ray/doc/source/tune/api_docs/schedulers.rst

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

293 lines
12 KiB
ReStructuredText
Raw Normal View History

.. _tune-schedulers:
Trial Schedulers (tune.schedulers)
==================================
In Tune, some hyperparameter optimization algorithms are written as "scheduling algorithms".
These Trial Schedulers can early terminate bad trials, pause trials, clone trials,
and alter hyperparameters of a running trial.
All Trial Schedulers take in a ``metric``, which is a value returned in the result dict of your
Trainable and is maximized or minimized according to ``mode``.
.. code-block:: python
from ray import tune
tuner = tune.Tuner( ... , tune_config=tune.TuneConfig(scheduler=Scheduler(metric="accuracy", mode="max")))
results = tuner.fit()
.. _tune-scheduler-hyperband:
ASHA (tune.schedulers.ASHAScheduler)
------------------------------------
The `ASHA <https://openreview.net/forum?id=S1Y7OOlRZ>`__ scheduler can be used by
setting the ``scheduler`` parameter of ``tune.TuneConfig``, which is taken in by ``Tuner``, e.g.
.. code-block:: python
from ray import tune
asha_scheduler = ASHAScheduler(
time_attr='training_iteration',
metric='episode_reward_mean',
mode='max',
max_t=100,
grace_period=10,
reduction_factor=3,
brackets=1)
tuner = tune.Tuner( ... , tune_config=tune.TuneConfig(scheduler=asha_scheduler))
results = tuner.fit()
Compared to the original version of HyperBand, this implementation provides better
parallelism and avoids straggler issues during eliminations.
**We recommend using this over the standard HyperBand scheduler.**
An example of this can be found here: :doc:`/tune/examples/includes/async_hyperband_example`.
Even though the original paper mentions a bracket count of 3, discussions with the authors concluded
that the value should be left to 1 bracket.
This is the default used if no value is provided for the ``brackets`` argument.
.. autoclass:: ray.tune.schedulers.AsyncHyperBandScheduler
.. autoclass:: ray.tune.schedulers.ASHAScheduler
.. _tune-original-hyperband:
HyperBand (tune.schedulers.HyperBandScheduler)
----------------------------------------------
Tune implements the `standard version of HyperBand <https://arxiv.org/abs/1603.06560>`__.
**We recommend using the ASHA Scheduler over the standard HyperBand scheduler.**
.. autoclass:: ray.tune.schedulers.HyperBandScheduler
HyperBand Implementation Details
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Implementation details may deviate slightly from theory but are focused on increasing usability.
Note: ``R``, ``s_max``, and ``eta`` are parameters of HyperBand given by the paper.
See `this post <https://blog.ml.cmu.edu/2018/12/12/massively-parallel-hyperparameter-optimization/>`_ for context.
1. Both ``s_max`` (representing the ``number of brackets - 1``) and ``eta``, representing the downsampling rate, are fixed.
In many practical settings, ``R``, which represents some resource unit and often the number of training iterations,
can be set reasonably large, like ``R >= 200``.
For simplicity, assume ``eta = 3``. Varying ``R`` between ``R = 200`` and ``R = 1000``
creates a huge range of the number of trials needed to fill up all brackets.
.. image:: /images/hyperband_bracket.png
On the other hand, holding ``R`` constant at ``R = 300`` and varying ``eta`` also leads to
HyperBand configurations that are not very intuitive:
.. image:: /images/hyperband_eta.png
The implementation takes the same configuration as the example given in the paper
and exposes ``max_t``, which is not a parameter in the paper.
2. The example in the `post <https://blog.ml.cmu.edu/2018/12/12/massively-parallel-hyperparameter-optimization/>`_ to calculate ``n_0``
is actually a little different than the algorithm given in the paper.
In this implementation, we implement ``n_0`` according to the paper (which is `n` in the below example):
.. image:: /images/hyperband_allocation.png
3. There are also implementation specific details like how trials are placed into brackets which are not covered in the paper.
This implementation places trials within brackets according to smaller bracket first - meaning
that with low number of trials, there will be less early stopping.
.. _tune-scheduler-msr:
Median Stopping Rule (tune.schedulers.MedianStoppingRule)
---------------------------------------------------------
The Median Stopping Rule implements the simple strategy of stopping a trial if its performance falls
below the median of other trials at similar points in time.
.. autoclass:: ray.tune.schedulers.MedianStoppingRule
.. _tune-scheduler-pbt:
Population Based Training (tune.schedulers.PopulationBasedTraining)
-------------------------------------------------------------------
Tune includes a distributed implementation of `Population Based Training (PBT) <https://www.deepmind.com/blog/population-based-training-of-neural-networks>`__.
This can be enabled by setting the ``scheduler`` parameter of ``tune.TuneConfig``, which is taken in by ``Tuner``, e.g.
.. code-block:: python
pbt_scheduler = PopulationBasedTraining(
time_attr='time_total_s',
metric='mean_accuracy',
mode='max',
perturbation_interval=600.0,
hyperparam_mutations={
"lr": [1e-3, 5e-4, 1e-4, 5e-5, 1e-5],
"alpha": lambda: random.uniform(0.0, 1.0),
...
})
tuner = tune.Tuner( ... , tune_config=tune.TuneConfig(scheduler=pbt_scheduler))
tuner.fit()
When the PBT scheduler is enabled, each trial variant is treated as a member of the population.
Periodically, top-performing trials are checkpointed
(this requires your Trainable to support :ref:`save and restore <tune-checkpoint-syncing>`).
Low-performing trials clone the checkpoints of top performers and perturb the configurations
in the hope of discovering an even better variation.
You can run this :doc:`toy PBT example </tune/examples/includes/pbt_function>` to get an idea of how how PBT operates.
When training in PBT mode, a single trial may see many different hyperparameters over its lifetime,
which is recorded in its ``result.json`` file.
The following figure generated by the example shows PBT with optimizing a LR schedule over
the course of a single experiment:
.. image:: ../images/pbt.png
.. autoclass:: ray.tune.schedulers.PopulationBasedTraining
.. _tune-scheduler-pbt-replay:
Population Based Training Replay (tune.schedulers.PopulationBasedTrainingReplay)
--------------------------------------------------------------------------------
Tune includes a utility to replay hyperparameter schedules of Population Based Training runs.
You just specify an existing experiment directory and the ID of the trial you would
like to replay. The scheduler accepts only one trial, and it will update its
config according to the obtained schedule.
.. code-block:: python
replay = PopulationBasedTrainingReplay(
experiment_dir="~/ray_results/pbt_experiment/",
trial_id="XXXXX_00001")
tuner = tune.Tuner(
...,
tune_config=tune.TuneConfig(scheduler=replay)
)
results = tuner.fit()
See :ref:`here for an example <tune-advanced-tutorial-pbt-replay>` on how to use the
replay utility in practice.
.. autoclass:: ray.tune.schedulers.PopulationBasedTrainingReplay
.. _tune-scheduler-pb2:
Population Based Bandits (PB2) (tune.schedulers.pb2.PB2)
--------------------------------------------------------
Tune includes a distributed implementation of `Population Based Bandits (PB2) <https://arxiv.org/abs/2002.02518>`__.
This algorithm builds upon PBT, with the main difference being that instead of using random perturbations,
PB2 selects new hyperparameter configurations using a Gaussian Process model.
The Tune implementation of PB2 requires GPy and sklearn to be installed:
.. code-block:: bash
pip install GPy sklearn
PB2 can be enabled by setting the ``scheduler`` parameter of ``tune.TuneConfig`` which is taken in by ``Tuner``, e.g.:
.. code-block:: python
from ray.tune.schedulers.pb2 import PB2
pb2_scheduler = PB2(
time_attr='time_total_s',
metric='mean_accuracy',
mode='max',
perturbation_interval=600.0,
hyperparam_bounds={
"lr": [1e-3, 1e-5],
"alpha": [0.0, 1.0],
...
})
tuner = tune.Tuner( ... , tune_config=tune.TuneConfig(scheduler=pb2_scheduler))
results = tuner.fit()
When the PB2 scheduler is enabled, each trial variant is treated as a member of the population.
Periodically, top-performing trials are checkpointed (this requires your Trainable to
support :ref:`save and restore <tune-checkpoint-syncing>`).
Low-performing trials clone the checkpoints of top performers and perturb the configurations
in the hope of discovering an even better variation.
The primary motivation for PB2 is the ability to find promising hyperparamters with only a small population size.
With that in mind, you can run this :doc:`PB2 PPO example </tune/examples/includes/pb2_ppo_example>` to compare PB2 vs. PBT,
with a population size of ``4`` (as in the paper).
The example uses the ``BipedalWalker`` environment so does not require any additional licenses.
.. autoclass:: ray.tune.schedulers.pb2.PB2
.. _tune-scheduler-bohb:
BOHB (tune.schedulers.HyperBandForBOHB)
---------------------------------------
This class is a variant of HyperBand that enables the `BOHB Algorithm <https://arxiv.org/abs/1807.01774>`_.
This implementation is true to the original HyperBand implementation and does not implement pipelining nor
straggler mitigation.
This is to be used in conjunction with the Tune BOHB search algorithm.
See :ref:`TuneBOHB <suggest-TuneBOHB>` for package requirements, examples, and details.
An example of this in use can be found here: :doc:`/tune/examples/includes/bohb_example`.
.. autoclass:: ray.tune.schedulers.HyperBandForBOHB
.. _tune-resource-changing-scheduler:
ResourceChangingScheduler
-------------------------
This class is a utility scheduler, allowing for trial resource requirements to be changed during tuning.
It wraps around another scheduler and uses its decisions.
* If you are using the Trainable (class) API for tuning, your Trainable must implement ``Trainable.update_resources``,
which will let your model know about the new resources assigned. You can also obtain the current trial resources
by calling ``Trainable.trial_resources``.
* If you are using the functional API for tuning, the current trial resources can be
obtained by calling `tune.get_trial_resources()` inside the training function.
The function should be able to :ref:`load and save checkpoints <tune-checkpoint-syncing>`
(the latter preferably every iteration).
An example of this in use can be found here: :doc:`/tune/examples/includes/xgboost_dynamic_resources_example`.
.. autoclass:: ray.tune.schedulers.ResourceChangingScheduler
DistributeResources
~~~~~~~~~~~~~~~~~~~
.. autoclass:: ray.tune.schedulers.resource_changing_scheduler.DistributeResources
DistributeResourcesToTopJob
~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: ray.tune.schedulers.resource_changing_scheduler.DistributeResourcesToTopJob
FIFOScheduler
-------------
.. autoclass:: ray.tune.schedulers.FIFOScheduler
TrialScheduler
--------------
.. autoclass:: ray.tune.schedulers.TrialScheduler
:members:
[tune] implement shim instantiation (#10456) * Create ray.tune.suggest.create.create_scheduler * Update __init__.py * Resolve conflict in __init__.py * Create ray.tune.schedulers.create.create_scheduler * Update __init__.py * Move create_scheduler to tune.schedulers.__init__ * Move create_searcher to tune.suggest.__init__ * Delete tune.suggest.create * Delete tune.schedulers.create * Update imports for shim functions in tune.__init__ * Remove shim from tune.suggest.__init__.__all__ * Remove shim from tune.schedulers.__init__.__all__ * Add ShimCreationTest * Move ShimCreationTest to test_api * Delete test_shim.py * Add docstring for ray.tune.create_scheduler * Add docstring to ray.tune.create_searcher * Fix typo in ray.tune.create_scheduler docstring * Fix lint errors in tune.schedulers.__init__ * Fix lint errors in tune.suggest.__init__ * Fix lint errors in tune.suggest.__init__ * Fix lint errors in tune.schedulers.__init__ * Fix imports in test_api * Fix lint errors in test_api * Fix kwargs in create_searcher * Fix kwargs in create_scheduler * Merge branch 'master' into shim-instantiation * Update use-case in docs in tune.create_scheduler * Update use-case in docs in tune.create_searcher * Remove duplicate pytest run from test_api * Add check to create_searcher Co-authored-by: Richard Liaw <rliaw@berkeley.edu> * Add check to create_scheduler * lint * Compare types of instances in test_api Co-authored-by: Richard Liaw <rliaw@berkeley.edu> * Add tune.create_searcher to docs * Fix doc build * Fix tests * Add tune.create_scheduler to docs * Fix tests * Fix lint errors * Update Ax search for master * Fix metric kwarg for Ax in test_api * Fix doc build * Fix HyperOptSearch import in test_api * Fix HyperOptSearch import in create_searcher Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-09-05 12:36:42 -04:00
Shim Instantiation (tune.create_scheduler)
------------------------------------------
There is also a shim function that constructs the scheduler based on the provided string.
This can be useful if the scheduler you want to use changes often (e.g., specifying the scheduler
via a CLI option or config file).
[tune] implement shim instantiation (#10456) * Create ray.tune.suggest.create.create_scheduler * Update __init__.py * Resolve conflict in __init__.py * Create ray.tune.schedulers.create.create_scheduler * Update __init__.py * Move create_scheduler to tune.schedulers.__init__ * Move create_searcher to tune.suggest.__init__ * Delete tune.suggest.create * Delete tune.schedulers.create * Update imports for shim functions in tune.__init__ * Remove shim from tune.suggest.__init__.__all__ * Remove shim from tune.schedulers.__init__.__all__ * Add ShimCreationTest * Move ShimCreationTest to test_api * Delete test_shim.py * Add docstring for ray.tune.create_scheduler * Add docstring to ray.tune.create_searcher * Fix typo in ray.tune.create_scheduler docstring * Fix lint errors in tune.schedulers.__init__ * Fix lint errors in tune.suggest.__init__ * Fix lint errors in tune.suggest.__init__ * Fix lint errors in tune.schedulers.__init__ * Fix imports in test_api * Fix lint errors in test_api * Fix kwargs in create_searcher * Fix kwargs in create_scheduler * Merge branch 'master' into shim-instantiation * Update use-case in docs in tune.create_scheduler * Update use-case in docs in tune.create_searcher * Remove duplicate pytest run from test_api * Add check to create_searcher Co-authored-by: Richard Liaw <rliaw@berkeley.edu> * Add check to create_scheduler * lint * Compare types of instances in test_api Co-authored-by: Richard Liaw <rliaw@berkeley.edu> * Add tune.create_searcher to docs * Fix doc build * Fix tests * Add tune.create_scheduler to docs * Fix tests * Fix lint errors * Update Ax search for master * Fix metric kwarg for Ax in test_api * Fix doc build * Fix HyperOptSearch import in test_api * Fix HyperOptSearch import in create_searcher Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-09-05 12:36:42 -04:00
.. automethod:: ray.tune.create_scheduler