+ jaR t0t^RIt^RIt^RIt^RIHtHt^RIt^RI H t ^RI H t H t ]'d^RIHtRR.t]P"!]4t]^kRs]^kRs]^kR RR lltR R ltR#)N)Literal TYPE_CHECKING) _WorkerServer)get_ranktcpstore_client) DebugHandlerstart_debug_serverstop_debug_servercV^8dQhR\R\R\R ,R,R\R,R\R\\,R,R\ R ,R,R \R R/ #) port worker_port start_methodNdump_dir dump_interval enabled_dumpshandlersr fetch_timeoutreturn)forkspawn forkserver)intrstrfloatsetlist)formats"a/home/luca/screen-checking/.venv/lib/python3.14/site-packages/torch/distributed/debug/__init__.py __annotate__r sd#d# d#d#784?d#Dj d#  d# s8d? d#>"T)d#d# d#c V\e \R4h\e \R4h\P RV4\ 4p\ RV4s\4p VPRV 2R\P!4 R\P 24V ^8Xd^RI H p ^R IHp VfV !4pVf0RmpR VR VR VR VRVRV/p Ve+\ P"!V4p V P%WRR7sM\ P$!WRR7s\P'4R#R#)a Start the debug server stack on all workers. The frontend debug server is only started on rank0 while the per rank worker servers are started on all ranks. This server provides an HTTP frontend that allows for debugging slow and deadlocked distributed jobs across all ranks simultaneously. This collects data such as stack traces, FlightRecorder events, and performance profiles. This depends on dependencies which are not installed by default. Dependencies: - Jinja2 - aiohttp WARNING: This is intended to only be used in trusted network environments. The debug server is not designed to be secure and should not be exposed to the public internet. See SECURITY.md for more details. WARNING: This is an experimental feature and may change at any time. Args: port (int): The port to start the frontend debug server on. worker_port (int): The port to start the worker server on. Defaults to 0, which will cause the worker server to bind to an ephemeral port. start_method (str | None): The multiprocessing start method to use for the frontend server process. One of "fork", "spawn", or "forkserver". If None, uses the default start method. Using "spawn" is recommended when using CUDA or when fork safety is a concern. dump_dir (str | None): Directory to write periodic debug dumps to. If None, periodic dumping is disabled. dump_interval (float): Seconds between periodic dumps. Defaults to 60. enabled_dumps (set[str] | None): Set of handler dump filenames to enable (e.g. {"stacks", "fr_trace", "tcpstore"}). If None, all handlers that implement dump() are enabled. handlers (list[DebugHandler] | None): List of debug handlers to use. If None, uses the default handlers. See torch.distributed.debug._handlers for the default handlers. fetch_timeout (float): Timeout in seconds for fetching data from individual workers. Defaults to 60. Workers that don't respond within this time will be reported as unavailable. Nzdebug server already startedz Starting debug server on port %dz::rankzhttp://:)default_handlers)mainr rrrrrT)targetkwargsdaemon>stacksfr_tracetorchcomms_health_check)_WORKER_SERVERAssertionError_DEBUG_SERVER_PROCloggerinforrrrsocket gethostnamer 'torch.distributed.debug._debug_handlersr%!torch.distributed.debug._frontendr&multiprocessing get_contextProcessstart)r rrrrrrrstoreRANKr%r& main_kwargsctxs&&&&&&&& rr r s1l!;<<%;<< KK2D9  E"45N :D IITFmwv'9'9';&Q>Q=RST qyL:  ')H  M D  ] ]  ]    #!--l;C!$"-" "1!8!8"    "Ar!cV^8dQhRR/#)r rN)rs"rr r s4r!c\f\h\f\h\P R4\P 4\P 4\P4RsRsR#)zG Shutdown the debug server and stop the frontend debug server process. NzStopping debug server)r/r.r-r0r1 terminateshutdownjoinr?r!rr r sX ! KK'(  "Nr!cV^8dQh/^\9d\P;R&^\9d\R,;R&^\9d\P R,;R&#)r r0Nr-r/)__conditional_annotations__loggingLoggerrr6r8)rs"rr r sP(544),,+ $+-.:9O++d29/r!)ierNNN@NNrH)rErFr6r2typingrr!torch.distributed.debug._handlerstorchtorch._C._distributed_c10drtorch.distributed.debug._storerrr5r__all__ getLogger__name__r0r-r/r r r )rEs@rrQsh ))4D>  !**844'++599d#Nr!