|j%fddlmZddlZddlZddlZgZdZdZdZ dZ dZ dZ d Z d Zdd ZdS)) annotationsNc tjdgdd}tjddgdtjtjjd }tjdgd }tjj ||| }tj |}|||fS) zD Define a simple network composed by a single linear layer. input)Nrfloat32)nameshapedtyperg?) initializer)r r attr)r r )xweightbias) paddlestaticdatacreate_parameter ParamAttrnnr Constant functionallineartensorsum)rrr linear_outouts j/lsinfo/ai/hellotax_ai/data_center/backend/venv/lib/python3.11/site-packages/paddle/utils/install_check.py_simple_networkrs M   LLL    E $!f  &)*?*H*H*M*M N N NF  "!I > > >D%,,uV$,OOJ -  J ' 'C #v c|tjddgddggtj}|dddS)zK Prepare feeding data for simple network. The shape is [1, 2, 2]. ?g@g@g@)r r)nparrayrreshape)np_input_singles r _prepare_datar(,s? hc S#J7rzJJJO  " "1a + ++r c ttjdksJdS#t$r"}t jd|Yd}~dSd}~wwxYw)z* Check whether CUDA is available. rTzYou are using GPU version PaddlePaddle, but there is no GPU detected on your machine. Maybe CUDA devices is not set properly. Original Error is NF)lenrr cuda_places Exceptionloggingwarninges r_is_cuda_availabler16s 6=,,..//!3333t  ($% ( (   uuuuu 15 A!AA!c ttjdksJdS#t$r"}t jd|Yd}~dSd}~wwxYw)z) Check whether XPU is available. rTzYou are using XPU version PaddlePaddle, but there is no XPU detected on your machine. Maybe XPU devices is not set properly. Original Error is NF)r*rr xpu_placesr,r-r.r/s r_is_xpu_availabler5Fs 6=++--..2222t  ($% ( (   uuuuu r2cVtj|rtjdnB|rtjdn+|rtj|ntjdtjdtjjd}tjdtjjd }tjd d || }t}tj |}||} tj | } | tj d |} | dS)z Testing the simple network in dygraph mode using one CPU/GPU/XPU. Args: use_cuda (bool): Whether running with CUDA. use_xpu (bool): Whether running with XPU. gpuxpucpurg?)value)rr rr"r) weight_attr bias_attrMbP? learning_rate parametersN)rdisable_static set_devicerrr rLinearr( to_tensorrrbackward optimizerAdamrAstep) use_cudause_xpu use_customcustom_device_namer<r=rinput_np input_tensorrropts r_run_dygraph_singlerQVs !%    !%    !,----%   " 69#8#A#A#A#L#LK !6!?!?c!?!J!JIY   1+FH#H--L %%J -  J ' 'CLLNNN    (9(9(;(;   CHHJJJJJr ctjtjtj5tj}tj}d|_tj||5t\}}}tj ||gd} dddn #1swxYwY|rtj d} nB|rtj d} n+|rtj |d} ntj } tj| } | || ||jt#i|| dgdddn #1swxYwYtjdS)z Testing the simple network with executor running directly, using one CPU/GPU/XPU. Args: use_cuda (bool): Whether running with CUDA. use_xpu (bool): Whether running with XPU. r#)parameter_listrN)feed fetch_list)r enable_staticr scope_guardScopeProgram random_seed program_guardrappend_backward CUDAPlaceXPUPlace CustomPlaceCPUPlaceExecutorrunrr(rB) rJrKrLrM train_prog startup_progrrr param_gradsplaceexes r_run_static_singlerh{s9   " "6=#6#6#8#8 9 9  ]**,, },,.. #$  ] ( (\ B B  !0!2!2 E3 -77VH8K                 &$Q''EE  &OA&&EE  &&'91==EEO%%Em$$U++    *moo.[^,    +               4 s8A$G3;C:. G:C> >GC> CGGGc\Gddtjj}tj|}tj|}tj}tjd| }tj ddgd}||}tj ddgd}|||}| | | dS) z2 train script for parallel training check c(eZdZdZfdZdZxZS))train_for_run_parallel..LinearNetz? simple fc network for parallel training check cttjdd|_tjdd|_dS)N r#)super__init__rrrD_linear1_linear2)self __class__s rroz2train_for_run_parallel..LinearNet.__init__sO GG     "I,,R44DM"I,,R33DMMMr cR|||S)z! forward )rqrp)rrrs rforwardz1train_for_run_parallel..LinearNet.forwards"==q!1!122 2r )__name__ __module__ __qualname____doc__roru __classcell__)rss@r LinearNetrksQ   4 4 4 4 4  3 3 3 3 3 3 3r r{r>r?rmrr#N)rrLayer distributedinit_parallel_env DataParallelMSELossrGrHrArandnrFrI clear_grad) r{layerdp_layerloss_fnadaminputsoutputslabelslosss rtrain_for_run_parallelrs 33333FIO333  ((*** IKKE"5))Hi!!G   (;(;(=(= !  D\2r(I . .FhvG \2q'9 - -F 77F # #DMMOOOIIKKKOOr cltjtt |dS)z Testing the simple network in data parallel mode, using multiple CPU/GPU. Args: use_cuda (bool): Whether running with CUDA. use_xpu (bool): Whether running with XPU. device_list (int): The specified devices. )nprocsN)rr}spawnrr*) device_lists r _run_parallelrs/ 3C >> import paddle >>> # doctest: +SKIP('the output will change in different run') >>> paddle.utils.run_check() Running verify PaddlePaddle program ... I0818 15:35:08.335391 30540 program_interpreter.cc:173] New Executor is Running. I0818 15:35:08.398319 30540 interpreter_util.cc:529] Standalone Executor is Used. PaddlePaddle works well on 1 CPU. PaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now. z(Running verify PaddlePaddle program ... FNrTr#zWMore than one kind of custom devices detected, but run check would only be executed on .GPUXPUCPU) device_countzPaddlePaddle works well on 1 xcclPADDLE_DISTRI_BACKENDzPaddlePaddle works well on  zs.zXPaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now.z%PaddlePaddle meets some problem with aqs. This may be caused by: 1. There is not enough GPUs visible on your system 2. Some GPUs are occupied by other process now 3. NVIDIA-NCCL2 is not installed correctly on your system. Please follow instruction on https://github.com/NVIDIA/nccl-tests to test your NCCL, or reinstall it following https://docs.nvidia.com/deeplearning/sdk/nccl-install-guide/index.htmlz Original Error is: z7PaddlePaddle is installed successfully ONLY for single z2! Let's start deep learning with PaddlePaddle now.)printris_compiled_with_cudar1is_compiled_with_xpur5r* frameworkcoreget_all_custom_device_typer-r.rr+r4listrangeget_custom_device_count cpu_placesrhrQosenvironrr,) rJrKrLrM device_strrrrr0s r run_checkrsX& 4555HGJ #%% %''  $ & &#%% V  " = = ? ? @ @1 D D v$??AA B BQ F F Ocjpjzjk[k[k]k]^_k`ccc   ? m//11 ? m..00  ?%*EEGGJ '  %==&     m..A.>> {##Lx*6HIII':7IJJJ 7* 7 7 7888 {  a   = 6< 23 + & & & M MMzMMM N N N f       EL E E: E E E    444555  ?j ? ? ?   s AI99 KAK  K)rr) __future__rr-numpyr$r__all__rr(r1r5rQrhrrrr rrs#""""" $,,,       """J###L'''T N N NWWWWWWr