.hTdZddlmZddlZddlZddlZddlZddlZddlmZddl Z ddl Z ddl m Z mZddlmZmZmZmZmZmZddlmZddlmZdd lmZGd d Zy) a5 Module provides functionalities for hyperparameter tuning of the Ultralytics YOLO models for object detection, instance segmentation, image classification, pose estimation, and multi-object tracking. 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 YOLO11n on COCO8 at imgsz=640 and epochs=30 for 300 tuning iterations. >>> from ultralytics import YOLO >>> model = YOLO("yolo11n.pt") >>> model.tune(data="coco8.yaml", epochs=10, iterations=300, optimizer="AdamW", plots=False, save=False, val=False) ) annotationsN)datetime)get_cfg get_save_dir) DEFAULT_CFGLOGGERYAML callbackscolorstrremove_colorstr)check_requirements) torch_load)plot_tune_resultsczeZdZdZedfd dZd ddZddZdddZddZ dZ ddd Z d dd Z ddd Z y)TuneraF 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 CSV 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_csv (Path): Path to the CSV file where evolution logs are saved. args (dict): Configuration arguments for the tuning process. callbacks (list): 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 YOLO11n on COCO8 at imgsz=640 and epochs=30 for 300 tuning iterations. >>> from ultralytics import YOLO >>> model = YOLO("yolo11n.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-1), "momentum": (0.6, 0.98)}) Nc |jddxs@iddddddd d d d d dddddddddddddddddddddd d d!d!d!d!d!d!d!dd"|_|jd#d}|jd$d%}|jd&d'}t|(|_|jj|j_t |j|jjxsd)*|_d+\|j_|j_|j_|jd,z |_ |xstj|_ td-|_ tj|d|_|r|j!|||t#j$|jd.|jd/|jd0y)1z Initialize the Tuner with configurations. Args: args (dict): Configuration for hyperparameter evolution. _callbacks (list | None, optional): Callback functions to be executed during tuning. spaceNlr0)gh㈵>皙?lrf)g-C6?rmomentum)gffffff?g\(\?g333333? weight_decay)gMbP? warmup_epochs)rg@warmup_momentum)rgffffff?box)?g4@cls)r@dfl)g?g@hsv_h)rrhsv_s)rg?hsv_vdegrees)rgF@ translatescaleshear)rg$@ perspective)rr)flipudfliplrbgrmosaicmixupcutmix copy_paste close_mosaic mongodb_uri mongodb_db ultralyticsmongodb_collection tuner_results) overridestune)name)NFFztune_results.csvzTuner: z*Initialized Tuner instance with 'tune_dir=z' uT💡 Learn about tuning at https://docs.ultralytics.com/guides/hyperparameter-tuning)poprrargsresumeexist_okrr8tune_dirtune_csvr get_default_callbacksr prefixadd_integration_callbacksmongodb _init_mongodbrinfo)selfr: _callbacksr1r2r4s V/var/www/html/ai-service/venv/lib/python3.12/site-packages/ultralytics/engine/tuner.py__init__zTuner.__init__Us3XXgt, 1 <1  =1  ( 1 L 1 Z 1  {1  ;1  :1  :1  Z1  Z1  Z1  {1  1 [!1 " [#1 $ <%1 &!   $'51  8hh}d3 XXlM: !XX&:OLd+ !YY-- $TYYTYY^^5MvN ?S<  *DII,< (:: #Hy'F'F'Hy) ++D1    {J8J K {{mEdmm_TW{{mo q c tdddlm}ddlm}m}t |D]Z} ||dddddd d d }|jjd tj|jd|dzd|cSy#||f$rQ||dz k(rd|z}tj|jd|dzd|dtj|YwxYw)ao 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. pymongor) MongoClient)ConnectionFailureServerSelectionTimeoutErrori0ui Ni@Ti`)serverSelectionTimeoutMSconnectTimeoutMSsocketTimeoutMS retryWrites retryReads maxPoolSize minPoolSize maxIdleTimeMSpingz$Connected to MongoDB Atlas (attempt )z#MongoDB connection failed (attempt z), retrying in zs...N)r rKrLpymongo.errorsrMrNrangeadmincommandrrDr@warningtimesleep) rEuri max_retriesrLrMrNattemptclient wait_times rG_connectzTuner._connects 9%'Q[) &G &$-2%*$) $# " !"'  $$V, t{{m+OPWZ[P[}\]^_  & &'BC &kAo-wJ {{m#FwQR{mSbclbmmqr 9% &sABACCc|j||_|j|||_|jjdgdt j |j dy)a 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, e.g. 'mongodb+srv://username:password@cluster.mongodb.net/'. mongodb_db (str, optional): Database name. mongodb_collection (str, optional): Collection name. Notes: - Creates a fitness index for fast queries of top results - Falls back to CSV-only mode if connection fails - Uses connection pooling and retry logic for production reliability )fitnessT) backgroundz*Using MongoDB Atlas for distributed tuningN)rirB collection create_indexrrDr@)rEr1r2r4s rGrCzTuner._init_mongodbs\$}}[1 ,,z23EF $$o%64$H t{{m#MNOrIc t|jjjddj |S#t $rgcYSwxYw)z Get top N results from MongoDB sorted by fitness. Args: n (int): Number of top results to retrieve. Returns: (list[dict]): List of result documents with fitness scores and hyperparameters. rkrl)listrnfindsortlimit Exception)rEns rG_get_mongodb_resultszTuner._get_mongodb_resultssK ,,.33IrBHHKL L I sAA AAc z |jjt||jDcic]$\}}|t |dr|j n|&c}}|t j|dycc}}w#t$r.}tj|jd|Yd}~yd}~wwxYw)a Save results to MongoDB with proper type conversion. Args: fitness (float): Fitness score achieved with these hyperparameters. hyperparameters (dict[str, float]): Dictionary of hyperparameter values. metrics (dict): Complete training metrics dictionary (mAP, precision, recall, losses, etc.). iteration (int): Current iteration number. item)rkhyperparametersmetrics timestamp iterationzMongoDB save failed: N) rn insert_onefloatitemshasattrryrnowrurrar@)rErkrzr{r}kves rG_save_to_mongodbzTuner._save_to_mongodbs E OO & &$W~_n_t_t_v'wW[WXZ[68JAFFHPQ(Q'w&!)!*  (x  E NNdkk]*?sC D D Es(2B)A= B=B B: $B55B:c  t|jjjdd}|sydj dgt|j j zdz}t|jdd 5}|j||D]u}|d}|j j Dcgc] }|d | }}t|d g|z}|jdj tt|dzw dddycc}w#1swYyxYw#t$r.} tj|j d | Yd} ~ yd} ~ wwxYw) a Sync MongoDB results to CSV for plotting compatibility. Downloads all results from MongoDB and writes them to the local CSV file in chronological order. This enables the existing plotting functions to work seamlessly with distributed MongoDB data. r}rZN,rk wutf-8encodingrzzMongoDB to CSV sync failed: )rqrnrrrsjoinrkeysopenr>writeroundmapstrrurrar@) rE all_resultsheadersfresultrkr hyp_valueslog_rowrs rG_sync_mongodb_to_csvzTuner._sync_mongodb_to_csvsL Lt335::;JKKhh {T$**//2C-DDELGdmmS7; @q )@F$Y/GHL HY!Z1&):";A">!ZJ!Z$Wa01J>GGGCHHSg%67$>? @ @ @"[  @ @ L NNdkk]*FqcJ K K LsN5D0AD0 8D$DAD$D0D$$D-)D0-D00 E'9$E""E'cbt|t|}|dddf|dddfjz dz}tj|j r|j dk(rtj |}tjtt|||}tj|Dcgc] }||dd c}d}|jd|jd} }| |z } tjj||| zz | || zzScc}w)uGBLX-α crossover from up to top-k parents (x[:,0]=fitness, rest=genes).Nrgư>)weightsrrZ) minlennpisfiniteallsum ones_likerandomchoicesr^stackmaxuniform) rExalpharridxsi parents_matlohispans rG _crossoverzTuner._crossovers 3q6NAqD'AadGKKM)D0{{7#'')W[[]a-?ll7+G~~eCFmWBhh$7Q!QR7; #[__Q%7BBwyy  edl!2B4EFF 8sD,c d}|jr|j|}|rYtj|Dcgc]7}|dg|jj Dcgc] }|d| c}z9c}}}n|j j|j jjvrTtjdg|jj Dcgc]}t|j|c}zg}|r|jjrXtj|jddd}t|d kDr'|ddd f} tj | } || d|}|tj"j%t't)j(t|j} |j+|} tj|jj-D cgc]} t| d k(r| dnd c} }tj.| }tj0|dk(rtj"j#| |k}tj"j3| ||zz}tj4|tj6|d j9d d }tj0|dk(rt;|jj Dcic]\}}|t=| |||z}}}n<|jj Dcic]}|t|j|}}|jj?D]1\}}tAtCtE|||d |dd||<3d|vrt'tA|d|d<|Scc}wcc}}wcc}wcc} wcc}}wcc}w)a# Mutate hyperparameters based on bounds and scaling factors specified in `self.space`. Args: parent (str): Parent selection method (kept for API compatibility, unused in BLX mode). n (int): Number of top parents to consider. mutation (float): Probability of a parameter mutation in any given iteration. sigma (float): Standard deviation for Gaussian random number generator. Returns: (dict[str, float]): A dictionary containing mutated hyperparameters. Nrkrzrr\rrZndmin delimiterskiprowsrrPrg?rrr0)#rBrwrarrayrrrnr8databaselist_collection_namesgetattrr:r>existsloadtxtrargsortrseedintrbrvaluesonesrrandnwhereexpclip enumeraterrrrr)rErvmutationsigmarresultsrrcsv_datarkordernggenesrgainsfactorsmasksteprhypboundss rG_mutatez Tuner._mutates=$  <<//2GHHmtuhiq|nQUQ[Q[Q`Q`Qb/cA2C0DQ0G/ccuv%%)A)A)W)W)YYHHsedjjooFW&Xwtyy!'<&XXYZ 9--/zz$--qCRSTH8}q "1a4. G8,UOBQ' = IINN3tyy{+ ,TZZBOOA&EHHDJJDUDUDWXqc!fkads:XYEggbkG&&A&yy''+h6yyr*eem<((4s;@@sK&&A&@IIZ?[\tq!1eE!Hwqz122\C\59ZZ__5FG1gdii++GCG))+ FIAv3s3q66!95vayA1ECF F S "%eC,?&@"AC  O0du&X&Y ]Gs/&O- O(+O- "O3 O8O=P(O- c tj}d\}}|jdz jdd|jr|j d}|j j rgtj|j ddd }|jd}tj|jd |jd |dzd t||D]} t| d z d} dd| zz } |j| } tj|jd| dzd|d| i} it!|j"| }t%t'|}|dz } t)dj*ddg}g|dd|j-D}t/j0|dj2}||dz j rdndz }t5|d} |dk(sJdtj6dt9j:t<j>jA| jGd"d#}|jrr|jI|| | | dz|j |jJjMi}||k\r tj|jd$|d%|d&y!tO|d'g|jPjSDcgc]}| | c}z}|j j rd(n8djUd"gtW|jPjSzd z}tY|j d)d*+5}|j[|djUt]t^|zd zd!d!d!tj|j ddd }|d!d!df}|ja}|| k(}|rrt_|}| j-Dcic]\}}|tO|d'}}}|jcd,D]%}tejf||jdz 'n|r|rtejh|d-tkt_|j |j| dzd|d.tj|z d/d0|jd1tmd2|jd |jd3||d4|dzd |jd5|d |jd6|} tjd | zto|jPjSD cic]\} }|tq||| dzf}!} }tsjt|jd7z |!tw| jy|jd8d z9tsjz|jd7z y!#tB$r)}tjDd| dzd |Yd!}~d!}~wwxYwcc}w#1swYxYwcc}}wcc}} w):aV Execute the hyperparameter evolution process when the Tuner instance is called. This method iterates through the specified number of iterations, performing the following steps: 1. Sync MongoDB results to CSV (if using distributed mode) 2. Mutate hyperparameters using the best previous results or defaults 3. Train a YOLO model with the mutated hyperparameters 4. Log fitness scores and hyperparameters to MongoDB and/or CSV 5. Track the best performing configuration across all iterations Args: model (Model | None, optional): A pre-initialized YOLO model to be used for training. iterations (int): The number of generations to run the evolution for. cleanup (bool): Whether to delete iteration weights to reduce storage space during tuning. )NNrT)parentsr<rr\rrZrzResuming tuning run z from iteration z...gr@r皙?r)rzStarting iteration /z with hyperparameters: sysz-mzultralytics.cfg.__init__trainc30K|]\}}|d|yw)=N).0rrs rG z!Tuner.__call__..s*U$!QaS!:*Us)checkzbest.ptzlast.pt train_metricsztraining failedz5training failure for hyperparameter tuning iteration rNrkrzTarget iterations (z) reached in MongoDB (z ). Stopping.rarrz*.pt) ignore_errorsu iterations complete ✅ (z.2fzs) zResults saved to boldz Best fitness=z observed at iteration zBest fitness metrics are zBest fitness model is zbest_hyperparameters.yamlz# )dataheader)>rbr=mkdirrBrr>rrrshaperrDr@r^rrvarsr:rr __import__ executabler subprocessrun returncoderrcgccollecttorchcuda empty_cacheruerrorgetrrncount_documentsrrrrrqrrrrargmaxglobshutilcopy2rmtreerr rrr saver replaceprint)"rEmodel iterationscleanupt0 best_save_dir best_metricsstartrrfracsigma_i mutated_hypr{ train_argssave_dir weights_dirlaunchcmd return_code ckpt_filerrktotal_mongo_iterationsrrrrbest_idxbest_is_currentrckptrrs" rG__call__zTuner.__call___s YY[&0# |  "))$)F <<  % % ' ==   ! 4==S1MAGGAJE KK4;;-';DMM?JZ[`cd[dZeehi juj)O DAq5y#&DC$J&G,,W,5K KK4;;-':1q5':,Nefqers tG;DO;{;J#GJ$78H"Y.K c$U+66>XYVVV*U*BRBRBT*UV(nnS=HH 'y9P8X8X8Z9`ij $Y/@"a':)::' 1   &&( kk)S1G||%%g{GQUK))+)-)H)H)L&)Z7KK;;-'::,F\]s\tuAB!!,-IZ0[AQ0[[ $ 4 4 6"SXXykTXY]YcYcYhYhYjTkFk=los=s$--w?J1GGGchhs3/@&AADHIJ 4==S1MA1gG~~'H&!mO #H ;B==?K41a5A;K K',,V4BDLLt}}y'@AB] m4@ c$--0 1;;-Awa |3Mdiik\^N^_bMccg;;-0&$--1P0QQS;;-}WX->,??VW_bcWcVddf;;-8b;;-5m_ F  KKv &=FtzzGX=YZTQAuQxQ/00ZDZ II ;;&v~~dkk4'HIDP  JJt}}'BB C_O D6 c TUVYZUZT[[]^_]`abb c 1\JJ L$[s74CX= X7;5X<Y 0Y X4 X//X4<Y )rFz list | None)z4mongodb+srv://username:password@cluster.mongodb.net/rP)rdrrer)rrr)r)rvrreturnrq)rkrrzdict[str, float]r{dictr}r)r )r np.ndarrayrrrrrr)rg?r)rvrrrrrrr)N T)rrrbool)__name__ __module__ __qualname____doc__rrHrirCrwrrrrrrrIrGrr$s{.`(49 v'&RP. E.L4 G @ @@ @  @DlDrIr)r# __future__rrrrrrbrnumpyrrultralytics.cfgrrultralytics.utilsrrr r r r ultralytics.utils.checksr ultralytics.utils.patchesrultralytics.utils.plottingrrrrIrGr+sG #   1]]708gDgDrI