============================= test session starts ============================== platform linux -- Python 3.7.5, pytest-5.4.3, py-1.8.1, pluggy-0.13.1 rootdir: /home/jenkins/mindspore/testcases/testcases/tests/st/dyn_shape_dev, inifile: /home/jenkins/sault/virtual_test/virtualenv_006/sault/config/pytest.ini plugins: anyio-3.7.1, forked-1.1.3, xdist-1.32.0 [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:09.526.286 [trace_attr.c:105](tid:32458) platform is 1. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:09.526.433 [trace_recorder.c:114](tid:32458) use root path: /home/jenkins/ascend/atrace [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:09.526.458 [trace_signal.c:133](tid:32458) register signal handler for signo 2 succeed. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:09.526.469 [trace_signal.c:133](tid:32458) register signal handler for signo 15 succeed. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:09.966.427 [runtime.cc:1159] 32458 GetAicoreNumByLevel: workingDev_=0 [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:09.966.482 [runtime.cc:4719] 32458 GetVisibleDevices: ASCEND_RT_VISIBLE_DEVICES param was not set collected 22 items test_cummax.py [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.140.122 [process_mode_manager.cpp:109][OpenProcess][tid:32458] [ProcessModeManager] enter into open process deviceId[5] rankSize[0] [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.143.771 [process_mode_manager.cpp:379][InitTsdClient][tid:32458] [TsdClient] deviceId[5] begin to init hdc client [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.143.890 [version_verify.cpp:34][SetVersionInfo][tid:32458] VersionVerify: send client version to server [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.143.917 [version_verify.cpp:50][SetVersionInfo][tid:32458] send feature_info:{msg_type:35, features:{check before send aicpu package,}} [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.143.930 [version_verify.cpp:50][SetVersionInfo][tid:32458] send feature_info:{msg_type:37, features:{check before send open qs message,}} [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.186 [version_verify.cpp:66][PeerVersionCheck][tid:32458] VersionVerify: Check client version info, server[1230], client[1230] [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.202 [version_verify.cpp:87][ParseVersionInfo][tid:32458] VersionVerify: pass client version info success [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.211 [hdc_client.cpp:276][CheckHdcConnection][tid:32458] Service[2] create hdc success [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.226 [version_verify.cpp:120][SpecialFeatureCheck][tid:32458] VersionVerify: new type[35], supported [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.268 [process_mode_manager.cpp:748][GetDeviceCheckCode][tid:32458] [TsdClient][deviceId=5] [sessionId=1] wait package info respond [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.364 [process_mode_manager.cpp:379][InitTsdClient][tid:32458] [TsdClient] deviceId[5] begin to init hdc client [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.471 [version_verify.cpp:34][SetVersionInfo][tid:32458] VersionVerify: send client version to server [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.483 [version_verify.cpp:50][SetVersionInfo][tid:32458] send feature_info:{msg_type:35, features:{check before send aicpu package,}} [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.494 [version_verify.cpp:50][SetVersionInfo][tid:32458] send feature_info:{msg_type:37, features:{check before send open qs message,}} [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.731 [version_verify.cpp:66][PeerVersionCheck][tid:32458] VersionVerify: Check client version info, server[1230], client[1230] [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.742 [version_verify.cpp:87][ParseVersionInfo][tid:32458] VersionVerify: pass client version info success [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.751 [hdc_client.cpp:276][CheckHdcConnection][tid:32458] Service[2] create hdc success [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.762 [process_mode_manager.cpp:426][ConstructOpenMsg][tid:32458] [TsdClient] tsd get process sign successfully, procpid[32458] signSize[48] [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.788 [version_verify.cpp:112][SpecialFeatureCheck][tid:32458] VersionVerify: previous type[6], supported [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.144.809 [process_mode_manager.cpp:126][OpenProcess][tid:32458] [ProcessModeManager] deviceId[5] sessionId[1] rankSize[0], wait sub process start respond [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.336.267 [stub_process_mode_nowin.cpp:63][ProcessQueueForMdc][tid:32458] [TsdClient] it is unnecessary of current mode[0] chiptype[1] to grant queue auth to aicpusd [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.336.300 [stub_process_mode_nowin.cpp:101][OpenInHost][tid:32458] enter into OpenInHost deviceid[5] [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.336.311 [stub_process_mode_nowin.cpp:105][OpenInHost][tid:32458] host cpu not support [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.336.319 [process_mode_manager.cpp:156][OpenProcess][tid:32458] [TsdClient][deviceId=5] [sessionId=1] start hccp and computer process success [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:14.338.972 [device.cc:340] 32458 Init: isDoubledie:0, topologytype:0 [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:14.355.228 [npu_driver.cc:5428] 33320 GetDeviceStatus: GetDeviceStatus status=1. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:14.355.262 [atrace_api.c:28](tid:32458) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:14.355.360 [trace_rb_log.c:84](tid:32458) [RUNTIME_ATRACE_DEV5_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:14.355.374 [atrace_api.c:32](tid:32458) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:14.355.389 [client_manager.cpp:157][SetProfilingCallback][tid:32458] [TsdClient] set profiling callback success [TRACE] GE(32458,python3.7):2024-01-11-05:34:14.504.681 [status:INIT] [ge_api.cc:144]32458 GEInitializeImpl:GEInitialize start [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:14.718.094 [msprofiler_impl.cpp:156] >>> (tid:32458) ProfNotifySetDevice called, is open: 1, devId: 5 [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:14.718.213 [platform.cpp:38] >>> (tid:32458) Profiling platform version: 1.0. [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:14.718.228 [ai_drv_dev_api.cpp:384] >>> (tid:32458) Succeeded to DrvGetApiVersion version: 0x72313 [TRACE] GE(32458,python3.7):2024-01-11-05:34:14.771.846 [status:RUNNING] [ge_api.cc:211]32458 GEInitializeImpl:Initializing environment [INFO] GE(32458,python3.7):2024-01-11-05:34:14.771.912 [gelib.cc:98][EVENT]32458 Initialize:[GEPERFTRACE] GE Init Start [INFO] GE(32458,python3.7):2024-01-11-05:34:14.772.221 [gelib.cc:307][EVENT]32458 SystemInitialize:Online infer init GELib success, device id :5 [INFO] DVPP(32458,python3.7):2024-01-11-05:34:15.140.351 [dvpp_engine.cc:41][ENGINE][Initialize:41][tid:32458]dvpp engine do not support [INFO] TUNE(32458,python3.7):2024-01-11-05:34:15.143.746 [cann_kb_pyfunc_mgr.cpp:72][CANNKB][Tid:32458]"CannKbPyfuncMgr: Enter PyObjectInit, reference_ is 0!" [INFO] TUNE(32458,python3.7):2024-01-11-05:34:15.143.785 [handle_manager.cpp:115][CANNKB][Tid:32458]"Start to run init functions to load dynamic python lib!" [INFO] TUNE(32458,python3.7):2024-01-11-05:34:15.143.842 [handle_manager.cpp:407][CANNKB][Tid:32458]"Init functions of loading dynamic python lib end!" [INFO] TUNE(32458,python3.7):2024-01-11-05:34:15.143.852 [cann_kb_pyfunc_mgr.cpp:24][CANNKB][Tid:32458]"CANN_KB_Py has already been initialized." [INFO] TUNE(32458,python3.7):2024-01-11-05:34:15.143.924 [cann_kb_pyfunc_mgr.cpp:117][CANNKB][Tid:32458]"CannKbPyfuncMgr: Run PyObjectInit successfully!" [INFO] HCCL(32458,python3.7):2024-01-11-05:34:27.174.201 [plugin_manager.cc:42][32458]hcom running normal mode. [INFO] DVPP(32458,python3.7):2024-01-11-05:34:27.174.867 [dvpp_engine.cc:92][ENGINE][GetOpsKernelInfoStores:92][tid:32458]dvpp ops kernel info store do not support [INFO] DVPP(32458,python3.7):2024-01-11-05:34:27.175.036 [dvpp_engine.cc:69][ENGINE][GetGraphOptimizerObjs:69][tid:32458]dvpp graph optimizer do not support [INFO] DVPP(32458,python3.7):2024-01-11-05:34:27.705.489 [dvpp_ops_kernel_builder.cc:48][ENGINE][Initialize:48][tid:32458]dvpp ops kernel builder do not support [INFO] GE(32458,python3.7):2024-01-11-05:34:27.713.974 [gelib.cc:169][EVENT]32458 Initialize:[GEPERFTRACE] The time cost of GELib::Initialize is [12942006] micro second. [TRACE] GE(32458,python3.7):2024-01-11-05:34:27.796.735 [status:STOP] [ge_api.cc:255]32458 GEInitializeImpl:GEInitialize finished [TRACE] GE(32458,python3.7):2024-01-11-05:34:27.796.871 [status:INIT] [ge_api.cc:398]32458 Session:Start to construct session. [TRACE] GE(32458,python3.7):2024-01-11-05:34:27.796.889 [status:RUNNING] [ge_api.cc:408]32458 Session:Creating session [INFO] GE(32458,python3.7):2024-01-11-05:34:27.797.269 [graph_var_manager.cc:1445][EVENT]32458 SetMemoryMallocSize:Total memory size is 34359738368 [INFO] GE(32458,python3.7):2024-01-11-05:34:27.797.289 [graph_var_manager.cc:1424][EVENT]32458 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:27.797.580 [msprofiler_impl.cpp:156] >>> (tid:32458) ProfNotifySetDevice called, is open: 1, devId: 5 [TRACE] GE(32458,python3.7):2024-01-11-05:34:27.798.410 [status:RUNNING] [ge_api.cc:411]32458 Session:Session id is 0 [TRACE] GE(32458,python3.7):2024-01-11-05:34:27.798.430 [status:STOP] [ge_api.cc:420]32458 Session:Session Constructor finished [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:27.808.536 [platform.cpp:38] >>> (tid:32458) Profiling platform version: 1.0. [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:27.808.563 [ai_drv_dev_api.cpp:384] >>> (tid:32458) Succeeded to DrvGetApiVersion version: 0x72313 [TRACE] GE(32458,python3.7):2024-01-11-05:34:27.808.743 [status:INIT] [ge_api.cc:144]32458 GEInitializeImpl:GEInitialize start TotalTime = 0.464454, [20] [parse]: 0.0132317 [symbol_resolve]: 0.0288836, [1] [Cycle 1]: 0.028807, [1] [resolve]: 0.0287847 [combine_like_graphs]: 1.16e-06 [graph_reusing]: 3.15e-06 [meta_unpack_prepare]: 0.00015891 [pre_cconv]: 4.92e-06 [abstract_specialize]: 0.00522014 [pack_expand]: 1.538e-05 [auto_monad]: 0.00011322 [inline]: 1.72e-06 [pre_auto_parallel]: 1.427e-05 [pipeline_split]: 2.70001e-06 [optimize]: 0.40877, [35] [py_interpret_to_execute]: 3.34e-06 [rewriter_before_opt_a]: 0.00019811 [opt_a]: 0.406853, [4] [Cycle 1]: 0.339798, [30] [expand_dump_flag]: 4.01e-06 [switch_simplify]: 2.446e-05 [a_1]: 0.00040485 [recompute_prepare]: 9.32001e-06 [updatestate_depend_eliminate]: 1.098e-05 [updatestate_assign_eliminate]: 7.07e-06 [updatestate_loads_eliminate]: 6.34e-06 [parameter_eliminate]: 4.72e-06 [a_2]: 7.961e-05 [accelerated_algorithm]: 5.4e-06 [pynative_shard]: 1.96e-06 [auto_parallel]: 3.51e-06 [parallel]: 1.889e-05 [merge_comm]: 1.2e-05 [allreduce_fusion]: 1.97e-06 [virtual_dataset]: 5.37001e-06 [get_grad_eliminate_]: 4.34e-06 [virtual_output]: 4.07e-06 [merge_forward]: 8.77e-06 [cell_reuse_recompute_pass]: 7.40001e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.213e-05 [meta_fg_expand]: 0.259465, [1] [Cycle 1]: 0.00548333, [1] [resolve]: 0.00545095 [after_resolve]: 5.264e-05 [a_after_grad]: 0.00012093 [renormalize]: 0.078467 [real_op_eliminate]: 4.385e-05 [auto_monad_grad]: 7.219e-05 [auto_monad_eliminator]: 8.13e-05 [cse]: 0.00036224 [a_3]: 0.00031645 [Cycle 2]: 0.0544592, [30] [expand_dump_flag]: 5.19e-06 [switch_simplify]: 0.00010309 [a_1]: 0.00077709 [recompute_prepare]: 1.377e-05 [updatestate_depend_eliminate]: 1.706e-05 [updatestate_assign_eliminate]: 1.354e-05 [updatestate_loads_eliminate]: 1.339e-05 [parameter_eliminate]: 4.71e-06 [a_2]: 0.00019907 [accelerated_algorithm]: 2.22e-05 [pynative_shard]: 2.25e-06 [auto_parallel]: 8.37e-06 [parallel]: 7.68001e-06 [merge_comm]: 4.22e-06 [allreduce_fusion]: 1.79e-06 [virtual_dataset]: 1.037e-05 [get_grad_eliminate_]: 9.51e-06 [virtual_output]: 9.46e-06 [merge_forward]: 1.484e-05 [cell_reuse_recompute_pass]: 1.03001e-06 [cell_reuse_handle_not_recompute_node_pass]: 2.428e-05 [meta_fg_expand]: 0.0133905, [5] [Cycle 1]: 0.00032359, [1] [resolve]: 0.00030588 [Cycle 1]: 0.00035078, [1] [resolve]: 0.0003324 [Cycle 1]: 0.00168399, [1] [resolve]: 0.00166572 [Cycle 1]: 0.00031687, [1] [resolve]: 0.00029936 [Cycle 1]: 0.00036793, [1] [resolve]: 0.00034993 [after_resolve]: 7.037e-05 [a_after_grad]: 0.00015971 [renormalize]: 0.0382861 [real_op_eliminate]: 5.28e-05 [auto_monad_grad]: 0.00019534 [auto_monad_eliminator]: 0.00010819 [cse]: 0.00030882 [a_3]: 0.00044098 [Cycle 3]: 0.00464809, [30] [expand_dump_flag]: 4.09e-06 [switch_simplify]: 0.00011335 [a_1]: 0.00119989 [recompute_prepare]: 1.753e-05 [updatestate_depend_eliminate]: 2.674e-05 [updatestate_assign_eliminate]: 1.822e-05 [updatestate_loads_eliminate]: 1.792e-05 [parameter_eliminate]: 4.06001e-06 [a_2]: 0.00027481 [accelerated_algorithm]: 2.471e-05 [pynative_shard]: 1.32e-06 [auto_parallel]: 4.56e-06 [parallel]: 4.58e-06 [merge_comm]: 3.49e-06 [allreduce_fusion]: 2.44e-06 [virtual_dataset]: 1.364e-05 [get_grad_eliminate_]: 1.31e-05 [virtual_output]: 1.228e-05 [merge_forward]: 1.944e-05 [cell_reuse_recompute_pass]: 4.99997e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.277e-05 [meta_fg_expand]: 4.676e-05 [after_resolve]: 1.641e-05 [a_after_grad]: 2.02e-05 [renormalize]: 0.00221805 [real_op_eliminate]: 1.918e-05 [auto_monad_grad]: 6e-06 [auto_monad_eliminator]: 3.413e-05 [cse]: 0.00019961 [a_3]: 0.000121 [Cycle 4]: 0.00116778, [30] [expand_dump_flag]: 1.23e-06 [switch_simplify]: 1.442e-05 [a_1]: 0.00028212 [recompute_prepare]: 1.488e-05 [updatestate_depend_eliminate]: 1.961e-05 [updatestate_assign_eliminate]: 1.671e-05 [updatestate_loads_eliminate]: 1.597e-05 [parameter_eliminate]: 2.23e-06 [a_2]: 0.00027224 [accelerated_algorithm]: 2.452e-05 [pynative_shard]: 1.4e-06 [auto_parallel]: 3.88e-06 [parallel]: 3.77e-06 [merge_comm]: 3.11e-06 [allreduce_fusion]: 1.97e-06 [virtual_dataset]: 1.402e-05 [get_grad_eliminate_]: 1.381e-05 [virtual_output]: 1.245e-05 [merge_forward]: 1.722e-05 [cell_reuse_recompute_pass]: 4.70005e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.259e-05 [meta_fg_expand]: 1.314e-05 [after_resolve]: 1.551e-05 [a_after_grad]: 2.026e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.326e-05 [auto_monad_grad]: 2.18e-06 [auto_monad_eliminator]: 2.976e-05 [cse]: 8.025e-05 [a_3]: 0.00011182 [py_interpret_to_execute_after_opt_a]: 4.17e-06 [slice_cell_reuse_recomputed_activation]: 2.59e-06 [rewriter_after_opt_a]: 0.000102 [convert_after_rewriter]: 2.426e-05 [order_py_execute_after_rewriter]: 1.73e-05 [opt_b]: 0.00109487, [2] [Cycle 1]: 0.00092794, [7] [b_1]: 0.00083589 [b_2]: 5.9e-06 [updatestate_depend_eliminate]: 5.79e-06 [updatestate_assign_eliminate]: 4.09e-06 [updatestate_loads_eliminate]: 4.09e-06 [renormalize]: 4.69998e-07 [cse]: 3.615e-05 [Cycle 2]: 0.00015704, [7] [b_1]: 9.659e-05 [b_2]: 4.22e-06 [updatestate_depend_eliminate]: 4.62e-06 [updatestate_assign_eliminate]: 3.82e-06 [updatestate_loads_eliminate]: 3.55e-06 [renormalize]: 5.99975e-08 [cse]: 1.787e-05 [cconv]: 2.193e-05 [opt_after_cconv]: 6.965e-05, [1] [Cycle 1]: 6.519e-05, [7] [c_1]: 8.76e-06 [parameter_eliminate]: 2.23e-06 [updatestate_depend_eliminate]: 4.43e-06 [updatestate_assign_eliminate]: 3.74e-06 [updatestate_loads_eliminate]: 3.49e-06 [cse]: 1.604e-05 [renormalize]: 3.70004e-07 [remove_dup_value]: 1.754e-05 [tuple_transform]: 6.62e-05, [1] [Cycle 1]: 6.238e-05, [3] [d_1]: 4.06e-05 [d_2]: 9.1e-06 [renormalize]: 1.69995e-07 [add_cache_embedding]: 1.301e-05 [add_recomputation]: 6.631e-05 [cse_after_recomputation]: 2.53e-05, [1] [Cycle 1]: 2.113e-05, [1] [cse]: 1.662e-05 [environ_conv]: 2.603e-05 [label_micro_interleaved_index]: 2.45e-06 [label_fine_grained_interleaved_index]: 2.43e-06 [assign_add_opt]: 2.73e-06 [slice_recompute_activation]: 1.94e-06 [micro_interleaved_order_control]: 1.63e-06 [full_micro_interleaved_order_control]: 1.89e-06 [comp_comm_scheduling]: 2.35e-06 [reorder_send_recv_between_fp_bp]: 2.37e-06 [comm_op_add_attrs]: 9.70002e-07 [add_comm_op_reuse_tag]: 8.40002e-07 [overlap_opt_shard_in_pipeline]: 9.40003e-07 [grouped_pairwise_exchange_alltoall]: 1.29e-06 [overlap_recompute_and_grad_model_parallel]: 1.87e-06 [overlap_grad_matmul_and_grad_allreduce]: 7.10002e-07 [split_matmul_comm_elemetwise]: 2.56e-06 [split_layernorm_comm]: 1.73999e-06 [process_send_recv_for_ge]: 2.14e-06 [handle_group_info]: 1.01e-06 [auto_monad_reorder]: 2.885e-05 [get_jit_bprop_graph]: 4.20005e-07 [eliminate_special_op_node]: 0.000542 [validate]: 5.616e-05 [distribtued_split]: 1.16e-06 [task_emit]: 0.00718868 [execute]: 7.38e-06 Sums parse : 0.013232s : 6.89% symbol_resolve.resolve : 0.028785s : 14.99% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000159s : 0.08% pre_cconv : 0.000005s : 0.00% abstract_specialize : 0.005220s : 2.72% pack_expand : 0.000015s : 0.01% auto_monad : 0.000113s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000014s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000003s : 0.00% optimize.rewriter_before_opt_a : 0.000198s : 0.10% optimize.opt_a.expand_dump_flag : 0.000015s : 0.01% optimize.opt_a.switch_simplify : 0.000255s : 0.13% optimize.opt_a.a_1 : 0.002664s : 1.39% optimize.opt_a.recompute_prepare : 0.000056s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000074s : 0.04% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.03% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.03% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000826s : 0.43% optimize.opt_a.accelerated_algorithm : 0.000077s : 0.04% optimize.opt_a.pynative_shard : 0.000007s : 0.00% optimize.opt_a.auto_parallel : 0.000020s : 0.01% optimize.opt_a.parallel : 0.000035s : 0.02% optimize.opt_a.merge_comm : 0.000023s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.00% optimize.opt_a.virtual_dataset : 0.000043s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.02% optimize.opt_a.virtual_output : 0.000038s : 0.02% optimize.opt_a.merge_forward : 0.000060s : 0.03% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.05% optimize.opt_a.meta_fg_expand : 0.000060s : 0.03% optimize.opt_a.meta_fg_expand.resolve : 0.008404s : 4.38% optimize.opt_a.after_resolve : 0.000155s : 0.08% optimize.opt_a.a_after_grad : 0.000321s : 0.17% optimize.opt_a.renormalize : 0.118971s : 61.96% optimize.opt_a.real_op_eliminate : 0.000129s : 0.07% optimize.opt_a.auto_monad_grad : 0.000276s : 0.14% optimize.opt_a.auto_monad_eliminator : 0.000253s : 0.13% optimize.opt_a.cse : 0.000951s : 0.50% optimize.opt_a.a_3 : 0.000990s : 0.52% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000102s : 0.05% optimize.convert_after_rewriter : 0.000024s : 0.01% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000932s : 0.49% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000054s : 0.03% optimize.cconv : 0.000022s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000041s : 0.02% optimize.tuple_transform.d_2 : 0.000009s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000066s : 0.03% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000026s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000003s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000029s : 0.02% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000542s : 0.28% validate : 0.000056s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.007189s : 3.74% execute : 0.000007s : 0.00% Time group info: ------[substitution.] 0.037422 880 0.01% : 0.000004s : 5: substitution.float_depend_g_call 0.06% : 0.000024s : 49: substitution.float_tuple_getitem_switch 95.18% : 0.035616s : 59: substitution.getattr_setattr_resolve 0.02% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 2.81% : 0.001051s : 97: substitution.inline 0.02% : 0.000008s : 23: substitution.less_batch_normalization 0.13% : 0.000048s : 23: substitution.meta_unpack_prepare 0.08% : 0.000032s : 40: substitution.minmaximum_grad 0.04% : 0.000014s : 5: substitution.partial_eliminate 0.00% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.02% : 0.000009s : 81: substitution.remove_not_recompute_node 0.24% : 0.000092s : 63: substitution.replace_applicator 0.04% : 0.000014s : 36: substitution.replace_old_param 0.01% : 0.000003s : 2: substitution.reset_defer_inline 0.02% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.02% : 0.000008s : 5: substitution.specialize_transform 0.03% : 0.000012s : 10: substitution.switch_simplify 0.04% : 0.000015s : 4: substitution.transpose_eliminate 0.30% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.13% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.18% : 0.000067s : 60: substitution.tuple_list_get_item_depend_reorder 0.42% : 0.000155s : 112: substitution.tuple_list_get_item_eliminator 0.18% : 0.000068s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.118955 6 94.14% : 0.111986s : 3: renormalize.infer 5.86% : 0.006969s : 3: renormalize.specialize ------[replace.] 0.001242 141 54.43% : 0.000676s : 55: replace.getattr_setattr_resolve 26.84% : 0.000333s : 56: replace.inline 3.48% : 0.000043s : 2: replace.meta_unpack_prepare 7.31% : 0.000091s : 10: replace.switch_simplify 1.43% : 0.000018s : 4: replace.transpose_eliminate 6.51% : 0.000081s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.036373 141 97.35% : 0.035408s : 55: match.getattr_setattr_resolve 2.40% : 0.000873s : 56: match.inline 0.10% : 0.000037s : 2: match.meta_unpack_prepare 0.03% : 0.000012s : 10: match.switch_simplify 0.04% : 0.000015s : 4: match.transpose_eliminate 0.08% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.007074 119 68.23% : 0.004826s : 53: func_graph_cloner_run.FuncGraphClonerGraph 31.77% : 0.002247s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.043561 259 4.28% : 0.001865s : 104: opt.transform.opt_a 2.07% : 0.000901s : 92: opt.transform.opt_b 84.60% : 0.036851s : 14: opt.transform.opt_resolve 0.29% : 0.000126s : 1: opt.transforms.meta_unpack_prepare 8.59% : 0.003740s : 40: opt.transforms.opt_a 0.02% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.02% : 0.000008s : 2: opt.transforms.opt_b 0.11% : 0.000048s : 2: opt.transforms.opt_trans_graph 0.03% : 0.000014s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:28.386.233 [scalable_config.cc:55][EVENT]36564 ScalableConfig:device total max size: 34359738368, page_mem_size_total_thresold: 32641751449, uncacheable_size_threshold: 17179869184 [INFO] GE(32458,python3.7):2024-01-11-05:34:28.466.881 [graph_var_manager.cc:1424][EVENT]36564 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:28.466.976 [graph_manager.cc:1248][EVENT]36564 PreRun:PreRun start: graph node size 3, session id 1, graph id 0, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.467.877 [atrace_api.c:28](tid:36564) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.467.960 [trace_rb_log.c:84](tid:36564) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.467.976 [atrace_api.c:32](tid:36564) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:28.468.010 [client_manager.cpp:157][SetProfilingCallback][tid:36564] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:28.468.964 [parallel_partitioner.cc:165][EVENT]36564 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.469.009 [parallel_partitioner.cc:178][EVENT]36564 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.469.067 [graph_prepare.cc:1378][EVENT]36564 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.469.811 [graph_manager.cc:1050][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [765] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.469.844 [graph_manager.cc:1052][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.469.991 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.022 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.095 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.109 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.199 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.212 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.234 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [12] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.344 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.470.377 [graph_manager.cc:1054][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [516] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.478.302 [graph_manager.cc:1055][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [7910] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.410 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.434 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.445 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.455 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [338] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.464 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [16] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.473 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.481 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [18] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.490 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [23] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.479.498 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.480.850 [graph_manager.cc:1056][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2514] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.480.911 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.480.928 [graph_prepare.cc:1982][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.366 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.388 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.398 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.407 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [256] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.416 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.425 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.433 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.441 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.476 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.501 [graph_prepare.cc:1983][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [559] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.525 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.536 [graph_prepare.cc:1984][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.550 [graph_prepare.cc:1985][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.577 [graph_prepare.cc:1986][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.589 [graph_prepare.cc:1987][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.604 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.617 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.630 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.712 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.724 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.733 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrintOpPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.742 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.750 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.758 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.767 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.775 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.783 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of StopGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.791 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.799 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.808 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SnapshotPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.822 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.831 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.839 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.847 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.869 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.881 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.913 [graph_prepare.cc:1988][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [314] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.481.926 [graph_manager.cc:1065][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [1048] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.494.882 [graph_manager.cc:1077][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12936] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.494.948 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.495.003 [graph_manager.cc:1080][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [89] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.801 [graph_manager.cc:1081][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2783] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.838 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.853 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.865 [graph_manager.cc:1082][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.895 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.911 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.925 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.497.997 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.014 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.048 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [22] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.063 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.116 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.135 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.153 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.201 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [38] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.219 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.232 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.242 [graph_manager.cc:2700][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [351] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.353 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.367 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.377 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.386 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.394 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.403 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CastRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.411 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.420 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.428 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.436 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.444 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.453 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.461 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.469 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.477 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.495 [graph_manager.cc:2741][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [235] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.505 [graph_manager.cc:2752][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.527 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.540 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.556 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.572 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.583 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.595 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.614 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.628 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.641 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.651 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.664 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.675 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.693 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.705 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.714 [graph_manager.cc:2810][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [191] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.742 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.755 [graph_manager.cc:2821][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [32] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.783 [graph_manager.cc:1087][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [900] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.915 [graph_manager.cc:1088][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [120] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.955 [graph_manager.cc:1089][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.980 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.498.996 [graph_manager.cc:1097][EVENT]36564 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.017 [graph_manager.cc:3325][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.239 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.255 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.265 [engine_place.cc:144][EVENT]36564 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [116] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.345 [graph_manager.cc:3351][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [314] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.362 [graph_manager.cc:3364][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.443 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [22] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.459 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.611 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [142] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.653 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.705 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [39] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.739 [graph_manager.cc:3405][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [364] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.499.758 [graph_manager.cc:3412][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.420 [graph_manager.cc:3422][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1646] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.450 [graph_manager.cc:3428][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.576 [graph_manager.cc:3467][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [105] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.593 [graph_manager.cc:3377][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [2219] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.609 [graph_manager.cc:1106][EVENT]36564 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2600] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.622 [graph_manager.cc:1115][EVENT]36564 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.652 [graph_manager.cc:1130][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.685 [graph_manager.cc:1131][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.713 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.730 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.740 [graph_manager.cc:2837][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [39] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.813 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.826 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.836 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.844 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.853 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.861 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.871 [graph_manager.cc:2864][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [115] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.883 [graph_manager.cc:2872][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.904 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.919 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.935 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.949 [compile_nodes_pass.cc:88][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.959 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.501.969 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.057 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [80] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.087 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.105 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.118 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.131 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.140 [graph_manager.cc:2927][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [238] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.158 [graph_manager.cc:2937][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.174 [graph_manager.cc:2943][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.502.186 [graph_manager.cc:2950][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.309 [graph_manager.cc:2958][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [45] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.352 [graph_manager.cc:1132][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [10653] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.431 [graph_manager.cc:1135][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [64] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.475 [graph_manager.cc:2975][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [25] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.517 [graph_manager.cc:2981][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.534 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.544 [graph_manager.cc:2986][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.554 [graph_manager.cc:1136][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [104] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.672 [graph_manager.cc:3555][EVENT]36564 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [83] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.766 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.782 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.907 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [115] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.938 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.512.990 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [31] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.513.011 [graph_builder.cc:865][EVENT]36564 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [277] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:28.513.614 [logger.cc:1071] 36564 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.513.654 [task_generator.cc:804][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [173] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.513.727 [task_generator.cc:805][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [60] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.226 [task_generator.cc:814][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [484] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.240 [task_generator.cc:954][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [760] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.308 [task_generator.cc:967][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [38] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:28.514.327 [logger.cc:1084] 36564 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.492 [graph_manager.cc:1152][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1912] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.509 [graph_manager.cc:1164][EVENT]36564 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.545 [graph_manager.cc:1271][EVENT]36564 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [45705] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.514.555 [graph_manager.cc:1272][EVENT]36564 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.514.868 [atrace_api.c:93](tid:36564) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.514.896 [atrace_api.c:95](tid:36564) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:28.519.426 [graph_converter.cc:838][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1245] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.519.621 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [153] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.015 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [371] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.099 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.117 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [80] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.422 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [291] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.532 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [90] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.569 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.725 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [141] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.797 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [54] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.817 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [75] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.846 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.872 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.898 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.520.960 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.018 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [48] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.029 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [59] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.054 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.079 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.100 [graph_converter.cc:849][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1637] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.310 [graph_converter.cc:853][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [200] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.521.968 [graph_converter.cc:857][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [642] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.522.106 [graph_converter.cc:862][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [117] micro second. . TotalTime = 0.145095, [20] [parse]: 0.00145855 [symbol_resolve]: 0.0122904, [1] [Cycle 1]: 0.0122215, [1] [resolve]: 0.0122019 [combine_like_graphs]: 8.90002e-07 [graph_reusing]: 3.54e-06 [meta_unpack_prepare]: 0.00016016 [pre_cconv]: 6.69999e-07 [abstract_specialize]: 0.0041851 [pack_expand]: 1.509e-05 [auto_monad]: 8.78e-05 [inline]: 1.58e-06 [pre_auto_parallel]: 1.009e-05 [pipeline_split]: 2.99e-06 [optimize]: 0.122626, [35] [py_interpret_to_execute]: 4.32001e-06 [rewriter_before_opt_a]: 0.00018992 [opt_a]: 0.120708, [4] [Cycle 1]: 0.0577684, [30] [expand_dump_flag]: 3.8e-06 [switch_simplify]: 2.514e-05 [a_1]: 0.00072649 [recompute_prepare]: 8.16e-06 [updatestate_depend_eliminate]: 1.022e-05 [updatestate_assign_eliminate]: 7.83e-06 [updatestate_loads_eliminate]: 6.87e-06 [parameter_eliminate]: 5.26e-06 [a_2]: 7.546e-05 [accelerated_algorithm]: 5.38e-06 [pynative_shard]: 1.66e-06 [auto_parallel]: 3.17e-06 [parallel]: 9.23e-06 [merge_comm]: 3.68e-06 [allreduce_fusion]: 1.91999e-06 [virtual_dataset]: 5.6e-06 [get_grad_eliminate_]: 4.95e-06 [virtual_output]: 4.44e-06 [merge_forward]: 8.12e-06 [cell_reuse_recompute_pass]: 7.7e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.152e-05 [meta_fg_expand]: 0.0068932, [1] [Cycle 1]: 0.00314085, [1] [resolve]: 0.00312188 [after_resolve]: 4.783e-05 [a_after_grad]: 0.00012944 [renormalize]: 0.0487469 [real_op_eliminate]: 4.328e-05 [auto_monad_grad]: 6.564e-05 [auto_monad_eliminator]: 7.992e-05 [cse]: 0.00034255 [a_3]: 0.00030787 [Cycle 2]: 0.0530977, [30] [expand_dump_flag]: 3.91999e-06 [switch_simplify]: 0.00012931 [a_1]: 0.001613 [recompute_prepare]: 1.25e-05 [updatestate_depend_eliminate]: 1.677e-05 [updatestate_assign_eliminate]: 1.274e-05 [updatestate_loads_eliminate]: 1.246e-05 [parameter_eliminate]: 4.35e-06 [a_2]: 0.00019185 [accelerated_algorithm]: 1.896e-05 [pynative_shard]: 1.34e-06 [auto_parallel]: 4.82e-06 [parallel]: 4.79e-06 [merge_comm]: 2.46e-06 [allreduce_fusion]: 1.46e-06 [virtual_dataset]: 1.083e-05 [get_grad_eliminate_]: 1.033e-05 [virtual_output]: 1.031e-05 [merge_forward]: 1.377e-05 [cell_reuse_recompute_pass]: 7.7e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.33e-05 [meta_fg_expand]: 0.0116789, [5] [Cycle 1]: 0.00031566, [1] [resolve]: 0.00029854 [Cycle 1]: 0.00032422, [1] [resolve]: 0.0003062 [Cycle 1]: 0.00165074, [1] [resolve]: 0.00163233 [Cycle 1]: 0.00031008, [1] [resolve]: 0.00029188 [Cycle 1]: 0.00030824, [1] [resolve]: 0.00029031 [after_resolve]: 7.489e-05 [a_after_grad]: 0.00019287 [renormalize]: 0.0377077 [real_op_eliminate]: 6.038e-05 [auto_monad_grad]: 0.00020259 [auto_monad_eliminator]: 0.00010753 [cse]: 0.0003151 [a_3]: 0.00046861 [Cycle 3]: 0.00585433, [30] [expand_dump_flag]: 4.49e-06 [switch_simplify]: 0.00015678 [a_1]: 0.00230231 [recompute_prepare]: 1.639e-05 [updatestate_depend_eliminate]: 3.178e-05 [updatestate_assign_eliminate]: 1.873e-05 [updatestate_loads_eliminate]: 1.851e-05 [parameter_eliminate]: 4.58e-06 [a_2]: 0.00027419 [accelerated_algorithm]: 2.516e-05 [pynative_shard]: 1.54e-06 [auto_parallel]: 5.47e-06 [parallel]: 4.68999e-06 [merge_comm]: 4e-06 [allreduce_fusion]: 2.44001e-06 [virtual_dataset]: 1.496e-05 [get_grad_eliminate_]: 1.432e-05 [virtual_output]: 1.381e-05 [merge_forward]: 2.009e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.258e-05 [meta_fg_expand]: 4.815e-05 [after_resolve]: 1.815e-05 [a_after_grad]: 3.205e-05 [renormalize]: 0.00224395 [real_op_eliminate]: 2.191e-05 [auto_monad_grad]: 5.63001e-06 [auto_monad_eliminator]: 3.435e-05 [cse]: 0.0002021 [a_3]: 0.0001214 [Cycle 4]: 0.00165193, [30] [expand_dump_flag]: 1.36e-06 [switch_simplify]: 1.528e-05 [a_1]: 0.00074508 [recompute_prepare]: 1.53e-05 [updatestate_depend_eliminate]: 1.96e-05 [updatestate_assign_eliminate]: 1.689e-05 [updatestate_loads_eliminate]: 1.654e-05 [parameter_eliminate]: 2.26e-06 [a_2]: 0.00027371 [accelerated_algorithm]: 2.508e-05 [pynative_shard]: 1.49e-06 [auto_parallel]: 3.97e-06 [parallel]: 3.58e-06 [merge_comm]: 3.3e-06 [allreduce_fusion]: 1.92e-06 [virtual_dataset]: 1.479e-05 [get_grad_eliminate_]: 1.474e-05 [virtual_output]: 1.412e-05 [merge_forward]: 1.735e-05 [cell_reuse_recompute_pass]: 3.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.225e-05 [meta_fg_expand]: 1.303e-05 [after_resolve]: 1.643e-05 [a_after_grad]: 3.14e-05 [renormalize]: 6.00048e-08 [real_op_eliminate]: 1.444e-05 [auto_monad_grad]: 2.1e-06 [auto_monad_eliminator]: 2.981e-05 [cse]: 8.113e-05 [a_3]: 0.00011165 [py_interpret_to_execute_after_opt_a]: 4.14e-06 [slice_cell_reuse_recomputed_activation]: 2.68e-06 [rewriter_after_opt_a]: 9.891e-05 [convert_after_rewriter]: 2.305e-05 [order_py_execute_after_rewriter]: 1.733e-05 [opt_b]: 0.0011255, [2] [Cycle 1]: 0.00095592, [7] [b_1]: 0.00086118 [b_2]: 5.04999e-06 [updatestate_depend_eliminate]: 6.2e-06 [updatestate_assign_eliminate]: 4.12e-06 [updatestate_loads_eliminate]: 3.99e-06 [renormalize]: 5.4e-07 [cse]: 3.727e-05 [Cycle 2]: 0.00015982, [7] [b_1]: 9.801e-05 [b_2]: 3.98e-06 [updatestate_depend_eliminate]: 5.12e-06 [updatestate_assign_eliminate]: 3.88e-06 [updatestate_loads_eliminate]: 3.72e-06 [renormalize]: 7.0002e-08 [cse]: 1.765e-05 [cconv]: 2.176e-05 [opt_after_cconv]: 8.708e-05, [1] [Cycle 1]: 8.262e-05, [7] [c_1]: 2.494e-05 [parameter_eliminate]: 2.42001e-06 [updatestate_depend_eliminate]: 4.39e-06 [updatestate_assign_eliminate]: 4.19e-06 [updatestate_loads_eliminate]: 3.54e-06 [cse]: 1.588e-05 [renormalize]: 4.1e-07 [remove_dup_value]: 1.786e-05 [tuple_transform]: 8.277e-05, [1] [Cycle 1]: 7.908e-05, [3] [d_1]: 5.781e-05 [d_2]: 9.18e-06 [renormalize]: 2.19996e-07 [add_cache_embedding]: 1.247e-05 [add_recomputation]: 5.881e-05 [cse_after_recomputation]: 2.598e-05, [1] [Cycle 1]: 2.181e-05, [1] [cse]: 1.728e-05 [environ_conv]: 9.92999e-06 [label_micro_interleaved_index]: 2.48e-06 [label_fine_grained_interleaved_index]: 2.29e-06 [assign_add_opt]: 1.67e-06 [slice_recompute_activation]: 2.27e-06 [micro_interleaved_order_control]: 1.75e-06 [full_micro_interleaved_order_control]: 1.63e-06 [comp_comm_scheduling]: 2.12e-06 [reorder_send_recv_between_fp_bp]: 2.07e-06 [comm_op_add_attrs]: 1.11001e-06 [add_comm_op_reuse_tag]: 8.2e-07 [overlap_opt_shard_in_pipeline]: 1.32999e-06 [grouped_pairwise_exchange_alltoall]: 1.18e-06 [overlap_recompute_and_grad_model_parallel]: 1.65e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.99998e-07 [split_matmul_comm_elemetwise]: 2.06e-06 [split_layernorm_comm]: 1.89e-06 [process_send_recv_for_ge]: 7.7e-07 [handle_group_info]: 9.5e-07 [auto_monad_reorder]: 2.235e-05 [get_jit_bprop_graph]: 3.12e-06 [eliminate_special_op_node]: 0.00052241 [validate]: 3.81e-05 [distribtued_split]: 1.08e-06 [task_emit]: 0.00346792 [execute]: 5.94e-06 Sums parse : 0.001459s : 1.13% symbol_resolve.resolve : 0.012202s : 9.46% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000004s : 0.00% meta_unpack_prepare : 0.000160s : 0.12% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004185s : 3.25% pack_expand : 0.000015s : 0.01% auto_monad : 0.000088s : 0.07% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000190s : 0.15% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000327s : 0.25% optimize.opt_a.a_1 : 0.005387s : 4.18% optimize.opt_a.recompute_prepare : 0.000052s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000078s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000815s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000022s : 0.02% optimize.opt_a.merge_comm : 0.000013s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000046s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000100s : 0.08% optimize.opt_a.meta_fg_expand : 0.000061s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.005941s : 4.61% optimize.opt_a.after_resolve : 0.000157s : 0.12% optimize.opt_a.a_after_grad : 0.000386s : 0.30% optimize.opt_a.renormalize : 0.088699s : 68.78% optimize.opt_a.real_op_eliminate : 0.000140s : 0.11% optimize.opt_a.auto_monad_grad : 0.000276s : 0.21% optimize.opt_a.auto_monad_eliminator : 0.000252s : 0.20% optimize.opt_a.cse : 0.000941s : 0.73% optimize.opt_a.a_3 : 0.001010s : 0.78% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000099s : 0.08% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000959s : 0.74% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000055s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000025s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000058s : 0.04% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000012s : 0.01% optimize.add_recomputation : 0.000059s : 0.05% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000022s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000522s : 0.41% validate : 0.000038s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003468s : 2.69% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018408 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.26% : 0.016616s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.47% : 0.001006s : 103: substitution.inline 0.05% : 0.000009s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.50% : 0.000092s : 69: substitution.replace_applicator 0.06% : 0.000011s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000012s : 10: substitution.switch_simplify 0.07% : 0.000012s : 4: substitution.transpose_eliminate 0.79% : 0.000145s : 70: substitution.tuple_list_convert_item_index_to_positive 0.31% : 0.000057s : 70: substitution.tuple_list_get_item_const_eliminator 0.41% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.87% : 0.000160s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000076s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.088682 6 92.44% : 0.081973s : 3: renormalize.infer 7.56% : 0.006709s : 3: renormalize.specialize ------[replace.] 0.001207 141 54.17% : 0.000654s : 55: replace.getattr_setattr_resolve 26.21% : 0.000316s : 56: replace.inline 3.66% : 0.000044s : 2: replace.meta_unpack_prepare 7.47% : 0.000090s : 10: replace.switch_simplify 1.75% : 0.000021s : 4: replace.transpose_eliminate 6.75% : 0.000081s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017391 141 94.80% : 0.016487s : 55: match.getattr_setattr_resolve 4.81% : 0.000836s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.07% : 0.000012s : 10: match.switch_simplify 0.07% : 0.000012s : 4: match.transpose_eliminate 0.15% : 0.000026s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006807 119 69.27% : 0.004715s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.73% : 0.002091s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027371 589 0.50% : 0.000138s : 2: opt.transform.meta_unpack_prepare 30.64% : 0.008388s : 461: opt.transform.opt_a 0.08% : 0.000021s : 7: opt.transform.opt_after_cconv 3.41% : 0.000933s : 94: opt.transform.opt_b 65.08% : 0.017814s : 14: opt.transform.opt_resolve 0.23% : 0.000062s : 8: opt.transform.opt_trans_graph 0.06% : 0.000016s : 3: opt.transform.special_op_eliminate . TotalTime = 0.1437, [20] [parse]: 0.0012949 [symbol_resolve]: 0.012308, [1] [Cycle 1]: 0.0122474, [1] [resolve]: 0.0122309 [combine_like_graphs]: 8.50006e-07 [graph_reusing]: 2.99e-06 [meta_unpack_prepare]: 0.00012425 [pre_cconv]: 4.60001e-07 [abstract_specialize]: 0.00395819 [pack_expand]: 1.629e-05 [auto_monad]: 7.345e-05 [inline]: 1.53e-06 [pre_auto_parallel]: 7.22001e-06 [pipeline_split]: 2.1e-06 [optimize]: 0.119721, [35] [py_interpret_to_execute]: 4.03e-06 [rewriter_before_opt_a]: 0.00018705 [opt_a]: 0.117915, [4] [Cycle 1]: 0.0579959, [30] [expand_dump_flag]: 3.01e-06 [switch_simplify]: 2.452e-05 [a_1]: 0.00037862 [recompute_prepare]: 8.48e-06 [updatestate_depend_eliminate]: 9.83e-06 [updatestate_assign_eliminate]: 6.38999e-06 [updatestate_loads_eliminate]: 5.71e-06 [parameter_eliminate]: 4.11e-06 [a_2]: 7.834e-05 [accelerated_algorithm]: 5.51e-06 [pynative_shard]: 9.59997e-07 [auto_parallel]: 3.24e-06 [parallel]: 5.72e-06 [merge_comm]: 2.61e-06 [allreduce_fusion]: 1.8e-06 [virtual_dataset]: 5.07e-06 [get_grad_eliminate_]: 4.53e-06 [virtual_output]: 4.22e-06 [merge_forward]: 7.27e-06 [cell_reuse_recompute_pass]: 4.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.132e-05 [meta_fg_expand]: 0.00690439, [1] [Cycle 1]: 0.0031822, [1] [resolve]: 0.00316342 [after_resolve]: 4.495e-05 [a_after_grad]: 0.00010647 [renormalize]: 0.0493707 [real_op_eliminate]: 4.083e-05 [auto_monad_grad]: 6.608e-05 [auto_monad_eliminator]: 7.759e-05 [cse]: 0.00032731 [a_3]: 0.00030567 [Cycle 2]: 0.0516922, [30] [expand_dump_flag]: 3.58e-06 [switch_simplify]: 0.00010088 [a_1]: 0.00076003 [recompute_prepare]: 1.35e-05 [updatestate_depend_eliminate]: 1.62e-05 [updatestate_assign_eliminate]: 1.268e-05 [updatestate_loads_eliminate]: 1.276e-05 [parameter_eliminate]: 4.17999e-06 [a_2]: 0.0002121 [accelerated_algorithm]: 1.805e-05 [pynative_shard]: 1.23e-06 [auto_parallel]: 4.28e-06 [parallel]: 4.74e-06 [merge_comm]: 2.44e-06 [allreduce_fusion]: 1.57e-06 [virtual_dataset]: 1.058e-05 [get_grad_eliminate_]: 9.73e-06 [virtual_output]: 9.74e-06 [merge_forward]: 1.439e-05 [cell_reuse_recompute_pass]: 4.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.346e-05 [meta_fg_expand]: 0.0112314, [5] [Cycle 1]: 0.0003226, [1] [resolve]: 0.00030488 [Cycle 1]: 0.00031574, [1] [resolve]: 0.0002985 [Cycle 1]: 0.00167474, [1] [resolve]: 0.00165665 [Cycle 1]: 0.00031693, [1] [resolve]: 0.00029936 [Cycle 1]: 0.000313, [1] [resolve]: 0.00029608 [after_resolve]: 6.973e-05 [a_after_grad]: 0.00015874 [renormalize]: 0.0377254 [real_op_eliminate]: 5.334e-05 [auto_monad_grad]: 0.00019952 [auto_monad_eliminator]: 0.00010536 [cse]: 0.00030398 [a_3]: 0.00042442 [Cycle 3]: 0.00467122, [30] [expand_dump_flag]: 3.98e-06 [switch_simplify]: 0.00011345 [a_1]: 0.00122597 [recompute_prepare]: 1.991e-05 [updatestate_depend_eliminate]: 2.787e-05 [updatestate_assign_eliminate]: 1.844e-05 [updatestate_loads_eliminate]: 1.819e-05 [parameter_eliminate]: 4.50001e-06 [a_2]: 0.0002774 [accelerated_algorithm]: 2.456e-05 [pynative_shard]: 1.19e-06 [auto_parallel]: 3.89e-06 [parallel]: 4.29e-06 [merge_comm]: 3.82e-06 [allreduce_fusion]: 2.40999e-06 [virtual_dataset]: 1.405e-05 [get_grad_eliminate_]: 1.349e-05 [virtual_output]: 1.284e-05 [merge_forward]: 1.889e-05 [cell_reuse_recompute_pass]: 4.30002e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.316e-05 [meta_fg_expand]: 4.797e-05 [after_resolve]: 1.657e-05 [a_after_grad]: 2.12e-05 [renormalize]: 0.00218071 [real_op_eliminate]: 2.004e-05 [auto_monad_grad]: 5.71e-06 [auto_monad_eliminator]: 3.412e-05 [cse]: 0.00020172 [a_3]: 0.00012287 [Cycle 4]: 0.00121697, [30] [expand_dump_flag]: 1.41e-06 [switch_simplify]: 1.481e-05 [a_1]: 0.00028841 [recompute_prepare]: 1.512e-05 [updatestate_depend_eliminate]: 1.929e-05 [updatestate_assign_eliminate]: 1.651e-05 [updatestate_loads_eliminate]: 1.625e-05 [parameter_eliminate]: 2.26e-06 [a_2]: 0.00027509 [accelerated_algorithm]: 5.053e-05 [pynative_shard]: 1.53e-06 [auto_parallel]: 3.85e-06 [parallel]: 3.78e-06 [merge_comm]: 3.32e-06 [allreduce_fusion]: 2.06e-06 [virtual_dataset]: 1.441e-05 [get_grad_eliminate_]: 1.359e-05 [virtual_output]: 1.288e-05 [merge_forward]: 1.786e-05 [cell_reuse_recompute_pass]: 4.30002e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.395e-05 [meta_fg_expand]: 1.326e-05 [after_resolve]: 1.576e-05 [a_after_grad]: 2.045e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.362e-05 [auto_monad_grad]: 2.5e-06 [auto_monad_eliminator]: 3e-05 [cse]: 8.4e-05 [a_3]: 0.00011337 [py_interpret_to_execute_after_opt_a]: 3.84e-06 [slice_cell_reuse_recomputed_activation]: 1.18e-06 [rewriter_after_opt_a]: 9.661e-05 [convert_after_rewriter]: 2.279e-05 [order_py_execute_after_rewriter]: 1.584e-05 [opt_b]: 0.00109485, [2] [Cycle 1]: 0.00092479, [7] [b_1]: 0.00083413 [b_2]: 5.69e-06 [updatestate_depend_eliminate]: 5.62e-06 [updatestate_assign_eliminate]: 4.05e-06 [updatestate_loads_eliminate]: 4.16e-06 [renormalize]: 3.70004e-07 [cse]: 3.58e-05 [Cycle 2]: 0.0001605, [7] [b_1]: 9.876e-05 [b_2]: 4.17e-06 [updatestate_depend_eliminate]: 4.65e-06 [updatestate_assign_eliminate]: 4.1e-06 [updatestate_loads_eliminate]: 3.61e-06 [renormalize]: 5.99975e-08 [cse]: 1.833e-05 [cconv]: 1.723e-05 [opt_after_cconv]: 6.921e-05, [1] [Cycle 1]: 6.481e-05, [7] [c_1]: 8.66e-06 [parameter_eliminate]: 2.16e-06 [updatestate_depend_eliminate]: 4.33e-06 [updatestate_assign_eliminate]: 3.66e-06 [updatestate_loads_eliminate]: 3.57e-06 [cse]: 1.577e-05 [renormalize]: 3.20004e-07 [remove_dup_value]: 1.339e-05 [tuple_transform]: 6.563e-05, [1] [Cycle 1]: 6.196e-05, [3] [d_1]: 3.95e-05 [d_2]: 9.4e-06 [renormalize]: 2.3e-07 [add_cache_embedding]: 9.23e-06 [add_recomputation]: 4.897e-05 [cse_after_recomputation]: 2.506e-05, [1] [Cycle 1]: 2.121e-05, [1] [cse]: 1.659e-05 [environ_conv]: 8.62e-06 [label_micro_interleaved_index]: 1.48e-06 [label_fine_grained_interleaved_index]: 1.33e-06 [assign_add_opt]: 1.06e-06 [slice_recompute_activation]: 1.6e-06 [micro_interleaved_order_control]: 1.65e-06 [full_micro_interleaved_order_control]: 9.59997e-07 [comp_comm_scheduling]: 1.42e-06 [reorder_send_recv_between_fp_bp]: 1.65e-06 [comm_op_add_attrs]: 6.40001e-07 [add_comm_op_reuse_tag]: 6.69999e-07 [overlap_opt_shard_in_pipeline]: 6.30003e-07 [grouped_pairwise_exchange_alltoall]: 5.60001e-07 [overlap_recompute_and_grad_model_parallel]: 1.25e-06 [overlap_grad_matmul_and_grad_allreduce]: 5.20005e-07 [split_matmul_comm_elemetwise]: 1.45e-06 [split_layernorm_comm]: 1.09e-06 [process_send_recv_for_ge]: 7.2e-07 [handle_group_info]: 5.69999e-07 [auto_monad_reorder]: 1.927e-05 [get_jit_bprop_graph]: 7.40001e-07 [eliminate_special_op_node]: 0.0004753 [validate]: 3.585e-05 [distribtued_split]: 1.1e-06 [task_emit]: 0.00547762 [execute]: 4.86e-06 Sums parse : 0.001295s : 1.01% symbol_resolve.resolve : 0.012231s : 9.55% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000124s : 0.10% pre_cconv : 0.000000s : 0.00% abstract_specialize : 0.003958s : 3.09% pack_expand : 0.000016s : 0.01% auto_monad : 0.000073s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000007s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000187s : 0.15% optimize.opt_a.expand_dump_flag : 0.000012s : 0.01% optimize.opt_a.switch_simplify : 0.000254s : 0.20% optimize.opt_a.a_1 : 0.002653s : 2.07% optimize.opt_a.recompute_prepare : 0.000057s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000073s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000054s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000053s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000843s : 0.66% optimize.opt_a.accelerated_algorithm : 0.000099s : 0.08% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000015s : 0.01% optimize.opt_a.parallel : 0.000019s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000040s : 0.03% optimize.opt_a.merge_forward : 0.000058s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000061s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006019s : 4.70% optimize.opt_a.after_resolve : 0.000147s : 0.11% optimize.opt_a.a_after_grad : 0.000307s : 0.24% optimize.opt_a.renormalize : 0.089277s : 69.69% optimize.opt_a.real_op_eliminate : 0.000128s : 0.10% optimize.opt_a.auto_monad_grad : 0.000274s : 0.21% optimize.opt_a.auto_monad_eliminator : 0.000247s : 0.19% optimize.opt_a.cse : 0.000917s : 0.72% optimize.opt_a.a_3 : 0.000966s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000001s : 0.00% optimize.rewriter_after_opt_a : 0.000097s : 0.08% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000933s : 0.73% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000054s : 0.04% optimize.cconv : 0.000017s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000040s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000009s : 0.01% optimize.add_recomputation : 0.000049s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000001s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000001s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000001s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000019s : 0.02% get_jit_bprop_graph : 0.000001s : 0.00% eliminate_special_op_node : 0.000475s : 0.37% validate : 0.000036s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005478s : 4.28% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018461 880 0.02% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.65% : 0.016734s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.59% : 0.001032s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000028s : 23: substitution.meta_unpack_prepare 0.16% : 0.000029s : 40: substitution.minmaximum_grad 0.02% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000008s : 81: substitution.remove_not_recompute_node 0.48% : 0.000088s : 63: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.61% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_item_depend_reorder 0.82% : 0.000152s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000066s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089263 6 92.65% : 0.082703s : 3: renormalize.infer 7.35% : 0.006559s : 3: renormalize.specialize ------[replace.] 0.001218 141 53.90% : 0.000656s : 55: replace.getattr_setattr_resolve 26.85% : 0.000327s : 56: replace.inline 3.61% : 0.000044s : 2: replace.meta_unpack_prepare 7.48% : 0.000091s : 10: replace.switch_simplify 1.52% : 0.000019s : 4: replace.transpose_eliminate 6.64% : 0.000081s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017537 141 94.68% : 0.016605s : 55: match.getattr_setattr_resolve 4.94% : 0.000867s : 56: match.inline 0.09% : 0.000016s : 2: match.meta_unpack_prepare 0.05% : 0.000010s : 10: match.switch_simplify 0.08% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000026s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006583 119 69.25% : 0.004559s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.75% : 0.002024s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.024609 259 7.55% : 0.001858s : 104: opt.transform.opt_a 3.66% : 0.000901s : 92: opt.transform.opt_b 72.87% : 0.017933s : 14: opt.transform.opt_resolve 0.44% : 0.000108s : 1: opt.transforms.meta_unpack_prepare 15.16% : 0.003731s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000047s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:28.913.816 [graph_var_manager.cc:1424][EVENT]36565 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:28.913.909 [graph_manager.cc:1248][EVENT]36565 PreRun:PreRun start: graph node size 3, session id 2, graph id 1, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.914.736 [atrace_api.c:28](tid:36565) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.914.816 [trace_rb_log.c:84](tid:36565) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.914.831 [atrace_api.c:32](tid:36565) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:28.914.845 [client_manager.cpp:157][SetProfilingCallback][tid:36565] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:28.915.721 [parallel_partitioner.cc:165][EVENT]36565 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.915.762 [parallel_partitioner.cc:178][EVENT]36565 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.915.811 [graph_prepare.cc:1378][EVENT]36565 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.479 [graph_manager.cc:1050][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [684] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.509 [graph_manager.cc:1052][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.638 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.670 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.727 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [43] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.742 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.786 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.800 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.817 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.923 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.916.945 [graph_manager.cc:1054][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [423] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.917.195 [graph_manager.cc:1055][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [237] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.167 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.190 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.201 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.220 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [308] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.230 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.239 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.247 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.256 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [16] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.918.265 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.919.616 [graph_manager.cc:1056][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2399] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.919.677 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.919.695 [graph_prepare.cc:1982][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.092 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.113 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.123 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.132 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [220] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.141 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.150 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.158 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.166 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.175 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.200 [graph_prepare.cc:1983][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [491] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.223 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.236 [graph_prepare.cc:1984][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.249 [graph_prepare.cc:1985][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.268 [graph_prepare.cc:1986][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.288 [graph_prepare.cc:1987][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.304 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.317 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.331 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.414 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.427 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.436 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrintOpPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.444 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.453 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.461 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.469 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.478 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.486 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.494 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.503 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.511 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SnapshotPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.519 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.527 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.535 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.544 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.565 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.578 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.609 [graph_prepare.cc:1988][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [311] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.920.627 [graph_manager.cc:1065][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [983] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.932.969 [graph_manager.cc:1077][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12321] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.933.037 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.933.093 [graph_manager.cc:1080][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [89] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.776 [graph_manager.cc:1081][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2666] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.813 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.828 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.839 [graph_manager.cc:1082][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.871 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.887 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.901 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.972 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.935.989 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.021 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.037 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.077 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.095 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.113 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.138 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.154 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.166 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.187 [graph_manager.cc:2700][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [321] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.293 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.307 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AddNPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.316 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.325 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.334 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.343 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CastRemovePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.351 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.359 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.368 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.376 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.384 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.393 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.401 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.409 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.417 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.427 [graph_manager.cc:2741][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [221] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.437 [graph_manager.cc:2752][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.459 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.472 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.488 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.504 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.522 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.534 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.552 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.566 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.579 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.589 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.603 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.613 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.632 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.645 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.655 [graph_manager.cc:2810][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [201] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.683 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.696 [graph_manager.cc:2821][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.724 [graph_manager.cc:1087][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [866] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.862 [graph_manager.cc:1088][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [124] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.899 [graph_manager.cc:1089][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.917 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.932 [graph_manager.cc:1097][EVENT]36565 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.936.954 [graph_manager.cc:3325][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.193 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.211 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.220 [engine_place.cc:144][EVENT]36565 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [149] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.291 [graph_manager.cc:3351][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [324] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.316 [graph_manager.cc:3364][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.377 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.395 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.549 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [145] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.590 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.640 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [38] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.675 [graph_manager.cc:3405][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [346] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.937.694 [graph_manager.cc:3412][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.100 [graph_manager.cc:3422][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1392] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.128 [graph_manager.cc:3428][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.246 [graph_manager.cc:3467][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [99] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.263 [graph_manager.cc:3377][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1934] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.279 [graph_manager.cc:1106][EVENT]36565 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2332] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.292 [graph_manager.cc:1115][EVENT]36565 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.314 [graph_manager.cc:1130][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.345 [graph_manager.cc:1131][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.368 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.386 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.395 [graph_manager.cc:2837][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.471 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.484 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.493 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.502 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of BitcastPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.510 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.519 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.528 [graph_manager.cc:2864][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [110] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.540 [graph_manager.cc:2872][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.560 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.575 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.590 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.603 [compile_nodes_pass.cc:88][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.614 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.624 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.702 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [69] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.730 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.744 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.757 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.770 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.781 [graph_manager.cc:2927][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [223] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.793 [graph_manager.cc:2937][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.811 [graph_manager.cc:2943][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.823 [graph_manager.cc:2950][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.939.980 [graph_manager.cc:2958][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.011 [graph_manager.cc:1132][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [652] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.084 [graph_manager.cc:1135][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.121 [graph_manager.cc:2975][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.152 [graph_manager.cc:2981][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.167 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.177 [graph_manager.cc:2986][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.186 [graph_manager.cc:1136][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.289 [graph_manager.cc:3555][EVENT]36565 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [71] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.372 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.386 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.506 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [110] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.537 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.577 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.940.598 [graph_builder.cc:865][EVENT]36565 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [256] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:28.941.077 [logger.cc:1071] 36565 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.107 [task_generator.cc:804][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [165] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.183 [task_generator.cc:805][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [45] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.619 [task_generator.cc:814][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [419] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.639 [task_generator.cc:954][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [698] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.695 [task_generator.cc:967][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [32] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:28.941.713 [logger.cc:1084] 36565 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.862 [graph_manager.cc:1152][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1652] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.880 [graph_manager.cc:1164][EVENT]36565 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.913 [graph_manager.cc:1271][EVENT]36565 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [26278] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.941.924 [graph_manager.cc:1272][EVENT]36565 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.942.227 [atrace_api.c:93](tid:36565) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:28.942.242 [atrace_api.c:95](tid:36565) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:28.946.526 [graph_converter.cc:838][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1221] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.946.716 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [149] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.103 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [364] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.184 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.198 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [76] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.483 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [274] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.589 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [88] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.623 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.772 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [136] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.840 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [53] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.852 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [65] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.880 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.906 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.932 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.947.992 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.049 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.059 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [57] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.093 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.119 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.139 [graph_converter.cc:849][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1577] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.333 [graph_converter.cc:853][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [184] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.948.962 [graph_converter.cc:857][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [616] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:28.949.082 [graph_converter.cc:862][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [99] micro second. . TotalTime = 0.145873, [20] [parse]: 0.0014794 [symbol_resolve]: 0.0124051, [1] [Cycle 1]: 0.01233, [1] [resolve]: 0.0123101 [combine_like_graphs]: 9.89996e-07 [graph_reusing]: 3.36e-06 [meta_unpack_prepare]: 0.00017821 [pre_cconv]: 6.79996e-07 [abstract_specialize]: 0.00423313 [pack_expand]: 1.943e-05 [auto_monad]: 8.298e-05 [inline]: 1.68e-06 [pre_auto_parallel]: 1.091e-05 [pipeline_split]: 2.81e-06 [optimize]: 0.12313, [35] [py_interpret_to_execute]: 4.30999e-06 [rewriter_before_opt_a]: 0.00019011 [opt_a]: 0.121227, [4] [Cycle 1]: 0.0579938, [30] [expand_dump_flag]: 4.1e-06 [switch_simplify]: 2.639e-05 [a_1]: 0.00071894 [recompute_prepare]: 7.81e-06 [updatestate_depend_eliminate]: 1.075e-05 [updatestate_assign_eliminate]: 6.94999e-06 [updatestate_loads_eliminate]: 6.71e-06 [parameter_eliminate]: 5.01e-06 [a_2]: 7.578e-05 [accelerated_algorithm]: 5.06e-06 [pynative_shard]: 1.76e-06 [auto_parallel]: 3.26e-06 [parallel]: 8.69e-06 [merge_comm]: 4.03e-06 [allreduce_fusion]: 2.29e-06 [virtual_dataset]: 5.5e-06 [get_grad_eliminate_]: 4.62e-06 [virtual_output]: 4.43e-06 [merge_forward]: 8.51e-06 [cell_reuse_recompute_pass]: 9.39996e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.664e-05 [meta_fg_expand]: 0.00691248, [1] [Cycle 1]: 0.00318271, [1] [resolve]: 0.00316329 [after_resolve]: 4.726e-05 [a_after_grad]: 0.00013025 [renormalize]: 0.0489319 [real_op_eliminate]: 4.409e-05 [auto_monad_grad]: 6.321e-05 [auto_monad_eliminator]: 7.921e-05 [cse]: 0.00035456 [a_3]: 0.00031031 [Cycle 2]: 0.0532335, [30] [expand_dump_flag]: 3.79e-06 [switch_simplify]: 0.00013032 [a_1]: 0.00159966 [recompute_prepare]: 1.282e-05 [updatestate_depend_eliminate]: 1.679e-05 [updatestate_assign_eliminate]: 1.32e-05 [updatestate_loads_eliminate]: 1.281e-05 [parameter_eliminate]: 4.04e-06 [a_2]: 0.00019396 [accelerated_algorithm]: 1.925e-05 [pynative_shard]: 1.99e-06 [auto_parallel]: 5.34e-06 [parallel]: 4.76e-06 [merge_comm]: 2.57e-06 [allreduce_fusion]: 1.52e-06 [virtual_dataset]: 1.11e-05 [get_grad_eliminate_]: 1.025e-05 [virtual_output]: 1.014e-05 [merge_forward]: 1.419e-05 [cell_reuse_recompute_pass]: 4.69998e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.363e-05 [meta_fg_expand]: 0.0114087, [5] [Cycle 1]: 0.00032058, [1] [resolve]: 0.00030143 [Cycle 1]: 0.00031254, [1] [resolve]: 0.00029407 [Cycle 1]: 0.00164881, [1] [resolve]: 0.00162916 [Cycle 1]: 0.0003157, [1] [resolve]: 0.00029695 [Cycle 1]: 0.00031008, [1] [resolve]: 0.00029128 [after_resolve]: 7.498e-05 [a_after_grad]: 0.00019644 [renormalize]: 0.0381354 [real_op_eliminate]: 6.032e-05 [auto_monad_grad]: 0.00021563 [auto_monad_eliminator]: 0.0001101 [cse]: 0.00032557 [a_3]: 0.00042388 [Cycle 3]: 0.00598957, [30] [expand_dump_flag]: 4.91e-06 [switch_simplify]: 0.0001578 [a_1]: 0.00233254 [recompute_prepare]: 1.684e-05 [updatestate_depend_eliminate]: 3.324e-05 [updatestate_assign_eliminate]: 1.88e-05 [updatestate_loads_eliminate]: 1.758e-05 [parameter_eliminate]: 4.7e-06 [a_2]: 0.00027373 [accelerated_algorithm]: 2.528e-05 [pynative_shard]: 1.97e-06 [auto_parallel]: 6.3e-06 [parallel]: 5.92e-06 [merge_comm]: 4.1e-06 [allreduce_fusion]: 2.71e-06 [virtual_dataset]: 1.487e-05 [get_grad_eliminate_]: 1.434e-05 [virtual_output]: 1.396e-05 [merge_forward]: 1.938e-05 [cell_reuse_recompute_pass]: 9.29998e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.314e-05 [meta_fg_expand]: 5.177e-05 [after_resolve]: 1.834e-05 [a_after_grad]: 3.156e-05 [renormalize]: 0.00231898 [real_op_eliminate]: 2.109e-05 [auto_monad_grad]: 5.86001e-06 [auto_monad_eliminator]: 3.45e-05 [cse]: 0.00022285 [a_3]: 0.00012234 [Cycle 4]: 0.00165532, [30] [expand_dump_flag]: 1.42e-06 [switch_simplify]: 1.521e-05 [a_1]: 0.00074141 [recompute_prepare]: 1.549e-05 [updatestate_depend_eliminate]: 2.008e-05 [updatestate_assign_eliminate]: 1.703e-05 [updatestate_loads_eliminate]: 1.65e-05 [parameter_eliminate]: 2.35e-06 [a_2]: 0.00027379 [accelerated_algorithm]: 2.503e-05 [pynative_shard]: 1.41e-06 [auto_parallel]: 4.95e-06 [parallel]: 3.72e-06 [merge_comm]: 3.26e-06 [allreduce_fusion]: 2.02e-06 [virtual_dataset]: 1.522e-05 [get_grad_eliminate_]: 1.492e-05 [virtual_output]: 1.394e-05 [merge_forward]: 1.689e-05 [cell_reuse_recompute_pass]: 3.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.246e-05 [meta_fg_expand]: 1.289e-05 [after_resolve]: 1.659e-05 [a_after_grad]: 3.149e-05 [renormalize]: 5.99975e-08 [real_op_eliminate]: 1.433e-05 [auto_monad_grad]: 2.03e-06 [auto_monad_eliminator]: 3.008e-05 [cse]: 8.284e-05 [a_3]: 0.00011254 [py_interpret_to_execute_after_opt_a]: 4.14e-06 [slice_cell_reuse_recomputed_activation]: 2.27e-06 [rewriter_after_opt_a]: 0.00010008 [convert_after_rewriter]: 2.33e-05 [order_py_execute_after_rewriter]: 1.696e-05 [opt_b]: 0.00110482, [2] [Cycle 1]: 0.00093495, [7] [b_1]: 0.00083956 [b_2]: 5.2e-06 [updatestate_depend_eliminate]: 5.67e-06 [updatestate_assign_eliminate]: 4.2e-06 [updatestate_loads_eliminate]: 3.73001e-06 [renormalize]: 4.89999e-07 [cse]: 3.902e-05 [Cycle 2]: 0.00015882, [7] [b_1]: 9.722e-05 [b_2]: 3.72e-06 [updatestate_depend_eliminate]: 4.81e-06 [updatestate_assign_eliminate]: 3.93e-06 [updatestate_loads_eliminate]: 3.63999e-06 [renormalize]: 6.00048e-08 [cse]: 1.844e-05 [cconv]: 2.215e-05 [opt_after_cconv]: 8.665e-05, [1] [Cycle 1]: 8.21e-05, [7] [c_1]: 2.421e-05 [parameter_eliminate]: 2.36e-06 [updatestate_depend_eliminate]: 4.11001e-06 [updatestate_assign_eliminate]: 4.04e-06 [updatestate_loads_eliminate]: 3.60001e-06 [cse]: 1.671e-05 [renormalize]: 4.19997e-07 [remove_dup_value]: 1.78e-05 [tuple_transform]: 8.338e-05, [1] [Cycle 1]: 7.94e-05, [3] [d_1]: 5.764e-05 [d_2]: 9.25e-06 [renormalize]: 2.20003e-07 [add_cache_embedding]: 1.284e-05 [add_recomputation]: 5.857e-05 [cse_after_recomputation]: 2.667e-05, [1] [Cycle 1]: 2.252e-05, [1] [cse]: 1.773e-05 [environ_conv]: 9.7e-06 [label_micro_interleaved_index]: 2.73e-06 [label_fine_grained_interleaved_index]: 2.5e-06 [assign_add_opt]: 1.44e-06 [slice_recompute_activation]: 1.96999e-06 [micro_interleaved_order_control]: 1.84e-06 [full_micro_interleaved_order_control]: 1.78999e-06 [comp_comm_scheduling]: 1.93e-06 [reorder_send_recv_between_fp_bp]: 2.28e-06 [comm_op_add_attrs]: 1.05e-06 [add_comm_op_reuse_tag]: 8.39995e-07 [overlap_opt_shard_in_pipeline]: 1.28e-06 [grouped_pairwise_exchange_alltoall]: 1.53e-06 [overlap_recompute_and_grad_model_parallel]: 1.66e-06 [overlap_grad_matmul_and_grad_allreduce]: 7.00005e-07 [split_matmul_comm_elemetwise]: 1.99e-06 [split_layernorm_comm]: 1.68e-06 [process_send_recv_for_ge]: 9.70002e-07 [handle_group_info]: 9.70002e-07 [auto_monad_reorder]: 2.274e-05 [get_jit_bprop_graph]: 4.07999e-06 [eliminate_special_op_node]: 0.00055529 [validate]: 3.954e-05 [distribtued_split]: 1.15e-06 [task_emit]: 0.00349495 [execute]: 6.57e-06 Sums parse : 0.001479s : 1.14% symbol_resolve.resolve : 0.012310s : 9.47% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000178s : 0.14% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004233s : 3.26% pack_expand : 0.000019s : 0.01% auto_monad : 0.000083s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000011s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000190s : 0.15% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000330s : 0.25% optimize.opt_a.a_1 : 0.005393s : 4.15% optimize.opt_a.recompute_prepare : 0.000053s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000081s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000817s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.06% optimize.opt_a.pynative_shard : 0.000007s : 0.01% optimize.opt_a.auto_parallel : 0.000020s : 0.02% optimize.opt_a.parallel : 0.000023s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000009s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000042s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000116s : 0.09% optimize.opt_a.meta_fg_expand : 0.000065s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.005976s : 4.60% optimize.opt_a.after_resolve : 0.000157s : 0.12% optimize.opt_a.a_after_grad : 0.000390s : 0.30% optimize.opt_a.renormalize : 0.089386s : 68.76% optimize.opt_a.real_op_eliminate : 0.000140s : 0.11% optimize.opt_a.auto_monad_grad : 0.000287s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000254s : 0.20% optimize.opt_a.cse : 0.000986s : 0.76% optimize.opt_a.a_3 : 0.000969s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000100s : 0.08% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000937s : 0.72% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000007s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000057s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000058s : 0.04% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000059s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000003s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000023s : 0.02% get_jit_bprop_graph : 0.000004s : 0.00% eliminate_special_op_node : 0.000555s : 0.43% validate : 0.000040s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003495s : 2.69% execute : 0.000007s : 0.01% Time group info: ------[substitution.] 0.018532 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.30% : 0.016734s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.02% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.52% : 0.001023s : 103: substitution.inline 0.05% : 0.000010s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000034s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.51% : 0.000094s : 69: substitution.replace_applicator 0.06% : 0.000011s : 36: substitution.replace_old_param 0.02% : 0.000004s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.05% : 0.000009s : 5: substitution.specialize_transform 0.07% : 0.000013s : 10: substitution.switch_simplify 0.07% : 0.000012s : 4: substitution.transpose_eliminate 0.66% : 0.000121s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.41% : 0.000076s : 70: substitution.tuple_list_get_item_depend_reorder 0.89% : 0.000164s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000077s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089369 6 92.31% : 0.082496s : 3: renormalize.infer 7.69% : 0.006873s : 3: renormalize.specialize ------[replace.] 0.001230 141 54.83% : 0.000674s : 55: replace.getattr_setattr_resolve 25.89% : 0.000318s : 56: replace.inline 3.58% : 0.000044s : 2: replace.meta_unpack_prepare 7.37% : 0.000091s : 10: replace.switch_simplify 1.59% : 0.000019s : 4: replace.transpose_eliminate 6.75% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017519 141 94.76% : 0.016601s : 55: match.getattr_setattr_resolve 4.85% : 0.000849s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.07% : 0.000013s : 10: match.switch_simplify 0.07% : 0.000012s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006845 119 68.62% : 0.004697s : 53: func_graph_cloner_run.FuncGraphClonerGraph 31.38% : 0.002148s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027510 589 0.56% : 0.000155s : 2: opt.transform.meta_unpack_prepare 30.46% : 0.008379s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.31% : 0.000911s : 94: opt.transform.opt_b 65.31% : 0.017968s : 14: opt.transform.opt_resolve 0.23% : 0.000062s : 8: opt.transform.opt_trans_graph 0.06% : 0.000016s : 3: opt.transform.special_op_eliminate . TotalTime = 0.14545, [20] [parse]: 0.0013675 [symbol_resolve]: 0.0124905, [1] [Cycle 1]: 0.0124246, [1] [resolve]: 0.0124044 [combine_like_graphs]: 8.70001e-07 [graph_reusing]: 3.49e-06 [meta_unpack_prepare]: 0.00013275 [pre_cconv]: 6.10002e-07 [abstract_specialize]: 0.00414989 [pack_expand]: 1.476e-05 [auto_monad]: 8.032e-05 [inline]: 1.54e-06 [pre_auto_parallel]: 8.45001e-06 [pipeline_split]: 2.9e-06 [optimize]: 0.120825, [35] [py_interpret_to_execute]: 4.07999e-06 [rewriter_before_opt_a]: 0.00019132 [opt_a]: 0.118978, [4] [Cycle 1]: 0.0581788, [30] [expand_dump_flag]: 3.5e-06 [switch_simplify]: 2.679e-05 [a_1]: 0.00038951 [recompute_prepare]: 9.45001e-06 [updatestate_depend_eliminate]: 9.89001e-06 [updatestate_assign_eliminate]: 6.55e-06 [updatestate_loads_eliminate]: 6.12e-06 [parameter_eliminate]: 4.78e-06 [a_2]: 7.975e-05 [accelerated_algorithm]: 5.48e-06 [pynative_shard]: 1.49e-06 [auto_parallel]: 3.26e-06 [parallel]: 6.23e-06 [merge_comm]: 2.85e-06 [allreduce_fusion]: 1.76e-06 [virtual_dataset]: 5.3e-06 [get_grad_eliminate_]: 4.62e-06 [virtual_output]: 4.05e-06 [merge_forward]: 7.79e-06 [cell_reuse_recompute_pass]: 3.80001e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.127e-05 [meta_fg_expand]: 0.0071091, [1] [Cycle 1]: 0.00331606, [1] [resolve]: 0.00329718 [after_resolve]: 4.833e-05 [a_after_grad]: 0.00011001 [renormalize]: 0.0493033 [real_op_eliminate]: 4.03e-05 [auto_monad_grad]: 7e-05 [auto_monad_eliminator]: 7.794e-05 [cse]: 0.00033941 [a_3]: 0.00030781 [Cycle 2]: 0.0522639, [30] [expand_dump_flag]: 3.74e-06 [switch_simplify]: 0.00010104 [a_1]: 0.00075444 [recompute_prepare]: 1.362e-05 [updatestate_depend_eliminate]: 1.685e-05 [updatestate_assign_eliminate]: 1.273e-05 [updatestate_loads_eliminate]: 1.236e-05 [parameter_eliminate]: 3.7e-06 [a_2]: 0.00019761 [accelerated_algorithm]: 1.812e-05 [pynative_shard]: 1.11001e-06 [auto_parallel]: 4.8e-06 [parallel]: 4.7e-06 [merge_comm]: 2.42e-06 [allreduce_fusion]: 1.52e-06 [virtual_dataset]: 1.064e-05 [get_grad_eliminate_]: 9.69999e-06 [virtual_output]: 9.52e-06 [merge_forward]: 1.432e-05 [cell_reuse_recompute_pass]: 4.20005e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.363e-05 [meta_fg_expand]: 0.0113169, [5] [Cycle 1]: 0.00032094, [1] [resolve]: 0.00030315 [Cycle 1]: 0.00031584, [1] [resolve]: 0.00029753 [Cycle 1]: 0.0016858, [1] [resolve]: 0.00166697 [Cycle 1]: 0.00031645, [1] [resolve]: 0.00029811 [Cycle 1]: 0.00031661, [1] [resolve]: 0.00029842 [after_resolve]: 7.116e-05 [a_after_grad]: 0.00016057 [renormalize]: 0.038219 [real_op_eliminate]: 5.277e-05 [auto_monad_grad]: 0.00020569 [auto_monad_eliminator]: 0.00010709 [cse]: 0.00030645 [a_3]: 0.00042354 [Cycle 3]: 0.00475336, [30] [expand_dump_flag]: 4.61e-06 [switch_simplify]: 0.0001134 [a_1]: 0.00129302 [recompute_prepare]: 1.815e-05 [updatestate_depend_eliminate]: 2.792e-05 [updatestate_assign_eliminate]: 1.823e-05 [updatestate_loads_eliminate]: 1.759e-05 [parameter_eliminate]: 4.35e-06 [a_2]: 0.00027669 [accelerated_algorithm]: 2.457e-05 [pynative_shard]: 1.25e-06 [auto_parallel]: 4.33e-06 [parallel]: 4.09e-06 [merge_comm]: 3.8e-06 [allreduce_fusion]: 2.34e-06 [virtual_dataset]: 1.426e-05 [get_grad_eliminate_]: 1.315e-05 [virtual_output]: 1.259e-05 [merge_forward]: 1.927e-05 [cell_reuse_recompute_pass]: 5.00004e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.292e-05 [meta_fg_expand]: 4.633e-05 [after_resolve]: 1.665e-05 [a_after_grad]: 2.054e-05 [renormalize]: 0.00221338 [real_op_eliminate]: 1.969e-05 [auto_monad_grad]: 5.69e-06 [auto_monad_eliminator]: 3.523e-05 [cse]: 0.0002037 [a_3]: 0.00012262 [Cycle 4]: 0.0012067, [30] [expand_dump_flag]: 1.26001e-06 [switch_simplify]: 1.483e-05 [a_1]: 0.00028461 [recompute_prepare]: 1.515e-05 [updatestate_depend_eliminate]: 2.009e-05 [updatestate_assign_eliminate]: 1.664e-05 [updatestate_loads_eliminate]: 1.631e-05 [parameter_eliminate]: 2.11001e-06 [a_2]: 0.00029125 [accelerated_algorithm]: 2.491e-05 [pynative_shard]: 1.55999e-06 [auto_parallel]: 3.73e-06 [parallel]: 3.60001e-06 [merge_comm]: 3.09e-06 [allreduce_fusion]: 2.03e-06 [virtual_dataset]: 1.41e-05 [get_grad_eliminate_]: 1.397e-05 [virtual_output]: 1.294e-05 [merge_forward]: 1.815e-05 [cell_reuse_recompute_pass]: 3.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.364e-05 [meta_fg_expand]: 1.337e-05 [after_resolve]: 1.573e-05 [a_after_grad]: 1.998e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.352e-05 [auto_monad_grad]: 2.19999e-06 [auto_monad_eliminator]: 3.026e-05 [cse]: 8.517e-05 [a_3]: 0.0001133 [py_interpret_to_execute_after_opt_a]: 3.84e-06 [slice_cell_reuse_recomputed_activation]: 1.76e-06 [rewriter_after_opt_a]: 9.575e-05 [convert_after_rewriter]: 2.304e-05 [order_py_execute_after_rewriter]: 1.639e-05 [opt_b]: 0.00110584, [2] [Cycle 1]: 0.00093644, [7] [b_1]: 0.00084216 [b_2]: 5.38e-06 [updatestate_depend_eliminate]: 6.23e-06 [updatestate_assign_eliminate]: 4.39e-06 [updatestate_loads_eliminate]: 4.3e-06 [renormalize]: 4.79995e-07 [cse]: 3.786e-05 [Cycle 2]: 0.00015918, [7] [b_1]: 9.744e-05 [b_2]: 4.12e-06 [updatestate_depend_eliminate]: 4.75e-06 [updatestate_assign_eliminate]: 3.83e-06 [updatestate_loads_eliminate]: 3.78e-06 [renormalize]: 6.00048e-08 [cse]: 1.842e-05 [cconv]: 1.978e-05 [opt_after_cconv]: 6.958e-05, [1] [Cycle 1]: 6.499e-05, [7] [c_1]: 8.44e-06 [parameter_eliminate]: 2.21e-06 [updatestate_depend_eliminate]: 4.31e-06 [updatestate_assign_eliminate]: 3.52e-06 [updatestate_loads_eliminate]: 3.5e-06 [cse]: 1.633e-05 [renormalize]: 4.60001e-07 [remove_dup_value]: 1.509e-05 [tuple_transform]: 6.65e-05, [1] [Cycle 1]: 6.252e-05, [3] [d_1]: 4.089e-05 [d_2]: 9.04e-06 [renormalize]: 2.09999e-07 [add_cache_embedding]: 1.113e-05 [add_recomputation]: 5.558e-05 [cse_after_recomputation]: 2.62e-05, [1] [Cycle 1]: 2.217e-05, [1] [cse]: 1.777e-05 [environ_conv]: 9.32e-06 [label_micro_interleaved_index]: 2.55e-06 [label_fine_grained_interleaved_index]: 1.67e-06 [assign_add_opt]: 1.14999e-06 [slice_recompute_activation]: 1.79e-06 [micro_interleaved_order_control]: 1.84e-06 [full_micro_interleaved_order_control]: 1.34e-06 [comp_comm_scheduling]: 2.12e-06 [reorder_send_recv_between_fp_bp]: 1.60999e-06 [comm_op_add_attrs]: 8.90002e-07 [add_comm_op_reuse_tag]: 8.70001e-07 [overlap_opt_shard_in_pipeline]: 7.40001e-07 [grouped_pairwise_exchange_alltoall]: 9.5e-07 [overlap_recompute_and_grad_model_parallel]: 1.88001e-06 [overlap_grad_matmul_and_grad_allreduce]: 5.10001e-07 [split_matmul_comm_elemetwise]: 1.81e-06 [split_layernorm_comm]: 1.57e-06 [process_send_recv_for_ge]: 6.99998e-07 [handle_group_info]: 9.59997e-07 [auto_monad_reorder]: 1.985e-05 [get_jit_bprop_graph]: 4.49996e-07 [eliminate_special_op_node]: 0.00051245 [validate]: 3.788e-05 [distribtued_split]: 1.18e-06 [task_emit]: 0.00561486 [execute]: 6.14e-06 Sums parse : 0.001368s : 1.06% symbol_resolve.resolve : 0.012404s : 9.58% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000133s : 0.10% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004150s : 3.21% pack_expand : 0.000015s : 0.01% auto_monad : 0.000080s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000008s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000191s : 0.15% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000256s : 0.20% optimize.opt_a.a_1 : 0.002722s : 2.10% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000075s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000054s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000052s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000845s : 0.65% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000016s : 0.01% optimize.opt_a.parallel : 0.000019s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000039s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000101s : 0.08% optimize.opt_a.meta_fg_expand : 0.000060s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006161s : 4.76% optimize.opt_a.after_resolve : 0.000152s : 0.12% optimize.opt_a.a_after_grad : 0.000311s : 0.24% optimize.opt_a.renormalize : 0.089736s : 69.31% optimize.opt_a.real_op_eliminate : 0.000126s : 0.10% optimize.opt_a.auto_monad_grad : 0.000284s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000251s : 0.19% optimize.opt_a.cse : 0.000935s : 0.72% optimize.opt_a.a_3 : 0.000967s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000096s : 0.07% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000940s : 0.73% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000056s : 0.04% optimize.cconv : 0.000020s : 0.02% optimize.opt_after_cconv.c_1 : 0.000008s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000015s : 0.01% optimize.tuple_transform.d_1 : 0.000041s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000011s : 0.01% optimize.add_recomputation : 0.000056s : 0.04% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000003s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000020s : 0.02% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000512s : 0.40% validate : 0.000038s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005615s : 4.34% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018756 880 0.02% : 0.000003s : 5: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.59% : 0.016991s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.62% : 0.001053s : 97: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.17% : 0.000031s : 23: substitution.meta_unpack_prepare 0.16% : 0.000030s : 40: substitution.minmaximum_grad 0.02% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.04% : 0.000008s : 81: substitution.remove_not_recompute_node 0.48% : 0.000090s : 63: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000014s : 4: substitution.transpose_eliminate 0.60% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000068s : 60: substitution.tuple_list_get_item_depend_reorder 0.83% : 0.000156s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000068s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089720 6 92.50% : 0.082990s : 3: renormalize.infer 7.50% : 0.006731s : 3: renormalize.specialize ------[replace.] 0.001234 141 54.06% : 0.000667s : 55: replace.getattr_setattr_resolve 26.99% : 0.000333s : 56: replace.inline 3.57% : 0.000044s : 2: replace.meta_unpack_prepare 7.31% : 0.000090s : 10: replace.switch_simplify 1.46% : 0.000018s : 4: replace.transpose_eliminate 6.62% : 0.000082s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017815 141 94.64% : 0.016859s : 55: match.getattr_setattr_resolve 4.97% : 0.000885s : 56: match.inline 0.10% : 0.000018s : 2: match.meta_unpack_prepare 0.06% : 0.000010s : 10: match.switch_simplify 0.08% : 0.000014s : 4: match.transpose_eliminate 0.16% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006858 119 68.87% : 0.004723s : 53: func_graph_cloner_run.FuncGraphClonerGraph 31.13% : 0.002135s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.024985 259 7.45% : 0.001861s : 104: opt.transform.opt_a 3.63% : 0.000907s : 92: opt.transform.opt_b 73.02% : 0.018245s : 14: opt.transform.opt_resolve 0.45% : 0.000112s : 1: opt.transforms.meta_unpack_prepare 15.14% : 0.003783s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000007s : 2: opt.transforms.opt_b 0.19% : 0.000048s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:29.338.712 [graph_var_manager.cc:1424][EVENT]36565 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:29.338.789 [graph_manager.cc:1248][EVENT]36565 PreRun:PreRun start: graph node size 3, session id 3, graph id 2, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.339.105 [atrace_api.c:28](tid:36565) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.339.160 [trace_rb_log.c:84](tid:36565) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.339.173 [atrace_api.c:32](tid:36565) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:29.339.186 [client_manager.cpp:157][SetProfilingCallback][tid:36565] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:29.339.613 [parallel_partitioner.cc:165][EVENT]36565 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.339.649 [parallel_partitioner.cc:178][EVENT]36565 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [12] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.339.713 [graph_prepare.cc:1378][EVENT]36565 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.339.930 [graph_manager.cc:1050][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [234] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.339.956 [graph_manager.cc:1052][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.082 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.112 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.165 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [42] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.179 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.225 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.239 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.257 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.353 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.374 [graph_manager.cc:1054][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [404] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.340.598 [graph_manager.cc:1055][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [210] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.578 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.602 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.613 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.623 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [301] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.632 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.642 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.650 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.659 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [16] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.341.668 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.005 [graph_manager.cc:1056][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2387] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.066 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.082 [graph_prepare.cc:1982][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [49] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.477 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.498 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.509 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.518 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [224] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.526 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.535 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.543 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.552 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.560 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.585 [graph_prepare.cc:1983][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [489] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.608 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.619 [graph_prepare.cc:1984][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.632 [graph_prepare.cc:1985][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.646 [graph_prepare.cc:1986][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.659 [graph_prepare.cc:1987][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.674 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.685 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.698 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.781 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.801 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.811 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrintOpPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.820 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.828 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DropOutPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.837 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.845 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.854 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.862 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of StopGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.870 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.878 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.886 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SnapshotPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.894 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.903 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.911 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.919 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.941 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.953 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.985 [graph_prepare.cc:1988][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [316] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.343.998 [graph_manager.cc:1065][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [964] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.356.368 [graph_manager.cc:1077][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12350] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.356.435 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.356.489 [graph_manager.cc:1080][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [88] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.189 [graph_manager.cc:1081][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2675] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.226 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.241 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.252 [graph_manager.cc:1082][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.284 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.300 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.314 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.385 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.402 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.434 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.450 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.490 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.508 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.526 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.551 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.567 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.579 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.588 [graph_manager.cc:2700][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [310] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.695 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.709 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.718 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.727 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.746 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.755 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CastRemovePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.763 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.772 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.780 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.788 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.796 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.804 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.813 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.821 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.829 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.839 [graph_manager.cc:2741][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [233] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.848 [graph_manager.cc:2752][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.871 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.884 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.901 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.917 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.929 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.941 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.961 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.976 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.359.990 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.000 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.019 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.030 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.049 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.062 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.071 [graph_manager.cc:2810][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [205] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.101 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.113 [graph_manager.cc:2821][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.141 [graph_manager.cc:1087][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [870] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.276 [graph_manager.cc:1088][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [122] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.315 [graph_manager.cc:1089][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.332 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.347 [graph_manager.cc:1097][EVENT]36565 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.367 [graph_manager.cc:3325][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.574 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.591 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.599 [engine_place.cc:144][EVENT]36565 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [118] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.669 [graph_manager.cc:3351][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [288] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.686 [graph_manager.cc:3364][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.749 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.766 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.920 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [141] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.360.970 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.361.020 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [38] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.361.055 [graph_manager.cc:3405][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [355] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.361.073 [graph_manager.cc:3412][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.473 [graph_manager.cc:3422][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1386] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.502 [graph_manager.cc:3428][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.621 [graph_manager.cc:3467][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [100] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.639 [graph_manager.cc:3377][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1941] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.655 [graph_manager.cc:1106][EVENT]36565 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2294] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.667 [graph_manager.cc:1115][EVENT]36565 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.689 [graph_manager.cc:1130][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.722 [graph_manager.cc:1131][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.746 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.762 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.772 [graph_manager.cc:2837][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.843 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.856 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.865 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.874 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.883 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.891 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.907 [graph_manager.cc:2864][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [118] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.919 [graph_manager.cc:2872][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.938 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.953 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.968 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.982 [compile_nodes_pass.cc:88][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.362.993 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.003 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.083 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [70] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.110 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.123 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.136 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.148 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.156 [graph_manager.cc:2927][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [221] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.169 [graph_manager.cc:2937][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.182 [graph_manager.cc:2943][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.193 [graph_manager.cc:2950][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.352 [graph_manager.cc:2958][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.382 [graph_manager.cc:1132][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [647] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.458 [graph_manager.cc:1135][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.488 [graph_manager.cc:2975][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.525 [graph_manager.cc:2981][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.539 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.550 [graph_manager.cc:2986][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.559 [graph_manager.cc:1136][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [86] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.664 [graph_manager.cc:3555][EVENT]36565 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [74] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.747 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.761 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.874 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.903 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.941 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.363.963 [graph_builder.cc:865][EVENT]36565 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [247] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:29.364.312 [logger.cc:1071] 36565 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.364.341 [task_generator.cc:804][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [77] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.364.398 [task_generator.cc:805][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [45] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.364.842 [task_generator.cc:814][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [429] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.364.856 [task_generator.cc:954][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [593] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.364.912 [task_generator.cc:967][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [33] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:29.364.930 [logger.cc:1084] 36565 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:29.365.074 [graph_manager.cc:1152][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1491] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.365.092 [graph_manager.cc:1164][EVENT]36565 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.365.136 [graph_manager.cc:1271][EVENT]36565 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [25611] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.365.155 [graph_manager.cc:1272][EVENT]36565 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.365.461 [atrace_api.c:93](tid:36565) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.365.476 [atrace_api.c:95](tid:36565) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:29.369.714 [graph_converter.cc:838][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1216] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.369.905 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [148] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.291 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [363] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.373 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.389 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [77] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.674 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [274] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.782 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [90] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.818 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.370.968 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [137] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.038 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [54] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.051 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [67] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.078 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.106 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.133 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.192 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.249 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.259 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [57] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.285 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.309 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.330 [graph_converter.cc:849][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1578] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.371.525 [graph_converter.cc:853][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [185] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.372.164 [graph_converter.cc:857][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [623] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.372.296 [graph_converter.cc:862][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [99] micro second. . TotalTime = 0.145502, [20] [parse]: 0.00146911 [symbol_resolve]: 0.012436, [1] [Cycle 1]: 0.0123563, [1] [resolve]: 0.012336 [combine_like_graphs]: 8.50006e-07 [graph_reusing]: 3.25e-06 [meta_unpack_prepare]: 0.00016417 [pre_cconv]: 6.99998e-07 [abstract_specialize]: 0.00430712 [pack_expand]: 1.941e-05 [auto_monad]: 8.629e-05 [inline]: 1.96e-06 [pre_auto_parallel]: 9.68e-06 [pipeline_split]: 3.18e-06 [optimize]: 0.122827, [35] [py_interpret_to_execute]: 4.46e-06 [rewriter_before_opt_a]: 0.00020603 [opt_a]: 0.120917, [4] [Cycle 1]: 0.0583424, [30] [expand_dump_flag]: 4.24e-06 [switch_simplify]: 2.586e-05 [a_1]: 0.00072461 [recompute_prepare]: 8.19e-06 [updatestate_depend_eliminate]: 1.1e-05 [updatestate_assign_eliminate]: 7.1e-06 [updatestate_loads_eliminate]: 6.99001e-06 [parameter_eliminate]: 4.73e-06 [a_2]: 7.641e-05 [accelerated_algorithm]: 5.27e-06 [pynative_shard]: 1.59e-06 [auto_parallel]: 3.85e-06 [parallel]: 7.99e-06 [merge_comm]: 4.1e-06 [allreduce_fusion]: 1.97e-06 [virtual_dataset]: 5.28e-06 [get_grad_eliminate_]: 4.52e-06 [virtual_output]: 4.2e-06 [merge_forward]: 8.74e-06 [cell_reuse_recompute_pass]: 1e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.183e-05 [meta_fg_expand]: 0.00695768, [1] [Cycle 1]: 0.00316185, [1] [resolve]: 0.00314298 [after_resolve]: 4.851e-05 [a_after_grad]: 0.00013251 [renormalize]: 0.0491942 [real_op_eliminate]: 4.378e-05 [auto_monad_grad]: 7.075e-05 [auto_monad_eliminator]: 0.00010433 [cse]: 0.00036439 [a_3]: 0.00031021 [Cycle 2]: 0.0526515, [30] [expand_dump_flag]: 3.86e-06 [switch_simplify]: 0.00013045 [a_1]: 0.00160115 [recompute_prepare]: 1.288e-05 [updatestate_depend_eliminate]: 1.67e-05 [updatestate_assign_eliminate]: 1.287e-05 [updatestate_loads_eliminate]: 1.223e-05 [parameter_eliminate]: 4.11e-06 [a_2]: 0.00019498 [accelerated_algorithm]: 1.842e-05 [pynative_shard]: 1.41e-06 [auto_parallel]: 5.06e-06 [parallel]: 5.21e-06 [merge_comm]: 2.57e-06 [allreduce_fusion]: 1.64e-06 [virtual_dataset]: 1.09e-05 [get_grad_eliminate_]: 1.034e-05 [virtual_output]: 1.042e-05 [merge_forward]: 1.377e-05 [cell_reuse_recompute_pass]: 4.1e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.347e-05 [meta_fg_expand]: 0.011316, [5] [Cycle 1]: 0.00031868, [1] [resolve]: 0.00030085 [Cycle 1]: 0.00030824, [1] [resolve]: 0.00029022 [Cycle 1]: 0.00163986, [1] [resolve]: 0.00162129 [Cycle 1]: 0.00030757, [1] [resolve]: 0.00028972 [Cycle 1]: 0.00030917, [1] [resolve]: 0.00029122 [after_resolve]: 7.475e-05 [a_after_grad]: 0.00019498 [renormalize]: 0.037686 [real_op_eliminate]: 5.674e-05 [auto_monad_grad]: 0.00020776 [auto_monad_eliminator]: 0.00010632 [cse]: 0.00030943 [a_3]: 0.00042153 [Cycle 3]: 0.0058885, [30] [expand_dump_flag]: 3.96001e-06 [switch_simplify]: 0.00015507 [a_1]: 0.0023168 [recompute_prepare]: 1.634e-05 [updatestate_depend_eliminate]: 3.087e-05 [updatestate_assign_eliminate]: 1.882e-05 [updatestate_loads_eliminate]: 3.625e-05 [parameter_eliminate]: 4.83e-06 [a_2]: 0.00027255 [accelerated_algorithm]: 2.46e-05 [pynative_shard]: 1.41e-06 [auto_parallel]: 4.03e-06 [parallel]: 4.18e-06 [merge_comm]: 3.78e-06 [allreduce_fusion]: 2.29001e-06 [virtual_dataset]: 1.448e-05 [get_grad_eliminate_]: 1.438e-05 [virtual_output]: 1.38e-05 [merge_forward]: 1.978e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.243e-05 [meta_fg_expand]: 4.945e-05 [after_resolve]: 1.82e-05 [a_after_grad]: 3.173e-05 [renormalize]: 0.00224833 [real_op_eliminate]: 2.107e-05 [auto_monad_grad]: 5.99999e-06 [auto_monad_eliminator]: 3.529e-05 [cse]: 0.0002041 [a_3]: 0.00012083 [Cycle 4]: 0.0016775, [30] [expand_dump_flag]: 1.31e-06 [switch_simplify]: 1.518e-05 [a_1]: 0.00075975 [recompute_prepare]: 1.649e-05 [updatestate_depend_eliminate]: 2.067e-05 [updatestate_assign_eliminate]: 1.703e-05 [updatestate_loads_eliminate]: 1.67e-05 [parameter_eliminate]: 2.3e-06 [a_2]: 0.00027258 [accelerated_algorithm]: 2.512e-05 [pynative_shard]: 1.53e-06 [auto_parallel]: 4.08e-06 [parallel]: 3.56e-06 [merge_comm]: 2.98e-06 [allreduce_fusion]: 2.02e-06 [virtual_dataset]: 1.543e-05 [get_grad_eliminate_]: 1.482e-05 [virtual_output]: 1.4e-05 [merge_forward]: 1.752e-05 [cell_reuse_recompute_pass]: 3.80001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.283e-05 [meta_fg_expand]: 1.383e-05 [after_resolve]: 1.65e-05 [a_after_grad]: 3.187e-05 [renormalize]: 6.99947e-08 [real_op_eliminate]: 1.432e-05 [auto_monad_grad]: 2.15e-06 [auto_monad_eliminator]: 3.028e-05 [cse]: 8.315e-05 [a_3]: 0.00011174 [py_interpret_to_execute_after_opt_a]: 3.96001e-06 [slice_cell_reuse_recomputed_activation]: 2.6e-06 [rewriter_after_opt_a]: 9.869e-05 [convert_after_rewriter]: 2.364e-05 [order_py_execute_after_rewriter]: 1.734e-05 [opt_b]: 0.00109643, [2] [Cycle 1]: 0.00092926, [7] [b_1]: 0.00083565 [b_2]: 5.16e-06 [updatestate_depend_eliminate]: 5.76999e-06 [updatestate_assign_eliminate]: 4.27e-06 [updatestate_loads_eliminate]: 3.95e-06 [renormalize]: 4.60001e-07 [cse]: 3.847e-05 [Cycle 2]: 0.00015781, [7] [b_1]: 9.656e-05 [b_2]: 4.25e-06 [updatestate_depend_eliminate]: 5.25e-06 [updatestate_assign_eliminate]: 3.97e-06 [updatestate_loads_eliminate]: 3.71e-06 [renormalize]: 7.0002e-08 [cse]: 1.745e-05 [cconv]: 2.143e-05 [opt_after_cconv]: 8.577e-05, [1] [Cycle 1]: 8.15e-05, [7] [c_1]: 2.44e-05 [parameter_eliminate]: 2.19e-06 [updatestate_depend_eliminate]: 4.52e-06 [updatestate_assign_eliminate]: 4.31e-06 [updatestate_loads_eliminate]: 3.51e-06 [cse]: 1.593e-05 [renormalize]: 4.30002e-07 [remove_dup_value]: 1.742e-05 [tuple_transform]: 8.459e-05, [1] [Cycle 1]: 8.088e-05, [3] [d_1]: 5.842e-05 [d_2]: 9.58e-06 [renormalize]: 1.90004e-07 [add_cache_embedding]: 1.317e-05 [add_recomputation]: 5.913e-05 [cse_after_recomputation]: 2.65e-05, [1] [Cycle 1]: 2.238e-05, [1] [cse]: 1.75e-05 [environ_conv]: 9.69999e-06 [label_micro_interleaved_index]: 2.42e-06 [label_fine_grained_interleaved_index]: 2.54e-06 [assign_add_opt]: 1.39e-06 [slice_recompute_activation]: 1.98e-06 [micro_interleaved_order_control]: 1.59e-06 [full_micro_interleaved_order_control]: 2.06001e-06 [comp_comm_scheduling]: 1.92e-06 [reorder_send_recv_between_fp_bp]: 2.14e-06 [comm_op_add_attrs]: 1.01e-06 [add_comm_op_reuse_tag]: 8.59996e-07 [overlap_opt_shard_in_pipeline]: 1.29001e-06 [grouped_pairwise_exchange_alltoall]: 1.07e-06 [overlap_recompute_and_grad_model_parallel]: 1.98e-06 [overlap_grad_matmul_and_grad_allreduce]: 7.30004e-07 [split_matmul_comm_elemetwise]: 2.15e-06 [split_layernorm_comm]: 1.85e-06 [process_send_recv_for_ge]: 7.90002e-07 [handle_group_info]: 1.27e-06 [auto_monad_reorder]: 2.223e-05 [get_jit_bprop_graph]: 3.27e-06 [eliminate_special_op_node]: 0.00050539 [validate]: 3.807e-05 [distribtued_split]: 1.15e-06 [task_emit]: 0.00340703 [execute]: 5.92e-06 Sums parse : 0.001469s : 1.13% symbol_resolve.resolve : 0.012336s : 9.52% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000164s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004307s : 3.32% pack_expand : 0.000019s : 0.01% auto_monad : 0.000086s : 0.07% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000206s : 0.16% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000327s : 0.25% optimize.opt_a.a_1 : 0.005402s : 4.17% optimize.opt_a.recompute_prepare : 0.000054s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000079s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000072s : 0.06% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000817s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000021s : 0.02% optimize.opt_a.merge_comm : 0.000013s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000046s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000042s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000101s : 0.08% optimize.opt_a.meta_fg_expand : 0.000063s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.005936s : 4.58% optimize.opt_a.after_resolve : 0.000158s : 0.12% optimize.opt_a.a_after_grad : 0.000391s : 0.30% optimize.opt_a.renormalize : 0.089129s : 68.76% optimize.opt_a.real_op_eliminate : 0.000136s : 0.10% optimize.opt_a.auto_monad_grad : 0.000287s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000276s : 0.21% optimize.opt_a.cse : 0.000961s : 0.74% optimize.opt_a.a_3 : 0.000964s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000099s : 0.08% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000932s : 0.72% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000056s : 0.04% optimize.cconv : 0.000021s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000058s : 0.05% optimize.tuple_transform.d_2 : 0.000010s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000059s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000022s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000505s : 0.39% validate : 0.000038s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003407s : 2.63% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018522 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.34% : 0.016733s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.54% : 0.001026s : 103: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.51% : 0.000094s : 69: substitution.replace_applicator 0.06% : 0.000011s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000012s : 10: substitution.switch_simplify 0.06% : 0.000012s : 4: substitution.transpose_eliminate 0.66% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.86% : 0.000160s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000076s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089113 6 92.54% : 0.082468s : 3: renormalize.infer 7.46% : 0.006645s : 3: renormalize.specialize ------[replace.] 0.001233 141 54.20% : 0.000669s : 55: replace.getattr_setattr_resolve 26.25% : 0.000324s : 56: replace.inline 3.92% : 0.000048s : 2: replace.meta_unpack_prepare 7.22% : 0.000089s : 10: replace.switch_simplify 1.59% : 0.000020s : 4: replace.transpose_eliminate 6.81% : 0.000084s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017519 141 94.76% : 0.016601s : 55: match.getattr_setattr_resolve 4.85% : 0.000850s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.07% : 0.000012s : 10: match.switch_simplify 0.07% : 0.000012s : 4: match.transpose_eliminate 0.15% : 0.000026s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006839 119 69.43% : 0.004748s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.57% : 0.002091s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027461 589 0.52% : 0.000142s : 2: opt.transform.meta_unpack_prepare 30.44% : 0.008360s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.30% : 0.000907s : 94: opt.transform.opt_b 65.38% : 0.017953s : 14: opt.transform.opt_resolve 0.23% : 0.000064s : 8: opt.transform.opt_trans_graph 0.06% : 0.000015s : 3: opt.transform.special_op_eliminate . TotalTime = 0.142677, [20] [parse]: 0.00126938 [symbol_resolve]: 0.0125173, [1] [Cycle 1]: 0.0124542, [1] [resolve]: 0.0124361 [combine_like_graphs]: 7.59996e-07 [graph_reusing]: 3.06e-06 [meta_unpack_prepare]: 0.00012674 [pre_cconv]: 6.19999e-07 [abstract_specialize]: 0.00403891 [pack_expand]: 1.364e-05 [auto_monad]: 6.698e-05 [inline]: 1.39e-06 [pre_auto_parallel]: 6.4e-06 [pipeline_split]: 1.8e-06 [optimize]: 0.118538, [35] [py_interpret_to_execute]: 4.22e-06 [rewriter_before_opt_a]: 0.00020836 [opt_a]: 0.116702, [4] [Cycle 1]: 0.0574866, [30] [expand_dump_flag]: 3.47001e-06 [switch_simplify]: 2.553e-05 [a_1]: 0.00038369 [recompute_prepare]: 8.76e-06 [updatestate_depend_eliminate]: 9.45e-06 [updatestate_assign_eliminate]: 6.22e-06 [updatestate_loads_eliminate]: 5.89999e-06 [parameter_eliminate]: 4.14e-06 [a_2]: 7.839e-05 [accelerated_algorithm]: 5.53e-06 [pynative_shard]: 8.49999e-07 [auto_parallel]: 3.31e-06 [parallel]: 5.57e-06 [merge_comm]: 2.66e-06 [allreduce_fusion]: 1.60999e-06 [virtual_dataset]: 5.43e-06 [get_grad_eliminate_]: 4.56e-06 [virtual_output]: 4.01e-06 [merge_forward]: 7.32e-06 [cell_reuse_recompute_pass]: 3.80001e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.174e-05 [meta_fg_expand]: 0.00700519, [1] [Cycle 1]: 0.00328105, [1] [resolve]: 0.00326293 [after_resolve]: 4.748e-05 [a_after_grad]: 0.00011054 [renormalize]: 0.0487431 [real_op_eliminate]: 3.942e-05 [auto_monad_grad]: 6.913e-05 [auto_monad_eliminator]: 7.476e-05 [cse]: 0.00033147 [a_3]: 0.00030407 [Cycle 2]: 0.0510466, [30] [expand_dump_flag]: 3.32e-06 [switch_simplify]: 0.00013202 [a_1]: 0.00081278 [recompute_prepare]: 1.407e-05 [updatestate_depend_eliminate]: 1.645e-05 [updatestate_assign_eliminate]: 1.256e-05 [updatestate_loads_eliminate]: 1.248e-05 [parameter_eliminate]: 3.9e-06 [a_2]: 0.00019835 [accelerated_algorithm]: 1.758e-05 [pynative_shard]: 1.08e-06 [auto_parallel]: 4.09e-06 [parallel]: 4.62e-06 [merge_comm]: 2.52e-06 [allreduce_fusion]: 1.45e-06 [virtual_dataset]: 1.045e-05 [get_grad_eliminate_]: 9.59e-06 [virtual_output]: 9.6e-06 [merge_forward]: 1.415e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.352e-05 [meta_fg_expand]: 0.0113674, [5] [Cycle 1]: 0.00033207, [1] [resolve]: 0.00031433 [Cycle 1]: 0.00031733, [1] [resolve]: 0.00029988 [Cycle 1]: 0.00170527, [1] [resolve]: 0.00168707 [Cycle 1]: 0.00031799, [1] [resolve]: 0.0003004 [Cycle 1]: 0.00031678, [1] [resolve]: 0.00029936 [after_resolve]: 7.528e-05 [a_after_grad]: 0.00016519 [renormalize]: 0.0368923 [real_op_eliminate]: 5.11e-05 [auto_monad_grad]: 0.00019597 [auto_monad_eliminator]: 0.00010096 [cse]: 0.00029644 [a_3]: 0.00040875 [Cycle 3]: 0.00467528, [30] [expand_dump_flag]: 3.88001e-06 [switch_simplify]: 0.00010981 [a_1]: 0.00129704 [recompute_prepare]: 1.77e-05 [updatestate_depend_eliminate]: 2.643e-05 [updatestate_assign_eliminate]: 1.86e-05 [updatestate_loads_eliminate]: 1.771e-05 [parameter_eliminate]: 4.36e-06 [a_2]: 0.00027583 [accelerated_algorithm]: 2.456e-05 [pynative_shard]: 1.52e-06 [auto_parallel]: 4.04e-06 [parallel]: 4.25e-06 [merge_comm]: 3.57e-06 [allreduce_fusion]: 2.44e-06 [virtual_dataset]: 1.399e-05 [get_grad_eliminate_]: 1.283e-05 [virtual_output]: 1.249e-05 [merge_forward]: 1.958e-05 [cell_reuse_recompute_pass]: 5.30003e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.323e-05 [meta_fg_expand]: 2.889e-05 [after_resolve]: 1.577e-05 [a_after_grad]: 2.063e-05 [renormalize]: 0.00216942 [real_op_eliminate]: 2.025e-05 [auto_monad_grad]: 5.65001e-06 [auto_monad_eliminator]: 3.466e-05 [cse]: 0.00019558 [a_3]: 0.0001212 [Cycle 4]: 0.00120142, [30] [expand_dump_flag]: 1.24999e-06 [switch_simplify]: 1.488e-05 [a_1]: 0.00028418 [recompute_prepare]: 1.522e-05 [updatestate_depend_eliminate]: 1.971e-05 [updatestate_assign_eliminate]: 3.44e-05 [updatestate_loads_eliminate]: 1.622e-05 [parameter_eliminate]: 2.34e-06 [a_2]: 0.00027534 [accelerated_algorithm]: 2.499e-05 [pynative_shard]: 1.55e-06 [auto_parallel]: 3.88e-06 [parallel]: 3.64e-06 [merge_comm]: 3.14e-06 [allreduce_fusion]: 2.05e-06 [virtual_dataset]: 1.409e-05 [get_grad_eliminate_]: 1.366e-05 [virtual_output]: 1.269e-05 [merge_forward]: 1.763e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.372e-05 [meta_fg_expand]: 1.311e-05 [after_resolve]: 1.578e-05 [a_after_grad]: 2.043e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.34e-05 [auto_monad_grad]: 2.14999e-06 [auto_monad_eliminator]: 3.056e-05 [cse]: 8.131e-05 [a_3]: 0.00011215 [py_interpret_to_execute_after_opt_a]: 4.2e-06 [slice_cell_reuse_recomputed_activation]: 1.51e-06 [rewriter_after_opt_a]: 9.608e-05 [convert_after_rewriter]: 2.21e-05 [order_py_execute_after_rewriter]: 1.616e-05 [opt_b]: 0.00109513, [2] [Cycle 1]: 0.00092685, [7] [b_1]: 0.00083291 [b_2]: 5.67e-06 [updatestate_depend_eliminate]: 5.78e-06 [updatestate_assign_eliminate]: 4.37e-06 [updatestate_loads_eliminate]: 4.28e-06 [renormalize]: 4.60001e-07 [cse]: 3.757e-05 [Cycle 2]: 0.00015907, [7] [b_1]: 9.795e-05 [b_2]: 4.37e-06 [updatestate_depend_eliminate]: 4.56e-06 [updatestate_assign_eliminate]: 3.91e-06 [updatestate_loads_eliminate]: 3.55e-06 [renormalize]: 6.99947e-08 [cse]: 1.774e-05 [cconv]: 1.776e-05 [opt_after_cconv]: 7.212e-05, [1] [Cycle 1]: 6.76e-05, [7] [c_1]: 8.67e-06 [parameter_eliminate]: 2.21e-06 [updatestate_depend_eliminate]: 4.34e-06 [updatestate_assign_eliminate]: 3.58e-06 [updatestate_loads_eliminate]: 3.56e-06 [cse]: 1.722e-05 [renormalize]: 4.30002e-07 [remove_dup_value]: 1.321e-05 [tuple_transform]: 6.668e-05, [1] [Cycle 1]: 6.304e-05, [3] [d_1]: 4.041e-05 [d_2]: 9.42e-06 [renormalize]: 2.59999e-07 [add_cache_embedding]: 1.078e-05 [add_recomputation]: 5.022e-05 [cse_after_recomputation]: 2.557e-05, [1] [Cycle 1]: 2.181e-05, [1] [cse]: 1.693e-05 [environ_conv]: 8.88e-06 [label_micro_interleaved_index]: 1.7e-06 [label_fine_grained_interleaved_index]: 1.32e-06 [assign_add_opt]: 1e-06 [slice_recompute_activation]: 1.37e-06 [micro_interleaved_order_control]: 1.08e-06 [full_micro_interleaved_order_control]: 9.5e-07 [comp_comm_scheduling]: 1.36e-06 [reorder_send_recv_between_fp_bp]: 1.50999e-06 [comm_op_add_attrs]: 7.50006e-07 [add_comm_op_reuse_tag]: 1e-06 [overlap_opt_shard_in_pipeline]: 6.30003e-07 [grouped_pairwise_exchange_alltoall]: 5.80003e-07 [overlap_recompute_and_grad_model_parallel]: 1.07e-06 [overlap_grad_matmul_and_grad_allreduce]: 3.69997e-07 [split_matmul_comm_elemetwise]: 1.49001e-06 [split_layernorm_comm]: 1.4e-06 [process_send_recv_for_ge]: 7.2e-07 [handle_group_info]: 5.4e-07 [auto_monad_reorder]: 1.858e-05 [get_jit_bprop_graph]: 4.49996e-07 [eliminate_special_op_node]: 0.00050479 [validate]: 3.677e-05 [distribtued_split]: 1.17e-06 [task_emit]: 0.00534678 [execute]: 4.95e-06 Sums parse : 0.001269s : 1.00% symbol_resolve.resolve : 0.012436s : 9.79% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000127s : 0.10% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004039s : 3.18% pack_expand : 0.000014s : 0.01% auto_monad : 0.000067s : 0.05% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000006s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000208s : 0.16% optimize.opt_a.expand_dump_flag : 0.000012s : 0.01% optimize.opt_a.switch_simplify : 0.000282s : 0.22% optimize.opt_a.a_1 : 0.002778s : 2.19% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000072s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000072s : 0.06% optimize.opt_a.updatestate_loads_eliminate : 0.000052s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000828s : 0.65% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000015s : 0.01% optimize.opt_a.parallel : 0.000018s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000039s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000042s : 0.03% optimize.opt_a.meta_fg_expand.resolve : 0.006164s : 4.85% optimize.opt_a.after_resolve : 0.000154s : 0.12% optimize.opt_a.a_after_grad : 0.000317s : 0.25% optimize.opt_a.renormalize : 0.087805s : 69.12% optimize.opt_a.real_op_eliminate : 0.000124s : 0.10% optimize.opt_a.auto_monad_grad : 0.000273s : 0.21% optimize.opt_a.auto_monad_eliminator : 0.000241s : 0.19% optimize.opt_a.cse : 0.000905s : 0.71% optimize.opt_a.a_3 : 0.000946s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000096s : 0.08% optimize.convert_after_rewriter : 0.000022s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000931s : 0.73% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000055s : 0.04% optimize.cconv : 0.000018s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000040s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000011s : 0.01% optimize.add_recomputation : 0.000050s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000001s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000001s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000000s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000001s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000019s : 0.01% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000505s : 0.40% validate : 0.000037s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005347s : 4.21% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018826 880 0.02% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.38% : 0.017014s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.93% : 0.001116s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000029s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.02% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.47% : 0.000089s : 63: substitution.replace_applicator 0.05% : 0.000009s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.06% : 0.000011s : 4: substitution.transpose_eliminate 0.60% : 0.000112s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000068s : 60: substitution.tuple_list_get_item_depend_reorder 0.82% : 0.000154s : 112: substitution.tuple_list_get_item_eliminator 0.35% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.087791 6 92.81% : 0.081477s : 3: renormalize.infer 7.19% : 0.006314s : 3: renormalize.specialize ------[replace.] 0.001262 141 52.96% : 0.000668s : 55: replace.getattr_setattr_resolve 28.63% : 0.000361s : 56: replace.inline 3.54% : 0.000045s : 2: replace.meta_unpack_prepare 7.08% : 0.000089s : 10: replace.switch_simplify 1.35% : 0.000017s : 4: replace.transpose_eliminate 6.45% : 0.000081s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017893 141 94.36% : 0.016884s : 55: match.getattr_setattr_resolve 5.28% : 0.000944s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.05% : 0.000010s : 10: match.switch_simplify 0.06% : 0.000011s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006634 118 69.33% : 0.004600s : 52: func_graph_cloner_run.FuncGraphClonerGraph 0.87% : 0.000058s : 3: func_graph_cloner_run.FuncGraphClonerNode 29.80% : 0.001977s : 63: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.025022 259 7.29% : 0.001823s : 104: opt.transform.opt_a 3.59% : 0.000899s : 92: opt.transform.opt_b 72.91% : 0.018244s : 14: opt.transform.opt_resolve 0.44% : 0.000111s : 1: opt.transforms.meta_unpack_prepare 15.46% : 0.003868s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000048s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:29.760.344 [graph_var_manager.cc:1424][EVENT]36563 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:29.760.429 [graph_manager.cc:1248][EVENT]36563 PreRun:PreRun start: graph node size 3, session id 4, graph id 3, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.761.345 [atrace_api.c:28](tid:36563) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.761.414 [trace_rb_log.c:84](tid:36563) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.761.428 [atrace_api.c:32](tid:36563) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:29.761.441 [client_manager.cpp:157][SetProfilingCallback][tid:36563] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:29.762.278 [parallel_partitioner.cc:165][EVENT]36563 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.762.316 [parallel_partitioner.cc:178][EVENT]36563 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [12] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.762.363 [graph_prepare.cc:1378][EVENT]36563 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.762.990 [graph_manager.cc:1050][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [643] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.017 [graph_manager.cc:1052][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.143 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.173 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.239 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [42] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.253 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.298 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.311 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.328 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.428 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.449 [graph_manager.cc:1054][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [419] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.763.665 [graph_manager.cc:1055][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [203] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.628 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.654 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.665 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.676 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of InferShapePass is [309] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.688 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.697 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.706 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.715 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [17] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.764.724 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of InferValuePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.136 [graph_manager.cc:1056][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2452] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.201 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of CondRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.219 [graph_prepare.cc:1982][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.616 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.637 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.655 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of MergePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.665 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of InferShapePass is [218] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.674 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.683 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.692 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.700 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.708 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of InferValuePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.733 [graph_prepare.cc:1983][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [500] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.757 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.769 [graph_prepare.cc:1984][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.783 [graph_prepare.cc:1985][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.803 [graph_prepare.cc:1986][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.815 [graph_prepare.cc:1987][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.830 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.842 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.857 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.940 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.955 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of CondPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.964 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrintOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.972 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.981 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.989 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.766.998 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.012 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.021 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.029 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.037 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.045 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SnapshotPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.054 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.062 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.070 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.078 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of IdentityPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.100 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.113 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.145 [graph_prepare.cc:1988][EVENT]36563 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [320] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.767.157 [graph_manager.cc:1065][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [989] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.779.401 [graph_manager.cc:1077][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12223] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.779.468 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.779.522 [graph_manager.cc:1080][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [88] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.209 [graph_manager.cc:1081][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2670] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.246 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.262 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.273 [graph_manager.cc:1082][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.304 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.320 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.345 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.418 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.435 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.466 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.480 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.520 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.538 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.555 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.581 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.595 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.607 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.615 [graph_manager.cc:2700][EVENT]36563 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [316] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.720 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.734 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of AddNPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.744 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.752 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.761 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.769 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of CastRemovePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.778 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.786 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.794 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.802 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.817 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.826 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.834 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.842 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.850 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.860 [graph_manager.cc:2741][EVENT]36563 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [226] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.869 [graph_manager.cc:2752][EVENT]36563 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.892 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.904 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.920 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.936 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.948 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.961 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.979 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.782.996 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.010 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.020 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.032 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.043 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.061 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.074 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.084 [graph_manager.cc:2810][EVENT]36563 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [197] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.120 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of IdentityPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.132 [graph_manager.cc:2821][EVENT]36563 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.160 [graph_manager.cc:1087][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [868] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.294 [graph_manager.cc:1088][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [121] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.331 [graph_manager.cc:1089][EVENT]36563 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.348 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.363 [graph_manager.cc:1097][EVENT]36563 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.385 [graph_manager.cc:3325][EVENT]36563 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.596 [engine_place.cc:144][EVENT]36563 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.612 [engine_place.cc:144][EVENT]36563 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.621 [engine_place.cc:144][EVENT]36563 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [120] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.691 [graph_manager.cc:3351][EVENT]36563 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [292] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.709 [graph_manager.cc:3364][EVENT]36563 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.776 [engine_partitioner.cc:1139][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.793 [engine_partitioner.cc:1142][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.948 [engine_partitioner.cc:1148][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [146] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.783.990 [engine_partitioner.cc:1155][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.784.039 [engine_partitioner.cc:1164][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [39] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.784.074 [graph_manager.cc:3405][EVENT]36563 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [352] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.784.093 [graph_manager.cc:3412][EVENT]36563 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.499 [graph_manager.cc:3422][EVENT]36563 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1392] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.535 [graph_manager.cc:3428][EVENT]36563 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.659 [graph_manager.cc:3467][EVENT]36563 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [103] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.676 [graph_manager.cc:3377][EVENT]36563 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1956] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.693 [graph_manager.cc:1106][EVENT]36563 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2315] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.706 [graph_manager.cc:1115][EVENT]36563 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.729 [graph_manager.cc:1130][EVENT]36563 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.762 [graph_manager.cc:1131][EVENT]36563 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.785 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.801 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.811 [graph_manager.cc:2837][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.882 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.895 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.904 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of CondRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.913 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.921 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.930 [base_pass.cc:339][EVENT]36563 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.940 [graph_manager.cc:2864][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [112] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.952 [graph_manager.cc:2872][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.973 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.785.988 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.003 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.023 [compile_nodes_pass.cc:88][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.034 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.044 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.122 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [68] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.149 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.162 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.175 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.188 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.197 [graph_manager.cc:2927][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [226] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.209 [graph_manager.cc:2937][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.222 [graph_manager.cc:2943][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.234 [graph_manager.cc:2950][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.390 [graph_manager.cc:2958][EVENT]36563 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.421 [graph_manager.cc:1132][EVENT]36563 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [645] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.497 [graph_manager.cc:1135][EVENT]36563 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [64] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.535 [graph_manager.cc:2975][EVENT]36563 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.566 [graph_manager.cc:2981][EVENT]36563 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.580 [pass_manager.cc:82][EVENT]36563 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.590 [graph_manager.cc:2986][EVENT]36563 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.599 [graph_manager.cc:1136][EVENT]36563 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [85] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.701 [graph_manager.cc:3555][EVENT]36563 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [71] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.787 [engine_partitioner.cc:1139][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.803 [engine_partitioner.cc:1142][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.924 [engine_partitioner.cc:1148][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [111] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.952 [engine_partitioner.cc:1155][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.786.991 [engine_partitioner.cc:1164][EVENT]36563 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.787.013 [graph_builder.cc:865][EVENT]36563 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [255] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:29.787.490 [logger.cc:1071] 36563 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.787.521 [task_generator.cc:804][EVENT]36563 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [158] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.787.581 [task_generator.cc:805][EVENT]36563 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [48] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.032 [task_generator.cc:814][EVENT]36563 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [437] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.046 [task_generator.cc:954][EVENT]36563 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [684] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.104 [task_generator.cc:967][EVENT]36563 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [35] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:29.788.121 [logger.cc:1084] 36563 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.269 [graph_manager.cc:1152][EVENT]36563 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1645] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.286 [graph_manager.cc:1164][EVENT]36563 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.317 [graph_manager.cc:1271][EVENT]36563 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [26131] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.788.329 [graph_manager.cc:1272][EVENT]36563 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.788.630 [atrace_api.c:93](tid:36563) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:29.788.648 [atrace_api.c:95](tid:36563) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:29.792.907 [graph_converter.cc:838][EVENT]36563 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1181] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.793.100 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of ZeroCopy is [150] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.793.502 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of CEM is [353] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.793.588 [copy_flow_launch_fuse.cc:395][EVENT]36563 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [64] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.793.611 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [88] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.793.895 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [273] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.793.999 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.034 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.186 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of CEM is [139] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.254 [copy_flow_launch_fuse.cc:395][EVENT]36563 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [53] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.266 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [65] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.294 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.320 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.347 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.407 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of CEM is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.465 [copy_flow_launch_fuse.cc:395][EVENT]36563 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.475 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [58] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.501 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.525 [base_optimizer.cc:70][EVENT]36563 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.545 [graph_converter.cc:849][EVENT]36563 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1600] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.794.736 [graph_converter.cc:853][EVENT]36563 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [181] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.795.385 [graph_converter.cc:857][EVENT]36563 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [635] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:29.795.508 [graph_converter.cc:862][EVENT]36563 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [102] micro second. . TotalTime = 0.143882, [20] [parse]: 0.00146964 [symbol_resolve]: 0.0124356, [1] [Cycle 1]: 0.0123564, [1] [resolve]: 0.0123356 [combine_like_graphs]: 7.90002e-07 [graph_reusing]: 3.31e-06 [meta_unpack_prepare]: 0.00016476 [pre_cconv]: 6.49998e-07 [abstract_specialize]: 0.00429036 [pack_expand]: 1.702e-05 [auto_monad]: 8.373e-05 [inline]: 1.43e-06 [pre_auto_parallel]: 1.109e-05 [pipeline_split]: 3.03e-06 [optimize]: 0.121245, [35] [py_interpret_to_execute]: 4.47e-06 [rewriter_before_opt_a]: 0.00019202 [opt_a]: 0.119342, [4] [Cycle 1]: 0.0577791, [30] [expand_dump_flag]: 4.07e-06 [switch_simplify]: 2.65e-05 [a_1]: 0.0007401 [recompute_prepare]: 8.34e-06 [updatestate_depend_eliminate]: 1.069e-05 [updatestate_assign_eliminate]: 6.93e-06 [updatestate_loads_eliminate]: 7.56e-06 [parameter_eliminate]: 5.09e-06 [a_2]: 7.427e-05 [accelerated_algorithm]: 5.23e-06 [pynative_shard]: 1.68e-06 [auto_parallel]: 3.54e-06 [parallel]: 8.66e-06 [merge_comm]: 4.09e-06 [allreduce_fusion]: 1.99e-06 [virtual_dataset]: 5.53e-06 [get_grad_eliminate_]: 5.17e-06 [virtual_output]: 4.36e-06 [merge_forward]: 8.58e-06 [cell_reuse_recompute_pass]: 1.03001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.137e-05 [meta_fg_expand]: 0.0069524, [1] [Cycle 1]: 0.00317498, [1] [resolve]: 0.00315572 [after_resolve]: 4.889e-05 [a_after_grad]: 0.00012829 [renormalize]: 0.048663 [real_op_eliminate]: 4.356e-05 [auto_monad_grad]: 6.976e-05 [auto_monad_eliminator]: 7.82e-05 [cse]: 0.00036153 [a_3]: 0.00030308 [Cycle 2]: 0.0516901, [30] [expand_dump_flag]: 3.43e-06 [switch_simplify]: 0.00012846 [a_1]: 0.00165268 [recompute_prepare]: 1.339e-05 [updatestate_depend_eliminate]: 1.676e-05 [updatestate_assign_eliminate]: 1.303e-05 [updatestate_loads_eliminate]: 1.249e-05 [parameter_eliminate]: 4.21e-06 [a_2]: 0.00019428 [accelerated_algorithm]: 1.906e-05 [pynative_shard]: 1.60999e-06 [auto_parallel]: 4.67e-06 [parallel]: 4.78e-06 [merge_comm]: 2.45999e-06 [allreduce_fusion]: 1.49e-06 [virtual_dataset]: 1.103e-05 [get_grad_eliminate_]: 1.038e-05 [virtual_output]: 1.02e-05 [merge_forward]: 1.402e-05 [cell_reuse_recompute_pass]: 5.19998e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.341e-05 [meta_fg_expand]: 0.0112665, [5] [Cycle 1]: 0.00031751, [1] [resolve]: 0.00029934 [Cycle 1]: 0.00031226, [1] [resolve]: 0.00029439 [Cycle 1]: 0.00164813, [1] [resolve]: 0.00162937 [Cycle 1]: 0.00030969, [1] [resolve]: 0.00029121 [Cycle 1]: 0.00030812, [1] [resolve]: 0.00029012 [after_resolve]: 7.475e-05 [a_after_grad]: 0.00019673 [renormalize]: 0.0367507 [real_op_eliminate]: 5.593e-05 [auto_monad_grad]: 0.00020467 [auto_monad_eliminator]: 0.00010105 [cse]: 0.00029961 [a_3]: 0.00040597 [Cycle 3]: 0.00584506, [30] [expand_dump_flag]: 4.14e-06 [switch_simplify]: 0.0001532 [a_1]: 0.00235463 [recompute_prepare]: 1.588e-05 [updatestate_depend_eliminate]: 3.059e-05 [updatestate_assign_eliminate]: 1.913e-05 [updatestate_loads_eliminate]: 1.837e-05 [parameter_eliminate]: 4.36e-06 [a_2]: 0.0002696 [accelerated_algorithm]: 2.491e-05 [pynative_shard]: 1.24999e-06 [auto_parallel]: 4.21e-06 [parallel]: 4.08e-06 [merge_comm]: 3.76e-06 [allreduce_fusion]: 2.61e-06 [virtual_dataset]: 1.465e-05 [get_grad_eliminate_]: 1.398e-05 [virtual_output]: 1.361e-05 [merge_forward]: 1.987e-05 [cell_reuse_recompute_pass]: 4.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.226e-05 [meta_fg_expand]: 3.006e-05 [after_resolve]: 1.714e-05 [a_after_grad]: 3.123e-05 [renormalize]: 0.00221871 [real_op_eliminate]: 2.135e-05 [auto_monad_grad]: 5.54e-06 [auto_monad_eliminator]: 3.442e-05 [cse]: 0.00019691 [a_3]: 0.00012114 [Cycle 4]: 0.00165061, [30] [expand_dump_flag]: 1.27e-06 [switch_simplify]: 1.481e-05 [a_1]: 0.00073878 [recompute_prepare]: 1.537e-05 [updatestate_depend_eliminate]: 1.998e-05 [updatestate_assign_eliminate]: 1.675e-05 [updatestate_loads_eliminate]: 1.626e-05 [parameter_eliminate]: 2.34e-06 [a_2]: 0.00027117 [accelerated_algorithm]: 2.512e-05 [pynative_shard]: 1.55e-06 [auto_parallel]: 3.79e-06 [parallel]: 3.71e-06 [merge_comm]: 3.2e-06 [allreduce_fusion]: 2.05e-06 [virtual_dataset]: 1.52e-05 [get_grad_eliminate_]: 1.487e-05 [virtual_output]: 1.402e-05 [merge_forward]: 1.721e-05 [cell_reuse_recompute_pass]: 3.99996e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.269e-05 [meta_fg_expand]: 1.328e-05 [after_resolve]: 1.618e-05 [a_after_grad]: 3.181e-05 [renormalize]: 6.99947e-08 [real_op_eliminate]: 1.455e-05 [auto_monad_grad]: 2.22e-06 [auto_monad_eliminator]: 3.03e-05 [cse]: 8.417e-05 [a_3]: 0.00011263 [py_interpret_to_execute_after_opt_a]: 4.18e-06 [slice_cell_reuse_recomputed_activation]: 2.4e-06 [rewriter_after_opt_a]: 9.732e-05 [convert_after_rewriter]: 2.326e-05 [order_py_execute_after_rewriter]: 1.713e-05 [opt_b]: 0.00110009, [2] [Cycle 1]: 0.00093158, [7] [b_1]: 0.00083781 [b_2]: 5.11e-06 [updatestate_depend_eliminate]: 5.93e-06 [updatestate_assign_eliminate]: 4.21e-06 [updatestate_loads_eliminate]: 4.14001e-06 [renormalize]: 4.49996e-07 [cse]: 3.712e-05 [Cycle 2]: 0.0001587, [7] [b_1]: 9.65e-05 [b_2]: 4.03e-06 [updatestate_depend_eliminate]: 5.11e-06 [updatestate_assign_eliminate]: 3.89e-06 [updatestate_loads_eliminate]: 3.65e-06 [renormalize]: 7.0002e-08 [cse]: 1.83e-05 [cconv]: 2.183e-05 [opt_after_cconv]: 8.674e-05, [1] [Cycle 1]: 8.25e-05, [7] [c_1]: 2.443e-05 [parameter_eliminate]: 2.08e-06 [updatestate_depend_eliminate]: 4.65e-06 [updatestate_assign_eliminate]: 4.21e-06 [updatestate_loads_eliminate]: 3.69e-06 [cse]: 1.608e-05 [renormalize]: 4.39999e-07 [remove_dup_value]: 1.788e-05 [tuple_transform]: 8.148e-05, [1] [Cycle 1]: 7.799e-05, [3] [d_1]: 5.706e-05 [d_2]: 8.8e-06 [renormalize]: 2.40005e-07 [add_cache_embedding]: 1.282e-05 [add_recomputation]: 7.03e-05 [cse_after_recomputation]: 2.774e-05, [1] [Cycle 1]: 2.336e-05, [1] [cse]: 1.865e-05 [environ_conv]: 9.47e-06 [label_micro_interleaved_index]: 2.4e-06 [label_fine_grained_interleaved_index]: 2.17e-06 [assign_add_opt]: 1.73e-06 [slice_recompute_activation]: 2.01e-06 [micro_interleaved_order_control]: 1.72e-06 [full_micro_interleaved_order_control]: 1.65e-06 [comp_comm_scheduling]: 2.03e-06 [reorder_send_recv_between_fp_bp]: 2.16e-06 [comm_op_add_attrs]: 9.70002e-07 [add_comm_op_reuse_tag]: 8.2e-07 [overlap_opt_shard_in_pipeline]: 1e-06 [grouped_pairwise_exchange_alltoall]: 1.1e-06 [overlap_recompute_and_grad_model_parallel]: 1.87e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.99998e-07 [split_matmul_comm_elemetwise]: 2.12e-06 [split_layernorm_comm]: 1.69e-06 [process_send_recv_for_ge]: 8.79998e-07 [handle_group_info]: 9e-07 [auto_monad_reorder]: 2.203e-05 [get_jit_bprop_graph]: 3.43e-06 [eliminate_special_op_node]: 0.00049143 [validate]: 3.794e-05 [distribtued_split]: 1.4e-06 [task_emit]: 0.00340031 [execute]: 5.85e-06 Sums parse : 0.001470s : 1.15% symbol_resolve.resolve : 0.012336s : 9.63% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000165s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004290s : 3.35% pack_expand : 0.000017s : 0.01% auto_monad : 0.000084s : 0.07% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000011s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000192s : 0.15% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000323s : 0.25% optimize.opt_a.a_1 : 0.005486s : 4.28% optimize.opt_a.recompute_prepare : 0.000053s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000078s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000055s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000809s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000074s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000016s : 0.01% optimize.opt_a.parallel : 0.000021s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000046s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000042s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000100s : 0.08% optimize.opt_a.meta_fg_expand : 0.000043s : 0.03% optimize.opt_a.meta_fg_expand.resolve : 0.005960s : 4.65% optimize.opt_a.after_resolve : 0.000157s : 0.12% optimize.opt_a.a_after_grad : 0.000388s : 0.30% optimize.opt_a.renormalize : 0.087632s : 68.43% optimize.opt_a.real_op_eliminate : 0.000135s : 0.11% optimize.opt_a.auto_monad_grad : 0.000282s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000244s : 0.19% optimize.opt_a.cse : 0.000942s : 0.74% optimize.opt_a.a_3 : 0.000943s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000097s : 0.08% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000934s : 0.73% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000055s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000057s : 0.04% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000070s : 0.05% optimize.cse_after_recomputation.cse : 0.000019s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000022s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000491s : 0.38% validate : 0.000038s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003400s : 2.66% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018640 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000023s : 49: substitution.float_tuple_getitem_switch 89.83% : 0.016745s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.02% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 6.09% : 0.001135s : 103: substitution.inline 0.05% : 0.000009s : 23: substitution.less_batch_normalization 0.20% : 0.000038s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.50% : 0.000092s : 69: substitution.replace_applicator 0.06% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000011s : 10: substitution.switch_simplify 0.05% : 0.000010s : 4: substitution.transpose_eliminate 0.65% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000057s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.86% : 0.000160s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000076s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.087617 6 92.62% : 0.081150s : 3: renormalize.infer 7.38% : 0.006467s : 3: renormalize.specialize ------[replace.] 0.001241 141 53.67% : 0.000666s : 55: replace.getattr_setattr_resolve 27.21% : 0.000338s : 56: replace.inline 3.73% : 0.000046s : 2: replace.meta_unpack_prepare 7.19% : 0.000089s : 10: replace.switch_simplify 1.52% : 0.000019s : 4: replace.transpose_eliminate 6.68% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017643 141 94.17% : 0.016614s : 55: match.getattr_setattr_resolve 5.46% : 0.000962s : 56: match.inline 0.10% : 0.000018s : 2: match.meta_unpack_prepare 0.06% : 0.000011s : 10: match.switch_simplify 0.06% : 0.000010s : 4: match.transpose_eliminate 0.16% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006817 118 69.33% : 0.004726s : 52: func_graph_cloner_run.FuncGraphClonerGraph 0.80% : 0.000054s : 3: func_graph_cloner_run.FuncGraphClonerNode 29.87% : 0.002036s : 63: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027520 589 0.52% : 0.000142s : 2: opt.transform.meta_unpack_prepare 30.54% : 0.008405s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.30% : 0.000908s : 94: opt.transform.opt_b 65.29% : 0.017969s : 14: opt.transform.opt_resolve 0.22% : 0.000061s : 8: opt.transform.opt_trans_graph 0.05% : 0.000015s : 3: opt.transform.special_op_eliminate . TotalTime = 0.144531, [20] [parse]: 0.00129455 [symbol_resolve]: 0.012347, [1] [Cycle 1]: 0.0122823, [1] [resolve]: 0.0122653 [combine_like_graphs]: 8.29998e-07 [graph_reusing]: 2.86e-06 [meta_unpack_prepare]: 0.0001276 [pre_cconv]: 3.99996e-07 [abstract_specialize]: 0.00404749 [pack_expand]: 1.456e-05 [auto_monad]: 6.389e-05 [inline]: 1.4e-06 [pre_auto_parallel]: 6.99e-06 [pipeline_split]: 1.6e-06 [optimize]: 0.120662, [35] [py_interpret_to_execute]: 4.26e-06 [rewriter_before_opt_a]: 0.00018828 [opt_a]: 0.11885, [4] [Cycle 1]: 0.058278, [30] [expand_dump_flag]: 3.2e-06 [switch_simplify]: 2.559e-05 [a_1]: 0.00042699 [recompute_prepare]: 8.73e-06 [updatestate_depend_eliminate]: 9.92e-06 [updatestate_assign_eliminate]: 6.39e-06 [updatestate_loads_eliminate]: 6.04001e-06 [parameter_eliminate]: 4.35e-06 [a_2]: 7.795e-05 [accelerated_algorithm]: 5.45e-06 [pynative_shard]: 1.09e-06 [auto_parallel]: 3.47e-06 [parallel]: 6.34e-06 [merge_comm]: 2.99e-06 [allreduce_fusion]: 1.66e-06 [virtual_dataset]: 5.26e-06 [get_grad_eliminate_]: 4.44e-06 [virtual_output]: 3.99999e-06 [merge_forward]: 7.35001e-06 [cell_reuse_recompute_pass]: 4.19997e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.203e-05 [meta_fg_expand]: 0.0070043, [1] [Cycle 1]: 0.0032143, [1] [resolve]: 0.00319493 [after_resolve]: 4.857e-05 [a_after_grad]: 0.00011124 [renormalize]: 0.0494661 [real_op_eliminate]: 4.098e-05 [auto_monad_grad]: 6.445e-05 [auto_monad_eliminator]: 7.64e-05 [cse]: 0.00034698 [a_3]: 0.00031008 [Cycle 2]: 0.0523383, [30] [expand_dump_flag]: 3.67e-06 [switch_simplify]: 0.00010209 [a_1]: 0.0007621 [recompute_prepare]: 1.384e-05 [updatestate_depend_eliminate]: 1.626e-05 [updatestate_assign_eliminate]: 1.26e-05 [updatestate_loads_eliminate]: 1.223e-05 [parameter_eliminate]: 3.86e-06 [a_2]: 0.00019782 [accelerated_algorithm]: 1.774e-05 [pynative_shard]: 1.14e-06 [auto_parallel]: 3.89e-06 [parallel]: 4.95e-06 [merge_comm]: 2.31e-06 [allreduce_fusion]: 1.84e-06 [virtual_dataset]: 1.046e-05 [get_grad_eliminate_]: 9.54e-06 [virtual_output]: 9.54e-06 [merge_forward]: 1.384e-05 [cell_reuse_recompute_pass]: 4.50003e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.37e-05 [meta_fg_expand]: 0.0113414, [5] [Cycle 1]: 0.00032262, [1] [resolve]: 0.0003044 [Cycle 1]: 0.00031709, [1] [resolve]: 0.00029857 [Cycle 1]: 0.0016663, [1] [resolve]: 0.00164711 [Cycle 1]: 0.00031841, [1] [resolve]: 0.00029976 [Cycle 1]: 0.00031383, [1] [resolve]: 0.00029637 [after_resolve]: 7.065e-05 [a_after_grad]: 0.00016017 [renormalize]: 0.0382672 [real_op_eliminate]: 5.258e-05 [auto_monad_grad]: 0.00020464 [auto_monad_eliminator]: 0.00010493 [cse]: 0.00030555 [a_3]: 0.00042205 [Cycle 3]: 0.00472762, [30] [expand_dump_flag]: 4.06e-06 [switch_simplify]: 0.00011246 [a_1]: 0.00122816 [recompute_prepare]: 1.71e-05 [updatestate_depend_eliminate]: 2.733e-05 [updatestate_assign_eliminate]: 4.393e-05 [updatestate_loads_eliminate]: 1.917e-05 [parameter_eliminate]: 4.45e-06 [a_2]: 0.00027696 [accelerated_algorithm]: 2.428e-05 [pynative_shard]: 1.22e-06 [auto_parallel]: 3.56e-06 [parallel]: 3.92e-06 [merge_comm]: 3.73001e-06 [allreduce_fusion]: 2.41e-06 [virtual_dataset]: 1.389e-05 [get_grad_eliminate_]: 1.311e-05 [virtual_output]: 1.286e-05 [merge_forward]: 1.945e-05 [cell_reuse_recompute_pass]: 4.69998e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.345e-05 [meta_fg_expand]: 4.761e-05 [after_resolve]: 1.67e-05 [a_after_grad]: 2.045e-05 [renormalize]: 0.00222997 [real_op_eliminate]: 1.961e-05 [auto_monad_grad]: 5.53e-06 [auto_monad_eliminator]: 3.496e-05 [cse]: 0.00020358 [a_3]: 0.00012129 [Cycle 4]: 0.00120497, [30] [expand_dump_flag]: 1.35e-06 [switch_simplify]: 1.517e-05 [a_1]: 0.00028308 [recompute_prepare]: 1.571e-05 [updatestate_depend_eliminate]: 1.954e-05 [updatestate_assign_eliminate]: 1.716e-05 [updatestate_loads_eliminate]: 1.655e-05 [parameter_eliminate]: 2.22e-06 [a_2]: 0.00027135 [accelerated_algorithm]: 2.512e-05 [pynative_shard]: 1.48e-06 [auto_parallel]: 4.33e-06 [parallel]: 3.66e-06 [merge_comm]: 3.24e-06 [allreduce_fusion]: 2.05e-06 [virtual_dataset]: 1.495e-05 [get_grad_eliminate_]: 1.364e-05 [virtual_output]: 1.313e-05 [merge_forward]: 1.802e-05 [cell_reuse_recompute_pass]: 4.00003e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.363e-05 [meta_fg_expand]: 1.329e-05 [after_resolve]: 1.555e-05 [a_after_grad]: 2.047e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.34e-05 [auto_monad_grad]: 2.33e-06 [auto_monad_eliminator]: 3.095e-05 [cse]: 8.274e-05 [a_3]: 0.00011192 [py_interpret_to_execute_after_opt_a]: 4.12e-06 [slice_cell_reuse_recomputed_activation]: 1.33e-06 [rewriter_after_opt_a]: 9.395e-05 [convert_after_rewriter]: 2.246e-05 [order_py_execute_after_rewriter]: 1.583e-05 [opt_b]: 0.00109987, [2] [Cycle 1]: 0.00093042, [7] [b_1]: 0.00083435 [b_2]: 5.68e-06 [updatestate_depend_eliminate]: 5.98e-06 [updatestate_assign_eliminate]: 4.29e-06 [updatestate_loads_eliminate]: 4.3e-06 [renormalize]: 5.10001e-07 [cse]: 3.869e-05 [Cycle 2]: 0.00015971, [7] [b_1]: 9.861e-05 [b_2]: 3.99e-06 [updatestate_depend_eliminate]: 4.49e-06 [updatestate_assign_eliminate]: 3.83e-06 [updatestate_loads_eliminate]: 3.83e-06 [renormalize]: 6.00048e-08 [cse]: 1.816e-05 [cconv]: 1.743e-05 [opt_after_cconv]: 7.05e-05, [1] [Cycle 1]: 6.554e-05, [7] [c_1]: 8.55e-06 [parameter_eliminate]: 2.06e-06 [updatestate_depend_eliminate]: 4.36e-06 [updatestate_assign_eliminate]: 3.65001e-06 [updatestate_loads_eliminate]: 3.62e-06 [cse]: 1.611e-05 [renormalize]: 5.29995e-07 [remove_dup_value]: 1.294e-05 [tuple_transform]: 6.592e-05, [1] [Cycle 1]: 6.219e-05, [3] [d_1]: 3.997e-05 [d_2]: 9.13e-06 [renormalize]: 2.00002e-07 [add_cache_embedding]: 9.1e-06 [add_recomputation]: 4.915e-05 [cse_after_recomputation]: 2.432e-05, [1] [Cycle 1]: 2.055e-05, [1] [cse]: 1.613e-05 [environ_conv]: 8.77e-06 [label_micro_interleaved_index]: 1.6e-06 [label_fine_grained_interleaved_index]: 2.04e-06 [assign_add_opt]: 9.79999e-07 [slice_recompute_activation]: 1.27e-06 [micro_interleaved_order_control]: 1.39e-06 [full_micro_interleaved_order_control]: 7.89994e-07 [comp_comm_scheduling]: 1.2e-06 [reorder_send_recv_between_fp_bp]: 1.66e-06 [comm_op_add_attrs]: 6.69999e-07 [add_comm_op_reuse_tag]: 6.40001e-07 [overlap_opt_shard_in_pipeline]: 9.00007e-07 [grouped_pairwise_exchange_alltoall]: 7.7e-07 [overlap_recompute_and_grad_model_parallel]: 1.03e-06 [overlap_grad_matmul_and_grad_allreduce]: 5.19998e-07 [split_matmul_comm_elemetwise]: 1.83e-06 [split_layernorm_comm]: 8.70001e-07 [process_send_recv_for_ge]: 9.70002e-07 [handle_group_info]: 6.69999e-07 [auto_monad_reorder]: 1.836e-05 [get_jit_bprop_graph]: 3.89999e-07 [eliminate_special_op_node]: 0.00049978 [validate]: 3.63e-05 [distribtued_split]: 1.26e-06 [task_emit]: 0.00521978 [execute]: 4.85e-06 Sums parse : 0.001295s : 1.01% symbol_resolve.resolve : 0.012265s : 9.53% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000128s : 0.10% pre_cconv : 0.000000s : 0.00% abstract_specialize : 0.004047s : 3.14% pack_expand : 0.000015s : 0.01% auto_monad : 0.000064s : 0.05% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000007s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000188s : 0.15% optimize.opt_a.expand_dump_flag : 0.000012s : 0.01% optimize.opt_a.switch_simplify : 0.000255s : 0.20% optimize.opt_a.a_1 : 0.002700s : 2.10% optimize.opt_a.recompute_prepare : 0.000055s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000073s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000080s : 0.06% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000824s : 0.64% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000015s : 0.01% optimize.opt_a.parallel : 0.000019s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000040s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000103s : 0.08% optimize.opt_a.meta_fg_expand : 0.000061s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006041s : 4.69% optimize.opt_a.after_resolve : 0.000151s : 0.12% optimize.opt_a.a_after_grad : 0.000312s : 0.24% optimize.opt_a.renormalize : 0.089963s : 69.87% optimize.opt_a.real_op_eliminate : 0.000127s : 0.10% optimize.opt_a.auto_monad_grad : 0.000277s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000247s : 0.19% optimize.opt_a.cse : 0.000939s : 0.73% optimize.opt_a.a_3 : 0.000965s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000001s : 0.00% optimize.rewriter_after_opt_a : 0.000094s : 0.07% optimize.convert_after_rewriter : 0.000022s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000933s : 0.72% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000057s : 0.04% optimize.cconv : 0.000017s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000040s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000009s : 0.01% optimize.add_recomputation : 0.000049s : 0.04% optimize.cse_after_recomputation.cse : 0.000016s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000001s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000018s : 0.01% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000500s : 0.39% validate : 0.000036s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005220s : 4.05% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018537 880 0.02% : 0.000003s : 5: substitution.float_depend_g_call 0.13% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.38% : 0.016754s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.87% : 0.001087s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000028s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.02% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.49% : 0.000090s : 63: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.01% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.60% : 0.000111s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_item_depend_reorder 0.82% : 0.000152s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089949 6 92.72% : 0.083403s : 3: renormalize.infer 7.28% : 0.006547s : 3: renormalize.specialize ------[replace.] 0.001234 141 54.11% : 0.000668s : 55: replace.getattr_setattr_resolve 26.62% : 0.000328s : 56: replace.inline 3.71% : 0.000046s : 2: replace.meta_unpack_prepare 7.44% : 0.000092s : 10: replace.switch_simplify 1.47% : 0.000018s : 4: replace.transpose_eliminate 6.65% : 0.000082s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017611 141 94.40% : 0.016625s : 55: match.getattr_setattr_resolve 5.22% : 0.000920s : 56: match.inline 0.09% : 0.000016s : 2: match.meta_unpack_prepare 0.05% : 0.000010s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006735 119 69.70% : 0.004695s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.30% : 0.002041s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.024666 259 7.45% : 0.001838s : 104: opt.transform.opt_a 3.65% : 0.000901s : 92: opt.transform.opt_b 72.90% : 0.017980s : 14: opt.transform.opt_resolve 0.45% : 0.000111s : 1: opt.transforms.meta_unpack_prepare 15.24% : 0.003758s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000007s : 2: opt.transforms.opt_b 0.19% : 0.000047s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:30.163.696 [graph_var_manager.cc:1424][EVENT]36565 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:30.163.778 [graph_manager.cc:1248][EVENT]36565 PreRun:PreRun start: graph node size 3, session id 5, graph id 4, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.164.084 [atrace_api.c:28](tid:36565) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.164.121 [trace_rb_log.c:84](tid:36565) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.164.134 [atrace_api.c:32](tid:36565) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:30.164.147 [client_manager.cpp:157][SetProfilingCallback][tid:36565] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:30.164.568 [parallel_partitioner.cc:165][EVENT]36565 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.164.609 [parallel_partitioner.cc:178][EVENT]36565 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.164.660 [graph_prepare.cc:1378][EVENT]36565 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.164.868 [graph_manager.cc:1050][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [228] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.164.894 [graph_manager.cc:1052][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.018 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.048 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.100 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [40] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.114 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.178 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.192 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.209 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.306 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.341 [graph_manager.cc:1054][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [434] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.165.559 [graph_manager.cc:1055][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [203] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.495 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.520 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.531 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.541 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [297] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.550 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.559 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.567 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.576 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [17] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.166.584 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.167.918 [graph_manager.cc:1056][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2340] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.167.980 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.167.999 [graph_prepare.cc:1982][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.398 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.419 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.429 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.438 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [225] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.446 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.455 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.463 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.472 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.487 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.514 [graph_prepare.cc:1983][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [502] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.537 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.548 [graph_prepare.cc:1984][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.562 [graph_prepare.cc:1985][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.577 [graph_prepare.cc:1986][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.588 [graph_prepare.cc:1987][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.602 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.614 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.628 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.712 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.725 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.733 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrintOpPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.742 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.750 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.759 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.767 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.775 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.784 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of StopGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.792 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.800 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.809 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SnapshotPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.822 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.831 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.839 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.847 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.870 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.884 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.916 [graph_prepare.cc:1988][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [320] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.168.929 [graph_manager.cc:1065][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [982] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.181.261 [graph_manager.cc:1077][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12313] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.181.327 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.181.381 [graph_manager.cc:1080][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [88] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.043 [graph_manager.cc:1081][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2646] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.079 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.095 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.106 [graph_manager.cc:1082][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.137 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.153 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.167 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.237 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [60] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.254 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.285 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.301 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.352 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.371 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.390 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.416 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.433 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.445 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.454 [graph_manager.cc:2700][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [322] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.558 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.572 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.582 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.591 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.599 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.608 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CastRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.616 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.624 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.633 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.641 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.649 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.657 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.665 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.674 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.682 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.705 [graph_manager.cc:2741][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [233] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.714 [graph_manager.cc:2752][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.737 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.750 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.766 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.781 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.793 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.808 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.828 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.841 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.854 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.864 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.877 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.887 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.906 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.919 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.927 [graph_manager.cc:2810][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [195] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.957 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.969 [graph_manager.cc:2821][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.184.997 [graph_manager.cc:1087][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [873] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.143 [graph_manager.cc:1088][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [134] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.186 [graph_manager.cc:1089][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.212 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.228 [graph_manager.cc:1097][EVENT]36565 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.250 [graph_manager.cc:3325][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.458 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.475 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.484 [engine_place.cc:144][EVENT]36565 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [120] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.555 [graph_manager.cc:3351][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [292] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.572 [graph_manager.cc:3364][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.636 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.653 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.804 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [140] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.844 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.891 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.926 [graph_manager.cc:3405][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [340] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.185.944 [graph_manager.cc:3412][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.342 [graph_manager.cc:3422][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1384] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.372 [graph_manager.cc:3428][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.493 [graph_manager.cc:3467][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [101] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.511 [graph_manager.cc:3377][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1926] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.527 [graph_manager.cc:1106][EVENT]36565 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2284] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.539 [graph_manager.cc:1115][EVENT]36565 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.569 [graph_manager.cc:1130][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.603 [graph_manager.cc:1131][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.626 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.645 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.655 [graph_manager.cc:2837][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.725 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.738 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.748 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.756 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.765 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.773 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.783 [graph_manager.cc:2864][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [111] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.794 [graph_manager.cc:2872][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.813 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.827 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.843 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.857 [compile_nodes_pass.cc:88][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.868 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.878 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.956 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [68] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.187.983 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.002 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.015 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.028 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.037 [graph_manager.cc:2927][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [227] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.049 [graph_manager.cc:2937][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.062 [graph_manager.cc:2943][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.073 [graph_manager.cc:2950][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.229 [graph_manager.cc:2958][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [36] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.260 [graph_manager.cc:1132][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [643] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.336 [graph_manager.cc:1135][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.366 [graph_manager.cc:2975][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.397 [graph_manager.cc:2981][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.412 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.421 [graph_manager.cc:2986][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.430 [graph_manager.cc:1136][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [80] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.535 [graph_manager.cc:3555][EVENT]36565 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [73] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.618 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.632 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.743 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.772 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.818 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [26] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.188.839 [graph_builder.cc:865][EVENT]36565 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [251] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:30.189.213 [logger.cc:1071] 36565 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.244 [task_generator.cc:804][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [85] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.302 [task_generator.cc:805][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.737 [task_generator.cc:814][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [421] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.751 [task_generator.cc:954][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [593] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.806 [task_generator.cc:967][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [31] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:30.189.823 [logger.cc:1084] 36565 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.970 [graph_manager.cc:1152][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1515] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.189.988 [graph_manager.cc:1164][EVENT]36565 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.190.019 [graph_manager.cc:1271][EVENT]36565 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [25538] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.190.029 [graph_manager.cc:1272][EVENT]36565 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.190.334 [atrace_api.c:93](tid:36565) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.190.349 [atrace_api.c:95](tid:36565) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:30.194.595 [graph_converter.cc:838][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1199] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.194.785 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [149] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.164 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [358] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.246 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [62] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.260 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [77] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.543 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [272] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.648 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.683 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.831 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [136] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.899 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [53] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.919 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [73] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.948 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.195.975 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.001 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.063 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.121 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.132 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [58] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.157 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.181 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.201 [graph_converter.cc:849][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1572] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.196.394 [graph_converter.cc:853][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [182] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.197.023 [graph_converter.cc:857][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [616] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.197.164 [graph_converter.cc:862][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [119] micro second. . TotalTime = 0.146112, [20] [parse]: 0.0014834 [symbol_resolve]: 0.0125122, [1] [Cycle 1]: 0.0124059, [1] [resolve]: 0.0123854 [combine_like_graphs]: 1e-06 [graph_reusing]: 3.41e-06 [meta_unpack_prepare]: 0.0001659 [pre_cconv]: 9.20001e-07 [abstract_specialize]: 0.00435224 [pack_expand]: 1.608e-05 [auto_monad]: 8.32e-05 [inline]: 1.75e-06 [pre_auto_parallel]: 1.01e-05 [pipeline_split]: 3.27e-06 [optimize]: 0.123327, [35] [py_interpret_to_execute]: 4.82e-06 [rewriter_before_opt_a]: 0.00019092 [opt_a]: 0.121402, [4] [Cycle 1]: 0.0586129, [30] [expand_dump_flag]: 4.37e-06 [switch_simplify]: 2.725e-05 [a_1]: 0.00074845 [recompute_prepare]: 8.59e-06 [updatestate_depend_eliminate]: 1.06e-05 [updatestate_assign_eliminate]: 7.32e-06 [updatestate_loads_eliminate]: 6.66e-06 [parameter_eliminate]: 4.99e-06 [a_2]: 7.571e-05 [accelerated_algorithm]: 5.48e-06 [pynative_shard]: 1.67e-06 [auto_parallel]: 3.51e-06 [parallel]: 9.24e-06 [merge_comm]: 3.87e-06 [allreduce_fusion]: 2.12e-06 [virtual_dataset]: 5.71e-06 [get_grad_eliminate_]: 4.9e-06 [virtual_output]: 4.55e-06 [merge_forward]: 8.78e-06 [cell_reuse_recompute_pass]: 7.40001e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.212e-05 [meta_fg_expand]: 0.00695986, [1] [Cycle 1]: 0.00316349, [1] [resolve]: 0.00314421 [after_resolve]: 4.916e-05 [a_after_grad]: 0.0001334 [renormalize]: 0.0494492 [real_op_eliminate]: 4.477e-05 [auto_monad_grad]: 7.185e-05 [auto_monad_eliminator]: 8.093e-05 [cse]: 0.00036899 [a_3]: 0.00031056 [Cycle 2]: 0.0527952, [30] [expand_dump_flag]: 3.83e-06 [switch_simplify]: 0.00013415 [a_1]: 0.00161546 [recompute_prepare]: 1.328e-05 [updatestate_depend_eliminate]: 1.682e-05 [updatestate_assign_eliminate]: 1.273e-05 [updatestate_loads_eliminate]: 1.31e-05 [parameter_eliminate]: 3.97e-06 [a_2]: 0.00021528 [accelerated_algorithm]: 2.168e-05 [pynative_shard]: 1.6e-06 [auto_parallel]: 5.17e-06 [parallel]: 4.85e-06 [merge_comm]: 2.51e-06 [allreduce_fusion]: 1.49e-06 [virtual_dataset]: 1.128e-05 [get_grad_eliminate_]: 1.025e-05 [virtual_output]: 1.05e-05 [merge_forward]: 1.462e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.414e-05 [meta_fg_expand]: 0.0115106, [5] [Cycle 1]: 0.00033983, [1] [resolve]: 0.00032172 [Cycle 1]: 0.00032093, [1] [resolve]: 0.00030247 [Cycle 1]: 0.00170873, [1] [resolve]: 0.00169032 [Cycle 1]: 0.00030601, [1] [resolve]: 0.00028835 [Cycle 1]: 0.00030746, [1] [resolve]: 0.00028981 [after_resolve]: 7.434e-05 [a_after_grad]: 0.00020977 [renormalize]: 0.037566 [real_op_eliminate]: 5.682e-05 [auto_monad_grad]: 0.00021058 [auto_monad_eliminator]: 0.00010572 [cse]: 0.00030994 [a_3]: 0.0004214 [Cycle 3]: 0.00594279, [30] [expand_dump_flag]: 3.81e-06 [switch_simplify]: 0.00015612 [a_1]: 0.0023438 [recompute_prepare]: 1.673e-05 [updatestate_depend_eliminate]: 3.089e-05 [updatestate_assign_eliminate]: 1.875e-05 [updatestate_loads_eliminate]: 1.777e-05 [parameter_eliminate]: 4.52e-06 [a_2]: 0.00027303 [accelerated_algorithm]: 2.546e-05 [pynative_shard]: 1.66e-06 [auto_parallel]: 4.62e-06 [parallel]: 4.29e-06 [merge_comm]: 3.67e-06 [allreduce_fusion]: 2.48e-06 [virtual_dataset]: 1.521e-05 [get_grad_eliminate_]: 1.449e-05 [virtual_output]: 1.394e-05 [merge_forward]: 1.931e-05 [cell_reuse_recompute_pass]: 4.49996e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.282e-05 [meta_fg_expand]: 4.858e-05 [after_resolve]: 1.754e-05 [a_after_grad]: 3.239e-05 [renormalize]: 0.00228997 [real_op_eliminate]: 2.197e-05 [auto_monad_grad]: 5.86001e-06 [auto_monad_eliminator]: 3.516e-05 [cse]: 0.00020471 [a_3]: 0.00012171 [Cycle 4]: 0.00166079, [30] [expand_dump_flag]: 1.32e-06 [switch_simplify]: 1.572e-05 [a_1]: 0.00074179 [recompute_prepare]: 1.58e-05 [updatestate_depend_eliminate]: 2.011e-05 [updatestate_assign_eliminate]: 1.683e-05 [updatestate_loads_eliminate]: 1.633e-05 [parameter_eliminate]: 2.26001e-06 [a_2]: 0.00027511 [accelerated_algorithm]: 2.547e-05 [pynative_shard]: 1.51e-06 [auto_parallel]: 4.01e-06 [parallel]: 3.62e-06 [merge_comm]: 3.38e-06 [allreduce_fusion]: 1.97e-06 [virtual_dataset]: 1.526e-05 [get_grad_eliminate_]: 1.535e-05 [virtual_output]: 1.442e-05 [merge_forward]: 1.736e-05 [cell_reuse_recompute_pass]: 4.00003e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.305e-05 [meta_fg_expand]: 1.316e-05 [after_resolve]: 1.697e-05 [a_after_grad]: 3.214e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.455e-05 [auto_monad_grad]: 2.33e-06 [auto_monad_eliminator]: 3.082e-05 [cse]: 8.276e-05 [a_3]: 0.00011223 [py_interpret_to_execute_after_opt_a]: 4.31e-06 [slice_cell_reuse_recomputed_activation]: 2.24e-06 [rewriter_after_opt_a]: 9.863e-05 [convert_after_rewriter]: 2.396e-05 [order_py_execute_after_rewriter]: 1.726e-05 [opt_b]: 0.00110135, [2] [Cycle 1]: 0.00093246, [7] [b_1]: 0.00083812 [b_2]: 5.05e-06 [updatestate_depend_eliminate]: 5.96e-06 [updatestate_assign_eliminate]: 4.14e-06 [updatestate_loads_eliminate]: 3.86e-06 [renormalize]: 3.99996e-07 [cse]: 3.94e-05 [Cycle 2]: 0.00015921, [7] [b_1]: 9.823e-05 [b_2]: 3.99999e-06 [updatestate_depend_eliminate]: 5.24e-06 [updatestate_assign_eliminate]: 3.73e-06 [updatestate_loads_eliminate]: 3.48e-06 [renormalize]: 7.0002e-08 [cse]: 1.758e-05 [cconv]: 2.195e-05 [opt_after_cconv]: 8.44e-05, [1] [Cycle 1]: 8.033e-05, [7] [c_1]: 2.403e-05 [parameter_eliminate]: 2.08e-06 [updatestate_depend_eliminate]: 4.19e-06 [updatestate_assign_eliminate]: 4.09e-06 [updatestate_loads_eliminate]: 3.54e-06 [cse]: 1.581e-05 [renormalize]: 3.20004e-07 [remove_dup_value]: 1.655e-05 [tuple_transform]: 8.22e-05, [1] [Cycle 1]: 7.852e-05, [3] [d_1]: 5.704e-05 [d_2]: 9.05e-06 [renormalize]: 1.69995e-07 [add_cache_embedding]: 1.378e-05 [add_recomputation]: 5.909e-05 [cse_after_recomputation]: 2.712e-05, [1] [Cycle 1]: 2.285e-05, [1] [cse]: 1.806e-05 [environ_conv]: 3.913e-05 [label_micro_interleaved_index]: 3.61e-06 [label_fine_grained_interleaved_index]: 2.37e-06 [assign_add_opt]: 1.85e-06 [slice_recompute_activation]: 2.17e-06 [micro_interleaved_order_control]: 1.61e-06 [full_micro_interleaved_order_control]: 2.16e-06 [comp_comm_scheduling]: 1.98e-06 [reorder_send_recv_between_fp_bp]: 2.13e-06 [comm_op_add_attrs]: 1e-06 [add_comm_op_reuse_tag]: 8.29998e-07 [overlap_opt_shard_in_pipeline]: 1.21e-06 [grouped_pairwise_exchange_alltoall]: 1.11001e-06 [overlap_recompute_and_grad_model_parallel]: 1.63e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.89994e-07 [split_matmul_comm_elemetwise]: 2.13e-06 [split_layernorm_comm]: 2.21e-06 [process_send_recv_for_ge]: 9.80006e-07 [handle_group_info]: 1.1e-06 [auto_monad_reorder]: 2.253e-05 [get_jit_bprop_graph]: 3.27e-06 [eliminate_special_op_node]: 0.00049749 [validate]: 3.913e-05 [distribtued_split]: 1.28e-06 [task_emit]: 0.00339039 [execute]: 6.14e-06 Sums parse : 0.001483s : 1.14% symbol_resolve.resolve : 0.012385s : 9.52% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000166s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004352s : 3.35% pack_expand : 0.000016s : 0.01% auto_monad : 0.000083s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000191s : 0.15% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000333s : 0.26% optimize.opt_a.a_1 : 0.005450s : 4.19% optimize.opt_a.recompute_prepare : 0.000054s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000078s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000839s : 0.65% optimize.opt_a.accelerated_algorithm : 0.000078s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000022s : 0.02% optimize.opt_a.merge_comm : 0.000013s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000045s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000062s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006037s : 4.64% optimize.opt_a.after_resolve : 0.000158s : 0.12% optimize.opt_a.a_after_grad : 0.000408s : 0.31% optimize.opt_a.renormalize : 0.089305s : 68.65% optimize.opt_a.real_op_eliminate : 0.000138s : 0.11% optimize.opt_a.auto_monad_grad : 0.000291s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000253s : 0.19% optimize.opt_a.cse : 0.000966s : 0.74% optimize.opt_a.a_3 : 0.000966s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000099s : 0.08% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000936s : 0.72% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000007s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000057s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000057s : 0.04% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000014s : 0.01% optimize.add_recomputation : 0.000059s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000039s : 0.03% optimize.label_micro_interleaved_index : 0.000004s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000023s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000497s : 0.38% validate : 0.000039s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003390s : 2.61% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018603 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.34% : 0.016806s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.55% : 0.001033s : 103: substitution.inline 0.05% : 0.000009s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.50% : 0.000092s : 69: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000012s : 10: substitution.switch_simplify 0.06% : 0.000011s : 4: substitution.transpose_eliminate 0.65% : 0.000120s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000057s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.87% : 0.000162s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000077s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089288 6 92.50% : 0.082592s : 3: renormalize.infer 7.50% : 0.006696s : 3: renormalize.specialize ------[replace.] 0.001241 141 54.75% : 0.000679s : 55: replace.getattr_setattr_resolve 25.71% : 0.000319s : 56: replace.inline 3.76% : 0.000047s : 2: replace.meta_unpack_prepare 7.39% : 0.000092s : 10: replace.switch_simplify 1.73% : 0.000021s : 4: replace.transpose_eliminate 6.66% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017598 141 94.74% : 0.016673s : 55: match.getattr_setattr_resolve 4.87% : 0.000857s : 56: match.inline 0.10% : 0.000018s : 2: match.meta_unpack_prepare 0.07% : 0.000012s : 10: match.switch_simplify 0.06% : 0.000011s : 4: match.transpose_eliminate 0.16% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006924 119 69.28% : 0.004797s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.72% : 0.002127s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027713 589 0.52% : 0.000143s : 2: opt.transform.meta_unpack_prepare 30.55% : 0.008467s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.28% : 0.000910s : 94: opt.transform.opt_b 65.30% : 0.018097s : 14: opt.transform.opt_resolve 0.22% : 0.000061s : 8: opt.transform.opt_trans_graph 0.05% : 0.000015s : 3: opt.transform.special_op_eliminate . TotalTime = 0.144689, [20] [parse]: 0.00131265 [symbol_resolve]: 0.01253, [1] [Cycle 1]: 0.0124655, [1] [resolve]: 0.0124476 [combine_like_graphs]: 1.09e-06 [graph_reusing]: 3.25e-06 [meta_unpack_prepare]: 0.00012667 [pre_cconv]: 5.4e-07 [abstract_specialize]: 0.00406769 [pack_expand]: 1.422e-05 [auto_monad]: 6.586e-05 [inline]: 1.3e-06 [pre_auto_parallel]: 6.45e-06 [pipeline_split]: 1.68e-06 [optimize]: 0.120541, [35] [py_interpret_to_execute]: 4.44e-06 [rewriter_before_opt_a]: 0.00018643 [opt_a]: 0.118735, [4] [Cycle 1]: 0.0582411, [30] [expand_dump_flag]: 3.53999e-06 [switch_simplify]: 2.721e-05 [a_1]: 0.00040777 [recompute_prepare]: 8.62e-06 [updatestate_depend_eliminate]: 9.43e-06 [updatestate_assign_eliminate]: 6.82e-06 [updatestate_loads_eliminate]: 6.31e-06 [parameter_eliminate]: 4.24001e-06 [a_2]: 7.771e-05 [accelerated_algorithm]: 5.29e-06 [pynative_shard]: 1.15e-06 [auto_parallel]: 3.32999e-06 [parallel]: 5.51e-06 [merge_comm]: 2.84e-06 [allreduce_fusion]: 1.76e-06 [virtual_dataset]: 5.06e-06 [get_grad_eliminate_]: 4.32e-06 [virtual_output]: 4.02e-06 [merge_forward]: 7.27001e-06 [cell_reuse_recompute_pass]: 4.50003e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.161e-05 [meta_fg_expand]: 0.00701219, [1] [Cycle 1]: 0.00327715, [1] [resolve]: 0.00325826 [after_resolve]: 4.974e-05 [a_after_grad]: 0.00010958 [renormalize]: 0.0494393 [real_op_eliminate]: 4.028e-05 [auto_monad_grad]: 6.993e-05 [auto_monad_eliminator]: 7.79e-05 [cse]: 0.00034076 [a_3]: 0.00031013 [Cycle 2]: 0.0522274, [30] [expand_dump_flag]: 3.29e-06 [switch_simplify]: 9.945e-05 [a_1]: 0.00075531 [recompute_prepare]: 1.389e-05 [updatestate_depend_eliminate]: 1.617e-05 [updatestate_assign_eliminate]: 1.257e-05 [updatestate_loads_eliminate]: 1.237e-05 [parameter_eliminate]: 3.88e-06 [a_2]: 0.00019797 [accelerated_algorithm]: 1.76e-05 [pynative_shard]: 1.21e-06 [auto_parallel]: 3.91e-06 [parallel]: 4.36e-06 [merge_comm]: 2.45e-06 [allreduce_fusion]: 1.57e-06 [virtual_dataset]: 1.035e-05 [get_grad_eliminate_]: 9.64e-06 [virtual_output]: 9.33e-06 [merge_forward]: 1.348e-05 [cell_reuse_recompute_pass]: 4.79995e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.352e-05 [meta_fg_expand]: 0.0113343, [5] [Cycle 1]: 0.00032227, [1] [resolve]: 0.00030427 [Cycle 1]: 0.00031753, [1] [resolve]: 0.00029977 [Cycle 1]: 0.00165885, [1] [resolve]: 0.00163993 [Cycle 1]: 0.00031661, [1] [resolve]: 0.0002988 [Cycle 1]: 0.00031715, [1] [resolve]: 0.00030019 [after_resolve]: 7.154e-05 [a_after_grad]: 0.0001609 [renormalize]: 0.0381677 [real_op_eliminate]: 5.295e-05 [auto_monad_grad]: 0.00020608 [auto_monad_eliminator]: 0.0001062 [cse]: 0.00030721 [a_3]: 0.00042102 [Cycle 3]: 0.00474228, [30] [expand_dump_flag]: 3.8e-06 [switch_simplify]: 0.00011408 [a_1]: 0.00122598 [recompute_prepare]: 1.743e-05 [updatestate_depend_eliminate]: 2.706e-05 [updatestate_assign_eliminate]: 1.825e-05 [updatestate_loads_eliminate]: 1.761e-05 [parameter_eliminate]: 4.21e-06 [a_2]: 0.00029536 [accelerated_algorithm]: 2.542e-05 [pynative_shard]: 1.19e-06 [auto_parallel]: 4.09e-06 [parallel]: 3.85e-06 [merge_comm]: 3.94e-06 [allreduce_fusion]: 2.32e-06 [virtual_dataset]: 1.401e-05 [get_grad_eliminate_]: 1.334e-05 [virtual_output]: 1.264e-05 [merge_forward]: 1.977e-05 [cell_reuse_recompute_pass]: 4.40006e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.293e-05 [meta_fg_expand]: 4.898e-05 [after_resolve]: 1.664e-05 [a_after_grad]: 2.125e-05 [renormalize]: 0.00224828 [real_op_eliminate]: 1.981e-05 [auto_monad_grad]: 5.76e-06 [auto_monad_eliminator]: 3.495e-05 [cse]: 0.00020536 [a_3]: 0.00012107 [Cycle 4]: 0.00122085, [30] [expand_dump_flag]: 1.35999e-06 [switch_simplify]: 1.461e-05 [a_1]: 0.00028303 [recompute_prepare]: 1.522e-05 [updatestate_depend_eliminate]: 2.042e-05 [updatestate_assign_eliminate]: 1.697e-05 [updatestate_loads_eliminate]: 1.637e-05 [parameter_eliminate]: 2.22e-06 [a_2]: 0.00027271 [accelerated_algorithm]: 2.55e-05 [pynative_shard]: 1.51e-06 [auto_parallel]: 3.57e-06 [parallel]: 3.69e-06 [merge_comm]: 3.28e-06 [allreduce_fusion]: 2.1e-06 [virtual_dataset]: 1.419e-05 [get_grad_eliminate_]: 1.341e-05 [virtual_output]: 1.269e-05 [merge_forward]: 1.802e-05 [cell_reuse_recompute_pass]: 4.20005e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.349e-05 [meta_fg_expand]: 1.364e-05 [after_resolve]: 1.848e-05 [a_after_grad]: 2.069e-05 [renormalize]: 6.99947e-08 [real_op_eliminate]: 1.393e-05 [auto_monad_grad]: 2.58e-06 [auto_monad_eliminator]: 3.109e-05 [cse]: 8.314e-05 [a_3]: 0.0001121 [py_interpret_to_execute_after_opt_a]: 4.08e-06 [slice_cell_reuse_recomputed_activation]: 1.32e-06 [rewriter_after_opt_a]: 9.576e-05 [convert_after_rewriter]: 2.276e-05 [order_py_execute_after_rewriter]: 1.641e-05 [opt_b]: 0.00109569, [2] [Cycle 1]: 0.00092678, [7] [b_1]: 0.00083367 [b_2]: 5.48999e-06 [updatestate_depend_eliminate]: 5.98e-06 [updatestate_assign_eliminate]: 4.21e-06 [updatestate_loads_eliminate]: 4.06e-06 [renormalize]: 4.39999e-07 [cse]: 3.744e-05 [Cycle 2]: 0.00015926, [7] [b_1]: 9.785e-05 [b_2]: 4.17e-06 [updatestate_depend_eliminate]: 4.58e-06 [updatestate_assign_eliminate]: 3.83e-06 [updatestate_loads_eliminate]: 3.72e-06 [renormalize]: 7.0002e-08 [cse]: 1.785e-05 [cconv]: 1.677e-05 [opt_after_cconv]: 6.915e-05, [1] [Cycle 1]: 6.476e-05, [7] [c_1]: 8.46e-06 [parameter_eliminate]: 2.04e-06 [updatestate_depend_eliminate]: 4.43e-06 [updatestate_assign_eliminate]: 3.63e-06 [updatestate_loads_eliminate]: 3.6e-06 [cse]: 1.602e-05 [renormalize]: 3.30001e-07 [remove_dup_value]: 1.274e-05 [tuple_transform]: 6.585e-05, [1] [Cycle 1]: 6.219e-05, [3] [d_1]: 3.982e-05 [d_2]: 9.18e-06 [renormalize]: 2.19996e-07 [add_cache_embedding]: 9.03e-06 [add_recomputation]: 4.938e-05 [cse_after_recomputation]: 2.48e-05, [1] [Cycle 1]: 2.079e-05, [1] [cse]: 1.628e-05 [environ_conv]: 8.97e-06 [label_micro_interleaved_index]: 1.57e-06 [label_fine_grained_interleaved_index]: 1.35e-06 [assign_add_opt]: 1.07e-06 [slice_recompute_activation]: 1.39e-06 [micro_interleaved_order_control]: 1.03e-06 [full_micro_interleaved_order_control]: 1e-06 [comp_comm_scheduling]: 1.25e-06 [reorder_send_recv_between_fp_bp]: 1.15e-06 [comm_op_add_attrs]: 6.60002e-07 [add_comm_op_reuse_tag]: 6.90001e-07 [overlap_opt_shard_in_pipeline]: 6.40001e-07 [grouped_pairwise_exchange_alltoall]: 6.00005e-07 [overlap_recompute_and_grad_model_parallel]: 9.29998e-07 [overlap_grad_matmul_and_grad_allreduce]: 5.4e-07 [split_matmul_comm_elemetwise]: 2.07e-06 [split_layernorm_comm]: 1.04e-06 [process_send_recv_for_ge]: 5.30003e-07 [handle_group_info]: 5.80003e-07 [auto_monad_reorder]: 1.938e-05 [get_jit_bprop_graph]: 3.80001e-07 [eliminate_special_op_node]: 0.00049678 [validate]: 3.627e-05 [distribtued_split]: 1.09e-06 [task_emit]: 0.00528521 [execute]: 4.94e-06 Sums parse : 0.001313s : 1.02% symbol_resolve.resolve : 0.012448s : 9.65% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000127s : 0.10% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004068s : 3.15% pack_expand : 0.000014s : 0.01% auto_monad : 0.000066s : 0.05% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000006s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000186s : 0.14% optimize.opt_a.expand_dump_flag : 0.000012s : 0.01% optimize.opt_a.switch_simplify : 0.000255s : 0.20% optimize.opt_a.a_1 : 0.002672s : 2.07% optimize.opt_a.recompute_prepare : 0.000055s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000073s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000055s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000053s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000844s : 0.65% optimize.opt_a.accelerated_algorithm : 0.000074s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000015s : 0.01% optimize.opt_a.parallel : 0.000017s : 0.01% optimize.opt_a.merge_comm : 0.000013s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000039s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000063s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006101s : 4.73% optimize.opt_a.after_resolve : 0.000156s : 0.12% optimize.opt_a.a_after_grad : 0.000312s : 0.24% optimize.opt_a.renormalize : 0.089855s : 69.67% optimize.opt_a.real_op_eliminate : 0.000127s : 0.10% optimize.opt_a.auto_monad_grad : 0.000284s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000250s : 0.19% optimize.opt_a.cse : 0.000936s : 0.73% optimize.opt_a.a_3 : 0.000964s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000001s : 0.00% optimize.rewriter_after_opt_a : 0.000096s : 0.07% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000932s : 0.72% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000055s : 0.04% optimize.cconv : 0.000017s : 0.01% optimize.opt_after_cconv.c_1 : 0.000008s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000040s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000009s : 0.01% optimize.add_recomputation : 0.000049s : 0.04% optimize.cse_after_recomputation.cse : 0.000016s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000001s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000001s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000001s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000019s : 0.02% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000497s : 0.39% validate : 0.000036s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005285s : 4.10% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018707 880 0.02% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.63% : 0.016955s : 59: substitution.getattr_setattr_resolve 0.02% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.66% : 0.001059s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000029s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.01% : 0.000002s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000008s : 81: substitution.remove_not_recompute_node 0.48% : 0.000089s : 63: substitution.replace_applicator 0.05% : 0.000009s : 36: substitution.replace_old_param 0.01% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.60% : 0.000112s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.35% : 0.000066s : 60: substitution.tuple_list_get_item_depend_reorder 0.82% : 0.000154s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089841 6 92.70% : 0.083279s : 3: renormalize.infer 7.30% : 0.006561s : 3: renormalize.specialize ------[replace.] 0.001266 141 55.63% : 0.000704s : 55: replace.getattr_setattr_resolve 25.90% : 0.000328s : 56: replace.inline 3.49% : 0.000044s : 2: replace.meta_unpack_prepare 7.07% : 0.000090s : 10: replace.switch_simplify 1.39% : 0.000018s : 4: replace.transpose_eliminate 6.52% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017785 141 94.62% : 0.016828s : 55: match.getattr_setattr_resolve 5.00% : 0.000890s : 56: match.inline 0.09% : 0.000017s : 2: match.meta_unpack_prepare 0.06% : 0.000010s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006745 119 69.43% : 0.004683s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.57% : 0.002062s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.024893 259 7.46% : 0.001857s : 104: opt.transform.opt_a 3.61% : 0.000899s : 92: opt.transform.opt_b 73.17% : 0.018215s : 14: opt.transform.opt_resolve 0.44% : 0.000110s : 1: opt.transforms.meta_unpack_prepare 15.00% : 0.003735s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000047s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:30.569.164 [graph_var_manager.cc:1424][EVENT]36565 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:30.569.257 [graph_manager.cc:1248][EVENT]36565 PreRun:PreRun start: graph node size 3, session id 6, graph id 5, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.569.547 [atrace_api.c:28](tid:36565) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.569.580 [trace_rb_log.c:84](tid:36565) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.569.592 [atrace_api.c:32](tid:36565) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:30.569.605 [client_manager.cpp:157][SetProfilingCallback][tid:36565] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.005 [parallel_partitioner.cc:165][EVENT]36565 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.041 [parallel_partitioner.cc:178][EVENT]36565 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.089 [graph_prepare.cc:1378][EVENT]36565 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.284 [graph_manager.cc:1050][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [214] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.310 [graph_manager.cc:1052][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.439 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.469 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.522 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [40] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.536 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.581 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.595 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.615 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.712 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.733 [graph_manager.cc:1054][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [409] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.570.953 [graph_manager.cc:1055][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [207] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.909 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.932 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.943 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.962 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [299] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.972 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.980 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.989 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.571.997 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [16] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.572.005 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.369 [graph_manager.cc:1056][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2396] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.431 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.449 [graph_prepare.cc:1982][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.850 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.871 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.881 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.890 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [227] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.899 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.907 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.916 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.924 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.932 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.957 [graph_prepare.cc:1983][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [495] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.980 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.573.991 [graph_prepare.cc:1984][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.005 [graph_prepare.cc:1985][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.019 [graph_prepare.cc:1986][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.038 [graph_prepare.cc:1987][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.053 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.065 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.079 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.162 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.173 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.182 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrintOpPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.191 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.199 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.207 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.216 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.224 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.232 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.240 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.249 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.257 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SnapshotPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.265 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.273 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.281 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.289 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.311 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.325 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.362 [graph_prepare.cc:1988][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [314] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.574.375 [graph_manager.cc:1065][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [977] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.586.637 [graph_manager.cc:1077][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12242] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.586.703 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.586.759 [graph_manager.cc:1080][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [91] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.425 [graph_manager.cc:1081][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2650] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.462 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.477 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.488 [graph_manager.cc:1082][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.519 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.536 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.549 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.619 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.636 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.666 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.681 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.721 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.739 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.756 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.782 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.797 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.821 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.831 [graph_manager.cc:2700][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [316] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.936 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.949 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.959 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.967 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.976 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.984 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CastRemovePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.589.993 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.001 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.009 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.017 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.026 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.034 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.042 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.050 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.058 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.068 [graph_manager.cc:2741][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [219] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.077 [graph_manager.cc:2752][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.099 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.111 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.128 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.144 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.163 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.175 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.194 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.208 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.221 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.231 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.244 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.255 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.274 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.287 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.295 [graph_manager.cc:2810][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [201] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.324 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.337 [graph_manager.cc:2821][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [32] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.363 [graph_manager.cc:1087][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [856] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.495 [graph_manager.cc:1088][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [117] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.533 [graph_manager.cc:1089][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.552 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.568 [graph_manager.cc:1097][EVENT]36565 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.589 [graph_manager.cc:3325][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.796 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.813 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.822 [engine_place.cc:144][EVENT]36565 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [116] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.899 [graph_manager.cc:3351][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [296] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.917 [graph_manager.cc:3364][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.980 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.590.997 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.591.145 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [139] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.591.185 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.591.234 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.591.269 [graph_manager.cc:3405][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [339] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.591.287 [graph_manager.cc:3412][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.690 [graph_manager.cc:3422][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1389] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.719 [graph_manager.cc:3428][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.841 [graph_manager.cc:3467][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.857 [graph_manager.cc:3377][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1928] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.873 [graph_manager.cc:1106][EVENT]36565 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2290] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.885 [graph_manager.cc:1115][EVENT]36565 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.907 [graph_manager.cc:1130][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.939 [graph_manager.cc:1131][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.963 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.978 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.592.988 [graph_manager.cc:2837][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.067 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.080 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.089 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.098 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of BitcastPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.107 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.115 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.141 [graph_manager.cc:2864][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [129] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.153 [graph_manager.cc:2872][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.172 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.186 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.202 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.216 [compile_nodes_pass.cc:88][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.227 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.238 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.316 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [69] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.343 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.356 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.370 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.382 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.392 [graph_manager.cc:2927][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [222] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.404 [graph_manager.cc:2937][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.424 [graph_manager.cc:2943][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.436 [graph_manager.cc:2950][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.593 [graph_manager.cc:2958][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.624 [graph_manager.cc:1132][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [670] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.698 [graph_manager.cc:1135][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [60] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.729 [graph_manager.cc:2975][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.761 [graph_manager.cc:2981][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.777 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.786 [graph_manager.cc:2986][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.795 [graph_manager.cc:1136][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [82] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.901 [graph_manager.cc:3555][EVENT]36565 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [75] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.982 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.593.997 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.594.109 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.594.138 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.594.176 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [26] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.594.196 [graph_builder.cc:865][EVENT]36565 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [244] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:30.594.545 [logger.cc:1071] 36565 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.594.575 [task_generator.cc:804][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [77] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.594.633 [task_generator.cc:805][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.063 [task_generator.cc:814][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [416] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.083 [task_generator.cc:954][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [585] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.139 [task_generator.cc:967][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [32] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:30.595.157 [logger.cc:1084] 36565 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.303 [graph_manager.cc:1152][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1484] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.321 [graph_manager.cc:1164][EVENT]36565 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.351 [graph_manager.cc:1271][EVENT]36565 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [25433] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.595.362 [graph_manager.cc:1272][EVENT]36565 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.595.665 [atrace_api.c:93](tid:36565) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.595.680 [atrace_api.c:95](tid:36565) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:30.599.921 [graph_converter.cc:838][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1200] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.600.115 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [150] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.600.497 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [361] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.600.578 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [60] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.600.593 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [76] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.600.875 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [271] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.600.979 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.015 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.190 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [163] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.260 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [54] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.272 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [66] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.299 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.326 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.351 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.412 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.469 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.487 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [64] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.512 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.536 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.558 [graph_converter.cc:849][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1599] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.601.752 [graph_converter.cc:853][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [184] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.602.376 [graph_converter.cc:857][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [611] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:30.602.497 [graph_converter.cc:862][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [100] micro second. . TotalTime = 0.148753, [20] [parse]: 0.0015192 [symbol_resolve]: 0.0125021, [1] [Cycle 1]: 0.0124236, [1] [resolve]: 0.0124032 [combine_like_graphs]: 7.99999e-07 [graph_reusing]: 3.24e-06 [meta_unpack_prepare]: 0.00016431 [pre_cconv]: 5.99997e-07 [abstract_specialize]: 0.00436121 [pack_expand]: 1.672e-05 [auto_monad]: 8.304e-05 [inline]: 1.67e-06 [pre_auto_parallel]: 1.029e-05 [pipeline_split]: 2.89e-06 [optimize]: 0.125798, [35] [py_interpret_to_execute]: 4.89e-06 [rewriter_before_opt_a]: 0.00019056 [opt_a]: 0.123878, [4] [Cycle 1]: 0.0586607, [30] [expand_dump_flag]: 4.45e-06 [switch_simplify]: 2.642e-05 [a_1]: 0.00073171 [recompute_prepare]: 8.35001e-06 [updatestate_depend_eliminate]: 1.057e-05 [updatestate_assign_eliminate]: 7.01e-06 [updatestate_loads_eliminate]: 6.94999e-06 [parameter_eliminate]: 4.87e-06 [a_2]: 7.493e-05 [accelerated_algorithm]: 5.77e-06 [pynative_shard]: 1.66e-06 [auto_parallel]: 3.74e-06 [parallel]: 8.47e-06 [merge_comm]: 3.66e-06 [allreduce_fusion]: 1.75e-06 [virtual_dataset]: 5.57e-06 [get_grad_eliminate_]: 4.65001e-06 [virtual_output]: 4.37e-06 [merge_forward]: 8.62e-06 [cell_reuse_recompute_pass]: 9.60004e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.231e-05 [meta_fg_expand]: 0.00695672, [1] [Cycle 1]: 0.00317534, [1] [resolve]: 0.00315609 [after_resolve]: 4.922e-05 [a_after_grad]: 0.00013288 [renormalize]: 0.0495244 [real_op_eliminate]: 4.42e-05 [auto_monad_grad]: 7.308e-05 [auto_monad_eliminator]: 8.028e-05 [cse]: 0.00036739 [a_3]: 0.00031022 [Cycle 2]: 0.0549949, [30] [expand_dump_flag]: 3.71e-06 [switch_simplify]: 0.00013224 [a_1]: 0.00164205 [recompute_prepare]: 1.269e-05 [updatestate_depend_eliminate]: 1.718e-05 [updatestate_assign_eliminate]: 1.28e-05 [updatestate_loads_eliminate]: 1.264e-05 [parameter_eliminate]: 4.5e-06 [a_2]: 0.00019534 [accelerated_algorithm]: 1.882e-05 [pynative_shard]: 1.33e-06 [auto_parallel]: 4.89e-06 [parallel]: 4.48e-06 [merge_comm]: 2.57001e-06 [allreduce_fusion]: 1.67e-06 [virtual_dataset]: 1.09e-05 [get_grad_eliminate_]: 1.025e-05 [virtual_output]: 1.063e-05 [merge_forward]: 1.385e-05 [cell_reuse_recompute_pass]: 4.00003e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.341e-05 [meta_fg_expand]: 0.0113564, [5] [Cycle 1]: 0.00031998, [1] [resolve]: 0.00030132 [Cycle 1]: 0.00031321, [1] [resolve]: 0.0002951 [Cycle 1]: 0.00164614, [1] [resolve]: 0.0016274 [Cycle 1]: 0.00031563, [1] [resolve]: 0.00029742 [Cycle 1]: 0.00030865, [1] [resolve]: 0.00029117 [after_resolve]: 7.512e-05 [a_after_grad]: 0.00019617 [renormalize]: 0.0399044 [real_op_eliminate]: 5.978e-05 [auto_monad_grad]: 0.00021988 [auto_monad_eliminator]: 0.00010608 [cse]: 0.00031524 [a_3]: 0.00043153 [Cycle 3]: 0.00615898, [30] [expand_dump_flag]: 4.7e-06 [switch_simplify]: 0.00015899 [a_1]: 0.00235536 [recompute_prepare]: 1.901e-05 [updatestate_depend_eliminate]: 3.288e-05 [updatestate_assign_eliminate]: 1.923e-05 [updatestate_loads_eliminate]: 1.777e-05 [parameter_eliminate]: 5.37001e-06 [a_2]: 0.00027351 [accelerated_algorithm]: 2.756e-05 [pynative_shard]: 2.50999e-06 [auto_parallel]: 8.77e-06 [parallel]: 5.07e-06 [merge_comm]: 3.99e-06 [allreduce_fusion]: 2.34001e-06 [virtual_dataset]: 1.56e-05 [get_grad_eliminate_]: 1.421e-05 [virtual_output]: 1.386e-05 [merge_forward]: 1.996e-05 [cell_reuse_recompute_pass]: 4.69998e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.341e-05 [meta_fg_expand]: 5.649e-05 [after_resolve]: 1.924e-05 [a_after_grad]: 3.259e-05 [renormalize]: 0.00245299 [real_op_eliminate]: 2.183e-05 [auto_monad_grad]: 6.04e-06 [auto_monad_eliminator]: 3.524e-05 [cse]: 0.000208 [a_3]: 0.00012098 [Cycle 4]: 0.00165121, [30] [expand_dump_flag]: 1.3e-06 [switch_simplify]: 1.58e-05 [a_1]: 0.0007403 [recompute_prepare]: 1.552e-05 [updatestate_depend_eliminate]: 1.964e-05 [updatestate_assign_eliminate]: 1.662e-05 [updatestate_loads_eliminate]: 1.653e-05 [parameter_eliminate]: 2.38e-06 [a_2]: 0.00027198 [accelerated_algorithm]: 2.517e-05 [pynative_shard]: 1.61e-06 [auto_parallel]: 4.15e-06 [parallel]: 3.73e-06 [merge_comm]: 3.12e-06 [allreduce_fusion]: 1.99e-06 [virtual_dataset]: 1.534e-05 [get_grad_eliminate_]: 1.501e-05 [virtual_output]: 1.384e-05 [merge_forward]: 1.713e-05 [cell_reuse_recompute_pass]: 4.39999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.289e-05 [meta_fg_expand]: 1.33e-05 [after_resolve]: 1.673e-05 [a_after_grad]: 3.177e-05 [renormalize]: 5.99975e-08 [real_op_eliminate]: 1.445e-05 [auto_monad_grad]: 2.24001e-06 [auto_monad_eliminator]: 2.99e-05 [cse]: 8.262e-05 [a_3]: 0.00011113 [py_interpret_to_execute_after_opt_a]: 4.24e-06 [slice_cell_reuse_recomputed_activation]: 2.62e-06 [rewriter_after_opt_a]: 0.0001001 [convert_after_rewriter]: 2.419e-05 [order_py_execute_after_rewriter]: 1.692e-05 [opt_b]: 0.00110383, [2] [Cycle 1]: 0.00093399, [7] [b_1]: 0.00083746 [b_2]: 4.91001e-06 [updatestate_depend_eliminate]: 6.1e-06 [updatestate_assign_eliminate]: 4.18e-06 [updatestate_loads_eliminate]: 3.93e-06 [renormalize]: 4.39999e-07 [cse]: 3.77e-05 [Cycle 2]: 0.00015988, [7] [b_1]: 9.783e-05 [b_2]: 4.04999e-06 [updatestate_depend_eliminate]: 5.51e-06 [updatestate_assign_eliminate]: 3.8e-06 [updatestate_loads_eliminate]: 3.6e-06 [renormalize]: 6.99947e-08 [cse]: 1.833e-05 [cconv]: 2.315e-05 [opt_after_cconv]: 9.205e-05, [1] [Cycle 1]: 8.443e-05, [7] [c_1]: 2.471e-05 [parameter_eliminate]: 2.17e-06 [updatestate_depend_eliminate]: 4.15e-06 [updatestate_assign_eliminate]: 4.23e-06 [updatestate_loads_eliminate]: 3.54e-06 [cse]: 1.626e-05 [renormalize]: 2.80001e-07 [remove_dup_value]: 1.764e-05 [tuple_transform]: 8.835e-05, [1] [Cycle 1]: 8.451e-05, [3] [d_1]: 5.892e-05 [d_2]: 1.049e-05 [renormalize]: 2.19996e-07 [add_cache_embedding]: 1.345e-05 [add_recomputation]: 5.948e-05 [cse_after_recomputation]: 2.739e-05, [1] [Cycle 1]: 2.342e-05, [1] [cse]: 1.886e-05 [environ_conv]: 1.001e-05 [label_micro_interleaved_index]: 2.65e-06 [label_fine_grained_interleaved_index]: 2.74e-06 [assign_add_opt]: 1.54e-06 [slice_recompute_activation]: 1.91e-06 [micro_interleaved_order_control]: 1.63e-06 [full_micro_interleaved_order_control]: 1.73e-06 [comp_comm_scheduling]: 2.39e-06 [reorder_send_recv_between_fp_bp]: 2.09e-06 [comm_op_add_attrs]: 1e-06 [add_comm_op_reuse_tag]: 1.38e-06 [overlap_opt_shard_in_pipeline]: 1.01e-06 [grouped_pairwise_exchange_alltoall]: 1.72e-06 [overlap_recompute_and_grad_model_parallel]: 1.62001e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.79996e-07 [split_matmul_comm_elemetwise]: 2.08e-06 [split_layernorm_comm]: 1.64e-06 [process_send_recv_for_ge]: 7.59996e-07 [handle_group_info]: 9.5e-07 [auto_monad_reorder]: 2.332e-05 [get_jit_bprop_graph]: 6.16e-06 [eliminate_special_op_node]: 0.00055999 [validate]: 4.19e-05 [distribtued_split]: 1.16e-06 [task_emit]: 0.00343523 [execute]: 6.3e-06 Sums parse : 0.001519s : 1.14% symbol_resolve.resolve : 0.012403s : 9.34% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000164s : 0.12% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004361s : 3.28% pack_expand : 0.000017s : 0.01% auto_monad : 0.000083s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000191s : 0.14% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000333s : 0.25% optimize.opt_a.a_1 : 0.005469s : 4.12% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000080s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000017s : 0.01% optimize.opt_a.a_2 : 0.000816s : 0.61% optimize.opt_a.accelerated_algorithm : 0.000077s : 0.06% optimize.opt_a.pynative_shard : 0.000007s : 0.01% optimize.opt_a.auto_parallel : 0.000022s : 0.02% optimize.opt_a.parallel : 0.000022s : 0.02% optimize.opt_a.merge_comm : 0.000013s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000070s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.005969s : 4.50% optimize.opt_a.after_resolve : 0.000160s : 0.12% optimize.opt_a.a_after_grad : 0.000393s : 0.30% optimize.opt_a.renormalize : 0.091882s : 69.21% optimize.opt_a.real_op_eliminate : 0.000140s : 0.11% optimize.opt_a.auto_monad_grad : 0.000301s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000252s : 0.19% optimize.opt_a.cse : 0.000973s : 0.73% optimize.opt_a.a_3 : 0.000974s : 0.73% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000100s : 0.08% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000935s : 0.70% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000012s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000056s : 0.04% optimize.cconv : 0.000023s : 0.02% optimize.opt_after_cconv.c_1 : 0.000025s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000059s : 0.04% optimize.tuple_transform.d_2 : 0.000010s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000059s : 0.04% optimize.cse_after_recomputation.cse : 0.000019s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000003s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000023s : 0.02% get_jit_bprop_graph : 0.000006s : 0.00% eliminate_special_op_node : 0.000560s : 0.42% validate : 0.000042s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003435s : 2.59% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018638 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.27% : 0.016825s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.58% : 0.001040s : 103: substitution.inline 0.05% : 0.000009s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000034s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.52% : 0.000097s : 69: substitution.replace_applicator 0.06% : 0.000011s : 36: substitution.replace_old_param 0.02% : 0.000004s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000008s : 5: substitution.specialize_transform 0.06% : 0.000012s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.65% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.86% : 0.000161s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000077s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.091864 6 92.15% : 0.084654s : 3: renormalize.infer 7.85% : 0.007210s : 3: renormalize.specialize ------[replace.] 0.001255 141 53.02% : 0.000665s : 55: replace.getattr_setattr_resolve 27.73% : 0.000348s : 56: replace.inline 3.75% : 0.000047s : 2: replace.meta_unpack_prepare 7.29% : 0.000091s : 10: replace.switch_simplify 1.60% : 0.000020s : 4: replace.transpose_eliminate 6.60% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017632 141 94.69% : 0.016696s : 55: match.getattr_setattr_resolve 4.91% : 0.000865s : 56: match.inline 0.11% : 0.000019s : 2: match.meta_unpack_prepare 0.07% : 0.000012s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.007268 119 68.41% : 0.004972s : 53: func_graph_cloner_run.FuncGraphClonerGraph 31.59% : 0.002296s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027648 589 0.51% : 0.000142s : 2: opt.transform.meta_unpack_prepare 30.56% : 0.008448s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.29% : 0.000909s : 94: opt.transform.opt_b 65.28% : 0.018049s : 14: opt.transform.opt_resolve 0.23% : 0.000064s : 8: opt.transform.opt_trans_graph 0.06% : 0.000015s : 3: opt.transform.special_op_eliminate . TotalTime = 0.145463, [20] [parse]: 0.00127417 [symbol_resolve]: 0.0124761, [1] [Cycle 1]: 0.0124114, [1] [resolve]: 0.0123944 [combine_like_graphs]: 7.80004e-07 [graph_reusing]: 3.25e-06 [meta_unpack_prepare]: 0.00012809 [pre_cconv]: 4.69998e-07 [abstract_specialize]: 0.0041042 [pack_expand]: 1.408e-05 [auto_monad]: 7.391e-05 [inline]: 1.39e-06 [pre_auto_parallel]: 7.1e-06 [pipeline_split]: 2.09999e-06 [optimize]: 0.121321, [35] [py_interpret_to_execute]: 4.52e-06 [rewriter_before_opt_a]: 0.00019169 [opt_a]: 0.119499, [4] [Cycle 1]: 0.058708, [30] [expand_dump_flag]: 3.51e-06 [switch_simplify]: 2.757e-05 [a_1]: 0.00038939 [recompute_prepare]: 9.12001e-06 [updatestate_depend_eliminate]: 9e-06 [updatestate_assign_eliminate]: 6.71e-06 [updatestate_loads_eliminate]: 6.22e-06 [parameter_eliminate]: 4.33e-06 [a_2]: 7.836e-05 [accelerated_algorithm]: 5.53e-06 [pynative_shard]: 9.10004e-07 [auto_parallel]: 3.37001e-06 [parallel]: 5.35e-06 [merge_comm]: 2.77e-06 [allreduce_fusion]: 1.67e-06 [virtual_dataset]: 5.31e-06 [get_grad_eliminate_]: 4.46e-06 [virtual_output]: 4.13e-06 [merge_forward]: 7.58e-06 [cell_reuse_recompute_pass]: 4.19997e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.214e-05 [meta_fg_expand]: 0.00702737, [1] [Cycle 1]: 0.00327297, [1] [resolve]: 0.00325415 [after_resolve]: 4.935e-05 [a_after_grad]: 0.00010858 [renormalize]: 0.0498993 [real_op_eliminate]: 4.033e-05 [auto_monad_grad]: 7.199e-05 [auto_monad_eliminator]: 7.834e-05 [cse]: 0.00034683 [a_3]: 0.00031141 [Cycle 2]: 0.0524857, [30] [expand_dump_flag]: 3.73e-06 [switch_simplify]: 0.0001025 [a_1]: 0.00079909 [recompute_prepare]: 1.41e-05 [updatestate_depend_eliminate]: 1.645e-05 [updatestate_assign_eliminate]: 1.251e-05 [updatestate_loads_eliminate]: 1.218e-05 [parameter_eliminate]: 3.8e-06 [a_2]: 0.00019835 [accelerated_algorithm]: 1.782e-05 [pynative_shard]: 1.19999e-06 [auto_parallel]: 4.31e-06 [parallel]: 4.98e-06 [merge_comm]: 2.49e-06 [allreduce_fusion]: 1.49e-06 [virtual_dataset]: 1.066e-05 [get_grad_eliminate_]: 9.46e-06 [virtual_output]: 9.47e-06 [merge_forward]: 1.38e-05 [cell_reuse_recompute_pass]: 5.20005e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.399e-05 [meta_fg_expand]: 0.0114056, [5] [Cycle 1]: 0.0003253, [1] [resolve]: 0.0003074 [Cycle 1]: 0.0003369, [1] [resolve]: 0.00031881 [Cycle 1]: 0.00168293, [1] [resolve]: 0.00166459 [Cycle 1]: 0.00031716, [1] [resolve]: 0.00029917 [Cycle 1]: 0.00035884, [1] [resolve]: 0.00034074 [after_resolve]: 7.174e-05 [a_after_grad]: 0.00016281 [renormalize]: 0.0382873 [real_op_eliminate]: 5.256e-05 [auto_monad_grad]: 0.00020624 [auto_monad_eliminator]: 0.00010664 [cse]: 0.00030775 [a_3]: 0.00043279 [Cycle 3]: 0.00474111, [30] [expand_dump_flag]: 4.29e-06 [switch_simplify]: 0.00011558 [a_1]: 0.00123279 [recompute_prepare]: 1.767e-05 [updatestate_depend_eliminate]: 2.642e-05 [updatestate_assign_eliminate]: 1.766e-05 [updatestate_loads_eliminate]: 1.739e-05 [parameter_eliminate]: 4.22e-06 [a_2]: 0.0002743 [accelerated_algorithm]: 2.462e-05 [pynative_shard]: 1.27e-06 [auto_parallel]: 3.89e-06 [parallel]: 4.2e-06 [merge_comm]: 3.6e-06 [allreduce_fusion]: 2.49e-06 [virtual_dataset]: 1.415e-05 [get_grad_eliminate_]: 1.313e-05 [virtual_output]: 1.275e-05 [merge_forward]: 1.934e-05 [cell_reuse_recompute_pass]: 4.19997e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.3e-05 [meta_fg_expand]: 4.687e-05 [after_resolve]: 1.652e-05 [a_after_grad]: 2.073e-05 [renormalize]: 0.00226958 [real_op_eliminate]: 1.932e-05 [auto_monad_grad]: 5.52e-06 [auto_monad_eliminator]: 3.477e-05 [cse]: 0.00020057 [a_3]: 0.00012192 [Cycle 4]: 0.00118828, [30] [expand_dump_flag]: 1.36e-06 [switch_simplify]: 1.48e-05 [a_1]: 0.00028475 [recompute_prepare]: 1.523e-05 [updatestate_depend_eliminate]: 2.03e-05 [updatestate_assign_eliminate]: 1.716e-05 [updatestate_loads_eliminate]: 1.653e-05 [parameter_eliminate]: 2.22e-06 [a_2]: 0.00027588 [accelerated_algorithm]: 2.482e-05 [pynative_shard]: 1.41e-06 [auto_parallel]: 3.73001e-06 [parallel]: 3.71999e-06 [merge_comm]: 3.11e-06 [allreduce_fusion]: 1.99e-06 [virtual_dataset]: 1.452e-05 [get_grad_eliminate_]: 1.404e-05 [virtual_output]: 1.305e-05 [merge_forward]: 1.736e-05 [cell_reuse_recompute_pass]: 4.20005e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.367e-05 [meta_fg_expand]: 1.312e-05 [after_resolve]: 1.574e-05 [a_after_grad]: 2.068e-05 [renormalize]: 1.00001e-07 [real_op_eliminate]: 1.387e-05 [auto_monad_grad]: 2.31e-06 [auto_monad_eliminator]: 2.992e-05 [cse]: 8.302e-05 [a_3]: 0.00011275 [py_interpret_to_execute_after_opt_a]: 4.21e-06 [slice_cell_reuse_recomputed_activation]: 1.34e-06 [rewriter_after_opt_a]: 9.485e-05 [convert_after_rewriter]: 2.234e-05 [order_py_execute_after_rewriter]: 1.593e-05 [opt_b]: 0.00110372, [2] [Cycle 1]: 0.00093321, [7] [b_1]: 0.00083942 [b_2]: 5.74e-06 [updatestate_depend_eliminate]: 6.06e-06 [updatestate_assign_eliminate]: 4.04e-06 [updatestate_loads_eliminate]: 4.33e-06 [renormalize]: 4.1e-07 [cse]: 3.744e-05 [Cycle 2]: 0.00016096, [7] [b_1]: 9.855e-05 [b_2]: 4.14e-06 [updatestate_depend_eliminate]: 4.85e-06 [updatestate_assign_eliminate]: 3.9e-06 [updatestate_loads_eliminate]: 3.64e-06 [renormalize]: 6.00048e-08 [cse]: 1.883e-05 [cconv]: 1.723e-05 [opt_after_cconv]: 7.069e-05, [1] [Cycle 1]: 6.645e-05, [7] [c_1]: 8.59e-06 [parameter_eliminate]: 2.32e-06 [updatestate_depend_eliminate]: 4.44e-06 [updatestate_assign_eliminate]: 3.66e-06 [updatestate_loads_eliminate]: 3.61e-06 [cse]: 1.701e-05 [renormalize]: 3.30001e-07 [remove_dup_value]: 1.29e-05 [tuple_transform]: 6.552e-05, [1] [Cycle 1]: 6.173e-05, [3] [d_1]: 4.008e-05 [d_2]: 8.78e-06 [renormalize]: 2.00002e-07 [add_cache_embedding]: 9.69e-06 [add_recomputation]: 5.032e-05 [cse_after_recomputation]: 2.51e-05, [1] [Cycle 1]: 2.125e-05, [1] [cse]: 1.694e-05 [environ_conv]: 8.27e-06 [label_micro_interleaved_index]: 1.89e-06 [label_fine_grained_interleaved_index]: 1.29e-06 [assign_add_opt]: 8.49999e-07 [slice_recompute_activation]: 1.57e-06 [micro_interleaved_order_control]: 1.13e-06 [full_micro_interleaved_order_control]: 1.13e-06 [comp_comm_scheduling]: 1.2e-06 [reorder_send_recv_between_fp_bp]: 1.50999e-06 [comm_op_add_attrs]: 4.60001e-07 [add_comm_op_reuse_tag]: 7.20007e-07 [overlap_opt_shard_in_pipeline]: 6.29996e-07 [grouped_pairwise_exchange_alltoall]: 5.69999e-07 [overlap_recompute_and_grad_model_parallel]: 1.26e-06 [overlap_grad_matmul_and_grad_allreduce]: 5.19998e-07 [split_matmul_comm_elemetwise]: 1.55e-06 [split_layernorm_comm]: 1.49e-06 [process_send_recv_for_ge]: 1.12e-06 [handle_group_info]: 5.30003e-07 [auto_monad_reorder]: 1.796e-05 [get_jit_bprop_graph]: 3.80001e-07 [eliminate_special_op_node]: 0.00052416 [validate]: 3.566e-05 [distribtued_split]: 1.13e-06 [task_emit]: 0.00529005 [execute]: 5.24e-06 Sums parse : 0.001274s : 0.98% symbol_resolve.resolve : 0.012394s : 9.56% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000128s : 0.10% pre_cconv : 0.000000s : 0.00% abstract_specialize : 0.004104s : 3.16% pack_expand : 0.000014s : 0.01% auto_monad : 0.000074s : 0.06% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000007s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000192s : 0.15% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000260s : 0.20% optimize.opt_a.a_1 : 0.002706s : 2.09% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000072s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000054s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000052s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000827s : 0.64% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000015s : 0.01% optimize.opt_a.parallel : 0.000018s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000039s : 0.03% optimize.opt_a.merge_forward : 0.000058s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000103s : 0.08% optimize.opt_a.meta_fg_expand : 0.000060s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006185s : 4.77% optimize.opt_a.after_resolve : 0.000153s : 0.12% optimize.opt_a.a_after_grad : 0.000313s : 0.24% optimize.opt_a.renormalize : 0.090456s : 69.75% optimize.opt_a.real_op_eliminate : 0.000126s : 0.10% optimize.opt_a.auto_monad_grad : 0.000286s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000250s : 0.19% optimize.opt_a.cse : 0.000938s : 0.72% optimize.opt_a.a_3 : 0.000979s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000001s : 0.00% optimize.rewriter_after_opt_a : 0.000095s : 0.07% optimize.convert_after_rewriter : 0.000022s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000938s : 0.72% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000056s : 0.04% optimize.cconv : 0.000017s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000040s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000010s : 0.01% optimize.add_recomputation : 0.000050s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000008s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000001s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000000s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000018s : 0.01% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000524s : 0.40% validate : 0.000036s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005290s : 4.08% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018766 880 0.01% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.55% : 0.016992s : 59: substitution.getattr_setattr_resolve 0.02% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.74% : 0.001078s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000028s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.01% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.04% : 0.000008s : 81: substitution.remove_not_recompute_node 0.47% : 0.000089s : 63: substitution.replace_applicator 0.05% : 0.000009s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.60% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_item_depend_reorder 0.82% : 0.000154s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.090441 6 92.59% : 0.083738s : 3: renormalize.infer 7.41% : 0.006703s : 3: renormalize.specialize ------[replace.] 0.001273 141 55.31% : 0.000704s : 55: replace.getattr_setattr_resolve 25.89% : 0.000330s : 56: replace.inline 3.59% : 0.000046s : 2: replace.meta_unpack_prepare 7.25% : 0.000092s : 10: replace.switch_simplify 1.41% : 0.000018s : 4: replace.transpose_eliminate 6.55% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017840 141 94.53% : 0.016863s : 55: match.getattr_setattr_resolve 5.10% : 0.000910s : 56: match.inline 0.09% : 0.000017s : 2: match.meta_unpack_prepare 0.05% : 0.000010s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006854 119 69.25% : 0.004746s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.75% : 0.002108s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.024982 259 7.43% : 0.001855s : 104: opt.transform.opt_a 3.63% : 0.000906s : 92: opt.transform.opt_b 73.09% : 0.018260s : 14: opt.transform.opt_resolve 0.45% : 0.000111s : 1: opt.transforms.meta_unpack_prepare 15.10% : 0.003773s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000047s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:30.990.906 [graph_var_manager.cc:1424][EVENT]36564 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:30.990.985 [graph_manager.cc:1248][EVENT]36564 PreRun:PreRun start: graph node size 3, session id 7, graph id 6, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.991.339 [atrace_api.c:28](tid:36564) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.991.372 [trace_rb_log.c:84](tid:36564) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:30.991.385 [atrace_api.c:32](tid:36564) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:30.991.398 [client_manager.cpp:157][SetProfilingCallback][tid:36564] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:31.737.919 [parallel_partitioner.cc:165][EVENT]36564 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [22] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.737.983 [parallel_partitioner.cc:178][EVENT]36564 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.060 [graph_prepare.cc:1378][EVENT]36564 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.272 [graph_manager.cc:1050][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [232] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.298 [graph_manager.cc:1052][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.435 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.467 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.525 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [44] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.539 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.602 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.616 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.634 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.733 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.754 [graph_manager.cc:1054][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [441] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.738.981 [graph_manager.cc:1055][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [214] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.739.982 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.005 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.016 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.026 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [299] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.035 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.044 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.052 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.061 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [17] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.740.069 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.741.571 [graph_manager.cc:1056][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2571] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.741.633 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.741.650 [graph_prepare.cc:1982][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [49] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.047 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.067 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.078 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.087 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [222] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.096 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.104 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.113 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.121 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.130 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.155 [graph_prepare.cc:1983][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [491] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.180 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.194 [graph_prepare.cc:1984][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [24] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.209 [graph_prepare.cc:1985][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.223 [graph_prepare.cc:1986][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.234 [graph_prepare.cc:1987][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.249 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.261 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.275 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.357 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.377 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.386 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrintOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.394 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.403 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.411 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.419 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.427 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.436 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.444 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.452 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.460 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SnapshotPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.468 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.476 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.485 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.493 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.515 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.527 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.558 [graph_prepare.cc:1988][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [316] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.742.571 [graph_manager.cc:1065][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [972] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.755.967 [graph_manager.cc:1077][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [13376] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.756.033 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.756.090 [graph_manager.cc:1080][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [91] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.861 [graph_manager.cc:1081][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2748] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.898 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.913 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.925 [graph_manager.cc:1082][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.955 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.971 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.758.985 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.064 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [69] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.082 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.115 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.130 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.170 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.189 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.207 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.234 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.249 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.260 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.270 [graph_manager.cc:2700][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [319] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.376 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.389 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AddNPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.398 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.407 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.423 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.432 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CastRemovePass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.440 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.448 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.457 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.465 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.473 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.481 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.489 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.498 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.506 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.515 [graph_manager.cc:2741][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [227] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.524 [graph_manager.cc:2752][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.548 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.561 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.578 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.594 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.606 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.618 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.639 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.655 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.668 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.679 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.697 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.708 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.727 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [11] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.741 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.750 [graph_manager.cc:2810][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [207] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.780 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.792 [graph_manager.cc:2821][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.821 [graph_manager.cc:1087][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [877] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.958 [graph_manager.cc:1088][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [122] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.759.996 [graph_manager.cc:1089][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.014 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.030 [graph_manager.cc:1097][EVENT]36564 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.052 [graph_manager.cc:3325][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.305 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [12] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.322 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [11] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.331 [engine_place.cc:144][EVENT]36564 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [119] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.402 [graph_manager.cc:3351][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [337] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.420 [graph_manager.cc:3364][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.482 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.498 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.648 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [140] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.699 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [30] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.746 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [36] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.779 [graph_manager.cc:3405][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [346] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.760.797 [graph_manager.cc:3412][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.264 [graph_manager.cc:3422][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1453] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.293 [graph_manager.cc:3428][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.417 [graph_manager.cc:3467][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [104] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.435 [graph_manager.cc:3377][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [2003] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.451 [graph_manager.cc:1106][EVENT]36564 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2406] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.464 [graph_manager.cc:1115][EVENT]36564 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.486 [graph_manager.cc:1130][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.520 [graph_manager.cc:1131][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.545 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.561 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.571 [graph_manager.cc:2837][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.642 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.655 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.665 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.673 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.682 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.690 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.707 [graph_manager.cc:2864][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [119] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.719 [graph_manager.cc:2872][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.739 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.753 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.769 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.785 [compile_nodes_pass.cc:88][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.796 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.806 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.883 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [68] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.911 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.924 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.937 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.950 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.959 [graph_manager.cc:2927][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [223] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.971 [graph_manager.cc:2937][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.984 [graph_manager.cc:2943][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.762.996 [graph_manager.cc:2950][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.200 [graph_manager.cc:2958][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [36] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.229 [graph_manager.cc:1132][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [694] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.309 [graph_manager.cc:1135][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [66] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.341 [graph_manager.cc:2975][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.379 [graph_manager.cc:2981][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.394 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.404 [graph_manager.cc:2986][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [12] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.413 [graph_manager.cc:1136][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [88] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.519 [graph_manager.cc:3555][EVENT]36564 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [74] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.600 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.615 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.725 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [100] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.754 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.793 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.763.815 [graph_builder.cc:865][EVENT]36564 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [245] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:31.764.168 [logger.cc:1071] 36564 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.198 [task_generator.cc:804][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [71] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.255 [task_generator.cc:805][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [45] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.698 [task_generator.cc:814][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [428] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.712 [task_generator.cc:954][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [585] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.766 [task_generator.cc:967][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [31] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:31.764.783 [logger.cc:1084] 36564 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.929 [graph_manager.cc:1152][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1492] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.947 [graph_manager.cc:1164][EVENT]36564 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.978 [graph_manager.cc:1271][EVENT]36564 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [27165] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.764.997 [graph_manager.cc:1272][EVENT]36564 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:31.765.328 [atrace_api.c:93](tid:36564) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:31.765.345 [atrace_api.c:95](tid:36564) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:31.769.599 [graph_converter.cc:838][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1182] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.769.787 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [148] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.161 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [352] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.243 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.258 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [77] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.543 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [274] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.647 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.680 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.831 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [138] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.898 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [53] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.910 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [65] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.938 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.965 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.770.992 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.052 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.109 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.120 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [57] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.145 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.169 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.189 [graph_converter.cc:849][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1555] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.771.385 [graph_converter.cc:853][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [185] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.772.016 [graph_converter.cc:857][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [618] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:31.772.145 [graph_converter.cc:862][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [100] micro second. . TotalTime = 0.147158, [20] [parse]: 0.00149493 [symbol_resolve]: 0.0124893, [1] [Cycle 1]: 0.0124116, [1] [resolve]: 0.0123918 [combine_like_graphs]: 9.79999e-07 [graph_reusing]: 3.52001e-06 [meta_unpack_prepare]: 0.00016652 [pre_cconv]: 8.39995e-07 [abstract_specialize]: 0.00438164 [pack_expand]: 1.663e-05 [auto_monad]: 0.00012542 [inline]: 1.49e-06 [pre_auto_parallel]: 1.054e-05 [pipeline_split]: 2.56e-06 [optimize]: 0.124154, [35] [py_interpret_to_execute]: 4.46e-06 [rewriter_before_opt_a]: 0.000194 [opt_a]: 0.122239, [4] [Cycle 1]: 0.0589241, [30] [expand_dump_flag]: 5.08e-06 [switch_simplify]: 2.932e-05 [a_1]: 0.00075413 [recompute_prepare]: 8.46e-06 [updatestate_depend_eliminate]: 1.085e-05 [updatestate_assign_eliminate]: 7.21e-06 [updatestate_loads_eliminate]: 6.67e-06 [parameter_eliminate]: 4.7e-06 [a_2]: 7.49e-05 [accelerated_algorithm]: 5.46e-06 [pynative_shard]: 1.69e-06 [auto_parallel]: 3.44e-06 [parallel]: 9.19e-06 [merge_comm]: 3.95e-06 [allreduce_fusion]: 2.08e-06 [virtual_dataset]: 5.29e-06 [get_grad_eliminate_]: 4.61e-06 [virtual_output]: 4.31e-06 [merge_forward]: 9.40001e-06 [cell_reuse_recompute_pass]: 1.04e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.374e-05 [meta_fg_expand]: 0.00704527, [1] [Cycle 1]: 0.00322155, [1] [resolve]: 0.00320338 [after_resolve]: 4.945e-05 [a_after_grad]: 0.00013205 [renormalize]: 0.0496481 [real_op_eliminate]: 4.452e-05 [auto_monad_grad]: 7.401e-05 [auto_monad_eliminator]: 8.01e-05 [cse]: 0.00038127 [a_3]: 0.00031423 [Cycle 2]: 0.0532745, [30] [expand_dump_flag]: 3.91e-06 [switch_simplify]: 0.00013594 [a_1]: 0.00162839 [recompute_prepare]: 1.312e-05 [updatestate_depend_eliminate]: 1.708e-05 [updatestate_assign_eliminate]: 1.291e-05 [updatestate_loads_eliminate]: 1.303e-05 [parameter_eliminate]: 4.01e-06 [a_2]: 0.00019615 [accelerated_algorithm]: 1.899e-05 [pynative_shard]: 1.83e-06 [auto_parallel]: 5.78e-06 [parallel]: 5.2e-06 [merge_comm]: 2.53e-06 [allreduce_fusion]: 1.49e-06 [virtual_dataset]: 1.095e-05 [get_grad_eliminate_]: 1.022e-05 [virtual_output]: 1.068e-05 [merge_forward]: 1.369e-05 [cell_reuse_recompute_pass]: 4.79995e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.302e-05 [meta_fg_expand]: 0.0113918, [5] [Cycle 1]: 0.00031872, [1] [resolve]: 0.0003009 [Cycle 1]: 0.00030948, [1] [resolve]: 0.00029128 [Cycle 1]: 0.00168411, [1] [resolve]: 0.00166539 [Cycle 1]: 0.00031182, [1] [resolve]: 0.00029357 [Cycle 1]: 0.00030813, [1] [resolve]: 0.00028982 [after_resolve]: 7.695e-05 [a_after_grad]: 0.00019827 [renormalize]: 0.0381749 [real_op_eliminate]: 5.749e-05 [auto_monad_grad]: 0.00021894 [auto_monad_eliminator]: 0.0001066 [cse]: 0.00030944 [a_3]: 0.00042409 [Cycle 3]: 0.00598456, [30] [expand_dump_flag]: 4.05e-06 [switch_simplify]: 0.00015659 [a_1]: 0.00233045 [recompute_prepare]: 1.682e-05 [updatestate_depend_eliminate]: 3.128e-05 [updatestate_assign_eliminate]: 1.889e-05 [updatestate_loads_eliminate]: 1.786e-05 [parameter_eliminate]: 4.64e-06 [a_2]: 0.00026984 [accelerated_algorithm]: 2.506e-05 [pynative_shard]: 1.47e-06 [auto_parallel]: 4.41e-06 [parallel]: 4.95e-06 [merge_comm]: 3.95e-06 [allreduce_fusion]: 2.43e-06 [virtual_dataset]: 1.483e-05 [get_grad_eliminate_]: 1.408e-05 [virtual_output]: 1.404e-05 [merge_forward]: 1.948e-05 [cell_reuse_recompute_pass]: 5.10001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.284e-05 [meta_fg_expand]: 4.889e-05 [after_resolve]: 1.798e-05 [a_after_grad]: 3.211e-05 [renormalize]: 0.00234524 [real_op_eliminate]: 2.167e-05 [auto_monad_grad]: 5.61e-06 [auto_monad_eliminator]: 3.51e-05 [cse]: 0.00020825 [a_3]: 0.00011943 [Cycle 4]: 0.00163851, [30] [expand_dump_flag]: 1.26e-06 [switch_simplify]: 1.52e-05 [a_1]: 0.00073369 [recompute_prepare]: 1.527e-05 [updatestate_depend_eliminate]: 2.003e-05 [updatestate_assign_eliminate]: 1.68e-05 [updatestate_loads_eliminate]: 1.624e-05 [parameter_eliminate]: 2.23e-06 [a_2]: 0.0002725 [accelerated_algorithm]: 2.488e-05 [pynative_shard]: 1.51e-06 [auto_parallel]: 3.8e-06 [parallel]: 3.65001e-06 [merge_comm]: 3.21e-06 [allreduce_fusion]: 2.08e-06 [virtual_dataset]: 1.498e-05 [get_grad_eliminate_]: 1.461e-05 [virtual_output]: 1.399e-05 [merge_forward]: 1.73e-05 [cell_reuse_recompute_pass]: 4.1e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.358e-05 [meta_fg_expand]: 1.357e-05 [after_resolve]: 1.645e-05 [a_after_grad]: 3.133e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.43e-05 [auto_monad_grad]: 2.19e-06 [auto_monad_eliminator]: 3.085e-05 [cse]: 7.988e-05 [a_3]: 0.00010928 [py_interpret_to_execute_after_opt_a]: 4.48e-06 [slice_cell_reuse_recomputed_activation]: 2e-06 [rewriter_after_opt_a]: 9.853e-05 [convert_after_rewriter]: 2.33e-05 [order_py_execute_after_rewriter]: 1.671e-05 [opt_b]: 0.00109998, [2] [Cycle 1]: 0.00093198, [7] [b_1]: 0.00083955 [b_2]: 5.08e-06 [updatestate_depend_eliminate]: 6.03e-06 [updatestate_assign_eliminate]: 4.23e-06 [updatestate_loads_eliminate]: 3.88e-06 [renormalize]: 4.20005e-07 [cse]: 3.724e-05 [Cycle 2]: 0.00015826, [7] [b_1]: 9.77e-05 [b_2]: 3.82e-06 [updatestate_depend_eliminate]: 5.15e-06 [updatestate_assign_eliminate]: 3.83e-06 [updatestate_loads_eliminate]: 3.62e-06 [renormalize]: 6.00048e-08 [cse]: 1.771e-05 [cconv]: 2.163e-05 [opt_after_cconv]: 8.65e-05, [1] [Cycle 1]: 8.202e-05, [7] [c_1]: 2.411e-05 [parameter_eliminate]: 2.2e-06 [updatestate_depend_eliminate]: 4.33e-06 [updatestate_assign_eliminate]: 4.44e-06 [updatestate_loads_eliminate]: 3.75e-06 [cse]: 1.576e-05 [renormalize]: 3.99996e-07 [remove_dup_value]: 1.724e-05 [tuple_transform]: 9.736e-05, [1] [Cycle 1]: 9.36e-05, [3] [d_1]: 7.079e-05 [d_2]: 9.62e-06 [renormalize]: 1.99994e-07 [add_cache_embedding]: 1.445e-05 [add_recomputation]: 6.025e-05 [cse_after_recomputation]: 2.635e-05, [1] [Cycle 1]: 2.206e-05, [1] [cse]: 1.751e-05 [environ_conv]: 1.011e-05 [label_micro_interleaved_index]: 2.52e-06 [label_fine_grained_interleaved_index]: 2.60001e-06 [assign_add_opt]: 1.37999e-06 [slice_recompute_activation]: 1.99e-06 [micro_interleaved_order_control]: 1.56e-06 [full_micro_interleaved_order_control]: 1.75e-06 [comp_comm_scheduling]: 2.09e-06 [reorder_send_recv_between_fp_bp]: 2.04e-06 [comm_op_add_attrs]: 1.02e-06 [add_comm_op_reuse_tag]: 8.49999e-07 [overlap_opt_shard_in_pipeline]: 1.11e-06 [grouped_pairwise_exchange_alltoall]: 1.42e-06 [overlap_recompute_and_grad_model_parallel]: 1.61e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.06001e-06 [split_matmul_comm_elemetwise]: 2.36e-06 [split_layernorm_comm]: 1.60999e-06 [process_send_recv_for_ge]: 7.30004e-07 [handle_group_info]: 9.29998e-07 [auto_monad_reorder]: 2.353e-05 [get_jit_bprop_graph]: 3.31e-06 [eliminate_special_op_node]: 0.00051592 [validate]: 3.803e-05 [distribtued_split]: 1.44001e-06 [task_emit]: 0.00352787 [execute]: 6.37e-06 Sums parse : 0.001495s : 1.14% symbol_resolve.resolve : 0.012392s : 9.45% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000004s : 0.00% meta_unpack_prepare : 0.000167s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004382s : 3.34% pack_expand : 0.000017s : 0.01% auto_monad : 0.000125s : 0.10% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000011s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000194s : 0.15% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000337s : 0.26% optimize.opt_a.a_1 : 0.005447s : 4.15% optimize.opt_a.recompute_prepare : 0.000054s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000079s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000813s : 0.62% optimize.opt_a.accelerated_algorithm : 0.000074s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000023s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000046s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000103s : 0.08% optimize.opt_a.meta_fg_expand : 0.000062s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006044s : 4.61% optimize.opt_a.after_resolve : 0.000161s : 0.12% optimize.opt_a.a_after_grad : 0.000394s : 0.30% optimize.opt_a.renormalize : 0.090168s : 68.74% optimize.opt_a.real_op_eliminate : 0.000138s : 0.11% optimize.opt_a.auto_monad_grad : 0.000301s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000253s : 0.19% optimize.opt_a.cse : 0.000979s : 0.75% optimize.opt_a.a_3 : 0.000967s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000099s : 0.08% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000937s : 0.71% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000007s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000055s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000071s : 0.05% optimize.tuple_transform.d_2 : 0.000010s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000014s : 0.01% optimize.add_recomputation : 0.000060s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000003s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000024s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000516s : 0.39% validate : 0.000038s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003528s : 2.69% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018668 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.23% : 0.016844s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.60% : 0.001045s : 103: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.21% : 0.000039s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000010s : 81: substitution.remove_not_recompute_node 0.51% : 0.000094s : 69: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000012s : 10: substitution.switch_simplify 0.06% : 0.000011s : 4: substitution.transpose_eliminate 0.65% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.86% : 0.000161s : 122: substitution.tuple_list_get_item_eliminator 0.47% : 0.000088s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.090152 6 92.43% : 0.083328s : 3: renormalize.infer 7.57% : 0.006824s : 3: renormalize.specialize ------[replace.] 0.001253 141 54.57% : 0.000684s : 55: replace.getattr_setattr_resolve 25.78% : 0.000323s : 56: replace.inline 3.76% : 0.000047s : 2: replace.meta_unpack_prepare 7.45% : 0.000093s : 10: replace.switch_simplify 1.60% : 0.000020s : 4: replace.transpose_eliminate 6.84% : 0.000086s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017654 141 94.68% : 0.016715s : 55: match.getattr_setattr_resolve 4.92% : 0.000869s : 56: match.inline 0.11% : 0.000019s : 2: match.meta_unpack_prepare 0.07% : 0.000012s : 10: match.switch_simplify 0.06% : 0.000011s : 4: match.transpose_eliminate 0.16% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.007019 119 69.26% : 0.004861s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.74% : 0.002158s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027673 589 0.52% : 0.000143s : 2: opt.transform.meta_unpack_prepare 30.43% : 0.008421s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.29% : 0.000912s : 94: opt.transform.opt_b 65.36% : 0.018086s : 14: opt.transform.opt_resolve 0.27% : 0.000075s : 8: opt.transform.opt_trans_graph 0.06% : 0.000016s : 3: opt.transform.special_op_eliminate . TotalTime = 0.146588, [20] [parse]: 0.00131588 [symbol_resolve]: 0.0124554, [1] [Cycle 1]: 0.0123893, [1] [resolve]: 0.012372 [combine_like_graphs]: 7.59996e-07 [graph_reusing]: 3.24e-06 [meta_unpack_prepare]: 0.00012849 [pre_cconv]: 5.10001e-07 [abstract_specialize]: 0.00407952 [pack_expand]: 1.524e-05 [auto_monad]: 7.191e-05 [inline]: 1.36e-06 [pre_auto_parallel]: 7.38e-06 [pipeline_split]: 2e-06 [optimize]: 0.121998, [35] [py_interpret_to_execute]: 4.39e-06 [rewriter_before_opt_a]: 0.00019034 [opt_a]: 0.12014, [4] [Cycle 1]: 0.0587405, [30] [expand_dump_flag]: 3.41e-06 [switch_simplify]: 2.714e-05 [a_1]: 0.00039029 [recompute_prepare]: 9.21e-06 [updatestate_depend_eliminate]: 9.17001e-06 [updatestate_assign_eliminate]: 6.29e-06 [updatestate_loads_eliminate]: 5.82e-06 [parameter_eliminate]: 4.4e-06 [a_2]: 7.898e-05 [accelerated_algorithm]: 5.25e-06 [pynative_shard]: 9e-07 [auto_parallel]: 3.45999e-06 [parallel]: 5.98e-06 [merge_comm]: 3.18e-06 [allreduce_fusion]: 1.66e-06 [virtual_dataset]: 5.83e-06 [get_grad_eliminate_]: 4.47001e-06 [virtual_output]: 4.14e-06 [merge_forward]: 7.18e-06 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.218e-05 [meta_fg_expand]: 0.00699698, [1] [Cycle 1]: 0.0032428, [1] [resolve]: 0.00322415 [after_resolve]: 4.952e-05 [a_after_grad]: 0.00011174 [renormalize]: 0.0499537 [real_op_eliminate]: 4.108e-05 [auto_monad_grad]: 7.342e-05 [auto_monad_eliminator]: 7.823e-05 [cse]: 0.00034974 [a_3]: 0.00031263 [Cycle 2]: 0.0529059, [30] [expand_dump_flag]: 3.85e-06 [switch_simplify]: 0.00010188 [a_1]: 0.00078777 [recompute_prepare]: 1.41e-05 [updatestate_depend_eliminate]: 1.606e-05 [updatestate_assign_eliminate]: 1.304e-05 [updatestate_loads_eliminate]: 1.308e-05 [parameter_eliminate]: 3.97e-06 [a_2]: 0.00019959 [accelerated_algorithm]: 1.796e-05 [pynative_shard]: 1.19e-06 [auto_parallel]: 4.32e-06 [parallel]: 4.54e-06 [merge_comm]: 2.24001e-06 [allreduce_fusion]: 1.41e-06 [virtual_dataset]: 1.063e-05 [get_grad_eliminate_]: 9.49e-06 [virtual_output]: 9.61e-06 [merge_forward]: 1.431e-05 [cell_reuse_recompute_pass]: 4.19997e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.393e-05 [meta_fg_expand]: 0.0115186, [5] [Cycle 1]: 0.00032207, [1] [resolve]: 0.00030428 [Cycle 1]: 0.00032015, [1] [resolve]: 0.00030222 [Cycle 1]: 0.0017141, [1] [resolve]: 0.00169525 [Cycle 1]: 0.00032093, [1] [resolve]: 0.0003027 [Cycle 1]: 0.00031978, [1] [resolve]: 0.00030226 [after_resolve]: 7.276e-05 [a_after_grad]: 0.00016267 [renormalize]: 0.0385389 [real_op_eliminate]: 5.757e-05 [auto_monad_grad]: 0.00022418 [auto_monad_eliminator]: 0.00010825 [cse]: 0.00032009 [a_3]: 0.00045465 [Cycle 3]: 0.00494918, [30] [expand_dump_flag]: 5.09001e-06 [switch_simplify]: 0.00011985 [a_1]: 0.00125435 [recompute_prepare]: 1.803e-05 [updatestate_depend_eliminate]: 2.871e-05 [updatestate_assign_eliminate]: 1.926e-05 [updatestate_loads_eliminate]: 1.804e-05 [parameter_eliminate]: 4.42e-06 [a_2]: 0.00027619 [accelerated_algorithm]: 2.572e-05 [pynative_shard]: 1.38e-06 [auto_parallel]: 5.71e-06 [parallel]: 5.3e-06 [merge_comm]: 4.44e-06 [allreduce_fusion]: 2.73e-06 [virtual_dataset]: 1.428e-05 [get_grad_eliminate_]: 1.317e-05 [virtual_output]: 1.293e-05 [merge_forward]: 2.014e-05 [cell_reuse_recompute_pass]: 4.39999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.347e-05 [meta_fg_expand]: 5.182e-05 [after_resolve]: 1.695e-05 [a_after_grad]: 2.072e-05 [renormalize]: 0.00241403 [real_op_eliminate]: 2.027e-05 [auto_monad_grad]: 5.96e-06 [auto_monad_eliminator]: 3.59e-05 [cse]: 0.00021276 [a_3]: 0.0001227 [Cycle 4]: 0.0011842, [30] [expand_dump_flag]: 1.39e-06 [switch_simplify]: 1.512e-05 [a_1]: 0.00028311 [recompute_prepare]: 1.558e-05 [updatestate_depend_eliminate]: 1.966e-05 [updatestate_assign_eliminate]: 1.684e-05 [updatestate_loads_eliminate]: 1.7e-05 [parameter_eliminate]: 2.11e-06 [a_2]: 0.0002726 [accelerated_algorithm]: 2.557e-05 [pynative_shard]: 1.42e-06 [auto_parallel]: 4.45e-06 [parallel]: 3.64e-06 [merge_comm]: 3.25e-06 [allreduce_fusion]: 2.12e-06 [virtual_dataset]: 1.448e-05 [get_grad_eliminate_]: 1.362e-05 [virtual_output]: 1.306e-05 [merge_forward]: 1.735e-05 [cell_reuse_recompute_pass]: 4.30002e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.369e-05 [meta_fg_expand]: 1.337e-05 [after_resolve]: 1.574e-05 [a_after_grad]: 2.069e-05 [renormalize]: 7.99992e-08 [real_op_eliminate]: 1.34e-05 [auto_monad_grad]: 2.41e-06 [auto_monad_eliminator]: 3.064e-05 [cse]: 8.209e-05 [a_3]: 0.00011274 [py_interpret_to_execute_after_opt_a]: 4.51e-06 [slice_cell_reuse_recomputed_activation]: 2.11e-06 [rewriter_after_opt_a]: 0.00010096 [convert_after_rewriter]: 2.298e-05 [order_py_execute_after_rewriter]: 1.667e-05 [opt_b]: 0.00110253, [2] [Cycle 1]: 0.00093537, [7] [b_1]: 0.00083934 [b_2]: 5.86001e-06 [updatestate_depend_eliminate]: 6.18e-06 [updatestate_assign_eliminate]: 4.37e-06 [updatestate_loads_eliminate]: 4.14001e-06 [renormalize]: 3.80001e-07 [cse]: 3.844e-05 [Cycle 2]: 0.00015776, [7] [b_1]: 9.78e-05 [b_2]: 4.29e-06 [updatestate_depend_eliminate]: 4.68e-06 [updatestate_assign_eliminate]: 3.82e-06 [updatestate_loads_eliminate]: 3.56e-06 [renormalize]: 6.00048e-08 [cse]: 1.735e-05 [cconv]: 2.069e-05 [opt_after_cconv]: 7.138e-05, [1] [Cycle 1]: 6.685e-05, [7] [c_1]: 8.66e-06 [parameter_eliminate]: 2.12e-06 [updatestate_depend_eliminate]: 4.52e-06 [updatestate_assign_eliminate]: 3.77e-06 [updatestate_loads_eliminate]: 3.58e-06 [cse]: 1.671e-05 [renormalize]: 3.80001e-07 [remove_dup_value]: 1.621e-05 [tuple_transform]: 6.827e-05, [1] [Cycle 1]: 6.451e-05, [3] [d_1]: 4.208e-05 [d_2]: 9.29e-06 [renormalize]: 1.60006e-07 [add_cache_embedding]: 1.14e-05 [add_recomputation]: 5.81e-05 [cse_after_recomputation]: 2.654e-05, [1] [Cycle 1]: 2.194e-05, [1] [cse]: 1.736e-05 [environ_conv]: 1.032e-05 [label_micro_interleaved_index]: 1.91999e-06 [label_fine_grained_interleaved_index]: 1.89e-06 [assign_add_opt]: 1.39e-06 [slice_recompute_activation]: 1.92e-06 [micro_interleaved_order_control]: 1.60999e-06 [full_micro_interleaved_order_control]: 1.69e-06 [comp_comm_scheduling]: 2.23e-06 [reorder_send_recv_between_fp_bp]: 2.09e-06 [comm_op_add_attrs]: 1.01e-06 [add_comm_op_reuse_tag]: 8.49999e-07 [overlap_opt_shard_in_pipeline]: 8.49999e-07 [grouped_pairwise_exchange_alltoall]: 9.89996e-07 [overlap_recompute_and_grad_model_parallel]: 1.63e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.90001e-07 [split_matmul_comm_elemetwise]: 1.82e-06 [split_layernorm_comm]: 1.67e-06 [process_send_recv_for_ge]: 6.59995e-07 [handle_group_info]: 9.29998e-07 [auto_monad_reorder]: 3.653e-05 [get_jit_bprop_graph]: 4.1e-07 [eliminate_special_op_node]: 0.00052157 [validate]: 3.851e-05 [distribtued_split]: 1.19e-06 [task_emit]: 0.00571972 [execute]: 6.71e-06 Sums parse : 0.001316s : 1.01% symbol_resolve.resolve : 0.012372s : 9.47% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000128s : 0.10% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004080s : 3.12% pack_expand : 0.000015s : 0.01% auto_monad : 0.000072s : 0.06% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000007s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000190s : 0.15% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000264s : 0.20% optimize.opt_a.a_1 : 0.002716s : 2.08% optimize.opt_a.recompute_prepare : 0.000057s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000074s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000055s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000827s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000018s : 0.01% optimize.opt_a.parallel : 0.000019s : 0.01% optimize.opt_a.merge_comm : 0.000013s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.03% optimize.opt_a.virtual_output : 0.000040s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000103s : 0.08% optimize.opt_a.meta_fg_expand : 0.000065s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006131s : 4.69% optimize.opt_a.after_resolve : 0.000155s : 0.12% optimize.opt_a.a_after_grad : 0.000316s : 0.24% optimize.opt_a.renormalize : 0.090907s : 69.56% optimize.opt_a.real_op_eliminate : 0.000132s : 0.10% optimize.opt_a.auto_monad_grad : 0.000306s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000253s : 0.19% optimize.opt_a.cse : 0.000965s : 0.74% optimize.opt_a.a_3 : 0.001003s : 0.77% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000101s : 0.08% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000937s : 0.72% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000056s : 0.04% optimize.cconv : 0.000021s : 0.02% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.01% optimize.tuple_transform.d_1 : 0.000042s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000011s : 0.01% optimize.add_recomputation : 0.000058s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000037s : 0.03% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000522s : 0.40% validate : 0.000039s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005720s : 4.38% execute : 0.000007s : 0.01% Time group info: ------[substitution.] 0.018704 880 0.01% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.52% : 0.016931s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.70% : 0.001066s : 97: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.16% : 0.000029s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.02% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.49% : 0.000092s : 63: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.06% : 0.000011s : 10: substitution.switch_simplify 0.07% : 0.000014s : 4: substitution.transpose_eliminate 0.61% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000068s : 60: substitution.tuple_list_get_item_depend_reorder 0.82% : 0.000153s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.090891 6 92.53% : 0.084100s : 3: renormalize.infer 7.47% : 0.006791s : 3: renormalize.specialize ------[replace.] 0.001252 141 54.22% : 0.000679s : 55: replace.getattr_setattr_resolve 26.66% : 0.000334s : 56: replace.inline 3.59% : 0.000045s : 2: replace.meta_unpack_prepare 7.44% : 0.000093s : 10: replace.switch_simplify 1.47% : 0.000018s : 4: replace.transpose_eliminate 6.63% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017767 141 94.57% : 0.016803s : 55: match.getattr_setattr_resolve 5.04% : 0.000895s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.06% : 0.000011s : 10: match.switch_simplify 0.08% : 0.000014s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006961 119 68.96% : 0.004800s : 53: func_graph_cloner_run.FuncGraphClonerGraph 31.04% : 0.002161s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.024953 259 7.53% : 0.001880s : 104: opt.transform.opt_a 3.63% : 0.000905s : 92: opt.transform.opt_b 72.85% : 0.018178s : 14: opt.transform.opt_resolve 0.45% : 0.000111s : 1: opt.transforms.meta_unpack_prepare 15.23% : 0.003799s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.20% : 0.000049s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:32.150.981 [graph_var_manager.cc:1424][EVENT]36565 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:32.151.058 [graph_manager.cc:1248][EVENT]36565 PreRun:PreRun start: graph node size 3, session id 8, graph id 7, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.151.682 [atrace_api.c:28](tid:36565) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.151.737 [trace_rb_log.c:84](tid:36565) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.151.750 [atrace_api.c:32](tid:36565) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:32.151.763 [client_manager.cpp:157][SetProfilingCallback][tid:36565] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.134 [parallel_partitioner.cc:165][EVENT]36565 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.203 [parallel_partitioner.cc:178][EVENT]36565 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.252 [graph_prepare.cc:1378][EVENT]36565 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.529 [graph_manager.cc:1050][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [292] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.559 [graph_manager.cc:1052][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.692 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.725 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.798 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [44] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.816 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.868 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.883 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.900 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.461.998 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.462.020 [graph_manager.cc:1054][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [446] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.462.240 [graph_manager.cc:1055][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [205] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.207 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.233 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.245 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.255 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [304] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.264 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.273 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [3] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.281 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.289 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [18] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.463.298 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.464.659 [graph_manager.cc:1056][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2398] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.464.723 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.464.741 [graph_prepare.cc:1982][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.176 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.199 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.209 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.225 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [256] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.235 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.243 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.252 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.261 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.269 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.294 [graph_prepare.cc:1983][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [539] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.318 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.329 [graph_prepare.cc:1984][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.343 [graph_prepare.cc:1985][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.358 [graph_prepare.cc:1986][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.369 [graph_prepare.cc:1987][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.385 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.397 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.411 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.496 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.508 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.517 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrintOpPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.525 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.534 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DropOutPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.542 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.550 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.564 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.573 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.581 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.589 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.598 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SnapshotPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.606 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.614 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.622 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.630 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.654 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [11] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.667 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.698 [graph_prepare.cc:1988][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [319] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.465.711 [graph_manager.cc:1065][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [1022] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.478.129 [graph_manager.cc:1077][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12397] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.478.196 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.478.253 [graph_manager.cc:1080][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [91] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.480.939 [graph_manager.cc:1081][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2669] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.480.977 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.480.992 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.004 [graph_manager.cc:1082][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.036 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.051 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.080 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.183 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [94] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.204 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.236 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.251 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.293 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [30] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.311 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.329 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.356 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.372 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.384 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.393 [graph_manager.cc:2700][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [363] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.499 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.513 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.522 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.531 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.540 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.548 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CastRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.556 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.565 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.573 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.581 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.596 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.605 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.614 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.622 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.630 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.640 [graph_manager.cc:2741][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [228] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.649 [graph_manager.cc:2752][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.673 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.685 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.702 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.717 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.728 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.742 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.760 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.777 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.791 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.801 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.814 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.825 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.844 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.858 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.867 [graph_manager.cc:2810][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [198] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.897 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.915 [graph_manager.cc:2821][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [39] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.481.942 [graph_manager.cc:1087][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [920] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.081 [graph_manager.cc:1088][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [124] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.119 [graph_manager.cc:1089][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.137 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.153 [graph_manager.cc:1097][EVENT]36565 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.174 [graph_manager.cc:3325][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.380 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.397 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.406 [engine_place.cc:144][EVENT]36565 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [118] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.477 [graph_manager.cc:3351][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [290] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.494 [graph_manager.cc:3364][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.559 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.577 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.730 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [142] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.773 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.821 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.859 [graph_manager.cc:3405][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [352] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.482.879 [graph_manager.cc:3412][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.364 [graph_manager.cc:3422][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1470] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.401 [graph_manager.cc:3428][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.522 [graph_manager.cc:3467][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [101] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.542 [graph_manager.cc:3377][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [2035] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.560 [graph_manager.cc:1106][EVENT]36565 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2394] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.572 [graph_manager.cc:1115][EVENT]36565 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.597 [graph_manager.cc:1130][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.630 [graph_manager.cc:1131][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.656 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.673 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.683 [graph_manager.cc:2837][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.752 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.767 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.778 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.788 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.797 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.808 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.818 [graph_manager.cc:2864][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [120] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.832 [graph_manager.cc:2872][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.853 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.869 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.884 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.906 [compile_nodes_pass.cc:88][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.920 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.484.932 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.011 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [68] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.040 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.053 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.067 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.079 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.088 [graph_manager.cc:2927][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [237] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.100 [graph_manager.cc:2937][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.113 [graph_manager.cc:2943][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.191 [graph_manager.cc:2950][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.353 [graph_manager.cc:2958][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.387 [graph_manager.cc:1132][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [740] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.463 [graph_manager.cc:1135][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.496 [graph_manager.cc:2975][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.534 [graph_manager.cc:2981][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [22] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.547 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.557 [graph_manager.cc:2986][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.566 [graph_manager.cc:1136][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [86] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.668 [graph_manager.cc:3555][EVENT]36565 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [71] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.760 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.778 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.893 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.924 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.962 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.485.986 [graph_builder.cc:865][EVENT]36565 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [257] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:32.486.346 [logger.cc:1071] 36565 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.486.378 [task_generator.cc:804][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [85] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.486.436 [task_generator.cc:805][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.486.882 [task_generator.cc:814][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [432] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.486.898 [task_generator.cc:954][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [605] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.486.953 [task_generator.cc:967][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [32] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:32.486.973 [logger.cc:1084] 36565 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:32.487.123 [graph_manager.cc:1152][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1532] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.487.142 [graph_manager.cc:1164][EVENT]36565 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.487.177 [graph_manager.cc:1271][EVENT]36565 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [26146] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.487.187 [graph_manager.cc:1272][EVENT]36565 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.487.490 [atrace_api.c:93](tid:36565) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.487.505 [atrace_api.c:95](tid:36565) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:32.491.785 [graph_converter.cc:838][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1198] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.491.978 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [151] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.492.360 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [358] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.492.443 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.492.467 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [86] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.492.755 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [278] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.492.862 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [90] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.492.897 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.050 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [140] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.139 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [74] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.159 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [95] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.190 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.215 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.241 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.303 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [52] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.360 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.370 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [56] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.396 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.423 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.444 [graph_converter.cc:849][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1622] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.493.643 [graph_converter.cc:853][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [188] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.494.288 [graph_converter.cc:857][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [632] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.494.412 [graph_converter.cc:862][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [103] micro second. . TotalTime = 0.147944, [20] [parse]: 0.00153337 [symbol_resolve]: 0.0126478, [1] [Cycle 1]: 0.0125672, [1] [resolve]: 0.0125463 [combine_like_graphs]: 8.49999e-07 [graph_reusing]: 3.23e-06 [meta_unpack_prepare]: 0.00016605 [pre_cconv]: 7.40001e-07 [abstract_specialize]: 0.00439688 [pack_expand]: 1.647e-05 [auto_monad]: 8.736e-05 [inline]: 1.87e-06 [pre_auto_parallel]: 1.022e-05 [pipeline_split]: 2.72e-06 [optimize]: 0.124766, [35] [py_interpret_to_execute]: 4.35e-06 [rewriter_before_opt_a]: 0.00019471 [opt_a]: 0.122846, [4] [Cycle 1]: 0.0593108, [30] [expand_dump_flag]: 4.83001e-06 [switch_simplify]: 2.717e-05 [a_1]: 0.00073952 [recompute_prepare]: 8.56e-06 [updatestate_depend_eliminate]: 1.088e-05 [updatestate_assign_eliminate]: 7.25e-06 [updatestate_loads_eliminate]: 7.31e-06 [parameter_eliminate]: 5.45001e-06 [a_2]: 7.654e-05 [accelerated_algorithm]: 5.48e-06 [pynative_shard]: 1.97e-06 [auto_parallel]: 3.73e-06 [parallel]: 8.45e-06 [merge_comm]: 4.33e-06 [allreduce_fusion]: 2.28e-06 [virtual_dataset]: 5.49e-06 [get_grad_eliminate_]: 4.87e-06 [virtual_output]: 4.45e-06 [merge_forward]: 8.77e-06 [cell_reuse_recompute_pass]: 8.70001e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.159e-05 [meta_fg_expand]: 0.00718337, [1] [Cycle 1]: 0.0033197, [1] [resolve]: 0.00330035 [after_resolve]: 4.988e-05 [a_after_grad]: 0.00013149 [renormalize]: 0.0498876 [real_op_eliminate]: 4.428e-05 [auto_monad_grad]: 7.396e-05 [auto_monad_eliminator]: 8.197e-05 [cse]: 0.0003763 [a_3]: 0.00034223 [Cycle 2]: 0.053484, [30] [expand_dump_flag]: 4.06e-06 [switch_simplify]: 0.00013449 [a_1]: 0.00162562 [recompute_prepare]: 1.317e-05 [updatestate_depend_eliminate]: 1.694e-05 [updatestate_assign_eliminate]: 1.292e-05 [updatestate_loads_eliminate]: 1.257e-05 [parameter_eliminate]: 4.72e-06 [a_2]: 0.00019645 [accelerated_algorithm]: 1.905e-05 [pynative_shard]: 1.55e-06 [auto_parallel]: 4.83e-06 [parallel]: 5.02e-06 [merge_comm]: 2.82e-06 [allreduce_fusion]: 1.5e-06 [virtual_dataset]: 1.134e-05 [get_grad_eliminate_]: 1.039e-05 [virtual_output]: 1.045e-05 [merge_forward]: 1.368e-05 [cell_reuse_recompute_pass]: 4.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.324e-05 [meta_fg_expand]: 0.0116364, [5] [Cycle 1]: 0.00035335, [1] [resolve]: 0.00033483 [Cycle 1]: 0.00032103, [1] [resolve]: 0.00030208 [Cycle 1]: 0.00166316, [1] [resolve]: 0.00164482 [Cycle 1]: 0.0003169, [1] [resolve]: 0.00029882 [Cycle 1]: 0.0003122, [1] [resolve]: 0.00029482 [after_resolve]: 7.67e-05 [a_after_grad]: 0.00019431 [renormalize]: 0.038143 [real_op_eliminate]: 5.7e-05 [auto_monad_grad]: 0.00022103 [auto_monad_eliminator]: 0.00010587 [cse]: 0.00031291 [a_3]: 0.00042398 [Cycle 3]: 0.0059676, [30] [expand_dump_flag]: 4.32e-06 [switch_simplify]: 0.00015595 [a_1]: 0.00235531 [recompute_prepare]: 1.667e-05 [updatestate_depend_eliminate]: 3.228e-05 [updatestate_assign_eliminate]: 1.866e-05 [updatestate_loads_eliminate]: 1.788e-05 [parameter_eliminate]: 4.65e-06 [a_2]: 0.00027388 [accelerated_algorithm]: 2.49e-05 [pynative_shard]: 1.22e-06 [auto_parallel]: 4.31e-06 [parallel]: 4.18e-06 [merge_comm]: 3.86e-06 [allreduce_fusion]: 2.36e-06 [virtual_dataset]: 1.494e-05 [get_grad_eliminate_]: 1.443e-05 [virtual_output]: 1.422e-05 [merge_forward]: 1.918e-05 [cell_reuse_recompute_pass]: 4.50003e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.226e-05 [meta_fg_expand]: 4.978e-05 [after_resolve]: 1.821e-05 [a_after_grad]: 3.227e-05 [renormalize]: 0.00229806 [real_op_eliminate]: 2.21e-05 [auto_monad_grad]: 6.12e-06 [auto_monad_eliminator]: 3.486e-05 [cse]: 0.00020716 [a_3]: 0.00012087 [Cycle 4]: 0.00168218, [30] [expand_dump_flag]: 1.34001e-06 [switch_simplify]: 1.538e-05 [a_1]: 0.00075902 [recompute_prepare]: 1.633e-05 [updatestate_depend_eliminate]: 2.021e-05 [updatestate_assign_eliminate]: 1.718e-05 [updatestate_loads_eliminate]: 1.695e-05 [parameter_eliminate]: 2.39e-06 [a_2]: 0.00027267 [accelerated_algorithm]: 2.571e-05 [pynative_shard]: 1.51e-06 [auto_parallel]: 3.89e-06 [parallel]: 3.79e-06 [merge_comm]: 3.32e-06 [allreduce_fusion]: 2.04e-06 [virtual_dataset]: 1.512e-05 [get_grad_eliminate_]: 1.507e-05 [virtual_output]: 1.407e-05 [merge_forward]: 1.737e-05 [cell_reuse_recompute_pass]: 4.19997e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.331e-05 [meta_fg_expand]: 1.392e-05 [after_resolve]: 1.666e-05 [a_after_grad]: 3.184e-05 [renormalize]: 7.99992e-08 [real_op_eliminate]: 1.476e-05 [auto_monad_grad]: 2.05e-06 [auto_monad_eliminator]: 3.029e-05 [cse]: 8.407e-05 [a_3]: 0.00011379 [py_interpret_to_execute_after_opt_a]: 4.18e-06 [slice_cell_reuse_recomputed_activation]: 2.14e-06 [rewriter_after_opt_a]: 9.853e-05 [convert_after_rewriter]: 2.39e-05 [order_py_execute_after_rewriter]: 1.732e-05 [opt_b]: 0.00111799, [2] [Cycle 1]: 0.0009469, [7] [b_1]: 0.00085046 [b_2]: 5.35e-06 [updatestate_depend_eliminate]: 6.07e-06 [updatestate_assign_eliminate]: 4.16e-06 [updatestate_loads_eliminate]: 3.96001e-06 [renormalize]: 4.30002e-07 [cse]: 3.905e-05 [Cycle 2]: 0.00016158, [7] [b_1]: 9.835e-05 [b_2]: 4.08e-06 [updatestate_depend_eliminate]: 5.28e-06 [updatestate_assign_eliminate]: 4e-06 [updatestate_loads_eliminate]: 3.74e-06 [renormalize]: 9.00036e-08 [cse]: 1.891e-05 [cconv]: 2.256e-05 [opt_after_cconv]: 8.719e-05, [1] [Cycle 1]: 8.281e-05, [7] [c_1]: 2.441e-05 [parameter_eliminate]: 2.28e-06 [updatestate_depend_eliminate]: 4.33e-06 [updatestate_assign_eliminate]: 4.44e-06 [updatestate_loads_eliminate]: 3.64e-06 [cse]: 1.617e-05 [renormalize]: 4.60001e-07 [remove_dup_value]: 1.729e-05 [tuple_transform]: 8.382e-05, [1] [Cycle 1]: 7.993e-05, [3] [d_1]: 5.858e-05 [d_2]: 9.4e-06 [renormalize]: 2.3e-07 [add_cache_embedding]: 1.28e-05 [add_recomputation]: 5.929e-05 [cse_after_recomputation]: 2.726e-05, [1] [Cycle 1]: 2.314e-05, [1] [cse]: 1.839e-05 [environ_conv]: 1.04e-05 [label_micro_interleaved_index]: 2.27e-06 [label_fine_grained_interleaved_index]: 2.26e-06 [assign_add_opt]: 1.42e-06 [slice_recompute_activation]: 1.75e-06 [micro_interleaved_order_control]: 1.84e-06 [full_micro_interleaved_order_control]: 2.19e-06 [comp_comm_scheduling]: 1.98e-06 [reorder_send_recv_between_fp_bp]: 2.25e-06 [comm_op_add_attrs]: 1.08e-06 [add_comm_op_reuse_tag]: 8.70001e-07 [overlap_opt_shard_in_pipeline]: 9.89996e-07 [grouped_pairwise_exchange_alltoall]: 1.42999e-06 [overlap_recompute_and_grad_model_parallel]: 1.67e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.99998e-07 [split_matmul_comm_elemetwise]: 2.5e-06 [split_layernorm_comm]: 2.06e-06 [process_send_recv_for_ge]: 7.30004e-07 [handle_group_info]: 9.39996e-07 [auto_monad_reorder]: 2.211e-05 [get_jit_bprop_graph]: 3.33e-06 [eliminate_special_op_node]: 0.00055535 [validate]: 3.893e-05 [distribtued_split]: 1.35e-06 [task_emit]: 0.00348791 [execute]: 6.21e-06 Sums parse : 0.001533s : 1.16% symbol_resolve.resolve : 0.012546s : 9.53% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000166s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004397s : 3.34% pack_expand : 0.000016s : 0.01% auto_monad : 0.000087s : 0.07% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000195s : 0.15% optimize.opt_a.expand_dump_flag : 0.000015s : 0.01% optimize.opt_a.switch_simplify : 0.000333s : 0.25% optimize.opt_a.a_1 : 0.005479s : 4.16% optimize.opt_a.recompute_prepare : 0.000055s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000080s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000055s : 0.04% optimize.opt_a.parameter_eliminate : 0.000017s : 0.01% optimize.opt_a.a_2 : 0.000820s : 0.62% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000021s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000045s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000059s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000100s : 0.08% optimize.opt_a.meta_fg_expand : 0.000064s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006176s : 4.69% optimize.opt_a.after_resolve : 0.000161s : 0.12% optimize.opt_a.a_after_grad : 0.000390s : 0.30% optimize.opt_a.renormalize : 0.090329s : 68.58% optimize.opt_a.real_op_eliminate : 0.000138s : 0.10% optimize.opt_a.auto_monad_grad : 0.000303s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000253s : 0.19% optimize.opt_a.cse : 0.000980s : 0.74% optimize.opt_a.a_3 : 0.001001s : 0.76% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000099s : 0.07% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000949s : 0.72% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000058s : 0.04% optimize.cconv : 0.000023s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000059s : 0.04% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000059s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000022s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000555s : 0.42% validate : 0.000039s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003488s : 2.65% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018973 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.33% : 0.017139s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.49% : 0.001042s : 103: substitution.inline 0.05% : 0.000009s : 23: substitution.less_batch_normalization 0.19% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.62% : 0.000118s : 69: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000006s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000011s : 10: substitution.switch_simplify 0.06% : 0.000012s : 4: substitution.transpose_eliminate 0.64% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000075s : 70: substitution.tuple_list_get_item_depend_reorder 0.86% : 0.000163s : 122: substitution.tuple_list_get_item_eliminator 0.40% : 0.000076s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.090312 6 92.47% : 0.083515s : 3: renormalize.infer 7.53% : 0.006797s : 3: renormalize.specialize ------[replace.] 0.001256 141 54.61% : 0.000686s : 55: replace.getattr_setattr_resolve 26.04% : 0.000327s : 56: replace.inline 3.78% : 0.000047s : 2: replace.meta_unpack_prepare 7.41% : 0.000093s : 10: replace.switch_simplify 1.57% : 0.000020s : 4: replace.transpose_eliminate 6.59% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017946 141 94.77% : 0.017008s : 55: match.getattr_setattr_resolve 4.85% : 0.000870s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.06% : 0.000011s : 10: match.switch_simplify 0.07% : 0.000012s : 4: match.transpose_eliminate 0.16% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006975 119 69.35% : 0.004837s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.65% : 0.002138s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.028046 589 0.51% : 0.000142s : 2: opt.transform.meta_unpack_prepare 30.26% : 0.008488s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.29% : 0.000923s : 94: opt.transform.opt_b 65.58% : 0.018393s : 14: opt.transform.opt_resolve 0.23% : 0.000063s : 8: opt.transform.opt_trans_graph 0.06% : 0.000016s : 3: opt.transform.special_op_eliminate . TotalTime = 0.146571, [20] [parse]: 0.00127519 [symbol_resolve]: 0.0125124, [1] [Cycle 1]: 0.0124467, [1] [resolve]: 0.0124292 [combine_like_graphs]: 8.49999e-07 [graph_reusing]: 2.93e-06 [meta_unpack_prepare]: 0.00012643 [pre_cconv]: 4.60001e-07 [abstract_specialize]: 0.00410791 [pack_expand]: 1.465e-05 [auto_monad]: 6.62e-05 [inline]: 1.46e-06 [pre_auto_parallel]: 7.52e-06 [pipeline_split]: 2.09e-06 [optimize]: 0.122164, [35] [py_interpret_to_execute]: 4.65e-06 [rewriter_before_opt_a]: 0.00018988 [opt_a]: 0.120325, [4] [Cycle 1]: 0.0590276, [30] [expand_dump_flag]: 3.8e-06 [switch_simplify]: 2.712e-05 [a_1]: 0.00039474 [recompute_prepare]: 8.75e-06 [updatestate_depend_eliminate]: 9.44e-06 [updatestate_assign_eliminate]: 7.34e-06 [updatestate_loads_eliminate]: 6.07e-06 [parameter_eliminate]: 4.34e-06 [a_2]: 7.962e-05 [accelerated_algorithm]: 5.38e-06 [pynative_shard]: 1.07e-06 [auto_parallel]: 3.23e-06 [parallel]: 5.60001e-06 [merge_comm]: 2.78e-06 [allreduce_fusion]: 1.66e-06 [virtual_dataset]: 5.31e-06 [get_grad_eliminate_]: 4.63e-06 [virtual_output]: 4.67e-06 [merge_forward]: 7.53e-06 [cell_reuse_recompute_pass]: 4.49996e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.242e-05 [meta_fg_expand]: 0.00706471, [1] [Cycle 1]: 0.00330008, [1] [resolve]: 0.00328088 [after_resolve]: 4.94e-05 [a_after_grad]: 0.00011156 [renormalize]: 0.0501437 [real_op_eliminate]: 4.129e-05 [auto_monad_grad]: 7.202e-05 [auto_monad_eliminator]: 7.782e-05 [cse]: 0.00035427 [a_3]: 0.00031181 [Cycle 2]: 0.0528242, [30] [expand_dump_flag]: 3.65001e-06 [switch_simplify]: 0.00010162 [a_1]: 0.00077325 [recompute_prepare]: 1.374e-05 [updatestate_depend_eliminate]: 1.643e-05 [updatestate_assign_eliminate]: 1.292e-05 [updatestate_loads_eliminate]: 1.286e-05 [parameter_eliminate]: 4.21e-06 [a_2]: 0.00019692 [accelerated_algorithm]: 1.756e-05 [pynative_shard]: 1.17e-06 [auto_parallel]: 4.11e-06 [parallel]: 4.51e-06 [merge_comm]: 2.32e-06 [allreduce_fusion]: 1.62e-06 [virtual_dataset]: 1.027e-05 [get_grad_eliminate_]: 9.69e-06 [virtual_output]: 9.56999e-06 [merge_forward]: 1.386e-05 [cell_reuse_recompute_pass]: 4.69998e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.37e-05 [meta_fg_expand]: 0.0115052, [5] [Cycle 1]: 0.00032423, [1] [resolve]: 0.0003061 [Cycle 1]: 0.00031922, [1] [resolve]: 0.00030097 [Cycle 1]: 0.00168619, [1] [resolve]: 0.00166683 [Cycle 1]: 0.00031969, [1] [resolve]: 0.00030187 [Cycle 1]: 0.00031674, [1] [resolve]: 0.00029866 [after_resolve]: 7.257e-05 [a_after_grad]: 0.0001668 [renormalize]: 0.0385571 [real_op_eliminate]: 5.329e-05 [auto_monad_grad]: 0.00021036 [auto_monad_eliminator]: 0.00010586 [cse]: 0.00030758 [a_3]: 0.0004247 [Cycle 3]: 0.00493128, [30] [expand_dump_flag]: 4.09e-06 [switch_simplify]: 0.00011345 [a_1]: 0.00126576 [recompute_prepare]: 1.815e-05 [updatestate_depend_eliminate]: 2.694e-05 [updatestate_assign_eliminate]: 1.892e-05 [updatestate_loads_eliminate]: 1.785e-05 [parameter_eliminate]: 5.06e-06 [a_2]: 0.00028059 [accelerated_algorithm]: 2.566e-05 [pynative_shard]: 1.46e-06 [auto_parallel]: 6.49e-06 [parallel]: 4.42001e-06 [merge_comm]: 3.84e-06 [allreduce_fusion]: 2.43e-06 [virtual_dataset]: 1.416e-05 [get_grad_eliminate_]: 1.342e-05 [virtual_output]: 1.267e-05 [merge_forward]: 1.951e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.324e-05 [meta_fg_expand]: 4.825e-05 [after_resolve]: 1.675e-05 [a_after_grad]: 2.048e-05 [renormalize]: 0.00236454 [real_op_eliminate]: 2.035e-05 [auto_monad_grad]: 5.44e-06 [auto_monad_eliminator]: 3.484e-05 [cse]: 0.00020575 [a_3]: 0.0001638 [Cycle 4]: 0.00120165, [30] [expand_dump_flag]: 1.4e-06 [switch_simplify]: 1.521e-05 [a_1]: 0.00029061 [recompute_prepare]: 1.553e-05 [updatestate_depend_eliminate]: 2.001e-05 [updatestate_assign_eliminate]: 1.729e-05 [updatestate_loads_eliminate]: 1.687e-05 [parameter_eliminate]: 2.22e-06 [a_2]: 0.00027633 [accelerated_algorithm]: 2.563e-05 [pynative_shard]: 1.63e-06 [auto_parallel]: 4e-06 [parallel]: 3.74e-06 [merge_comm]: 3.29e-06 [allreduce_fusion]: 1.99e-06 [virtual_dataset]: 1.455e-05 [get_grad_eliminate_]: 1.405e-05 [virtual_output]: 1.326e-05 [merge_forward]: 1.747e-05 [cell_reuse_recompute_pass]: 4.20005e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.322e-05 [meta_fg_expand]: 1.375e-05 [after_resolve]: 1.56e-05 [a_after_grad]: 2.047e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.354e-05 [auto_monad_grad]: 2.36e-06 [auto_monad_eliminator]: 3.066e-05 [cse]: 8.392e-05 [a_3]: 0.00011439 [py_interpret_to_execute_after_opt_a]: 4.17e-06 [slice_cell_reuse_recomputed_activation]: 1.39e-06 [rewriter_after_opt_a]: 9.712e-05 [convert_after_rewriter]: 2.255e-05 [order_py_execute_after_rewriter]: 1.618e-05 [opt_b]: 0.00111367, [2] [Cycle 1]: 0.0009431, [7] [b_1]: 0.00084731 [b_2]: 5.99e-06 [updatestate_depend_eliminate]: 5.99e-06 [updatestate_assign_eliminate]: 4.32e-06 [updatestate_loads_eliminate]: 4.24e-06 [renormalize]: 4.99997e-07 [cse]: 3.827e-05 [Cycle 2]: 0.00016062, [7] [b_1]: 9.908e-05 [b_2]: 4.35e-06 [updatestate_depend_eliminate]: 4.57e-06 [updatestate_assign_eliminate]: 3.94e-06 [updatestate_loads_eliminate]: 3.59e-06 [renormalize]: 6.99947e-08 [cse]: 1.84e-05 [cconv]: 1.771e-05 [opt_after_cconv]: 6.963e-05, [1] [Cycle 1]: 6.523e-05, [7] [c_1]: 8.92e-06 [parameter_eliminate]: 2.45e-06 [updatestate_depend_eliminate]: 4.31e-06 [updatestate_assign_eliminate]: 3.68e-06 [updatestate_loads_eliminate]: 3.55e-06 [cse]: 1.579e-05 [renormalize]: 4.00003e-07 [remove_dup_value]: 1.35e-05 [tuple_transform]: 6.661e-05, [1] [Cycle 1]: 6.296e-05, [3] [d_1]: 4.111e-05 [d_2]: 8.88e-06 [renormalize]: 1.50001e-07 [add_cache_embedding]: 1.068e-05 [add_recomputation]: 5.158e-05 [cse_after_recomputation]: 2.563e-05, [1] [Cycle 1]: 2.143e-05, [1] [cse]: 1.685e-05 [environ_conv]: 8.99e-06 [label_micro_interleaved_index]: 1.65e-06 [label_fine_grained_interleaved_index]: 1.53e-06 [assign_add_opt]: 9.20001e-07 [slice_recompute_activation]: 1.45e-06 [micro_interleaved_order_control]: 1.18e-06 [full_micro_interleaved_order_control]: 1.15e-06 [comp_comm_scheduling]: 1.4e-06 [reorder_send_recv_between_fp_bp]: 1.88e-06 [comm_op_add_attrs]: 6.40001e-07 [add_comm_op_reuse_tag]: 6.69999e-07 [overlap_opt_shard_in_pipeline]: 8.70001e-07 [grouped_pairwise_exchange_alltoall]: 7.00005e-07 [overlap_recompute_and_grad_model_parallel]: 1.11001e-06 [overlap_grad_matmul_and_grad_allreduce]: 5.10001e-07 [split_matmul_comm_elemetwise]: 2.11001e-06 [split_layernorm_comm]: 8.90002e-07 [process_send_recv_for_ge]: 7.10002e-07 [handle_group_info]: 5.60001e-07 [auto_monad_reorder]: 1.825e-05 [get_jit_bprop_graph]: 4.1e-07 [eliminate_special_op_node]: 0.00068183 [validate]: 3.629e-05 [distribtued_split]: 1.18e-06 [task_emit]: 0.00536527 [execute]: 5.32001e-06 Sums parse : 0.001275s : 0.98% symbol_resolve.resolve : 0.012429s : 9.51% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000126s : 0.10% pre_cconv : 0.000000s : 0.00% abstract_specialize : 0.004108s : 3.14% pack_expand : 0.000015s : 0.01% auto_monad : 0.000066s : 0.05% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000008s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000190s : 0.15% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000257s : 0.20% optimize.opt_a.a_1 : 0.002724s : 2.09% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000073s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000054s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000833s : 0.64% optimize.opt_a.accelerated_algorithm : 0.000074s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000018s : 0.01% optimize.opt_a.parallel : 0.000018s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.03% optimize.opt_a.virtual_output : 0.000040s : 0.03% optimize.opt_a.merge_forward : 0.000058s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000103s : 0.08% optimize.opt_a.meta_fg_expand : 0.000062s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006155s : 4.71% optimize.opt_a.after_resolve : 0.000154s : 0.12% optimize.opt_a.a_after_grad : 0.000319s : 0.24% optimize.opt_a.renormalize : 0.091065s : 69.70% optimize.opt_a.real_op_eliminate : 0.000128s : 0.10% optimize.opt_a.auto_monad_grad : 0.000290s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000249s : 0.19% optimize.opt_a.cse : 0.000952s : 0.73% optimize.opt_a.a_3 : 0.001015s : 0.78% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000001s : 0.00% optimize.rewriter_after_opt_a : 0.000097s : 0.07% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000946s : 0.72% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000057s : 0.04% optimize.cconv : 0.000018s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000041s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000011s : 0.01% optimize.add_recomputation : 0.000052s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000001s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000018s : 0.01% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000682s : 0.52% validate : 0.000036s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005365s : 4.11% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018769 880 0.01% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.56% : 0.016998s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.63% : 0.001057s : 97: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.15% : 0.000028s : 23: substitution.meta_unpack_prepare 0.15% : 0.000028s : 40: substitution.minmaximum_grad 0.01% : 0.000003s : 5: substitution.partial_eliminate 0.00% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.48% : 0.000089s : 63: substitution.replace_applicator 0.05% : 0.000009s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.59% : 0.000111s : 60: substitution.tuple_list_convert_item_index_to_positive 0.36% : 0.000067s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_item_depend_reorder 0.83% : 0.000156s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.091050 6 92.69% : 0.084395s : 3: renormalize.infer 7.31% : 0.006655s : 3: renormalize.specialize ------[replace.] 0.001250 141 54.34% : 0.000679s : 55: replace.getattr_setattr_resolve 26.58% : 0.000332s : 56: replace.inline 3.58% : 0.000045s : 2: replace.meta_unpack_prepare 7.25% : 0.000091s : 10: replace.switch_simplify 1.43% : 0.000018s : 4: replace.transpose_eliminate 6.83% : 0.000085s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017819 141 94.67% : 0.016869s : 55: match.getattr_setattr_resolve 4.96% : 0.000884s : 56: match.inline 0.09% : 0.000016s : 2: match.meta_unpack_prepare 0.05% : 0.000010s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006878 119 69.60% : 0.004787s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.40% : 0.002091s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.025045 259 7.57% : 0.001895s : 104: opt.transform.opt_a 3.64% : 0.000913s : 92: opt.transform.opt_b 72.87% : 0.018249s : 14: opt.transform.opt_resolve 0.44% : 0.000110s : 1: opt.transforms.meta_unpack_prepare 15.17% : 0.003800s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000048s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:32.876.181 [graph_var_manager.cc:1424][EVENT]36564 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:32.876.259 [graph_manager.cc:1248][EVENT]36564 PreRun:PreRun start: graph node size 3, session id 9, graph id 8, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.876.896 [atrace_api.c:28](tid:36564) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.876.952 [trace_rb_log.c:84](tid:36564) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.876.965 [atrace_api.c:32](tid:36564) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:32.876.978 [client_manager.cpp:157][SetProfilingCallback][tid:36564] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:32.877.694 [parallel_partitioner.cc:165][EVENT]36564 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.877.738 [parallel_partitioner.cc:178][EVENT]36564 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.877.785 [graph_prepare.cc:1378][EVENT]36564 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.170 [graph_manager.cc:1050][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [401] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.199 [graph_manager.cc:1052][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.330 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.360 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.413 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [41] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.428 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.478 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.491 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.509 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.617 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.654 [graph_manager.cc:1054][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [442] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.878.876 [graph_manager.cc:1055][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [207] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.846 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [6] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.870 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.881 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.891 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [302] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.900 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.909 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [6] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.917 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.925 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [16] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.879.934 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.335 [graph_manager.cc:1056][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2441] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.400 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.418 [graph_prepare.cc:1982][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.815 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.837 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.848 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.857 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [224] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.866 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.875 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.883 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.892 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.908 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.933 [graph_prepare.cc:1983][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [502] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.957 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.968 [graph_prepare.cc:1984][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.981 [graph_prepare.cc:1985][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.881.997 [graph_prepare.cc:1986][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.010 [graph_prepare.cc:1987][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.024 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.036 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.049 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.132 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.147 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.156 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrintOpPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.165 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.173 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.181 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.190 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.198 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.206 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.214 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.223 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.231 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SnapshotPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.239 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.253 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.262 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.270 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.293 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [11] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.306 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.338 [graph_prepare.cc:1988][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [319] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.882.352 [graph_manager.cc:1065][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [985] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.894.707 [graph_manager.cc:1077][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12335] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.894.774 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.894.829 [graph_manager.cc:1080][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [91] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.544 [graph_manager.cc:1081][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2699] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.582 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.598 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.609 [graph_manager.cc:1082][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [36] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.640 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.656 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.670 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.742 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.759 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.792 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.807 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.854 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.873 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.890 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.919 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.934 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.945 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.897.955 [graph_manager.cc:2700][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [319] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.060 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.074 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.084 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.093 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.102 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.110 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CastRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.118 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.127 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.135 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.143 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.151 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.159 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.167 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.175 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.184 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.193 [graph_manager.cc:2741][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [220] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.213 [graph_manager.cc:2752][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.237 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.249 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.266 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.282 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.293 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.305 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.325 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.339 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.354 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.364 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.377 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.388 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.407 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.420 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.430 [graph_manager.cc:2810][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [197] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.459 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.472 [graph_manager.cc:2821][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.499 [graph_manager.cc:1087][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [872] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.634 [graph_manager.cc:1088][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [121] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.672 [graph_manager.cc:1089][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.691 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.713 [graph_manager.cc:1097][EVENT]36564 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.735 [graph_manager.cc:3325][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.942 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.958 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.898.968 [engine_place.cc:144][EVENT]36564 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [120] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.039 [graph_manager.cc:3351][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [291] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.056 [graph_manager.cc:3364][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.120 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.137 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.289 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [142] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.331 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.381 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.417 [graph_manager.cc:3405][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [349] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.899.436 [graph_manager.cc:3412][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.900.859 [graph_manager.cc:3422][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1411] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.900.888 [graph_manager.cc:3428][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.010 [graph_manager.cc:3467][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [103] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.028 [graph_manager.cc:3377][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1960] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.044 [graph_manager.cc:1106][EVENT]36564 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2315] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.057 [graph_manager.cc:1115][EVENT]36564 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.078 [graph_manager.cc:1130][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.141 [graph_manager.cc:1131][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [43] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.166 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.184 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.194 [graph_manager.cc:2837][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.266 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [13] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.279 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.289 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.297 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.306 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.315 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.324 [graph_manager.cc:2864][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [114] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.336 [graph_manager.cc:2872][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.356 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.372 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.387 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.401 [compile_nodes_pass.cc:88][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.411 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.420 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.500 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [71] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.527 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.540 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.559 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.572 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.580 [graph_manager.cc:2927][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [227] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.593 [graph_manager.cc:2937][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.606 [graph_manager.cc:2943][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.617 [graph_manager.cc:2950][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.774 [graph_manager.cc:2958][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.803 [graph_manager.cc:1132][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [646] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.881 [graph_manager.cc:1135][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [64] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.912 [graph_manager.cc:2975][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.941 [graph_manager.cc:2981][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.955 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.966 [graph_manager.cc:2986][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.901.975 [graph_manager.cc:1136][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [78] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.079 [graph_manager.cc:3555][EVENT]36564 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [73] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.162 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.177 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.297 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [110] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.325 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.365 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.393 [graph_builder.cc:865][EVENT]36564 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [262] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:32.902.750 [logger.cc:1071] 36564 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.781 [task_generator.cc:804][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [76] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.902.841 [task_generator.cc:805][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [48] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.294 [task_generator.cc:814][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [438] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.308 [task_generator.cc:954][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [603] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.364 [task_generator.cc:967][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [32] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:32.903.381 [logger.cc:1084] 36564 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.528 [graph_manager.cc:1152][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1529] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.545 [graph_manager.cc:1164][EVENT]36564 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.577 [graph_manager.cc:1271][EVENT]36564 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [25973] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.903.588 [graph_manager.cc:1272][EVENT]36564 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.903.888 [atrace_api.c:93](tid:36564) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:32.903.902 [atrace_api.c:95](tid:36564) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:32.908.169 [graph_converter.cc:838][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1213] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.908.361 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [151] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.908.748 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [364] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.908.830 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.908.845 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [76] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.144 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [289] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.251 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.287 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.440 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [138] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.513 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [58] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.525 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [71] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.561 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.588 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.614 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.675 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [52] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.734 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.744 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [58] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.769 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.793 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.909.812 [graph_converter.cc:849][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1608] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.910.010 [graph_converter.cc:853][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [188] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.910.644 [graph_converter.cc:857][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [621] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:32.910.769 [graph_converter.cc:862][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [103] micro second. . TotalTime = 0.147175, [20] [parse]: 0.00151451 [symbol_resolve]: 0.012544, [1] [Cycle 1]: 0.0124626, [1] [resolve]: 0.0124426 [combine_like_graphs]: 7.99999e-07 [graph_reusing]: 3.27e-06 [meta_unpack_prepare]: 0.0001649 [pre_cconv]: 6.99998e-07 [abstract_specialize]: 0.00438671 [pack_expand]: 1.711e-05 [auto_monad]: 8.43e-05 [inline]: 1.31e-06 [pre_auto_parallel]: 1.021e-05 [pipeline_split]: 2.99e-06 [optimize]: 0.124155, [35] [py_interpret_to_execute]: 4.71e-06 [rewriter_before_opt_a]: 0.00019248 [opt_a]: 0.122251, [4] [Cycle 1]: 0.0590499, [30] [expand_dump_flag]: 4.84e-06 [switch_simplify]: 2.771e-05 [a_1]: 0.00072989 [recompute_prepare]: 8.37e-06 [updatestate_depend_eliminate]: 1.028e-05 [updatestate_assign_eliminate]: 7.2e-06 [updatestate_loads_eliminate]: 7.01e-06 [parameter_eliminate]: 4.58e-06 [a_2]: 7.736e-05 [accelerated_algorithm]: 5.33e-06 [pynative_shard]: 1.82e-06 [auto_parallel]: 3.38e-06 [parallel]: 9.28e-06 [merge_comm]: 4.24e-06 [allreduce_fusion]: 2.21e-06 [virtual_dataset]: 5.38e-06 [get_grad_eliminate_]: 4.85e-06 [virtual_output]: 4.73e-06 [merge_forward]: 8.66e-06 [cell_reuse_recompute_pass]: 7.89994e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.182e-05 [meta_fg_expand]: 0.00711817, [1] [Cycle 1]: 0.00330588, [1] [resolve]: 0.00328505 [after_resolve]: 4.937e-05 [a_after_grad]: 0.00013781 [renormalize]: 0.0497152 [real_op_eliminate]: 4.418e-05 [auto_monad_grad]: 7.556e-05 [auto_monad_eliminator]: 8.126e-05 [cse]: 0.00037744 [a_3]: 0.00032432 [Cycle 2]: 0.0531564, [30] [expand_dump_flag]: 4.18e-06 [switch_simplify]: 0.00013551 [a_1]: 0.00161019 [recompute_prepare]: 1.264e-05 [updatestate_depend_eliminate]: 1.648e-05 [updatestate_assign_eliminate]: 1.273e-05 [updatestate_loads_eliminate]: 1.246e-05 [parameter_eliminate]: 4.44e-06 [a_2]: 0.00019437 [accelerated_algorithm]: 1.846e-05 [pynative_shard]: 1.34e-06 [auto_parallel]: 4.75e-06 [parallel]: 5.44e-06 [merge_comm]: 2.55999e-06 [allreduce_fusion]: 1.41e-06 [virtual_dataset]: 1.149e-05 [get_grad_eliminate_]: 1.046e-05 [virtual_output]: 1.082e-05 [merge_forward]: 1.436e-05 [cell_reuse_recompute_pass]: 4.19997e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.31e-05 [meta_fg_expand]: 0.0115006, [5] [Cycle 1]: 0.00033111, [1] [resolve]: 0.00031268 [Cycle 1]: 0.00031342, [1] [resolve]: 0.00029583 [Cycle 1]: 0.00165438, [1] [resolve]: 0.00163577 [Cycle 1]: 0.00032647, [1] [resolve]: 0.00030769 [Cycle 1]: 0.00031075, [1] [resolve]: 0.00029298 [after_resolve]: 7.557e-05 [a_after_grad]: 0.00019513 [renormalize]: 0.0379706 [real_op_eliminate]: 5.845e-05 [auto_monad_grad]: 0.00021886 [auto_monad_eliminator]: 0.00010693 [cse]: 0.00031215 [a_3]: 0.00042311 [Cycle 3]: 0.00593597, [30] [expand_dump_flag]: 4.51e-06 [switch_simplify]: 0.00015732 [a_1]: 0.00230061 [recompute_prepare]: 1.596e-05 [updatestate_depend_eliminate]: 3.227e-05 [updatestate_assign_eliminate]: 1.885e-05 [updatestate_loads_eliminate]: 1.777e-05 [parameter_eliminate]: 4.62e-06 [a_2]: 0.00028069 [accelerated_algorithm]: 2.458e-05 [pynative_shard]: 1.31e-06 [auto_parallel]: 4.14e-06 [parallel]: 4.24e-06 [merge_comm]: 3.67e-06 [allreduce_fusion]: 2.71e-06 [virtual_dataset]: 1.528e-05 [get_grad_eliminate_]: 1.463e-05 [virtual_output]: 1.399e-05 [merge_forward]: 1.97e-05 [cell_reuse_recompute_pass]: 4.69998e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.275e-05 [meta_fg_expand]: 5.063e-05 [after_resolve]: 1.756e-05 [a_after_grad]: 3.166e-05 [renormalize]: 0.0023114 [real_op_eliminate]: 2.127e-05 [auto_monad_grad]: 5.99e-06 [auto_monad_eliminator]: 3.457e-05 [cse]: 0.00020852 [a_3]: 0.00012186 [Cycle 4]: 0.00168828, [30] [expand_dump_flag]: 1.36e-06 [switch_simplify]: 1.542e-05 [a_1]: 0.00076775 [recompute_prepare]: 1.596e-05 [updatestate_depend_eliminate]: 2e-05 [updatestate_assign_eliminate]: 1.672e-05 [updatestate_loads_eliminate]: 1.599e-05 [parameter_eliminate]: 2.43e-06 [a_2]: 0.00027524 [accelerated_algorithm]: 2.538e-05 [pynative_shard]: 1.50999e-06 [auto_parallel]: 3.78e-06 [parallel]: 3.88e-06 [merge_comm]: 3.2e-06 [allreduce_fusion]: 2.19e-06 [virtual_dataset]: 1.527e-05 [get_grad_eliminate_]: 1.475e-05 [virtual_output]: 1.406e-05 [merge_forward]: 1.717e-05 [cell_reuse_recompute_pass]: 4.39999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.342e-05 [meta_fg_expand]: 1.386e-05 [after_resolve]: 1.691e-05 [a_after_grad]: 3.154e-05 [renormalize]: 5.99975e-08 [real_op_eliminate]: 1.465e-05 [auto_monad_grad]: 2.45e-06 [auto_monad_eliminator]: 3.007e-05 [cse]: 8.31e-05 [a_3]: 0.00011195 [py_interpret_to_execute_after_opt_a]: 4.17e-06 [slice_cell_reuse_recomputed_activation]: 2.26e-06 [rewriter_after_opt_a]: 9.814e-05 [convert_after_rewriter]: 2.36e-05 [order_py_execute_after_rewriter]: 1.696e-05 [opt_b]: 0.00110847, [2] [Cycle 1]: 0.0009375, [7] [b_1]: 0.00084364 [b_2]: 4.86999e-06 [updatestate_depend_eliminate]: 5.92e-06 [updatestate_assign_eliminate]: 4.29001e-06 [updatestate_loads_eliminate]: 3.93e-06 [renormalize]: 4.30002e-07 [cse]: 3.821e-05 [Cycle 2]: 0.00016149, [7] [b_1]: 9.847e-05 [b_2]: 3.94e-06 [updatestate_depend_eliminate]: 5.23e-06 [updatestate_assign_eliminate]: 4.05e-06 [updatestate_loads_eliminate]: 3.63e-06 [renormalize]: 7.99992e-08 [cse]: 1.857e-05 [cconv]: 2.144e-05 [opt_after_cconv]: 8.632e-05, [1] [Cycle 1]: 8.202e-05, [7] [c_1]: 2.45e-05 [parameter_eliminate]: 2.21e-06 [updatestate_depend_eliminate]: 4.44e-06 [updatestate_assign_eliminate]: 4.43e-06 [updatestate_loads_eliminate]: 3.63e-06 [cse]: 1.601e-05 [renormalize]: 3.09999e-07 [remove_dup_value]: 1.778e-05 [tuple_transform]: 8.481e-05, [1] [Cycle 1]: 8.104e-05, [3] [d_1]: 5.944e-05 [d_2]: 9.07999e-06 [renormalize]: 1.70003e-07 [add_cache_embedding]: 1.318e-05 [add_recomputation]: 5.898e-05 [cse_after_recomputation]: 2.654e-05, [1] [Cycle 1]: 2.215e-05, [1] [cse]: 1.736e-05 [environ_conv]: 1.004e-05 [label_micro_interleaved_index]: 2.72e-06 [label_fine_grained_interleaved_index]: 2.03001e-06 [assign_add_opt]: 1.4e-06 [slice_recompute_activation]: 2.06001e-06 [micro_interleaved_order_control]: 1.89e-06 [full_micro_interleaved_order_control]: 1.95e-06 [comp_comm_scheduling]: 2.14999e-06 [reorder_send_recv_between_fp_bp]: 2.4e-06 [comm_op_add_attrs]: 1.11e-06 [add_comm_op_reuse_tag]: 8.40002e-07 [overlap_opt_shard_in_pipeline]: 9.70002e-07 [grouped_pairwise_exchange_alltoall]: 1.1e-06 [overlap_recompute_and_grad_model_parallel]: 1.63e-06 [overlap_grad_matmul_and_grad_allreduce]: 8.60004e-07 [split_matmul_comm_elemetwise]: 2e-06 [split_layernorm_comm]: 1.9e-06 [process_send_recv_for_ge]: 7.49998e-07 [handle_group_info]: 9.09997e-07 [auto_monad_reorder]: 2.282e-05 [get_jit_bprop_graph]: 3.48e-06 [eliminate_special_op_node]: 0.00052128 [validate]: 3.864e-05 [distribtued_split]: 1.24001e-06 [task_emit]: 0.00349863 [execute]: 6.31e-06 Sums parse : 0.001515s : 1.16% symbol_resolve.resolve : 0.012443s : 9.49% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000165s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004387s : 3.35% pack_expand : 0.000017s : 0.01% auto_monad : 0.000084s : 0.06% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000192s : 0.15% optimize.opt_a.expand_dump_flag : 0.000015s : 0.01% optimize.opt_a.switch_simplify : 0.000336s : 0.26% optimize.opt_a.a_1 : 0.005408s : 4.13% optimize.opt_a.recompute_prepare : 0.000053s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000079s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000053s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000828s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000074s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000016s : 0.01% optimize.opt_a.parallel : 0.000023s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000009s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000045s : 0.03% optimize.opt_a.virtual_output : 0.000044s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000101s : 0.08% optimize.opt_a.meta_fg_expand : 0.000064s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006130s : 4.68% optimize.opt_a.after_resolve : 0.000159s : 0.12% optimize.opt_a.a_after_grad : 0.000396s : 0.30% optimize.opt_a.renormalize : 0.089997s : 68.66% optimize.opt_a.real_op_eliminate : 0.000139s : 0.11% optimize.opt_a.auto_monad_grad : 0.000303s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000253s : 0.19% optimize.opt_a.cse : 0.000981s : 0.75% optimize.opt_a.a_3 : 0.000981s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000098s : 0.07% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000942s : 0.72% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000057s : 0.04% optimize.cconv : 0.000021s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000059s : 0.05% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000013s : 0.01% optimize.add_recomputation : 0.000059s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000003s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000023s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000521s : 0.40% validate : 0.000039s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003499s : 2.67% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018783 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.35% : 0.016971s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.55% : 0.001042s : 103: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.20% : 0.000038s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.50% : 0.000094s : 69: substitution.replace_applicator 0.06% : 0.000011s : 36: substitution.replace_old_param 0.02% : 0.000004s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000011s : 10: substitution.switch_simplify 0.06% : 0.000012s : 4: substitution.transpose_eliminate 0.65% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.41% : 0.000076s : 70: substitution.tuple_list_get_item_depend_reorder 0.88% : 0.000165s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000077s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.089981 6 92.47% : 0.083208s : 3: renormalize.infer 7.53% : 0.006773s : 3: renormalize.specialize ------[replace.] 0.001256 141 54.93% : 0.000690s : 55: replace.getattr_setattr_resolve 25.66% : 0.000322s : 56: replace.inline 3.66% : 0.000046s : 2: replace.meta_unpack_prepare 7.47% : 0.000094s : 10: replace.switch_simplify 1.63% : 0.000020s : 4: replace.transpose_eliminate 6.65% : 0.000084s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017769 141 94.76% : 0.016837s : 55: match.getattr_setattr_resolve 4.86% : 0.000864s : 56: match.inline 0.10% : 0.000018s : 2: match.meta_unpack_prepare 0.06% : 0.000011s : 10: match.switch_simplify 0.06% : 0.000012s : 4: match.transpose_eliminate 0.15% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006965 119 69.32% : 0.004828s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.68% : 0.002137s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027801 589 0.51% : 0.000142s : 2: opt.transform.meta_unpack_prepare 30.27% : 0.008414s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.29% : 0.000915s : 94: opt.transform.opt_b 65.57% : 0.018229s : 14: opt.transform.opt_resolve 0.23% : 0.000064s : 8: opt.transform.opt_trans_graph 0.06% : 0.000016s : 3: opt.transform.special_op_eliminate . TotalTime = 0.145942, [20] [parse]: 0.00128342 [symbol_resolve]: 0.0125684, [1] [Cycle 1]: 0.0124908, [1] [resolve]: 0.0124731 [combine_like_graphs]: 6.29996e-07 [graph_reusing]: 3.09e-06 [meta_unpack_prepare]: 0.00012968 [pre_cconv]: 4.50003e-07 [abstract_specialize]: 0.00410588 [pack_expand]: 1.475e-05 [auto_monad]: 6.73e-05 [inline]: 1.35e-06 [pre_auto_parallel]: 6.74e-06 [pipeline_split]: 2.01e-06 [optimize]: 0.121525, [35] [py_interpret_to_execute]: 4.65001e-06 [rewriter_before_opt_a]: 0.00018842 [opt_a]: 0.11969, [4] [Cycle 1]: 0.0588874, [30] [expand_dump_flag]: 3.36e-06 [switch_simplify]: 2.68e-05 [a_1]: 0.00039204 [recompute_prepare]: 8.83e-06 [updatestate_depend_eliminate]: 9.51e-06 [updatestate_assign_eliminate]: 6.04e-06 [updatestate_loads_eliminate]: 6.16e-06 [parameter_eliminate]: 4.24001e-06 [a_2]: 7.871e-05 [accelerated_algorithm]: 5.43e-06 [pynative_shard]: 1.09e-06 [auto_parallel]: 3.93e-06 [parallel]: 5.48e-06 [merge_comm]: 3.02e-06 [allreduce_fusion]: 1.6e-06 [virtual_dataset]: 5.05e-06 [get_grad_eliminate_]: 4.65e-06 [virtual_output]: 4.01e-06 [merge_forward]: 7.4e-06 [cell_reuse_recompute_pass]: 4.30002e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.19e-05 [meta_fg_expand]: 0.00727435, [1] [Cycle 1]: 0.00347802, [1] [resolve]: 0.00345932 [after_resolve]: 4.945e-05 [a_after_grad]: 0.00011014 [renormalize]: 0.0498238 [real_op_eliminate]: 4.046e-05 [auto_monad_grad]: 6.998e-05 [auto_monad_eliminator]: 7.844e-05 [cse]: 0.00035199 [a_3]: 0.00031226 [Cycle 2]: 0.0524636, [30] [expand_dump_flag]: 3.85e-06 [switch_simplify]: 0.00010323 [a_1]: 0.00077008 [recompute_prepare]: 1.438e-05 [updatestate_depend_eliminate]: 1.577e-05 [updatestate_assign_eliminate]: 1.251e-05 [updatestate_loads_eliminate]: 1.208e-05 [parameter_eliminate]: 3.88e-06 [a_2]: 0.00019547 [accelerated_algorithm]: 1.728e-05 [pynative_shard]: 1.15e-06 [auto_parallel]: 4.07e-06 [parallel]: 4.53e-06 [merge_comm]: 2.55e-06 [allreduce_fusion]: 1.22e-06 [virtual_dataset]: 1.037e-05 [get_grad_eliminate_]: 9.65e-06 [virtual_output]: 9.46e-06 [merge_forward]: 1.363e-05 [cell_reuse_recompute_pass]: 4.89999e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.374e-05 [meta_fg_expand]: 0.0115397, [5] [Cycle 1]: 0.00032468, [1] [resolve]: 0.00030718 [Cycle 1]: 0.00031645, [1] [resolve]: 0.00029881 [Cycle 1]: 0.00167956, [1] [resolve]: 0.00166098 [Cycle 1]: 0.00032229, [1] [resolve]: 0.00030452 [Cycle 1]: 0.00031459, [1] [resolve]: 0.00029746 [after_resolve]: 7.284e-05 [a_after_grad]: 0.00016227 [renormalize]: 0.0381757 [real_op_eliminate]: 5.307e-05 [auto_monad_grad]: 0.00021118 [auto_monad_eliminator]: 0.00010648 [cse]: 0.00030571 [a_3]: 0.00042176 [Cycle 3]: 0.00478908, [30] [expand_dump_flag]: 4.35e-06 [switch_simplify]: 0.00011546 [a_1]: 0.00123991 [recompute_prepare]: 1.759e-05 [updatestate_depend_eliminate]: 2.77e-05 [updatestate_assign_eliminate]: 1.831e-05 [updatestate_loads_eliminate]: 1.764e-05 [parameter_eliminate]: 4.35e-06 [a_2]: 0.0002875 [accelerated_algorithm]: 2.495e-05 [pynative_shard]: 1.14e-06 [auto_parallel]: 3.85e-06 [parallel]: 4.37e-06 [merge_comm]: 3.6e-06 [allreduce_fusion]: 2.3e-06 [virtual_dataset]: 1.366e-05 [get_grad_eliminate_]: 1.321e-05 [virtual_output]: 1.264e-05 [merge_forward]: 1.947e-05 [cell_reuse_recompute_pass]: 4.29995e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.316e-05 [meta_fg_expand]: 5.048e-05 [after_resolve]: 1.689e-05 [a_after_grad]: 2.121e-05 [renormalize]: 0.00228191 [real_op_eliminate]: 2.026e-05 [auto_monad_grad]: 5.46e-06 [auto_monad_eliminator]: 3.484e-05 [cse]: 0.00020442 [a_3]: 0.00012224 [Cycle 4]: 0.00121515, [30] [expand_dump_flag]: 1.59e-06 [switch_simplify]: 1.509e-05 [a_1]: 0.0002835 [recompute_prepare]: 1.54e-05 [updatestate_depend_eliminate]: 2.026e-05 [updatestate_assign_eliminate]: 1.666e-05 [updatestate_loads_eliminate]: 1.617e-05 [parameter_eliminate]: 2.27e-06 [a_2]: 0.00027746 [accelerated_algorithm]: 2.548e-05 [pynative_shard]: 1.49e-06 [auto_parallel]: 3.84e-06 [parallel]: 3.68999e-06 [merge_comm]: 3.16e-06 [allreduce_fusion]: 2.15e-06 [virtual_dataset]: 1.448e-05 [get_grad_eliminate_]: 1.405e-05 [virtual_output]: 1.291e-05 [merge_forward]: 1.748e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.368e-05 [meta_fg_expand]: 3.377e-05 [after_resolve]: 1.674e-05 [a_after_grad]: 2.095e-05 [renormalize]: 6.99947e-08 [real_op_eliminate]: 1.36e-05 [auto_monad_grad]: 2.83001e-06 [auto_monad_eliminator]: 3.071e-05 [cse]: 8.241e-05 [a_3]: 0.00011391 [py_interpret_to_execute_after_opt_a]: 4.12e-06 [slice_cell_reuse_recomputed_activation]: 1.78e-06 [rewriter_after_opt_a]: 9.484e-05 [convert_after_rewriter]: 2.25e-05 [order_py_execute_after_rewriter]: 1.596e-05 [opt_b]: 0.00111298, [2] [Cycle 1]: 0.0009404, [7] [b_1]: 0.00084347 [b_2]: 5.58e-06 [updatestate_depend_eliminate]: 6.22001e-06 [updatestate_assign_eliminate]: 4.23e-06 [updatestate_loads_eliminate]: 4.29e-06 [renormalize]: 4.60001e-07 [cse]: 3.929e-05 [Cycle 2]: 0.0001626, [7] [b_1]: 9.961e-05 [b_2]: 4.68e-06 [updatestate_depend_eliminate]: 4.69e-06 [updatestate_assign_eliminate]: 3.94e-06 [updatestate_loads_eliminate]: 3.69e-06 [renormalize]: 7.99992e-08 [cse]: 1.867e-05 [cconv]: 1.798e-05 [opt_after_cconv]: 7.191e-05, [1] [Cycle 1]: 6.712e-05, [7] [c_1]: 8.78e-06 [parameter_eliminate]: 2.01e-06 [updatestate_depend_eliminate]: 4.39e-06 [updatestate_assign_eliminate]: 3.60001e-06 [updatestate_loads_eliminate]: 3.55e-06 [cse]: 1.73e-05 [renormalize]: 3.30001e-07 [remove_dup_value]: 1.328e-05 [tuple_transform]: 6.772e-05, [1] [Cycle 1]: 6.404e-05, [3] [d_1]: 4.11e-05 [d_2]: 9.18e-06 [renormalize]: 1.90004e-07 [add_cache_embedding]: 9.83e-06 [add_recomputation]: 5.191e-05 [cse_after_recomputation]: 2.511e-05, [1] [Cycle 1]: 2.138e-05, [1] [cse]: 1.693e-05 [environ_conv]: 8.88e-06 [label_micro_interleaved_index]: 1.57e-06 [label_fine_grained_interleaved_index]: 1.41e-06 [assign_add_opt]: 7.90002e-07 [slice_recompute_activation]: 1.58e-06 [micro_interleaved_order_control]: 1.27e-06 [full_micro_interleaved_order_control]: 8.90002e-07 [comp_comm_scheduling]: 1.12e-06 [reorder_send_recv_between_fp_bp]: 1.56e-06 [comm_op_add_attrs]: 6.50005e-07 [add_comm_op_reuse_tag]: 6.40001e-07 [overlap_opt_shard_in_pipeline]: 6.00005e-07 [grouped_pairwise_exchange_alltoall]: 5.29995e-07 [overlap_recompute_and_grad_model_parallel]: 1.07e-06 [overlap_grad_matmul_and_grad_allreduce]: 5.20005e-07 [split_matmul_comm_elemetwise]: 1.41e-06 [split_layernorm_comm]: 1.06e-06 [process_send_recv_for_ge]: 7.40001e-07 [handle_group_info]: 5.80003e-07 [auto_monad_reorder]: 1.808e-05 [get_jit_bprop_graph]: 4.39999e-07 [eliminate_special_op_node]: 0.00068445 [validate]: 3.657e-05 [distribtued_split]: 1.23e-06 [task_emit]: 0.00530765 [execute]: 4.94e-06 Sums parse : 0.001283s : 0.99% symbol_resolve.resolve : 0.012473s : 9.60% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000130s : 0.10% pre_cconv : 0.000000s : 0.00% abstract_specialize : 0.004106s : 3.16% pack_expand : 0.000015s : 0.01% auto_monad : 0.000067s : 0.05% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000007s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000188s : 0.14% optimize.opt_a.expand_dump_flag : 0.000013s : 0.01% optimize.opt_a.switch_simplify : 0.000261s : 0.20% optimize.opt_a.a_1 : 0.002686s : 2.07% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000073s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000054s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000052s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000839s : 0.65% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.06% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000016s : 0.01% optimize.opt_a.parallel : 0.000018s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000007s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.03% optimize.opt_a.virtual_output : 0.000039s : 0.03% optimize.opt_a.merge_forward : 0.000058s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000084s : 0.06% optimize.opt_a.meta_fg_expand.resolve : 0.006328s : 4.87% optimize.opt_a.after_resolve : 0.000156s : 0.12% optimize.opt_a.a_after_grad : 0.000315s : 0.24% optimize.opt_a.renormalize : 0.090282s : 69.47% optimize.opt_a.real_op_eliminate : 0.000127s : 0.10% optimize.opt_a.auto_monad_grad : 0.000289s : 0.22% optimize.opt_a.auto_monad_eliminator : 0.000250s : 0.19% optimize.opt_a.cse : 0.000945s : 0.73% optimize.opt_a.a_3 : 0.000970s : 0.75% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000095s : 0.07% optimize.convert_after_rewriter : 0.000023s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000943s : 0.73% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000058s : 0.04% optimize.cconv : 0.000018s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000041s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000010s : 0.01% optimize.add_recomputation : 0.000052s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000001s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000001s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000018s : 0.01% get_jit_bprop_graph : 0.000000s : 0.00% eliminate_special_op_node : 0.000684s : 0.53% validate : 0.000037s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005308s : 4.08% execute : 0.000005s : 0.00% Time group info: ------[substitution.] 0.018985 880 0.01% : 0.000003s : 5: substitution.float_depend_g_call 0.12% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.75% : 0.017228s : 59: substitution.getattr_setattr_resolve 0.02% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000001s : 3: substitution.incorporate_call_switch 5.57% : 0.001057s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000029s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.01% : 0.000003s : 5: substitution.partial_eliminate 0.00% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.04% : 0.000008s : 81: substitution.remove_not_recompute_node 0.47% : 0.000090s : 63: substitution.replace_applicator 0.05% : 0.000009s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.60% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.26% : 0.000050s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000068s : 60: substitution.tuple_list_get_item_depend_reorder 0.81% : 0.000154s : 112: substitution.tuple_list_get_item_eliminator 0.35% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.090268 6 92.67% : 0.083651s : 3: renormalize.infer 7.33% : 0.006616s : 3: renormalize.specialize ------[replace.] 0.001252 141 54.41% : 0.000681s : 55: replace.getattr_setattr_resolve 26.60% : 0.000333s : 56: replace.inline 3.69% : 0.000046s : 2: replace.meta_unpack_prepare 7.28% : 0.000091s : 10: replace.switch_simplify 1.46% : 0.000018s : 4: replace.transpose_eliminate 6.56% : 0.000082s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.018054 141 94.71% : 0.017100s : 55: match.getattr_setattr_resolve 4.92% : 0.000888s : 56: match.inline 0.09% : 0.000016s : 2: match.meta_unpack_prepare 0.06% : 0.000010s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006818 119 69.53% : 0.004741s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.47% : 0.002078s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.025193 259 7.37% : 0.001858s : 104: opt.transform.opt_a 3.62% : 0.000911s : 92: opt.transform.opt_b 73.34% : 0.018476s : 14: opt.transform.opt_resolve 0.45% : 0.000113s : 1: opt.transforms.meta_unpack_prepare 14.91% : 0.003756s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000048s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000015s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:33.285.450 [graph_var_manager.cc:1424][EVENT]36564 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:33.285.531 [graph_manager.cc:1248][EVENT]36564 PreRun:PreRun start: graph node size 3, session id 10, graph id 9, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.285.859 [atrace_api.c:28](tid:36564) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.285.895 [trace_rb_log.c:84](tid:36564) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.285.909 [atrace_api.c:32](tid:36564) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:33.285.921 [client_manager.cpp:157][SetProfilingCallback][tid:36564] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.347 [parallel_partitioner.cc:165][EVENT]36564 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.384 [parallel_partitioner.cc:178][EVENT]36564 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.431 [graph_prepare.cc:1378][EVENT]36564 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.623 [graph_manager.cc:1050][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [209] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.649 [graph_manager.cc:1052][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.776 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.807 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.860 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [41] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.874 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.921 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.935 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.286.954 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.287.051 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.287.072 [graph_manager.cc:1054][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [410] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.287.293 [graph_manager.cc:1055][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [207] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.249 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.273 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.283 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.293 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [298] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.312 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.321 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.330 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [9] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.339 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [16] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.288.347 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.289.754 [graph_manager.cc:1056][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2441] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.289.819 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.289.836 [graph_prepare.cc:1982][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.232 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.254 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.264 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.273 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferShapePass is [223] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.282 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.291 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [2] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.299 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.308 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.316 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of InferValuePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.341 [graph_prepare.cc:1983][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [491] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.365 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.378 [graph_prepare.cc:1984][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.392 [graph_prepare.cc:1985][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.405 [graph_prepare.cc:1986][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.423 [graph_prepare.cc:1987][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.438 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.451 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.465 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.551 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.564 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.573 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrintOpPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.582 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.591 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.599 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.607 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.616 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.624 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.632 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.640 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.649 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SnapshotPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.657 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.665 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.673 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.681 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.704 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.719 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.749 [graph_prepare.cc:1988][EVENT]36564 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [316] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.290.768 [graph_manager.cc:1065][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [983] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.303.043 [graph_manager.cc:1077][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12254] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.303.108 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.303.164 [graph_manager.cc:1080][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [89] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.863 [graph_manager.cc:1081][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2683] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.899 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.915 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.926 [graph_manager.cc:1082][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.958 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.973 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.305.987 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.058 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.075 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.108 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.123 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.163 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.181 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.199 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.224 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.240 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.253 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.274 [graph_manager.cc:2700][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [321] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.380 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of EnterPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.393 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.403 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.412 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.420 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of MergePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.428 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CastRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.437 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.445 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.453 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.462 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.470 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.478 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.487 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.495 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.503 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.512 [graph_manager.cc:2741][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [220] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.522 [graph_manager.cc:2752][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.545 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.557 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.573 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.589 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.600 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.626 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.645 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.660 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.673 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.683 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.696 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.708 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.726 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.740 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.749 [graph_manager.cc:2810][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [208] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.778 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.791 [graph_manager.cc:2821][EVENT]36564 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.818 [graph_manager.cc:1087][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [873] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.954 [graph_manager.cc:1088][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [122] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.306.992 [graph_manager.cc:1089][EVENT]36564 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.009 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.024 [graph_manager.cc:1097][EVENT]36564 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.045 [graph_manager.cc:3325][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.254 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.271 [engine_place.cc:144][EVENT]36564 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.281 [engine_place.cc:144][EVENT]36564 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [119] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.351 [graph_manager.cc:3351][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [292] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.376 [graph_manager.cc:3364][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.441 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.458 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.610 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [141] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.652 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [28] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.699 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.731 [graph_manager.cc:3405][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [342] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.307.750 [graph_manager.cc:3412][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.259 [graph_manager.cc:3422][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1495] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.290 [graph_manager.cc:3428][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.412 [graph_manager.cc:3467][EVENT]36564 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.428 [graph_manager.cc:3377][EVENT]36564 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [2040] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.445 [graph_manager.cc:1106][EVENT]36564 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2406] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.457 [graph_manager.cc:1115][EVENT]36564 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.480 [graph_manager.cc:1130][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.512 [graph_manager.cc:1131][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.536 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.554 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.564 [graph_manager.cc:2837][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [35] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.635 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [11] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.656 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.665 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of CondRemovePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.674 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.683 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.691 [base_pass.cc:339][EVENT]36564 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.701 [graph_manager.cc:2864][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [120] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.713 [graph_manager.cc:2872][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.733 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.748 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.763 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.777 [compile_nodes_pass.cc:88][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.789 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.799 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.877 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [69] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.904 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.917 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.930 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.944 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.954 [graph_manager.cc:2927][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [224] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.966 [graph_manager.cc:2937][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.981 [graph_manager.cc:2943][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.309.998 [graph_manager.cc:2950][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.158 [graph_manager.cc:2958][EVENT]36564 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.189 [graph_manager.cc:1132][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [662] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.266 [graph_manager.cc:1135][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [63] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.297 [graph_manager.cc:2975][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.329 [graph_manager.cc:2981][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.344 [pass_manager.cc:82][EVENT]36564 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.354 [graph_manager.cc:2986][EVENT]36564 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.363 [graph_manager.cc:1136][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [82] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.467 [graph_manager.cc:3555][EVENT]36564 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [73] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.550 [engine_partitioner.cc:1139][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.565 [engine_partitioner.cc:1142][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.678 [engine_partitioner.cc:1148][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [103] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.706 [engine_partitioner.cc:1155][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.744 [engine_partitioner.cc:1164][EVENT]36564 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.310.765 [graph_builder.cc:865][EVENT]36564 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [246] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:33.311.103 [logger.cc:1071] 36564 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.135 [task_generator.cc:804][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [68] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.195 [task_generator.cc:805][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [48] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.635 [task_generator.cc:814][EVENT]36564 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [424] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.648 [task_generator.cc:954][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [583] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.710 [task_generator.cc:967][EVENT]36564 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [31] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:33.311.728 [logger.cc:1084] 36564 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.877 [graph_manager.cc:1152][EVENT]36564 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1490] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.895 [graph_manager.cc:1164][EVENT]36564 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.927 [graph_manager.cc:1271][EVENT]36564 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [25669] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.311.938 [graph_manager.cc:1272][EVENT]36564 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.312.239 [atrace_api.c:93](tid:36564) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.312.254 [atrace_api.c:95](tid:36564) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:33.316.550 [graph_converter.cc:838][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1190] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.316.740 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [148] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.135 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [373] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.219 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.234 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [78] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.520 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [276] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.627 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [90] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.665 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.820 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [142] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.889 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [53] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.901 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [66] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.929 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [18] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.956 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.317.981 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of ZeroCopy is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.041 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CEM is [50] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.099 [copy_flow_launch_fuse.cc:395][EVENT]36564 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.110 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [58] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.143 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.167 [base_optimizer.cc:70][EVENT]36564 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.188 [graph_converter.cc:849][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1602] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.318.384 [graph_converter.cc:853][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [186] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.319.015 [graph_converter.cc:857][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [618] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.319.136 [graph_converter.cc:862][EVENT]36564 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [102] micro second. . TotalTime = 0.147396, [20] [parse]: 0.00153146 [symbol_resolve]: 0.0126333, [1] [Cycle 1]: 0.0125516, [1] [resolve]: 0.0125303 [combine_like_graphs]: 8.39995e-07 [graph_reusing]: 3.63e-06 [meta_unpack_prepare]: 0.00016525 [pre_cconv]: 6.99998e-07 [abstract_specialize]: 0.00440734 [pack_expand]: 1.707e-05 [auto_monad]: 8.601e-05 [inline]: 1.8e-06 [pre_auto_parallel]: 9.83e-06 [pipeline_split]: 2.65e-06 [optimize]: 0.124256, [35] [py_interpret_to_execute]: 4.44e-06 [rewriter_before_opt_a]: 0.00019299 [opt_a]: 0.122337, [4] [Cycle 1]: 0.0588698, [30] [expand_dump_flag]: 4.44e-06 [switch_simplify]: 2.684e-05 [a_1]: 0.00073291 [recompute_prepare]: 8.28e-06 [updatestate_depend_eliminate]: 1.089e-05 [updatestate_assign_eliminate]: 6.97e-06 [updatestate_loads_eliminate]: 6.78e-06 [parameter_eliminate]: 4.97e-06 [a_2]: 7.758e-05 [accelerated_algorithm]: 5.38e-06 [pynative_shard]: 1.97e-06 [auto_parallel]: 3.76e-06 [parallel]: 8.79e-06 [merge_comm]: 4.03e-06 [allreduce_fusion]: 2.23e-06 [virtual_dataset]: 5.24e-06 [get_grad_eliminate_]: 4.62e-06 [virtual_output]: 4.40999e-06 [merge_forward]: 8.77e-06 [cell_reuse_recompute_pass]: 1.21001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.164e-05 [meta_fg_expand]: 0.00700675, [1] [Cycle 1]: 0.00320011, [1] [resolve]: 0.00318129 [after_resolve]: 4.951e-05 [a_after_grad]: 0.00013487 [renormalize]: 0.0496355 [real_op_eliminate]: 4.499e-05 [auto_monad_grad]: 7.167e-05 [auto_monad_eliminator]: 8.168e-05 [cse]: 0.00039961 [a_3]: 0.00031394 [Cycle 2]: 0.0533214, [30] [expand_dump_flag]: 3.71999e-06 [switch_simplify]: 0.00013168 [a_1]: 0.00162851 [recompute_prepare]: 1.342e-05 [updatestate_depend_eliminate]: 1.725e-05 [updatestate_assign_eliminate]: 1.259e-05 [updatestate_loads_eliminate]: 1.201e-05 [parameter_eliminate]: 4.33e-06 [a_2]: 0.00019441 [accelerated_algorithm]: 1.911e-05 [pynative_shard]: 1.38e-06 [auto_parallel]: 5.66e-06 [parallel]: 5.09e-06 [merge_comm]: 2.91e-06 [allreduce_fusion]: 1.44e-06 [virtual_dataset]: 1.127e-05 [get_grad_eliminate_]: 1.064e-05 [virtual_output]: 1.068e-05 [merge_forward]: 1.411e-05 [cell_reuse_recompute_pass]: 4.60001e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.352e-05 [meta_fg_expand]: 0.0115692, [5] [Cycle 1]: 0.00032136, [1] [resolve]: 0.00030283 [Cycle 1]: 0.00031548, [1] [resolve]: 0.00029733 [Cycle 1]: 0.00165359, [1] [resolve]: 0.00163488 [Cycle 1]: 0.00031279, [1] [resolve]: 0.00029434 [Cycle 1]: 0.00030575, [1] [resolve]: 0.00028855 [after_resolve]: 7.592e-05 [a_after_grad]: 0.00019966 [renormalize]: 0.0380395 [real_op_eliminate]: 5.85e-05 [auto_monad_grad]: 0.00022194 [auto_monad_eliminator]: 0.00010722 [cse]: 0.00031398 [a_3]: 0.00042411 [Cycle 3]: 0.00605317, [30] [expand_dump_flag]: 4.16001e-06 [switch_simplify]: 0.00015725 [a_1]: 0.00233832 [recompute_prepare]: 1.648e-05 [updatestate_depend_eliminate]: 3.299e-05 [updatestate_assign_eliminate]: 1.85e-05 [updatestate_loads_eliminate]: 1.791e-05 [parameter_eliminate]: 4.76e-06 [a_2]: 0.00027248 [accelerated_algorithm]: 2.502e-05 [pynative_shard]: 1.53999e-06 [auto_parallel]: 4.58e-06 [parallel]: 4.34e-06 [merge_comm]: 4.22e-06 [allreduce_fusion]: 2.49e-06 [virtual_dataset]: 1.49e-05 [get_grad_eliminate_]: 1.426e-05 [virtual_output]: 1.391e-05 [merge_forward]: 1.933e-05 [cell_reuse_recompute_pass]: 4.80002e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.228e-05 [meta_fg_expand]: 5.067e-05 [after_resolve]: 1.795e-05 [a_after_grad]: 3.2e-05 [renormalize]: 0.00239623 [real_op_eliminate]: 2.146e-05 [auto_monad_grad]: 6.24e-06 [auto_monad_eliminator]: 3.47e-05 [cse]: 0.00020883 [a_3]: 0.00012171 [Cycle 4]: 0.00167654, [30] [expand_dump_flag]: 1.4e-06 [switch_simplify]: 1.608e-05 [a_1]: 0.00075478 [recompute_prepare]: 1.571e-05 [updatestate_depend_eliminate]: 2.067e-05 [updatestate_assign_eliminate]: 1.681e-05 [updatestate_loads_eliminate]: 1.656e-05 [parameter_eliminate]: 2.59e-06 [a_2]: 0.0002726 [accelerated_algorithm]: 2.551e-05 [pynative_shard]: 1.53e-06 [auto_parallel]: 3.81e-06 [parallel]: 3.57e-06 [merge_comm]: 3.2e-06 [allreduce_fusion]: 2.03e-06 [virtual_dataset]: 1.529e-05 [get_grad_eliminate_]: 1.493e-05 [virtual_output]: 1.388e-05 [merge_forward]: 1.768e-05 [cell_reuse_recompute_pass]: 3.90006e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.288e-05 [meta_fg_expand]: 1.346e-05 [after_resolve]: 1.694e-05 [a_after_grad]: 3.195e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.48e-05 [auto_monad_grad]: 2.3e-06 [auto_monad_eliminator]: 3.079e-05 [cse]: 8.33e-05 [a_3]: 0.00011187 [py_interpret_to_execute_after_opt_a]: 4.97e-06 [slice_cell_reuse_recomputed_activation]: 2.09999e-06 [rewriter_after_opt_a]: 9.919e-05 [convert_after_rewriter]: 2.375e-05 [order_py_execute_after_rewriter]: 1.721e-05 [opt_b]: 0.00111189, [2] [Cycle 1]: 0.00094106, [7] [b_1]: 0.00084448 [b_2]: 5.21e-06 [updatestate_depend_eliminate]: 6.4e-06 [updatestate_assign_eliminate]: 4.23e-06 [updatestate_loads_eliminate]: 3.94e-06 [renormalize]: 4.1e-07 [cse]: 3.924e-05 [Cycle 2]: 0.00016086, [7] [b_1]: 9.777e-05 [b_2]: 3.99e-06 [updatestate_depend_eliminate]: 5.2e-06 [updatestate_assign_eliminate]: 3.91e-06 [updatestate_loads_eliminate]: 3.71e-06 [renormalize]: 6.00048e-08 [cse]: 1.847e-05 [cconv]: 2.151e-05 [opt_after_cconv]: 8.935e-05, [1] [Cycle 1]: 8.494e-05, [7] [c_1]: 2.449e-05 [parameter_eliminate]: 2.19e-06 [updatestate_depend_eliminate]: 4.45e-06 [updatestate_assign_eliminate]: 4.33e-06 [updatestate_loads_eliminate]: 3.57e-06 [cse]: 1.705e-05 [renormalize]: 4.00003e-07 [remove_dup_value]: 1.871e-05 [tuple_transform]: 8.583e-05, [1] [Cycle 1]: 8.192e-05, [3] [d_1]: 5.982e-05 [d_2]: 9.32e-06 [renormalize]: 2.40005e-07 [add_cache_embedding]: 1.369e-05 [add_recomputation]: 5.971e-05 [cse_after_recomputation]: 2.724e-05, [1] [Cycle 1]: 2.284e-05, [1] [cse]: 1.807e-05 [environ_conv]: 1.027e-05 [label_micro_interleaved_index]: 2.36e-06 [label_fine_grained_interleaved_index]: 2.2e-06 [assign_add_opt]: 1.81999e-06 [slice_recompute_activation]: 2.2e-06 [micro_interleaved_order_control]: 1.9e-06 [full_micro_interleaved_order_control]: 2.15e-06 [comp_comm_scheduling]: 2.1e-06 [reorder_send_recv_between_fp_bp]: 2.29e-06 [comm_op_add_attrs]: 1.04e-06 [add_comm_op_reuse_tag]: 8.30005e-07 [overlap_opt_shard_in_pipeline]: 1.02e-06 [grouped_pairwise_exchange_alltoall]: 1.32e-06 [overlap_recompute_and_grad_model_parallel]: 1.69e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.79996e-07 [split_matmul_comm_elemetwise]: 2.21e-06 [split_layernorm_comm]: 1.56e-06 [process_send_recv_for_ge]: 7.99999e-07 [handle_group_info]: 9.79999e-07 [auto_monad_reorder]: 2.252e-05 [get_jit_bprop_graph]: 3.46e-06 [eliminate_special_op_node]: 0.00051759 [validate]: 3.911e-05 [distribtued_split]: 1.32e-06 [task_emit]: 0.00349508 [execute]: 6.15e-06 Sums parse : 0.001531s : 1.17% symbol_resolve.resolve : 0.012530s : 9.55% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000004s : 0.00% meta_unpack_prepare : 0.000165s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004407s : 3.36% pack_expand : 0.000017s : 0.01% auto_monad : 0.000086s : 0.07% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000010s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000004s : 0.00% optimize.rewriter_before_opt_a : 0.000193s : 0.15% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000332s : 0.25% optimize.opt_a.a_1 : 0.005455s : 4.16% optimize.opt_a.recompute_prepare : 0.000054s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000082s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000055s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000053s : 0.04% optimize.opt_a.parameter_eliminate : 0.000017s : 0.01% optimize.opt_a.a_2 : 0.000817s : 0.62% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000018s : 0.01% optimize.opt_a.parallel : 0.000022s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000100s : 0.08% optimize.opt_a.meta_fg_expand : 0.000064s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.005999s : 4.57% optimize.opt_a.after_resolve : 0.000160s : 0.12% optimize.opt_a.a_after_grad : 0.000398s : 0.30% optimize.opt_a.renormalize : 0.090071s : 68.65% optimize.opt_a.real_op_eliminate : 0.000140s : 0.11% optimize.opt_a.auto_monad_grad : 0.000302s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000254s : 0.19% optimize.opt_a.cse : 0.001006s : 0.77% optimize.opt_a.a_3 : 0.000972s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000099s : 0.08% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000942s : 0.72% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000012s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000058s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000024s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000019s : 0.01% optimize.tuple_transform.d_1 : 0.000060s : 0.05% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000014s : 0.01% optimize.add_recomputation : 0.000060s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000023s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000518s : 0.39% validate : 0.000039s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003495s : 2.66% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018778 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.12% : 0.000023s : 49: substitution.float_tuple_getitem_switch 90.33% : 0.016963s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.58% : 0.001049s : 103: substitution.inline 0.04% : 0.000008s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000034s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.51% : 0.000095s : 69: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.06% : 0.000012s : 10: substitution.switch_simplify 0.06% : 0.000011s : 4: substitution.transpose_eliminate 0.65% : 0.000122s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000056s : 70: substitution.tuple_list_get_item_const_eliminator 0.41% : 0.000076s : 70: substitution.tuple_list_get_item_depend_reorder 0.86% : 0.000161s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000077s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.090055 6 92.49% : 0.083291s : 3: renormalize.infer 7.51% : 0.006764s : 3: renormalize.specialize ------[replace.] 0.001242 141 54.60% : 0.000678s : 55: replace.getattr_setattr_resolve 26.00% : 0.000323s : 56: replace.inline 3.78% : 0.000047s : 2: replace.meta_unpack_prepare 7.30% : 0.000091s : 10: replace.switch_simplify 1.62% : 0.000020s : 4: replace.transpose_eliminate 6.69% : 0.000083s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017769 141 94.72% : 0.016831s : 55: match.getattr_setattr_resolve 4.90% : 0.000871s : 56: match.inline 0.10% : 0.000018s : 2: match.meta_unpack_prepare 0.07% : 0.000012s : 10: match.switch_simplify 0.06% : 0.000011s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.007010 119 69.51% : 0.004873s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.49% : 0.002137s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027804 589 0.51% : 0.000142s : 2: opt.transform.meta_unpack_prepare 30.36% : 0.008441s : 461: opt.transform.opt_a 0.07% : 0.000020s : 7: opt.transform.opt_after_cconv 3.29% : 0.000916s : 94: opt.transform.opt_b 65.48% : 0.018205s : 14: opt.transform.opt_resolve 0.23% : 0.000064s : 8: opt.transform.opt_trans_graph 0.06% : 0.000016s : 3: opt.transform.special_op_eliminate . TotalTime = 0.146646, [20] [parse]: 0.00131455 [symbol_resolve]: 0.0125158, [1] [Cycle 1]: 0.0124506, [1] [resolve]: 0.0124327 [combine_like_graphs]: 7.59996e-07 [graph_reusing]: 2.84e-06 [meta_unpack_prepare]: 0.00012787 [pre_cconv]: 5.00004e-07 [abstract_specialize]: 0.00416509 [pack_expand]: 1.514e-05 [auto_monad]: 6.861e-05 [inline]: 1.29001e-06 [pre_auto_parallel]: 6.82e-06 [pipeline_split]: 1.79e-06 [optimize]: 0.122173, [35] [py_interpret_to_execute]: 4.8e-06 [rewriter_before_opt_a]: 0.00019049 [opt_a]: 0.120336, [4] [Cycle 1]: 0.0591413, [30] [expand_dump_flag]: 3.91e-06 [switch_simplify]: 2.699e-05 [a_1]: 0.00044497 [recompute_prepare]: 9.2e-06 [updatestate_depend_eliminate]: 9.31e-06 [updatestate_assign_eliminate]: 6.94999e-06 [updatestate_loads_eliminate]: 6.15e-06 [parameter_eliminate]: 4.5e-06 [a_2]: 7.958e-05 [accelerated_algorithm]: 5.66999e-06 [pynative_shard]: 1.07e-06 [auto_parallel]: 3.38e-06 [parallel]: 5.47e-06 [merge_comm]: 2.99e-06 [allreduce_fusion]: 1.62e-06 [virtual_dataset]: 5.24e-06 [get_grad_eliminate_]: 4.68001e-06 [virtual_output]: 4.27e-06 [merge_forward]: 7.2e-06 [cell_reuse_recompute_pass]: 4.1e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.207e-05 [meta_fg_expand]: 0.00710831, [1] [Cycle 1]: 0.0033371, [1] [resolve]: 0.00331795 [after_resolve]: 4.953e-05 [a_after_grad]: 0.00010897 [renormalize]: 0.0501825 [real_op_eliminate]: 4.107e-05 [auto_monad_grad]: 7.202e-05 [auto_monad_eliminator]: 7.981e-05 [cse]: 0.00035434 [a_3]: 0.00030954 [Cycle 2]: 0.0527044, [30] [expand_dump_flag]: 3.88e-06 [switch_simplify]: 0.00010505 [a_1]: 0.00076906 [recompute_prepare]: 1.378e-05 [updatestate_depend_eliminate]: 1.635e-05 [updatestate_assign_eliminate]: 1.282e-05 [updatestate_loads_eliminate]: 1.231e-05 [parameter_eliminate]: 4.04e-06 [a_2]: 0.00019774 [accelerated_algorithm]: 3.51e-05 [pynative_shard]: 1.25e-06 [auto_parallel]: 4.16e-06 [parallel]: 4.71e-06 [merge_comm]: 2.35e-06 [allreduce_fusion]: 1.35e-06 [virtual_dataset]: 1.112e-05 [get_grad_eliminate_]: 9.86e-06 [virtual_output]: 9.44e-06 [merge_forward]: 1.431e-05 [cell_reuse_recompute_pass]: 5.19998e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.388e-05 [meta_fg_expand]: 0.0114665, [5] [Cycle 1]: 0.00032553, [1] [resolve]: 0.00030723 [Cycle 1]: 0.00031882, [1] [resolve]: 0.00030125 [Cycle 1]: 0.00171734, [1] [resolve]: 0.00169951 [Cycle 1]: 0.00032305, [1] [resolve]: 0.00030565 [Cycle 1]: 0.00031826, [1] [resolve]: 0.0003004 [after_resolve]: 7.355e-05 [a_after_grad]: 0.00016108 [renormalize]: 0.0384607 [real_op_eliminate]: 5.219e-05 [auto_monad_grad]: 0.00021525 [auto_monad_eliminator]: 0.00010596 [cse]: 0.00030923 [a_3]: 0.00041941 [Cycle 3]: 0.00496079, [30] [expand_dump_flag]: 4.30999e-06 [switch_simplify]: 0.00011337 [a_1]: 0.00128814 [recompute_prepare]: 1.8e-05 [updatestate_depend_eliminate]: 2.677e-05 [updatestate_assign_eliminate]: 1.78e-05 [updatestate_loads_eliminate]: 1.684e-05 [parameter_eliminate]: 4.35e-06 [a_2]: 0.00027543 [accelerated_algorithm]: 2.478e-05 [pynative_shard]: 1.2e-06 [auto_parallel]: 3.99999e-06 [parallel]: 4.02e-06 [merge_comm]: 3.55e-06 [allreduce_fusion]: 2.26e-06 [virtual_dataset]: 1.424e-05 [get_grad_eliminate_]: 1.348e-05 [virtual_output]: 1.255e-05 [merge_forward]: 1.905e-05 [cell_reuse_recompute_pass]: 4.39999e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.312e-05 [meta_fg_expand]: 4.855e-05 [after_resolve]: 1.63e-05 [a_after_grad]: 2.055e-05 [renormalize]: 0.0023839 [real_op_eliminate]: 2.335e-05 [auto_monad_grad]: 6.25e-06 [auto_monad_eliminator]: 3.531e-05 [cse]: 0.00020575 [a_3]: 0.00012243 [Cycle 4]: 0.00119554, [30] [expand_dump_flag]: 1.45e-06 [switch_simplify]: 1.49e-05 [a_1]: 0.00028517 [recompute_prepare]: 1.549e-05 [updatestate_depend_eliminate]: 2.051e-05 [updatestate_assign_eliminate]: 1.673e-05 [updatestate_loads_eliminate]: 1.633e-05 [parameter_eliminate]: 2.14e-06 [a_2]: 0.00027751 [accelerated_algorithm]: 2.519e-05 [pynative_shard]: 1.56e-06 [auto_parallel]: 3.9e-06 [parallel]: 3.76e-06 [merge_comm]: 3.4e-06 [allreduce_fusion]: 2.09e-06 [virtual_dataset]: 1.433e-05 [get_grad_eliminate_]: 1.425e-05 [virtual_output]: 1.291e-05 [merge_forward]: 1.791e-05 [cell_reuse_recompute_pass]: 4.00003e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.374e-05 [meta_fg_expand]: 1.422e-05 [after_resolve]: 1.585e-05 [a_after_grad]: 2.075e-05 [renormalize]: 8.99963e-08 [real_op_eliminate]: 1.352e-05 [auto_monad_grad]: 2.02e-06 [auto_monad_eliminator]: 3.077e-05 [cse]: 8.361e-05 [a_3]: 0.00011301 [py_interpret_to_execute_after_opt_a]: 4.3e-06 [slice_cell_reuse_recomputed_activation]: 1.71e-06 [rewriter_after_opt_a]: 9.451e-05 [convert_after_rewriter]: 2.206e-05 [order_py_execute_after_rewriter]: 1.616e-05 [opt_b]: 0.00111266, [2] [Cycle 1]: 0.00094214, [7] [b_1]: 0.00084624 [b_2]: 5.83e-06 [updatestate_depend_eliminate]: 5.79e-06 [updatestate_assign_eliminate]: 4.42e-06 [updatestate_loads_eliminate]: 4.5e-06 [renormalize]: 4.30002e-07 [cse]: 3.853e-05 [Cycle 2]: 0.00016088, [7] [b_1]: 9.913e-05 [b_2]: 4.09999e-06 [updatestate_depend_eliminate]: 4.85e-06 [updatestate_assign_eliminate]: 3.84e-06 [updatestate_loads_eliminate]: 3.59e-06 [renormalize]: 7.0002e-08 [cse]: 1.851e-05 [cconv]: 1.841e-05 [opt_after_cconv]: 7.14e-05, [1] [Cycle 1]: 6.687e-05, [7] [c_1]: 8.6e-06 [parameter_eliminate]: 2.14e-06 [updatestate_depend_eliminate]: 4.53e-06 [updatestate_assign_eliminate]: 3.73e-06 [updatestate_loads_eliminate]: 3.61e-06 [cse]: 1.719e-05 [renormalize]: 3.40005e-07 [remove_dup_value]: 1.393e-05 [tuple_transform]: 6.764e-05, [1] [Cycle 1]: 6.388e-05, [3] [d_1]: 4.173e-05 [d_2]: 8.79999e-06 [renormalize]: 1.59998e-07 [add_cache_embedding]: 9.69999e-06 [add_recomputation]: 5.154e-05 [cse_after_recomputation]: 2.578e-05, [1] [Cycle 1]: 2.188e-05, [1] [cse]: 1.725e-05 [environ_conv]: 9.1e-06 [label_micro_interleaved_index]: 1.54e-06 [label_fine_grained_interleaved_index]: 1.36e-06 [assign_add_opt]: 9.79999e-07 [slice_recompute_activation]: 1.49e-06 [micro_interleaved_order_control]: 1.32999e-06 [full_micro_interleaved_order_control]: 9.69994e-07 [comp_comm_scheduling]: 1.23e-06 [reorder_send_recv_between_fp_bp]: 1.47e-06 [comm_op_add_attrs]: 4.69998e-07 [add_comm_op_reuse_tag]: 8.30005e-07 [overlap_opt_shard_in_pipeline]: 6.19999e-07 [grouped_pairwise_exchange_alltoall]: 5.59994e-07 [overlap_recompute_and_grad_model_parallel]: 1.24e-06 [overlap_grad_matmul_and_grad_allreduce]: 4.99997e-07 [split_matmul_comm_elemetwise]: 1.98e-06 [split_layernorm_comm]: 1.03e-06 [process_send_recv_for_ge]: 7.2e-07 [handle_group_info]: 5.60001e-07 [auto_monad_reorder]: 1.863e-05 [get_jit_bprop_graph]: 5.49997e-07 [eliminate_special_op_node]: 0.00071276 [validate]: 3.745e-05 [distribtued_split]: 1.12e-06 [task_emit]: 0.00529388 [execute]: 5.93e-06 Sums parse : 0.001315s : 1.01% symbol_resolve.resolve : 0.012433s : 9.51% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000128s : 0.10% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004165s : 3.18% pack_expand : 0.000015s : 0.01% auto_monad : 0.000069s : 0.05% inline : 0.000001s : 0.00% pre_auto_parallel : 0.000007s : 0.01% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000190s : 0.15% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000260s : 0.20% optimize.opt_a.a_1 : 0.002787s : 2.13% optimize.opt_a.recompute_prepare : 0.000056s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000073s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000054s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000052s : 0.04% optimize.opt_a.parameter_eliminate : 0.000015s : 0.01% optimize.opt_a.a_2 : 0.000830s : 0.63% optimize.opt_a.accelerated_algorithm : 0.000091s : 0.07% optimize.opt_a.pynative_shard : 0.000005s : 0.00% optimize.opt_a.auto_parallel : 0.000015s : 0.01% optimize.opt_a.parallel : 0.000018s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000007s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.03% optimize.opt_a.virtual_output : 0.000039s : 0.03% optimize.opt_a.merge_forward : 0.000058s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000103s : 0.08% optimize.opt_a.meta_fg_expand : 0.000063s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006232s : 4.77% optimize.opt_a.after_resolve : 0.000155s : 0.12% optimize.opt_a.a_after_grad : 0.000311s : 0.24% optimize.opt_a.renormalize : 0.091027s : 69.61% optimize.opt_a.real_op_eliminate : 0.000130s : 0.10% optimize.opt_a.auto_monad_grad : 0.000296s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000252s : 0.19% optimize.opt_a.cse : 0.000953s : 0.73% optimize.opt_a.a_3 : 0.000964s : 0.74% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000095s : 0.07% optimize.convert_after_rewriter : 0.000022s : 0.02% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.opt_b.b_1 : 0.000945s : 0.72% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000057s : 0.04% optimize.cconv : 0.000018s : 0.01% optimize.opt_after_cconv.c_1 : 0.000009s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000042s : 0.03% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000010s : 0.01% optimize.add_recomputation : 0.000052s : 0.04% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% optimize.environ_conv : 0.000009s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000001s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000001s : 0.00% optimize.micro_interleaved_order_control : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000001s : 0.00% optimize.comp_comm_scheduling : 0.000001s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000001s : 0.00% optimize.comm_op_add_attrs : 0.000000s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000001s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000000s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000001s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000019s : 0.01% get_jit_bprop_graph : 0.000001s : 0.00% eliminate_special_op_node : 0.000713s : 0.55% validate : 0.000037s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.005294s : 4.05% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018844 880 0.01% : 0.000002s : 5: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.44% : 0.017043s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000005s : 6: substitution.graph_param_transform 0.01% : 0.000002s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.59% : 0.001054s : 97: substitution.inline 0.04% : 0.000007s : 23: substitution.less_batch_normalization 0.15% : 0.000029s : 23: substitution.meta_unpack_prepare 0.15% : 0.000029s : 40: substitution.minmaximum_grad 0.01% : 0.000003s : 5: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.47% : 0.000089s : 63: substitution.replace_applicator 0.05% : 0.000009s : 36: substitution.replace_old_param 0.02% : 0.000003s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.03% : 0.000006s : 5: substitution.specialize_transform 0.05% : 0.000010s : 10: substitution.switch_simplify 0.31% : 0.000059s : 4: substitution.transpose_eliminate 0.60% : 0.000113s : 60: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000051s : 60: substitution.tuple_list_get_item_const_eliminator 0.36% : 0.000068s : 60: substitution.tuple_list_get_item_depend_reorder 0.83% : 0.000156s : 112: substitution.tuple_list_get_item_eliminator 0.36% : 0.000067s : 60: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.091012 6 92.68% : 0.084348s : 3: renormalize.infer 7.32% : 0.006664s : 3: renormalize.specialize ------[replace.] 0.001282 141 55.17% : 0.000707s : 55: replace.getattr_setattr_resolve 25.87% : 0.000332s : 56: replace.inline 3.55% : 0.000046s : 2: replace.meta_unpack_prepare 7.26% : 0.000093s : 10: replace.switch_simplify 1.54% : 0.000020s : 4: replace.transpose_eliminate 6.61% : 0.000085s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017913 141 94.42% : 0.016914s : 55: match.getattr_setattr_resolve 4.95% : 0.000886s : 56: match.inline 0.09% : 0.000017s : 2: match.meta_unpack_prepare 0.06% : 0.000010s : 10: match.switch_simplify 0.33% : 0.000059s : 4: match.transpose_eliminate 0.15% : 0.000027s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.006944 119 68.99% : 0.004791s : 53: func_graph_cloner_run.FuncGraphClonerGraph 31.01% : 0.002153s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.025152 259 7.33% : 0.001844s : 104: opt.transform.opt_a 3.63% : 0.000913s : 92: opt.transform.opt_b 72.86% : 0.018327s : 14: opt.transform.opt_resolve 0.44% : 0.000112s : 1: opt.transforms.meta_unpack_prepare 15.42% : 0.003878s : 40: opt.transforms.opt_a 0.03% : 0.000007s : 1: opt.transforms.opt_after_cconv 0.03% : 0.000008s : 2: opt.transforms.opt_b 0.19% : 0.000049s : 2: opt.transforms.opt_trans_graph 0.06% : 0.000016s : 3: opt.transforms.special_op_eliminate [INFO] GE(32458,python3.7):2024-01-11-05:34:33.697.628 [graph_var_manager.cc:1424][EVENT]36565 SetAllMemoryMaxValue:The graph_mem_max_size is 27917287424 and the var_mem_max_size is 5368709120 [INFO] GE(32458,python3.7):2024-01-11-05:34:33.697.708 [graph_manager.cc:1248][EVENT]36565 PreRun:PreRun start: graph node size 3, session id 11, graph id 10, graph name online. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.698.040 [atrace_api.c:28](tid:36565) AtraceCreate start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.698.074 [trace_rb_log.c:84](tid:36565) [RUNTIME_ATRACE_DEV64_TS0] create ring buffer success, buffer size : 131152. [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.698.086 [atrace_api.c:32](tid:36565) AtraceCreate end [INFO] TDT(32458,python3.7):2024-01-11-05:34:33.698.099 [client_manager.cpp:157][SetProfilingCallback][tid:36565] [TsdClient] set profiling callback success [INFO] GE(32458,python3.7):2024-01-11-05:34:33.698.517 [parallel_partitioner.cc:165][EVENT]36565 DoPipelinePartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::PipelinePartition is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.698.556 [parallel_partitioner.cc:178][EVENT]36565 DoFlowGraphPartition:[GEPERFTRACE] The time cost of OptimizeSubgraph::FlowGraphPartition is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.698.619 [graph_prepare.cc:1378][EVENT]36565 Init:[GEPERFTRACE] The time cost of FileConstantUtils::ConvertFileConstToConst is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.698.842 [graph_manager.cc:1050][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareInit is [240] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.698.870 [graph_manager.cc:1052][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.HandleSummaryOp is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.006 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ForToWhilePass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.039 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::SavePass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.098 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::NetOutputPass is [46] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.112 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of ProcessNetOutput::DataPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.162 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of CreateSubGraphWithScopePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.177 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubgraphMultiDimsClonePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.196 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of MultiBatchClonePass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.304 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitVariableIntoSubgraphPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.327 [graph_manager.cc:1054][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.NormalizeGraph is [444] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.699.548 [graph_manager.cc:1055][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeGraphInit is [208] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.501 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.525 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.536 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.546 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [300] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.555 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.564 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.573 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.581 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [15] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.700.589 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.003 [graph_manager.cc:1056][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphForQuantize is [2434] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.067 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.085 [graph_prepare.cc:1982][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::ProcessBeforeInfershape is [51] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.480 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.501 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.512 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.521 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferShapePass is [222] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.530 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [8] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.539 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SplitShapeNPass is [1] micro second, call num is [6] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.547 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.556 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.564 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of InferValuePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.590 [graph_prepare.cc:1983][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::FormatAndShapeProcess is [491] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.614 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::MarkForceUnknownForCondPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.626 [graph_prepare.cc:1984][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::CtrlFlowPreProcess is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.640 [graph_prepare.cc:1985][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::multibatch::GetDynamicOutputShape is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.654 [graph_prepare.cc:1986][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::InsertAippOpUtil::Instance().UpdateDataNodeByAipp is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.666 [graph_prepare.cc:1987][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::SaveOriginalGraphToOmModel is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.682 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ShapeOperateOpRemovePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.694 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::ReplaceTransShapePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.708 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::MarkAgnosticPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.790 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.810 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.820 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrintOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.829 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of NoUseReshapeRemovePass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.837 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DropOutPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.846 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssertPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.854 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.862 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedConstPass is [0] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.870 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of StopGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.879 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreventGradientPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.887 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of PlaceholderWithDefaultPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.895 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SnapshotPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.904 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of GuaranteeConstPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.912 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of VarIsInitializedOpPass is [5] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.920 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ParallelConcatStartOpPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.928 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.951 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::PrunePass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.964 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareOptimize::HcclMemcpyPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.702.995 [graph_prepare.cc:1988][EVENT]36565 PrepareDynShape:[GEPERFTRACE] The time cost of Prepare::PrepareOptimize is [319] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.703.008 [graph_manager.cc:1065][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareDynShape is [975] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.715.336 [graph_manager.cc:1077][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraph is [12308] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.715.401 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PrepareRunningFormatRefiner::VariablePrepareOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.715.455 [graph_manager.cc:1080][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.preparer.PrepareRunningFormatRefiner is [87] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.149 [graph_manager.cc:1081][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeOriginalGraphJudgeInsert is [2678] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.194 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of SubexpressionMigrationPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.210 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of UnusedArgsCleanPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.220 [graph_manager.cc:1082][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::SubexpressionMigration is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.251 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeInputMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.267 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SwitchDataEdgesBypass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.280 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::ConstantFuseSamePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.350 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CSEBeforeFuseDataNodesWithCommonInputPass is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.366 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::FuseDataNodesWithCommonInputPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.399 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::CommonSubexpressionEliminationPass is [21] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.415 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::PermutePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.454 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::SameTransdataBreadthFusionPass is [29] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.473 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::VariableOpPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.490 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpWithoutReshapeFusionPass is [7] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.515 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::TransOpBreadthFusionPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.530 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::DataFlowPreparePass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.542 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_1::MergeUnknownShapeNPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.551 [graph_manager.cc:2700][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_1 is [304] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.658 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of EnterPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.672 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AddNPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.681 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchDeadBranchElimination is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.690 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of SwitchLogicRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.699 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of MergePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.718 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CastRemovePass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.727 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransposeTransDataPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.736 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.744 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpSymmetryEliminationPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.752 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of TransOpNearbyAllreduceFusionPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.760 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReplaceWithEmptyConstPass is [7] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.768 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionComputePass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.777 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [10] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.785 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.793 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of UselessControlOutRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.802 [graph_manager.cc:2741][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_2 is [233] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.811 [graph_manager.cc:2752][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of extern constant folding is [0] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.834 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::Migration is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.847 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ArgsClean is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.863 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::PrunePass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.879 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::NextIterationPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.891 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ControlTriggerPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.903 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MergeToStreamMergePass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.923 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SwitchToStreamSwitchPass is [9] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.938 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::AttachStreamLabelPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.951 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::MultiBatchPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.961 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::SubgraphMultiDimsPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.979 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::IteratorOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.718.991 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::VariableRefUselessControlOutDeletePass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.010 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::ReshapeRecoveryPass is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.023 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage1_3::RemoveSameConstPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.033 [graph_manager.cc:2810][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1_3 is [203] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.063 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of IdentityPass is [2] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.075 [graph_manager.cc:2821][EVENT]36565 OptimizeStage1:[GEPERFTRACE] The time cost of GraphPrepare::node_pass is [33] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.103 [graph_manager.cc:1087][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage1 is [864] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.237 [graph_manager.cc:1088][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeAfterStage1 is [121] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.275 [graph_manager.cc:1089][EVENT]36565 PreRunOptimizeOriginalGraph:[GEPERFTRACE] The time cost of GraphManager::GraphUtilsEx::InferShapeInNeed is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.292 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of PreRun::CtrlEdgeTransferPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.307 [graph_manager.cc:1097][EVENT]36565 PreRunOptimizeOriginalGraph:PreRun:PreRunOptimizeOriginalGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.328 [graph_manager.cc:3325][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::StagePartition is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.540 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_GE_LOCAL_OP_STORE::CheckSupported is [8] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.558 [engine_place.cc:144][EVENT]36565 Run:The time cost of DNN_VM_RTS_OP_STORE::CheckSupported is [10] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.567 [engine_place.cc:144][EVENT]36565 Run:The time cost of aicpu_ascend_kernel::CheckSupported is [121] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.638 [graph_manager.cc:3351][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::GraphPartitionDynamicShape is [296] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.655 [graph_manager.cc:3364][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::CompositeEngine is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.718 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.734 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.882 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [138] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.921 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.719.977 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [37] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.720.009 [graph_manager.cc:3405][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::Partition1 is [342] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.720.028 [graph_manager.cc:3412][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPreProc is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.454 [graph_manager.cc:3422][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubGraph is [1411] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.484 [graph_manager.cc:3428][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::SetSubgraphPostProc is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.603 [graph_manager.cc:3467][EVENT]36565 SubgraphPartitionAndOptimization:[GEPERFTRACE] The time cost of OptimizeSubgraph::MergeSubGraph is [100] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.620 [graph_manager.cc:3377][EVENT]36565 OptimizeSubgraph:[GEPERFTRACE] The time cost of OptimizeSubgraph::SubgraphPartitionAndOptimization::AtomicEngine is [1954] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.635 [graph_manager.cc:1106][EVENT]36565 PreRunOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeSubgraph is [2315] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.649 [graph_manager.cc:1115][EVENT]36565 PreRunOptimizeSubGraph:PreRun:PreRunOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.670 [graph_manager.cc:1130][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.ReplacePrecompiledNodeWithOmGraph is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.704 [graph_manager.cc:1131][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::stages.optimizer.OptimizeWholeGraph is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.728 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::LinkGenMaskNodesPass is [6] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.744 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::HcclContinuousMemcpyPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.754 [graph_manager.cc:2837][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses is [32] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.824 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ConstantFoldingPass is [12] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.837 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of ReshapeRemovePass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.847 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of CondRemovePass is [4] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.855 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of BitcastPass is [1] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.864 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of AssignRemovePass is [6] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.872 [base_pass.cc:339][EVENT]36565 Run:[GEPERFTRACE] The time cost of DimensionAdjustPass is [3] micro second, call num is [3] [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.889 [graph_manager.cc:2864][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::MergedGraphNameToPasses is [119] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.901 [graph_manager.cc:2872][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::RemoveIsolatedConst is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.921 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::MultiBatchPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.935 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::RefIdentityDeleteOpPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.950 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::VariableRefDeleteOpPass is [5] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.965 [compile_nodes_pass.cc:88][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.976 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::CompileNodesPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.721.986 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::SwapSpacePass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.064 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::InputOutputConnectionIdentifyPass is [68] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.090 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::AtomicAddrCleanPass is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.104 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::AfterMergePasses::EndOfSequenceAddControlPass is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.117 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::SubgraphPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.129 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize::AttachStreamLabelPass is [3] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.138 [graph_manager.cc:2927][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of OptimizeStage2::ControlAttrOptimize is [220] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.151 [graph_manager.cc:2937][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ModelBuilder::AssignFunctionalLabels is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.164 [graph_manager.cc:2943][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of MemcpyAddrAsyncPass::Run. is [4] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.175 [graph_manager.cc:2950][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of BufferPoolMemoryPass::Run. is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.333 [graph_manager.cc:2958][EVENT]36565 OptimizeStage2:[GEPERFTRACE] The time cost of ParallelGroupPass::Run. is [34] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.364 [graph_manager.cc:1132][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::OptimizeStage2 is [646] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.439 [graph_manager.cc:1135][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::GetCompilerStages(graph_node->GetGraphId()).optimizer.OptimizeGraphBeforeBuild is [60] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.469 [graph_manager.cc:2975][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of HandleMemoryRWConflict is [14] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.507 [graph_manager.cc:2981][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of MemLayoutConflictOptimizer::Run. is [17] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.520 [pass_manager.cc:82][EVENT]36565 Run:[GEPERFTRACE] The time cost of OptimizeStage2::SetFftsPlusAttrPass is [1] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.531 [graph_manager.cc:2986][EVENT]36565 MemConflictProc:[GEPERFTRACE] The time cost of SetFftsPlusAttrPass::last_passes.Run is [13] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.540 [graph_manager.cc:1136][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::MemConflictProc is [86] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.644 [graph_manager.cc:3555][EVENT]36565 Build:[GEPERFTRACE] The time cost of GraphManager::RecoverIrDefinitionAndModifyAippData is [72] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.726 [engine_partitioner.cc:1139][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionInitialize is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.741 [engine_partitioner.cc:1142][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionMarkClusters is [2] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.852 [engine_partitioner.cc:1148][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSplitSubGraphs is [102] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.880 [engine_partitioner.cc:1155][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionSortSubGraphs is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.919 [engine_partitioner.cc:1164][EVENT]36565 PartitionSubGraph:[GEPERFTRACE] The time cost of EnginePartitioner::PartitionAddPartitionsToGraphNode is [27] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.722.940 [graph_builder.cc:865][EVENT]36565 SecondPartition:[GEPERFTRACE] The time cost of EnginePartitioner::Partition2 is [245] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:33.723.296 [logger.cc:1071] 36565 ModelBindStream: model_id=832, stream_id=1089, flag=0. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.723.326 [task_generator.cc:804][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::SetStreamCtx is [84] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.723.385 [task_generator.cc:805][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::PrepareForGenerateTask is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.723.841 [task_generator.cc:814][EVENT]36565 GenerateTask:[GEPERFTRACE] The time cost of TaskGenerator::DoGenerateTask is [442] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.723.855 [task_generator.cc:954][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::GenerateTask is [614] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.723.911 [task_generator.cc:967][EVENT]36565 GetTaskInfo:[GEPERFTRACE] The time cost of TaskGenerator::AddModelTaskToModel is [32] micro second. [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:33.723.928 [logger.cc:1084] 36565 ModelUnbindStream: model_id=832, stream_id=1089, [INFO] GE(32458,python3.7):2024-01-11-05:34:33.724.075 [graph_manager.cc:1152][EVENT]36565 PreRunAfterOptimizeSubGraph:[GEPERFTRACE] The time cost of GraphManager::Build is [1511] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.724.092 [graph_manager.cc:1164][EVENT]36565 PreRunAfterOptimizeSubGraph:PreRun:PreRunAfterOptimizeSubGraph success. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.724.125 [graph_manager.cc:1271][EVENT]36565 PreRun:[GEPERFTRACE] The time cost of FlowModelBuild is [25696] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.724.136 [graph_manager.cc:1272][EVENT]36565 PreRun:[GEPERFTRACE] GE PreRun End [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.724.448 [atrace_api.c:93](tid:36565) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:33.724.464 [atrace_api.c:95](tid:36565) AtraceDestroy end [INFO] GE(32458,python3.7):2024-01-11-05:34:33.728.722 [graph_converter.cc:838][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CreateMainNode is [1209] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.728.915 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [151] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.729.334 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [397] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.729.417 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [61] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.729.431 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [77] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.729.716 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [274] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.729.822 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [89] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.729.857 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [20] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.008 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [139] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.077 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [54] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.088 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [66] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.116 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [19] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.142 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.167 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of ZeroCopy is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.226 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CEM is [49] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.283 [copy_flow_launch_fuse.cc:395][EVENT]36565 Run:[GEPERFTRACE] The time cost of Pass::CopyFlowLaunchFuse is [47] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.293 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of CopyFlowLaunch is [58] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.317 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of TrustOutTensor is [16] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.341 [base_optimizer.cc:70][EVENT]36565 Run:[GEPERFTRACE] The time cost of AicpuFuseHostInputs is [15] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.362 [graph_converter.cc:849][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::RunAllPass is [1603] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.730.557 [graph_converter.cc:853][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::TopologicalSorting is [184] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.731.187 [graph_converter.cc:857][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::AppendGraphLevelData is [616] micro second. [INFO] GE(32458,python3.7):2024-01-11-05:34:33.731.316 [graph_converter.cc:862][EVENT]36565 ConvertComputeGraphToExecuteGraph:[GEPERFTRACE] The time cost of ConvertComputeGraphToExecuteGraph::CalculatePriority is [100] micro second. . TotalTime = 0.148994, [20] [parse]: 0.00148916 [symbol_resolve]: 0.0126214, [1] [Cycle 1]: 0.0125375, [1] [resolve]: 0.0125178 [combine_like_graphs]: 8.2e-07 [graph_reusing]: 3.43e-06 [meta_unpack_prepare]: 0.00016665 [pre_cconv]: 6.40001e-07 [abstract_specialize]: 0.00438855 [pack_expand]: 1.745e-05 [auto_monad]: 8.398e-05 [inline]: 1.63e-06 [pre_auto_parallel]: 1.085e-05 [pipeline_split]: 2.51e-06 [optimize]: 0.125935, [35] [py_interpret_to_execute]: 4.65e-06 [rewriter_before_opt_a]: 0.00019363 [opt_a]: 0.123977, [4] [Cycle 1]: 0.0593084, [30] [expand_dump_flag]: 4.53e-06 [switch_simplify]: 2.735e-05 [a_1]: 0.0007343 [recompute_prepare]: 8.31e-06 [updatestate_depend_eliminate]: 1.054e-05 [updatestate_assign_eliminate]: 7.41e-06 [updatestate_loads_eliminate]: 7.37e-06 [parameter_eliminate]: 5.44e-06 [a_2]: 7.569e-05 [accelerated_algorithm]: 5.22e-06 [pynative_shard]: 1.69e-06 [auto_parallel]: 3.41e-06 [parallel]: 8.94001e-06 [merge_comm]: 4.65e-06 [allreduce_fusion]: 2.15e-06 [virtual_dataset]: 5.66e-06 [get_grad_eliminate_]: 4.83e-06 [virtual_output]: 4.39e-06 [merge_forward]: 8.72e-06 [cell_reuse_recompute_pass]: 9.09997e-07 [cell_reuse_handle_not_recompute_node_pass]: 1.224e-05 [meta_fg_expand]: 0.00705419, [1] [Cycle 1]: 0.00320354, [1] [resolve]: 0.00318441 [after_resolve]: 4.931e-05 [a_after_grad]: 0.0001328 [renormalize]: 0.0500479 [real_op_eliminate]: 4.64e-05 [auto_monad_grad]: 7.435e-05 [auto_monad_eliminator]: 8.121e-05 [cse]: 0.00037613 [a_3]: 0.00031269 [Cycle 2]: 0.0545757, [30] [expand_dump_flag]: 4e-06 [switch_simplify]: 0.00013636 [a_1]: 0.0016258 [recompute_prepare]: 1.294e-05 [updatestate_depend_eliminate]: 1.648e-05 [updatestate_assign_eliminate]: 1.287e-05 [updatestate_loads_eliminate]: 1.259e-05 [parameter_eliminate]: 3.81e-06 [a_2]: 0.00019403 [accelerated_algorithm]: 1.899e-05 [pynative_shard]: 1.3e-06 [auto_parallel]: 4.63e-06 [parallel]: 5.06001e-06 [merge_comm]: 2.5e-06 [allreduce_fusion]: 1.52e-06 [virtual_dataset]: 1.131e-05 [get_grad_eliminate_]: 1.049e-05 [virtual_output]: 1.051e-05 [merge_forward]: 1.332e-05 [cell_reuse_recompute_pass]: 4.79995e-07 [cell_reuse_handle_not_recompute_node_pass]: 2.359e-05 [meta_fg_expand]: 0.0117904, [5] [Cycle 1]: 0.00031662, [1] [resolve]: 0.00029814 [Cycle 1]: 0.00031012, [1] [resolve]: 0.00029246 [Cycle 1]: 0.00166452, [1] [resolve]: 0.00164559 [Cycle 1]: 0.00034651, [1] [resolve]: 0.00032774 [Cycle 1]: 0.00031359, [1] [resolve]: 0.00029501 [after_resolve]: 7.644e-05 [a_after_grad]: 0.00020085 [renormalize]: 0.0390609 [real_op_eliminate]: 5.953e-05 [auto_monad_grad]: 0.00022761 [auto_monad_eliminator]: 0.00010681 [cse]: 0.00031117 [a_3]: 0.00042521 [Cycle 3]: 0.00601039, [30] [expand_dump_flag]: 5.03e-06 [switch_simplify]: 0.00015915 [a_1]: 0.00233453 [recompute_prepare]: 1.64e-05 [updatestate_depend_eliminate]: 3.151e-05 [updatestate_assign_eliminate]: 1.826e-05 [updatestate_loads_eliminate]: 1.697e-05 [parameter_eliminate]: 4.87e-06 [a_2]: 0.00027341 [accelerated_algorithm]: 2.58e-05 [pynative_shard]: 1.29e-06 [auto_parallel]: 9.2e-06 [parallel]: 4.76e-06 [merge_comm]: 4.03e-06 [allreduce_fusion]: 2.45e-06 [virtual_dataset]: 1.531e-05 [get_grad_eliminate_]: 1.467e-05 [virtual_output]: 1.43e-05 [merge_forward]: 2.039e-05 [cell_reuse_recompute_pass]: 4.49996e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.309e-05 [meta_fg_expand]: 5.082e-05 [after_resolve]: 1.784e-05 [a_after_grad]: 3.201e-05 [renormalize]: 0.00234859 [real_op_eliminate]: 2.151e-05 [auto_monad_grad]: 6.27e-06 [auto_monad_eliminator]: 3.515e-05 [cse]: 0.00020982 [a_3]: 0.00012141 [Cycle 4]: 0.00166278, [30] [expand_dump_flag]: 1.3e-06 [switch_simplify]: 1.527e-05 [a_1]: 0.00074365 [recompute_prepare]: 1.601e-05 [updatestate_depend_eliminate]: 2.057e-05 [updatestate_assign_eliminate]: 1.711e-05 [updatestate_loads_eliminate]: 1.643e-05 [parameter_eliminate]: 2.1e-06 [a_2]: 0.00027452 [accelerated_algorithm]: 2.577e-05 [pynative_shard]: 1.45e-06 [auto_parallel]: 3.92e-06 [parallel]: 3.51e-06 [merge_comm]: 3.19001e-06 [allreduce_fusion]: 1.99e-06 [virtual_dataset]: 1.516e-05 [get_grad_eliminate_]: 1.486e-05 [virtual_output]: 1.408e-05 [merge_forward]: 1.737e-05 [cell_reuse_recompute_pass]: 5.10001e-07 [cell_reuse_handle_not_recompute_node_pass]: 3.284e-05 [meta_fg_expand]: 1.373e-05 [after_resolve]: 1.658e-05 [a_after_grad]: 3.182e-05 [renormalize]: 7.0002e-08 [real_op_eliminate]: 1.434e-05 [auto_monad_grad]: 2.52e-06 [auto_monad_eliminator]: 3.08e-05 [cse]: 8.332e-05 [a_3]: 0.00011204 [py_interpret_to_execute_after_opt_a]: 4.43e-06 [slice_cell_reuse_recomputed_activation]: 2.65e-06 [rewriter_after_opt_a]: 0.00010474 [convert_after_rewriter]: 2.436e-05 [order_py_execute_after_rewriter]: 1.688e-05 [opt_b]: 0.00114036, [2] [Cycle 1]: 0.00097021, [7] [b_1]: 0.00087431 [b_2]: 5.04e-06 [updatestate_depend_eliminate]: 5.94e-06 [updatestate_assign_eliminate]: 4.22999e-06 [updatestate_loads_eliminate]: 3.91999e-06 [renormalize]: 4.19997e-07 [cse]: 4.021e-05 [Cycle 2]: 0.0001601, [7] [b_1]: 9.795e-05 [b_2]: 3.93e-06 [updatestate_depend_eliminate]: 5.06e-06 [updatestate_assign_eliminate]: 3.86e-06 [updatestate_loads_eliminate]: 3.82e-06 [renormalize]: 7.0002e-08 [cse]: 1.891e-05 [cconv]: 2.202e-05 [opt_after_cconv]: 9.011e-05, [1] [Cycle 1]: 8.532e-05, [7] [c_1]: 2.57e-05 [parameter_eliminate]: 2.11001e-06 [updatestate_depend_eliminate]: 4.64e-06 [updatestate_assign_eliminate]: 4.4e-06 [updatestate_loads_eliminate]: 3.58e-06 [cse]: 1.711e-05 [renormalize]: 3.00002e-07 [remove_dup_value]: 1.823e-05 [tuple_transform]: 8.524e-05, [1] [Cycle 1]: 8.132e-05, [3] [d_1]: 6.01e-05 [d_2]: 9.31e-06 [renormalize]: 2.09999e-07 [add_cache_embedding]: 1.431e-05 [add_recomputation]: 5.994e-05 [cse_after_recomputation]: 2.74e-05, [1] [Cycle 1]: 2.293e-05, [1] [cse]: 1.802e-05 [environ_conv]: 1.023e-05 [label_micro_interleaved_index]: 2.47e-06 [label_fine_grained_interleaved_index]: 2.32999e-06 [assign_add_opt]: 1.5e-06 [slice_recompute_activation]: 1.99e-06 [micro_interleaved_order_control]: 1.88001e-06 [full_micro_interleaved_order_control]: 2.12e-06 [comp_comm_scheduling]: 2.65001e-06 [reorder_send_recv_between_fp_bp]: 2.24e-06 [comm_op_add_attrs]: 9.70002e-07 [add_comm_op_reuse_tag]: 8.49999e-07 [overlap_opt_shard_in_pipeline]: 1.34001e-06 [grouped_pairwise_exchange_alltoall]: 1.08e-06 [overlap_recompute_and_grad_model_parallel]: 1.67e-06 [overlap_grad_matmul_and_grad_allreduce]: 7.09995e-07 [split_matmul_comm_elemetwise]: 2.03e-06 [split_layernorm_comm]: 1.60999e-06 [process_send_recv_for_ge]: 9.90003e-07 [handle_group_info]: 1.22e-06 [auto_monad_reorder]: 2.19e-05 [get_jit_bprop_graph]: 3.31e-06 [eliminate_special_op_node]: 0.00051385 [validate]: 3.962e-05 [distribtued_split]: 1.22e-06 [task_emit]: 0.00348868 [execute]: 6.37e-06 Sums parse : 0.001489s : 1.12% symbol_resolve.resolve : 0.012518s : 9.44% combine_like_graphs : 0.000001s : 0.00% graph_reusing : 0.000003s : 0.00% meta_unpack_prepare : 0.000167s : 0.13% pre_cconv : 0.000001s : 0.00% abstract_specialize : 0.004389s : 3.31% pack_expand : 0.000017s : 0.01% auto_monad : 0.000084s : 0.06% inline : 0.000002s : 0.00% pre_auto_parallel : 0.000011s : 0.01% pipeline_split : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000194s : 0.15% optimize.opt_a.expand_dump_flag : 0.000015s : 0.01% optimize.opt_a.switch_simplify : 0.000338s : 0.26% optimize.opt_a.a_1 : 0.005438s : 4.10% optimize.opt_a.recompute_prepare : 0.000054s : 0.04% optimize.opt_a.updatestate_depend_eliminate : 0.000079s : 0.06% optimize.opt_a.updatestate_assign_eliminate : 0.000056s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000053s : 0.04% optimize.opt_a.parameter_eliminate : 0.000016s : 0.01% optimize.opt_a.a_2 : 0.000818s : 0.62% optimize.opt_a.accelerated_algorithm : 0.000076s : 0.06% optimize.opt_a.pynative_shard : 0.000006s : 0.00% optimize.opt_a.auto_parallel : 0.000021s : 0.02% optimize.opt_a.parallel : 0.000022s : 0.02% optimize.opt_a.merge_comm : 0.000014s : 0.01% optimize.opt_a.allreduce_fusion : 0.000008s : 0.01% optimize.opt_a.virtual_dataset : 0.000047s : 0.04% optimize.opt_a.get_grad_eliminate_ : 0.000045s : 0.03% optimize.opt_a.virtual_output : 0.000043s : 0.03% optimize.opt_a.merge_forward : 0.000060s : 0.05% optimize.opt_a.cell_reuse_recompute_pass : 0.000002s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.08% optimize.opt_a.meta_fg_expand : 0.000065s : 0.05% optimize.opt_a.meta_fg_expand.resolve : 0.006043s : 4.56% optimize.opt_a.after_resolve : 0.000160s : 0.12% optimize.opt_a.a_after_grad : 0.000397s : 0.30% optimize.opt_a.renormalize : 0.091457s : 68.99% optimize.opt_a.real_op_eliminate : 0.000142s : 0.11% optimize.opt_a.auto_monad_grad : 0.000311s : 0.23% optimize.opt_a.auto_monad_eliminator : 0.000254s : 0.19% optimize.opt_a.cse : 0.000980s : 0.74% optimize.opt_a.a_3 : 0.000971s : 0.73% optimize.py_interpret_to_execute_after_opt_a : 0.000004s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000105s : 0.08% optimize.convert_after_rewriter : 0.000024s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.opt_b.b_1 : 0.000972s : 0.73% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.01% optimize.opt_b.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000059s : 0.04% optimize.cconv : 0.000022s : 0.02% optimize.opt_after_cconv.c_1 : 0.000026s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000060s : 0.05% optimize.tuple_transform.d_2 : 0.000009s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.add_cache_embedding : 0.000014s : 0.01% optimize.add_recomputation : 0.000060s : 0.05% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000010s : 0.01% optimize.label_micro_interleaved_index : 0.000002s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.comp_comm_scheduling : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000002s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.process_send_recv_for_ge : 0.000001s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% auto_monad_reorder : 0.000022s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% eliminate_special_op_node : 0.000514s : 0.39% validate : 0.000040s : 0.03% distribtued_split : 0.000001s : 0.00% task_emit : 0.003489s : 2.63% execute : 0.000006s : 0.00% Time group info: ------[substitution.] 0.018809 973 0.02% : 0.000004s : 6: substitution.float_depend_g_call 0.13% : 0.000024s : 49: substitution.float_tuple_getitem_switch 90.29% : 0.016982s : 59: substitution.getattr_setattr_resolve 0.03% : 0.000006s : 6: substitution.graph_param_transform 0.01% : 0.000003s : 3: substitution.incorporate_call 0.01% : 0.000002s : 3: substitution.incorporate_call_switch 5.58% : 0.001049s : 103: substitution.inline 0.05% : 0.000009s : 23: substitution.less_batch_normalization 0.20% : 0.000037s : 42: substitution.meta_unpack_prepare 0.18% : 0.000033s : 50: substitution.minmaximum_grad 0.02% : 0.000004s : 6: substitution.partial_eliminate 0.01% : 0.000001s : 6: substitution.partial_unused_args_eliminate 0.05% : 0.000009s : 81: substitution.remove_not_recompute_node 0.51% : 0.000096s : 69: substitution.replace_applicator 0.05% : 0.000010s : 36: substitution.replace_old_param 0.02% : 0.000004s : 2: substitution.reset_defer_inline 0.03% : 0.000005s : 8: substitution.set_cell_output_no_recompute 0.04% : 0.000007s : 5: substitution.specialize_transform 0.07% : 0.000013s : 10: substitution.switch_simplify 0.07% : 0.000013s : 4: substitution.transpose_eliminate 0.66% : 0.000123s : 70: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000057s : 70: substitution.tuple_list_get_item_const_eliminator 0.40% : 0.000076s : 70: substitution.tuple_list_get_item_depend_reorder 0.87% : 0.000165s : 122: substitution.tuple_list_get_item_eliminator 0.41% : 0.000077s : 70: substitution.tuple_list_get_set_item_eliminator ------[renormalize.] 0.091441 6 92.50% : 0.084581s : 3: renormalize.infer 7.50% : 0.006860s : 3: renormalize.specialize ------[replace.] 0.001251 141 54.67% : 0.000684s : 55: replace.getattr_setattr_resolve 25.65% : 0.000321s : 56: replace.inline 3.78% : 0.000047s : 2: replace.meta_unpack_prepare 7.52% : 0.000094s : 10: replace.switch_simplify 1.60% : 0.000020s : 4: replace.transpose_eliminate 6.77% : 0.000085s : 14: replace.tuple_list_get_item_eliminator ------[match.] 0.017790 141 94.70% : 0.016848s : 55: match.getattr_setattr_resolve 4.90% : 0.000871s : 56: match.inline 0.10% : 0.000017s : 2: match.meta_unpack_prepare 0.07% : 0.000013s : 10: match.switch_simplify 0.07% : 0.000013s : 4: match.transpose_eliminate 0.16% : 0.000028s : 14: match.tuple_list_get_item_eliminator ------[func_graph_cloner_run.] 0.007014 119 69.30% : 0.004861s : 53: func_graph_cloner_run.FuncGraphClonerGraph 30.70% : 0.002153s : 66: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.027857 589 0.51% : 0.000143s : 2: opt.transform.meta_unpack_prepare 30.28% : 0.008434s : 461: opt.transform.opt_a 0.08% : 0.000021s : 7: opt.transform.opt_after_cconv 3.39% : 0.000945s : 94: opt.transform.opt_b 65.45% : 0.018233s : 14: opt.transform.opt_resolve 0.23% : 0.000065s : 8: opt.transform.opt_trans_graph 0.05% : 0.000015s : 3: opt.transform.special_op_eliminate . ============================= 22 passed in 25.77s ============================== [TRACE] GE(32458,python3.7):2024-01-11-05:34:37.172.243 [status:INIT] [ge_api.cc:463]32458 ~Session:Start to destruct session. [TRACE] GE(32458,python3.7):2024-01-11-05:34:37.172.300 [status:RUNNING] [ge_api.cc:475]32458 ~Session:Session id is 0 [TRACE] GE(32458,python3.7):2024-01-11-05:34:37.172.311 [status:RUNNING] [ge_api.cc:476]32458 ~Session:Destroying session [TRACE] GE(32458,python3.7):2024-01-11-05:34:37.173.272 [status:STOP] [ge_api.cc:491]32458 ~Session:Session Destructor finished [TRACE] GE(32458,python3.7):2024-01-11-05:34:37.173.300 [status:INIT] [ge_api.cc:301]32458 GEFinalize:GEFinalize start [INFO] GE(32458,python3.7):2024-01-11-05:34:37.173.375 [execution_runtime.cc:80][EVENT]32458 FinalizeExecutionRuntime:Execution runtime finalize begin. [INFO] GE(32458,python3.7):2024-01-11-05:34:37.173.393 [execution_runtime.cc:92][EVENT]32458 FinalizeExecutionRuntime:Execution runtime finalized. [TRACE] GE(32458,python3.7):2024-01-11-05:34:37.173.404 [status:RUNNING] [ge_api.cc:313]32458 GEFinalize:Finalizing environment [INFO] TUNE(32458,python3.7):2024-01-11-05:34:37.469.197 [cann_kb_pyfunc_mgr.cpp:127][CANNKB][Tid:32458]"CannKbPyfuncMgr: enter PyObjectDeinit function, reference_[1]" [INFO] TUNE(32458,python3.7):2024-01-11-05:34:37.469.251 [cann_kb_pyfunc_mgr.cpp:138][CANNKB][Tid:32458]"CannKbPyfuncMgr: PyObjectDeinit function end successfully!" [INFO] GE(32458,python3.7):2024-01-11-05:34:37.470.609 [gelib.cc:324][EVENT]32458 SystemFinalize:Online infer finalize GELib success. [TRACE] GE(32458,python3.7):2024-01-11-05:34:38.473.244 [status:STOP] [ge_api.cc:341]32458 GEFinalize:GEFinalize finished [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.814.466 [process_mode_manager.cpp:184][Close][tid:32458] [TsdClient] Close [deviceId=5][sessionId=1] hccp and computer enter [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.814.500 [version_verify.cpp:112][SpecialFeatureCheck][tid:32458] VersionVerify: previous type[7], supported [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.814.545 [process_mode_manager.cpp:192][Close][tid:32458] [TsdClient][deviceId=5] [sessionId=1] wait hccp and computer process close respond [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.845.691 [process_mode_manager.cpp:197][Close][tid:32458] [TsdClient][logicDeviceId_=5]has recv close hccp and computer process respond [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.845.707 [stub_process_mode_nowin.cpp:151][CloseInHost][tid:32458] enter into CloseInHost deviceid[5] [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.845.717 [stub_process_mode_nowin.cpp:154][CloseInHost][tid:32458] host cpu not support [INFO] TDT(32458,python3.7):2024-01-11-05:34:38.845.753 [process_mode_manager.cpp:208][Close][tid:32458] [TsdClient][deviceId=5] [sessionId=1] close hccp and computer process success [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:38.845.766 [atrace_api.c:93](tid:32458) AtraceDestroy start [INFO] ATRACE(32458,python3.7):2024-01-11-05:34:38.845.783 [atrace_api.c:95](tid:32458) AtraceDestroy end [INFO] PROFILING(32458,python3.7):2024-01-11-05:34:38.845.808 [msprofiler_impl.cpp:156] >>> (tid:32458) ProfNotifySetDevice called, is open: 0, devId: 5 [INFO] RUNTIME(32458,python3.7):2024-01-11-05:34:40.400.190 [runtime.cc:1737] 32458 ~Runtime: deconstruct runtime.