§
    l¬¸jær  ã                  ó¨   — d Z ddlmZ ddlZddlZddlZddlmZ ddlZddl	m
Z
mZmZmZ ddlmZmZmZmZmZ ddlmZ ddlmZ  G d	„ d
¦  «        ZdS )a  
Module provides functionalities for hyperparameter tuning of the Ultralytics YOLO models for object detection, instance
segmentation, image classification, and pose estimation.

Hyperparameter tuning is the process of systematically searching for the optimal set of hyperparameters
that yield the best model performance. This is particularly crucial in deep learning models like YOLO,
where small changes in hyperparameters can lead to significant differences in model accuracy and efficiency.

Examples:
    Tune hyperparameters for YOLO26n on COCO8 at imgsz=640 and epochs=10 for 300 tuning iterations.
    >>> from ultralytics import YOLO
    >>> model = YOLO("yolo26n.pt")
    >>> model.tune(data="coco8.yaml", epochs=10, iterations=300, optimizer="AdamW", plots=False, save=False, val=False)
é    )ÚannotationsN)Údatetime)ÚCFG_INT_KEYSÚTASK2METRICÚget_cfgÚget_save_dir)ÚLOGGERÚYAMLÚ	callbacksÚcolorstrÚremove_colorstr)Úcheck_requirements)Úplot_tune_resultsc                  óè   — e Zd ZdZd=ddœd>d„Zd?d@d„ZdAd„Zed„ ¦   «         Z	 	 dBdCd„Z	dDd!„Z
d"„ ZdEd$„ZdFd'„ZdGd)„ZedHd*„¦   «         ZedIdJd.„¦   «         ZedKd0„¦   «         Z	 	 dLdMd5„ZdNd7„ZdOdPd<„ZdS )QÚTunera¢  A class for hyperparameter tuning of YOLO models.

    The class evolves YOLO model hyperparameters over a given number of iterations by mutating them according to the
    search space and retraining the model to evaluate their performance. Supports both local NDJSON storage and
    distributed MongoDB Atlas coordination for multi-machine hyperparameter optimization.

    Attributes:
        space (dict[str, tuple]): Hyperparameter search space containing bounds and scaling factors for mutation.
        tune_dir (Path): Directory where evolution logs and results will be saved.
        tune_file (Path): Path to the NDJSON file where evolution logs are saved.
        args (SimpleNamespace): Configuration arguments for the tuning process.
        model (torch.nn.Module): Base model whose weights seed each iteration.
        callbacks (dict): Callback functions to be executed during tuning.
        prefix (str): Prefix string for logging messages.
        mongodb (MongoClient): Optional MongoDB client for distributed tuning.
        collection (Collection): MongoDB collection for storing tuning results.

    Methods:
        _mutate: Mutate hyperparameters based on bounds and scaling factors.
        __call__: Execute the hyperparameter evolution across multiple iterations.

    Examples:
        Tune hyperparameters for YOLO26n on COCO8 at imgsz=640 and epochs=10 for 300 tuning iterations.
        >>> from ultralytics import YOLO
        >>> model = YOLO("yolo26n.pt")
        >>> model.tune(
        >>>     data="coco8.yaml",
        >>>     epochs=10,
        >>>     iterations=300,
        >>>     plots=False,
        >>>     save=False,
        >>>     val=False
        >>> )

        Tune with distributed MongoDB Atlas coordination across multiple machines:
        >>> model.tune(
        >>>     data="coco8.yaml",
        >>>     epochs=10,
        >>>     iterations=300,
        >>>     mongodb_uri="mongodb+srv://user:pass@cluster.mongodb.net/",
        >>>     mongodb_db="ultralytics",
        >>>     mongodb_collection="tune_results"
        >>> )

        Tune with custom search space:
        >>> model.tune(space={"lr0": (1e-5, 1e-2), "momentum": (0.7, 0.98)})
    N)ÚmodelÚ
_callbacksúdict | Nonec               óº  — |                      dd¦  «        p@i dd“dd“dd“d	d
“dd“dd“dd“dd“dd“dd“dd“dd“dd“dd“dd“dd“d d!“d
dddddddd!d"œ	¥| _        |                      d#d¦  «        }|                      d$d%¦  «        }|                      d&d'¦  «        }t          |¬(¦  «        | _        |€ d)d*lm}  || j        j        ¦  «        j        }|| _        | j        j        | j        _        t          | j        | j        j
        pd+¬,¦  «        | _        d-\  | j        _
        | j        _        | j        _        | j        d.z  | _        |pt          j        ¦   «         | _        t          d/¦  «        | _        t          j        | ¦  «         d| _        |r|                      |||¦  «         t)          j        | j        › d0| j        › d1| j        › d2�¦  «         dS )3a]  Initialize the Tuner with configurations.

        Args:
            args (dict): Configuration for hyperparameter evolution.
            _callbacks (dict | None, optional): Callback functions to be executed during tuning.
            model (torch.nn.Module, optional): Base model whose weights seed each iteration, built from args if None.
        ÚspaceNÚlr0)çñhãˆµøä>ç{®Gáz„?Úlrf)r   ç      ð?Úmomentum)gffffffæ?g\�Âõ(\ï?g333333Ó?Úweight_decay)ç        gü©ñÒMbP?Úwarmup_epochs)r   g      @Úwarmup_momentum)r   gffffffî?Úbox)r   g      4@Úcls)çš™™™™™¹?g      @Úcls_pw)r   r   Údfl)çš™™™™™Ù?g      (@Úhsv_h)r   r#   Úhsv_s)r   gÍÌÌÌÌÌì?Úhsv_vÚdegrees)r   g     €F@Ú	translateÚscaleÚshear)r   g      $@)	ÚperspectiveÚflipudÚfliplrÚbgrÚmosaicÚmixupÚcutmixÚ
copy_pasteÚclose_mosaicÚmongodb_uriÚ
mongodb_dbÚultralyticsÚmongodb_collectionÚtuner_results)Ú	overridesr   )ÚYOLOÚtune)Úname)NFFztune_results.ndjsonzTuner: z*Initialized Tuner instance with 'tune_dir=z'
uT   ðŸ’¡ Learn about tuning at https://docs.ultralytics.com/guides/hyperparameter-tuning)Úpopr   r   Úargsr9   r=   r   ÚresumeÚexist_okr   r?   Útune_dirÚ	tune_filer   Úget_default_callbacksr   ÚprefixÚadd_integration_callbacksÚmongodbÚ_init_mongodbr	   Úinfo)ÚselfrA   r   r   r7   r8   r:   r=   s           úh/root/private/project/pallet_detection_v2/.venv/lib/python3.11/site-packages/ultralytics/engine/tuner.pyÚ__init__zTuner.__init__Q   sª  € ð —X’X˜g tÑ,Ô,ð 
ð 1
à�<ð1
ð �;ð1
ð Ð(ð	1
ð
 ˜Lð1
ð ˜Zð1
ð ˜{ð1
ð �;ð1
ð �:ð1
ð �jð1
ð �;ð1
ð �Zð1
ð �Zð1
ð �Zð1
ð �{ð1
ð  ˜ð!1
ð" �[ð#1
ð$ �[ð%1
ð& (Ø Ø ØØ ØØ Ø$Ø'ð71
ð 1
ð 1
ˆŒ
ð: —h’h˜}¨dÑ3Ô3ˆØ—X’X˜l¨MÑ:Ô:ˆ
Ø!ŸXšXÐ&:¸OÑLÔLÐå dÐ+Ñ+Ô+ˆŒ	Øˆ=Ø(Ð(Ð(Ð(Ð(Ð(à�D˜œœÑ)Ô)Ô/ˆEØˆŒ
Ø!œYÔ-ˆŒ	ÔÝ$ T¤Y°T´Y´^Ð5MÀvÐNÑNÔNˆŒØ?SÑ<ˆŒ	Œ˜œ	Ô*¨D¬IÔ,<ØœÐ)>Ñ>ˆŒØ#ÐH¥yÔ'FÑ'HÔ'HˆŒÝ˜yÑ)Ô)ˆŒÝÔ+¨DÑ1Ô1Ð1ð ˆŒØð 	LØ×Ò˜{¨JÐ8JÑKÔKÐKåŒØŒ{ð qð qÀdÄmð qð qØŒ{ðqð qð qñ	
ô 	
ð 	
ð 	
ð 	
ó    Ú é   ÚuriÚstrÚmax_retriesÚintc                ó¶  — t          d¦  «         ddlm} ddlm}m} t          |¦  «        D ]«}	  ||dddddd	d
d¬¦	  «	        }|j                             d¦  «         t          j
        | j        › d|dz   › d�¦  «         |c S # ||f$ rK ||dz
  k    r‚ d|z  }t          j        | j        › d|dz   › d|› d�¦  «         t          j        |¦  «         Y Œ¨w xY wdS )af  Create MongoDB client with exponential backoff retry on connection failures.

        Args:
            uri (str): MongoDB connection string with credentials and cluster information.
            max_retries (int): Maximum number of connection attempts before giving up.

        Returns:
            (MongoClient): Connected MongoDB client instance.
        Úpymongor   )ÚMongoClient)ÚConnectionFailureÚServerSelectionTimeoutErrori0u  i N  i@œ  Té   rQ   i`ê  )ÚserverSelectionTimeoutMSÚconnectTimeoutMSÚsocketTimeoutMSÚretryWritesÚ
retryReadsÚmaxPoolSizeÚminPoolSizeÚmaxIdleTimeMSÚpingz$Connected to MongoDB Atlas (attempt é   ú)é   z#MongoDB connection failed (attempt z), retrying in zs...N)r   rW   rX   Úpymongo.errorsrY   rZ   ÚrangeÚadminÚcommandr	   rK   rG   ÚwarningÚtimeÚsleep)	rL   rR   rT   rX   rY   rZ   ÚattemptÚclientÚ	wait_times	            rM   Ú_connectzTuner._connect’   sj  € õ 	˜9Ñ%Ô%Ð%à'Ð'Ð'Ð'Ð'Ð'ØQÐQÐQÐQÐQÐQÐQÐQå˜[Ñ)Ô)ð 	&ð 	&ˆGð&Ø$˜ØØ-2Ø%*Ø$)Ø $Ø#Ø "Ø !Ø"'ð
ñ 
ô 
�ð ”×$Ò$ VÑ,Ô,Ð,Ý”˜tœ{Ð^Ð^ÐPWÐZ[ÑP[Ð^Ð^Ð^Ñ_Ô_Ð_Ø���øØ%Ð'BÐCð &ð &ð &Ø˜k¨A™oÒ-Ð-ØØ˜w™J�	Ý”Ø”{ÐrÐrÀwÐQRÁ{ÐrÐrÐclÐrÐrÐrñô ð õ ”
˜9Ñ%Ô%Ð%Ð%Ð%ð&øøøð!	&ð 	&s   °ABÂACÃCc                ó¢   — |                       |¦  «        | _        | j        |         |         | _        t          j        | j        › d�¦  «         dS )aØ  Initialize MongoDB connection for distributed tuning.

        Connects to MongoDB Atlas for distributed hyperparameter optimization across multiple machines. Each worker
        saves results to a shared collection and reads the latest best hyperparameters from all workers for evolution.

        Args:
            mongodb_uri (str): MongoDB connection string.
            mongodb_db (str, optional): Database name.
            mongodb_collection (str, optional): Collection name.

        Notes:
            - Creates a fitness index when workers start a new collection
            - Falls back to local NDJSON mode if connection fails
            - Uses connection pooling and retry logic for production reliability
        z*Using MongoDB Atlas for distributed tuningN)rr   rI   Ú
collectionr	   rK   rG   )rL   r7   r8   r:   s       rM   rJ   zTuner._init_mongodbº   sL   € ð  —}’} [Ñ1Ô1ˆŒØœ, zÔ2Ð3EÔFˆŒÝŒ�t”{ÐNÐNÐNÑOÔOÐOÐOÐOrO   c                óh   — t          | d¦  «        r|                      ¦   «         nt          | ¦  «        S )z2Convert tensor-like values for JSON serialization.Úitem)Úhasattrrv   rS   )Úxs    rM   Ú_json_defaultzTuner._json_defaultÎ   s+   € õ # 1 fÑ-Ô-Ð9ˆq�vŠv‰xŒxˆxµ3°q±6´6Ð9rO   Ú	iterationÚfitnessÚfloatÚhyperparametersúdict[str, float]Údatasetsúdict[str, dict]Ú	save_dirsúdict[str, str] | NoneÚfailed_datasetsúlist[str] | NoneÚreturnÚdictc                óL   — |t          |d¦  «        ||dœ}|r||d<   |r||d<   |S )z%Build one local tuning result record.é   )rz   r{   r}   r   r�   rƒ   )Úround)rL   rz   r{   r}   r   r�   rƒ   Úresults           rM   Ú_result_recordzTuner._result_recordÓ   sR   € ð #Ý˜W aÑ(Ô(Ø.Ø ð	
ð 
ˆð ð 	,Ø"+ˆF�;ÑØð 	8Ø(7ˆFÐ$Ñ%ØˆrO   Úmetricsúdict[str, str]ú	list[str]c                ó†  — 	 | j                              |d„ |                     ¦   «         D ¦   «         ||||t          j        ¦   «                              ¦   «         | j                              ddidddiid¬¦  «        d         d	œ¦  «         dS # t          $ r)}t          j	        | j
        › d
|› �¦  «         Y d}~dS d}~ww xY w)a\  Save results to MongoDB with proper type conversion.

        Args:
            fitness (float): Fitness score achieved with these hyperparameters.
            hyperparameters (dict[str, float]): Dictionary of hyperparameter values.
            metrics (dict): Complete training metrics dictionary (mAP, precision, recall, losses, etc.).
            datasets (dict[str, dict]): Per-dataset metrics for the iteration.
            save_dirs (dict[str, str]): Per-dataset training directories for cleanup.
            failed_datasets (list[str]): Dataset runs that did not produce training metrics.
        c                ób   — i | ],\  }}|t          |d ¦  «        r|                     ¦   «         n|“Œ-S )rv   )rw   rv   ©Ú.0ÚkÚvs      rM   ú
<dictcomp>z*Tuner._save_to_mongodb.<locals>.<dictcomp>   s=   € Ð'wÐ'wÐ'wÑW[ÐWXÐZ[¨½ÀÀ6Ñ8JÔ8JÐ,Q¨A¯FªF©H¬H¨HÐPQÐ'wÐ'wÐ'wrO   Ú_idÚdefaultsz$incÚlast_iterationre   T)Úreturn_document)r{   r}   rŒ   r   r�   rƒ   Ú	timestamprz   zMongoDB save failed: N)rt   Ú
insert_oneÚitemsr   ÚnowÚ
astimezoneÚfind_one_and_updateÚ	Exceptionr	   rl   rG   )rL   r{   r}   rŒ   r   r�   rƒ   Úes           rM   Ú_save_to_mongodbzTuner._save_to_mongodbé   s
  € ð&	EØŒO×&Ò&à&Ø'wÐ'wÐ_n×_tÒ_tÑ_vÔ_vÐ'wÑ'wÔ'wØ&Ø (Ø!*Ø'6Ý!)¤¡¤×!:Ò!:Ñ!<Ô!<Ø!%¤×!DÒ!DØ 
Ð+¨fÐ7GÈÐ6KÐ-LÐ^bð "Eñ "ô "à&ô"(ðð ñô ð ð ð øõ ð 	Eð 	Eð 	EÝŒN˜dœkÐCÐCÀÐCÐCÑDÔDÐDÐDÐDÐDÐDÐDÐDøøøøð	Eøøøs   ‚B	B Â
C ÂB;Â;C c                ó  ‡ — 	 t          ‰ j                             dddii¦  «                             dd¦  «        ¦  «        }|sdS t	          d„ |D ¦   «         ¦  «        }‰ j                             dd	id
d|iid¬¦  «         t          ‰ j        dd¬¦  «        5 }|                     ˆ fd„|D ¦   «         ¦  «         ddd¦  «         dS # 1 swxY w Y   dS # t          $ r)}t          j        ‰ j        › d|› �¦  «         Y d}~dS d}~ww xY w)a  Sync MongoDB results to the local NDJSON tuning log.

        Downloads all results from MongoDB and writes them to the local NDJSON file in chronological order. This keeps
        resume, mutation, and plotting on the same local source of truth when using distributed tuning.
        r{   ú$existsTrz   re   Nc              3  ó&   K  — | ]}|d          V — ŒdS )rz   N© )r’   Úrs     rM   ú	<genexpr>z.Tuner._sync_mongodb_to_file.<locals>.<genexpr>  s&   è è € Ð EÐ E°A  ;¤Ð EÐ EÐ EÐ EÐ EÐ ErO   r–   r—   z$maxr˜   )ÚupsertÚwúutf-8©Úencodingc              3  ó<  •K  — | ]–}t          j        ‰                     |d          |d         pd|                     di ¦  «        |                     di ¦  «        |                     d¦  «        |                     d¦  «        ¦  «        ‰j        ¬¦  «        dz   V — Œ—d	S )
rz   r{   r   r}   r   r�   rƒ   ©Údefaultú
N)ÚjsonÚdumpsr‹   Úgetry   )r’   rŠ   rL   s     €rM   r¨   z.Tuner._sync_mongodb_to_file.<locals>.<genexpr>  sÆ   øè è € ð ð ð õ ”JØ×+Ò+Ø" ;Ô/Ø" 9Ô-Ð4°Ø"ŸJšJÐ'8¸"Ñ=Ô=Ø"ŸJšJ z°2Ñ6Ô6Ø"ŸJšJ {Ñ3Ô3Ø"ŸJšJÐ'8Ñ9Ô9ñô ð !%Ô 2ð
ñ 
ô 
ð ñðð ð ð ð ð rO   zMongoDB to NDJSON sync failed: )Úlistrt   ÚfindÚsortÚmaxÚ
update_oneÚopenrE   Ú
writelinesr    r	   rl   rG   )rL   Úall_resultsr˜   Úfr¡   s   `    rM   Ú_sync_mongodb_to_filezTuner._sync_mongodb_to_file  s§  ø€ ð	OÝ˜tœ×3Ò3°YÀÈDÐ@QÐ4RÑSÔS×XÒXÐYdÐfgÑhÔhÑiÔiˆKØð Ø�Ý Ð EÐ E¸Ð EÑ EÔ EÑEÔEˆNØŒO×&Ò&¨¨zÐ':¸VÐFVÐXfÐEgÐ<hÐquÐ&ÑvÔvÐvå�d”n c°GÐ<Ñ<Ô<ð ÀØ—’ð ð ð ð ð #.ðñ ô ñ ô ð ðð ð ñ ô ð ð ð ð ð ð ð øøøð ð ð ð ð ð øõ" ð 	Oð 	Oð 	OÝŒN˜dœkÐMÐMÈ!ÐMÐMÑNÔNÐNÐNÐNÐNÐNÐNÐNøøøøð	OøøøsC   ƒAC ÁAC Â"CÂ;C ÃCÃC ÃCÃC Ã
DÃDÄDú
list[dict]c                ó°   — | j                              ¦   «         sg S t          | j         d¬¦  «        5 }d„ |D ¦   «         cddd¦  «         S # 1 swxY w Y   dS )z.Load local tuning results from the NDJSON log.r«   r¬   c                ó^   — g | ]*}|                      ¦   «         ¯t          j        |¦  «        ‘Œ+S r¦   )Ústripr²   Úloads)r’   Úlines     rM   ú
<listcomp>z-Tuner._load_local_results.<locals>.<listcomp>4  s/   € ÐCÐCÐC¨°d·j²j±l´lÐC•D”J˜tÑ$Ô$ÐCÐCÐCrO   N)rE   Úexistsrº   )rL   r½   s     rM   Ú_load_local_resultszTuner._load_local_results/  s·   € àŒ~×$Ò$Ñ&Ô&ð 	ØˆIÝ�$”.¨7Ð3Ñ3Ô3ð 	D°qØCÐC°ÐCÑCÔCð	Dð 	Dð 	Dð 	Dñ 	Dô 	Dð 	Dð 	Dð 	Dð 	Dð 	Dð 	Døøøð 	Dð 	Dð 	Dð 	Dð 	Dð 	Ds   ²AÁAÁAÚresultsúnp.ndarray | Nonec                óZ   ‡ — |sdS t          j        ˆ fd„|D ¦   «         t          ¬¦  «        S )zKConvert local NDJSON records to a fitness-plus-hyperparameters numpy array.Nc                ój   •‡— g | ].Š‰                      d d¦  «        gˆˆfd„‰j        D ¦   «         z   ‘Œ/S )r{   r   c           	     óŠ   •— g | ]?}‰                      d i ¦  «                              |t          ‰j        |¦  «        ¦  «        ‘Œ@S ©r}   )r´   ÚgetattrrA   ©r’   r“   r§   rL   s     €€rM   rÅ   z<Tuner._local_results_to_array.<locals>.<listcomp>.<listcomp>=  sE   ø€ ÐbÐbÐbÐRS�1—5’5Ð*¨BÑ/Ô/×3Ò3°Aµw¸t¼yÈ!Ñ7LÔ7LÑMÔMÐbÐbÐbrO   )r´   r   ©r’   r§   rL   s    @€rM   rÅ   z1Tuner._local_results_to_array.<locals>.<listcomp>;  sa   øø€ ð ð ð ð ð —’�y #Ñ&Ô&Ð'ØbÐbÐbÐbÐbÐW[ÔWaÐbÑbÔbñcðð ð rO   )Údtype)ÚnpÚarrayr|   )rL   rÈ   s   ` rM   Ú_local_results_to_arrayzTuner._local_results_to_array6  sV   ø€ àð 	Ø�4ÝŒxðð ð ð ð !ðñ ô õ
 ð
ñ 
ô 
ð 	
rO   rŠ   c                óÈ   — t          | j        dd¬¦  «        5 }|                     t          j        || j        ¬¦  «        dz   ¦  «         ddd¦  «         dS # 1 swxY w Y   dS )z1Append one tuning result to the local NDJSON log.Úar«   r¬   r¯   r±   N)rº   rE   Úwriter²   r³   ry   )rL   rŠ   r½   s      rM   Ú_save_local_resultzTuner._save_local_resultC  s¶   € å�$”. #°Ð8Ñ8Ô8ð 	K¸AØ�GŠG•D”J˜v¨tÔ/AÐBÑBÔBÀTÑIÑJÔJÐJð	Kð 	Kð 	Kñ 	Kô 	Kð 	Kð 	Kð 	Kð 	Kð 	Kð 	Kð 	Køøøð 	Kð 	Kð 	Kð 	Kð 	Kð 	Ks   ˜2AÁAÁAc                ó  — |                       di ¦  «        }t          |¦  «        dk    r.t          t          |                     ¦   «         ¦  «        ¦  «        S t          |¦  «        dk    rd„ |                     ¦   «         D ¦   «         S dS )z*Summarize best-result metrics for logging.r   re   c                ó`   — i | ]+\  }}|t          |                     d ¦  «        pdd¦  «        “Œ,S )r{   r   rˆ   )r‰   r´   r‘   s      rM   r•   z'Tuner._best_metrics.<locals>.<dictcomp>O  s9   € ÐVÐVÐV¹T¸QÀ�A•u˜QŸUšU 9Ñ-Ô-Ð4°°aÑ8Ô8ÐVÐVÐVrO   N)r´   ÚlenÚnextÚiterÚvaluesrœ   )rŠ   r   s     rM   Ú_best_metricszTuner._best_metricsH  s|   € ð —:’:˜j¨"Ñ-Ô-ˆÝˆx‰=Œ=˜AÒÐÝ�˜XŸ_š_Ñ.Ô.Ñ/Ô/Ñ0Ô0Ð0Ýˆx‰=Œ=˜1ÒÐØVÐVÀXÇ^Â^ÑEUÔEUÐVÑVÔVÐVØˆtrO   FÚrequire_allÚboolc                ó  ‡— |                       di ¦  «        }t          |                       dd¦  «        ¦  «        Šˆfd„|                     ¦   «         D ¦   «         }t          |¦  «        o|rt	          |¦  «        nt          |¦  «        S )z9Return whether a tuning result contains training metrics.r   rƒ   r¦   c                ó<   •— g | ]\  }}t          |¦  «        o|‰v‘ŒS r¦   )rá   )r’   ÚdatasetrŒ   Úfaileds      €rM   rÅ   z/Tuner._has_training_metrics.<locals>.<listcomp>W  s.   ø€ ÐdÐdÐdÑ?O¸wÈ•4˜‘=”=Ð: W°FÐ%:ÐdÐdÐdrO   )r´   Úsetrœ   rá   ÚallÚany)rŠ   rà   r   Útrainedrå   s       @rM   Ú_has_training_metricszTuner._has_training_metricsR  s€   ø€ ð —:’:˜j¨"Ñ-Ô-ˆÝ�V—Z’ZÐ 1°2Ñ6Ô6Ñ7Ô7ˆØdÐdÐdÐdÐS[×SaÒSaÑScÔScÐdÑdÔdˆÝ�G‰}Œ}ÐP°+Ð"O¥# g¡,¤, ,Å3ÀwÁ<Ä<ÐPrO   ú
np.ndarrayc                óÚ   ‡ — ˆ fd„t          |¦  «        D ¦   «         }|r-|t          ||                              ¦   «         ¦  «                 n t          |                     ¦   «         ¦  «        S )zDReturn the best result index, preferring rows with training metrics.c                óD   •— g | ]\  }}‰                      |¦  «        ¯|‘ŒS r¦   )rê   )r’   ÚirŠ   r"   s      €rM   rÅ   z,Tuner._best_result_index.<locals>.<listcomp>]  s1   ø€ Ð\Ð\Ð\‘y�q˜&¸#×:SÒ:SÐTZÑ:[Ô:[Ð\�Ð\Ð\Ð\rO   )Ú	enumeraterU   Úargmax)r"   rÈ   r{   Úvalids   `   rM   Ú_best_result_indexzTuner._best_result_indexZ  sf   ø€ ð ]Ð\Ð\Ð\¥I¨gÑ$6Ô$6Ð\Ñ\Ô\ˆØ6;ÐVˆu•S˜ œ×.Ò.Ñ0Ô0Ñ1Ô1Ô2Ð2ÅÀWÇ^Â^ÑEUÔEUÑAVÔAVÐVrO   é	   çš™™™™™É?ÚnÚsigmac           
     óÜ  ‡ ‡‡‡— d}‰ j         �r*t          ‰ j                             dddiidddœ¦  «                             dd¦  «        ¦  «        x}r!t          j        ˆ fd„|D ¦   «         ¦  «        }nÄd	d
lm} ‰  	                    ˆ fd„‰ j
        D ¦   «         ¦  «        }	 ‰ j                             dt          j        ¦   «                              ¦   «         dœ¦  «         n5# |$ r- t          j        dg|                     ¦   «         ¢g¦  «        }Y nw xY w‰ j                             dgd¬¦  «         |€|S |€)‰                      ¦   «         }‰                      |¦  «        }|��ßt
          j                             ¦   «         }t+          ‰ j
        ¦  «        }t          j        |dd…d	f         d¦  «        }	t+          |¦  «        dz
  t/          t          j        |	¦  «        ¦  «        z
  }
t          j        |dd…d	f          ¦  «        }||         d|…         }t          j        d„ ‰ j
                             ¦   «         D ¦   «         ¦  «        Št          j        ‰d¬¦  «        Š‰d	k    }|                     ¦   «         �r®t          j        |dd…dd…f         ‰dd…d	f         z
  ‰t          j        |dd…dd…f         ¦  «        |¬¦  «        }|dd…d	f         |dd…d	f                              ¦   «         z
  dz   }t          j        |¦  «                              ¦   «         r| !                    ¦   «         r|nt          j"        |¦  «        }t          j        d„ ‰ j
                             ¦   «         D ¦   «         ¦  «        }t          j        d„ ‰ j
        D ¦   «         ¦  «        }ddt=          |
dz  d¦  «        z  z
  }||z  |z  }t          j#        |t          j        |‰t          j$        |¦  «        |¬¦  «        ¦  «        }d„ |D ¦   «         }dx}}t+          |¦  «        dk    �r
t=          t/          t          j%        t+          |¦  «        dz  ¦  «        ¦  «        d¦  «        }t=          || !                    ¦   «         z  d¦  «        }t          j        ||d|…         dd…f         ‰dd…d	f         z
  ‰t          j$        ||f¦  «        |¬¦  «        }t          j&        |d¬¦  «        |dz  z  |z  t          j'        t          j(        |¦  «        | !                    ¦   «         z  ¦  «        z   }tS          d¦  «        D �]Õ}|d k     ró|| *                    t+          |¦  «        || !                    ¦   «         z  ¬!¦  «                 Š|                     |¦  «        d"k     |z  }|�c|                     ¦   «         d#|z  k     rHt          j+        || ,                    ‰|¦  «        ‰¦  «        Šdt          j-        ‰dz  dz
  ¦  «        z
  Šn®t          j.        ‰|| /                    |¦  «        z  |z  z   d	d¦  «        Šn{|| *                    t+          |¦  «        || !                    ¦   «         z  ¬!¦  «                  0                    ¦   «         Š|                     | !                    ¦   «         ¦  «        ‰|<   ˆˆˆfd$„tc          ‰ j
        ¦  «        D ¦   «         }‰  	                    |¦  «        }te          |                     ¦   «         ¦  «        |vr|c S �Œ×tg          ‰ j4        › d%�¦  «        ‚tg          ‰ j4        › d&�¦  «        ‚ˆ fd'„‰ j
        D ¦   «         }‰  	                    |¦  «        S )(aS  Mutate hyperparameters based on bounds and scaling factors specified in `self.space`.

        Args:
            n (int): Number of top parents to consider.
            sigma (float): Initial normalized mutation standard deviation.

        Returns:
            (dict[str, float]): A dictionary containing mutated hyperparameters.
        Nr{   r¤   Tre   )r{   r}   r–   c                óN   •‡— g | ] Š‰d          gˆˆfd„‰j         D ¦   «         z   ‘Œ!S )r{   c                óx   •— g | ]6}‰d                                |‰j                              |¦  «        ¦  «        ‘Œ7S rÍ   )r´   rA   rÏ   s     €€rM   rÅ   z,Tuner._mutate.<locals>.<listcomp>.<listcomp>y  s?   ø€ Ð)lÐ)lÐ)lÐ\]¨!Ð,=Ô*>×*BÒ*BÀ1ÀdÄiÇmÂmÐTUÑFVÔFVÑ*WÔ*WÐ)lÐ)lÐ)lrO   )r   rÐ   s    @€rM   rÅ   z!Tuner._mutate.<locals>.<listcomp>x  sQ   øø€ ð ð ð àð ˜9œ˜Ð)lÐ)lÐ)lÐ)lÐ)lÐaeÔakÐ)lÑ)lÔ)lÑlðð ð rO   r   )ÚDuplicateKeyErrorc                ó<   •— i | ]}|t          ‰j        |¦  «        “ŒS r¦   ©rÎ   rA   ©r’   r“   rL   s     €rM   r•   z!Tuner._mutate.<locals>.<dictcomp>€  s'   ø€ Ð.\Ð.\Ð.\ÈA¨qµ'¸$¼)ÀQÑ2GÔ2GÐ.\Ð.\Ð.\rO   r—   )r–   rš   r   )r{   éÿÿÿÿ)Ú
backgroundrˆ   c                ó"   — g | ]}|d d…         ‘ŒS )Nrg   r¦   ©r’   r”   s     rM   rÅ   z!Tuner._mutate.<locals>.<listcomp>–  s    € ÐBÐBÐB¨˜q  ! œuÐBÐBÐBrO   )Úaxis)ÚoutÚwhereg�íµ ÷Æ°>c                óH   — g | ]}t          |¦  «        d k    r|d         nd‘Œ S )rQ   rg   r   )rÛ   r  s     rM   rÅ   z!Tuner._mutate.<locals>.<listcomp>�  s.   € Ð!\Ð!\Ð!\À1­#¨a©&¬&°Aª+¨+ ! A¤$ $¸3Ð!\Ð!\Ð!\rO   c                ó(   — g | ]}|t           v rd nd‘ŒS )re   r   )r   )r’   r“   s     rM   rÅ   z!Tuner._mutate.<locals>.<listcomp>ž  s&   € Ð&\Ð&\Ð&\ÈA¨AµÐ,=Ð,= q qÀ4Ð&\Ð&\Ð&\rO   rô   é   c                ó<   — h | ]}t          |d d…         ¦  «        ’ŒS )re   N)Útuple)r’   Úrows     rM   ú	<setcomp>z Tuner._mutate.<locals>.<setcomp>¢  s&   € Ð>Ð>Ð>¨s�E # a b b¤'™NœNÐ>Ð>Ð>rO   r[   F)Úrowvarrg   éÈ   éd   )Úpg      à?r&   c           	     ój   •— i | ]/\  }}|t          ‰|d f         ‰|         ‰|         z  z   ¦  «        “Œ0S )r   )r|   )r’   rî   r“   ÚboundsÚgenesÚspans      €€€rM   r•   z!Tuner._mutate.<locals>.<dictcomp>¼  sA   ø€ ÐiÐiÐiÉ4È1Èa˜1�e F¨1¨a¨4¤L°5¸´8¸dÀ1¼gÑ3EÑ$EÑFÔFÐiÐiÐirO   z3Unable to generate a unique hyperparameter mutationz(Hyperparameter search space is exhaustedc                ó<   •— i | ]}|t          ‰j        |¦  «        “ŒS r¦   rü   rý   s     €rM   r•   z!Tuner._mutate.<locals>.<dictcomp>Ã  s'   ø€ Ð@Ð@Ð@°�1•g˜dœi¨Ñ+Ô+Ð@Ð@Ð@rO   )5rI   rµ   rt   r¶   r·   rÒ   rÓ   rh   rú   Ú
_constrainr   r›   r   r�   rž   rÞ   Úcreate_indexrÇ   rÔ   ÚrandomÚdefault_rngrÛ   r‰   rU   rð   ÚargsortÚptprè   ÚdivideÚ
zeros_likeÚminÚisfiniterç   ÚsumÚ	ones_likeÚmaximumÚzerosÚceilÚcovÚdiagÚsquareri   Úchoicer  Úmultivariate_normalÚabsÚclipÚstandard_normalÚcopyrï   r	  ÚRuntimeErrorrG   )rL   rõ   rö   ÚhistoryrÈ   rú   Údefault_hypÚrngÚngr{   ÚstaleÚorderrx   ÚmutableÚ
populationÚweightsÚgainsÚ
resolutionÚdecayr,   ÚexistingÚ
covarianceÚ
confidenceÚn_eliteÚelitero   ÚmaskÚhypr  r  r  s   `                           @@@rM   Ú_mutatezTuner._mutate`  sa  øøøø€ ð ˆð Œ<ñ 	'ÝØ”×$Ò$ i°)¸TÐ1BÐ%CÐQRÐghÐEiÐEiÑjÔj×oÒoØ˜1ñô ñô ð ˆwð 'õ
 œ(ðð ð ð à!(ðñ ô ñô ��ð =Ð<Ð<Ð<Ð<Ð<à"ŸošoÐ.\Ð.\Ð.\Ð.\ÐQUÔQ[Ð.\Ñ.\Ô.\Ñ]Ô]�ðGØ”O×.Ò.°zÕPXÔP\ÑP^ÔP^×PiÒPiÑPkÔPkÐ/lÐ/lÑmÔmÐmÐmøØ(ð Gð Gð GÝ œh¨Ð(D¨{×/AÒ/AÑ/CÔ/CÐ(DÐ'EÑFÔF�G�G�GðGøøøà”×,Ò,¨oÐ->È4Ð,ÑPÔPÐPØ�?Ø&Ð&ð ˆ?Ø×.Ò.Ñ0Ô0ˆGØ×2Ò2°7Ñ;Ô;ˆGð ÑÝ”)×'Ò'Ñ)Ô)ˆCÝ�T”Z‘”ˆBÝ”h˜w q q q¨! tœ}¨aÑ0Ô0ˆGÝ˜‘L”L 1Ñ$¥s­2¬9°WÑ+=Ô+=Ñ'>Ô'>Ñ>ˆEÝ”J ¨¨¨¨1¨¤˜~Ñ.Ô.ˆEØ˜”˜r ˜rÔ"ˆAÝ”XÐBÐB¨d¬j×.?Ò.?Ñ.AÔ.AÐBÑBÔBÑCÔCˆFÝ”6˜& qÐ)Ñ)Ô)ˆDØ˜Q’hˆGØ�{Š{‰}Œ}ñ 'hÝœY q¨¨¨¨A¨B¨B¨¤x°&¸¸¸¸A¸´,Ñ'>ÀÍ"Ì-ÐXYÐZ[ÐZ[ÐZ[Ð]^Ð]_Ð]_ÐZ_ÔX`ÑJaÔJaÐipÐqÑqÔq�
Ø˜A˜A˜A˜q˜Dœ' A a a a¨ d¤G§K¢K¡M¤MÑ1°DÑ8�Ý%'¤[°Ñ%9Ô%9×%=Ò%=Ñ%?Ô%?ÐlÀGÇKÂKÁMÄMÐl˜'˜'ÕWYÔWcÐdkÑWlÔWl�ÝœÐ!\Ð!\ÈÌ
×HYÒHYÑH[ÔH[Ð!\Ñ!\Ô!\Ñ]Ô]�ÝœXÐ&\Ð&\ÐQUÔQ[Ð&\Ñ&\Ô&\Ñ]Ô]�
Ø˜C¥# e¨b¡j°!Ñ"4Ô"4Ñ4Ñ4�Ø ™¨Ñ-�Ýœ
 5­"¬)°JÀÍ"Ì(ÐSUÉ,Ì,Ð^eÐ*fÑ*fÔ*fÑgÔg�Ø>Ð>°gÐ>Ñ>Ô>�Ø*.Ð.�
˜ZÝ�w‘<”< 2Ò%Ñ%Ý!¥#¥b¤g­c°'©l¬l¸SÑ.@Ñ&AÔ&AÑ"BÔ"BÀBÑGÔG�GÝ!$ W¨w¯{ª{©}¬}Ñ%<¸aÑ!@Ô!@�JÝœIØ  h w h¤°°°Ð 3Ô4°v¸a¸a¸aÀ¸d´|ÑCØÝœH g¨r ]Ñ3Ô3Ø%ð	ñ ô �Eõ "$¤¨°eÐ!<Ñ!<Ô!<¸uÀa¹xÑ!GÈ*Ñ!TÕWYÔW^Ýœ	 %Ñ(Ô(¨7¯;ª;©=¬=Ñ8ñXô Xñ "�Jõ  % S™zœzð #ñ #�GØ ’}�}Ø *¨3¯:ª:µc¸!±f´fÀÈ'Ï+Ê+É-Ì-Ñ@W¨:Ñ+XÔ+XÔ Y˜Ø #§
¢
¨2¡¤°Ò 4¸Ñ?˜Ø%Ð1°c·j²j±l´lÀSÈ:ÑEUÒ6UÐ6UÝ$&¤H¨T°3×3JÒ3JÈ5ÐR\Ñ3]Ô3]Ð_dÑ$eÔ$e˜EØ$%­¬¨u°q©y¸1©}Ñ(=Ô(=Ñ$=˜E˜Eå$&¤G¨E°D¸3×;NÒ;NÈrÑ;RÔ;RÑ4RÐUZÑ4ZÑ,ZÐ\]Ð_`Ñ$aÔ$a˜E˜Eà *¨3¯:ª:µc¸!±f´fÀÈ'Ï+Ê+É-Ì-Ñ@W¨:Ñ+XÔ+XÔ Y× ^Ò ^Ñ `Ô `˜Ø),¯ª°G·K²K±M´MÑ)BÔ)B˜˜g™ØiÐiÐiÐiÐiÐiÕS\Ð]aÔ]gÑShÔShÐiÑiÔi�CØŸ/š/¨#Ñ.Ô.�CÝ˜SŸZšZ™\œ\Ñ*Ô*°(Ð:Ð:Ø"˜
˜
˜
ñ ;å" d¤kÐ#fÐ#fÐ#fÑgÔgÐgÝ $¤+ÐWÐWÐWÑXÔXÐXà@Ð@Ð@Ð@°T´ZÐ@Ñ@Ô@ˆCà�Š˜sÑ#Ô#Ð#s   Â"A C# Ã#/DÄDr@  c           	     ó  — | j                              ¦   «         D ]ò\  }}|t          v r£t          t	          j        |d         ¦  «        ¦  «        t          t	          j        |d         ¦  «        ¦  «        }}||k    rt          | j        › d|› d�¦  «        ‚t          t          t          ||         ¦  «        |¦  «        |¦  «        ||<   Œ±t          t          t          ||         d¦  «        |d         ¦  «        |d         ¦  «        ||<   Œó|S )zFConstrain hyperparameters to their search bounds and configured types.r   re   zSearch space for 'z' contains no integer valuesrˆ   )r   rœ   r   rU   rÒ   r#  ÚfloorÚ
ValueErrorrG   r  r¸   r‰   )rL   r@  r“   r  ÚlowerÚuppers         rM   r  zTuner._constrainÇ  sò   € àœ×)Ò)Ñ+Ô+ð 	Jð 	J‰IˆAˆvØ•LÐ Ð Ý"¥2¤7¨6°!¬9Ñ#5Ô#5Ñ6Ô6½½B¼HÀVÈAÄYÑ<OÔ<OÑ8PÔ8P�u�Ø˜5’=�=Ý$¨¬Ð%fÐ%fÀqÐ%fÐ%fÐ%fÑgÔgÐgÝ�S¥ s¨1¤v¡¤°Ñ6Ô6¸Ñ>Ô>��A‘�å�S¥ s¨1¤v¨qÑ!1Ô!1°6¸!´9Ñ=Ô=¸vÀa¼yÑIÔI��A‘�àˆ
rO   é,  TÚ
iterationsÚcleanupc                óª  ‡&‡'— ddl m} t          j        ¦   «         }| j                             dd¬¦  «         | j        dz                       dd¬¦  «         i }d}| j        r|                      ¦   «          d}| j                             ¦   «         rKt          |  
                    ¦   «         ¦  «        }t          j        | j        › d| j        › d|dz   › d	�¦  «         t          ||¦  «        D �]å}|                      ¦   «         }	t          j        | j        › d
|dz   › d|› d|	› �¦  «         i t!          | j        ¦  «        ¥|	¥}
|
                     d¦  «        }t'          |t(          t*          f¦  «        s|g} |di |
¥d|i¥| j        ¦  «        Š'‰'                     ¦   «         }d„ |                     ¦   «         D ¦   «         }t2          |
d                  Š&ˆ&fd„|                     ¦   «         D ¦   «         }ˆ'fd„|D ¦   «         }d„ |D ¦   «         }‰'j        }t7          d„ |                     ¦   «         D ¦   «         ¦  «        t          |¦  «        z  }||d<   |                      |dz   ||	|d„ t=          ||¦  «        D ¦   «         |¦  «        }|                      |d¬¦  «        r|dz  }d}| j        r[|                       ||	|||d         |¦  «         |                      ¦   «          | j!         "                    dddii¦  «        }||k    rd}n|  #                    |¦  «         |  
                    ¦   «         }|  $                    |¦  «        }|dd…df         }|  %                    ||¦  «        }||         }|dz   |z
  }| &                    di ¦  «        }| &                    d¦  «        |d         k    }|rÌ|rC|                     ¦   «         D ].}||                     ¦   «         vrtO          j(        |d¬¦  «         Œ/t=          ||¦  «        D ]s\  }}t          |¦  «        dk    r
| j        dz  n| j        dz  |z  } |                      dd¬¦  «         | )                    d¦  «        D ]}!tO          j*        |!| ¦  «         ŒŒt|}n|r|D ]}tO          j(        |d¬¦  «         Œ|}tW          tY          | j        ¦  «        ¦  «         ||k    rd}"n|dk    rd}"n	d|› d|› d �}"|                      |¦  «        }#| j        › |dz   › d|› d!|"› d"t          j        ¦   «         |z
  d#›d$�	| j        › d%t[          d&| j        ¦  «        › �g}$|#r†|$ .                    | j        › d'||         › d(|dz   › �| j        › d)|  /                    |¦  «        › �| j        › d*t          | &                    d+i ¦  «        ¦  «        dk    r
| j        dz  nd,› �g¦  «         d- 0                    |$¦  «        }%t          j        d-|%z   ¦  «         |#st          j1        | j        › d.�¦  «         nŸd/„ t=          | j2         3                    ¦   «         ||dd…f         ¦  «        D ¦   «         }ti          j5        | j        d0z  |tm          |% 7                    | j        d1¦  «        ¦  «        d-z   ¬2¦  «         ti          j8        | j        d0z  ¦  «         |r%t          j        | j        › d3|› d4|› d5�¦  «          dS �ŒçdS )6a÷  Execute the hyperparameter evolution process when the Tuner instance is called.

        This method iterates through the specified number of iterations, performing the following steps:
        1. Sync MongoDB results to local NDJSON (if using distributed mode)
        2. Mutate hyperparameters using the best previous results or defaults
        3. Train a YOLO model with the mutated hyperparameters
        4. Log fitness scores and hyperparameters to MongoDB and/or NDJSON
        5. Track the best performing configuration across all iterations

        Args:
            iterations (int): The number of generations to run the evolution for.
            cleanup (bool): Whether to delete iteration weights to reduce storage space during tuning.
        r   )ÚMultiTrainerT)ÚparentsrC   r6  zResuming tuning run z from iteration re   z...zStarting iteration ú/z with hyperparameters: ÚdataNc                ó   — g | ]	\  }}|°|‘Œ
S r¦   r¦   )r’   rä   rŒ   s      rM   rÅ   z"Tuner.__call__.<locals>.<listcomp>ý  s#   € ÐbÐbÐbÑ+;¨7°GÐZaÐb˜wÐbÐbÐbrO   Útaskc                ó(   •— i | ]\  }}||p‰d dd i“ŒS )r   r{   r¦   )r’   rä   rŒ   Úmetrics      €rM   r•   z"Tuner.__call__.<locals>.<dictcomp>ÿ  s:   ø€ ð ð ð ÙFVÀgÈw�˜ÐA V¨S°)¸SÐ$Aðð ð rO   c                ó$   •— g | ]}‰j         |z  ‘ŒS r¦   )Úsave_dir)r’   rä   Útrainers     €rM   rÅ   z"Tuner.__call__.<locals>.<listcomp>  s!   ø€ ÐRÐRÐR°w˜Ô(¨7Ñ2ÐRÐRÐRrO   c                ó   — g | ]}|d z  ‘ŒS )r6  r¦   )r’   Úss     rM   rÅ   z"Tuner.__call__.<locals>.<listcomp>  s   € Ð;Ð;Ð;¨Q˜1˜y™=Ð;Ð;Ð;rO   c              3  óH   K  — | ]}|pi                       d ¦  «        pdV — ŒdS )r{   r   N)r´   )r’   rŒ   s     rM   r¨   z!Tuner.__call__.<locals>.<genexpr>  s9   è è € ÐhÐhÀG˜7˜= b×-Ò-¨iÑ8Ô8Ð?¸CÐhÐhÐhÐhÐhÐhrO   r{   c                ó4   — i | ]\  }}|t          |¦  «        “ŒS r¦   )rS   )r’   rä   rW  s      rM   r•   z"Tuner.__call__.<locals>.<dictcomp>  s$   € ÐRÐRÐR¡Z W¨a��#˜a™&œ&ÐRÐRÐRrO   )rà   Fr�   r¤   )Úignore_errorsz*.ptu   complete âœ…u   complete (all failed) â�Œz
complete (u    succeeded) âš ï¸�z iterations z (z.2fzs)zResults saved to ÚboldzBest fitness=z observed at iteration zBest fitness metrics are zBest fitness model is r   z"not saved for multi-dataset tuningr±   zKNo iterations produced training metrics; skipping best_hyperparameters.yamlc                ód   — i | ]-\  }}||t           v rt          |¦  «        nt          |¦  «        “Œ.S r¦   )r   rU   r|   r‘   s      rM   r•   z"Tuner.__call__.<locals>.<dictcomp>Y  sE   € ð ð ð ÙEIÀQÈ�A ¥lÐ!2Ð!2•s˜1‘v”v�v½¸a¹¼ðð ð rO   zbest_hyperparameters.yamlz# )rN  ÚheaderzTarget iterations (z) reached in MongoDB (z). Stopping.)9Úultralytics.engine.trainerrK  rm   rD   ÚmkdirrI   r¾   rE   rÆ   rÛ   rÇ   r	   rK   rG   ri   rA  ÚvarsrA   r@   Ú
isinstancerµ   r	  r   Útrainrœ   r   Úmean_metricsr  rÞ   r‹   Úziprê   r¢   rt   Úcount_documentsrØ   rÔ   rò   r´   ÚshutilÚrmtreeÚglobÚcopy2r   rS   r   Úextendrß   ÚjoinÚerrorr   Úkeysr
   Úsaver   ÚreplaceÚprint)(rL   rH  rI  rK  Út0Úbest_save_dirsÚn_successfulÚstartrî   Úmutated_hypÚ
train_argsrN  Úraw_metricsrƒ   Údataset_metricsrT  Úweights_dirrŒ   r{   rŠ   Ústop_after_iterationÚtotal_mongo_iterationsrÈ   rx   Úbest_idxÚbest_resultÚn_attemptedÚcurrent_best_save_dirsÚbest_is_currentrW  rä   Ú
weight_dirÚbest_weights_dirÚckptÚstatusÚhas_valid_bestÚheader_linesr]  rR  rU  s(                                         @@rM   Ú__call__zTuner.__call__Ô  sU  øø€ ð 	<Ð;Ð;Ð;Ð;Ð;åŒY‰[Œ[ˆØŒ×Ò D°4ÐÑ8Ô8Ð8Ø	Œ˜Ñ	"×)Ò)°$ÀÐ)ÑFÔFÐFØˆØˆð Œ<ð 	)Ø×&Ò&Ñ(Ô(Ð(àˆØŒ>× Ò Ñ"Ô"ð 	kÝ˜×0Ò0Ñ2Ô2Ñ3Ô3ˆEÝŒK˜4œ;ÐiÐi¸D¼MÐiÐiÐ[`ÐcdÑ[dÐiÐiÐiÑjÔjÐjÝ�u˜jÑ)Ô)ð t	ñ t	ˆAàŸ,š,™.œ.ˆKÝŒK˜4œ;ÐsÐs¸1¸q¹5ÐsÐsÀ:ÐsÐsÐfqÐsÐsÑtÔtÐtà;�D ¤™OœOÐ;¨{Ð;ˆJØ—>’> &Ñ)Ô)ˆDÝ˜d¥T­5 MÑ2Ô2ð Ø�v�Ø"�l 4Ð)E¨JÐ)E¸ÀÐ)EÐ)EÀtÄzÑRÔRˆGØ!Ÿ-š-™/œ/ˆKØbÐb¸{×?PÒ?PÑ?RÔ?RÐbÑbÔbˆOÝ  ¨FÔ!3Ô4ˆFðð ð ð ØZe×ZkÒZkÑZmÔZmðñ ô ˆOð SÐRÐRÐRÀ/ÐRÑRÔRˆHØ;Ð;°(Ð;Ñ;Ô;ˆKØÔ*ˆGÝÐhÐhÈ×OeÒOeÑOgÔOgÐhÑhÔhÑhÔhÕknÐosÑktÔktÑtˆGØ!(ˆG�IÑØ×(Ò(Ø�A‘ØØØØRÐRµ3°ÈÑ3QÔ3QÐRÑRÔRØñô ˆFð ×)Ò)¨&¸dÐ)ÑCÔCð "Ø Ñ!�Ø#(Ð ØŒ|ð 	0Ø×%Ò%Ø˜[¨'°?ÀFÈ;ÔDWÐYhñô ð ð ×*Ò*Ñ,Ô,Ð,Ø)-¬×)HÒ)HÈ)ÐV_ÐaeÐUfÐIgÑ)hÔ)hÐ&Ø)¨ZÒ7Ð7Ø+/Ð(øà×'Ò'¨Ñ/Ô/Ð/ð ×.Ò.Ñ0Ô0ˆGØ×,Ò,¨WÑ5Ô5ˆAØ˜˜˜˜1˜”gˆGØ×.Ò.¨w¸Ñ@Ô@ˆHØ! (Ô+ˆKØ˜q™5 E™/ˆKØ%0§_¢_°[À"Ñ%EÔ%EÐ"Ø)Ÿošo¨kÑ:Ô:¸fÀ[Ô>QÒQˆOØð 8Øð AØ+×2Ò2Ñ4Ô4ð Að A˜ØÐ$:×$AÒ$AÑ$CÔ$CÐCÐCÝ"œM¨!¸4Ð@Ñ@Ô@Ð@øÝ+.¨ÀÑ+LÔ+Lð =ð =Ñ'�G˜Zå58¸±Y´YÀ!²^°^˜œ¨	Ñ1Ð1ÈÌÐYbÑIbÐelÑIlð %ð %×*Ò*°4À$Ð*ÑGÔGÐGØ *§¢°Ñ 7Ô 7ð =ð =˜Ýœ TÐ+;Ñ<Ô<Ð<Ð<ð=à!7��Øð 8Ø!ð 9ð 9�AÝ”M !°4Ð8Ñ8Ô8Ð8Ð8Ø!7�õ �c $¤.Ñ1Ô1Ñ2Ô2Ð2ð ˜{Ò*Ð*Ø'��Ø Ò"Ð"Ø4��àT lÐTÐT°[ÐTÐTÐT�Ø!×7Ò7¸ÑDÔDˆNà”;Ðb  A¡ÐbÐb¨
ÐbÐbÀÐbÐbÍ$Ì)É+Ì+ÐXZÑJZÐbÐbÐbÐbØ”;ÐRÐRµ¸&À$Ä-Ñ1PÔ1PÐRÐRðˆLð ð 
Ø×#Ò#àœ;ÐmÐm°W¸XÔ5FÐmÐmÐ_gÐjkÑ_kÐmÐmØœ;ÐbÐbÀ×ASÒASÐT_ÑA`ÔA`ÐbÐbà#œ{ð Rð RÝ<?ÀÇÂÐPZÐ\^Ñ@_Ô@_Ñ<`Ô<`ÐdeÒ<eÐ<e˜tœ}¨yÑ8Ð8ð  lPðRð Rð	ñ	ô 	ð 	ð —Y’Y˜|Ñ,Ô,ˆFÝŒK˜˜v™Ñ&Ô&Ð&Ø!ð HÝ”Ø”{ÐoÐoÐoñô ð ð ðð ÝMPÐQUÔQ[×Q`ÒQ`ÑQbÔQbÐdeÐfnÐpqÐprÐprÐfrÔdsÑMtÔMtðñ ô �õ ”	Ø”MÐ$?Ñ?ØÝ*¨6¯>ª>¸$¼+ÀtÑ+LÔ+LÑMÔMÐPTÑTðñ ô ð õ
 ”
˜4œ=Ð+FÑFÑGÔGÐGØ#ð Ý”Ø”{Ð}Ð}°zÐ}Ð}ÐYoÐ}Ð}Ð}ñô ð ð ��ñ	ðat	ð t	rO   )N)r   r   )rP   rQ   )rR   rS   rT   rU   )rP   rP   rP   )NN)rz   rU   r{   r|   r}   r~   r   r€   r�   r‚   rƒ   r„   r…   r†   )r{   r|   r}   r~   rŒ   r†   r   r€   r�   r�   rƒ   rŽ   )r…   r¿   )rÈ   r¿   r…   rÉ   )rŠ   r†   )rŠ   r†   r…   r   )F)rŠ   r†   rà   rá   r…   rá   )rÈ   r¿   r{   rë   r…   rU   )ró   rô   )rõ   rU   rö   r|   r…   r~   )r@  r~   r…   r~   )rG  T)rH  rU   rI  rá   )Ú__name__Ú
__module__Ú__qualname__Ú__doc__rN   rr   rJ   Ústaticmethodry   r‹   r¢   r¾   rÇ   rÔ   rØ   rß   rê   Úclassmethodrò   rA  r  r‡  r¦   rO   rM   r   r       sì  € € € € € ð.ð .ð`?
Àdð ?
ð ?
ð ?
ð ?
ð ?
ð ?
ðB&&ð &&ð &&ð &&ð &&ðPPð Pð Pð Pð( ð:ð :ñ „\ð:ð ,0Ø,0ðð ð ð ð ð,#Eð #Eð #Eð #EðJOð Oð OðBDð Dð Dð Dð
ð 
ð 
ð 
ðKð Kð Kð Kð
 ðð ð ñ „\ðð ðQð Qð Qð Qñ „\ðQð ðWð Wð Wñ „[ðWð Øðe$ð e$ð e$ð e$ð e$ðNð ð ð ðRð Rð Rð Rð Rð Rð RrO   r   )r‹  Ú
__future__r   r²   rf  rm   r   ÚnumpyrÒ   Úultralytics.cfgr   r   r   r   Úultralytics.utilsr	   r
   r   r   r   Úultralytics.utils.checksr   Úultralytics.utils.plottingr   r   r¦   rO   rM   ú<module>r”     s
  ððð ð #Ð "Ð "Ð "Ð "Ð "à €€€Ø €€€Ø €€€Ø Ð Ð Ð Ð Ð à Ð Ð Ð à LÐ LÐ LÐ LÐ LÐ LÐ LÐ LÐ LÐ LÐ LÐ LØ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PÐ PØ 7Ð 7Ð 7Ð 7Ð 7Ð 7Ø 8Ð 8Ð 8Ð 8Ð 8Ð 8ðF	ð F	ð F	ð F	ð F	ñ F	ô F	ð F	ð F	ð F	rO   