模型转化失败，似乎是环境问题，但是使用mindspore模型训练并不受影响

端侧部署 Lite
1

这是一部分的错误日志

[WARNING] FE(3703,converter_lite):2025-11-26-09:45:17.158.606 [configuration.cc:1131]3703 GetStringValue:"Can not find the value of key rootdir."
[INFO] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.169.956 [fusion_api.cc:119]3703 TbeInitialize Begin to do TbeInitialize
[INFO] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.169.976 [fusion_api.cc:121]3703 TbeInitialize Get TeFusionManager lock.
[WARNING] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.170.477 [tbe_handler_manager.cc:105]3703 LoadSo The so file path [] is not valid
[WARNING] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.170.489 [tbe_handler_manager.cc:45]3703 Initialize Unable to initialize tbe functions.
[WARNING] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.178.123 [py_decouple.cc:117]3703 CheckCommandValid The current env could'n found command [python3-config --prefix],cased by[sh: 1: python3-config: not found
]
[WARNING] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.178.262 [py_decouple.cc:155]3703 LoadDynLibFromPyCfg Cannot fetch the return value from cmd[python3-config --prefix].
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.186.587 [error_manager.cc:673]3703 ParseJsonFile:Parse json file:/usr/local/Ascend/ascend-toolkit/8.2.RC1/aarch64-linux/lib64/../conf/error_manager/error_code.json success
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.303 [error_manager.cc:413]3703 ReportErrMessage:report error_message, error_code:E40001, work_stream_id:370303703, error_mode:0.
[ERROR] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.187.385 [py_decouple.cc:47]3703 Initialize Launch dynamic-handle failed.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.428 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E49999, work_stream_id:370303703, error_mode:0
[ERROR] TEFUSION(3703,converter_lite):2025-11-26-09:45:17.187.450 [python_adapter_manager.cc:40]3703 Initialize Fail to initialize handle manager.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.466 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E49999, work_stream_id:370303703, error_mode:0
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.487 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:370303703, error_mode:0
[ERROR] FE(3703,converter_lite):2025-11-26-09:45:17.187.498 [tbe_op_store_adapter.cc:1921]3703 InitializeTeFusion:"[GraphOpt][InitializeInner][InitTbeFunc] Failed to init tbe."
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.525 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:370303703, error_mode:0
[ERROR] FE(3703,converter_lite):2025-11-26-09:45:17.187.538 [tbe_op_store_adapter.cc:1888]3703 InitializeInner:"[GraphOpt][InitializeInner][InitTeFusion]: Failed to initialize TeFusion."
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.555 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:370303703, error_mode:0
[ERROR] FE(3703,converter_lite):2025-11-26-09:45:17.187.564 [op_store_adapter_manager.cc:79]3703 InitializeAdapter:"[SubGraphOpt][PreCompileOp][InitAdapter] InitializeAdapter adapter [tbe_op_adapter] failed! Ret [4294967295]"
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.647 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:370303703, error_mode:0
[ERROR] FE(3703,converter_lite):2025-11-26-09:45:17.187.658 [op_store_adapter_manager.cc:120]3703 Initialize:"[SubGraphOpt][PreCompileOp][Init] Initialize op store adapter failed, OpsStoreName[tbe-custom]."
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.672 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E29999, work_stream_id:370303703, error_mode:0
[ERROR] FE(3703,converter_lite):2025-11-26-09:45:17.187.681 [fusion_manager.cc:115]3703 Initialize:"[FusionMngr][Init] Op store adapter manager init failed."
[INFO] FE(3703,converter_lite):2025-11-26-09:45:17.187.712 [op_store_adapter_manager.cc:155]3703 Finalize:"OpStoreAdapterManager finalize successfully."
[INFO] FE(3703,converter_lite):2025-11-26-09:45:17.187.740 [configuration.cc:580]3703 Finalize:"Configuration finalize successfully."
[INFO] FE(3703,converter_lite):2025-11-26-09:45:17.187.761 [fusion_manager.cc:196]3703 Finalize:"[FE_PERFORMANCE]The time cost of FusionManager::Finalize is [70] micro second."
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.772 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Initialize] in [libfe.so] is [32213] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.791 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E19999, work_stream_id:370303703, error_mode:0
[ERROR] GE(3703,converter_lite):2025-11-26-09:45:17.187.816 [ops_kernel_manager.cc:83]3703 Initialize: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER]PluginManager InvokeAll failed.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.840 [gelib.cc:236]3703 InnerInitialize:[GEPERFTRACE] The time cost of InnerInitialize::OpsManagerInitialize is [1222635] micro seconds.
[ERROR] GE(3703,converter_lite):2025-11-26-09:45:17.187.848 [gelib.cc:238]3703 InnerInitialize: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER][Init][OpsManager]GE ops manager initial failed.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.862 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E19999, work_stream_id:370303703, error_mode:0
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.875 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: AIcoreEngine.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.881 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_HCCL.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.887 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_AICPU.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.892 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_AICPU_ASCEND.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.898 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_AICPU_ASCEND_FFTS_PLUS.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.903 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_AICPU_FFTS_PLUS.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.909 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_DVPP.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.914 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_GE_LOCAL.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.919 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_HOST_CPU.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.924 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_RTS.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.929 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DNN_VM_RTS_FFTS_PLUS.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.934 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: DSAEngine.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.939 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: VectorEngine.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.946 [dnnengine_manager.cc:177]3703 Finalize:DNNEngine name: ffts_plus.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.969 [ops_kernel_manager.cc:261]3703 Finalize:free ops kernel resource.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.187.976 [ops_kernel_manager.cc:417]3703 FinalizeOpsKernel:ge invoke ops kernel finalize.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.203.583 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libaicpu_ascend_engine.so] is [15594] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.203.645 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libaicpu_tf_engine.so] is [35] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.203.841 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libdvpp_engine.so] is [2] micro seconds.
[WARNING] FE(3703,converter_lite):2025-11-26-09:45:17.203.861 [fusion_manager.cc:200]3703 Finalize:"Already Finalized, directly return."
[WARNING] FE(3703,converter_lite):2025-11-26-09:45:17.203.869 [fusion_manager.cc:200]3703 Finalize:"Already Finalized, directly return."
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.207 [atrace_api.c:297](tid:3703) AtraceEventReport end, ret=0.
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.231 [atrace_api.c:118](tid:3703) AtraceDestroy start.
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.239 [tracer_mgr_operate.c:248](tid:3703) destroy object FE_Global_Trace, exitSave(false).
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.247 [atrace_api.c:120](tid:3703) AtraceDestroy end.
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.253 [atrace_api.c:118](tid:3703) AtraceDestroy start.
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.258 [tracer_mgr_operate.c:248](tid:3703) destroy object FE_Statistics_Trace, exitSave(false).
[INFO] ATRACE(3703,converter_lite):2025-11-26-09:45:17.213.264 [atrace_api.c:120](tid:3703) AtraceDestroy end.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.283 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libfe.so] is [9434] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.299 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libffts.so] is [7] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.314 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libge_local_engine.so] is [8] micro seconds.
[INFO] HCCL(3703,converter_lite):2025-11-26-09:45:17.213.324 [plugin_manager.cc:56] [3703]hccl ops plugin finalize start.
[INFO] HCCL(3703,converter_lite):2025-11-26-09:45:17.213.349 [plugin_manager.cc:62] [3703]hccl ops plugin finalize end. ret[0]
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.356 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libhcom_graph_adaptor.so] is [35] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.394 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [libhost_cpu_engine.so] is [31] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.417 [dnnengine_manager.cc:41]3703 ~InvokeFuncPerfRecorder:[GEPERFTRACE] The time cost of InvokeAll [Finalize] in [librts_engine.so] is [14] micro seconds.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.431 [process_node_engine_manager.cc:85]3703 Finalize:ProcessNodeEngine id:HOST_CPU.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.438 [process_node_engine_manager.cc:85]3703 Finalize:ProcessNodeEngine id:NPU.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.445 [process_node_engine_manager.cc:85]3703 Finalize:ProcessNodeEngine id:PS.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.450 [process_node_engine_manager.cc:85]3703 Finalize:ProcessNodeEngine id:UDF.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.483 [graph_external_weight_manager.cc:107]3703 Destroy:Success to destroy external weight manager pool
[ERROR] GE(3703,converter_lite):2025-11-26-09:45:17.213.502 [gelib.cc:163]3703 Initialize: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER][Init][GeLib]GeLib initial failed.
[INFO] GE(3703,converter_lite):2025-11-26-09:45:17.213.542 [error_manager.cc:360]3703 ReportInterErrMessage:report error_message, error_code:E19999, work_stream_id:370303703, error_mode:0
[ERROR] GE(3703,converter_lite):2025-11-26-09:45:17.213.576 [ge_ir_build.cc:338]3703 aclgrphBuildInitializeImpl: ErrorNo: 1343250441(There is no valid so about OpsKernelInfoStore or GraphOptimizer.) [INIT][OPS_KER][Init][GELib] failed!
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.213.637 [mindspore-lite/tools/converter/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:109] BuildAirModel] AclBuildInit failed!
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.221.049 [mindspore-lite/tools/converter/adapter/acl/cxx_api_lite/cxx_api/model/acl/model_converter.cc:221] LoadMindIR] Convert model from MindIR to OM failed
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.659 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1001] ConvertGraphToOm] Model converter load mindir failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.704 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1051] BuildGraph] Convert graph  to om failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.717 [mindspore-lite/tools/converter/adapter/acl/src/acl_pass_impl.cc:1456] Run] Build graph failed!
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.731 [mindspore-lite/tools/converter/adapter/acl/acl_pass.cc:42] Run] Acl pass impl run failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.743 [mindspore-lite/tools/converter/anf_transform.cc:489] RunConvertPass] Acl pass failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.760 [mindspore-lite/tools/converter/anf_transform.cc:682] RunPass] Run convert pass failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.231.771 [mindspore-lite/tools/converter/anf_transform.cc:783] TransformFuncGraph] Proc online transform failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.232.173 [mindspore-lite/tools/converter/anf_transform.cc:894] Transform] optimizer failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.232.187 [mindspore-lite/tools/converter/converter_funcgraph.cc:646] Optimize] Transform anf graph failed.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.232.213 [mindspore-lite/tools/converter/converter.cc:1211] HandleGraphCommon] Optimize func graph failed: -2 NULL pointer returned.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.232.229 [mindspore-lite/tools/converter/converter.cc:1160] Convert] Handle graph failed: -2 NULL pointer returned.
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.232.240 [mindspore-lite/tools/converter/converter.cc:1353] RunConverter] Convert model failed
[ERROR] LITE(3703,ffff83176020,converter_lite):2025-11-26-09:45:17.232.264 [mindspore-lite/tools/converter/cxx_api/converter.cc:361] Convert] Convert model failed, ret=NULL pointer returned.
ERROR [mindspore-lite/tools/converter/converter_lite/main.cc:107] main] Convert failed. Ret: NULL pointer returned.
2

来个大佬看看，是不是python环境的问题的啊？因为这是华为大模型镜像docker自带的，就直接用的这个python环境

3

能补充说一下相关背景吗？啥模型？啥问题？怎么操作的？什么硬件？什么系统相关信息补充完整

4
  1. 自己搭建的一个CNN模型，使用pytorch训练，转为ONNX格式
  2. 使用模型转换工具将ONNX格式转换为MINDIR格式
    转换命令：
./converter_lite --fmk=ONNX --modelFile=./torch_model.onnx --outputFile=invisight --device=Ascend
  1. 昇腾910B
  2. 系统信息
Linux f41e18047fe7 5.4.0-26-generic #30-Ubuntu SMP Mon Apr 20 16:59:40 UTC 2020 aarch64 aarch64 aarch64 GNU/Linux
5

如果使用ascend硬件推理的话，使用正确的命令，

./converter_lite --fmk=ONNX --modelFile=./torch_model.onnx --outputFile=invisight --optimize=ascend_oriented

6

可以参考这个教程：

7

排查出来了，就是因为python环境问题，华为docker镜像python环境不正常，替换掉就好了