:jp6fddlZddlZddlmZddlmZmZmZmZm Z m Z m Z ddl Z ddlZddlZddlZddlmZddlmZddlmZmZddlmZmZddlmZmZmZdd lmZddlmZmZmZdd lm Z d d l!m"Z"d d lm#Z#ejHjKejHjLGddZ'y)N)Path)IterableOptionalCallable GeneratorMappingUnionDict)ExperimentalWarning) BasePruner) BaseSampler TPESampler)Trial FixedTrial) RDBStorageJournalStorageJournalFileStorage)tqdm) bayes_mvs)Pipeline) PipelineInputcXeZdZdZ ddedeedeedeeee fdeeee fdee d e fd Z ed efd Zed efd Zed efdZ ddeedee efd eegeffdZ ddeede dedee efd ef dZ ddeededee efd eeddffdZy) OptimizeracPipeline optimizer Parameters ---------- pipeline : `Pipeline` Pipeline. db : `Path`, optional Path to trial database on disk. Use ".sqlite" extension for SQLite backend, and ".journal" for Journal backend (prefered for parallel optimization). study_name : `str`, optional Name of study. In case it already exists, study will continue from there. # TODO -- generate this automatically sampler : `str` or sampler instance, optional Algorithm for value suggestion. Must be one of "RandomSampler" or "TPESampler", or a sampler instance. Defaults to "TPESampler". pruner : `str` or pruner instance, optional Algorithm for early pruning of trials. Must be one of "MedianPruner" or "SuccessiveHalvingPruner", or a pruner instance. Defaults to no pruning. seed : `int`, optional Seed value for the random number generator of the sampler. Defaults to no seed. average_case : `bool`, optional Optimize for average case (default). Set to False to optimize for worst case. Npipelinedb study_namesamplerprunerseed average_casec >||_||_|d|_nt|jj}|dk(r3t j dtd|j|_nL|dk(rtd|j|_n)|dk(r$tt|j|_||_ t|tr||_ nKt|tr( tt j"|||_ n|t)||_ t|t*r||_n=t|tr& tt j.||_nd|_t!j0|jd |j|j|j,|jj3 |_||_y#t$$r} d} t'| d} ~ wwxYw#t$$r} d} t'| d} ~ wwxYw) Nz.dbzHStorage with '.db' extension has been deprecated. Use '.sqlite' instead.z sqlite:///z.sqlitez.journal)r z8`sampler` must be one of "RandomSampler" or "TPESampler"zC`pruner` must be one of "MedianPruner" or "SuccessiveHalvingPruner"T)rload_if_existsstoragerr direction)rrstorage_rsuffixwarningswarnrrrr isinstancer rstrgetattroptunasamplersAttributeError ValueErrorrr rpruners create_study get_directionstudy_r!) selfrrrrrr r! extensionemsgs p/home/mcse/projects/srt_converter/srt-converter-venv/lib/python3.12/site-packages/pyannote/pipeline/optimizer.py__init__zOptimizer.__init__Ss!  : DMTWW ,,IE! ^!+Zy+A B i' *Zy+A B j( ./ATWWI/O P $ g{ +"DL  % &@wv@dK _%40DL fj ) DK  $ &=gfnnf=? DK ))MMLL;;mm113   )=" &P o% &" &[ o% &s0,&G#$H# G>, G99G> H HHreturnc |jj}|S#t$r8|jj dk(rdnd}|t j z}Y|SwxYw)zReturn best loss so farminimizer)r4 best_value Exceptionrr3npinf)r5r?r%s r9 best_losszOptimizer.best_losss_ ,//J ,"&--"="="?:"MQSUI"RVV+J ,s=AAcxt|jj}|jj |S)zReturn best parameters so fartrial)rr4 best_paramsr parameters)r5rFs r9rGzOptimizer.best_paramss04;;223}}''e'44cL|jj|jS)z8Return pipeline instantiated with best parameters so far)r instantiaterG)r5s r9 best_pipelinezOptimizer.best_pipelines}}(()9)9::rIinputs show_progresscvtt}dk(rdddddtdtffd }|S) a Create objective function used by optuna Parameters ---------- inputs : `iterable` List of inputs to process. show_progress : bool or dict Show within-trial progress bar using tqdm progress bar. Can also be a **kwarg dict passed to tqdm. Returns ------- objective : `callable` Callable that takes trial as input and returns correspond loss. Tz Current trialFr)descleavepositionrFr;c jj}g}g}jjjj |}dk7r't ddt i}|jdtD]f\}} tj} t| tr| jdi} ni} || fi| } tj} |j| | z tj}|$|j| | }j|nddlm}|| d| ||  }tj}|j||z dk7rjd j"|j%|t'j(n t+|||j-sUt/j0dk7rj3|j5d t7||j5d t7||Ct t'j8d k(r |dx}x}}n0t;|d \\}\}}}}n|j=d \}\}}j>r||St+|SjjAdk(r|S|S#t$r}d}g}Yd}~d}~wwxYw)zCompute objective value Parameter --------- trial : `Trial` Current trial Returns ------- loss : `float` Loss NrEFtotalrpipeline_kwargs) get_annotated annotation)uemrprocessing_timeevaluation_timeg?)alphar=)!r get_metricNotImplementedErrorrKrHrlenupdate enumeratetimer*rgetappendlosspyannote.databaserVrreportrAmeanabs should_pruner- TrialPrunedclose set_user_attrsumuniquerconfidence_intervalr!r3)rFmetricr7lossesrYrZr progress_bariinputbefore_processingrUoutputafter_processingbefore_evaluationrerV_after_evaluationrh lower_bound upper_boundrMr5rNs r9 objectivez*Optimizer.get_objective..objectives 113 !O O}}001I1IPU1I1VWH%#G#f+GG ##A&&f-+ /5%)IIK! eW-&+ii0A2&FO&(O!%;?;#'99; &&'7:K'KL%)IIK!>#==7DMM$' @u\2F e@TUA#'99; &&'7:K'KL E) ''*;;& RWWV_CKQRS%%' ,,..W+ /Z%""$    133G H    133G H~ryy()Q.7=ay@D@;?Hc@<6T5K14:3M3MTW3M3X00{K  >Kv;&==..0J> ! g'  sK K3$K..K3)listr_rfloat)r5rMrNn_inputsr~s``` r9 get_objectivezOptimizer.get_objectivesK.fv; D %4uRSTMh Uh uh TrI n_iterations warm_startcd|j_|j||}|rn|jj|}t j 5t j dt|jj|ddd|jj||ddd|j_|j|jd S#1swYRxYw) aTune pipeline Parameters ---------- inputs : iterable List of inputs processed by the pipeline at each iteration. n_iterations : int, optional Number of iterations. Defaults to 10. warm_start : dict, optional Nested dictionary of initial parameters used to bootstrap tuning. Returns ------- result : dict ['loss'] ['params'] nested dictionary of optimal parameters TrNignorecategoryNrn_trialstimeoutn_jobsFreparams) rtrainingr_flattenr(catch_warningsfilterwarningsr r4 enqueue_trialoptimizerCrG)r5rMrrrNr~flattened_paramss r9tunezOptimizer.tune2s4"& &&v]&K #}}55jA ((* <'';NO ))*:; < YtTUV"' $2B2BCC <  C61D>7D 9D>D(D>6 D?DD>DD>D%!D>( D;1D>6D;;D>)NNNNNT)F) NT)NT)__name__ __module__ __qualname____doc__rrrr+r r r intboolr:propertyrrCdictrGrLrrr rrrrrrr\rIr9rr6s>"$(5937"!?)?) TN?)SM ?) %[ 012 ?) sJ/0 ?)sm?)?)B55T55 ;x;;,1G'GT4Z(G 5'5. ! GX+/ *D'*D*D *D T4Z( *D  *D^ +/ 3='3=3=T4Z( 3= 4t# $ 3=rIr)(rbr(pathlibrtypingrrrrrr r numpyrAoptuna.loggingr-optuna.prunersoptuna.samplersoptuna.exceptionsr r r r optuna.trialrroptuna.storagesrrrr scipy.statsrrrrlogging set_verbosityWARNINGrr\rIr9rsq< PPP1%3*JJJJ!!V^^334[=[=rI