&`iSddlZddlZddlZddlZddlZddlmZmZmZm Z m Z m Z ddl m Z ddlmZddlmZmZddlmZer ddlmZddlmZejeZeGd d ZeGd d eZdS) N) TYPE_CHECKINGAnyDictListOptionalUnion) tag_searcher)+_set_search_properties_backwards_compatible) DeveloperAPI PublicAPI)log_onceExperimentAnalysisTrialc eZdZdZdZdZ d*deedeefdZdeedeede d e fd Z d ed e d dfd Z d+d ed ee de d dfdZ d ed ee fdZ d,de dede de deeef dZdededdfdefdZdefdZdefdZded e fdZd e fd Zd!e fd"Zd-d$ed%efd&Zd$efd'Zed efd(Zed efd)ZdS).SearcheraAbstract 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. Not all implementations support multi objectives. Note to Tune developers: If a new searcher is added, please update `air/_internal/usage.py`. Args: metric: The training result objective value attribute. If list then list of training result objective value attributes mode: If string One of {min, max}. If list then list of max and min, determines whether objective is minimizing or maximizing the metric attribute. Must match type of metric. .. 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]) tuner = tune.Tuner( trainable_function, tune_config=tune.TuneConfig( search_alg=ExampleSearch() ) ) tuner.fit() FINISHEDzsearcher-state-{}.pklNmetricmodect|||_||_|r|sdSt|t |s Jdt|t r|dvs JddSt|t rOt|t|ks Jdtd|Ds JddStd)Nz(metric and mode must be of the same type)minmaxz*if `mode` is a str must be 'min' or 'max'!z'Metric and mode must be the same lengthc3K|]}|dvV dS))rrobsN).0mods l/home/jaya/work/projects/VOICE-AGENT/VIET/agent-env/lib/python3.11/site-packages/ray/tune/search/searcher.py z$Searcher.__init__..bs914,,z,All of mode must be 'min' or 'max' or 'obs'!z$Mode most either be a list or string) r _metric_mode isinstancetypestrlistlenall ValueError)selfrrs r__init__zSearcher.__init__Ns6 T   6  F DJJ   6 6 5 6 6  dC  E>)))+W))))) d # # Et99F +++-V+++8< > >= > > > >CDD Dr!configreturnc dS)aPass search properties to searcher. This method acts as an alternative to instantiating search algorithms with their own specific search spaces. Instead they can accept a Tune config through this method. A searcher should return ``True`` if setting the config was successful, or ``False`` if it was unsuccessful, e.g. when the search space has already been set. Args: metric: Metric to optimize mode: One of ["min", "max"]. Direction to optimize. config: Tune config dict. **spec: Any kwargs for forward compatiblity. Info like Experiment.PUBLIC_KEYS is provided through here. Frr+rrr-specs rset_search_propertieszSearcher.set_search_propertieshs $ur!trial_idresultcdS)aOptional 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: A unique string ID for the trial. result: 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. Nrr+r3r4s ron_trial_resultzSearcher.on_trial_result|s r!Ferrorct)aNotification for the completion of trial. Typically, this method is used for notifying the underlying optimizer of the result. Args: trial_id: A unique string ID for the trial. result: 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: True if the training process raised an error. NotImplementedErrorr+r3r4r8s ron_trial_completezSearcher.on_trial_completes &"!r!ct)aQueries the algorithm to retrieve the next set of parameters. Arguments: trial_id: 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. r:)r+r3s rsuggestzSearcher.suggests "!r! parametersvalueprunedintermediate_valuesct)aPass results from a point that has been evaluated separately. This method allows for information from outside the suggest - on_trial_complete loop to be passed to the search algorithm. This functionality depends on the underlying search algorithm and may not be always available. Args: parameters: Parameters used for the trial. value: Metric value obtained in the trial. error: True if the training process raised an error. pruned: True if trial was pruned. intermediate_values: List of metric values for intermediate iterations of the result. None if not applicable. r:r+r@rAr8rBrCs radd_evaluated_pointzSearcher.add_evaluated_points 4"!r!trials_or_analysisrrc |jtjkrtddlm}ddlm ddlmt|ttfr|}n>t| r|g}n*t||r|j }ntd|d d dtttff fd }|D]}||}|r |jd i| st!jd d Sd S) a5Pass results from trials that have been evaluated separately. This method allows for information from outside the suggest - on_trial_complete loop to be passed to the search algorithm. This functionality depends on the underlying search algorithm and may not be always available (same as ``add_evaluated_point``.) Args: trials_or_analysis: Trials to pass results form to the searcher. metric: Metric name reported by trials used for determining the objective value. rrr)DONEzRExpected input to be a `Trial`, a list of `Trial`s, or `ExperimentAnalysis`, got: Ftrialr.cp|jjko|jd }|jjko|jd}s |jvo|jr |jvrdSt |j|jd|jjk|dS)NF)r@rAr8rBrC)status TERMINATED last_resultgetdictr-ERROR)rJhas_trial_been_prunedhas_trial_finishedrIrany_trial_had_metricrs rtrial_to_pointsz6Searcher.add_evaluated_trials..trial_to_pointss  00;)--dE::: "   00WU5F5J5J4QV5W5W ( e//F4F% F%2C$C$Ct <'++FD99lek1,$(  r!zaNo completed trial returned the specified metric. Make sure the name you have passed is correct. Nr)rFrr;ray.tune.analysisrray.tune.experimentrray.tune.resultrIr$r'tupletrialsrr&rwarningswarn) r+rGrrrZrUrJkwargsrIrrTs ` @@@radd_evaluated_trialszSearcher.add_evaluated_trialss&  #x'C C C% % 988888------(((((( (4- 8 8 'FF *E 2 2 ()FF *,> ? ? '.FF%C.@CC  % 5 T#s(^         . 3 3E$_U++F 3((226222#  MB       r!checkpoint_pathct)awSave state to path for this search algorithm. Args: checkpoint_path: File where the search algorithm state is saved. This path should be used later when restoring from file. Example: .. code-block:: python search_alg = Searcher(...) tuner = tune.Tuner( cost, tune_config=tune.TuneConfig( search_alg=search_alg, num_samples=5 ), param_space=config ) results = tuner.fit() search_alg.save("./my_favorite_path.pkl") .. versionchanged:: 0.8.7 Save is automatically called by `Tuner().fit()`. You can use `Tuner().restore()` to restore from an experiment directory such as `~/ray_results/trainable`. r:r+r_s rsavez Searcher.saves @"!r!ct)aRestore state for this search algorithm Args: checkpoint_path: File where the search algorithm state is saved. This path should be the same as the one provided to "save". Example: .. code-block:: python search_alg.save("./my_favorite_path.pkl") search_alg2 = Searcher(...) search_alg2 = ConcurrencyLimiter(search_alg2, 1) search_alg2.restore(checkpoint_path) tuner = tune.Tuner( cost, tune_config=tune.TuneConfig( search_alg=search_alg2, num_samples=5 ), ) tuner.fit() r:ras rrestorezSearcher.restore<s 8"!r!max_concurrentcdS)aSet max concurrent trials this searcher can run. This method will be called on the wrapped searcher by the ``ConcurrencyLimiter``. It is intended to allow for searchers which have custom, internal logic handling max concurrent trials to inherit the value passed to ``ConcurrencyLimiter``. If this method returns False, it signifies that no special logic for handling this case is present in the searcher. Args: max_concurrent: Number of maximum concurrent trials. Frr+res rset_max_concurrencyzSearcher.set_max_concurrencyZs ur!ctNr:r+s r get_statezSearcher.get_statej!!r!statectrjr:r+rns r set_statezSearcher.set_statemrmr!defaultcheckpoint_dir session_strc tj|d}d} ||n;#t$r.t drt dd}YnwxYw|rltj|rOtj |tj||j |dSdSdS)a,Automatically saves the given searcher to the checkpoint_dir. This is automatically used by Tuner().fit() during a Tune job. Args: checkpoint_dir: Filepath to experiment dir. session_str: Unique identifier of the current run session. z.tmp_searcher_ckptTzsuggest:save_to_dirz1save not implemented for Searcher. Skipping save.FN) ospathjoinrbr;r loggerwarningexistsreplaceCKPT_FILE_TMPLformat)r+rsrttmp_search_ckpt_pathsuccesss r save_to_dirzSearcher.save_to_dirps "w||N|j|||||Srj)rrFrEs rrFz&ConcurrencyLimiter.add_evaluated_point=s*}00 uf.A   r!cb|j}|d=tj|S)Nr)__dict__copydeepcopyrps rrlzConcurrencyLimiter.get_stateIs- ""$$ * }U###r!rnc:|j|dSrj)rupdaterps rrqzConcurrencyLimiter.set_stateNs U#####r!r_c:|j|dSrj)rrbras rrbzConcurrencyLimiter.saveQs ?+++++r!c:|j|dSrj)rrdras rrdzConcurrencyLimiter.restoreTs o.....r!)Frr)rrrrrrrr,rrhrr&rr2r?r=r7rrrFrlrqrbrd __classcell__)rs@rrrs#  D   3 t      .   #$  sm +3C= BF     .KP111%-d^1CG1111888T8d888859           &d5k2     $4$$$$ $t$$$$,C,,,,/s////////r!r)rrloggingrvr[typingrrrrrrray.air._internal.usager ray.tune.search.utilr ray.util.annotationsr r ray.util.debugr rVrrWr getLoggerrryrrrr!rrsu  BBBBBBBBBBBBBBBB000000LLLLLL88888888######*444444))))))  8 $ $````````F  \/\/\/\/\/\/\/ \/\/\/r!