mindspore lite 使用Ascend310P推理失败

minder · October 21, 2025, 8:57am

使用官网云侧推理案例，使用CPU的情况下推理正常，而是用Ascend310P，即在创建上下文的C++部分，同样使用官网教程(使用Ascend后端)替换部分代码，即出现错误:

[ERROR] ME(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.976.559 [er/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:107] BuildAirModel] Call aclgrphBuildInitialize fail: E40001: [PID: 1493515] 2025-10-21-16:46:08.935.688 Value [/usr/local/bin:/opt/software/jdk1.8.0_301/bin:/usr/local/Ascend/ascend-toolkit/latest/bin:/usr/local/Ascend/ascend-toolkit/latest/compiler/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/tools/ccec_compiler/bin:/opt/software/jdk1.8.0_301/bin:/usr/local/Ascend/ascend-toolkit/latest/bin:/usr/local/Ascend/ascend-toolkit/latest/compiler/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/tools/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/bin:/usr/local/Ascend/ascend-toolkit/latest/compiler/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/tools/ccec_compiler/bin:/opt/software/mysql-8.0.35-linux-glibc2.28-aarch64/bin:/home/software/jdk1.8.0_301/bin:/home/haohan/tiss/bin/:/hdf/bin/:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/root/bin] for environment variable [PATH] is invalid when executing the cmd python3-config --prefix.Possible Cause: The current value of environment variable is not effective.Solution: Reset the environment variable by referring to the installation guide.TraceBack (most recent call last):Fail to initialize handle manager.[FUNC:Initialize][FILE:python_adapter_manager.cc][LINE:40]Failed to initialize TeConfigInfo.[GraphOpt][InitializeInner][InitTbeFunc] Failed to init tbe.[FUNC:InitializeTeFusion][FILE:tbe_op_store_adapter.cc][LINE:1921][GraphOpt][InitializeInner][InitTeFusion]: Failed to initialize TeFusion.[FUNC:InitializeInner][FILE:tbe_op_store_adapter.cc][LINE:1888][SubGraphOpt][PreCompileOp][InitAdapter] InitializeAdapter adapter [tbe_op_adapter] failed! Ret [4294967295][FUNC:InitializeAdapter][FILE:op_store_adapter_manager.cc][LINE:79][SubGraphOpt][PreCompileOp][Init] Initialize op store adapter failed, OpsStoreName[tbe-custom].[FUNC:Initialize][FILE:op_store_adapter_manager.cc][LINE:120][FusionMngr][Init] Op store adapter manager init failed.[FUNC:Initialize][FILE:fusion_manager.cc][LINE:115]PluginManager InvokeAll failed.[FUNC:Initialize][FILE:ops_kernel_manager.cc][LINE:83]OpsManager initialize failed.[FUNC:InnerInitialize][FILE:gelib.cc][LINE:239]GELib::InnerInitialize failed.[FUNC:Initialize][FILE:gelib.cc][LINE:164]ctx is NULL![FUNC:GetDevErrMsg][FILE:api_impl.cc][LINE:6147]The argument is invalid.Reason: rtGetDevMsg execute failed, reason=[context pointer null]

[ERROR] ME(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.982.876 [er/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:217] LoadMindIR] Convert model from MindIR to OM failed[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.994.695 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:979] ConvertGraphToOm] Model converter load mindir failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.994.858 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1028] BuildGraph] Convert graph  to om failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.994.952 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1322] Run] Build graph failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.995.031 [mindspore-lite/tools/converter/adapter/acl/acl_pass.cc:42] Run] Acl pass impl run failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.995.108 [mindspore-lite/tools/converter/anf_transform.cc:488] RunConvertPass] Acl pass failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.995.191 [mindspore-lite/tools/converter/anf_transform.cc:681] RunPass] Run convert pass failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.995.263 [mindspore-lite/tools/converter/anf_transform.cc:782] TransformFuncGraph] Proc online transform failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.995.592 [mindspore-lite/tools/converter/anf_transform.cc:892] Transform] optimizer failed.[ERROR] LITE(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.995.688 [mindspore-lite/tools/converter/converter_funcgraph.cc:557] Optimize] Transform anf graph failed.[ERROR] ME(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.999.635 [convert/runtime_convert.cc:218] RuntimeConvert] Convert model failed[ERROR] ME(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.999.767 [cxx_api/model/model_impl.cc:744] ConvertGraphOnline] Failed to converter graph[ERROR] ME(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:08.999.867 [cxx_api/model/model_impl.cc:521] BuildByBufferImpl] convert graph failed!ret = Common error code.[ERROR] ME(1493515,ffff8d5cf020,mslite):2025-10-21-16:46:09.005.708 [cxx_api/model/model.cc:101] Build] impl_->Build failed! ret = Common error code.Build model error Common error code.

YeFeng24 · October 21, 2025, 11:25am

看了相关报错，貌似是昇腾相关的环境有问题（包含驱动），可以提供一下详细的报错日志，我们再看一下详细的报错

YeFeng24 · October 21, 2025, 11:26am

日志收集方式：

export ASCEND_GLOBAL_LOG_LEVEL=1

export ASCEND_SLOG_PRINT_TO_STDOUT=1

chengxiaoli · October 22, 2025, 12:59am

@minder 用户您好，请参考上述日志收集方式，补充下信息~

minder · October 22, 2025, 2:36am

[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.278 [platform_utils.cc:95]175839 Initialize:“The value of [ge.socVersion] from ge options is [Ascend310P3].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.286 [platform_utils.cc:109]175839 Initialize:“Current soc version is [Ascend310P3].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.754 [platform_utils.cc:319]175839 ParseIsaArchVersion:“Isa arch version is [v200].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.768 [platform_utils.cc:331]175839 ParseSupportFixpipe:“Is support fixpipe is [0]”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.780 [platform_utils.cc:347]175839 ParseFixpipeSupportMultiOutput:“The flag of is_fixpipe_support_multi_output is [1]”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.786 [platform_utils.cc:359]175839 ParseContextSwitch:“Is support context switch is [0]”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.792 [platform_utils.cc:371]175839 ParseDsaWorkspaceSize:“Dsa workspace size is [0]”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.798 [platform_utils.cc:387]175839 ParseFftsMode:“Ffts mode is [0].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.803 [platform_utils.cc:406]175839 ParseCubeVecState:“Cube vector state is [0].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.914 [platform_utils.cc:440]175839 ParseAllowHf32:“Enable allow hf32 is [0].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.951 [platform_utils.cc:423]175839 ParseCubeHighPrecison:“Enable cube high precision is [0].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.969 [platform_utils.cc:453]175839 ParseL2Type:“L2 type is [0].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.975 [platform_utils.cc:468]175839 ParseL2CacheMode:“L2 cache mode is [0].”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.983 [platform_utils.cc:480]175839 ParseSupportVectorEngine:“Parameter[support_vector_engine] is [1]”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.990 [platform_utils.cc:492]175839 ParseSpecifiedMemBase:“Parameter[specified_mem_base] is [0]”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.527.995 [platform_utils.cc:505]175839 ParseHardwareCoreSync:“Is hardware support core sync is [0]”
[INFO] GE(175839,mslite):2025-10-22-10:15:05.528.050 [transfer_shape_utils.cc:159]175839 operator():input platform is null, now get m0k0c0 from platformmanager by self
[INFO] GE(175839,mslite):2025-10-22-10:15:05.528.589 [op_lib_register.cc:58]175839 GetCustomOpLibPath:get op lib path is
[INFO] FE(175839,mslite):2025-10-22-10:15:05.528.612 [fe_type_utils.cc:66]175839 GetRealPath:“path /usr/local/Ascend/ascend-toolkit/8.2.RC1/opp/vendors/config.ini is not exist.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.528.618 [configuration.cc:794]175839 GetCustomOppPathFromOppConfigFile:“The opp config file path[/usr/local/Ascend/ascend-toolkit/8.2.RC1/opp/vendors/config.ini] is not existed.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.528.796 [configuration.cc:861]175839 LoadConfigFile:“The real path of fe.ini is /usr/local/Ascend/ascend-toolkit/8.2.RC1/aarch64-linux/lib64/plugin/opskernel/fe_config/fe.ini.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.529.825 [fe_type_utils.cc:66]175839 GetRealPath:“path /usr/local/Ascend/ascend-toolkit/8.2.RC1/opp//built-in/op_impl/ai_core/tbe/kernel/config/Ascend310P3/op_info_config.json is not exist.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.530.031 [configuration.cc:956]175839 AssembleOpsStoreInfoVector:“OP store [tbe-builtin] has been added to ops_store_info_vector_ of AIcoreEngine.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.530.056 [configuration.cc:956]175839 AssembleOpsStoreInfoVector:“OP store [tbe-custom] has been added to ops_store_info_vector_ of AIcoreEngine.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.531.182 [configuration.cc:365]175839 InitializeConfigParser:“Initialization of Configuration end.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.531.201 [fusion_manager.cc:58]175839 InitBinKernelInfoStore:“[FusionMngr][InitBinKernelInfoStore] Binary config file path is empty.”
[WARNING] FE(175839,mslite):2025-10-22-10:15:05.531.363 [configuration.cc:1131]175839 GetStringValue:“Can not find the value of key rootdir.”
[INFO] TEFUSION(175839,mslite):2025-10-22-10:15:05.550.099 [fusion_api.cc:119]175839 TbeInitialize Begin to do TbeInitialize
[INFO] TEFUSION(175839,mslite):2025-10-22-10:15:05.550.123 [fusion_api.cc:121]175839 TbeInitialize Get TeFusionManager lock.
[WARNING] TEFUSION(175839,mslite):2025-10-22-10:15:05.550.461 [tbe_handler_manager.cc:105]175839 LoadSo The so file path is not valid
[WARNING] TEFUSION(175839,mslite):2025-10-22-10:15:05.550.471 [tbe_handler_manager.cc:45]175839 Initialize Unable to initialize tbe functions.
[WARNING] TEFUSION(175839,mslite):2025-10-22-10:15:05.581.163 [py_decouple.cc:150]175839 LoadDynLibFromPyCfg Python path[/usr/local/lib/libpython3.11.so.1.0] is invalid, please confirm ld_library_path of the python library.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.590.190 [error_manager.cc:673]175839 ParseJsonFile:Parse json file:/usr/local/Ascend/ascend-toolkit/8.2.RC1/aarch64-linux/lib64/../conf/error_manager/error_code.json success
[INFO] GE(175839,mslite):2025-10-22-10:15:05.590.880 [error_manager.cc:413]175839 ReportErrMessage:report error_message, error_code:E40001, work_stream_id:17584075839, error_mode:0.
[ERROR] TEFUSION(175839,mslite):2025-10-22-10:15:05.590.953 [py_decouple.cc:47]175839 Initialize Launch dynamic-handle failed.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.013 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E49999, work_stream_id:17584075839, error_mode:0
[ERROR] TEFUSION(175839,mslite):2025-10-22-10:15:05.591.037 [python_adapter_manager.cc:40]175839 Initialize Fail to initialize handle manager.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.053 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E49999, work_stream_id:17584075839, error_mode:0
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.076 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:17584075839, error_mode:0
[ERROR] FE(175839,mslite):2025-10-22-10:15:05.591.087 [tbe_op_store_adapter.cc:1921]175839 InitializeTeFusion:“[GraphOpt][InitializeInner][InitTbeFunc] Failed to init tbe.”
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.115 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:17584075839, error_mode:0
[ERROR] FE(175839,mslite):2025-10-22-10:15:05.591.127 [tbe_op_store_adapter.cc:1888]175839 InitializeInner:“[GraphOpt][InitializeInner][InitTeFusion]: Failed to initialize TeFusion.”
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.145 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:17584075839, error_mode:0
[ERROR] FE(175839,mslite):2025-10-22-10:15:05.591.154 [op_store_adapter_manager.cc:79]175839 InitializeAdapter:“[SubGraphOpt][PreCompileOp][InitAdapter] InitializeAdapter adapter [tbe_op_adapter] failed! Ret [4294967295]”
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.232 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:17584075839, error_mode:0
[ERROR] FE(175839,mslite):2025-10-22-10:15:05.591.244 [op_store_adapter_manager.cc:120]175839 Initialize:“[SubGraphOpt][PreCompileOp][Init] Initialize op store adapter failed, OpsStoreName[tbe-custom].”
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.259 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:17584075839, error_mode:0
[ERROR] FE(175839,mslite):2025-10-22-10:15:05.591.268 [fusion_manager.cc:115]175839 Initialize:“[FusionMngr][Init] Op store adapter manager init failed.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.591.294 [op_store_adapter_manager.cc:155]175839 Finalize:“OpStoreAdapterManager finalize successfully.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.591.323 [configuration.cc:580]175839 Finalize:“Configuration finalize successfully.”
[INFO] FE(175839,mslite):2025-10-22-10:15:05.591.346 [fusion_manager.cc:196]175839 Finalize:“[FE_PERFORMANCE]The time cost of FusionManager::Finalize is [69] micro second.”
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.354 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Initialize] in [libfe.so] is [64314] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.372 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E19999, work_stream_id:17584075839, error_mode:0
[ERROR] GE(175839,mslite):2025-10-22-10:15:05.591.397 [ops_kernel_manager.cc:83]175839 Initialize: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER]PluginManager InvokeAll failed.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.417 [gelib.cc:236]175839 InnerInitialize:[GEPERFTRACE] The time cost of InnerInitialize::OpsManagerInitialize is [1600427] micro seconds.
[ERROR] GE(175839,mslite):2025-10-22-10:15:05.591.425 [gelib.cc:238]175839 InnerInitialize: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER][Init][OpsManager]GE ops manager initial failed.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.437 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E19999, work_stream_id:17584075839, error_mode:0
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.447 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: AIcoreEngine.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.453 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_HCCL.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.457 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_AICPU.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.462 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_AICPU_ASCEND.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.466 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_AICPU_ASCEND_FFTS_PLUS.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.471 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_AICPU_FFTS_PLUS.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.475 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_DVPP.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.479 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_GE_LOCAL.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.484 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_HOST_CPU.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.488 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_RTS.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.493 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DNN_VM_RTS_FFTS_PLUS.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.497 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: DSAEngine.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.501 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: VectorEngine.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.506 [dnnengine_manager.cc:177]175839 Finalize:DNNEngine name: ffts_plus.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.539 [ops_kernel_manager.cc:261]175839 Finalize:free ops kernel resource.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.591.544 [ops_kernel_manager.cc:417]175839 FinalizeOpsKernel:ge invoke ops kernel finalize.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.606.537 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libaicpu_ascend_engine.so] is [14982] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.606.583 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libaicpu_tf_engine.so] is [17] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.606.594 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libdvpp_engine.so] is [2] micro seconds.
[WARNING] FE(175839,mslite):2025-10-22-10:15:05.606.613 [fusion_manager.cc:200]175839 Finalize:“Already Finalized, directly return.”
[WARNING] FE(175839,mslite):2025-10-22-10:15:05.606.622 [fusion_manager.cc:200]175839 Finalize:“Already Finalized, directly return.”
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.356 atrace_api.c:297 AtraceEventReport end, ret=0.
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.375 atrace_api.c:118 AtraceDestroy start.
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.382 tracer_mgr_operate.c:248 destroy object FE_Global_Trace, exitSave(false).
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.388 atrace_api.c:120 AtraceDestroy end.
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.393 atrace_api.c:118 AtraceDestroy start.
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.398 tracer_mgr_operate.c:248 destroy object FE_Statistics_Trace, exitSave(false).
[INFO] ATRACE(175839,mslite):2025-10-22-10:15:05.615.403 atrace_api.c:120 AtraceDestroy end.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.418 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libfe.so] is [8817] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.434 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libffts.so] is [8] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.448 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libge_local_engine.so] is [7] micro seconds.
[INFO] HCCL(175839,mslite):2025-10-22-10:15:05.615.458 [plugin_manager.cc:56] [175839]hccl ops plugin finalize start.
[INFO] HCCL(175839,mslite):2025-10-22-10:15:05.615.485 [plugin_manager.cc:62] [175839]hccl ops plugin finalize end. ret[0]
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.491 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libhcom_graph_adaptor.so] is [37] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.535 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libhost_cpu_engine.so] is [37] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.551 [dnnengine_manager.cc:41]175839 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [librts_engine.so] is [9] micro seconds.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.565 [process_node_engine_manager.cc:85]175839 Finalize:ProcessNodeEngine id:HOST_CPU.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.574 [process_node_engine_manager.cc:85]175839 Finalize:ProcessNodeEngine id:NPU.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.580 [process_node_engine_manager.cc:85]175839 Finalize:ProcessNodeEngine id:PS.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.585 [process_node_engine_manager.cc:85]175839 Finalize:ProcessNodeEngine id:UDF.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.607 [graph_external_weight_manager.cc:107]175839 Destroy:Success to destroy external weight manager pool
[ERROR] GE(175839,mslite):2025-10-22-10:15:05.615.626 [gelib.cc:163]175839 Initialize: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER][Init][GeLib]GeLib initial failed.
[INFO] GE(175839,mslite):2025-10-22-10:15:05.615.674 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:E19999, work_stream_id:17584075839, error_mode:0
[ERROR] GE(175839,mslite):2025-10-22-10:15:05.615.713 [ge_ir_build.cc:338]175839 aclgrphBuildInitializeImpl: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER][Init][GELib] failed!
[INFO] ASCENDCL(175839,mslite):2025-10-22-10:15:05.615.749 [acl.cpp:780]175839 aclGetRecentErrMsg: start to execute aclGetRecentErrMsg.
[WARNING] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.012 [api_impl.cc:258] 175839 CurrentContext: current ctx is nullptr!
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.026 [api_impl.cc:6147]175839 GetDevErrMsg:[INIT][OPS_KER]report error module_type=3, module_name=EE8888
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.033 [api_impl.cc:6147]175839 GetDevErrMsg:[INIT][OPS_KER]ctx is NULL!
[INFO] GE(175839,mslite):2025-10-22-10:15:05.616.056 [error_manager.cc:360]175839 ReportInterErrMessage:report error_message, error_code:EE8888, work_stream_id:17584075839, error_mode:0
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.071 [api_impl.cc:6228]175839 GetDevMsg:[INIT][OPS_KER]Failed to GetDeviceErrMsg, retCode=0x7070001.
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.083 [api_error.cc:3916]175839 GetDevMsg:[INIT][OPS_KER]GetDeviceMsg failed, getMsgType=0.
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.103 [api_c_device.cc:613]175839 rtGetDevMsg:[INIT][OPS_KER]ErrCode=107002, desc=[context pointer null], InnerCode=0x7070001
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.110 [error_message_manage.cc:48]175839 FuncErrorReason:[INIT][OPS_KER]report error module_name=EE1001
[ERROR] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.119 [error_message_manage.cc:48]175839 FuncErrorReason:[INIT][OPS_KER]rtGetDevMsg execute failed, reason=[context pointer null]
[INFO] GE(175839,mslite):2025-10-22-10:15:05.616.142 [error_manager.cc:413]175839 ReportErrMessage:report error_message, error_code:EE1001, work_stream_id:17584075839, error_mode:0.
[WARNING] RUNTIME(175839,mslite):2025-10-22-10:15:05.616.175 [api_impl.cc:258] 175839 CurrentContext: current ctx is nullptr!
[INFO] ASCENDCL(175839,mslite):2025-10-22-10:15:05.616.182 [acl.cpp:749]175839 GetFaultEventInfo: can not get device id, runtime errorCode is 107002
[INFO] GE(175839,mslite):2025-10-22-10:15:05.616.188 [error_manager.cc:989]175839 GetRawErrorMessages:current work_stream_id:17584075839
[ERROR] ME(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.616.294 [er/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:107] BuildAirModel] Call aclgrphBuildInitialize fail: E40001: [PID: 175839] 2025-10-22-10:15:05.581.333 Value [/usr/local/bin:/opt/software/jdk1.8.0_301/bin:/usr/local/Ascend/ascend-toolkit/latest/bin:/usr/local/Ascend/ascend-toolkit/latest/compiler/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/tools/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/bin:/usr/local/Ascend/ascend-toolkit/latest/compiler/ccec_compiler/bin:/usr/local/Ascend/ascend-toolkit/latest/tools/ccec_compiler/bin:/opt/software/mysql-8.0.35-linux-glibc2.28-aarch64/bin:/home/software/jdk1.8.0_301/bin:/home/haohan/tiss/bin/:/hdf/bin/:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/root/bin] for environment variable [PATH] is invalid when executing the cmd python3-config --prefix.
Possible Cause: The current value of environment variable is not effective.
Solution: Reset the environment variable by referring to the installation guide.
TraceBack (most recent call last):
Fail to initialize handle manager.[FUNC:Initialize][FILE:python_adapter_manager.cc][LINE:40]
Failed to initialize TeConfigInfo.
[GraphOpt][InitializeInner][InitTbeFunc] Failed to init tbe.[FUNC:InitializeTeFusion][FILE:tbe_op_store_adapter.cc][LINE:1921]
[GraphOpt][InitializeInner][InitTeFusion]: Failed to initialize TeFusion.[FUNC:InitializeInner][FILE:tbe_op_store_adapter.cc][LINE:1888]
[SubGraphOpt][PreCompileOp][InitAdapter] InitializeAdapter adapter [tbe_op_adapter] failed! Ret [4294967295][FUNC:InitializeAdapter][FILE:op_store_adapter_manager.cc][LINE:79]
[SubGraphOpt][PreCompileOp][Init] Initialize op store adapter failed, OpsStoreName[tbe-custom].[FUNC:Initialize][FILE:op_store_adapter_manager.cc][LINE:120]
[FusionMngr][Init] Op store adapter manager init failed.[FUNC:Initialize][FILE:fusion_manager.cc][LINE:115]
PluginManager InvokeAll failed.[FUNC:Initialize][FILE:ops_kernel_manager.cc][LINE:83]
OpsManager initialize failed.[FUNC:InnerInitialize][FILE:gelib.cc][LINE:239]
GELib::InnerInitialize failed.[FUNC:Initialize][FILE:gelib.cc][LINE:164]
ctx is NULL![FUNC:GetDevErrMsg][FILE:api_impl.cc][LINE:6147]
The argument is invalid.Reason: rtGetDevMsg execute failed, reason=[context pointer null]

[ERROR] ME(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.623.435 [er/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:217] LoadMindIR] Convert model from MindIR to OM failed
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.493 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:979] ConvertGraphToOm] Model converter load mindir failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.529 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1028] BuildGraph] Convert graph to om failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.541 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1322] Run] Build graph failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.559 [mindspore-lite/tools/converter/adapter/acl/acl_pass.cc:42] Run] Acl pass impl run failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.573 [mindspore-lite/tools/converter/anf_transform.cc:488] RunConvertPass] Acl pass failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.593 [mindspore-lite/tools/converter/anf_transform.cc:681] RunPass] Run convert pass failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.604 [mindspore-lite/tools/converter/anf_transform.cc:782] TransformFuncGraph] Proc online transform failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.805 [mindspore-lite/tools/converter/anf_transform.cc:892] Transform] optimizer failed.
[ERROR] LITE(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.632.817 [mindspore-lite/tools/converter/converter_funcgraph.cc:557] Optimize] Transform anf graph failed.
[ERROR] ME(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.636.373 [convert/runtime_convert.cc:218] RuntimeConvert] Convert model failed
[ERROR] ME(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.636.408 [cxx_api/model/model_impl.cc:744] ConvertGraphOnline] Failed to converter graph
[ERROR] ME(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.636.432 [cxx_api/model/model_impl.cc:521] BuildByBufferImpl] convert graph failed!ret = Common error code.
[ERROR] ME(175839,ffffa1c9c020,mslite):2025-10-22-10:15:05.641.818 [cxx_api/model/model.cc:101] Build] impl_->Build failed! ret = Common error code.
Build model error Common error code.

minder · October 22, 2025, 2:37am

已补充

YeFeng24 · October 22, 2025, 7:53am

看报错日志，你的机器上是不是没有安装CANN相关的版本包以及对应的驱动包？

你的机器上插了Ascend的推理卡吗？

如果你的机器是插入了Ascend卡的机器的话，确认一下你的CANN以及驱动的安装路径；

一般默认情况安装的话，在/usr/local/Ascend/ascend-toolkit/下应该有一个set_env.sh文件，请执行一下source set_env.sh，然后再尝试一下是否可以转换成功；

如果你的CANN包等安装路径，非默认路径，请找到对应的set_env.sh文件，并source生效即可。

minder · October 22, 2025, 8:39am

已经安装了，卡是有的，驱动等我觉得也没啥问题，因为已经成功使用mindspore并利用Ascend[Ascend310P]卡进行了训练，但是使用mindspore lite 推理就是有问题，错误不变

set_env的内容：
export LD_LIBRARY_PATH=/usr/local/Ascend/driver/lib64:/usr/local/Ascend/driver/lib64/common:/usr/local/Ascend/driver/lib64/driver:$LD_LIBRARY_PATH
export ASCEND_TOOLKIT_HOME=/usr/local/Ascend/ascend-toolkit/latest
export LD_LIBRARY_PATH=${ASCEND_TOOLKIT_HOME}/lib64:${ASCEND_TOOLKIT_HOME}/lib64/plugin/opskernel:${ASCEND_TOOLKIT_HOME}/lib64/plugin/nnengine:${ASCEND_TOOLKIT_HOME}/opp/built-in/op_impl/ai_core/tbe/op_tiling/lib/linux/$(arch):$LD_LIBRARY_PATH
export LD_LIBRARY_PATH=${ASCEND_TOOLKIT_HOME}/tools/aml/lib64:${ASCEND_TOOLKIT_HOME}/tools/aml/lib64/plugin:$LD_LIBRARY_PATH
export PYTHONPATH=${ASCEND_TOOLKIT_HOME}/python/site-packages:${ASCEND_TOOLKIT_HOME}/opp/built-in/op_impl/ai_core/tbe:$PYTHONPATH
export PATH=${ASCEND_TOOLKIT_HOME}/bin:${ASCEND_TOOLKIT_HOME}/compiler/ccec_compiler/bin:${ASCEND_TOOLKIT_HOME}/tools/ccec_compiler/bin:$PATH
export ASCEND_AICPU_PATH=${ASCEND_TOOLKIT_HOME}
export ASCEND_OPP_PATH=${ASCEND_TOOLKIT_HOME}/opp
export TOOLCHAIN_HOME=${ASCEND_TOOLKIT_HOME}/toolkit
export ASCEND_HOME_PATH=${ASCEND_TOOLKIT_HOME}

latest 目录:

YeFeng24 · October 22, 2025, 9:06am

你是直接拿mindspore训练导出的mindir模型，使用mindspore lite进行推理的吗？

流程能大概说一下吗？

按道理你应该是mindspore训练导出mindir，然后使用mindspore lite提供的converter_lite将这个mindir转换一下，得到一个新的mindir，然后新的mindir去部署推理，你现在的报错是在哪个流程报错的？

minder · October 22, 2025, 9:15am

使用mindspore训练模型保存了最好的checkpoint，然后使用mindspore的export函数导出模型（已做）：

export(model, flow, seq, seq_attention_mask, file_name=“flow_cnn_res_model”, file_format=“MINDIR”)
然后使用mindspore云侧推理C++方式进行推理(未做) 这是一个预想的思路。但是我得先把环境验证没有问题再进行我自己的推理。因此我完全按照官网的demo进行推理，demo的模型也是从官网下载的示例，相对于官网CPU示例，只对上下文时从CPU修改为了Ascend。就是直接使用官网的例子出错的我贴一下代码：

// mslite.cpp: 定义应用程序的入口点。
//

#include
#include
#include
#include
#include
#include
#include “include/api/model.h”
#include “include/api/context.h”
#include “include/api/status.h”
#include “include/api/types.h”

template <typename T, typename Distribution>
void GenerateRandomData(int size, void* data, Distribution distribution) {
std::mt19937 random_engine;
int elements_num = size / sizeof(T);
(void)std::generate_n(static_cast<T*>(data), elements_num,
&distribution, &random_engine { return static_cast(distribution(random_engine)); });
}

int GenerateInputDataWithRandom(std::vectormindspore::MSTensor inputs) {
for (auto tensor : inputs) {
auto input_data = tensor.MutableData();
if (input_data == nullptr) {
std::cerr << “MallocData for inTensor failed.” << std::endl;
return -1;
}
GenerateRandomData(tensor.DataSize(), input_data, std::uniform_real_distribution(0.1f, 1.0f));
}
return 0;
}

int QuickStart(int argc, const char** argv) {
if (argc < 3) {
std::cerr << “Model file and device id must be provided.\n”;
return -1;
}
// Read model file.
std::string model_path = argv[1];
if (model_path.empty()) {
std::cerr << “Model path " << model_path << " is invalid.”;
return -1;
}

// Create and init context, add CPU device info
auto context = std::make_shared<mindspore::Context>();
if (context == nullptr) {
    std::cerr << "New context failed." << std::endl;
    return -1;
}
auto& device_list = context->MutableDeviceInfo();
auto device_info = std::make_shared<mindspore::AscendDeviceInfo>();
if (device_info == nullptr) {
    std::cerr << "New AscendDeviceInfo failed." << std::endl;
    return -1;
}
// Set Atlas 200/300/500 inference product, Atlas inference series, Atlas training series device id.
int device_id = atoi(argv[2]);
device_info->SetDeviceID(device_id);
device_info->SetPrecisionMode("enforce_fp32");
device_list.push_back(device_info);

mindspore::Model model;
// Build model
auto build_ret = model.Build(model_path, mindspore::kMindIR, context);
if (build_ret != mindspore::kSuccess) {
    std::cerr << "Build model error " << build_ret << std::endl;
    return -1;
}

// Get Input
auto inputs = model.GetInputs();
// Generate random data as input data.
if (GenerateInputDataWithRandom(inputs) != 0) {
    std::cerr << "Generate Random Input Data failed." << std::endl;
    return -1;
}

// Model Predict
std::vector<mindspore::MSTensor> outputs;
auto predict_ret = model.Predict(inputs, &outputs);
if (predict_ret != mindspore::kSuccess) {
    std::cerr << "Predict error " << predict_ret << std::endl;
    return -1;
}

// Print Output Tensor Data.
constexpr int kNumPrintOfOutData = 50;
for (auto& tensor : outputs) {
    std::cout << "tensor name is:" << tensor.Name() << " tensor size is:" << tensor.DataSize()
        << " tensor elements num is:" << tensor.ElementNum() << std::endl;
    auto out_data = reinterpret_cast<const float*>(tensor.Data().get());
    std::cout << "output data is:";
    for (int i = 0; i < tensor.ElementNum() && i <= kNumPrintOfOutData; i++) {
        std::cout << out_data[i] << " ";
    }
    std::cout << std::endl;
}
return 0;

}

int main(int argc, const char** argv) { return QuickStart(argc, argv); }

YeFeng24 · October 22, 2025, 1:24pm

这个报错，应该是环境不对导致的，可以确认一下以下几点：

1、你安装的cann包的python环境和当前执行的python环境是否是同一个环境；

2、/usr/local/Ascend/ascend-toolkit/latest/lib64/te-0.4.0-py3-none-any.whl （这个是安装默认路径的对应文件，你如果安装到其他目录，自行参考替换）确认这个whl是否安装；

目前从报错来看就是环境问题；

你当前用的是镜像还是裸机环境，你的看一下你当前LD_LIBRARY_PATH对应的路径有哪些；以及执行一下env | grep ASCEND，看一下显示的有哪些环境变量路径。

minder · October 23, 2025, 2:24am

YeFeng24:

使用的裸机，非docker。你说的python环境问题我在检查了软连接之后重新安装了一下toolkit和kernel(冲洗运行还是一样的错误)

Python 3.11.9

whl已安装：

Processing /usr/local/Ascend/ascend-toolkit/latest/lib64/te-0.4.0-py3-none-any.whl
Requirement already satisfied: attrs in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (25.4.0)
Requirement already satisfied: cloudpickle in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (3.1.1)
Requirement already satisfied: decorator in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (5.2.1)
Requirement already satisfied: ml-dtypes in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (0.5.3)
Requirement already satisfied: numpy in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (1.26.4)
Requirement already satisfied: psutil in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (7.1.1)
Requirement already satisfied: scipy in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (1.16.2)
Requirement already satisfied: tornado in /usr/local/lib/python3.11/site-packages (from te==0.4.0) (6.5.2)

系统环境：

ASCEND_TOOLKIT_HOME=/usr/local/Ascend/ascend-toolkit/latest
ASCEND_OPP_PATH=/usr/local/Ascend/ascend-toolkit/latest/opp
ASCEND_AICPU_PATH=/usr/local/Ascend/ascend-toolkit/latest
LOCAL_ASCEND=/usr/local/Ascend
ASCEND_HOME_PATH=/usr/local/Ascend/ascend-toolkit/latest

YeFeng24 · October 23, 2025, 6:50am

你的看一下你当前LD_LIBRARY_PATH对应的路径有哪些

这个也提供一下

YeFeng24 · October 23, 2025, 6:50am

LD_LIBRARY_PATH对应的变量值也提供一下

minder · October 24, 2025, 2:32am

环境问题解决了，是因为环境中安装了两套CANN，一个8.0，一个是8.2，不知道为什么，推理时用的8.0的版本，卸载8.0之后，环境问题应该是没有了，不过出现了另一个问题

[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.906 [le_utils.cc:230] RealPath] file path not exists.
[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.940 [er/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:195] LoadMindIR] real_path is invalid.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.956 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:979] ConvertGraphToOm] Model converter load mindir failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.969 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1028] BuildGraph] Convert graph to om failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.980 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1322] Run] Build graph failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.992 [mindspore-lite/tools/converter/adapter/acl/acl_pass.cc:42] Run] Acl pass impl run failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.963.004 [mindspore-lite/tools/converter/anf_transform.cc:488] RunConvertPass] Acl pass failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.963.024 [mindspore-lite/tools/converter/anf_transform.cc:681] RunPass] Run convert pass failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.963.035 [mindspore-lite/tools/converter/anf_transform.cc:782] TransformFuncGraph] Proc online transform failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.963.126 [mindspore-lite/tools/converter/anf_transform.cc:892] Transform] optimizer failed.
[ERROR] LITE(260654,ffff97681020,mslite):2025-10-24-10:24:27.963.139 [mindspore-lite/tools/converter/converter_funcgraph.cc:557] Optimize] Transform anf graph failed.
[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.964.225 [convert/runtime_convert.cc:218] RuntimeConvert] Convert model failed
[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.964.248 [cxx_api/model/model_impl.cc:744] ConvertGraphOnline] Failed to converter graph
[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.964.268 [cxx_api/model/model_impl.cc:521] BuildByBufferImpl] convert graph failed!ret = Common error code.
[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.968.856 [cxx_api/model/model.cc:101] Build] impl_->Build failed! ret = Common error code.
Build model error Common error code.

YeFeng24 · October 24, 2025, 2:35am

[ERROR] ME(260654,ffff97681020,mslite):2025-10-24-10:24:27.962.906 [le_utils.cc:230] RealPath] file path not exists.
文件路径没有找到，你截图看一下你的转换命令，以及–modelFile对应的路径下的所有文件，看看是否缺少啥东西

minder · October 24, 2025, 2:44am

我没有使用转换命令，这个是直接加载的mobilenetv2.mindir（官网下载的模型案例）

mindspore::Model model;
// Build model
auto build_ret = model.Build(model_path, mindspore::kMindIR, context);
if (build_ret != mindspore::kSuccess) {
std::cerr << "Build model error " << build_ret << std::endl;
return -1;
}

意思是需要我再使用转换程序转一遍是么？

YeFeng24 · October 24, 2025, 4:02am

按照现在的报错日志，就是你的文件没有找到，看这个报错，应该就是mindir文件没有找到

minder · October 24, 2025, 6:29am

不是模型文件没找到，我修改了路径报的错误不一样
[ERROR] ME(638459,ffff919a0020,mslite):2025-10-24-14:22:47.226.487 [cxx_api/serialization.cc:55] ReadFile] File: /home/jy/ms-c/model/mobilenetv1.mindir does not exist.
[ERROR] ME(638459,ffff919a0020,mslite):2025-10-24-14:22:47.226.529 [cxx_api/model/model_impl.cc:706] Build] Failed to read buffer from model file.
[ERROR] ME(638459,ffff919a0020,mslite):2025-10-24-14:22:47.226.544 [cxx_api/model/model.cc:101] Build] impl_->Build failed! ret = Common error code.
Build model error Common error code.

YeFeng24 · October 24, 2025, 6:32am

从报错看，是这个文件没有找到`/home/jy/ms-c/model/mobilenetv1.mindir`，在你执行的目录下，可以访问到这个文件吗？

Topic		Replies	Views
MindSpore模型加载报错RuntimeError: build from file failed! Error is Common error code. 模型训练-Model Training	0	20	August 31, 2025
使用MindSpore Lite端侧模型转换工具将YOLOv8.onnx转为.ms报错Convert failed. Ret: Common error code. 推理经验-Inference Experience	1	39	October 31, 2025
模型转换失败：not support onnx data type IsNaN 端侧部署 Lite	28	157	October 27, 2025
MindSpore Lite模型加载报错RuntimeError: build from file failed! Error is Common error code. 推理经验-Inference Experience	0	11	September 9, 2025
win11 下使用C++运行官网推理示例出错端侧部署 Lite 推理	10	95	October 31, 2025

mindspore lite 使用Ascend310P推理失败

Related topics