^is'JUdZddlZddlZddlZddlZddlmZddlmZm Z ddl m Z m Z m Z mZmZddlZddlmZddlmZmZmZmZmZmZej6d Ze eeeeffZee d <e eeefZ!ee d <e ee!fZ"ee d <eGd dZ#de!ddfdZ$d&dZ%defdZ&defdZ'de(fdZ)de ddfdZ*defdZ+dededefdZ,dejZdefdZ.e GddeZ/dej`d e/dej`fd!Z1d&d"Z2de efd#Z3Gd$d%Z4y)'z-XGBoost collective communication related API.N) dataclass)IntEnumunique)AnyDictOptional TypeAliasUnion)_T)_LIB _check_call build_infoc_str make_jcargspy_strz[xgboost.collective]_Conf_ArgVals_ArgsceZdZUdZdZeeed<dZeeed<dZ ee ed<dZ eeed<dZ eeed<de d e fd Zy) ConfigaUser configuration for the communicator context. This is used for easier integration with distributed frameworks. Users of the collective module can pass the parameters directly into tracker and the communicator. .. versionadded:: 3.0 Attributes ---------- retry : See `dmlc_retry` in :py:meth:`init`. timeout : See `dmlc_timeout` in :py:meth:`init`. This is only used for communicators, not the tracker. They are different parameters since the timeout for tracker limits only the time for starting and finalizing the communication group, whereas the timeout for communicators limits the time used for collective operations, like :py:meth:`allreduce`. tracker_host_ip : See :py:class:`~xgboost.tracker.RabitTracker`. tracker_port : See :py:class:`~xgboost.tracker.RabitTracker`. tracker_timeout : See :py:class:`~xgboost.tracker.RabitTracker`. Nretrytimeouttracker_host_ip tracker_porttracker_timeoutargsreturncr|j|j|d<|j|j|d<|S)z*Update the arguments for the communicator. dmlc_retry dmlc_timeout)rrselfrs S/var/www/html/hubwallet-dev/venv/lib/python3.12/site-packages/xgboost/collective.pyget_comm_configzConfig.get_comm_config:s7 :: !!%D  << ##'<ttjy)zFinalize the communicator.N)rr XGCommunicatorFinalizer.r&r$finalizer4is++-.r&c.tj}|S)zjGet rank of current process. Returns ------- rank : int Rank of current process. )r XGCommunicatorGetRankrets r$get_rankr9ns  $ $ &C Jr&c.tj}|S)z`Get total number workers. Returns ------- n : Total number of process. )r XGCommunicatorGetWorldSizer7s r$get_world_sizer<zs  ) ) +C Jr&c@tj}t|S)z.If the collective communicator is distributed.)r XGCommunicatorIsDistributedbool)is_dists r$is_distributedrAs..0G =r&msgct|ts t|}tj}|dk7r6t tj t |jyt|jdy)zPrint message to the communicator. This function can be used to communicate the information of the progress to the communicator. Parameters ---------- msg : str The message to be printed to the communicator. rT)flushN) isinstancer-r r>rXGCommunicatorPrintrstripprint)rBr@s r$communicator_printrIsZ c3 #h..0G!|D,,U399;-?@A ciik&r&ctj}ttjtj ||j }t|S)zdGet the processor name. Returns ------- name : The name of processor(host) )ctypesc_char_prr XGCommunicatorGetProcessorNamebyrefvaluer)name_strrOs r$get_processor_namerQs? H33FLL4JKL NNE %=r&datarootc Xt}tj}||k(r>|Jdtj|tj }t ||_ttjtj|tjtj|||k7rtj|jz}ttjtj|tj|j|tj |j"}~|Sttjtjtj$tj|j|~|S)aSBroadcast object from one node to all other nodes. Parameters ---------- data : any type that can be pickled Input data, if current rank does not equal root, this can be None root : int Rank of the node to broadcast data from. Returns ------- object : int the result of broadcast. z&need to pass in data when broadcasting)protocol)r9rKc_ulongpickledumpsHIGHEST_PROTOCOLlenrOrr XGCommunicatorBroadcastrNsizeofc_charcastc_void_ploadsrawrL)rRrSranklengthsdptrs r$ broadcastrfs6 :D ^^ F t|I!II LL(?(? @1v  $$ LL &--"?   t|  ,/  ( ( D&//2FLL$  ||DHH%  K   ( ( FOOA.@&,,PT   Kr&dtypecptjddtjddtjddtjddtjd d tjd d tjd dtjddtjddtjddtjddi } |jtjddi||vrtd|d||S#t$rY#wxYw)Nfloat16rfloat32r float64int8int16int32int64uint8uint16 uint32 uint64 float128z data type z* is not supported on the current platform.)nprgupdate TypeError)rg dtype_maps r$ _map_dtypers Q Q Q ! 1 1 1 1 A B B I "((:.23 I*UG+UVWW U     s+&D)) D54D5c(eZdZdZdZdZdZdZdZdZ y) Opz#Supported operations for allreduce.rr rlr~rnrpN) r'r(r)r*MAXMINSUM BITWISE_AND BITWISE_OR BITWISE_XORr.r&r$rrs#- C C CKJKr&ropc ht|tjs td|j j }t tj|jjtj|jt|jt||S)a'Perform allreduce, return the result. Parameters ---------- data : Input data. op : Reduction operator. Returns ------- result : The result of allreduce, have same shape as data Notes ----- This function is not thread-safe. z%allreduce only takes in numpy.ndarray)rErndarrayrravelcopyrr XGCommunicatorAllreducerKdata_asr_sizerrgr+)rRrbufs r$ allreducers& dBJJ '?@@ **,   C $$ JJ  v / HH syy ! G   Jr&c>ttjy)zKill the process.N)rr XGCommunicatorSignalErrorr.r&r$ signal_errorr$s..01r&cddlm}|j*tjj |j}n6t |dr(t|jdkDr|jd}nd}|sytj|}|syd}|D]}|jds|}n|"tjj||}|Sy)Nr)lib__path__z libnccl.so) nvidia.ncclr__file__ospathdirnamehasattrrZrlistdir startswithjoin)rrfileslibnamenamers r$ _find_ncclr)s ||!#!> j !c#,,&7!&;,,q/  JJw E !G ??< (G  ww||GW- r&c<eZdZdZdeddfdZdefdZdeddfdZ y)CommunicatorContextzNA context controlling collective communicator initialization and finalization.rrNc ||_d}|j|dyt}|dsy t}|r||j|<yy#t$rYywxYw)Ndmlc_nccl_pathUSE_DLOPEN_NCCL)rgetrr ImportError)r#rkeybinfors r$__init__zCommunicatorContext.__init__Nsh  88C  *  &'  rs3  ! 88KK   1 2U38_,-y-uS#X/)/X &y& '' 'T#>#>d#>L/ #    'C'D'( C ,B,c,b,`bhh32BJJB2::B2 HSMDPPr&