Files
ray/python/ray/tune/suggest/suggestion.py
T

343 lines
12 KiB
Python

import copy
import logging
from ray.tune.error import TuneError
from ray.tune.experiment import convert_to_experiment_list
from ray.tune.config_parser import make_parser, create_trial_from_spec
from ray.tune.suggest.search import SearchAlgorithm
from ray.tune.suggest.variant_generator import format_vars, resolve_nested_dict
from ray.tune.trial import Trial
from ray.tune.utils import merge_dicts, flatten_dict
logger = logging.getLogger(__name__)
def _warn_on_repeater(searcher, total_samples):
from ray.tune.suggest.repeater import _warn_num_samples
_warn_num_samples(searcher, total_samples)
class Searcher:
"""Abstract class for wrapping suggesting algorithms.
Custom algorithms can extend this class easily by overriding the
`suggest` method provide generated parameters for the trials.
Any subclass that implements ``__init__`` must also call the
constructor of this class: ``super(Subclass, self).__init__(...)``.
To track suggestions and their corresponding evaluations, the method
`suggest` will be passed a trial_id, which will be used in
subsequent notifications.
Args:
metric (str): The training result objective value attribute.
mode (str): One of {min, max}. Determines whether objective is
minimizing or maximizing the metric attribute.
.. code-block:: python
class ExampleSearch(Searcher):
def __init__(self, metric="mean_loss", mode="min", **kwargs):
super(ExampleSearch, self).__init__(
metric=metric, mode=mode, **kwargs)
self.optimizer = Optimizer()
self.configurations = {}
def suggest(self, trial_id):
configuration = self.optimizer.query()
self.configurations[trial_id] = configuration
def on_trial_complete(self, trial_id, result, **kwargs):
configuration = self.configurations[trial_id]
if result and self.metric in result:
self.optimizer.update(configuration, result[self.metric])
tune.run(trainable_function, search_alg=ExampleSearch())
"""
FINISHED = "FINISHED"
def __init__(self,
metric="episode_reward_mean",
mode="max",
max_concurrent=None,
use_early_stopped_trials=None):
if use_early_stopped_trials is False:
raise DeprecationWarning(
"Early stopped trials are now always used. If this is a "
"problem, file an issue: https://github.com/ray-project/ray.")
if max_concurrent is not None:
logger.warning(
"DeprecationWarning: `max_concurrent` is deprecated for this "
"search algorithm. Use tune.suggest.ConcurrencyLimiter() "
"instead. This will raise an error in future versions of Ray.")
assert mode in ["min", "max"], "`mode` must be 'min' or 'max'!"
self._metric = metric
self._mode = mode
def on_trial_result(self, trial_id, result):
"""Optional notification for result during training.
Note that by default, the result dict may include NaNs or
may not include the optimization metric. It is up to the
subclass implementation to preprocess the result to
avoid breaking the optimization process.
Args:
trial_id (str): A unique string ID for the trial.
result (dict): Dictionary of metrics for current training progress.
Note that the result dict may include NaNs or
may not include the optimization metric. It is up to the
subclass implementation to preprocess the result to
avoid breaking the optimization process.
"""
pass
def on_trial_complete(self, trial_id, result=None, error=False):
"""Notification for the completion of trial.
Typically, this method is used for notifying the underlying
optimizer of the result.
Args:
trial_id (str): A unique string ID for the trial.
result (dict): Dictionary of metrics for current training progress.
Note that the result dict may include NaNs or
may not include the optimization metric. It is up to the
subclass implementation to preprocess the result to
avoid breaking the optimization process. Upon errors, this
may also be None.
error (bool): True if the training process raised an error.
"""
raise NotImplementedError
def suggest(self, trial_id):
"""Queries the algorithm to retrieve the next set of parameters.
Arguments:
trial_id (str): Trial ID used for subsequent notifications.
Returns:
dict | FINISHED | None: Configuration for a trial, if possible.
If FINISHED is returned, Tune will be notified that
no more suggestions/configurations will be provided.
If None is returned, Tune will skip the querying of the
searcher for this step.
"""
raise NotImplementedError
def save(self, checkpoint_dir):
"""Save function for this object."""
raise NotImplementedError
def restore(self, checkpoint_dir):
"""Restore function for this object."""
raise NotImplementedError
@property
def metric(self):
"""The training result objective value attribute."""
return self._metric
@property
def mode(self):
"""Specifies if minimizing or maximizing the metric."""
return self._mode
class ConcurrencyLimiter(Searcher):
"""A wrapper algorithm for limiting the number of concurrent trials.
Args:
searcher (Searcher): Searcher object that the
ConcurrencyLimiter will manage.
Example:
.. code-block:: python
from ray.tune.suggest import ConcurrencyLimiter
search_alg = HyperOptSearch(metric="accuracy")
search_alg = ConcurrencyLimiter(search_alg, max_concurrent=2)
tune.run(trainable, search_alg=search_alg)
"""
def __init__(self, searcher, max_concurrent):
assert type(max_concurrent) is int and max_concurrent > 0
self.searcher = searcher
self.max_concurrent = max_concurrent
self.live_trials = set()
super(ConcurrencyLimiter, self).__init__(
metric=self.searcher.metric, mode=self.searcher.mode)
def suggest(self, trial_id):
if len(self.live_trials) >= self.max_concurrent:
return
suggestion = self.searcher.suggest(trial_id)
if suggestion not in (None, Searcher.FINISHED):
self.live_trials.add(trial_id)
return suggestion
def on_trial_complete(self, trial_id, result=None, error=False):
if trial_id not in self.live_trials:
return
else:
self.searcher.on_trial_complete(
trial_id, result=result, error=error)
self.live_trials.remove(trial_id)
def save(self, checkpoint_dir):
self.searcher.save(checkpoint_dir)
def restore(self, checkpoint_dir):
self.searcher.restore(checkpoint_dir)
class SearchGenerator(SearchAlgorithm):
"""Generates trials to be passed to the TrialRunner.
Uses the provided ``searcher`` object to generate trials. This class
transparently handles repeating trials with score aggregation
without embedding logic into the Searcher.
Args:
searcher: Search object that subclasses the Searcher base class. This
is then used for generating new hyperparameter samples.
"""
def __init__(self, searcher):
assert issubclass(
type(searcher),
Searcher), ("Searcher should be subclassing Searcher.")
self.searcher = searcher
self._parser = make_parser()
self._experiment = None
self._counter = 0 # Keeps track of number of trials created.
self._total_samples = 0 # int: total samples to evaluate.
self._finished = False
def add_configurations(self, experiments):
"""Registers experiment specifications.
Arguments:
experiments (Experiment | list | dict): Experiments to run.
"""
logger.debug("added configurations")
experiment_list = convert_to_experiment_list(experiments)
assert len(experiment_list) == 1, (
"SearchAlgorithms can only support 1 experiment at a time.")
self._experiment = experiment_list[0]
experiment_spec = self._experiment.spec
self._total_samples = experiment_spec.get("num_samples", 1)
_warn_on_repeater(self.searcher, self._total_samples)
if "run" not in experiment_spec:
raise TuneError("Must specify `run` in {}".format(experiment_spec))
def next_trials(self):
"""Provides a batch of Trial objects to be queued into the TrialRunner.
Returns:
List[Trial]: A list of trials for the Runner to consume.
"""
trials = []
while not self.is_finished():
trial = self.create_trial_if_possible(self._experiment.spec,
self._experiment.name)
if trial is None:
break
trials.append(trial)
return trials
def create_trial_if_possible(self, experiment_spec, output_path):
logger.debug("creating trial")
trial_id = Trial.generate_id()
suggested_config = self.searcher.suggest(trial_id)
if suggested_config == Searcher.FINISHED:
self._finished = True
logger.debug("Searcher has finished.")
return
if suggested_config is None:
return
spec = copy.deepcopy(experiment_spec)
spec["config"] = merge_dicts(spec["config"],
copy.deepcopy(suggested_config))
# Create a new trial_id if duplicate trial is created
flattened_config = resolve_nested_dict(spec["config"])
self._counter += 1
tag = "{0}_{1}".format(
str(self._counter), format_vars(flattened_config))
trial = create_trial_from_spec(
spec,
output_path,
self._parser,
evaluated_params=flatten_dict(suggested_config),
experiment_tag=tag,
trial_id=trial_id)
return trial
def on_trial_result(self, trial_id, result):
"""Notifies the underlying searcher."""
self.searcher.on_trial_result(trial_id, result)
def on_trial_complete(self, trial_id, result=None, error=False):
self.searcher.on_trial_complete(
trial_id=trial_id, result=result, error=error)
def is_finished(self):
return self._counter >= self._total_samples or self._finished
class _MockSearcher(Searcher):
def __init__(self, **kwargs):
self.live_trials = {}
self.counter = {"result": 0, "complete": 0}
self.final_results = []
self.stall = False
self.results = []
super(_MockSearcher, self).__init__(**kwargs)
def suggest(self, trial_id):
if not self.stall:
self.live_trials[trial_id] = 1
return {"test_variable": 2}
return None
def on_trial_result(self, trial_id, result):
self.counter["result"] += 1
self.results += [result]
def on_trial_complete(self, trial_id, result=None, error=False):
self.counter["complete"] += 1
if result:
self._process_result(result)
if trial_id in self.live_trials:
del self.live_trials[trial_id]
def _process_result(self, result):
self.final_results += [result]
class _MockSuggestionAlgorithm(SearchGenerator):
def __init__(self, max_concurrent=None, **kwargs):
self.searcher = _MockSearcher(**kwargs)
if max_concurrent:
self.searcher = ConcurrencyLimiter(
self.searcher, max_concurrent=max_concurrent)
super(_MockSuggestionAlgorithm, self).__init__(self.searcher)
@property
def live_trials(self):
return self.searcher.live_trials
@property
def results(self):
return self.searcher.results