init()
tune.register_trainable("train_func", train_func)

tune.run_experiments({
all_trials = tune.run_experiments({
"my_experiment": {
"run": "train_func",
"stop": {"mean_accuracy": 99},
Expand Down Expand Up @@ -55,7 +55,7 @@ For the function you wish to tune, add a two-line modification (note that we use
accuracy = eval_accuracy(...)
reporter(timesteps_total=idx, mean_accuracy=accuracy) # report metrics

This PyTorch script runs a small grid search over the ``train_func`` function using Ray Tune, reporting status on the command line until the stopping condition of ``mean_accuracy >= 99`` is reached (for metrics like _loss_ that decrease over time, specify `neg_mean_loss <https://github.com/ray-project/ray/blob/master/python/ray/tune/result.py#L40>`__ as a condition instead):
This PyTorch script runs a small grid search over the ``train_func`` function using Ray Tune, reporting status on the command line until the stopping condition of ``mean_accuracy >= 99`` is reached (for metrics like `loss` that decrease over time, specify `neg_mean_loss <https://github.com/ray-project/ray/blob/master/python/ray/tune/result.py#L40>`__ as a condition instead):

::

Expand All @@ -72,7 +72,9 @@ This PyTorch script runs a small grid search over the ``train_func`` function us

In order to report incremental progress, ``train_func`` periodically calls the ``reporter`` function passed in by Ray Tune to return the current timestep and other metrics as defined in `ray.tune.result.TrainingResult <https://github.com/ray-project/ray/blob/master/python/ray/tune/result.py>`__. Incremental results will be synced to local disk on the head node of the cluster.

Learn more `about specifying experiments <tune-config.html>`__ .
``tune.run_experiments`` returns a list of Trial objects which you can inspect results of via ``trial.last_result``.

Learn more `about specifying experiments <tune-config.html>`__.


Features
Expand Down Expand Up @@ -242,6 +244,7 @@ Additionally, checkpointing can be used to provide fault-tolerance for experimen
The class interface that must be implemented to enable checkpointing is as follows:

.. autoclass:: ray.tune.trainable.Trainable
:members: _save, _restore, _train, _setup, _stop


Client API
Expand Down
3 changes: 3 additions & 0 deletions python/ray/tune/tune.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,6 +54,9 @@ def run_experiments(experiments,
not currently have enough resources to launch one. This should
be set to True when running on an autoscaling cluster to enable
automatic scale-up.

Returns:
List of Trial objects, holding data for each executed trial.
"""

if scheduler is None:
Expand Down