==================================================Ascend ============================= test session starts ============================== platform linux -- Python 3.9.19, pytest-6.2.5, py-1.11.0, pluggy-1.5.0 rootdir: /home/jenkins/mindspore/testcases/testcases/tests/st/infer/ops/test_internal_ops, configfile: ../../../../../../../../sault/virtual_test/virtualenv_005/sault/config/pytest.ini plugins: mock-3.14.0, hydra-core-1.3.2, forked-1.6.0, anyio-4.9.0, xdist-1.32.0 collected 16 items test_matmul_split.py [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:37:39.340.360 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 2.86054, [21] [bootstrap]: 0.00077835 [type_inference]: 2.73058 [event_method]: 1.934e-05 [auto_monad]: 0.0416186 [graph_reusing]: 7.41001e-06 [inline]: 3.49001e-06 [add_attr]: 0.00796515, [1] [add_attr_with_inline]: 0.00794964, [1] [Cycle 1]: 9.842e-05, [2] [tag_attr]: 2.216e-05 [meta_addattr_fg_expand]: 4.89e-06 [parallel-infer-symbol]: 3.63e-06 [pre_auto_parallel]: 0.00013465 [insert-virtual-dataset]: 3.33998e-06 [parallel-infer-symbol-second]: 1.10999e-06 [dataset_repeat_opt]: 2.59999e-06 [pipeline_split]: 1.81e-06 [optimize]: 0.0783609, [53] [py_interpret_to_execute]: 2.818e-05 [rewriter_before_opt_a]: 8.793e-05 [opt_a]: 0.0104425, [2] [Cycle 1]: 0.00887805, [45] [expand_dump_flag]: 2.61e-06 [switch_simplify]: 3.601e-05 [loop_unroll]: 1.977e-05 [a_1]: 0.000641 [with_stream_mark]: 2.226e-05 [recompute_prepare]: 1.624e-05 [updatestate_depend_eliminate]: 8.87999e-06 [updatestate_assign_eliminate]: 8.52998e-06 [updatestate_loads_eliminate]: 1.154e-05 [parameter_eliminate]: 1.96998e-06 [a_2]: 0.00022757 [accelerated_algorithm]: 3.396e-05 [shard]: 2.16e-06 [meta_shard_fg_expand]: 3.53e-06 [shard_inline]: 1.552e-05 [merge_send_recv]: 3.042e-05 [auto_parallel]: 1.197e-05 [parallel]: 4.14e-05 [flash_sp]: 1.947e-05 [merge_comm]: 9.67001e-06 [allreduce_fusion]: 8.72e-06 [matmul_add_comm_reduction]: 1.638e-05 [allreduce_slice_to_reducescatter]: 9.20001e-07 [virtual_shard_identity]: 1.894e-05 [virtual_dataset]: 1.605e-05 [get_grad_eliminate_]: 1.758e-05 [virtual_output]: 1.559e-05 [merge_forward]: 8.79998e-06 [cell_reuse_recompute_pass]: 2.28998e-06 [offload_activation]: 1.833e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.991e-05 [merge_recompute_call_nodes]: 1.49e-06 [before_grad]: 2.706e-05 [set_forward_comm_id_for_comm_node_pass]: 8.80001e-06 [meta_fg_expand]: 6.72002e-06 [flash_sp_send_recv_attached]: 4.90999e-06 [receive_attached]: 1.055e-05 [after_resolve]: 2.196e-05 [a_after_grad]: 2.522e-05 [renormalize]: 0.00675379 [add_forward_monad_depend]: 9.11002e-06 [auto_monad_grad]: 3.40998e-06 [auto_monad_eliminator]: 5.306e-05 [cse]: 0.00022704 [a_3]: 0.0001205 [Cycle 2]: 0.00155075, [45] [expand_dump_flag]: 2.19999e-06 [switch_simplify]: 1.928e-05 [loop_unroll]: 1.584e-05 [a_1]: 0.00043219 [with_stream_mark]: 2.165e-05 [recompute_prepare]: 1.533e-05 [updatestate_depend_eliminate]: 9.20001e-06 [updatestate_assign_eliminate]: 7.93001e-06 [updatestate_loads_eliminate]: 9.32001e-06 [parameter_eliminate]: 2.06e-06 [a_2]: 0.00021296 [accelerated_algorithm]: 2.042e-05 [shard]: 1.97001e-06 [meta_shard_fg_expand]: 4e-06 [shard_inline]: 1.513e-05 [merge_send_recv]: 1.223e-05 [auto_parallel]: 1.206e-05 [parallel]: 8.98002e-06 [flash_sp]: 4.05e-06 [merge_comm]: 8.25e-06 [allreduce_fusion]: 8.42998e-06 [matmul_add_comm_reduction]: 1.5e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 1.717e-05 [virtual_dataset]: 1.535e-05 [get_grad_eliminate_]: 1.529e-05 [virtual_output]: 1.463e-05 [merge_forward]: 8.65999e-06 [cell_reuse_recompute_pass]: 2.91e-06 [offload_activation]: 1.743e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.728e-05 [merge_recompute_call_nodes]: 1.29e-06 [before_grad]: 2.486e-05 [set_forward_comm_id_for_comm_node_pass]: 8.59e-06 [meta_fg_expand]: 5.67999e-06 [flash_sp_send_recv_attached]: 1.90001e-06 [receive_attached]: 2.48998e-06 [after_resolve]: 2.137e-05 [a_after_grad]: 2.5e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.48002e-06 [auto_monad_grad]: 1.69e-06 [auto_monad_eliminator]: 3.193e-05 [cse]: 0.00013041 [a_3]: 0.00010306 [py_interpret_to_execute_after_opt_a]: 2.289e-05 [slice_cell_reuse_recomputed_activation]: 2.66999e-06 [rewriter_after_opt_a]: 0.00024924 [convert_after_rewriter]: 1.757e-05 [order_py_execute_after_rewriter]: 1.066e-05 [mutable_eliminate]: 0.0649137 [opt_b]: 0.00061135, [1] [Cycle 1]: 0.00060006, [7] [b_1]: 0.00038123 [b_2]: 2.256e-05 [updatestate_depend_eliminate]: 2.088e-05 [updatestate_assign_eliminate]: 9.71e-06 [updatestate_loads_eliminate]: 1.287e-05 [renormalize]: 1.30001e-06 [cse]: 9.988e-05 [optimize_parallel_all_gather_comm]: 4.034e-05 [overlap_param_gather]: 5.36002e-06 [cconv]: 4.472e-05 [loop_unroll]: 0.00057213 [opt_after_cconv]: 0.00024811, [1] [Cycle 1]: 0.00024091, [7] [c_1]: 0.00010586 [parameter_eliminate]: 6.97002e-06 [updatestate_depend_eliminate]: 1.449e-05 [updatestate_assign_eliminate]: 7.90998e-06 [updatestate_loads_eliminate]: 9.62999e-06 [cse]: 5.458e-05 [renormalize]: 7.2e-07 [remove_dup_value]: 7.115e-05 [tuple_transform]: 0.00016417, [1] [Cycle 1]: 0.00015785, [4] [d_1]: 0.00011521 [none_parameter_eliminate]: 2.19001e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 1.747e-05 [partial_unused_args_eliminate]: 2.31e-06 [add_recomputation]: 0.00011848 [cse_after_recomputation]: 5.37e-05, [1] [Cycle 1]: 4.835e-05, [1] [cse]: 4.048e-05 [environ_conv]: 3.221e-05 [swap_dp_allreduce_reducescatter]: 1.335e-05 [bias_add_comm_swap]: 3.04001e-06 [label_micro_interleaved_index]: 5.25999e-06 [label_fine_grained_interleaved_index]: 3.2e-06 [merge_cast_opt]: 1.47999e-06 [slice_recompute_activation]: 2.49999e-06 [micro_interleaved_order_control]: 2.95002e-06 [assign_add_opt]: 1.60999e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.05999e-06 [full_micro_interleaved_order_control]: 2.46e-06 [reorder_send_recv_between_fp_bp]: 3.13e-06 [comm_op_add_attrs]: 1.05001e-06 [add_comm_op_reuse_tag]: 1.01002e-06 [interleave_split_concat_branches]: 1.29998e-06 [interleave_parallel_branches]: 1.06002e-06 [overlap_opt_shard_in_pipeline]: 2.452e-05 [overlap_opt_shard_grad_in_pipeline]: 2.06998e-06 [control_data_broadcast_order]: 2.723e-05 [grouped_pairwise_exchange_alltoall]: 1.59e-06 [offloading_packed_experts]: 7.61001e-06 [overlap_recompute_and_grad_model_parallel]: 9.00001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.17e-06 [overlap_recompute_allgather_and_fa_grad]: 1.70001e-06 [overlap_recompute_comm]: 2.59001e-06 [overlap_grad_ring_attention]: 6.93998e-06 [overlap_grad_flash_sp]: 5.668e-05 [begin_end_overlap_inline]: 5.89993e-07 [split_matmul_comm_elemetwise]: 2.57001e-06 [split_layernorm_comm]: 1.87001e-06 [handle_group_info]: 1.35001e-06 [symbol_engine_optimizer]: 0.00015694, [1] [Cycle 1]: 0.00015095, [6] [build]: 1.651e-05 [elim_shapecalc]: 2.487e-05 [elim_not_effective]: 2.936e-05 [opt_reshape]: 2.184e-05 [fold_const_symbol]: 2.461e-05 [renormalize]: 4.59986e-07 [detach_backward]: 2.32999e-06 [pipeline_parallel_scheduler]: 2.25002e-06 [auto_monad_reorder]: 5.194e-05 [get_jit_bprop_graph]: 2.20002e-06 [rewriter_after_jit_bprop_graph]: 7.10998e-06 [opt_after_jit_grad]: 0.00064376 [validate]: 0.0001064 Sums bootstrap : 0.000778s : 0.03% type_inference : 2.730579s : 95.76% event_method : 0.000019s : 0.00% auto_monad : 0.041619s : 1.46% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000022s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000005s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000135s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000028s : 0.00% optimize.rewriter_before_opt_a : 0.000088s : 0.00% optimize.opt_a.expand_dump_flag : 0.000005s : 0.00% optimize.opt_a.switch_simplify : 0.000055s : 0.00% optimize.opt_a.loop_unroll : 0.000036s : 0.00% optimize.opt_a.a_1 : 0.001073s : 0.04% optimize.opt_a.with_stream_mark : 0.000044s : 0.00% optimize.opt_a.recompute_prepare : 0.000032s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000018s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000016s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000021s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000441s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000054s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.00% optimize.opt_a.shard_inline : 0.000031s : 0.00% optimize.opt_a.merge_send_recv : 0.000043s : 0.00% optimize.opt_a.auto_parallel : 0.000024s : 0.00% optimize.opt_a.parallel : 0.000050s : 0.00% optimize.opt_a.flash_sp : 0.000024s : 0.00% optimize.opt_a.merge_comm : 0.000018s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000031s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000036s : 0.00% optimize.opt_a.virtual_dataset : 0.000031s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000033s : 0.00% optimize.opt_a.virtual_output : 0.000030s : 0.00% optimize.opt_a.merge_forward : 0.000017s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000036s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000057s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000052s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000017s : 0.00% optimize.opt_a.meta_fg_expand : 0.000012s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.00% optimize.opt_a.receive_attached : 0.000013s : 0.00% optimize.opt_a.after_resolve : 0.000043s : 0.00% optimize.opt_a.a_after_grad : 0.000050s : 0.00% optimize.opt_a.renormalize : 0.006754s : 0.24% optimize.opt_a.add_forward_monad_depend : 0.000011s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000085s : 0.00% optimize.opt_a.cse : 0.000357s : 0.01% optimize.opt_a.a_3 : 0.000224s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000023s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000249s : 0.01% optimize.convert_after_rewriter : 0.000018s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.064914s : 2.28% optimize.opt_b.b_1 : 0.000381s : 0.01% optimize.opt_b.b_2 : 0.000023s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000021s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000100s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000040s : 0.00% optimize.overlap_param_gather : 0.000005s : 0.00% optimize.cconv : 0.000045s : 0.00% optimize.loop_unroll : 0.000572s : 0.02% optimize.opt_after_cconv.c_1 : 0.000106s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000014s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.cse : 0.000055s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000071s : 0.00% optimize.tuple_transform.d_1 : 0.000115s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000017s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000118s : 0.00% optimize.cse_after_recomputation.cse : 0.000040s : 0.00% optimize.environ_conv : 0.000032s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000013s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000025s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000027s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000008s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000009s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000007s : 0.00% optimize.overlap_grad_flash_sp : 0.000057s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000017s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000025s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000029s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000022s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000025s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000052s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000644s : 0.02% validate : 0.000106s : 0.00% Time group info: ------[substitution.] 0.000266 96 4.34% : 0.000012s : 2: substitution.depend_value_elim 1.58% : 0.000004s : 8: substitution.elim_not_effective 1.33% : 0.000004s : 8: substitution.fold_const_symbol 4.76% : 0.000013s : 13: substitution.graph_param_transform 37.96% : 0.000101s : 1: substitution.inline 3.46% : 0.000009s : 16: substitution.j_node_and_user_rematch 6.56% : 0.000017s : 2: substitution.less_batch_normalization 2.54% : 0.000007s : 6: substitution.load_eliminater 0.77% : 0.000002s : 2: substitution.opt_reshape 4.65% : 0.000012s : 16: substitution.remove_not_recompute_node 2.60% : 0.000007s : 4: substitution.replace_old_param 12.70% : 0.000034s : 4: substitution.reshape_eliminate 9.74% : 0.000026s : 6: substitution.updatestate_pure_node_eliminater 7.00% : 0.000019s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.730461 2 99.85% : 2.726409s : 1: type_inference.infer 0.15% : 0.004052s : 1: type_inference.specialize ------[replace.] 0.000018 1 100.00% : 0.000018s : 1: replace.inline ------[match.] 0.000100 1 100.00% : 0.000100s : 1: match.inline ------[predicate.] 0.000430 3002 0.80% : 0.000003s : 27: predicate.accumulaten_eliminater 0.85% : 0.000004s : 13: predicate.ad_related_special_op_eliminate 0.75% : 0.000003s : 26: predicate.addn_check_dump 0.76% : 0.000003s : 27: predicate.addn_zero_filter 0.71% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.00% : 0.000009s : 53: predicate.arithmetic_simplify 0.83% : 0.000004s : 27: predicate.cast_eliminate 0.80% : 0.000003s : 26: predicate.check_bprop_eliminate 0.77% : 0.000003s : 26: predicate.compare_switch_simplify 0.26% : 0.000001s : 13: predicate.const_output_eliminate 0.84% : 0.000004s : 26: predicate.depend_value_elim 0.79% : 0.000003s : 27: predicate.dict_get_item_const_eliminator 0.86% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.79% : 0.000003s : 27: predicate.dict_set_item_eliminator 2.31% : 0.000010s : 26: predicate.dumpgradient_eliminate 0.33% : 0.000001s : 13: predicate.elim_not_effective 0.54% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.19% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.08% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.08% : 0.000005s : 40: predicate.environ_get_depend_swap 1.94% : 0.000008s : 66: predicate.environ_get_eliminate 1.12% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.76% : 0.000003s : 28: predicate.exchange_switch_depend_value 1.34% : 0.000006s : 28: predicate.float_depend_g_call 0.78% : 0.000003s : 26: predicate.float_environ_get_switch 1.25% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.24% : 0.000001s : 13: predicate.fold_const_symbol 0.98% : 0.000004s : 26: predicate.get_grad_eliminate 0.37% : 0.000002s : 13: predicate.graph_param_transform 0.79% : 0.000003s : 26: predicate.incorporate_call 0.72% : 0.000003s : 26: predicate.incorporate_call_switch 5.41% : 0.000023s : 133: predicate.inline 1.04% : 0.000004s : 26: predicate.inline_without_move 0.46% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.13% : 0.000005s : 27: predicate.less_batch_normalization 1.65% : 0.000007s : 53: predicate.list_to_tuple_eliminator_ 2.30% : 0.000010s : 80: predicate.load_eliminater 0.95% : 0.000004s : 13: predicate.loop_unroll_after_grad 1.00% : 0.000004s : 31: predicate.loop_unroll_before_grad 1.86% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.79% : 0.000003s : 26: predicate.merge_addn 0.80% : 0.000003s : 26: predicate.micro_step_allgather_replace 0.79% : 0.000003s : 26: predicate.mini_step_allgather_replace 0.72% : 0.000003s : 27: predicate.minmaximum_grad 2.12% : 0.000009s : 13: predicate.mutable_eliminate 0.50% : 0.000002s : 13: predicate.opt_reshape 0.45% : 0.000002s : 13: predicate.parallel_virtual_node 1.09% : 0.000005s : 28: predicate.partial_defer_inline 1.32% : 0.000006s : 40: predicate.partial_eliminate 0.78% : 0.000003s : 27: predicate.print_const_string_wrapper 0.75% : 0.000003s : 26: predicate.reduce_all_const_elim 1.07% : 0.000005s : 27: predicate.reduce_eliminate 2.43% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.53% : 0.000002s : 26: predicate.remove_not_recompute_node 1.37% : 0.000006s : 53: predicate.replace_applicator 0.51% : 0.000002s : 26: predicate.replace_old_param 0.43% : 0.000002s : 13: predicate.reset_defer_inline 0.91% : 0.000004s : 27: predicate.reshape_eliminate 0.80% : 0.000003s : 26: predicate.row_tensor_add_zeros_like 0.47% : 0.000002s : 13: predicate.row_tensor_eliminate 1.18% : 0.000005s : 26: predicate.same_eliminate 0.55% : 0.000002s : 26: predicate.set_cell_output_no_recompute 0.90% : 0.000004s : 26: predicate.shard_identity_eliminate 0.90% : 0.000004s : 26: predicate.special_op_eliminate 0.87% : 0.000004s : 26: predicate.specialize_transform 1.06% : 0.000005s : 26: predicate.split_environ_get_set_with_tuple_value 1.03% : 0.000004s : 26: predicate.stack_unstack_eliminate 0.48% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.88% : 0.000004s : 28: predicate.switch_defer_inline 1.66% : 0.000007s : 54: predicate.switch_layer_defer_inline 3.36% : 0.000014s : 98: predicate.switch_simplify 0.76% : 0.000003s : 27: predicate.tile_eliminate 0.77% : 0.000003s : 27: predicate.transpose_eliminate 1.59% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.68% : 0.000007s : 53: predicate.tuple_list_get_item_const_eliminator 1.52% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 3.03% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.61% : 0.000007s : 53: predicate.tuple_list_get_set_item_eliminator 2.54% : 0.000011s : 79: predicate.tuple_list_set_item_eliminator 1.70% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.33% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.41% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.46% : 0.000002s : 13: predicate.value_based_eliminate 0.97% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.86% : 0.000004s : 26: predicate.virtual_output_eliminate 0.40% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.50% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.009006 58 82.51% : 0.007431s : 55: func_graph_cloner_run.FuncGraphClonerGraph 17.49% : 0.001575s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.956393 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.27% : 0.007971s : 1: add_attr 0.27% : 0.007954s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000125s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 1.41% : 0.041652s : 1: auto_monad 0.00% : 0.000057s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.03% : 0.000819s : 1: bootstrap 0.00% : 0.000050s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000031s : 1: control_data_broadcast_order 0.00% : 0.000022s : 1: convert_after_rewriter 0.00% : 0.000057s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000036s : 1: environ_conv 0.00% : 0.000028s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000014s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000008s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.02% : 0.000583s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 2.20% : 0.064932s : 1: mutable_eliminate 0.00% : 0.000011s : 1: offloading_packed_experts 0.00% : 0.000032s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000053s : 1: opt.transform.mutable_eliminate 0.07% : 0.002176s : 78: opt.transform.opt_a 0.00% : 0.000105s : 1: opt.transform.opt_after_cconv 0.00% : 0.000055s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000369s : 28: opt.transform.opt_b 0.00% : 0.000130s : 2: opt.transform.opt_trans_graph 0.00% : 0.000096s : 4: opt.transform.symbol_engine_opt 0.35% : 0.010447s : 1: opt_a 0.01% : 0.000252s : 1: opt_after_cconv 0.02% : 0.000653s : 1: opt_after_jit_grad 0.02% : 0.000617s : 1: opt_b 2.65% : 0.078367s : 1: optimize 0.00% : 0.000046s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000014s : 1: order_py_execute_after_rewriter 0.00% : 0.000062s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000010s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000028s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000009s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000012s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.00% : 0.000144s : 1: pre_auto_parallel 0.00% : 0.000033s : 1: py_interpret_to_execute 0.00% : 0.000027s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000077s : 1: remove_dup_value 0.17% : 0.004989s : 1: renormalize.infer 0.06% : 0.001752s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000256s : 1: rewriter_after_opt_a 0.00% : 0.000092s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000017s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000160s : 1: symbol_engine_optimizer 0.01% : 0.000167s : 1: tuple_transform 92.36% : 2.730614s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape0-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape0-1],max_mem:120.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:38:19.405.96 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 5.97267, [21] [bootstrap]: 0.119098 [type_inference]: 5.49181 [event_method]: 2.063e-05 [auto_monad]: 0.00042468 [graph_reusing]: 7.25e-06 [inline]: 3.06001e-06 [add_attr]: 0.172293, [1] [add_attr_with_inline]: 0.172275, [1] [Cycle 1]: 7.481e-05, [2] [tag_attr]: 2.349e-05 [meta_addattr_fg_expand]: 4.74e-06 [parallel-infer-symbol]: 5.01002e-06 [pre_auto_parallel]: 3.721e-05 [insert-virtual-dataset]: 2.62001e-06 [parallel-infer-symbol-second]: 7.49977e-07 [dataset_repeat_opt]: 2.09e-06 [pipeline_split]: 1.50999e-06 [optimize]: 0.186919, [53] [py_interpret_to_execute]: 3.351e-05 [rewriter_before_opt_a]: 9.661e-05 [opt_a]: 0.183014, [2] [Cycle 1]: 0.181216, [45] [expand_dump_flag]: 3.04001e-06 [switch_simplify]: 3.586e-05 [loop_unroll]: 1.991e-05 [a_1]: 0.00081706 [with_stream_mark]: 2.64e-05 [recompute_prepare]: 2.232e-05 [updatestate_depend_eliminate]: 1.237e-05 [updatestate_assign_eliminate]: 8.19998e-06 [updatestate_loads_eliminate]: 1.078e-05 [parameter_eliminate]: 2.19999e-06 [a_2]: 0.00024838 [accelerated_algorithm]: 4.522e-05 [shard]: 3.14001e-06 [meta_shard_fg_expand]: 3.95998e-06 [shard_inline]: 1.645e-05 [merge_send_recv]: 1.37e-05 [auto_parallel]: 1.277e-05 [parallel]: 3.806e-05 [flash_sp]: 1.232e-05 [merge_comm]: 1.313e-05 [allreduce_fusion]: 8.39002e-06 [matmul_add_comm_reduction]: 1.768e-05 [allreduce_slice_to_reducescatter]: 8.50006e-07 [virtual_shard_identity]: 3.033e-05 [virtual_dataset]: 3.318e-05 [get_grad_eliminate_]: 3.688e-05 [virtual_output]: 3.965e-05 [merge_forward]: 9.44e-06 [cell_reuse_recompute_pass]: 1.87001e-06 [offload_activation]: 2.32e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.941e-05 [merge_recompute_call_nodes]: 1.94999e-06 [before_grad]: 3.067e-05 [set_forward_comm_id_for_comm_node_pass]: 9.72999e-06 [meta_fg_expand]: 5.98998e-06 [flash_sp_send_recv_attached]: 6.24001e-06 [receive_attached]: 2.71e-06 [after_resolve]: 2.565e-05 [a_after_grad]: 3.045e-05 [renormalize]: 0.0116447 [add_forward_monad_depend]: 9.71e-06 [auto_monad_grad]: 2.53998e-06 [auto_monad_eliminator]: 5.843e-05 [cse]: 0.167266 [a_3]: 0.00016108 [Cycle 2]: 0.00178211, [45] [expand_dump_flag]: 4.98001e-06 [switch_simplify]: 1.91e-05 [loop_unroll]: 1.68e-05 [a_1]: 0.00047749 [with_stream_mark]: 3.986e-05 [recompute_prepare]: 1.739e-05 [updatestate_depend_eliminate]: 1.147e-05 [updatestate_assign_eliminate]: 8.07e-06 [updatestate_loads_eliminate]: 1.085e-05 [parameter_eliminate]: 3.43999e-06 [a_2]: 0.00022298 [accelerated_algorithm]: 2.28e-05 [shard]: 2.93998e-06 [meta_shard_fg_expand]: 7.05e-06 [shard_inline]: 1.555e-05 [merge_send_recv]: 1.671e-05 [auto_parallel]: 1.649e-05 [parallel]: 1.101e-05 [flash_sp]: 4.50999e-06 [merge_comm]: 8.38999e-06 [allreduce_fusion]: 8.87e-06 [matmul_add_comm_reduction]: 1.937e-05 [allreduce_slice_to_reducescatter]: 1.22e-06 [virtual_shard_identity]: 1.729e-05 [virtual_dataset]: 1.567e-05 [get_grad_eliminate_]: 1.559e-05 [virtual_output]: 1.564e-05 [merge_forward]: 9.40001e-06 [cell_reuse_recompute_pass]: 3.54002e-06 [offload_activation]: 1.833e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.015e-05 [merge_recompute_call_nodes]: 1.60999e-06 [before_grad]: 2.608e-05 [set_forward_comm_id_for_comm_node_pass]: 9.09998e-06 [meta_fg_expand]: 7.32997e-06 [flash_sp_send_recv_attached]: 2.01998e-06 [receive_attached]: 2.68e-06 [after_resolve]: 2.235e-05 [a_after_grad]: 2.707e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 3.97002e-06 [auto_monad_grad]: 3.91999e-06 [auto_monad_eliminator]: 6.244e-05 [cse]: 0.00017969 [a_3]: 0.00010689 [py_interpret_to_execute_after_opt_a]: 2.756e-05 [slice_cell_reuse_recomputed_activation]: 2.56e-06 [rewriter_after_opt_a]: 0.00020756 [convert_after_rewriter]: 1.742e-05 [order_py_execute_after_rewriter]: 1.224e-05 [mutable_eliminate]: 0.00083924 [opt_b]: 0.00057611, [1] [Cycle 1]: 0.00056729, [7] [b_1]: 0.00038781 [b_2]: 2.173e-05 [updatestate_depend_eliminate]: 1.693e-05 [updatestate_assign_eliminate]: 8.56002e-06 [updatestate_loads_eliminate]: 1.119e-05 [renormalize]: 5.39992e-07 [cse]: 7.262e-05 [optimize_parallel_all_gather_comm]: 3.702e-05 [overlap_param_gather]: 3.7e-06 [cconv]: 4.624e-05 [loop_unroll]: 0.00061225 [opt_after_cconv]: 0.00026694, [1] [Cycle 1]: 0.00025764, [7] [c_1]: 0.00011962 [parameter_eliminate]: 4.99998e-06 [updatestate_depend_eliminate]: 1.493e-05 [updatestate_assign_eliminate]: 8.76002e-06 [updatestate_loads_eliminate]: 9.72999e-06 [cse]: 5.465e-05 [renormalize]: 6.09987e-07 [remove_dup_value]: 8.396e-05 [tuple_transform]: 0.00018536, [1] [Cycle 1]: 0.0001789, [4] [d_1]: 0.0001325 [none_parameter_eliminate]: 2.46998e-06 [renormalize]: 3.00002e-07 [switch_simplify]: 1.99e-05 [partial_unused_args_eliminate]: 2.52001e-06 [add_recomputation]: 0.00013147 [cse_after_recomputation]: 5.875e-05, [1] [Cycle 1]: 5.315e-05, [1] [cse]: 4.57e-05 [environ_conv]: 2.14e-05 [swap_dp_allreduce_reducescatter]: 1.3e-05 [bias_add_comm_swap]: 3.75003e-06 [label_micro_interleaved_index]: 6.46e-06 [label_fine_grained_interleaved_index]: 3.22002e-06 [merge_cast_opt]: 1.71002e-06 [slice_recompute_activation]: 2.29999e-06 [micro_interleaved_order_control]: 2.66999e-06 [assign_add_opt]: 1.92999e-06 [ForceFp32Comm]: 1.15001e-06 [remove_cast_before_assign_add]: 1.21002e-06 [full_micro_interleaved_order_control]: 2.50002e-06 [reorder_send_recv_between_fp_bp]: 3.03998e-06 [comm_op_add_attrs]: 1.14998e-06 [add_comm_op_reuse_tag]: 1.45999e-06 [interleave_split_concat_branches]: 1.41002e-06 [interleave_parallel_branches]: 1.17e-06 [overlap_opt_shard_in_pipeline]: 4.02e-06 [overlap_opt_shard_grad_in_pipeline]: 1.84e-06 [control_data_broadcast_order]: 2.901e-05 [grouped_pairwise_exchange_alltoall]: 1.78002e-06 [offloading_packed_experts]: 8.13001e-06 [overlap_recompute_and_grad_model_parallel]: 8.42998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.39e-06 [overlap_recompute_allgather_and_fa_grad]: 1.71e-06 [overlap_recompute_comm]: 2.42001e-06 [overlap_grad_ring_attention]: 9.04998e-06 [overlap_grad_flash_sp]: 4.313e-05 [begin_end_overlap_inline]: 7.7e-07 [split_matmul_comm_elemetwise]: 2.91e-06 [split_layernorm_comm]: 1.85001e-06 [handle_group_info]: 1.32e-06 [symbol_engine_optimizer]: 0.00017569, [1] [Cycle 1]: 0.00016887, [6] [build]: 1.867e-05 [elim_shapecalc]: 2.482e-05 [elim_not_effective]: 3.334e-05 [opt_reshape]: 2.493e-05 [fold_const_symbol]: 2.914e-05 [renormalize]: 3.10014e-07 [detach_backward]: 2.39999e-06 [pipeline_parallel_scheduler]: 1.61998e-06 [auto_monad_reorder]: 4.858e-05 [get_jit_bprop_graph]: 2.17999e-06 [rewriter_after_jit_bprop_graph]: 7.37997e-06 [opt_after_jit_grad]: 0.00063361 [validate]: 0.00010453 Sums bootstrap : 0.119098s : 2.05% type_inference : 5.491809s : 94.72% event_method : 0.000021s : 0.00% auto_monad : 0.000425s : 0.01% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000023s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000005s : 0.00% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000037s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000034s : 0.00% optimize.rewriter_before_opt_a : 0.000097s : 0.00% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000055s : 0.00% optimize.opt_a.loop_unroll : 0.000037s : 0.00% optimize.opt_a.a_1 : 0.001295s : 0.02% optimize.opt_a.with_stream_mark : 0.000066s : 0.00% optimize.opt_a.recompute_prepare : 0.000040s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000024s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000016s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000022s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000471s : 0.01% optimize.opt_a.accelerated_algorithm : 0.000068s : 0.00% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000011s : 0.00% optimize.opt_a.shard_inline : 0.000032s : 0.00% optimize.opt_a.merge_send_recv : 0.000030s : 0.00% optimize.opt_a.auto_parallel : 0.000029s : 0.00% optimize.opt_a.parallel : 0.000049s : 0.00% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000022s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000037s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000048s : 0.00% optimize.opt_a.virtual_dataset : 0.000049s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000052s : 0.00% optimize.opt_a.virtual_output : 0.000055s : 0.00% optimize.opt_a.merge_forward : 0.000019s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000042s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000070s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000057s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000019s : 0.00% optimize.opt_a.meta_fg_expand : 0.000013s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000048s : 0.00% optimize.opt_a.a_after_grad : 0.000058s : 0.00% optimize.opt_a.renormalize : 0.011645s : 0.20% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000121s : 0.00% optimize.opt_a.cse : 0.167445s : 2.89% optimize.opt_a.a_3 : 0.000268s : 0.00% optimize.py_interpret_to_execute_after_opt_a : 0.000028s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000208s : 0.00% optimize.convert_after_rewriter : 0.000017s : 0.00% optimize.order_py_execute_after_rewriter : 0.000012s : 0.00% optimize.mutable_eliminate : 0.000839s : 0.01% optimize.opt_b.b_1 : 0.000388s : 0.01% optimize.opt_b.b_2 : 0.000022s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000017s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000073s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000037s : 0.00% optimize.overlap_param_gather : 0.000004s : 0.00% optimize.cconv : 0.000046s : 0.00% optimize.loop_unroll : 0.000612s : 0.01% optimize.opt_after_cconv.c_1 : 0.000120s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.cse : 0.000055s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000084s : 0.00% optimize.tuple_transform.d_1 : 0.000133s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000020s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000131s : 0.00% optimize.cse_after_recomputation.cse : 0.000046s : 0.00% optimize.environ_conv : 0.000021s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000013s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000006s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000029s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000008s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000043s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000019s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000025s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000033s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000025s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000029s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000049s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000634s : 0.01% validate : 0.000105s : 0.00% Time group info: ------[substitution.] 0.000303 96 3.71% : 0.000011s : 2: substitution.depend_value_elim 1.30% : 0.000004s : 8: substitution.elim_not_effective 1.20% : 0.000004s : 8: substitution.fold_const_symbol 4.69% : 0.000014s : 13: substitution.graph_param_transform 42.60% : 0.000129s : 1: substitution.inline 3.11% : 0.000009s : 16: substitution.j_node_and_user_rematch 6.98% : 0.000021s : 2: substitution.less_batch_normalization 2.09% : 0.000006s : 6: substitution.load_eliminater 0.91% : 0.000003s : 2: substitution.opt_reshape 3.98% : 0.000012s : 16: substitution.remove_not_recompute_node 2.74% : 0.000008s : 4: substitution.replace_old_param 10.71% : 0.000032s : 4: substitution.reshape_eliminate 7.92% : 0.000024s : 6: substitution.updatestate_pure_node_eliminater 8.08% : 0.000024s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 5.491668 2 99.89% : 5.485491s : 1: type_inference.infer 0.11% : 0.006177s : 1: type_inference.specialize ------[replace.] 0.000020 1 100.00% : 0.000020s : 1: replace.inline ------[match.] 0.000128 1 100.00% : 0.000128s : 1: match.inline ------[predicate.] 0.000460 3002 0.81% : 0.000004s : 27: predicate.accumulaten_eliminater 1.01% : 0.000005s : 13: predicate.ad_related_special_op_eliminate 0.70% : 0.000003s : 26: predicate.addn_check_dump 0.76% : 0.000004s : 27: predicate.addn_zero_filter 0.68% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.03% : 0.000009s : 53: predicate.arithmetic_simplify 0.85% : 0.000004s : 27: predicate.cast_eliminate 0.79% : 0.000004s : 26: predicate.check_bprop_eliminate 0.71% : 0.000003s : 26: predicate.compare_switch_simplify 0.26% : 0.000001s : 13: predicate.const_output_eliminate 0.82% : 0.000004s : 26: predicate.depend_value_elim 0.83% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.88% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.80% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.90% : 0.000009s : 26: predicate.dumpgradient_eliminate 0.27% : 0.000001s : 13: predicate.elim_not_effective 0.54% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.21% : 0.000006s : 40: predicate.environ_add_const_eliminate 1.10% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.12% : 0.000005s : 40: predicate.environ_get_depend_swap 1.95% : 0.000009s : 66: predicate.environ_get_eliminate 1.09% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.75% : 0.000003s : 28: predicate.exchange_switch_depend_value 1.30% : 0.000006s : 28: predicate.float_depend_g_call 0.74% : 0.000003s : 26: predicate.float_environ_get_switch 1.08% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.24% : 0.000001s : 13: predicate.fold_const_symbol 0.88% : 0.000004s : 26: predicate.get_grad_eliminate 0.31% : 0.000001s : 13: predicate.graph_param_transform 0.78% : 0.000004s : 26: predicate.incorporate_call 0.68% : 0.000003s : 26: predicate.incorporate_call_switch 5.35% : 0.000025s : 133: predicate.inline 1.19% : 0.000005s : 26: predicate.inline_without_move 0.42% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.14% : 0.000005s : 27: predicate.less_batch_normalization 1.81% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.27% : 0.000010s : 80: predicate.load_eliminater 1.18% : 0.000005s : 13: predicate.loop_unroll_after_grad 1.07% : 0.000005s : 31: predicate.loop_unroll_before_grad 1.75% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.82% : 0.000004s : 26: predicate.merge_addn 0.78% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.82% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.71% : 0.000003s : 27: predicate.minmaximum_grad 1.27% : 0.000006s : 13: predicate.mutable_eliminate 0.58% : 0.000003s : 13: predicate.opt_reshape 0.53% : 0.000002s : 13: predicate.parallel_virtual_node 1.21% : 0.000006s : 28: predicate.partial_defer_inline 1.29% : 0.000006s : 40: predicate.partial_eliminate 0.76% : 0.000003s : 27: predicate.print_const_string_wrapper 0.88% : 0.000004s : 26: predicate.reduce_all_const_elim 1.16% : 0.000005s : 27: predicate.reduce_eliminate 2.35% : 0.000011s : 80: predicate.redundant_stop_gradient_eliminater 0.48% : 0.000002s : 26: predicate.remove_not_recompute_node 1.43% : 0.000007s : 53: predicate.replace_applicator 0.54% : 0.000002s : 26: predicate.replace_old_param 0.34% : 0.000002s : 13: predicate.reset_defer_inline 0.94% : 0.000004s : 27: predicate.reshape_eliminate 0.79% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.60% : 0.000003s : 13: predicate.row_tensor_eliminate 1.36% : 0.000006s : 26: predicate.same_eliminate 0.56% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.95% : 0.000004s : 26: predicate.shard_identity_eliminate 1.02% : 0.000005s : 26: predicate.special_op_eliminate 0.86% : 0.000004s : 26: predicate.specialize_transform 1.13% : 0.000005s : 26: predicate.split_environ_get_set_with_tuple_value 0.92% : 0.000004s : 26: predicate.stack_unstack_eliminate 0.49% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.92% : 0.000004s : 28: predicate.switch_defer_inline 1.58% : 0.000007s : 54: predicate.switch_layer_defer_inline 3.29% : 0.000015s : 98: predicate.switch_simplify 0.88% : 0.000004s : 27: predicate.tile_eliminate 0.75% : 0.000003s : 27: predicate.transpose_eliminate 1.61% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.68% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.65% : 0.000008s : 53: predicate.tuple_list_get_item_depend_reorder 2.82% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.85% : 0.000009s : 53: predicate.tuple_list_get_set_item_eliminator 2.64% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.60% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.39% : 0.000011s : 80: predicate.updatestate_pure_node_eliminater 3.54% : 0.000016s : 106: predicate.updatestate_useless_node_eliminater 0.50% : 0.000002s : 13: predicate.value_based_eliminate 0.84% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.82% : 0.000004s : 26: predicate.virtual_output_eliminate 0.50% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.47% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.017707 58 77.53% : 0.013728s : 55: func_graph_cloner_run.FuncGraphClonerGraph 22.47% : 0.003979s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 6.346761 192 0.00% : 0.000004s : 1: ForceFp32Comm 2.71% : 0.172300s : 1: add_attr 2.71% : 0.172280s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000137s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.01% : 0.000440s : 1: auto_monad 0.00% : 0.000054s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 1.89% : 0.120150s : 1: bootstrap 0.00% : 0.000050s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000033s : 1: control_data_broadcast_order 0.00% : 0.000022s : 1: convert_after_rewriter 0.00% : 0.000062s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000027s : 1: environ_conv 0.00% : 0.000031s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000012s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000005s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000007s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.01% : 0.000626s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.01% : 0.000860s : 1: mutable_eliminate 0.00% : 0.000012s : 1: offloading_packed_experts 0.00% : 0.000035s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000073s : 1: opt.transform.mutable_eliminate 0.04% : 0.002570s : 78: opt.transform.opt_a 0.00% : 0.000118s : 1: opt.transform.opt_after_cconv 0.00% : 0.000065s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000371s : 28: opt.transform.opt_b 0.00% : 0.000149s : 2: opt.transform.opt_trans_graph 0.00% : 0.000108s : 4: opt.transform.symbol_engine_opt 2.88% : 0.183018s : 1: opt_a 0.00% : 0.000271s : 1: opt_after_cconv 0.01% : 0.000646s : 1: opt_after_jit_grad 0.01% : 0.000581s : 1: opt_b 2.95% : 0.186927s : 1: optimize 0.00% : 0.000042s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000017s : 1: order_py_execute_after_rewriter 0.00% : 0.000047s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000014s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000007s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000006s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000012s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.00% : 0.000042s : 1: pre_auto_parallel 0.00% : 0.000038s : 1: py_interpret_to_execute 0.00% : 0.000032s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000090s : 1: remove_dup_value 0.12% : 0.007325s : 1: renormalize.infer 0.07% : 0.004304s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000216s : 1: rewriter_after_opt_a 0.00% : 0.000102s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000006s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000017s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000179s : 1: symbol_engine_optimizer 0.00% : 0.000189s : 1: tuple_transform 86.53% : 5.491850s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape0-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape0-16],max_mem:120.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:38:49.137.846 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 3.2854, [21] [bootstrap]: 0.00073926 [type_inference]: 2.92834 [event_method]: 1.641e-05 [auto_monad]: 0.00029917 [graph_reusing]: 7.11001e-06 [inline]: 2.68998e-06 [add_attr]: 0.15904, [1] [add_attr_with_inline]: 0.159024, [1] [Cycle 1]: 7.1e-05, [2] [tag_attr]: 2.38e-05 [meta_addattr_fg_expand]: 4.39002e-06 [parallel-infer-symbol]: 4.85001e-06 [pre_auto_parallel]: 0.00010137 [insert-virtual-dataset]: 3.08998e-06 [parallel-infer-symbol-second]: 1.20999e-06 [dataset_repeat_opt]: 2.43e-06 [pipeline_split]: 1.72001e-06 [optimize]: 0.195492, [53] [py_interpret_to_execute]: 3.728e-05 [rewriter_before_opt_a]: 8.723e-05 [opt_a]: 0.191351, [2] [Cycle 1]: 0.189523, [45] [expand_dump_flag]: 3.4e-06 [switch_simplify]: 3.523e-05 [loop_unroll]: 1.967e-05 [a_1]: 0.00084086 [with_stream_mark]: 2.826e-05 [recompute_prepare]: 2.216e-05 [updatestate_depend_eliminate]: 1.186e-05 [updatestate_assign_eliminate]: 8.42e-06 [updatestate_loads_eliminate]: 1.104e-05 [parameter_eliminate]: 2.41e-06 [a_2]: 0.0002458 [accelerated_algorithm]: 4.404e-05 [shard]: 2.15002e-06 [meta_shard_fg_expand]: 3.52002e-06 [shard_inline]: 1.741e-05 [merge_send_recv]: 1.464e-05 [auto_parallel]: 1.211e-05 [parallel]: 4.034e-05 [flash_sp]: 1.279e-05 [merge_comm]: 9.00001e-06 [allreduce_fusion]: 8.28999e-06 [matmul_add_comm_reduction]: 1.669e-05 [allreduce_slice_to_reducescatter]: 6.00005e-07 [virtual_shard_identity]: 2.039e-05 [virtual_dataset]: 1.795e-05 [get_grad_eliminate_]: 1.931e-05 [virtual_output]: 3.714e-05 [merge_forward]: 9.10999e-06 [cell_reuse_recompute_pass]: 1.89e-06 [offload_activation]: 1.908e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.976e-05 [merge_recompute_call_nodes]: 1.71002e-06 [before_grad]: 6.556e-05 [set_forward_comm_id_for_comm_node_pass]: 1.011e-05 [meta_fg_expand]: 7.3e-06 [flash_sp_send_recv_attached]: 7.61999e-06 [receive_attached]: 2.66e-06 [after_resolve]: 4.857e-05 [a_after_grad]: 0.00010002 [renormalize]: 0.185725 [add_forward_monad_depend]: 1.433e-05 [auto_monad_grad]: 3.25e-06 [auto_monad_eliminator]: 6.621e-05 [cse]: 0.00135026 [a_3]: 0.00015171 [Cycle 2]: 0.0018102, [45] [expand_dump_flag]: 4.72e-06 [switch_simplify]: 2.08e-05 [loop_unroll]: 1.722e-05 [a_1]: 0.00049354 [with_stream_mark]: 3.337e-05 [recompute_prepare]: 1.933e-05 [updatestate_depend_eliminate]: 1.086e-05 [updatestate_assign_eliminate]: 7.98001e-06 [updatestate_loads_eliminate]: 1.067e-05 [parameter_eliminate]: 2.53e-06 [a_2]: 0.00022415 [accelerated_algorithm]: 2.296e-05 [shard]: 2.98e-06 [meta_shard_fg_expand]: 7.5e-06 [shard_inline]: 1.581e-05 [merge_send_recv]: 1.672e-05 [auto_parallel]: 1.543e-05 [parallel]: 1.123e-05 [flash_sp]: 4.49998e-06 [merge_comm]: 8.89e-06 [allreduce_fusion]: 9.34e-06 [matmul_add_comm_reduction]: 2.02e-05 [allreduce_slice_to_reducescatter]: 9.00007e-07 [virtual_shard_identity]: 1.872e-05 [virtual_dataset]: 1.598e-05 [get_grad_eliminate_]: 1.537e-05 [virtual_output]: 1.572e-05 [merge_forward]: 9.42999e-06 [cell_reuse_recompute_pass]: 4.07003e-06 [offload_activation]: 1.835e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.997e-05 [merge_recompute_call_nodes]: 2.15002e-06 [before_grad]: 2.612e-05 [set_forward_comm_id_for_comm_node_pass]: 8.95999e-06 [meta_fg_expand]: 8.18999e-06 [flash_sp_send_recv_attached]: 2.57001e-06 [receive_attached]: 3.16001e-06 [after_resolve]: 2.372e-05 [a_after_grad]: 2.759e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 3.78001e-06 [auto_monad_grad]: 2.65997e-06 [auto_monad_eliminator]: 6.411e-05 [cse]: 0.00015108 [a_3]: 0.00011101 [py_interpret_to_execute_after_opt_a]: 3.245e-05 [slice_cell_reuse_recomputed_activation]: 2.94999e-06 [rewriter_after_opt_a]: 0.00025299 [convert_after_rewriter]: 1.522e-05 [order_py_execute_after_rewriter]: 1.238e-05 [mutable_eliminate]: 0.00089023 [opt_b]: 0.00059866, [1] [Cycle 1]: 0.00058702, [7] [b_1]: 0.00038843 [b_2]: 1.874e-05 [updatestate_depend_eliminate]: 1.865e-05 [updatestate_assign_eliminate]: 8.69e-06 [updatestate_loads_eliminate]: 1.301e-05 [renormalize]: 6.80011e-07 [cse]: 9.126e-05 [optimize_parallel_all_gather_comm]: 3.588e-05 [overlap_param_gather]: 2.11998e-06 [cconv]: 4.694e-05 [loop_unroll]: 0.00074914 [opt_after_cconv]: 0.00027895, [1] [Cycle 1]: 0.00026759, [7] [c_1]: 0.00011719 [parameter_eliminate]: 7.64002e-06 [updatestate_depend_eliminate]: 1.855e-05 [updatestate_assign_eliminate]: 7.64002e-06 [updatestate_loads_eliminate]: 1.084e-05 [cse]: 6.531e-05 [renormalize]: 4.09986e-07 [remove_dup_value]: 8.021e-05 [tuple_transform]: 0.0001765, [1] [Cycle 1]: 0.00017096, [4] [d_1]: 0.0001245 [none_parameter_eliminate]: 2.81e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 1.965e-05 [partial_unused_args_eliminate]: 2.16e-06 [add_recomputation]: 0.00013229 [cse_after_recomputation]: 6.551e-05, [1] [Cycle 1]: 6.05e-05, [1] [cse]: 5.281e-05 [environ_conv]: 2.265e-05 [swap_dp_allreduce_reducescatter]: 1.305e-05 [bias_add_comm_swap]: 3.73001e-06 [label_micro_interleaved_index]: 7.56999e-06 [label_fine_grained_interleaved_index]: 2.78998e-06 [merge_cast_opt]: 1.58997e-06 [slice_recompute_activation]: 2.17999e-06 [micro_interleaved_order_control]: 2.61e-06 [assign_add_opt]: 1.91998e-06 [ForceFp32Comm]: 8.89995e-07 [remove_cast_before_assign_add]: 1.09e-06 [full_micro_interleaved_order_control]: 2.22999e-06 [reorder_send_recv_between_fp_bp]: 2.76e-06 [comm_op_add_attrs]: 1.10999e-06 [add_comm_op_reuse_tag]: 1.08001e-06 [interleave_split_concat_branches]: 1.20001e-06 [interleave_parallel_branches]: 1.06997e-06 [overlap_opt_shard_in_pipeline]: 1.37999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.89e-06 [control_data_broadcast_order]: 2.844e-05 [grouped_pairwise_exchange_alltoall]: 1.68002e-06 [offloading_packed_experts]: 6.98998e-06 [overlap_recompute_and_grad_model_parallel]: 7.80998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.17999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.59e-06 [overlap_recompute_comm]: 2.68e-06 [overlap_grad_ring_attention]: 8.40001e-06 [overlap_grad_flash_sp]: 4.022e-05 [begin_end_overlap_inline]: 6.00005e-07 [split_matmul_comm_elemetwise]: 2.67001e-06 [split_layernorm_comm]: 1.87001e-06 [handle_group_info]: 1.05999e-06 [symbol_engine_optimizer]: 0.00016706, [1] [Cycle 1]: 0.00015974, [6] [build]: 2.062e-05 [elim_shapecalc]: 2.461e-05 [elim_not_effective]: 3.067e-05 [opt_reshape]: 2.453e-05 [fold_const_symbol]: 2.469e-05 [renormalize]: 2.00002e-07 [detach_backward]: 2.31e-06 [pipeline_parallel_scheduler]: 1.47001e-06 [auto_monad_reorder]: 5.03e-05 [get_jit_bprop_graph]: 2.83e-06 [rewriter_after_jit_bprop_graph]: 9.28002e-06 [opt_after_jit_grad]: 0.00093462 [validate]: 0.00011072 Sums bootstrap : 0.000739s : 0.02% type_inference : 2.928335s : 93.71% event_method : 0.000016s : 0.00% auto_monad : 0.000299s : 0.01% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000024s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000004s : 0.00% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000101s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000037s : 0.00% optimize.rewriter_before_opt_a : 0.000087s : 0.00% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000056s : 0.00% optimize.opt_a.loop_unroll : 0.000037s : 0.00% optimize.opt_a.a_1 : 0.001334s : 0.04% optimize.opt_a.with_stream_mark : 0.000062s : 0.00% optimize.opt_a.recompute_prepare : 0.000041s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000023s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000016s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000022s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000470s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000067s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000011s : 0.00% optimize.opt_a.shard_inline : 0.000033s : 0.00% optimize.opt_a.merge_send_recv : 0.000031s : 0.00% optimize.opt_a.auto_parallel : 0.000028s : 0.00% optimize.opt_a.parallel : 0.000052s : 0.00% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000018s : 0.00% optimize.opt_a.allreduce_fusion : 0.000018s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000037s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000039s : 0.00% optimize.opt_a.virtual_dataset : 0.000034s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000035s : 0.00% optimize.opt_a.virtual_output : 0.000053s : 0.00% optimize.opt_a.merge_forward : 0.000019s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000037s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000080s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000092s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000019s : 0.00% optimize.opt_a.meta_fg_expand : 0.000015s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000010s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000072s : 0.00% optimize.opt_a.a_after_grad : 0.000128s : 0.00% optimize.opt_a.renormalize : 0.185725s : 5.94% optimize.opt_a.add_forward_monad_depend : 0.000018s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000130s : 0.00% optimize.opt_a.cse : 0.001501s : 0.05% optimize.opt_a.a_3 : 0.000263s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000032s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000253s : 0.01% optimize.convert_after_rewriter : 0.000015s : 0.00% optimize.order_py_execute_after_rewriter : 0.000012s : 0.00% optimize.mutable_eliminate : 0.000890s : 0.03% optimize.opt_b.b_1 : 0.000388s : 0.01% optimize.opt_b.b_2 : 0.000019s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000019s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000091s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000036s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000047s : 0.00% optimize.loop_unroll : 0.000749s : 0.02% optimize.opt_after_cconv.c_1 : 0.000117s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000019s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.cse : 0.000065s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000080s : 0.00% optimize.tuple_transform.d_1 : 0.000125s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000020s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000132s : 0.00% optimize.cse_after_recomputation.cse : 0.000053s : 0.00% optimize.environ_conv : 0.000023s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000013s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000008s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000028s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000007s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000008s : 0.00% optimize.overlap_grad_flash_sp : 0.000040s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000021s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000025s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000031s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000025s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000025s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000050s : 0.00% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000009s : 0.00% opt_after_jit_grad : 0.000935s : 0.03% validate : 0.000111s : 0.00% Time group info: ------[substitution.] 0.000320 96 3.47% : 0.000011s : 2: substitution.depend_value_elim 1.36% : 0.000004s : 8: substitution.elim_not_effective 1.01% : 0.000003s : 8: substitution.fold_const_symbol 4.36% : 0.000014s : 13: substitution.graph_param_transform 39.36% : 0.000126s : 1: substitution.inline 2.76% : 0.000009s : 16: substitution.j_node_and_user_rematch 7.26% : 0.000023s : 2: substitution.less_batch_normalization 2.28% : 0.000007s : 6: substitution.load_eliminater 0.95% : 0.000003s : 2: substitution.opt_reshape 3.84% : 0.000012s : 16: substitution.remove_not_recompute_node 2.40% : 0.000008s : 4: substitution.replace_old_param 15.52% : 0.000050s : 4: substitution.reshape_eliminate 8.15% : 0.000026s : 6: substitution.updatestate_pure_node_eliminater 7.28% : 0.000023s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.928241 2 99.82% : 2.922845s : 1: type_inference.infer 0.18% : 0.005396s : 1: type_inference.specialize ------[replace.] 0.000018 1 100.00% : 0.000018s : 1: replace.inline ------[match.] 0.000125 1 100.00% : 0.000125s : 1: match.inline ------[predicate.] 0.000460 3002 0.78% : 0.000004s : 27: predicate.accumulaten_eliminater 1.55% : 0.000007s : 13: predicate.ad_related_special_op_eliminate 0.72% : 0.000003s : 26: predicate.addn_check_dump 0.80% : 0.000004s : 27: predicate.addn_zero_filter 0.78% : 0.000004s : 27: predicate.adjust_all_reduce_mul_add 2.04% : 0.000009s : 53: predicate.arithmetic_simplify 0.77% : 0.000004s : 27: predicate.cast_eliminate 0.77% : 0.000004s : 26: predicate.check_bprop_eliminate 0.74% : 0.000003s : 26: predicate.compare_switch_simplify 0.24% : 0.000001s : 13: predicate.const_output_eliminate 0.80% : 0.000004s : 26: predicate.depend_value_elim 0.81% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.85% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.76% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.21% : 0.000006s : 26: predicate.dumpgradient_eliminate 0.26% : 0.000001s : 13: predicate.elim_not_effective 0.58% : 0.000003s : 13: predicate.elim_shapecalc_of_broadcastargs 1.15% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.07% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.16% : 0.000005s : 40: predicate.environ_get_depend_swap 1.85% : 0.000009s : 66: predicate.environ_get_eliminate 1.06% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.73% : 0.000003s : 28: predicate.exchange_switch_depend_value 1.41% : 0.000006s : 28: predicate.float_depend_g_call 0.73% : 0.000003s : 26: predicate.float_environ_get_switch 1.08% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.20% : 0.000001s : 13: predicate.fold_const_symbol 0.85% : 0.000004s : 26: predicate.get_grad_eliminate 0.36% : 0.000002s : 13: predicate.graph_param_transform 0.73% : 0.000003s : 26: predicate.incorporate_call 0.67% : 0.000003s : 26: predicate.incorporate_call_switch 5.45% : 0.000025s : 133: predicate.inline 1.22% : 0.000006s : 26: predicate.inline_without_move 0.42% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.20% : 0.000006s : 27: predicate.less_batch_normalization 1.71% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.22% : 0.000010s : 80: predicate.load_eliminater 1.25% : 0.000006s : 13: predicate.loop_unroll_after_grad 0.98% : 0.000005s : 31: predicate.loop_unroll_before_grad 1.79% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.75% : 0.000003s : 26: predicate.merge_addn 0.80% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.78% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.69% : 0.000003s : 27: predicate.minmaximum_grad 1.65% : 0.000008s : 13: predicate.mutable_eliminate 0.71% : 0.000003s : 13: predicate.opt_reshape 0.45% : 0.000002s : 13: predicate.parallel_virtual_node 1.13% : 0.000005s : 28: predicate.partial_defer_inline 1.24% : 0.000006s : 40: predicate.partial_eliminate 0.79% : 0.000004s : 27: predicate.print_const_string_wrapper 0.75% : 0.000003s : 26: predicate.reduce_all_const_elim 1.01% : 0.000005s : 27: predicate.reduce_eliminate 2.27% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.46% : 0.000002s : 26: predicate.remove_not_recompute_node 1.46% : 0.000007s : 53: predicate.replace_applicator 0.58% : 0.000003s : 26: predicate.replace_old_param 0.29% : 0.000001s : 13: predicate.reset_defer_inline 0.95% : 0.000004s : 27: predicate.reshape_eliminate 0.94% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.47% : 0.000002s : 13: predicate.row_tensor_eliminate 1.45% : 0.000007s : 26: predicate.same_eliminate 0.53% : 0.000002s : 26: predicate.set_cell_output_no_recompute 0.97% : 0.000004s : 26: predicate.shard_identity_eliminate 0.98% : 0.000005s : 26: predicate.special_op_eliminate 0.84% : 0.000004s : 26: predicate.specialize_transform 1.21% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 1.02% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.51% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.87% : 0.000004s : 28: predicate.switch_defer_inline 1.69% : 0.000008s : 54: predicate.switch_layer_defer_inline 3.57% : 0.000016s : 98: predicate.switch_simplify 0.76% : 0.000003s : 27: predicate.tile_eliminate 0.75% : 0.000003s : 27: predicate.transpose_eliminate 1.81% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 1.92% : 0.000009s : 53: predicate.tuple_list_get_item_const_eliminator 1.62% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.93% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.63% : 0.000007s : 53: predicate.tuple_list_get_set_item_eliminator 2.70% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.67% : 0.000008s : 53: predicate.tuple_to_list_eliminator_ 2.24% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.24% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.51% : 0.000002s : 13: predicate.value_based_eliminate 0.89% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.85% : 0.000004s : 26: predicate.virtual_output_eliminate 0.39% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.53% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.285273 58 37.81% : 0.107869s : 55: func_graph_cloner_run.FuncGraphClonerGraph 62.19% : 0.177404s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 3.829007 192 0.00% : 0.000003s : 1: ForceFp32Comm 4.15% : 0.159048s : 1: add_attr 4.15% : 0.159029s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000138s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000310s : 1: auto_monad 0.00% : 0.000056s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.02% : 0.000802s : 1: bootstrap 0.00% : 0.000051s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000033s : 1: control_data_broadcast_order 0.00% : 0.000020s : 1: convert_after_rewriter 0.00% : 0.000069s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000027s : 1: environ_conv 0.00% : 0.000023s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000012s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000009s : 1: insert-virtual-dataset 0.00% : 0.000005s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.02% : 0.000766s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.02% : 0.000908s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000059s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000045s : 1: opt.transform.mutable_eliminate 0.07% : 0.002707s : 78: opt.transform.opt_a 0.00% : 0.000115s : 1: opt.transform.opt_after_cconv 0.00% : 0.000076s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000368s : 28: opt.transform.opt_b 0.00% : 0.000141s : 2: opt.transform.opt_trans_graph 0.00% : 0.000100s : 4: opt.transform.symbol_engine_opt 5.00% : 0.191357s : 1: opt_a 0.01% : 0.000283s : 1: opt_after_cconv 0.02% : 0.000955s : 1: opt_after_jit_grad 0.02% : 0.000603s : 1: opt_b 5.11% : 0.195499s : 1: optimize 0.00% : 0.000040s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000016s : 1: order_py_execute_after_rewriter 0.00% : 0.000044s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000007s : 1: overlap_param_gather 0.00% : 0.000006s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000011s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000006s : 1: pipeline_split 0.00% : 0.000110s : 1: pre_auto_parallel 0.00% : 0.000042s : 1: py_interpret_to_execute 0.00% : 0.000037s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000085s : 1: remove_dup_value 0.20% : 0.007818s : 1: renormalize.infer 4.65% : 0.177888s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000013s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000264s : 1: rewriter_after_opt_a 0.00% : 0.000092s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000006s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000016s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000170s : 1: symbol_engine_optimizer 0.00% : 0.000179s : 1: tuple_transform 76.48% : 2.928361s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape1-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape1-1],max_mem:140.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:39:12.960.014 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 3.13943, [21] [bootstrap]: 0.00077891 [type_inference]: 2.79662 [event_method]: 1.959e-05 [auto_monad]: 0.00038525 [graph_reusing]: 7.46001e-06 [inline]: 3.36001e-06 [add_attr]: 0.125181, [1] [add_attr_with_inline]: 0.125159, [1] [Cycle 1]: 8.375e-05, [2] [tag_attr]: 2.539e-05 [meta_addattr_fg_expand]: 4.34002e-06 [parallel-infer-symbol]: 6.17999e-06 [pre_auto_parallel]: 4.052e-05 [insert-virtual-dataset]: 2.88e-06 [parallel-infer-symbol-second]: 7.50006e-07 [dataset_repeat_opt]: 2.24001e-06 [pipeline_split]: 1.66e-06 [optimize]: 0.215322, [53] [py_interpret_to_execute]: 3.738e-05 [rewriter_before_opt_a]: 9.081e-05 [opt_a]: 0.211589, [2] [Cycle 1]: 0.0145164, [45] [expand_dump_flag]: 3.21999e-06 [switch_simplify]: 3.677e-05 [loop_unroll]: 2.059e-05 [a_1]: 0.0008775 [with_stream_mark]: 2.664e-05 [recompute_prepare]: 2.22e-05 [updatestate_depend_eliminate]: 1.133e-05 [updatestate_assign_eliminate]: 8.45001e-06 [updatestate_loads_eliminate]: 1.138e-05 [parameter_eliminate]: 2.61e-06 [a_2]: 0.00025027 [accelerated_algorithm]: 4.176e-05 [shard]: 2.44001e-06 [meta_shard_fg_expand]: 3.7e-06 [shard_inline]: 1.798e-05 [merge_send_recv]: 1.48e-05 [auto_parallel]: 1.262e-05 [parallel]: 3.883e-05 [flash_sp]: 1.252e-05 [merge_comm]: 8.97e-06 [allreduce_fusion]: 8.09997e-06 [matmul_add_comm_reduction]: 1.867e-05 [allreduce_slice_to_reducescatter]: 1.10999e-06 [virtual_shard_identity]: 2.559e-05 [virtual_dataset]: 1.723e-05 [get_grad_eliminate_]: 1.929e-05 [virtual_output]: 1.729e-05 [merge_forward]: 9.25999e-06 [cell_reuse_recompute_pass]: 1.84e-06 [offload_activation]: 1.977e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.493e-05 [merge_recompute_call_nodes]: 1.77999e-06 [before_grad]: 2.952e-05 [set_forward_comm_id_for_comm_node_pass]: 9.10001e-06 [meta_fg_expand]: 5.72999e-06 [flash_sp_send_recv_attached]: 6.72002e-06 [receive_attached]: 2.52001e-06 [after_resolve]: 2.52e-05 [a_after_grad]: 3.092e-05 [renormalize]: 0.0120379 [add_forward_monad_depend]: 1.237e-05 [auto_monad_grad]: 2.219e-05 [auto_monad_eliminator]: 6.556e-05 [cse]: 0.00021039 [a_3]: 0.00013885 [Cycle 2]: 0.197055, [45] [expand_dump_flag]: 2.73998e-06 [switch_simplify]: 2.028e-05 [loop_unroll]: 1.685e-05 [a_1]: 0.00046105 [with_stream_mark]: 2.719e-05 [recompute_prepare]: 4.437e-05 [updatestate_depend_eliminate]: 1.134e-05 [updatestate_assign_eliminate]: 8.35001e-06 [updatestate_loads_eliminate]: 1.2e-05 [parameter_eliminate]: 3.14001e-06 [a_2]: 0.00024526 [accelerated_algorithm]: 2.393e-05 [shard]: 3.339e-05 [meta_shard_fg_expand]: 5.18002e-06 [shard_inline]: 2.044e-05 [merge_send_recv]: 1.718e-05 [auto_parallel]: 1.605e-05 [parallel]: 1.032e-05 [flash_sp]: 4.62e-06 [merge_comm]: 9.56003e-06 [allreduce_fusion]: 9.32999e-06 [matmul_add_comm_reduction]: 1.832e-05 [allreduce_slice_to_reducescatter]: 9.09989e-07 [virtual_shard_identity]: 1.876e-05 [virtual_dataset]: 1.601e-05 [get_grad_eliminate_]: 1.606e-05 [virtual_output]: 1.505e-05 [merge_forward]: 8.84e-06 [cell_reuse_recompute_pass]: 3.55998e-06 [offload_activation]: 1.853e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.102e-05 [merge_recompute_call_nodes]: 1.57001e-06 [before_grad]: 2.617e-05 [set_forward_comm_id_for_comm_node_pass]: 9.59e-06 [meta_fg_expand]: 6.98e-06 [flash_sp_send_recv_attached]: 2.34001e-06 [receive_attached]: 2.49999e-06 [after_resolve]: 2.451e-05 [a_after_grad]: 2.831e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 3.93001e-06 [auto_monad_grad]: 2.57001e-06 [auto_monad_eliminator]: 5.006e-05 [cse]: 0.195206 [a_3]: 0.00021884 [py_interpret_to_execute_after_opt_a]: 3.739e-05 [slice_cell_reuse_recomputed_activation]: 4.04002e-06 [rewriter_after_opt_a]: 0.00024589 [convert_after_rewriter]: 1.77e-05 [order_py_execute_after_rewriter]: 1.145e-05 [mutable_eliminate]: 0.00083968 [opt_b]: 0.00057965, [1] [Cycle 1]: 0.00056961, [7] [b_1]: 0.00038896 [b_2]: 1.881e-05 [updatestate_depend_eliminate]: 1.555e-05 [updatestate_assign_eliminate]: 8.36002e-06 [updatestate_loads_eliminate]: 1.128e-05 [renormalize]: 6.39993e-07 [cse]: 8.127e-05 [optimize_parallel_all_gather_comm]: 3.551e-05 [overlap_param_gather]: 4.50999e-06 [cconv]: 4.383e-05 [loop_unroll]: 0.0005299 [opt_after_cconv]: 0.00023013, [1] [Cycle 1]: 0.00022154, [7] [c_1]: 9.98e-05 [parameter_eliminate]: 4.97e-06 [updatestate_depend_eliminate]: 1.324e-05 [updatestate_assign_eliminate]: 7.06999e-06 [updatestate_loads_eliminate]: 9.05999e-06 [cse]: 4.907e-05 [renormalize]: 6.00005e-07 [remove_dup_value]: 7.472e-05 [tuple_transform]: 0.00015944, [1] [Cycle 1]: 0.00015405, [4] [d_1]: 0.00011218 [none_parameter_eliminate]: 2.17001e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 1.715e-05 [partial_unused_args_eliminate]: 2.07999e-06 [add_recomputation]: 0.00012822 [cse_after_recomputation]: 5.784e-05, [1] [Cycle 1]: 5.271e-05, [1] [cse]: 4.656e-05 [environ_conv]: 1.932e-05 [swap_dp_allreduce_reducescatter]: 1.165e-05 [bias_add_comm_swap]: 4.37e-06 [label_micro_interleaved_index]: 5.05999e-06 [label_fine_grained_interleaved_index]: 3.35e-06 [merge_cast_opt]: 1.64e-06 [slice_recompute_activation]: 2.36998e-06 [micro_interleaved_order_control]: 2.61999e-06 [assign_add_opt]: 1.45001e-06 [ForceFp32Comm]: 9.09989e-07 [remove_cast_before_assign_add]: 1.16002e-06 [full_micro_interleaved_order_control]: 2.98e-06 [reorder_send_recv_between_fp_bp]: 2.71e-06 [comm_op_add_attrs]: 1.10999e-06 [add_comm_op_reuse_tag]: 1.05001e-06 [interleave_split_concat_branches]: 1.22e-06 [interleave_parallel_branches]: 1.09003e-06 [overlap_opt_shard_in_pipeline]: 3.51001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.19999e-06 [control_data_broadcast_order]: 2.646e-05 [grouped_pairwise_exchange_alltoall]: 2.06e-06 [offloading_packed_experts]: 6.79999e-06 [overlap_recompute_and_grad_model_parallel]: 7.98999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.34e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40999e-06 [overlap_recompute_comm]: 2.93e-06 [overlap_grad_ring_attention]: 8.12e-06 [overlap_grad_flash_sp]: 4.005e-05 [begin_end_overlap_inline]: 5.60016e-07 [split_matmul_comm_elemetwise]: 2.32001e-06 [split_layernorm_comm]: 1.97999e-06 [handle_group_info]: 1.00999e-06 [symbol_engine_optimizer]: 0.00015646, [1] [Cycle 1]: 0.00015008, [6] [build]: 1.893e-05 [elim_shapecalc]: 2.176e-05 [elim_not_effective]: 2.877e-05 [opt_reshape]: 2.258e-05 [fold_const_symbol]: 2.42e-05 [renormalize]: 2.00002e-07 [detach_backward]: 2.39001e-06 [pipeline_parallel_scheduler]: 1.50999e-06 [auto_monad_reorder]: 4.662e-05 [get_jit_bprop_graph]: 2.22999e-06 [rewriter_after_jit_bprop_graph]: 6.06998e-06 [opt_after_jit_grad]: 0.00064398 [validate]: 0.00010209 Sums bootstrap : 0.000779s : 0.03% type_inference : 2.796619s : 92.82% event_method : 0.000020s : 0.00% auto_monad : 0.000385s : 0.01% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000025s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000004s : 0.00% parallel-infer-symbol : 0.000006s : 0.00% pre_auto_parallel : 0.000041s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000037s : 0.00% optimize.rewriter_before_opt_a : 0.000091s : 0.00% optimize.opt_a.expand_dump_flag : 0.000006s : 0.00% optimize.opt_a.switch_simplify : 0.000057s : 0.00% optimize.opt_a.loop_unroll : 0.000037s : 0.00% optimize.opt_a.a_1 : 0.001339s : 0.04% optimize.opt_a.with_stream_mark : 0.000054s : 0.00% optimize.opt_a.recompute_prepare : 0.000067s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000023s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000017s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000023s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000496s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000066s : 0.00% optimize.opt_a.shard : 0.000036s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000009s : 0.00% optimize.opt_a.shard_inline : 0.000038s : 0.00% optimize.opt_a.merge_send_recv : 0.000032s : 0.00% optimize.opt_a.auto_parallel : 0.000029s : 0.00% optimize.opt_a.parallel : 0.000049s : 0.00% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000019s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000037s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000044s : 0.00% optimize.opt_a.virtual_dataset : 0.000033s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000035s : 0.00% optimize.opt_a.virtual_output : 0.000032s : 0.00% optimize.opt_a.merge_forward : 0.000018s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000038s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000066s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000056s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000019s : 0.00% optimize.opt_a.meta_fg_expand : 0.000013s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000050s : 0.00% optimize.opt_a.a_after_grad : 0.000059s : 0.00% optimize.opt_a.renormalize : 0.012038s : 0.40% optimize.opt_a.add_forward_monad_depend : 0.000016s : 0.00% optimize.opt_a.auto_monad_grad : 0.000025s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000116s : 0.00% optimize.opt_a.cse : 0.195417s : 6.49% optimize.opt_a.a_3 : 0.000358s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000037s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000004s : 0.00% optimize.rewriter_after_opt_a : 0.000246s : 0.01% optimize.convert_after_rewriter : 0.000018s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.000840s : 0.03% optimize.opt_b.b_1 : 0.000389s : 0.01% optimize.opt_b.b_2 : 0.000019s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000016s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000081s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000036s : 0.00% optimize.overlap_param_gather : 0.000005s : 0.00% optimize.cconv : 0.000044s : 0.00% optimize.loop_unroll : 0.000530s : 0.02% optimize.opt_after_cconv.c_1 : 0.000100s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.cse : 0.000049s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000075s : 0.00% optimize.tuple_transform.d_1 : 0.000112s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000017s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000128s : 0.00% optimize.cse_after_recomputation.cse : 0.000047s : 0.00% optimize.environ_conv : 0.000019s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000012s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000026s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000007s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000008s : 0.00% optimize.overlap_grad_flash_sp : 0.000040s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000019s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000022s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000029s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000023s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000024s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000047s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000644s : 0.02% validate : 0.000102s : 0.00% Time group info: ------[substitution.] 0.000309 96 3.48% : 0.000011s : 2: substitution.depend_value_elim 1.15% : 0.000004s : 8: substitution.elim_not_effective 1.07% : 0.000003s : 8: substitution.fold_const_symbol 3.79% : 0.000012s : 13: substitution.graph_param_transform 45.99% : 0.000142s : 1: substitution.inline 2.83% : 0.000009s : 16: substitution.j_node_and_user_rematch 6.91% : 0.000021s : 2: substitution.less_batch_normalization 2.43% : 0.000008s : 6: substitution.load_eliminater 1.05% : 0.000003s : 2: substitution.opt_reshape 3.70% : 0.000011s : 16: substitution.remove_not_recompute_node 2.90% : 0.000009s : 4: substitution.replace_old_param 10.13% : 0.000031s : 4: substitution.reshape_eliminate 8.13% : 0.000025s : 6: substitution.updatestate_pure_node_eliminater 6.43% : 0.000020s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.796507 2 99.80% : 2.790989s : 1: type_inference.infer 0.20% : 0.005518s : 1: type_inference.specialize ------[replace.] 0.000019 1 100.00% : 0.000019s : 1: replace.inline ------[match.] 0.000141 1 100.00% : 0.000141s : 1: match.inline ------[predicate.] 0.000464 3002 0.78% : 0.000004s : 27: predicate.accumulaten_eliminater 0.90% : 0.000004s : 13: predicate.ad_related_special_op_eliminate 0.76% : 0.000004s : 26: predicate.addn_check_dump 0.79% : 0.000004s : 27: predicate.addn_zero_filter 0.74% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.18% : 0.000010s : 53: predicate.arithmetic_simplify 0.82% : 0.000004s : 27: predicate.cast_eliminate 0.79% : 0.000004s : 26: predicate.check_bprop_eliminate 0.75% : 0.000003s : 26: predicate.compare_switch_simplify 0.24% : 0.000001s : 13: predicate.const_output_eliminate 0.82% : 0.000004s : 26: predicate.depend_value_elim 0.79% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.88% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.77% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.90% : 0.000009s : 26: predicate.dumpgradient_eliminate 0.29% : 0.000001s : 13: predicate.elim_not_effective 0.50% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 2.46% : 0.000011s : 40: predicate.environ_add_const_eliminate 1.12% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.15% : 0.000005s : 40: predicate.environ_get_depend_swap 1.88% : 0.000009s : 66: predicate.environ_get_eliminate 1.13% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.74% : 0.000003s : 28: predicate.exchange_switch_depend_value 1.36% : 0.000006s : 28: predicate.float_depend_g_call 0.75% : 0.000004s : 26: predicate.float_environ_get_switch 1.11% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.20% : 0.000001s : 13: predicate.fold_const_symbol 0.92% : 0.000004s : 26: predicate.get_grad_eliminate 0.29% : 0.000001s : 13: predicate.graph_param_transform 0.76% : 0.000004s : 26: predicate.incorporate_call 0.73% : 0.000003s : 26: predicate.incorporate_call_switch 5.25% : 0.000024s : 133: predicate.inline 0.98% : 0.000005s : 26: predicate.inline_without_move 0.48% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.17% : 0.000005s : 27: predicate.less_batch_normalization 1.82% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.15% : 0.000010s : 80: predicate.load_eliminater 0.94% : 0.000004s : 13: predicate.loop_unroll_after_grad 0.93% : 0.000004s : 31: predicate.loop_unroll_before_grad 1.78% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.80% : 0.000004s : 26: predicate.merge_addn 0.82% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.81% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.68% : 0.000003s : 27: predicate.minmaximum_grad 1.02% : 0.000005s : 13: predicate.mutable_eliminate 0.47% : 0.000002s : 13: predicate.opt_reshape 0.48% : 0.000002s : 13: predicate.parallel_virtual_node 1.12% : 0.000005s : 28: predicate.partial_defer_inline 1.24% : 0.000006s : 40: predicate.partial_eliminate 0.78% : 0.000004s : 27: predicate.print_const_string_wrapper 0.77% : 0.000004s : 26: predicate.reduce_all_const_elim 1.05% : 0.000005s : 27: predicate.reduce_eliminate 2.18% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.46% : 0.000002s : 26: predicate.remove_not_recompute_node 1.52% : 0.000007s : 53: predicate.replace_applicator 0.56% : 0.000003s : 26: predicate.replace_old_param 0.41% : 0.000002s : 13: predicate.reset_defer_inline 0.90% : 0.000004s : 27: predicate.reshape_eliminate 0.80% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.53% : 0.000002s : 13: predicate.row_tensor_eliminate 1.85% : 0.000009s : 26: predicate.same_eliminate 0.55% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.91% : 0.000004s : 26: predicate.shard_identity_eliminate 0.87% : 0.000004s : 26: predicate.special_op_eliminate 0.87% : 0.000004s : 26: predicate.specialize_transform 1.43% : 0.000007s : 26: predicate.split_environ_get_set_with_tuple_value 1.03% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.46% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.86% : 0.000004s : 28: predicate.switch_defer_inline 1.58% : 0.000007s : 54: predicate.switch_layer_defer_inline 3.30% : 0.000015s : 98: predicate.switch_simplify 0.78% : 0.000004s : 27: predicate.tile_eliminate 0.78% : 0.000004s : 27: predicate.transpose_eliminate 1.77% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 1.63% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.56% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 3.01% : 0.000014s : 79: predicate.tuple_list_get_item_eliminator 1.66% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.54% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.60% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.19% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.57% : 0.000017s : 106: predicate.updatestate_useless_node_eliminater 0.44% : 0.000002s : 13: predicate.value_based_eliminate 0.88% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.83% : 0.000004s : 26: predicate.virtual_output_eliminate 0.37% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.48% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.025463 58 90.90% : 0.023144s : 55: func_graph_cloner_run.FuncGraphClonerGraph 9.10% : 0.002318s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 3.495207 192 0.00% : 0.000003s : 1: ForceFp32Comm 3.58% : 0.125189s : 1: add_attr 3.58% : 0.125164s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000134s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000399s : 1: auto_monad 0.00% : 0.000052s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.02% : 0.000855s : 1: bootstrap 0.00% : 0.000048s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000030s : 1: control_data_broadcast_order 0.00% : 0.000022s : 1: convert_after_rewriter 0.00% : 0.000061s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000024s : 1: environ_conv 0.00% : 0.000028s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000012s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000003s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000008s : 1: insert-virtual-dataset 0.00% : 0.000006s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.02% : 0.000541s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.02% : 0.000851s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000028s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000034s : 1: opt.transform.mutable_eliminate 0.08% : 0.002684s : 78: opt.transform.opt_a 0.00% : 0.000098s : 1: opt.transform.opt_after_cconv 0.00% : 0.000059s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000369s : 28: opt.transform.opt_b 0.00% : 0.000127s : 2: opt.transform.opt_trans_graph 0.00% : 0.000093s : 4: opt.transform.symbol_engine_opt 6.05% : 0.211593s : 1: opt_a 0.01% : 0.000234s : 1: opt_after_cconv 0.02% : 0.000656s : 1: opt_after_jit_grad 0.02% : 0.000584s : 1: opt_b 6.16% : 0.215329s : 1: optimize 0.00% : 0.000039s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000015s : 1: order_py_execute_after_rewriter 0.00% : 0.000044s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000006s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000009s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000011s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000010s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000006s : 1: pipeline_split 0.00% : 0.000045s : 1: pre_auto_parallel 0.00% : 0.000042s : 1: py_interpret_to_execute 0.00% : 0.000042s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000079s : 1: remove_dup_value 0.27% : 0.009292s : 1: renormalize.infer 0.08% : 0.002727s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000253s : 1: rewriter_after_opt_a 0.00% : 0.000095s : 1: rewriter_before_opt_a 0.00% : 0.000007s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000006s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000015s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000159s : 1: symbol_engine_optimizer 0.00% : 0.000162s : 1: tuple_transform 80.01% : 2.796652s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape1-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape1-16],max_mem:140.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:39:51.554.20 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 4.69946, [21] [bootstrap]: 0.106797 [type_inference]: 3.94453 [event_method]: 1.939e-05 [auto_monad]: 0.00059252 [graph_reusing]: 7.05e-06 [inline]: 2.80002e-06 [add_attr]: 0.4136, [1] [add_attr_with_inline]: 0.413508, [1] [Cycle 1]: 9.684e-05, [2] [tag_attr]: 2.561e-05 [meta_addattr_fg_expand]: 4.68999e-06 [parallel-infer-symbol]: 5.39998e-06 [pre_auto_parallel]: 3.953e-05 [insert-virtual-dataset]: 2.88e-06 [parallel-infer-symbol-second]: 7.09988e-07 [dataset_repeat_opt]: 2.01e-06 [pipeline_split]: 1.59998e-06 [optimize]: 0.232617, [53] [py_interpret_to_execute]: 3.877e-05 [rewriter_before_opt_a]: 0.00011429 [opt_a]: 0.224499, [2] [Cycle 1]: 0.221334, [45] [expand_dump_flag]: 3.2e-06 [switch_simplify]: 3.609e-05 [loop_unroll]: 2.034e-05 [a_1]: 0.00148805 [with_stream_mark]: 2.593e-05 [recompute_prepare]: 2.268e-05 [updatestate_depend_eliminate]: 1.084e-05 [updatestate_assign_eliminate]: 8.68001e-06 [updatestate_loads_eliminate]: 1.155e-05 [parameter_eliminate]: 2.21e-06 [a_2]: 0.0006267 [accelerated_algorithm]: 0.00011196 [shard]: 2.51998e-06 [meta_shard_fg_expand]: 3.66001e-06 [shard_inline]: 2.642e-05 [merge_send_recv]: 1.774e-05 [auto_parallel]: 1.362e-05 [parallel]: 0.00039007 [flash_sp]: 1.397e-05 [merge_comm]: 9.33002e-06 [allreduce_fusion]: 8.46002e-06 [matmul_add_comm_reduction]: 1.698e-05 [allreduce_slice_to_reducescatter]: 7.30011e-07 [virtual_shard_identity]: 2.198e-05 [virtual_dataset]: 1.81e-05 [get_grad_eliminate_]: 1.916e-05 [virtual_output]: 1.782e-05 [merge_forward]: 8.87999e-06 [cell_reuse_recompute_pass]: 1.89e-06 [offload_activation]: 2.213e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.525e-05 [merge_recompute_call_nodes]: 1.62001e-06 [before_grad]: 2.875e-05 [set_forward_comm_id_for_comm_node_pass]: 8.87e-06 [meta_fg_expand]: 6.59001e-06 [flash_sp_send_recv_attached]: 6.21e-06 [receive_attached]: 2.74999e-06 [after_resolve]: 0.00023533 [a_after_grad]: 3.131e-05 [renormalize]: 0.215045 [add_forward_monad_depend]: 1.466e-05 [auto_monad_grad]: 2.90998e-06 [auto_monad_eliminator]: 6.548e-05 [cse]: 0.00015715 [a_3]: 0.00014664 [Cycle 2]: 0.00314904, [45] [expand_dump_flag]: 2.78998e-06 [switch_simplify]: 1.981e-05 [loop_unroll]: 1.691e-05 [a_1]: 0.00047004 [with_stream_mark]: 2.485e-05 [recompute_prepare]: 1.734e-05 [updatestate_depend_eliminate]: 1.113e-05 [updatestate_assign_eliminate]: 8.15e-06 [updatestate_loads_eliminate]: 1.169e-05 [parameter_eliminate]: 2.58e-06 [a_2]: 0.00031103 [accelerated_algorithm]: 2.098e-05 [shard]: 2.51e-06 [meta_shard_fg_expand]: 4.26001e-06 [shard_inline]: 1.64e-05 [merge_send_recv]: 1.789e-05 [auto_parallel]: 1.796e-05 [parallel]: 1.076e-05 [flash_sp]: 5.00001e-06 [merge_comm]: 8.80999e-06 [allreduce_fusion]: 9.09e-06 [matmul_add_comm_reduction]: 1.784e-05 [allreduce_slice_to_reducescatter]: 8.70001e-07 [virtual_shard_identity]: 2.151e-05 [virtual_dataset]: 1.638e-05 [get_grad_eliminate_]: 1.533e-05 [virtual_output]: 1.524e-05 [merge_forward]: 9.08002e-06 [cell_reuse_recompute_pass]: 3.26999e-06 [offload_activation]: 1.733e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.053e-05 [merge_recompute_call_nodes]: 1.77001e-06 [before_grad]: 2.601e-05 [set_forward_comm_id_for_comm_node_pass]: 9.37999e-06 [meta_fg_expand]: 7.14001e-06 [flash_sp_send_recv_attached]: 2.08002e-06 [receive_attached]: 2.93e-06 [after_resolve]: 2.424e-05 [a_after_grad]: 2.625e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.29001e-06 [auto_monad_grad]: 1.71002e-06 [auto_monad_eliminator]: 3.544e-05 [cse]: 0.00081578 [a_3]: 0.00011333 [py_interpret_to_execute_after_opt_a]: 2.839e-05 [slice_cell_reuse_recomputed_activation]: 2.32999e-06 [rewriter_after_opt_a]: 0.00239545 [convert_after_rewriter]: 4.802e-05 [order_py_execute_after_rewriter]: 1.097e-05 [mutable_eliminate]: 0.00103379 [opt_b]: 0.00051961, [1] [Cycle 1]: 0.00051113, [7] [b_1]: 0.0003692 [b_2]: 1.768e-05 [updatestate_depend_eliminate]: 1.166e-05 [updatestate_assign_eliminate]: 7.03998e-06 [updatestate_loads_eliminate]: 9.67001e-06 [renormalize]: 5.19998e-07 [cse]: 5.629e-05 [optimize_parallel_all_gather_comm]: 4.42e-05 [overlap_param_gather]: 2.658e-05 [cconv]: 3.748e-05 [loop_unroll]: 0.00073746 [opt_after_cconv]: 0.00021671, [1] [Cycle 1]: 0.00020888, [7] [c_1]: 9.811e-05 [parameter_eliminate]: 2.86e-06 [updatestate_depend_eliminate]: 1.224e-05 [updatestate_assign_eliminate]: 7e-06 [updatestate_loads_eliminate]: 8.33001e-06 [cse]: 4.35e-05 [renormalize]: 8.39995e-07 [remove_dup_value]: 6.758e-05 [tuple_transform]: 0.00015766, [1] [Cycle 1]: 0.00015208, [4] [d_1]: 0.00011218 [none_parameter_eliminate]: 2.02999e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 1.674e-05 [partial_unused_args_eliminate]: 2.21998e-06 [add_recomputation]: 0.00105527 [cse_after_recomputation]: 0.0008084, [1] [Cycle 1]: 0.00080168, [1] [cse]: 0.00079098 [environ_conv]: 1.751e-05 [swap_dp_allreduce_reducescatter]: 1.349e-05 [bias_add_comm_swap]: 2.55002e-06 [label_micro_interleaved_index]: 4.57e-06 [label_fine_grained_interleaved_index]: 2.74999e-06 [merge_cast_opt]: 1.53002e-06 [slice_recompute_activation]: 2.43e-06 [micro_interleaved_order_control]: 2.99999e-06 [assign_add_opt]: 1.34998e-06 [ForceFp32Comm]: 8.80013e-07 [remove_cast_before_assign_add]: 1.32999e-06 [full_micro_interleaved_order_control]: 2.94001e-06 [reorder_send_recv_between_fp_bp]: 3.03998e-06 [comm_op_add_attrs]: 1.41998e-06 [add_comm_op_reuse_tag]: 1.30001e-06 [interleave_split_concat_branches]: 1.17999e-06 [interleave_parallel_branches]: 1.30999e-06 [overlap_opt_shard_in_pipeline]: 9.89001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.80001e-06 [control_data_broadcast_order]: 2.752e-05 [grouped_pairwise_exchange_alltoall]: 1.69998e-06 [offloading_packed_experts]: 6.64999e-06 [overlap_recompute_and_grad_model_parallel]: 7.99002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.64e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40999e-06 [overlap_recompute_comm]: 2.12999e-06 [overlap_grad_ring_attention]: 7.62998e-06 [overlap_grad_flash_sp]: 8.507e-05 [begin_end_overlap_inline]: 5.59987e-07 [split_matmul_comm_elemetwise]: 2.44999e-06 [split_layernorm_comm]: 1.83997e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 0.00016359, [1] [Cycle 1]: 0.00015627, [6] [build]: 1.903e-05 [elim_shapecalc]: 2.381e-05 [elim_not_effective]: 3.114e-05 [opt_reshape]: 2.256e-05 [fold_const_symbol]: 2.487e-05 [renormalize]: 2.09984e-07 [detach_backward]: 2.48002e-06 [pipeline_parallel_scheduler]: 1.73002e-06 [auto_monad_reorder]: 4.683e-05 [get_jit_bprop_graph]: 1.89e-06 [rewriter_after_jit_bprop_graph]: 4.17e-06 [opt_after_jit_grad]: 0.00077662 [validate]: 0.00012209 Sums bootstrap : 0.106797s : 2.49% type_inference : 3.944534s : 92.13% event_method : 0.000019s : 0.00% auto_monad : 0.000593s : 0.01% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000026s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000005s : 0.00% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000040s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000039s : 0.00% optimize.rewriter_before_opt_a : 0.000114s : 0.00% optimize.opt_a.expand_dump_flag : 0.000006s : 0.00% optimize.opt_a.switch_simplify : 0.000056s : 0.00% optimize.opt_a.loop_unroll : 0.000037s : 0.00% optimize.opt_a.a_1 : 0.001958s : 0.05% optimize.opt_a.with_stream_mark : 0.000051s : 0.00% optimize.opt_a.recompute_prepare : 0.000040s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000022s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000017s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000023s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000938s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000133s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.00% optimize.opt_a.shard_inline : 0.000043s : 0.00% optimize.opt_a.merge_send_recv : 0.000036s : 0.00% optimize.opt_a.auto_parallel : 0.000032s : 0.00% optimize.opt_a.parallel : 0.000401s : 0.01% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000018s : 0.00% optimize.opt_a.allreduce_fusion : 0.000018s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000035s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000043s : 0.00% optimize.opt_a.virtual_dataset : 0.000034s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000034s : 0.00% optimize.opt_a.virtual_output : 0.000033s : 0.00% optimize.opt_a.merge_forward : 0.000018s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000039s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000066s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000055s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000018s : 0.00% optimize.opt_a.meta_fg_expand : 0.000014s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000260s : 0.01% optimize.opt_a.a_after_grad : 0.000058s : 0.00% optimize.opt_a.renormalize : 0.215045s : 5.02% optimize.opt_a.add_forward_monad_depend : 0.000017s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000101s : 0.00% optimize.opt_a.cse : 0.000973s : 0.02% optimize.opt_a.a_3 : 0.000260s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000028s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.002395s : 0.06% optimize.convert_after_rewriter : 0.000048s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.001034s : 0.02% optimize.opt_b.b_1 : 0.000369s : 0.01% optimize.opt_b.b_2 : 0.000018s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000056s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000044s : 0.00% optimize.overlap_param_gather : 0.000027s : 0.00% optimize.cconv : 0.000037s : 0.00% optimize.loop_unroll : 0.000737s : 0.02% optimize.opt_after_cconv.c_1 : 0.000098s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.cse : 0.000044s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000068s : 0.00% optimize.tuple_transform.d_1 : 0.000112s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000017s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.001055s : 0.02% optimize.cse_after_recomputation.cse : 0.000791s : 0.02% optimize.environ_conv : 0.000018s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000013s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000010s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000028s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000007s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000002s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000008s : 0.00% optimize.overlap_grad_flash_sp : 0.000085s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000019s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000024s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000031s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000023s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000025s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000047s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000777s : 0.02% validate : 0.000122s : 0.00% Time group info: ------[substitution.] 0.000426 96 2.74% : 0.000012s : 2: substitution.depend_value_elim 1.00% : 0.000004s : 8: substitution.elim_not_effective 0.81% : 0.000003s : 8: substitution.fold_const_symbol 2.86% : 0.000012s : 13: substitution.graph_param_transform 31.26% : 0.000133s : 1: substitution.inline 2.18% : 0.000009s : 16: substitution.j_node_and_user_rematch 21.31% : 0.000091s : 2: substitution.less_batch_normalization 1.51% : 0.000006s : 6: substitution.load_eliminater 0.73% : 0.000003s : 2: substitution.opt_reshape 2.70% : 0.000012s : 16: substitution.remove_not_recompute_node 1.69% : 0.000007s : 4: substitution.replace_old_param 6.76% : 0.000029s : 4: substitution.reshape_eliminate 5.30% : 0.000023s : 6: substitution.updatestate_pure_node_eliminater 19.15% : 0.000082s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 3.944178 2 99.71% : 3.932651s : 1: type_inference.infer 0.29% : 0.011527s : 1: type_inference.specialize ------[replace.] 0.000020 1 100.00% : 0.000020s : 1: replace.inline ------[match.] 0.000132 1 100.00% : 0.000132s : 1: match.inline ------[predicate.] 0.000458 3002 0.78% : 0.000004s : 27: predicate.accumulaten_eliminater 0.90% : 0.000004s : 13: predicate.ad_related_special_op_eliminate 0.74% : 0.000003s : 26: predicate.addn_check_dump 0.82% : 0.000004s : 27: predicate.addn_zero_filter 0.74% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.23% : 0.000010s : 53: predicate.arithmetic_simplify 0.89% : 0.000004s : 27: predicate.cast_eliminate 0.79% : 0.000004s : 26: predicate.check_bprop_eliminate 0.77% : 0.000004s : 26: predicate.compare_switch_simplify 0.23% : 0.000001s : 13: predicate.const_output_eliminate 0.89% : 0.000004s : 26: predicate.depend_value_elim 0.80% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.93% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.78% : 0.000004s : 27: predicate.dict_set_item_eliminator 4.63% : 0.000021s : 26: predicate.dumpgradient_eliminate 0.26% : 0.000001s : 13: predicate.elim_not_effective 0.46% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.11% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.11% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.09% : 0.000005s : 40: predicate.environ_get_depend_swap 1.98% : 0.000009s : 66: predicate.environ_get_eliminate 1.08% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.77% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.30% : 0.000006s : 28: predicate.float_depend_g_call 0.75% : 0.000003s : 26: predicate.float_environ_get_switch 1.12% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.22% : 0.000001s : 13: predicate.fold_const_symbol 0.83% : 0.000004s : 26: predicate.get_grad_eliminate 0.31% : 0.000001s : 13: predicate.graph_param_transform 0.76% : 0.000003s : 26: predicate.incorporate_call 0.67% : 0.000003s : 26: predicate.incorporate_call_switch 5.43% : 0.000025s : 133: predicate.inline 0.99% : 0.000005s : 26: predicate.inline_without_move 0.42% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.03% : 0.000005s : 27: predicate.less_batch_normalization 1.64% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.24% : 0.000010s : 80: predicate.load_eliminater 0.79% : 0.000004s : 13: predicate.loop_unroll_after_grad 1.04% : 0.000005s : 31: predicate.loop_unroll_before_grad 1.68% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.79% : 0.000004s : 26: predicate.merge_addn 0.81% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.79% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.74% : 0.000003s : 27: predicate.minmaximum_grad 0.90% : 0.000004s : 13: predicate.mutable_eliminate 0.59% : 0.000003s : 13: predicate.opt_reshape 0.42% : 0.000002s : 13: predicate.parallel_virtual_node 1.03% : 0.000005s : 28: predicate.partial_defer_inline 1.23% : 0.000006s : 40: predicate.partial_eliminate 0.79% : 0.000004s : 27: predicate.print_const_string_wrapper 0.76% : 0.000003s : 26: predicate.reduce_all_const_elim 1.13% : 0.000005s : 27: predicate.reduce_eliminate 2.25% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.49% : 0.000002s : 26: predicate.remove_not_recompute_node 1.39% : 0.000006s : 53: predicate.replace_applicator 0.62% : 0.000003s : 26: predicate.replace_old_param 0.30% : 0.000001s : 13: predicate.reset_defer_inline 0.88% : 0.000004s : 27: predicate.reshape_eliminate 0.91% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.46% : 0.000002s : 13: predicate.row_tensor_eliminate 1.14% : 0.000005s : 26: predicate.same_eliminate 0.57% : 0.000003s : 26: predicate.set_cell_output_no_recompute 1.02% : 0.000005s : 26: predicate.shard_identity_eliminate 0.85% : 0.000004s : 26: predicate.special_op_eliminate 0.85% : 0.000004s : 26: predicate.specialize_transform 1.23% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 1.00% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.44% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.83% : 0.000004s : 28: predicate.switch_defer_inline 1.66% : 0.000008s : 54: predicate.switch_layer_defer_inline 3.24% : 0.000015s : 98: predicate.switch_simplify 0.82% : 0.000004s : 27: predicate.tile_eliminate 0.78% : 0.000004s : 27: predicate.transpose_eliminate 1.66% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 1.79% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.51% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.90% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.76% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.54% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.56% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.27% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.33% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.44% : 0.000002s : 13: predicate.value_based_eliminate 0.88% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.82% : 0.000004s : 26: predicate.virtual_output_eliminate 0.41% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.45% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.607716 58 66.76% : 0.405702s : 55: func_graph_cloner_run.FuncGraphClonerGraph 33.24% : 0.202015s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 5.564992 192 0.00% : 0.000003s : 1: ForceFp32Comm 7.43% : 0.413608s : 1: add_attr 7.43% : 0.413513s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.001067s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000605s : 1: auto_monad 0.00% : 0.000051s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 1.92% : 0.106895s : 1: bootstrap 0.00% : 0.000041s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000031s : 1: control_data_broadcast_order 0.00% : 0.000053s : 1: convert_after_rewriter 0.01% : 0.000812s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000023s : 1: environ_conv 0.00% : 0.000028s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000012s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000008s : 1: insert-virtual-dataset 0.00% : 0.000005s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.01% : 0.000747s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000007s : 1: micro_interleaved_order_control 0.02% : 0.001045s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000027s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000031s : 1: opt.transform.mutable_eliminate 0.07% : 0.003923s : 78: opt.transform.opt_a 0.00% : 0.000097s : 1: opt.transform.opt_after_cconv 0.00% : 0.000061s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000355s : 28: opt.transform.opt_b 0.00% : 0.000127s : 2: opt.transform.opt_trans_graph 0.00% : 0.000098s : 4: opt.transform.symbol_engine_opt 4.03% : 0.224503s : 1: opt_a 0.00% : 0.000220s : 1: opt_after_cconv 0.01% : 0.000790s : 1: opt_after_jit_grad 0.01% : 0.000523s : 1: opt_b 4.18% : 0.232625s : 1: optimize 0.00% : 0.000048s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000015s : 1: order_py_execute_after_rewriter 0.00% : 0.000088s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000013s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000031s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000011s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000006s : 1: pipeline_split 0.00% : 0.000044s : 1: pre_auto_parallel 0.00% : 0.000044s : 1: py_interpret_to_execute 0.00% : 0.000032s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000073s : 1: remove_dup_value 0.23% : 0.012866s : 1: renormalize.infer 3.63% : 0.202159s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.04% : 0.002408s : 1: rewriter_after_opt_a 0.00% : 0.000119s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000006s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000017s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000166s : 1: symbol_engine_optimizer 0.00% : 0.000161s : 1: tuple_transform 70.88% : 3.944579s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape2-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape2-1],max_mem:196.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:40:48.995.859 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 3.19302, [21] [bootstrap]: 0.00152157 [type_inference]: 2.89076 [event_method]: 1.964e-05 [auto_monad]: 0.00037317 [graph_reusing]: 7.41999e-06 [inline]: 3.14999e-06 [add_attr]: 0.142976, [1] [add_attr_with_inline]: 0.142959, [1] [Cycle 1]: 7.525e-05, [2] [tag_attr]: 2.333e-05 [meta_addattr_fg_expand]: 4.23001e-06 [parallel-infer-symbol]: 5.18002e-06 [pre_auto_parallel]: 3.577e-05 [insert-virtual-dataset]: 2.80002e-06 [parallel-infer-symbol-second]: 7.50006e-07 [dataset_repeat_opt]: 1.96e-06 [pipeline_split]: 1.77001e-06 [optimize]: 0.155606, [53] [py_interpret_to_execute]: 3.505e-05 [rewriter_before_opt_a]: 9.022e-05 [opt_a]: 0.0382066, [2] [Cycle 1]: 0.0363567, [45] [expand_dump_flag]: 2.69001e-06 [switch_simplify]: 3.347e-05 [loop_unroll]: 1.974e-05 [a_1]: 0.0211565 [with_stream_mark]: 4.886e-05 [recompute_prepare]: 3.325e-05 [updatestate_depend_eliminate]: 1.346e-05 [updatestate_assign_eliminate]: 9.60001e-06 [updatestate_loads_eliminate]: 1.125e-05 [parameter_eliminate]: 3.16001e-06 [a_2]: 0.0002778 [accelerated_algorithm]: 5.18e-05 [shard]: 3.6e-06 [meta_shard_fg_expand]: 6.81999e-06 [shard_inline]: 1.931e-05 [merge_send_recv]: 1.661e-05 [auto_parallel]: 1.636e-05 [parallel]: 4.727e-05 [flash_sp]: 1.721e-05 [merge_comm]: 1.072e-05 [allreduce_fusion]: 8.35999e-06 [matmul_add_comm_reduction]: 1.842e-05 [allreduce_slice_to_reducescatter]: 8.59989e-07 [virtual_shard_identity]: 2.375e-05 [virtual_dataset]: 1.936e-05 [get_grad_eliminate_]: 2.09e-05 [virtual_output]: 1.963e-05 [merge_forward]: 1.096e-05 [cell_reuse_recompute_pass]: 3.39001e-06 [offload_activation]: 1.87e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.115e-05 [merge_recompute_call_nodes]: 2.16998e-06 [before_grad]: 3.147e-05 [set_forward_comm_id_for_comm_node_pass]: 1.055e-05 [meta_fg_expand]: 7.45998e-06 [flash_sp_send_recv_attached]: 8.94e-06 [receive_attached]: 2.39001e-06 [after_resolve]: 3.076e-05 [a_after_grad]: 3.329e-05 [renormalize]: 0.0123676 [add_forward_monad_depend]: 1.411e-05 [auto_monad_grad]: 2.81999e-06 [auto_monad_eliminator]: 6.52e-05 [cse]: 0.00120878 [a_3]: 0.00014632 [Cycle 2]: 0.00183176, [45] [expand_dump_flag]: 3.5e-06 [switch_simplify]: 2.051e-05 [loop_unroll]: 1.67e-05 [a_1]: 0.00054489 [with_stream_mark]: 3.495e-05 [recompute_prepare]: 2e-05 [updatestate_depend_eliminate]: 1.11e-05 [updatestate_assign_eliminate]: 7.89002e-06 [updatestate_loads_eliminate]: 1.126e-05 [parameter_eliminate]: 2.58998e-06 [a_2]: 0.00022903 [accelerated_algorithm]: 2.367e-05 [shard]: 3.25998e-06 [meta_shard_fg_expand]: 5.66e-06 [shard_inline]: 1.624e-05 [merge_send_recv]: 1.484e-05 [auto_parallel]: 1.509e-05 [parallel]: 1.113e-05 [flash_sp]: 3.87002e-06 [merge_comm]: 8.53001e-06 [allreduce_fusion]: 9.04e-06 [matmul_add_comm_reduction]: 1.828e-05 [allreduce_slice_to_reducescatter]: 8.99978e-07 [virtual_shard_identity]: 1.98e-05 [virtual_dataset]: 1.626e-05 [get_grad_eliminate_]: 1.611e-05 [virtual_output]: 1.609e-05 [merge_forward]: 9.94001e-06 [cell_reuse_recompute_pass]: 3.03e-06 [offload_activation]: 1.888e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.191e-05 [merge_recompute_call_nodes]: 1.72001e-06 [before_grad]: 2.665e-05 [set_forward_comm_id_for_comm_node_pass]: 1.069e-05 [meta_fg_expand]: 7.16001e-06 [flash_sp_send_recv_attached]: 1.83002e-06 [receive_attached]: 2.71999e-06 [after_resolve]: 2.484e-05 [a_after_grad]: 2.728e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 4.57e-06 [auto_monad_grad]: 2.34999e-06 [auto_monad_eliminator]: 5.772e-05 [cse]: 0.00011268 [a_3]: 0.00010949 [py_interpret_to_execute_after_opt_a]: 3.171e-05 [slice_cell_reuse_recomputed_activation]: 2.19001e-06 [rewriter_after_opt_a]: 0.00022574 [convert_after_rewriter]: 1.799e-05 [order_py_execute_after_rewriter]: 1.111e-05 [mutable_eliminate]: 0.00083765 [opt_b]: 0.00055844, [1] [Cycle 1]: 0.0005488, [7] [b_1]: 0.0003713 [b_2]: 2.048e-05 [updatestate_depend_eliminate]: 1.492e-05 [updatestate_assign_eliminate]: 7.81001e-06 [updatestate_loads_eliminate]: 1.196e-05 [renormalize]: 1.32e-06 [cse]: 7.284e-05 [optimize_parallel_all_gather_comm]: 3.472e-05 [overlap_param_gather]: 4.63001e-06 [cconv]: 3.98e-05 [loop_unroll]: 0.00062229 [opt_after_cconv]: 0.00029738, [1] [Cycle 1]: 0.00028556, [7] [c_1]: 0.00011463 [parameter_eliminate]: 4.97e-06 [updatestate_depend_eliminate]: 1.728e-05 [updatestate_assign_eliminate]: 8.18999e-06 [updatestate_loads_eliminate]: 1.215e-05 [cse]: 5.962e-05 [renormalize]: 1.19e-06 [remove_dup_value]: 7.814e-05 [tuple_transform]: 0.0001827, [1] [Cycle 1]: 0.00017625, [4] [d_1]: 0.0001205 [none_parameter_eliminate]: 3.01999e-06 [renormalize]: 1.40019e-07 [switch_simplify]: 2.271e-05 [partial_unused_args_eliminate]: 2.39999e-06 [add_recomputation]: 0.113478 [cse_after_recomputation]: 0.00011279, [1] [Cycle 1]: 0.0001015, [1] [cse]: 8.482e-05 [environ_conv]: 2.306e-05 [swap_dp_allreduce_reducescatter]: 1.289e-05 [bias_add_comm_swap]: 4.18001e-06 [label_micro_interleaved_index]: 7.14001e-06 [label_fine_grained_interleaved_index]: 2.65002e-06 [merge_cast_opt]: 1.40999e-06 [slice_recompute_activation]: 2.19999e-06 [micro_interleaved_order_control]: 2.86e-06 [assign_add_opt]: 1.61002e-06 [ForceFp32Comm]: 1.32e-06 [remove_cast_before_assign_add]: 1.17e-06 [full_micro_interleaved_order_control]: 2.14e-06 [reorder_send_recv_between_fp_bp]: 2.97002e-06 [comm_op_add_attrs]: 1.37e-06 [add_comm_op_reuse_tag]: 9.39996e-07 [interleave_split_concat_branches]: 1.19998e-06 [interleave_parallel_branches]: 1.35999e-06 [overlap_opt_shard_in_pipeline]: 2.79001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.82999e-06 [control_data_broadcast_order]: 3.143e-05 [grouped_pairwise_exchange_alltoall]: 1.76998e-06 [offloading_packed_experts]: 7.50998e-06 [overlap_recompute_and_grad_model_parallel]: 7.57002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.50999e-06 [overlap_recompute_comm]: 2.24001e-06 [overlap_grad_ring_attention]: 7.91001e-06 [overlap_grad_flash_sp]: 5.765e-05 [begin_end_overlap_inline]: 5.59987e-07 [split_matmul_comm_elemetwise]: 2.27999e-06 [split_layernorm_comm]: 1.87999e-06 [handle_group_info]: 1.35999e-06 [symbol_engine_optimizer]: 0.00019867, [1] [Cycle 1]: 0.00019074, [6] [build]: 2.01e-05 [elim_shapecalc]: 4.149e-05 [elim_not_effective]: 3.933e-05 [opt_reshape]: 2.803e-05 [fold_const_symbol]: 2.533e-05 [renormalize]: 4.2998e-07 [detach_backward]: 2.31e-06 [pipeline_parallel_scheduler]: 1.34e-06 [auto_monad_reorder]: 5.349e-05 [get_jit_bprop_graph]: 2.25002e-06 [rewriter_after_jit_bprop_graph]: 8.15999e-06 [opt_after_jit_grad]: 0.0009017 [validate]: 0.00010397 Sums bootstrap : 0.001522s : 0.05% type_inference : 2.890764s : 94.84% event_method : 0.000020s : 0.00% auto_monad : 0.000373s : 0.01% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000023s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000004s : 0.00% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000036s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000035s : 0.00% optimize.rewriter_before_opt_a : 0.000090s : 0.00% optimize.opt_a.expand_dump_flag : 0.000006s : 0.00% optimize.opt_a.switch_simplify : 0.000054s : 0.00% optimize.opt_a.loop_unroll : 0.000036s : 0.00% optimize.opt_a.a_1 : 0.021701s : 0.71% optimize.opt_a.with_stream_mark : 0.000084s : 0.00% optimize.opt_a.recompute_prepare : 0.000053s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000025s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000017s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000023s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000507s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.00% optimize.opt_a.shard : 0.000007s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000012s : 0.00% optimize.opt_a.shard_inline : 0.000036s : 0.00% optimize.opt_a.merge_send_recv : 0.000031s : 0.00% optimize.opt_a.auto_parallel : 0.000031s : 0.00% optimize.opt_a.parallel : 0.000058s : 0.00% optimize.opt_a.flash_sp : 0.000021s : 0.00% optimize.opt_a.merge_comm : 0.000019s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000037s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000044s : 0.00% optimize.opt_a.virtual_dataset : 0.000036s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000037s : 0.00% optimize.opt_a.virtual_output : 0.000036s : 0.00% optimize.opt_a.merge_forward : 0.000021s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000038s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000073s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000058s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000021s : 0.00% optimize.opt_a.meta_fg_expand : 0.000015s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000011s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000056s : 0.00% optimize.opt_a.a_after_grad : 0.000061s : 0.00% optimize.opt_a.renormalize : 0.012368s : 0.41% optimize.opt_a.add_forward_monad_depend : 0.000019s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000123s : 0.00% optimize.opt_a.cse : 0.001321s : 0.04% optimize.opt_a.a_3 : 0.000256s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000032s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000226s : 0.01% optimize.convert_after_rewriter : 0.000018s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.000838s : 0.03% optimize.opt_b.b_1 : 0.000371s : 0.01% optimize.opt_b.b_2 : 0.000020s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000073s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000035s : 0.00% optimize.overlap_param_gather : 0.000005s : 0.00% optimize.cconv : 0.000040s : 0.00% optimize.loop_unroll : 0.000622s : 0.02% optimize.opt_after_cconv.c_1 : 0.000115s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000017s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.cse : 0.000060s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000078s : 0.00% optimize.tuple_transform.d_1 : 0.000121s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000023s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.113478s : 3.72% optimize.cse_after_recomputation.cse : 0.000085s : 0.00% optimize.environ_conv : 0.000023s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000013s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000007s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000003s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000031s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000008s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000008s : 0.00% optimize.overlap_grad_flash_sp : 0.000058s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000020s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000041s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000039s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000028s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000025s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000053s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000902s : 0.03% validate : 0.000104s : 0.00% Time group info: ------[substitution.] 0.000326 96 3.95% : 0.000013s : 2: substitution.depend_value_elim 1.35% : 0.000004s : 8: substitution.elim_not_effective 0.97% : 0.000003s : 8: substitution.fold_const_symbol 4.37% : 0.000014s : 13: substitution.graph_param_transform 39.42% : 0.000128s : 1: substitution.inline 2.92% : 0.000010s : 16: substitution.j_node_and_user_rematch 7.99% : 0.000026s : 2: substitution.less_batch_normalization 3.09% : 0.000010s : 6: substitution.load_eliminater 1.05% : 0.000003s : 2: substitution.opt_reshape 4.00% : 0.000013s : 16: substitution.remove_not_recompute_node 2.67% : 0.000009s : 4: substitution.replace_old_param 13.42% : 0.000044s : 4: substitution.reshape_eliminate 8.30% : 0.000027s : 6: substitution.updatestate_pure_node_eliminater 6.51% : 0.000021s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.890569 2 99.82% : 2.885473s : 1: type_inference.infer 0.18% : 0.005096s : 1: type_inference.specialize ------[replace.] 0.000019 1 100.00% : 0.000019s : 1: replace.inline ------[match.] 0.000127 1 100.00% : 0.000127s : 1: match.inline ------[predicate.] 0.020682 3002 0.03% : 0.000006s : 27: predicate.accumulaten_eliminater 0.02% : 0.000005s : 13: predicate.ad_related_special_op_eliminate 0.02% : 0.000004s : 26: predicate.addn_check_dump 0.02% : 0.000004s : 27: predicate.addn_zero_filter 97.28% : 0.020120s : 27: predicate.adjust_all_reduce_mul_add 0.06% : 0.000011s : 53: predicate.arithmetic_simplify 0.02% : 0.000004s : 27: predicate.cast_eliminate 0.02% : 0.000004s : 26: predicate.check_bprop_eliminate 0.02% : 0.000003s : 26: predicate.compare_switch_simplify 0.01% : 0.000001s : 13: predicate.const_output_eliminate 0.02% : 0.000004s : 26: predicate.depend_value_elim 0.02% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.02% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.02% : 0.000004s : 27: predicate.dict_set_item_eliminator 0.04% : 0.000009s : 26: predicate.dumpgradient_eliminate 0.01% : 0.000002s : 13: predicate.elim_not_effective 0.01% : 0.000003s : 13: predicate.elim_shapecalc_of_broadcastargs 0.03% : 0.000007s : 40: predicate.environ_add_const_eliminate 0.02% : 0.000005s : 40: predicate.environ_get_add_eliminate 0.03% : 0.000005s : 40: predicate.environ_get_depend_swap 0.04% : 0.000009s : 66: predicate.environ_get_eliminate 0.02% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.02% : 0.000004s : 28: predicate.exchange_switch_depend_value 0.03% : 0.000007s : 28: predicate.float_depend_g_call 0.02% : 0.000003s : 26: predicate.float_environ_get_switch 0.03% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.01% : 0.000001s : 13: predicate.fold_const_symbol 0.02% : 0.000004s : 26: predicate.get_grad_eliminate 0.01% : 0.000001s : 13: predicate.graph_param_transform 0.02% : 0.000004s : 26: predicate.incorporate_call 0.02% : 0.000003s : 26: predicate.incorporate_call_switch 0.13% : 0.000027s : 133: predicate.inline 0.03% : 0.000006s : 26: predicate.inline_without_move 0.01% : 0.000002s : 26: predicate.j_node_and_user_rematch 0.03% : 0.000007s : 27: predicate.less_batch_normalization 0.04% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 0.05% : 0.000011s : 80: predicate.load_eliminater 0.03% : 0.000006s : 13: predicate.loop_unroll_after_grad 0.02% : 0.000005s : 31: predicate.loop_unroll_before_grad 0.04% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.02% : 0.000004s : 26: predicate.merge_addn 0.02% : 0.000003s : 26: predicate.micro_step_allgather_replace 0.02% : 0.000003s : 26: predicate.mini_step_allgather_replace 0.02% : 0.000004s : 27: predicate.minmaximum_grad 0.03% : 0.000006s : 13: predicate.mutable_eliminate 0.01% : 0.000003s : 13: predicate.opt_reshape 0.01% : 0.000002s : 13: predicate.parallel_virtual_node 0.05% : 0.000010s : 28: predicate.partial_defer_inline 0.03% : 0.000006s : 40: predicate.partial_eliminate 0.02% : 0.000004s : 27: predicate.print_const_string_wrapper 0.02% : 0.000004s : 26: predicate.reduce_all_const_elim 0.03% : 0.000005s : 27: predicate.reduce_eliminate 0.39% : 0.000081s : 80: predicate.redundant_stop_gradient_eliminater 0.01% : 0.000002s : 26: predicate.remove_not_recompute_node 0.03% : 0.000007s : 53: predicate.replace_applicator 0.01% : 0.000003s : 26: predicate.replace_old_param 0.01% : 0.000002s : 13: predicate.reset_defer_inline 0.02% : 0.000005s : 27: predicate.reshape_eliminate 0.02% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.01% : 0.000003s : 13: predicate.row_tensor_eliminate 0.03% : 0.000007s : 26: predicate.same_eliminate 0.01% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.03% : 0.000005s : 26: predicate.shard_identity_eliminate 0.02% : 0.000004s : 26: predicate.special_op_eliminate 0.02% : 0.000004s : 26: predicate.specialize_transform 0.03% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.02% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.01% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.02% : 0.000005s : 28: predicate.switch_defer_inline 0.04% : 0.000008s : 54: predicate.switch_layer_defer_inline 0.09% : 0.000019s : 98: predicate.switch_simplify 0.02% : 0.000005s : 27: predicate.tile_eliminate 0.02% : 0.000004s : 27: predicate.transpose_eliminate 0.04% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 0.04% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 0.04% : 0.000008s : 53: predicate.tuple_list_get_item_depend_reorder 0.08% : 0.000016s : 79: predicate.tuple_list_get_item_eliminator 0.04% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 0.06% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 0.04% : 0.000008s : 53: predicate.tuple_to_list_eliminator_ 0.05% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 0.08% : 0.000016s : 106: predicate.updatestate_useless_node_eliminater 0.01% : 0.000002s : 13: predicate.value_based_eliminate 0.02% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.02% : 0.000004s : 26: predicate.virtual_output_eliminate 0.01% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.01% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.018359 58 78.36% : 0.014386s : 55: func_graph_cloner_run.FuncGraphClonerGraph 21.64% : 0.003973s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 3.527428 192 0.00% : 0.000004s : 1: ForceFp32Comm 4.05% : 0.142983s : 1: add_attr 4.05% : 0.142963s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 3.22% : 0.113506s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000386s : 1: auto_monad 0.00% : 0.000058s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.05% : 0.001738s : 1: bootstrap 0.00% : 0.000044s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000035s : 1: control_data_broadcast_order 0.00% : 0.000023s : 1: convert_after_rewriter 0.00% : 0.000119s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000028s : 1: environ_conv 0.00% : 0.000030s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000031s : 1: get_jit_bprop_graph 0.00% : 0.000013s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000008s : 1: insert-virtual-dataset 0.00% : 0.000005s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.02% : 0.000637s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.02% : 0.000850s : 1: mutable_eliminate 0.00% : 0.000011s : 1: offloading_packed_experts 0.00% : 0.000039s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000038s : 1: opt.transform.mutable_eliminate 0.65% : 0.022979s : 78: opt.transform.opt_a 0.00% : 0.000113s : 1: opt.transform.opt_after_cconv 0.00% : 0.000063s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000357s : 28: opt.transform.opt_b 0.00% : 0.000139s : 2: opt.transform.opt_trans_graph 0.00% : 0.000127s : 4: opt.transform.symbol_engine_opt 1.08% : 0.038211s : 1: opt_a 0.01% : 0.000302s : 1: opt_after_cconv 0.03% : 0.000913s : 1: opt_after_jit_grad 0.02% : 0.000563s : 1: opt_b 4.41% : 0.155613s : 1: optimize 0.00% : 0.000039s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000015s : 1: order_py_execute_after_rewriter 0.00% : 0.000063s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000006s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000009s : 1: overlap_param_gather 0.00% : 0.000006s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000006s : 1: pipeline_split 0.00% : 0.000040s : 1: pre_auto_parallel 0.00% : 0.000039s : 1: py_interpret_to_execute 0.00% : 0.000037s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000085s : 1: remove_dup_value 0.22% : 0.007809s : 1: renormalize.infer 0.13% : 0.004538s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000012s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000236s : 1: rewriter_after_opt_a 0.00% : 0.000096s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000006s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000016s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000201s : 1: symbol_engine_optimizer 0.01% : 0.000186s : 1: tuple_transform 81.95% : 2.890885s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape2-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape2-16],max_mem:196.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:41:53.688.658 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 1.36628, [21] [bootstrap]: 0.00078318 [type_inference]: 1.24012 [event_method]: 2.132e-05 [auto_monad]: 0.00047136 [graph_reusing]: 7.44002e-06 [inline]: 4.98001e-06 [add_attr]: 0.00863887, [1] [add_attr_with_inline]: 0.00862128, [1] [Cycle 1]: 7.457e-05, [2] [tag_attr]: 2.275e-05 [meta_addattr_fg_expand]: 4.4e-06 [parallel-infer-symbol]: 5.12e-06 [pre_auto_parallel]: 3.711e-05 [insert-virtual-dataset]: 2.99001e-06 [parallel-infer-symbol-second]: 7.2e-07 [dataset_repeat_opt]: 2.09999e-06 [pipeline_split]: 1.64e-06 [optimize]: 0.114953, [53] [py_interpret_to_execute]: 3.481e-05 [rewriter_before_opt_a]: 0.00014417 [opt_a]: 0.0713713, [2] [Cycle 1]: 0.0696287, [45] [expand_dump_flag]: 2.78e-06 [switch_simplify]: 3.417e-05 [loop_unroll]: 2.008e-05 [a_1]: 0.00089621 [with_stream_mark]: 2.779e-05 [recompute_prepare]: 2.278e-05 [updatestate_depend_eliminate]: 1.18e-05 [updatestate_assign_eliminate]: 8.33001e-06 [updatestate_loads_eliminate]: 1.162e-05 [parameter_eliminate]: 2.19001e-06 [a_2]: 0.00025646 [accelerated_algorithm]: 4.433e-05 [shard]: 3.02002e-06 [meta_shard_fg_expand]: 3.55e-06 [shard_inline]: 1.693e-05 [merge_send_recv]: 1.379e-05 [auto_parallel]: 1.277e-05 [parallel]: 5.393e-05 [flash_sp]: 1.32e-05 [merge_comm]: 9.69999e-06 [allreduce_fusion]: 8.05e-06 [matmul_add_comm_reduction]: 2.036e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 2.287e-05 [virtual_dataset]: 1.772e-05 [get_grad_eliminate_]: 2.046e-05 [virtual_output]: 1.825e-05 [merge_forward]: 9.69999e-06 [cell_reuse_recompute_pass]: 1.60999e-06 [offload_activation]: 2.144e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.504e-05 [merge_recompute_call_nodes]: 1.75001e-06 [before_grad]: 2.929e-05 [set_forward_comm_id_for_comm_node_pass]: 9.20999e-06 [meta_fg_expand]: 6.55002e-06 [flash_sp_send_recv_attached]: 6.53e-06 [receive_attached]: 2.46998e-06 [after_resolve]: 2.484e-05 [a_after_grad]: 2.958e-05 [renormalize]: 0.0658945 [add_forward_monad_depend]: 1.277e-05 [auto_monad_grad]: 2.80997e-06 [auto_monad_eliminator]: 6.53e-05 [cse]: 0.00133561 [a_3]: 0.00014987 [Cycle 2]: 0.0017258, [45] [expand_dump_flag]: 4.47998e-06 [switch_simplify]: 1.947e-05 [loop_unroll]: 1.794e-05 [a_1]: 0.0004688 [with_stream_mark]: 3.402e-05 [recompute_prepare]: 1.847e-05 [updatestate_depend_eliminate]: 1.129e-05 [updatestate_assign_eliminate]: 9.00999e-06 [updatestate_loads_eliminate]: 1.091e-05 [parameter_eliminate]: 2.62001e-06 [a_2]: 0.00022019 [accelerated_algorithm]: 2.243e-05 [shard]: 2.34999e-06 [meta_shard_fg_expand]: 5.62999e-06 [shard_inline]: 1.635e-05 [merge_send_recv]: 1.499e-05 [auto_parallel]: 1.53e-05 [parallel]: 1.02e-05 [flash_sp]: 4.50001e-06 [merge_comm]: 9.04e-06 [allreduce_fusion]: 9.39998e-06 [matmul_add_comm_reduction]: 1.861e-05 [allreduce_slice_to_reducescatter]: 8.70001e-07 [virtual_shard_identity]: 1.911e-05 [virtual_dataset]: 1.616e-05 [get_grad_eliminate_]: 1.615e-05 [virtual_output]: 1.608e-05 [merge_forward]: 9.34998e-06 [cell_reuse_recompute_pass]: 4.38001e-06 [offload_activation]: 1.907e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.953e-05 [merge_recompute_call_nodes]: 1.72001e-06 [before_grad]: 2.576e-05 [set_forward_comm_id_for_comm_node_pass]: 9.58002e-06 [meta_fg_expand]: 7.05e-06 [flash_sp_send_recv_attached]: 2.19001e-06 [receive_attached]: 3.20002e-06 [after_resolve]: 2.398e-05 [a_after_grad]: 2.742e-05 [renormalize]: 1.50001e-07 [add_forward_monad_depend]: 3.4e-06 [auto_monad_grad]: 2.87002e-06 [auto_monad_eliminator]: 5.614e-05 [cse]: 0.00012798 [a_3]: 0.00010938 [py_interpret_to_execute_after_opt_a]: 2.994e-05 [slice_cell_reuse_recomputed_activation]: 2.67001e-06 [rewriter_after_opt_a]: 0.00022032 [convert_after_rewriter]: 1.595e-05 [order_py_execute_after_rewriter]: 1.075e-05 [mutable_eliminate]: 0.0402865 [opt_b]: 0.00066039, [1] [Cycle 1]: 0.00064642, [7] [b_1]: 0.00041722 [b_2]: 2.241e-05 [updatestate_depend_eliminate]: 2.372e-05 [updatestate_assign_eliminate]: 8.65001e-06 [updatestate_loads_eliminate]: 1.245e-05 [renormalize]: 1.17999e-06 [cse]: 9.907e-05 [optimize_parallel_all_gather_comm]: 4.26e-05 [overlap_param_gather]: 3.76999e-06 [cconv]: 4.377e-05 [loop_unroll]: 0.00074788 [opt_after_cconv]: 0.00026232, [1] [Cycle 1]: 0.0002525, [7] [c_1]: 0.00011224 [parameter_eliminate]: 7.27997e-06 [updatestate_depend_eliminate]: 1.655e-05 [updatestate_assign_eliminate]: 7.08998e-06 [updatestate_loads_eliminate]: 1.059e-05 [cse]: 5.972e-05 [renormalize]: 3.69997e-07 [remove_dup_value]: 7.649e-05 [tuple_transform]: 0.0001696, [1] [Cycle 1]: 0.00016421, [4] [d_1]: 0.00011935 [none_parameter_eliminate]: 2.86e-06 [renormalize]: 3.00002e-07 [switch_simplify]: 1.801e-05 [partial_unused_args_eliminate]: 2.24001e-06 [add_recomputation]: 0.00012476 [cse_after_recomputation]: 6.1e-05, [1] [Cycle 1]: 5.503e-05, [1] [cse]: 4.85e-05 [environ_conv]: 2.2e-05 [swap_dp_allreduce_reducescatter]: 1.239e-05 [bias_add_comm_swap]: 3.56001e-06 [label_micro_interleaved_index]: 7.87998e-06 [label_fine_grained_interleaved_index]: 2.62001e-06 [merge_cast_opt]: 1.77001e-06 [slice_recompute_activation]: 2.17999e-06 [micro_interleaved_order_control]: 2.77002e-06 [assign_add_opt]: 1.62999e-06 [ForceFp32Comm]: 9.09989e-07 [remove_cast_before_assign_add]: 1.29e-06 [full_micro_interleaved_order_control]: 2.78e-06 [reorder_send_recv_between_fp_bp]: 2.44999e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 1.07e-06 [interleave_split_concat_branches]: 1.20999e-06 [interleave_parallel_branches]: 1.11997e-06 [overlap_opt_shard_in_pipeline]: 3.86999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.85001e-06 [control_data_broadcast_order]: 2.795e-05 [grouped_pairwise_exchange_alltoall]: 1.68002e-06 [offloading_packed_experts]: 7.31999e-06 [overlap_recompute_and_grad_model_parallel]: 7.9e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.22e-06 [overlap_recompute_allgather_and_fa_grad]: 1.44e-06 [overlap_recompute_comm]: 2.39999e-06 [overlap_grad_ring_attention]: 8.08001e-06 [overlap_grad_flash_sp]: 4.198e-05 [begin_end_overlap_inline]: 5.69999e-07 [split_matmul_comm_elemetwise]: 2.27999e-06 [split_layernorm_comm]: 1.71998e-06 [handle_group_info]: 1.37999e-06 [symbol_engine_optimizer]: 0.00016101, [1] [Cycle 1]: 0.00015494, [6] [build]: 1.794e-05 [elim_shapecalc]: 2.357e-05 [elim_not_effective]: 3.106e-05 [opt_reshape]: 2.351e-05 [fold_const_symbol]: 2.468e-05 [renormalize]: 2.30008e-07 [detach_backward]: 2.70002e-06 [pipeline_parallel_scheduler]: 1.95001e-06 [auto_monad_reorder]: 5.023e-05 [get_jit_bprop_graph]: 2.06998e-06 [rewriter_after_jit_bprop_graph]: 7.94002e-06 [opt_after_jit_grad]: 0.00075875 [validate]: 0.00011165 Sums bootstrap : 0.000783s : 0.06% type_inference : 1.240121s : 91.44% event_method : 0.000021s : 0.00% auto_monad : 0.000471s : 0.03% graph_reusing : 0.000007s : 0.00% inline : 0.000005s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000023s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000004s : 0.00% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000037s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000035s : 0.00% optimize.rewriter_before_opt_a : 0.000144s : 0.01% optimize.opt_a.expand_dump_flag : 0.000007s : 0.00% optimize.opt_a.switch_simplify : 0.000054s : 0.00% optimize.opt_a.loop_unroll : 0.000038s : 0.00% optimize.opt_a.a_1 : 0.001365s : 0.10% optimize.opt_a.with_stream_mark : 0.000062s : 0.00% optimize.opt_a.recompute_prepare : 0.000041s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000023s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000017s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000023s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000477s : 0.04% optimize.opt_a.accelerated_algorithm : 0.000067s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000009s : 0.00% optimize.opt_a.shard_inline : 0.000033s : 0.00% optimize.opt_a.merge_send_recv : 0.000029s : 0.00% optimize.opt_a.auto_parallel : 0.000028s : 0.00% optimize.opt_a.parallel : 0.000064s : 0.00% optimize.opt_a.flash_sp : 0.000018s : 0.00% optimize.opt_a.merge_comm : 0.000019s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000039s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000042s : 0.00% optimize.opt_a.virtual_dataset : 0.000034s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000037s : 0.00% optimize.opt_a.virtual_output : 0.000034s : 0.00% optimize.opt_a.merge_forward : 0.000019s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000041s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000065s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000055s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000019s : 0.00% optimize.opt_a.meta_fg_expand : 0.000014s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000049s : 0.00% optimize.opt_a.a_after_grad : 0.000057s : 0.00% optimize.opt_a.renormalize : 0.065895s : 4.86% optimize.opt_a.add_forward_monad_depend : 0.000016s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000121s : 0.01% optimize.opt_a.cse : 0.001464s : 0.11% optimize.opt_a.a_3 : 0.000259s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000030s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000220s : 0.02% optimize.convert_after_rewriter : 0.000016s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.040286s : 2.97% optimize.opt_b.b_1 : 0.000417s : 0.03% optimize.opt_b.b_2 : 0.000022s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000024s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000099s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000043s : 0.00% optimize.overlap_param_gather : 0.000004s : 0.00% optimize.cconv : 0.000044s : 0.00% optimize.loop_unroll : 0.000748s : 0.06% optimize.opt_after_cconv.c_1 : 0.000112s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000017s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.cse : 0.000060s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000076s : 0.01% optimize.tuple_transform.d_1 : 0.000119s : 0.01% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000018s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000125s : 0.01% optimize.cse_after_recomputation.cse : 0.000049s : 0.00% optimize.environ_conv : 0.000022s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000012s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000008s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 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.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000028s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000007s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000008s : 0.00% optimize.overlap_grad_flash_sp : 0.000042s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000018s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000024s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000031s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000024s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000025s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000050s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000759s : 0.06% validate : 0.000112s : 0.01% Time group info: ------[substitution.] 0.000306 96 4.62% : 0.000014s : 2: substitution.depend_value_elim 1.32% : 0.000004s : 8: substitution.elim_not_effective 1.11% : 0.000003s : 8: substitution.fold_const_symbol 4.07% : 0.000012s : 13: substitution.graph_param_transform 42.51% : 0.000130s : 1: substitution.inline 2.89% : 0.000009s : 16: substitution.j_node_and_user_rematch 7.00% : 0.000021s : 2: substitution.less_batch_normalization 2.48% : 0.000008s : 6: substitution.load_eliminater 1.06% : 0.000003s : 2: substitution.opt_reshape 3.91% : 0.000012s : 16: substitution.remove_not_recompute_node 2.61% : 0.000008s : 4: substitution.replace_old_param 9.63% : 0.000030s : 4: substitution.reshape_eliminate 9.70% : 0.000030s : 6: substitution.updatestate_pure_node_eliminater 7.09% : 0.000022s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 1.240003 2 94.40% : 1.170514s : 1: type_inference.infer 5.60% : 0.069489s : 1: type_inference.specialize ------[replace.] 0.000020 1 100.00% : 0.000020s : 1: replace.inline ------[match.] 0.000129 1 100.00% : 0.000129s : 1: match.inline ------[predicate.] 0.000467 3002 0.84% : 0.000004s : 27: predicate.accumulaten_eliminater 0.95% : 0.000004s : 13: predicate.ad_related_special_op_eliminate 0.81% : 0.000004s : 26: predicate.addn_check_dump 0.82% : 0.000004s : 27: predicate.addn_zero_filter 0.73% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.00% : 0.000009s : 53: predicate.arithmetic_simplify 0.95% : 0.000004s : 27: predicate.cast_eliminate 0.79% : 0.000004s : 26: predicate.check_bprop_eliminate 0.71% : 0.000003s : 26: predicate.compare_switch_simplify 0.24% : 0.000001s : 13: predicate.const_output_eliminate 0.79% : 0.000004s : 26: predicate.depend_value_elim 0.81% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.86% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.81% : 0.000004s : 27: predicate.dict_set_item_eliminator 2.08% : 0.000010s : 26: predicate.dumpgradient_eliminate 0.27% : 0.000001s : 13: predicate.elim_not_effective 0.52% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.15% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.12% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.10% : 0.000005s : 40: predicate.environ_get_depend_swap 1.82% : 0.000009s : 66: predicate.environ_get_eliminate 1.07% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.74% : 0.000003s : 28: predicate.exchange_switch_depend_value 1.26% : 0.000006s : 28: predicate.float_depend_g_call 0.75% : 0.000004s : 26: predicate.float_environ_get_switch 1.11% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.21% : 0.000001s : 13: predicate.fold_const_symbol 0.87% : 0.000004s : 26: predicate.get_grad_eliminate 0.28% : 0.000001s : 13: predicate.graph_param_transform 0.71% : 0.000003s : 26: predicate.incorporate_call 0.67% : 0.000003s : 26: predicate.incorporate_call_switch 5.33% : 0.000025s : 133: predicate.inline 1.03% : 0.000005s : 26: predicate.inline_without_move 0.41% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.27% : 0.000006s : 27: predicate.less_batch_normalization 1.61% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.17% : 0.000010s : 80: predicate.load_eliminater 1.38% : 0.000006s : 13: predicate.loop_unroll_after_grad 0.92% : 0.000004s : 31: predicate.loop_unroll_before_grad 1.72% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.74% : 0.000003s : 26: predicate.merge_addn 0.76% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.86% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.68% : 0.000003s : 27: predicate.minmaximum_grad 2.70% : 0.000013s : 13: predicate.mutable_eliminate 0.54% : 0.000002s : 13: predicate.opt_reshape 0.46% : 0.000002s : 13: predicate.parallel_virtual_node 1.01% : 0.000005s : 28: predicate.partial_defer_inline 1.22% : 0.000006s : 40: predicate.partial_eliminate 0.74% : 0.000003s : 27: predicate.print_const_string_wrapper 0.73% : 0.000003s : 26: predicate.reduce_all_const_elim 1.05% : 0.000005s : 27: predicate.reduce_eliminate 2.17% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.46% : 0.000002s : 26: predicate.remove_not_recompute_node 1.37% : 0.000006s : 53: predicate.replace_applicator 0.53% : 0.000002s : 26: predicate.replace_old_param 0.52% : 0.000002s : 13: predicate.reset_defer_inline 0.91% : 0.000004s : 27: predicate.reshape_eliminate 0.90% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.44% : 0.000002s : 13: predicate.row_tensor_eliminate 1.29% : 0.000006s : 26: predicate.same_eliminate 0.52% : 0.000002s : 26: predicate.set_cell_output_no_recompute 0.88% : 0.000004s : 26: predicate.shard_identity_eliminate 1.09% : 0.000005s : 26: predicate.special_op_eliminate 0.85% : 0.000004s : 26: predicate.specialize_transform 1.24% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.91% : 0.000004s : 26: predicate.stack_unstack_eliminate 0.46% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.83% : 0.000004s : 28: predicate.switch_defer_inline 1.56% : 0.000007s : 54: predicate.switch_layer_defer_inline 3.35% : 0.000016s : 98: predicate.switch_simplify 0.80% : 0.000004s : 27: predicate.tile_eliminate 0.74% : 0.000003s : 27: predicate.transpose_eliminate 1.78% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 1.72% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.66% : 0.000008s : 53: predicate.tuple_list_get_item_depend_reorder 3.21% : 0.000015s : 79: predicate.tuple_list_get_item_eliminator 1.71% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.52% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.63% : 0.000008s : 53: predicate.tuple_to_list_eliminator_ 2.23% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.41% : 0.000016s : 106: predicate.updatestate_useless_node_eliminater 0.56% : 0.000003s : 13: predicate.value_based_eliminate 0.79% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.82% : 0.000004s : 26: predicate.virtual_output_eliminate 0.41% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.66% : 0.000003s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.013861 58 73.23% : 0.010151s : 55: func_graph_cloner_run.FuncGraphClonerGraph 26.77% : 0.003710s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.559035 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.55% : 0.008646s : 1: add_attr 0.55% : 0.008626s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000131s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000492s : 1: auto_monad 0.00% : 0.000055s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.06% : 0.000879s : 1: bootstrap 0.00% : 0.000048s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000032s : 1: control_data_broadcast_order 0.00% : 0.000020s : 1: convert_after_rewriter 0.00% : 0.000064s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000026s : 1: environ_conv 0.00% : 0.000030s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000012s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000008s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000005s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000011s : 1: label_micro_interleaved_index 0.05% : 0.000763s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 2.59% : 0.040318s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000040s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000112s : 1: opt.transform.mutable_eliminate 0.17% : 0.002578s : 78: opt.transform.opt_a 0.01% : 0.000111s : 1: opt.transform.opt_after_cconv 0.00% : 0.000066s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000390s : 28: opt.transform.opt_b 0.01% : 0.000134s : 2: opt.transform.opt_trans_graph 0.01% : 0.000099s : 4: opt.transform.symbol_engine_opt 4.58% : 0.071375s : 1: opt_a 0.02% : 0.000266s : 1: opt_after_cconv 0.05% : 0.000774s : 1: opt_after_jit_grad 0.04% : 0.000666s : 1: opt_b 7.37% : 0.114960s : 1: optimize 0.00% : 0.000048s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000014s : 1: order_py_execute_after_rewriter 0.00% : 0.000046s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000007s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000009s : 1: overlap_param_gather 0.00% : 0.000006s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000011s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000006s : 1: pipeline_split 0.00% : 0.000042s : 1: pre_auto_parallel 0.00% : 0.000039s : 1: py_interpret_to_execute 0.00% : 0.000034s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000082s : 1: remove_dup_value 3.96% : 0.061730s : 1: renormalize.infer 0.27% : 0.004145s : 1: renormalize.specialize 0.00% : 0.000005s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000229s : 1: rewriter_after_opt_a 0.01% : 0.000150s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000006s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000016s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000164s : 1: symbol_engine_optimizer 0.01% : 0.000172s : 1: tuple_transform 79.55% : 1.240165s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape3-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape3-1],max_mem:208.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:42:47.133.440 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 2.17908, [21] [bootstrap]: 0.00154808 [type_inference]: 2.02494 [event_method]: 1.948e-05 [auto_monad]: 0.00032628 [graph_reusing]: 7.41001e-06 [inline]: 2.99001e-06 [add_attr]: 0.00665901, [1] [add_attr_with_inline]: 0.00664103, [1] [Cycle 1]: 7.965e-05, [2] [tag_attr]: 2.622e-05 [meta_addattr_fg_expand]: 4.91002e-06 [parallel-infer-symbol]: 5.52999e-06 [pre_auto_parallel]: 3.707e-05 [insert-virtual-dataset]: 2.59999e-06 [parallel-infer-symbol-second]: 1.02998e-06 [dataset_repeat_opt]: 2.30002e-06 [pipeline_split]: 2.23998e-06 [optimize]: 0.143937, [53] [py_interpret_to_execute]: 3.587e-05 [rewriter_before_opt_a]: 8.711e-05 [opt_a]: 0.0192717, [2] [Cycle 1]: 0.0174081, [45] [expand_dump_flag]: 2.96001e-06 [switch_simplify]: 3.517e-05 [loop_unroll]: 2.02e-05 [a_1]: 0.00089556 [with_stream_mark]: 3.494e-05 [recompute_prepare]: 2.654e-05 [updatestate_depend_eliminate]: 1.825e-05 [updatestate_assign_eliminate]: 8.90001e-06 [updatestate_loads_eliminate]: 1.131e-05 [parameter_eliminate]: 2.29999e-06 [a_2]: 0.00026089 [accelerated_algorithm]: 5.192e-05 [shard]: 3.16001e-06 [meta_shard_fg_expand]: 4.23001e-06 [shard_inline]: 1.957e-05 [merge_send_recv]: 1.754e-05 [auto_parallel]: 1.639e-05 [parallel]: 4.135e-05 [flash_sp]: 1.501e-05 [merge_comm]: 1.106e-05 [allreduce_fusion]: 7.98001e-06 [matmul_add_comm_reduction]: 1.889e-05 [allreduce_slice_to_reducescatter]: 8.39995e-07 [virtual_shard_identity]: 2.31e-05 [virtual_dataset]: 1.881e-05 [get_grad_eliminate_]: 1.988e-05 [virtual_output]: 1.803e-05 [merge_forward]: 1.017e-05 [cell_reuse_recompute_pass]: 2.17001e-06 [offload_activation]: 1.893e-05 [cell_reuse_handle_not_recompute_node_pass]: 0.00029648 [merge_recompute_call_nodes]: 2.36e-06 [before_grad]: 4.344e-05 [set_forward_comm_id_for_comm_node_pass]: 1.818e-05 [meta_fg_expand]: 7.58999e-06 [flash_sp_send_recv_attached]: 8.78001e-06 [receive_attached]: 2.31e-06 [after_resolve]: 3.362e-05 [a_after_grad]: 3.635e-05 [renormalize]: 0.0133475 [add_forward_monad_depend]: 1.403e-05 [auto_monad_grad]: 3.46999e-06 [auto_monad_eliminator]: 6.896e-05 [cse]: 0.00129739 [a_3]: 0.00014518 [Cycle 2]: 0.00179321, [45] [expand_dump_flag]: 4.33999e-06 [switch_simplify]: 1.974e-05 [loop_unroll]: 1.76e-05 [a_1]: 0.00046728 [with_stream_mark]: 3.473e-05 [recompute_prepare]: 2.073e-05 [updatestate_depend_eliminate]: 1.081e-05 [updatestate_assign_eliminate]: 8.68001e-06 [updatestate_loads_eliminate]: 1.122e-05 [parameter_eliminate]: 2.91999e-06 [a_2]: 0.00022907 [accelerated_algorithm]: 2.413e-05 [shard]: 2.75002e-06 [meta_shard_fg_expand]: 5.75001e-06 [shard_inline]: 1.687e-05 [merge_send_recv]: 1.669e-05 [auto_parallel]: 1.601e-05 [parallel]: 1.25e-05 [flash_sp]: 4.93001e-06 [merge_comm]: 9.39e-06 [allreduce_fusion]: 9.24e-06 [matmul_add_comm_reduction]: 1.915e-05 [allreduce_slice_to_reducescatter]: 9.30013e-07 [virtual_shard_identity]: 2.223e-05 [virtual_dataset]: 1.598e-05 [get_grad_eliminate_]: 1.636e-05 [virtual_output]: 1.565e-05 [merge_forward]: 9.76998e-06 [cell_reuse_recompute_pass]: 3.94002e-06 [offload_activation]: 2.001e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.626e-05 [merge_recompute_call_nodes]: 1.75001e-06 [before_grad]: 2.653e-05 [set_forward_comm_id_for_comm_node_pass]: 1.02e-05 [meta_fg_expand]: 7.65998e-06 [flash_sp_send_recv_attached]: 2.34999e-06 [receive_attached]: 2.44001e-06 [after_resolve]: 2.648e-05 [a_after_grad]: 2.828e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 4.67e-06 [auto_monad_grad]: 2.66e-06 [auto_monad_eliminator]: 5.702e-05 [cse]: 0.00010293 [a_3]: 0.00010584 [py_interpret_to_execute_after_opt_a]: 3.476e-05 [slice_cell_reuse_recomputed_activation]: 2.98e-06 [rewriter_after_opt_a]: 0.00022083 [convert_after_rewriter]: 1.842e-05 [order_py_execute_after_rewriter]: 1.153e-05 [mutable_eliminate]: 0.00085085 [opt_b]: 0.00059899, [1] [Cycle 1]: 0.00058798, [7] [b_1]: 0.00038129 [b_2]: 2.205e-05 [updatestate_depend_eliminate]: 1.879e-05 [updatestate_assign_eliminate]: 8.15999e-06 [updatestate_loads_eliminate]: 1.074e-05 [renormalize]: 9.10019e-07 [cse]: 8.247e-05 [optimize_parallel_all_gather_comm]: 4.018e-05 [overlap_param_gather]: 3.30998e-06 [cconv]: 4.566e-05 [loop_unroll]: 0.00077046 [opt_after_cconv]: 0.00029104, [1] [Cycle 1]: 0.00027931, [7] [c_1]: 0.00011378 [parameter_eliminate]: 6.89999e-06 [updatestate_depend_eliminate]: 2.054e-05 [updatestate_assign_eliminate]: 7.55e-06 [updatestate_loads_eliminate]: 1.15e-05 [cse]: 7.145e-05 [renormalize]: 5.89993e-07 [remove_dup_value]: 8.479e-05 [tuple_transform]: 0.00018905, [1] [Cycle 1]: 0.00018187, [4] [d_1]: 0.0001255 [none_parameter_eliminate]: 3.42002e-06 [renormalize]: 3.80009e-07 [switch_simplify]: 2.058e-05 [partial_unused_args_eliminate]: 3.05998e-06 [add_recomputation]: 0.00014055 [cse_after_recomputation]: 7.88e-05, [1] [Cycle 1]: 7.291e-05, [1] [cse]: 5.9e-05 [environ_conv]: 2.207e-05 [swap_dp_allreduce_reducescatter]: 1.305e-05 [bias_add_comm_swap]: 3.41999e-06 [label_micro_interleaved_index]: 8.08001e-06 [label_fine_grained_interleaved_index]: 2.88e-06 [merge_cast_opt]: 1.52999e-06 [slice_recompute_activation]: 2.46998e-06 [micro_interleaved_order_control]: 3.22002e-06 [assign_add_opt]: 1.47001e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 1.15999e-06 [full_micro_interleaved_order_control]: 2.71e-06 [reorder_send_recv_between_fp_bp]: 3.13e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 1.01002e-06 [interleave_split_concat_branches]: 1.66e-06 [interleave_parallel_branches]: 1.16997e-06 [overlap_opt_shard_in_pipeline]: 4.12998e-06 [overlap_opt_shard_grad_in_pipeline]: 1.97999e-06 [control_data_broadcast_order]: 3.582e-05 [grouped_pairwise_exchange_alltoall]: 2.01998e-06 [offloading_packed_experts]: 8.80999e-06 [overlap_recompute_and_grad_model_parallel]: 7.66001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.48002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.47001e-06 [overlap_recompute_comm]: 2.71e-06 [overlap_grad_ring_attention]: 9.31002e-06 [overlap_grad_flash_sp]: 4.735e-05 [begin_end_overlap_inline]: 6.10016e-07 [split_matmul_comm_elemetwise]: 2.44001e-06 [split_layernorm_comm]: 1.81e-06 [handle_group_info]: 1.02e-06 [symbol_engine_optimizer]: 0.120602, [1] [Cycle 1]: 0.120588, [6] [build]: 2.194e-05 [elim_shapecalc]: 7.654e-05 [elim_not_effective]: 4.534e-05 [opt_reshape]: 2.733e-05 [fold_const_symbol]: 2.696e-05 [renormalize]: 1.35001e-06 [detach_backward]: 4.63999e-06 [pipeline_parallel_scheduler]: 2.53998e-06 [auto_monad_reorder]: 7.969e-05 [get_jit_bprop_graph]: 2.29999e-06 [rewriter_after_jit_bprop_graph]: 1.262e-05 [opt_after_jit_grad]: 0.00091606 [validate]: 0.00012372 Sums bootstrap : 0.001548s : 0.08% type_inference : 2.024939s : 98.77% event_method : 0.000019s : 0.00% auto_monad : 0.000326s : 0.02% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000026s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000005s : 0.00% parallel-infer-symbol : 0.000006s : 0.00% pre_auto_parallel : 0.000037s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000036s : 0.00% optimize.rewriter_before_opt_a : 0.000087s : 0.00% optimize.opt_a.expand_dump_flag : 0.000007s : 0.00% optimize.opt_a.switch_simplify : 0.000055s : 0.00% optimize.opt_a.loop_unroll : 0.000038s : 0.00% optimize.opt_a.a_1 : 0.001363s : 0.07% optimize.opt_a.with_stream_mark : 0.000070s : 0.00% optimize.opt_a.recompute_prepare : 0.000047s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000029s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000018s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000023s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000490s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000076s : 0.00% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000036s : 0.00% optimize.opt_a.merge_send_recv : 0.000034s : 0.00% optimize.opt_a.auto_parallel : 0.000032s : 0.00% optimize.opt_a.parallel : 0.000054s : 0.00% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000020s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000038s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000045s : 0.00% optimize.opt_a.virtual_dataset : 0.000035s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000036s : 0.00% optimize.opt_a.virtual_output : 0.000034s : 0.00% optimize.opt_a.merge_forward : 0.000020s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000039s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000333s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000070s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000028s : 0.00% optimize.opt_a.meta_fg_expand : 0.000015s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000011s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000060s : 0.00% optimize.opt_a.a_after_grad : 0.000065s : 0.00% optimize.opt_a.renormalize : 0.013348s : 0.65% optimize.opt_a.add_forward_monad_depend : 0.000019s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000126s : 0.01% optimize.opt_a.cse : 0.001400s : 0.07% optimize.opt_a.a_3 : 0.000251s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000035s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000221s : 0.01% optimize.convert_after_rewriter : 0.000018s : 0.00% optimize.order_py_execute_after_rewriter : 0.000012s : 0.00% optimize.mutable_eliminate : 0.000851s : 0.04% optimize.opt_b.b_1 : 0.000381s : 0.02% optimize.opt_b.b_2 : 0.000022s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000019s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000082s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000040s : 0.00% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000046s : 0.00% optimize.loop_unroll : 0.000770s : 0.04% optimize.opt_after_cconv.c_1 : 0.000114s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000021s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.cse : 0.000071s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000085s : 0.00% optimize.tuple_transform.d_1 : 0.000126s : 0.01% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000021s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000141s : 0.01% optimize.cse_after_recomputation.cse : 0.000059s : 0.00% optimize.environ_conv : 0.000022s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000013s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000008s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000002s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000036s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000009s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000047s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000022s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000077s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000045s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000027s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000027s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000001s : 0.00% detach_backward : 0.000005s : 0.00% pipeline_parallel_scheduler : 0.000003s : 0.00% auto_monad_reorder : 0.000080s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000013s : 0.00% opt_after_jit_grad : 0.000916s : 0.04% validate : 0.000124s : 0.01% Time group info: ------[substitution.] 0.000333 96 3.24% : 0.000011s : 2: substitution.depend_value_elim 1.59% : 0.000005s : 8: substitution.elim_not_effective 1.14% : 0.000004s : 8: substitution.fold_const_symbol 4.04% : 0.000013s : 13: substitution.graph_param_transform 41.35% : 0.000137s : 1: substitution.inline 3.40% : 0.000011s : 16: substitution.j_node_and_user_rematch 7.84% : 0.000026s : 2: substitution.less_batch_normalization 2.05% : 0.000007s : 6: substitution.load_eliminater 1.07% : 0.000004s : 2: substitution.opt_reshape 4.45% : 0.000015s : 16: substitution.remove_not_recompute_node 2.96% : 0.000010s : 4: substitution.replace_old_param 10.01% : 0.000033s : 4: substitution.reshape_eliminate 10.05% : 0.000033s : 6: substitution.updatestate_pure_node_eliminater 6.82% : 0.000023s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.024827 2 99.74% : 2.019550s : 1: type_inference.infer 0.26% : 0.005277s : 1: type_inference.specialize ------[replace.] 0.000020 1 100.00% : 0.000020s : 1: replace.inline ------[match.] 0.000135 1 100.00% : 0.000135s : 1: match.inline ------[predicate.] 0.000481 3002 0.73% : 0.000003s : 27: predicate.accumulaten_eliminater 1.05% : 0.000005s : 13: predicate.ad_related_special_op_eliminate 0.69% : 0.000003s : 26: predicate.addn_check_dump 0.84% : 0.000004s : 27: predicate.addn_zero_filter 0.75% : 0.000004s : 27: predicate.adjust_all_reduce_mul_add 2.00% : 0.000010s : 53: predicate.arithmetic_simplify 0.83% : 0.000004s : 27: predicate.cast_eliminate 0.75% : 0.000004s : 26: predicate.check_bprop_eliminate 0.71% : 0.000003s : 26: predicate.compare_switch_simplify 0.22% : 0.000001s : 13: predicate.const_output_eliminate 0.76% : 0.000004s : 26: predicate.depend_value_elim 0.77% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.81% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.75% : 0.000004s : 27: predicate.dict_set_item_eliminator 2.15% : 0.000010s : 26: predicate.dumpgradient_eliminate 0.41% : 0.000002s : 13: predicate.elim_not_effective 2.54% : 0.000012s : 13: predicate.elim_shapecalc_of_broadcastargs 1.14% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.06% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.01% : 0.000005s : 40: predicate.environ_get_depend_swap 1.81% : 0.000009s : 66: predicate.environ_get_eliminate 1.13% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.83% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.26% : 0.000006s : 28: predicate.float_depend_g_call 0.80% : 0.000004s : 26: predicate.float_environ_get_switch 1.02% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.21% : 0.000001s : 13: predicate.fold_const_symbol 0.96% : 0.000005s : 26: predicate.get_grad_eliminate 0.29% : 0.000001s : 13: predicate.graph_param_transform 0.79% : 0.000004s : 26: predicate.incorporate_call 0.65% : 0.000003s : 26: predicate.incorporate_call_switch 5.47% : 0.000026s : 133: predicate.inline 1.20% : 0.000006s : 26: predicate.inline_without_move 0.46% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.25% : 0.000006s : 27: predicate.less_batch_normalization 1.65% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.18% : 0.000010s : 80: predicate.load_eliminater 1.49% : 0.000007s : 13: predicate.loop_unroll_after_grad 1.04% : 0.000005s : 31: predicate.loop_unroll_before_grad 1.69% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.71% : 0.000003s : 26: predicate.merge_addn 0.69% : 0.000003s : 26: predicate.micro_step_allgather_replace 0.67% : 0.000003s : 26: predicate.mini_step_allgather_replace 0.67% : 0.000003s : 27: predicate.minmaximum_grad 1.43% : 0.000007s : 13: predicate.mutable_eliminate 0.68% : 0.000003s : 13: predicate.opt_reshape 0.52% : 0.000002s : 13: predicate.parallel_virtual_node 1.13% : 0.000005s : 28: predicate.partial_defer_inline 1.19% : 0.000006s : 40: predicate.partial_eliminate 0.71% : 0.000003s : 27: predicate.print_const_string_wrapper 0.81% : 0.000004s : 26: predicate.reduce_all_const_elim 0.99% : 0.000005s : 27: predicate.reduce_eliminate 2.41% : 0.000012s : 80: predicate.redundant_stop_gradient_eliminater 0.67% : 0.000003s : 26: predicate.remove_not_recompute_node 1.32% : 0.000006s : 53: predicate.replace_applicator 0.50% : 0.000002s : 26: predicate.replace_old_param 0.37% : 0.000002s : 13: predicate.reset_defer_inline 0.86% : 0.000004s : 27: predicate.reshape_eliminate 0.74% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.49% : 0.000002s : 13: predicate.row_tensor_eliminate 1.49% : 0.000007s : 26: predicate.same_eliminate 0.65% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.99% : 0.000005s : 26: predicate.shard_identity_eliminate 0.82% : 0.000004s : 26: predicate.special_op_eliminate 0.85% : 0.000004s : 26: predicate.specialize_transform 1.17% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.99% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.46% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.84% : 0.000004s : 28: predicate.switch_defer_inline 1.51% : 0.000007s : 54: predicate.switch_layer_defer_inline 3.80% : 0.000018s : 98: predicate.switch_simplify 0.79% : 0.000004s : 27: predicate.tile_eliminate 0.74% : 0.000004s : 27: predicate.transpose_eliminate 1.53% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.55% : 0.000007s : 53: predicate.tuple_list_get_item_const_eliminator 1.42% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.87% : 0.000014s : 79: predicate.tuple_list_get_item_eliminator 1.53% : 0.000007s : 53: predicate.tuple_list_get_set_item_eliminator 2.42% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.67% : 0.000008s : 53: predicate.tuple_to_list_eliminator_ 2.33% : 0.000011s : 80: predicate.updatestate_pure_node_eliminater 3.03% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.47% : 0.000002s : 13: predicate.value_based_eliminate 0.77% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.79% : 0.000004s : 26: predicate.virtual_output_eliminate 0.38% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.47% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.016413 58 76.13% : 0.012495s : 55: func_graph_cloner_run.FuncGraphClonerGraph 23.87% : 0.003918s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.346317 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.28% : 0.006666s : 1: add_attr 0.28% : 0.006645s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 0.01% : 0.000153s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000338s : 1: auto_monad 0.00% : 0.000086s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.08% : 0.001794s : 1: bootstrap 0.00% : 0.000051s : 1: cconv 0.00% : 0.000007s : 1: comm_op_add_attrs 0.00% : 0.000041s : 1: control_data_broadcast_order 0.00% : 0.000024s : 1: convert_after_rewriter 0.00% : 0.000082s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000008s : 1: detach_backward 0.00% : 0.000028s : 1: environ_conv 0.00% : 0.000028s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000012s : 1: graph_reusing 0.00% : 0.000006s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000008s : 1: insert-virtual-dataset 0.00% : 0.000006s : 1: interleave_parallel_branches 0.00% : 0.000005s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000012s : 1: label_micro_interleaved_index 0.03% : 0.000791s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.04% : 0.000872s : 1: mutable_eliminate 0.00% : 0.000012s : 1: offloading_packed_experts 0.00% : 0.000043s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000043s : 1: opt.transform.mutable_eliminate 0.11% : 0.002657s : 78: opt.transform.opt_a 0.00% : 0.000112s : 1: opt.transform.opt_after_cconv 0.00% : 0.000070s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000365s : 28: opt.transform.opt_b 0.01% : 0.000141s : 2: opt.transform.opt_trans_graph 0.01% : 0.000163s : 4: opt.transform.symbol_engine_opt 0.82% : 0.019278s : 1: opt_a 0.01% : 0.000295s : 1: opt_after_cconv 0.04% : 0.000935s : 1: opt_after_jit_grad 0.03% : 0.000604s : 1: opt_b 6.13% : 0.143945s : 1: optimize 0.00% : 0.000045s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000015s : 1: order_py_execute_after_rewriter 0.00% : 0.000055s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000013s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000009s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000009s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000012s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000010s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000006s : 1: pipeline_parallel_scheduler 0.00% : 0.000006s : 1: pipeline_split 0.00% : 0.000042s : 1: pre_auto_parallel 0.00% : 0.000041s : 1: py_interpret_to_execute 0.00% : 0.000040s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000096s : 1: remove_dup_value 0.38% : 0.008953s : 1: renormalize.infer 0.19% : 0.004373s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000016s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000231s : 1: rewriter_after_opt_a 0.00% : 0.000093s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000007s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000017s : 1: swap_dp_allreduce_reducescatter 5.14% : 0.120609s : 1: symbol_engine_optimizer 0.01% : 0.000193s : 1: tuple_transform 86.30% : 2.024977s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[False-mstype0-k_n_shape3-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[False-mstype0-k_n_shape3-16],max_mem:208.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:43:25.166.711 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 1.35138, [21] [bootstrap]: 0.00334627 [type_inference]: 1.17084 [event_method]: 0.00025755 [auto_monad]: 0.00080385 [graph_reusing]: 5.193e-05 [inline]: 3.74002e-06 [add_attr]: 0.041844, [1] [add_attr_with_inline]: 0.0418243, [1] [Cycle 1]: 0.00031006, [2] [tag_attr]: 0.00020236 [meta_addattr_fg_expand]: 5.084e-05 [parallel-infer-symbol]: 3.8e-06 [pre_auto_parallel]: 0.00024424 [insert-virtual-dataset]: 3.08e-06 [parallel-infer-symbol-second]: 1.26002e-06 [dataset_repeat_opt]: 2.10002e-06 [pipeline_split]: 1.50999e-06 [optimize]: 0.132411, [53] [py_interpret_to_execute]: 0.00023667 [rewriter_before_opt_a]: 0.0008556 [opt_a]: 0.125664, [2] [Cycle 1]: 0.121966, [45] [expand_dump_flag]: 2.404e-05 [switch_simplify]: 0.00081487 [loop_unroll]: 0.00031585 [a_1]: 0.0385644 [with_stream_mark]: 6.092e-05 [recompute_prepare]: 4.246e-05 [updatestate_depend_eliminate]: 2.042e-05 [updatestate_assign_eliminate]: 1.642e-05 [updatestate_loads_eliminate]: 1.957e-05 [parameter_eliminate]: 2.64001e-06 [a_2]: 0.00079047 [accelerated_algorithm]: 6.349e-05 [shard]: 2.41e-06 [meta_shard_fg_expand]: 2.234e-05 [shard_inline]: 2.321e-05 [merge_send_recv]: 2.099e-05 [auto_parallel]: 2.127e-05 [parallel]: 4.131e-05 [flash_sp]: 1.521e-05 [merge_comm]: 1.382e-05 [allreduce_fusion]: 1.269e-05 [matmul_add_comm_reduction]: 2.191e-05 [allreduce_slice_to_reducescatter]: 8.00006e-07 [virtual_shard_identity]: 2.666e-05 [virtual_dataset]: 2.332e-05 [get_grad_eliminate_]: 2.099e-05 [virtual_output]: 2.169e-05 [merge_forward]: 1.313e-05 [cell_reuse_recompute_pass]: 2.99001e-06 [offload_activation]: 2.413e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.438e-05 [merge_recompute_call_nodes]: 1.87999e-06 [before_grad]: 3.988e-05 [set_forward_comm_id_for_comm_node_pass]: 1.486e-05 [meta_fg_expand]: 1.569e-05 [flash_sp_send_recv_attached]: 5.82001e-06 [receive_attached]: 2.72001e-06 [after_resolve]: 2.918e-05 [a_after_grad]: 3.724e-05 [renormalize]: 0.0774984 [add_forward_monad_depend]: 1.425e-05 [auto_monad_grad]: 3.63e-06 [auto_monad_eliminator]: 9.386e-05 [cse]: 0.00244321 [a_3]: 0.00019571 [Cycle 2]: 0.00367621, [45] [expand_dump_flag]: 4.07e-06 [switch_simplify]: 3.009e-05 [loop_unroll]: 2.416e-05 [a_1]: 0.00072917 [with_stream_mark]: 4.382e-05 [recompute_prepare]: 2.58e-05 [updatestate_depend_eliminate]: 1.754e-05 [updatestate_assign_eliminate]: 1.257e-05 [updatestate_loads_eliminate]: 1.651e-05 [parameter_eliminate]: 3.03e-06 [a_2]: 0.00033245 [accelerated_algorithm]: 3.273e-05 [shard]: 2.99001e-06 [meta_shard_fg_expand]: 9.71e-06 [shard_inline]: 2.173e-05 [merge_send_recv]: 2.156e-05 [auto_parallel]: 2.165e-05 [parallel]: 1.06e-05 [flash_sp]: 5.25001e-06 [merge_comm]: 1.384e-05 [allreduce_fusion]: 1.274e-05 [matmul_add_comm_reduction]: 2.463e-05 [allreduce_slice_to_reducescatter]: 8.89995e-07 [virtual_shard_identity]: 2.676e-05 [virtual_dataset]: 2.099e-05 [get_grad_eliminate_]: 2.204e-05 [virtual_output]: 2.455e-05 [merge_forward]: 1.45e-05 [cell_reuse_recompute_pass]: 3.36001e-06 [offload_activation]: 2.922e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.667e-05 [merge_recompute_call_nodes]: 1.60001e-06 [before_grad]: 4.091e-05 [set_forward_comm_id_for_comm_node_pass]: 1.527e-05 [meta_fg_expand]: 0.00119467 [flash_sp_send_recv_attached]: 3.96001e-06 [receive_attached]: 3.88001e-06 [after_resolve]: 6.641e-05 [a_after_grad]: 3.741e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 6.87002e-06 [auto_monad_grad]: 3.28998e-06 [auto_monad_eliminator]: 8.773e-05 [cse]: 0.00010203 [a_3]: 0.00015172 [py_interpret_to_execute_after_opt_a]: 4.122e-05 [slice_cell_reuse_recomputed_activation]: 2.98e-06 [rewriter_after_opt_a]: 0.00088355 [convert_after_rewriter]: 2.875e-05 [order_py_execute_after_rewriter]: 1.454e-05 [mutable_eliminate]: 0.00087731 [opt_b]: 0.00086654, [1] [Cycle 1]: 0.00085648, [7] [b_1]: 0.00062476 [b_2]: 2.592e-05 [updatestate_depend_eliminate]: 2.35e-05 [updatestate_assign_eliminate]: 1.199e-05 [updatestate_loads_eliminate]: 1.64e-05 [renormalize]: 1.09e-06 [cse]: 9.86e-05 [optimize_parallel_all_gather_comm]: 4.38e-05 [overlap_param_gather]: 6.69999e-06 [cconv]: 4.609e-05 [loop_unroll]: 0.00073182 [opt_after_cconv]: 0.00033658, [1] [Cycle 1]: 0.00032632, [7] [c_1]: 0.00014526 [parameter_eliminate]: 6.23e-06 [updatestate_depend_eliminate]: 2.04e-05 [updatestate_assign_eliminate]: 1.116e-05 [updatestate_loads_eliminate]: 1.327e-05 [cse]: 8.448e-05 [renormalize]: 8.70001e-07 [remove_dup_value]: 0.00010675 [tuple_transform]: 0.00027023, [1] [Cycle 1]: 0.00026365, [4] [d_1]: 0.0002155 [none_parameter_eliminate]: 2.31e-06 [renormalize]: 2.59985e-07 [switch_simplify]: 2.352e-05 [partial_unused_args_eliminate]: 2.57001e-06 [add_recomputation]: 0.00016245 [cse_after_recomputation]: 7.687e-05, [1] [Cycle 1]: 6.909e-05, [1] [cse]: 6.153e-05 [environ_conv]: 2.777e-05 [swap_dp_allreduce_reducescatter]: 1.857e-05 [bias_add_comm_swap]: 3.93999e-06 [label_micro_interleaved_index]: 7.41001e-06 [label_fine_grained_interleaved_index]: 3.11999e-06 [merge_cast_opt]: 1.60999e-06 [slice_recompute_activation]: 2.31e-06 [micro_interleaved_order_control]: 3.16001e-06 [assign_add_opt]: 1.62999e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 1.35001e-06 [full_micro_interleaved_order_control]: 2.53e-06 [reorder_send_recv_between_fp_bp]: 2.89999e-06 [comm_op_add_attrs]: 1.05999e-06 [add_comm_op_reuse_tag]: 1.44e-06 [interleave_split_concat_branches]: 1.54e-06 [interleave_parallel_branches]: 1.15999e-06 [overlap_opt_shard_in_pipeline]: 3.80998e-06 [overlap_opt_shard_grad_in_pipeline]: 2.42001e-06 [control_data_broadcast_order]: 4.023e-05 [grouped_pairwise_exchange_alltoall]: 2.11998e-06 [offloading_packed_experts]: 1.182e-05 [overlap_recompute_and_grad_model_parallel]: 1.188e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.45999e-06 [overlap_recompute_comm]: 2.79999e-06 [overlap_grad_ring_attention]: 9.99001e-06 [overlap_grad_flash_sp]: 5.228e-05 [begin_end_overlap_inline]: 6.09987e-07 [split_matmul_comm_elemetwise]: 2.53998e-06 [split_layernorm_comm]: 1.96e-06 [handle_group_info]: 1.00999e-06 [symbol_engine_optimizer]: 0.00059438, [1] [Cycle 1]: 0.00058783, [6] [build]: 0.00032075 [elim_shapecalc]: 3.608e-05 [elim_not_effective]: 7.856e-05 [opt_reshape]: 3.609e-05 [fold_const_symbol]: 6.079e-05 [renormalize]: 4.20026e-07 [detach_backward]: 2.96001e-06 [pipeline_parallel_scheduler]: 1.44998e-06 [auto_monad_reorder]: 6.968e-05 [get_jit_bprop_graph]: 2.17999e-06 [rewriter_after_jit_bprop_graph]: 7.06001e-06 [opt_after_jit_grad]: 0.00095272 [validate]: 0.00011627 Sums bootstrap : 0.003346s : 0.26% type_inference : 1.170843s : 89.52% event_method : 0.000258s : 0.02% auto_monad : 0.000804s : 0.06% graph_reusing : 0.000052s : 0.00% inline : 0.000004s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000202s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000051s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000244s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000237s : 0.02% optimize.rewriter_before_opt_a : 0.000856s : 0.07% optimize.opt_a.expand_dump_flag : 0.000028s : 0.00% optimize.opt_a.switch_simplify : 0.000845s : 0.06% optimize.opt_a.loop_unroll : 0.000340s : 0.03% optimize.opt_a.a_1 : 0.039294s : 3.00% optimize.opt_a.with_stream_mark : 0.000105s : 0.01% optimize.opt_a.recompute_prepare : 0.000068s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000038s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000029s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000036s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001123s : 0.09% optimize.opt_a.accelerated_algorithm : 0.000096s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000032s : 0.00% optimize.opt_a.shard_inline : 0.000045s : 0.00% optimize.opt_a.merge_send_recv : 0.000043s : 0.00% optimize.opt_a.auto_parallel : 0.000043s : 0.00% optimize.opt_a.parallel : 0.000052s : 0.00% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000028s : 0.00% optimize.opt_a.allreduce_fusion : 0.000025s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000047s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000053s : 0.00% optimize.opt_a.virtual_dataset : 0.000044s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000043s : 0.00% optimize.opt_a.virtual_output : 0.000046s : 0.00% optimize.opt_a.merge_forward : 0.000028s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000053s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000091s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000081s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000030s : 0.00% optimize.opt_a.meta_fg_expand : 0.001210s : 0.09% optimize.opt_a.flash_sp_send_recv_attached : 0.000010s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000096s : 0.01% optimize.opt_a.a_after_grad : 0.000075s : 0.01% optimize.opt_a.renormalize : 0.077499s : 5.93% optimize.opt_a.add_forward_monad_depend : 0.000021s : 0.00% optimize.opt_a.auto_monad_grad : 0.000007s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000182s : 0.01% optimize.opt_a.cse : 0.002545s : 0.19% optimize.opt_a.a_3 : 0.000347s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000041s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000884s : 0.07% optimize.convert_after_rewriter : 0.000029s : 0.00% optimize.order_py_execute_after_rewriter : 0.000015s : 0.00% optimize.mutable_eliminate : 0.000877s : 0.07% optimize.opt_b.b_1 : 0.000625s : 0.05% optimize.opt_b.b_2 : 0.000026s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000024s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000016s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000099s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000044s : 0.00% optimize.overlap_param_gather : 0.000007s : 0.00% optimize.cconv : 0.000046s : 0.00% optimize.loop_unroll : 0.000732s : 0.06% optimize.opt_after_cconv.c_1 : 0.000145s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000020s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.cse : 0.000084s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000107s : 0.01% optimize.tuple_transform.d_1 : 0.000216s : 0.02% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000024s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000162s : 0.01% optimize.cse_after_recomputation.cse : 0.000062s : 0.00% optimize.environ_conv : 0.000028s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000019s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000007s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000002s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000040s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000012s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000012s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000010s : 0.00% optimize.overlap_grad_flash_sp : 0.000052s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000321s : 0.02% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000036s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000079s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000036s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000061s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000070s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000953s : 0.07% validate : 0.000116s : 0.01% Time group info: ------[substitution.] 0.003938 569 0.45% : 0.000018s : 8: substitution.depend_value_elim 0.88% : 0.000035s : 14: substitution.elim_not_effective 0.47% : 0.000018s : 12: substitution.float_tuple_getitem_switch 0.75% : 0.000029s : 14: substitution.fold_const_symbol 0.40% : 0.000016s : 17: substitution.graph_param_transform 66.29% : 0.002611s : 71: substitution.inline 0.38% : 0.000015s : 29: substitution.j_node_and_user_rematch 0.87% : 0.000034s : 6: substitution.less_batch_normalization 1.47% : 0.000058s : 2: substitution.list_to_tuple_eliminator_ 0.19% : 0.000007s : 6: substitution.load_eliminater 1.02% : 0.000040s : 21: substitution.minmaximum_grad 0.24% : 0.000009s : 2: substitution.opt_reshape 0.52% : 0.000020s : 29: substitution.remove_not_recompute_node 0.25% : 0.000010s : 4: substitution.replace_old_param 1.65% : 0.000065s : 18: substitution.reshape_eliminate 1.29% : 0.000051s : 24: substitution.switch_simplify 4.49% : 0.000177s : 51: substitution.tuple_list_convert_item_index_to_positive 2.52% : 0.000099s : 51: substitution.tuple_list_get_item_const_eliminator 3.36% : 0.000132s : 51: substitution.tuple_list_get_item_depend_reorder 6.27% : 0.000247s : 73: substitution.tuple_list_get_item_eliminator 5.04% : 0.000198s : 51: substitution.tuple_list_get_set_item_eliminator 0.69% : 0.000027s : 6: substitution.updatestate_pure_node_eliminater 0.55% : 0.000022s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 1.170307 2 93.63% : 1.095710s : 1: type_inference.infer 6.37% : 0.074597s : 1: type_inference.specialize ------[replace.] 0.001185 115 4.38% : 0.000052s : 5: replace.depend_value_elim 56.91% : 0.000675s : 71: replace.inline 3.61% : 0.000043s : 2: replace.list_to_tuple_eliminator_ 25.62% : 0.000304s : 24: replace.switch_simplify 9.48% : 0.000112s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002701 115 0.10% : 0.000003s : 5: match.depend_value_elim 95.08% : 0.002568s : 71: match.inline 2.05% : 0.000055s : 2: match.list_to_tuple_eliminator_ 1.32% : 0.000036s : 24: match.switch_simplify 1.45% : 0.000039s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001660 9701 1.15% : 0.000019s : 123: predicate.accumulaten_eliminater 0.39% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.52% : 0.000009s : 58: predicate.addn_check_dump 1.17% : 0.000019s : 123: predicate.addn_zero_filter 1.17% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 2.20% : 0.000037s : 181: predicate.arithmetic_simplify 1.23% : 0.000020s : 123: predicate.cast_eliminate 0.33% : 0.000005s : 34: predicate.check_bprop_eliminate 0.49% : 0.000008s : 58: predicate.compare_switch_simplify 0.09% : 0.000001s : 17: predicate.const_output_eliminate 0.59% : 0.000010s : 61: predicate.depend_value_elim 1.29% : 0.000021s : 123: predicate.dict_get_item_const_eliminator 1.42% : 0.000024s : 123: predicate.dict_get_item_eliminator 1.07% : 0.000018s : 123: predicate.dict_set_item_eliminator 0.41% : 0.000007s : 34: predicate.dumpgradient_eliminate 0.10% : 0.000002s : 17: predicate.elim_not_effective 0.20% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.20% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.20% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.24% : 0.000021s : 140: predicate.environ_get_depend_swap 2.20% : 0.000037s : 198: predicate.environ_get_eliminate 1.33% : 0.000022s : 140: predicate.environ_get_set_eliminate 1.96% : 0.000032s : 209: predicate.exchange_switch_depend_value 2.85% : 0.000047s : 209: predicate.float_depend_g_call 0.49% : 0.000008s : 58: predicate.float_environ_get_switch 0.72% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.07% : 0.000001s : 17: predicate.fold_const_symbol 0.34% : 0.000006s : 35: predicate.get_grad_eliminate 0.10% : 0.000002s : 17: predicate.graph_param_transform 0.48% : 0.000008s : 58: predicate.incorporate_call 0.45% : 0.000007s : 58: predicate.incorporate_call_switch 6.11% : 0.000101s : 459: predicate.inline 0.50% : 0.000008s : 35: predicate.inline_without_move 0.15% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.47% : 0.000008s : 36: predicate.less_batch_normalization 1.71% : 0.000028s : 172: predicate.list_to_tuple_eliminator_ 2.58% : 0.000043s : 295: predicate.load_eliminater 0.43% : 0.000007s : 17: predicate.loop_unroll_after_grad 2.97% : 0.000049s : 314: predicate.loop_unroll_before_grad 1.61% : 0.000027s : 157: predicate.make_slice_get_slice_eliminator 0.52% : 0.000009s : 58: predicate.merge_addn 0.31% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.30% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.12% : 0.000019s : 123: predicate.minmaximum_grad 0.43% : 0.000007s : 17: predicate.mutable_eliminate 0.19% : 0.000003s : 17: predicate.opt_reshape 0.18% : 0.000003s : 17: predicate.parallel_virtual_node 3.49% : 0.000058s : 209: predicate.partial_defer_inline 1.49% : 0.000025s : 155: predicate.partial_eliminate 1.13% : 0.000019s : 123: predicate.print_const_string_wrapper 0.47% : 0.000008s : 53: predicate.reduce_all_const_elim 1.68% : 0.000028s : 123: predicate.reduce_eliminate 2.57% : 0.000043s : 295: predicate.redundant_stop_gradient_eliminater 0.16% : 0.000003s : 35: predicate.remove_not_recompute_node 1.10% : 0.000018s : 172: predicate.replace_applicator 0.21% : 0.000004s : 35: predicate.replace_old_param 0.09% : 0.000001s : 17: predicate.reset_defer_inline 1.26% : 0.000021s : 123: predicate.reshape_eliminate 0.31% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.20% : 0.000003s : 17: predicate.row_tensor_eliminate 0.60% : 0.000010s : 34: predicate.same_eliminate 0.21% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.37% : 0.000006s : 35: predicate.shard_identity_eliminate 0.37% : 0.000006s : 34: predicate.special_op_eliminate 0.58% : 0.000010s : 58: predicate.specialize_transform 0.53% : 0.000009s : 34: predicate.split_environ_get_set_with_tuple_value 0.44% : 0.000007s : 35: predicate.stack_unstack_eliminate 0.18% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.12% : 0.000035s : 209: predicate.switch_defer_inline 2.41% : 0.000040s : 243: predicate.switch_layer_defer_inline 6.53% : 0.000108s : 646: predicate.switch_simplify 1.27% : 0.000021s : 123: predicate.tile_eliminate 1.18% : 0.000020s : 123: predicate.transpose_eliminate 1.67% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.70% : 0.000028s : 157: predicate.tuple_list_get_item_const_eliminator 1.69% : 0.000028s : 157: predicate.tuple_list_get_item_depend_reorder 3.03% : 0.000050s : 228: predicate.tuple_list_get_item_eliminator 1.84% : 0.000030s : 157: predicate.tuple_list_get_set_item_eliminator 2.39% : 0.000040s : 215: predicate.tuple_list_set_item_eliminator 1.54% : 0.000026s : 170: predicate.tuple_to_list_eliminator_ 2.65% : 0.000044s : 295: predicate.updatestate_pure_node_eliminater 3.20% : 0.000053s : 353: predicate.updatestate_useless_node_eliminater 0.18% : 0.000003s : 17: predicate.value_based_eliminate 0.33% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.39% : 0.000006s : 35: predicate.virtual_output_eliminate 0.16% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.21% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.061289 135 89.66% : 0.054951s : 55: func_graph_cloner_run.FuncGraphClonerGraph 10.34% : 0.006338s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.646805 209 0.00% : 0.000004s : 1: ForceFp32Comm 2.54% : 0.041851s : 1: add_attr 2.54% : 0.041831s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000168s : 1: add_recomputation 0.00% : 0.000006s : 1: assign_add_opt 0.05% : 0.000824s : 1: auto_monad 0.00% : 0.000076s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.21% : 0.003525s : 1: bootstrap 0.00% : 0.000051s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000044s : 1: control_data_broadcast_order 0.00% : 0.000035s : 1: convert_after_rewriter 0.00% : 0.000080s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000031s : 1: environ_conv 0.02% : 0.000272s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000061s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.05% : 0.000747s : 1: loop_unroll 0.00% : 0.000006s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.05% : 0.000891s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.01% : 0.000124s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000048s : 1: opt.transform.mutable_eliminate 2.58% : 0.042520s : 95: opt.transform.opt_a 0.01% : 0.000143s : 1: opt.transform.opt_after_cconv 0.00% : 0.000082s : 1: opt.transform.opt_after_jit_grad 0.04% : 0.000608s : 28: opt.transform.opt_b 0.01% : 0.000236s : 2: opt.transform.opt_trans_graph 0.01% : 0.000204s : 4: opt.transform.symbol_engine_opt 7.63% : 0.125670s : 1: opt_a 0.02% : 0.000341s : 1: opt_after_cconv 0.06% : 0.000968s : 1: opt_after_jit_grad 0.05% : 0.000871s : 1: opt_b 8.04% : 0.132418s : 1: optimize 0.00% : 0.000048s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000018s : 1: order_py_execute_after_rewriter 0.00% : 0.000056s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000013s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000007s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000012s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000015s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000007s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.02% : 0.000253s : 1: pre_auto_parallel 0.01% : 0.000246s : 1: py_interpret_to_execute 0.00% : 0.000046s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000113s : 1: remove_dup_value 0.38% : 0.006274s : 1: renormalize.infer 4.32% : 0.071203s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.05% : 0.000901s : 1: rewriter_after_opt_a 0.05% : 0.000874s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000022s : 1: swap_dp_allreduce_reducescatter 0.04% : 0.000599s : 1: symbol_engine_optimizer 0.02% : 0.000275s : 1: tuple_transform 71.10% : 1.170885s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape0-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape0-1],max_mem:208.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:43:34.523.326 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 1.10069, [21] [bootstrap]: 0.00083268 [type_inference]: 0.978187 [event_method]: 0.00021061 [auto_monad]: 0.00077722 [graph_reusing]: 5.171e-05 [inline]: 6.89001e-06 [add_attr]: 0.00569845, [1] [add_attr_with_inline]: 0.00567731, [1] [Cycle 1]: 0.00032161, [2] [tag_attr]: 0.00020945 [meta_addattr_fg_expand]: 5.094e-05 [parallel-infer-symbol]: 4.39002e-06 [pre_auto_parallel]: 0.0002586 [insert-virtual-dataset]: 3.3e-06 [parallel-infer-symbol-second]: 1.15001e-06 [dataset_repeat_opt]: 2.91e-06 [pipeline_split]: 1.80001e-06 [optimize]: 0.113201, [53] [py_interpret_to_execute]: 0.00023083 [rewriter_before_opt_a]: 0.00081031 [opt_a]: 0.105606, [2] [Cycle 1]: 0.103417, [45] [expand_dump_flag]: 1.947e-05 [switch_simplify]: 0.00081165 [loop_unroll]: 0.0296436 [a_1]: 0.0233381 [with_stream_mark]: 6.804e-05 [recompute_prepare]: 4.82e-05 [updatestate_depend_eliminate]: 2.26e-05 [updatestate_assign_eliminate]: 1.633e-05 [updatestate_loads_eliminate]: 2.039e-05 [parameter_eliminate]: 3.78999e-06 [a_2]: 0.00086388 [accelerated_algorithm]: 8.84e-05 [shard]: 3.01001e-06 [meta_shard_fg_expand]: 2.99e-05 [shard_inline]: 2.739e-05 [merge_send_recv]: 2.494e-05 [auto_parallel]: 2.402e-05 [parallel]: 4.242e-05 [flash_sp]: 1.675e-05 [merge_comm]: 1.513e-05 [allreduce_fusion]: 1.245e-05 [matmul_add_comm_reduction]: 2.508e-05 [allreduce_slice_to_reducescatter]: 9.5999e-07 [virtual_shard_identity]: 3.521e-05 [virtual_dataset]: 2.478e-05 [get_grad_eliminate_]: 2.306e-05 [virtual_output]: 2.347e-05 [merge_forward]: 1.602e-05 [cell_reuse_recompute_pass]: 2.71e-06 [offload_activation]: 2.755e-05 [cell_reuse_handle_not_recompute_node_pass]: 5.746e-05 [merge_recompute_call_nodes]: 1.94e-06 [before_grad]: 4.355e-05 [set_forward_comm_id_for_comm_node_pass]: 2.045e-05 [meta_fg_expand]: 1.787e-05 [flash_sp_send_recv_attached]: 8.26002e-06 [receive_attached]: 2.19999e-06 [after_resolve]: 4.067e-05 [a_after_grad]: 4.253e-05 [renormalize]: 0.0456919 [add_forward_monad_depend]: 1.648e-05 [auto_monad_grad]: 3.65e-06 [auto_monad_eliminator]: 9.055e-05 [cse]: 0.00122958 [a_3]: 0.00017618 [Cycle 2]: 0.00216818, [45] [expand_dump_flag]: 3.31001e-06 [switch_simplify]: 2.658e-05 [loop_unroll]: 2.192e-05 [a_1]: 0.00065707 [with_stream_mark]: 3.566e-05 [recompute_prepare]: 2.233e-05 [updatestate_depend_eliminate]: 1.596e-05 [updatestate_assign_eliminate]: 1.244e-05 [updatestate_loads_eliminate]: 1.373e-05 [parameter_eliminate]: 2.63998e-06 [a_2]: 0.00030964 [accelerated_algorithm]: 3.07e-05 [shard]: 2.59001e-06 [meta_shard_fg_expand]: 6.48e-06 [shard_inline]: 2.062e-05 [merge_send_recv]: 1.968e-05 [auto_parallel]: 2.104e-05 [parallel]: 9.89999e-06 [flash_sp]: 4.36002e-06 [merge_comm]: 1.279e-05 [allreduce_fusion]: 1.206e-05 [matmul_add_comm_reduction]: 2.404e-05 [allreduce_slice_to_reducescatter]: 9.39996e-07 [virtual_shard_identity]: 2.298e-05 [virtual_dataset]: 2.041e-05 [get_grad_eliminate_]: 3.338e-05 [virtual_output]: 2.104e-05 [merge_forward]: 1.333e-05 [cell_reuse_recompute_pass]: 3.4e-06 [offload_activation]: 2.77e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.264e-05 [merge_recompute_call_nodes]: 1.65001e-06 [before_grad]: 3.996e-05 [set_forward_comm_id_for_comm_node_pass]: 1.433e-05 [meta_fg_expand]: 1.073e-05 [flash_sp_send_recv_attached]: 2.03002e-06 [receive_attached]: 2.46998e-06 [after_resolve]: 3.134e-05 [a_after_grad]: 3.504e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 3.93999e-06 [auto_monad_grad]: 2.63e-06 [auto_monad_eliminator]: 5.913e-05 [cse]: 0.000111 [a_3]: 0.00014453 [py_interpret_to_execute_after_opt_a]: 3.947e-05 [slice_cell_reuse_recomputed_activation]: 2.49999e-06 [rewriter_after_opt_a]: 0.00085313 [convert_after_rewriter]: 2.675e-05 [order_py_execute_after_rewriter]: 1.595e-05 [mutable_eliminate]: 0.00172675 [opt_b]: 0.00083109, [1] [Cycle 1]: 0.00082115, [7] [b_1]: 0.0005957 [b_2]: 2.398e-05 [updatestate_depend_eliminate]: 2.191e-05 [updatestate_assign_eliminate]: 1.258e-05 [updatestate_loads_eliminate]: 1.428e-05 [renormalize]: 7.79983e-07 [cse]: 0.00010423 [optimize_parallel_all_gather_comm]: 4.723e-05 [overlap_param_gather]: 2.53e-06 [cconv]: 4.733e-05 [loop_unroll]: 0.00072 [opt_after_cconv]: 0.00034256, [1] [Cycle 1]: 0.00033121, [7] [c_1]: 0.00013586 [parameter_eliminate]: 6.78e-06 [updatestate_depend_eliminate]: 2.458e-05 [updatestate_assign_eliminate]: 1.174e-05 [updatestate_loads_eliminate]: 1.468e-05 [cse]: 8.731e-05 [renormalize]: 1.10999e-06 [remove_dup_value]: 0.0001084 [tuple_transform]: 0.00029521, [1] [Cycle 1]: 0.00028747, [4] [d_1]: 0.00021761 [none_parameter_eliminate]: 3.55998e-06 [renormalize]: 7.50006e-07 [switch_simplify]: 2.855e-05 [partial_unused_args_eliminate]: 2.86e-06 [add_recomputation]: 0.00018054 [cse_after_recomputation]: 9.772e-05, [1] [Cycle 1]: 8.746e-05, [1] [cse]: 7.277e-05 [environ_conv]: 2.934e-05 [swap_dp_allreduce_reducescatter]: 2.293e-05 [bias_add_comm_swap]: 3.51001e-06 [label_micro_interleaved_index]: 7.34002e-06 [label_fine_grained_interleaved_index]: 2.66e-06 [merge_cast_opt]: 1.75001e-06 [slice_recompute_activation]: 2.29999e-06 [micro_interleaved_order_control]: 3.04999e-06 [assign_add_opt]: 1.24e-06 [ForceFp32Comm]: 8.99978e-07 [remove_cast_before_assign_add]: 1.47001e-06 [full_micro_interleaved_order_control]: 2.27999e-06 [reorder_send_recv_between_fp_bp]: 2.61e-06 [comm_op_add_attrs]: 1.24e-06 [add_comm_op_reuse_tag]: 1.00001e-06 [interleave_split_concat_branches]: 1.31002e-06 [interleave_parallel_branches]: 1.16002e-06 [overlap_opt_shard_in_pipeline]: 1.36002e-06 [overlap_opt_shard_grad_in_pipeline]: 2.12001e-06 [control_data_broadcast_order]: 4.338e-05 [grouped_pairwise_exchange_alltoall]: 2.01e-06 [offloading_packed_experts]: 1.264e-05 [overlap_recompute_and_grad_model_parallel]: 1.06e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.43002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.43002e-06 [overlap_recompute_comm]: 2.63e-06 [overlap_grad_ring_attention]: 9.71e-06 [overlap_grad_flash_sp]: 5.786e-05 [begin_end_overlap_inline]: 7.90023e-07 [split_matmul_comm_elemetwise]: 2.57001e-06 [split_layernorm_comm]: 1.86e-06 [handle_group_info]: 1.01002e-06 [symbol_engine_optimizer]: 0.00061121, [1] [Cycle 1]: 0.00060302, [6] [build]: 0.00032922 [elim_shapecalc]: 4.706e-05 [elim_not_effective]: 5.839e-05 [opt_reshape]: 3.68e-05 [fold_const_symbol]: 6.059e-05 [renormalize]: 8.30012e-07 [detach_backward]: 2.78e-06 [pipeline_parallel_scheduler]: 1.54e-06 [auto_monad_reorder]: 6.914e-05 [get_jit_bprop_graph]: 2.30002e-06 [rewriter_after_jit_bprop_graph]: 8.2e-06 [opt_after_jit_grad]: 0.00091384 [validate]: 0.00012432 Sums bootstrap : 0.000833s : 0.08% type_inference : 0.978187s : 89.46% event_method : 0.000211s : 0.02% auto_monad : 0.000777s : 0.07% graph_reusing : 0.000052s : 0.00% inline : 0.000007s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000209s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000051s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000259s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000231s : 0.02% optimize.rewriter_before_opt_a : 0.000810s : 0.07% optimize.opt_a.expand_dump_flag : 0.000023s : 0.00% optimize.opt_a.switch_simplify : 0.000838s : 0.08% optimize.opt_a.loop_unroll : 0.029666s : 2.71% optimize.opt_a.a_1 : 0.023995s : 2.19% optimize.opt_a.with_stream_mark : 0.000104s : 0.01% optimize.opt_a.recompute_prepare : 0.000071s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000039s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000029s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000034s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001174s : 0.11% optimize.opt_a.accelerated_algorithm : 0.000119s : 0.01% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000036s : 0.00% optimize.opt_a.shard_inline : 0.000048s : 0.00% optimize.opt_a.merge_send_recv : 0.000045s : 0.00% optimize.opt_a.auto_parallel : 0.000045s : 0.00% optimize.opt_a.parallel : 0.000052s : 0.00% optimize.opt_a.flash_sp : 0.000021s : 0.00% optimize.opt_a.merge_comm : 0.000028s : 0.00% optimize.opt_a.allreduce_fusion : 0.000025s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000049s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000058s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000056s : 0.01% optimize.opt_a.virtual_output : 0.000045s : 0.00% optimize.opt_a.merge_forward : 0.000029s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000055s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000100s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000084s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000035s : 0.00% optimize.opt_a.meta_fg_expand : 0.000029s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000010s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000072s : 0.01% optimize.opt_a.a_after_grad : 0.000078s : 0.01% optimize.opt_a.renormalize : 0.045692s : 4.18% optimize.opt_a.add_forward_monad_depend : 0.000020s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000150s : 0.01% optimize.opt_a.cse : 0.001341s : 0.12% optimize.opt_a.a_3 : 0.000321s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000039s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000853s : 0.08% optimize.convert_after_rewriter : 0.000027s : 0.00% optimize.order_py_execute_after_rewriter : 0.000016s : 0.00% optimize.mutable_eliminate : 0.001727s : 0.16% optimize.opt_b.b_1 : 0.000596s : 0.05% optimize.opt_b.b_2 : 0.000024s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000022s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000013s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000014s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000104s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000047s : 0.00% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000047s : 0.00% optimize.loop_unroll : 0.000720s : 0.07% optimize.opt_after_cconv.c_1 : 0.000136s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000025s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_after_cconv.cse : 0.000087s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000108s : 0.01% optimize.tuple_transform.d_1 : 0.000218s : 0.02% optimize.tuple_transform.none_parameter_eliminate : 0.000004s : 0.00% optimize.tuple_transform.renormalize : 0.000001s : 0.00% optimize.tuple_transform.switch_simplify : 0.000029s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000181s : 0.02% optimize.cse_after_recomputation.cse : 0.000073s : 0.01% optimize.environ_conv : 0.000029s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000023s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000007s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000043s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000013s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000011s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000010s : 0.00% optimize.overlap_grad_flash_sp : 0.000058s : 0.01% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000329s : 0.03% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000047s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000058s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000037s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000061s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000001s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000069s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000914s : 0.08% validate : 0.000124s : 0.01% Time group info: ------[substitution.] 0.016333 569 0.11% : 0.000019s : 8: substitution.depend_value_elim 0.09% : 0.000015s : 14: substitution.elim_not_effective 0.12% : 0.000019s : 12: substitution.float_tuple_getitem_switch 0.17% : 0.000028s : 14: substitution.fold_const_symbol 0.10% : 0.000016s : 17: substitution.graph_param_transform 91.84% : 0.015001s : 71: substitution.inline 0.09% : 0.000015s : 29: substitution.j_node_and_user_rematch 0.27% : 0.000044s : 6: substitution.less_batch_normalization 0.44% : 0.000072s : 2: substitution.list_to_tuple_eliminator_ 0.04% : 0.000007s : 6: substitution.load_eliminater 0.29% : 0.000047s : 21: substitution.minmaximum_grad 0.05% : 0.000009s : 2: substitution.opt_reshape 0.13% : 0.000022s : 29: substitution.remove_not_recompute_node 0.06% : 0.000010s : 4: substitution.replace_old_param 0.53% : 0.000087s : 18: substitution.reshape_eliminate 0.29% : 0.000048s : 24: substitution.switch_simplify 1.11% : 0.000181s : 51: substitution.tuple_list_convert_item_index_to_positive 0.59% : 0.000097s : 51: substitution.tuple_list_get_item_const_eliminator 0.79% : 0.000130s : 51: substitution.tuple_list_get_item_depend_reorder 1.76% : 0.000288s : 73: substitution.tuple_list_get_item_eliminator 0.80% : 0.000130s : 51: substitution.tuple_list_get_set_item_eliminator 0.15% : 0.000025s : 6: substitution.updatestate_pure_node_eliminater 0.15% : 0.000024s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.977707 2 93.16% : 0.910839s : 1: type_inference.infer 6.84% : 0.066868s : 1: type_inference.specialize ------[replace.] 0.001608 115 4.09% : 0.000066s : 5: replace.depend_value_elim 65.54% : 0.001054s : 71: replace.inline 3.39% : 0.000055s : 2: replace.list_to_tuple_eliminator_ 18.74% : 0.000301s : 24: replace.switch_simplify 8.24% : 0.000132s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.015096 115 0.02% : 0.000003s : 5: match.depend_value_elim 98.93% : 0.014935s : 71: match.inline 0.45% : 0.000068s : 2: match.list_to_tuple_eliminator_ 0.24% : 0.000036s : 24: match.switch_simplify 0.36% : 0.000054s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001846 9701 1.15% : 0.000021s : 123: predicate.accumulaten_eliminater 0.35% : 0.000007s : 17: predicate.ad_related_special_op_eliminate 0.44% : 0.000008s : 58: predicate.addn_check_dump 1.26% : 0.000023s : 123: predicate.addn_zero_filter 1.05% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 2.16% : 0.000040s : 181: predicate.arithmetic_simplify 1.04% : 0.000019s : 123: predicate.cast_eliminate 0.27% : 0.000005s : 34: predicate.check_bprop_eliminate 0.44% : 0.000008s : 58: predicate.compare_switch_simplify 0.07% : 0.000001s : 17: predicate.const_output_eliminate 0.57% : 0.000010s : 61: predicate.depend_value_elim 1.07% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 1.20% : 0.000022s : 123: predicate.dict_get_item_eliminator 1.10% : 0.000020s : 123: predicate.dict_set_item_eliminator 0.59% : 0.000011s : 34: predicate.dumpgradient_eliminate 0.10% : 0.000002s : 17: predicate.elim_not_effective 0.21% : 0.000004s : 17: predicate.elim_shapecalc_of_broadcastargs 1.10% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.15% : 0.000021s : 140: predicate.environ_get_add_eliminate 1.11% : 0.000020s : 140: predicate.environ_get_depend_swap 1.69% : 0.000031s : 198: predicate.environ_get_eliminate 1.12% : 0.000021s : 140: predicate.environ_get_set_eliminate 1.66% : 0.000031s : 209: predicate.exchange_switch_depend_value 3.79% : 0.000070s : 209: predicate.float_depend_g_call 0.46% : 0.000009s : 58: predicate.float_environ_get_switch 0.62% : 0.000011s : 75: predicate.float_tuple_getitem_switch 0.07% : 0.000001s : 17: predicate.fold_const_symbol 0.32% : 0.000006s : 35: predicate.get_grad_eliminate 0.11% : 0.000002s : 17: predicate.graph_param_transform 0.48% : 0.000009s : 58: predicate.incorporate_call 0.42% : 0.000008s : 58: predicate.incorporate_call_switch 5.87% : 0.000108s : 459: predicate.inline 0.49% : 0.000009s : 35: predicate.inline_without_move 0.14% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.49% : 0.000009s : 36: predicate.less_batch_normalization 1.44% : 0.000027s : 172: predicate.list_to_tuple_eliminator_ 2.36% : 0.000044s : 295: predicate.load_eliminater 0.39% : 0.000007s : 17: predicate.loop_unroll_after_grad 6.64% : 0.000123s : 314: predicate.loop_unroll_before_grad 1.44% : 0.000027s : 157: predicate.make_slice_get_slice_eliminator 0.46% : 0.000008s : 58: predicate.merge_addn 0.27% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.28% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.07% : 0.000020s : 123: predicate.minmaximum_grad 0.53% : 0.000010s : 17: predicate.mutable_eliminate 0.18% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 3.77% : 0.000070s : 209: predicate.partial_defer_inline 1.34% : 0.000025s : 155: predicate.partial_eliminate 1.01% : 0.000019s : 123: predicate.print_const_string_wrapper 0.45% : 0.000008s : 53: predicate.reduce_all_const_elim 1.40% : 0.000026s : 123: predicate.reduce_eliminate 2.32% : 0.000043s : 295: predicate.redundant_stop_gradient_eliminater 0.19% : 0.000003s : 35: predicate.remove_not_recompute_node 1.00% : 0.000018s : 172: predicate.replace_applicator 0.17% : 0.000003s : 35: predicate.replace_old_param 0.11% : 0.000002s : 17: predicate.reset_defer_inline 1.08% : 0.000020s : 123: predicate.reshape_eliminate 0.27% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.16% : 0.000003s : 17: predicate.row_tensor_eliminate 0.44% : 0.000008s : 34: predicate.same_eliminate 0.21% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.40% : 0.000007s : 35: predicate.shard_identity_eliminate 0.32% : 0.000006s : 34: predicate.special_op_eliminate 0.58% : 0.000011s : 58: predicate.specialize_transform 0.35% : 0.000006s : 34: predicate.split_environ_get_set_with_tuple_value 0.40% : 0.000007s : 35: predicate.stack_unstack_eliminate 0.15% : 0.000003s : 17: predicate.switch_call_monad_eliminater 4.89% : 0.000090s : 209: predicate.switch_defer_inline 2.23% : 0.000041s : 243: predicate.switch_layer_defer_inline 5.85% : 0.000108s : 646: predicate.switch_simplify 1.07% : 0.000020s : 123: predicate.tile_eliminate 1.07% : 0.000020s : 123: predicate.transpose_eliminate 1.54% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.52% : 0.000028s : 157: predicate.tuple_list_get_item_const_eliminator 1.48% : 0.000027s : 157: predicate.tuple_list_get_item_depend_reorder 3.45% : 0.000064s : 228: predicate.tuple_list_get_item_eliminator 1.53% : 0.000028s : 157: predicate.tuple_list_get_set_item_eliminator 2.09% : 0.000038s : 215: predicate.tuple_list_set_item_eliminator 1.45% : 0.000027s : 170: predicate.tuple_to_list_eliminator_ 2.34% : 0.000043s : 295: predicate.updatestate_pure_node_eliminater 2.90% : 0.000054s : 353: predicate.updatestate_useless_node_eliminater 0.14% : 0.000003s : 17: predicate.value_based_eliminate 0.32% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.34% : 0.000006s : 35: predicate.virtual_output_eliminate 0.14% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.19% : 0.000004s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.081425 135 91.50% : 0.074500s : 55: func_graph_cloner_run.FuncGraphClonerGraph 8.50% : 0.006924s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.322912 209 0.00% : 0.000003s : 1: ForceFp32Comm 0.43% : 0.005707s : 1: add_attr 0.43% : 0.005683s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000193s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.06% : 0.000801s : 1: auto_monad 0.01% : 0.000079s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.07% : 0.000925s : 1: bootstrap 0.00% : 0.000052s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000050s : 1: control_data_broadcast_order 0.00% : 0.000034s : 1: convert_after_rewriter 0.01% : 0.000102s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.00% : 0.000034s : 1: environ_conv 0.02% : 0.000226s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000060s : 1: graph_reusing 0.00% : 0.000006s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000006s : 1: handle_group_info 0.00% : 0.000011s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000011s : 1: label_micro_interleaved_index 0.06% : 0.000739s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.13% : 0.001748s : 1: mutable_eliminate 0.00% : 0.000016s : 1: offloading_packed_experts 0.00% : 0.000045s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000052s : 1: opt.transform.mutable_eliminate 4.28% : 0.056592s : 95: opt.transform.opt_a 0.01% : 0.000134s : 1: opt.transform.opt_after_cconv 0.01% : 0.000088s : 1: opt.transform.opt_after_jit_grad 0.04% : 0.000579s : 28: opt.transform.opt_b 0.02% : 0.000240s : 2: opt.transform.opt_trans_graph 0.01% : 0.000194s : 4: opt.transform.symbol_engine_opt 7.98% : 0.105611s : 1: opt_a 0.03% : 0.000347s : 1: opt_after_cconv 0.07% : 0.000934s : 1: opt_after_jit_grad 0.06% : 0.000835s : 1: opt_b 8.56% : 0.113208s : 1: optimize 0.00% : 0.000052s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000020s : 1: order_py_execute_after_rewriter 0.00% : 0.000065s : 1: overlap_grad_flash_sp 0.00% : 0.000005s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000013s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000007s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000014s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000007s : 1: overlap_recompute_comm 0.00% : 0.000011s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000007s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.02% : 0.000270s : 1: pre_auto_parallel 0.02% : 0.000244s : 1: py_interpret_to_execute 0.00% : 0.000044s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000119s : 1: remove_dup_value 2.95% : 0.038970s : 1: renormalize.infer 0.51% : 0.006700s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000012s : 1: rewriter_after_jit_bprop_graph 0.07% : 0.000865s : 1: rewriter_after_opt_a 0.06% : 0.000828s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000027s : 1: swap_dp_allreduce_reducescatter 0.05% : 0.000616s : 1: symbol_engine_optimizer 0.02% : 0.000300s : 1: tuple_transform 73.94% : 0.978223s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape0-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape0-16],max_mem:208.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:43:46.501.681 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 1.1822, [21] [bootstrap]: 0.00084917 [type_inference]: 0.959493 [event_method]: 0.00020047 [auto_monad]: 0.00067458 [graph_reusing]: 4.374e-05 [inline]: 3.14001e-06 [add_attr]: 0.00519432, [1] [add_attr_with_inline]: 0.00517477, [1] [Cycle 1]: 0.00032715, [2] [tag_attr]: 0.0002122 [meta_addattr_fg_expand]: 5.896e-05 [parallel-infer-symbol]: 4.25e-06 [pre_auto_parallel]: 0.00026615 [insert-virtual-dataset]: 3.11999e-06 [parallel-infer-symbol-second]: 8.89995e-07 [dataset_repeat_opt]: 2.96999e-06 [pipeline_split]: 1.57001e-06 [optimize]: 0.214068, [53] [py_interpret_to_execute]: 0.00026186 [rewriter_before_opt_a]: 0.00091798 [opt_a]: 0.152457, [2] [Cycle 1]: 0.150143, [45] [expand_dump_flag]: 3.667e-05 [switch_simplify]: 0.0009686 [loop_unroll]: 0.00036665 [a_1]: 0.130832 [with_stream_mark]: 7.211e-05 [recompute_prepare]: 4.864e-05 [updatestate_depend_eliminate]: 2.254e-05 [updatestate_assign_eliminate]: 1.607e-05 [updatestate_loads_eliminate]: 2.176e-05 [parameter_eliminate]: 2.88e-06 [a_2]: 0.0008318 [accelerated_algorithm]: 6.529e-05 [shard]: 2.79001e-06 [meta_shard_fg_expand]: 2.84e-05 [shard_inline]: 2.454e-05 [merge_send_recv]: 2.24e-05 [auto_parallel]: 2.299e-05 [parallel]: 4.116e-05 [flash_sp]: 1.59e-05 [merge_comm]: 1.439e-05 [allreduce_fusion]: 1.28e-05 [matmul_add_comm_reduction]: 2.283e-05 [allreduce_slice_to_reducescatter]: 9.09989e-07 [virtual_shard_identity]: 2.789e-05 [virtual_dataset]: 2.378e-05 [get_grad_eliminate_]: 2.203e-05 [virtual_output]: 2.33e-05 [merge_forward]: 1.443e-05 [cell_reuse_recompute_pass]: 3.41999e-06 [offload_activation]: 2.631e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.917e-05 [merge_recompute_call_nodes]: 1.54998e-06 [before_grad]: 4.148e-05 [set_forward_comm_id_for_comm_node_pass]: 1.379e-05 [meta_fg_expand]: 2.35e-05 [flash_sp_send_recv_attached]: 6.02999e-06 [receive_attached]: 2.45002e-06 [after_resolve]: 3.192e-05 [a_after_grad]: 3.782e-05 [renormalize]: 0.0132675 [add_forward_monad_depend]: 1.318e-05 [auto_monad_grad]: 3.56001e-06 [auto_monad_eliminator]: 8.981e-05 [cse]: 0.00224356 [a_3]: 0.00018498 [Cycle 2]: 0.00229297, [45] [expand_dump_flag]: 3.49001e-06 [switch_simplify]: 2.741e-05 [loop_unroll]: 2.187e-05 [a_1]: 0.00067448 [with_stream_mark]: 4.264e-05 [recompute_prepare]: 2.663e-05 [updatestate_depend_eliminate]: 1.633e-05 [updatestate_assign_eliminate]: 1.181e-05 [updatestate_loads_eliminate]: 1.498e-05 [parameter_eliminate]: 2.99999e-06 [a_2]: 0.00032031 [accelerated_algorithm]: 3.278e-05 [shard]: 3.16001e-06 [meta_shard_fg_expand]: 9.31e-06 [shard_inline]: 2.245e-05 [merge_send_recv]: 2.207e-05 [auto_parallel]: 2.235e-05 [parallel]: 1.223e-05 [flash_sp]: 5.40999e-06 [merge_comm]: 1.371e-05 [allreduce_fusion]: 1.243e-05 [matmul_add_comm_reduction]: 2.37e-05 [allreduce_slice_to_reducescatter]: 8.50006e-07 [virtual_shard_identity]: 2.633e-05 [virtual_dataset]: 2.141e-05 [get_grad_eliminate_]: 2.2e-05 [virtual_output]: 2.082e-05 [merge_forward]: 1.362e-05 [cell_reuse_recompute_pass]: 3.70998e-06 [offload_activation]: 2.739e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.777e-05 [merge_recompute_call_nodes]: 1.63002e-06 [before_grad]: 5.61e-05 [set_forward_comm_id_for_comm_node_pass]: 1.799e-05 [meta_fg_expand]: 9.99999e-06 [flash_sp_send_recv_attached]: 1.94e-06 [receive_attached]: 2.99999e-06 [after_resolve]: 3.279e-05 [a_after_grad]: 3.433e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 4.16001e-06 [auto_monad_grad]: 3.40998e-06 [auto_monad_eliminator]: 7.453e-05 [cse]: 9.331e-05 [a_3]: 0.00014232 [py_interpret_to_execute_after_opt_a]: 3.793e-05 [slice_cell_reuse_recomputed_activation]: 2.22001e-06 [rewriter_after_opt_a]: 0.00089509 [convert_after_rewriter]: 3.31e-05 [order_py_execute_after_rewriter]: 1.533e-05 [mutable_eliminate]: 0.00086803 [opt_b]: 0.00077188, [1] [Cycle 1]: 0.00076255, [7] [b_1]: 0.00055708 [b_2]: 2.368e-05 [updatestate_depend_eliminate]: 1.956e-05 [updatestate_assign_eliminate]: 1.164e-05 [updatestate_loads_eliminate]: 1.511e-05 [renormalize]: 5.99975e-07 [cse]: 9.067e-05 [optimize_parallel_all_gather_comm]: 4.358e-05 [overlap_param_gather]: 1.96003e-06 [cconv]: 4.521e-05 [loop_unroll]: 0.00053471 [opt_after_cconv]: 0.0552613, [1] [Cycle 1]: 0.0552492, [7] [c_1]: 0.0549307 [parameter_eliminate]: 9.17001e-06 [updatestate_depend_eliminate]: 4.312e-05 [updatestate_assign_eliminate]: 1.334e-05 [updatestate_loads_eliminate]: 1.787e-05 [cse]: 0.00012581 [renormalize]: 1.54998e-06 [remove_dup_value]: 0.00011583 [tuple_transform]: 0.00030932, [1] [Cycle 1]: 0.00029742, [4] [d_1]: 0.00024147 [none_parameter_eliminate]: 2.96999e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 2.35e-05 [partial_unused_args_eliminate]: 2.74001e-06 [add_recomputation]: 0.00018194 [cse_after_recomputation]: 8.281e-05, [1] [Cycle 1]: 7.505e-05, [1] [cse]: 6.564e-05 [environ_conv]: 2.923e-05 [swap_dp_allreduce_reducescatter]: 2.031e-05 [bias_add_comm_swap]: 3.59002e-06 [label_micro_interleaved_index]: 8.2e-06 [label_fine_grained_interleaved_index]: 2.84999e-06 [merge_cast_opt]: 1.91e-06 [slice_recompute_activation]: 2.37999e-06 [micro_interleaved_order_control]: 2.90998e-06 [assign_add_opt]: 1.63997e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.15001e-06 [full_micro_interleaved_order_control]: 2.19999e-06 [reorder_send_recv_between_fp_bp]: 3.24001e-06 [comm_op_add_attrs]: 1.14e-06 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.25001e-06 [interleave_parallel_branches]: 1.15001e-06 [overlap_opt_shard_in_pipeline]: 1.28002e-06 [overlap_opt_shard_grad_in_pipeline]: 2.35002e-06 [control_data_broadcast_order]: 4.032e-05 [grouped_pairwise_exchange_alltoall]: 1.64e-06 [offloading_packed_experts]: 1.157e-05 [overlap_recompute_and_grad_model_parallel]: 1.065e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.42e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40999e-06 [overlap_recompute_comm]: 2.93998e-06 [overlap_grad_ring_attention]: 9.44e-06 [overlap_grad_flash_sp]: 5.624e-05 [begin_end_overlap_inline]: 5.59987e-07 [split_matmul_comm_elemetwise]: 2.29001e-06 [split_layernorm_comm]: 1.81003e-06 [handle_group_info]: 1.16997e-06 [symbol_engine_optimizer]: 0.00057626, [1] [Cycle 1]: 0.00056863, [6] [build]: 0.0003087 [elim_shapecalc]: 4.058e-05 [elim_not_effective]: 5.711e-05 [opt_reshape]: 3.678e-05 [fold_const_symbol]: 6.218e-05 [renormalize]: 4.10015e-07 [detach_backward]: 4.03001e-06 [pipeline_parallel_scheduler]: 1.64998e-06 [auto_monad_reorder]: 7.003e-05 [get_jit_bprop_graph]: 2.37001e-06 [rewriter_after_jit_bprop_graph]: 6.53e-06 [opt_after_jit_grad]: 0.00090224 [validate]: 0.00011754 Sums bootstrap : 0.000849s : 0.07% type_inference : 0.959493s : 81.63% event_method : 0.000200s : 0.02% auto_monad : 0.000675s : 0.06% graph_reusing : 0.000044s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000212s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000059s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000266s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000262s : 0.02% optimize.rewriter_before_opt_a : 0.000918s : 0.08% optimize.opt_a.expand_dump_flag : 0.000040s : 0.00% optimize.opt_a.switch_simplify : 0.000996s : 0.08% optimize.opt_a.loop_unroll : 0.000389s : 0.03% optimize.opt_a.a_1 : 0.131506s : 11.19% optimize.opt_a.with_stream_mark : 0.000115s : 0.01% optimize.opt_a.recompute_prepare : 0.000075s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000039s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000028s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000037s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001152s : 0.10% optimize.opt_a.accelerated_algorithm : 0.000098s : 0.01% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000038s : 0.00% optimize.opt_a.shard_inline : 0.000047s : 0.00% optimize.opt_a.merge_send_recv : 0.000044s : 0.00% optimize.opt_a.auto_parallel : 0.000045s : 0.00% optimize.opt_a.parallel : 0.000053s : 0.00% optimize.opt_a.flash_sp : 0.000021s : 0.00% optimize.opt_a.merge_comm : 0.000028s : 0.00% optimize.opt_a.allreduce_fusion : 0.000025s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000047s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000054s : 0.00% optimize.opt_a.virtual_dataset : 0.000045s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.00% optimize.opt_a.virtual_output : 0.000044s : 0.00% optimize.opt_a.merge_forward : 0.000028s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000007s : 0.00% optimize.opt_a.offload_activation : 0.000054s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000097s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000098s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000032s : 0.00% optimize.opt_a.meta_fg_expand : 0.000033s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000065s : 0.01% optimize.opt_a.a_after_grad : 0.000072s : 0.01% optimize.opt_a.renormalize : 0.013268s : 1.13% optimize.opt_a.add_forward_monad_depend : 0.000017s : 0.00% optimize.opt_a.auto_monad_grad : 0.000007s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000164s : 0.01% optimize.opt_a.cse : 0.002337s : 0.20% optimize.opt_a.a_3 : 0.000327s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000038s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000895s : 0.08% optimize.convert_after_rewriter : 0.000033s : 0.00% optimize.order_py_execute_after_rewriter : 0.000015s : 0.00% optimize.mutable_eliminate : 0.000868s : 0.07% optimize.opt_b.b_1 : 0.000557s : 0.05% optimize.opt_b.b_2 : 0.000024s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000020s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000091s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000044s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000045s : 0.00% optimize.loop_unroll : 0.000535s : 0.05% optimize.opt_after_cconv.c_1 : 0.054931s : 4.67% optimize.opt_after_cconv.parameter_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000043s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000018s : 0.00% optimize.opt_after_cconv.cse : 0.000126s : 0.01% optimize.opt_after_cconv.renormalize : 0.000002s : 0.00% optimize.remove_dup_value : 0.000116s : 0.01% optimize.tuple_transform.d_1 : 0.000241s : 0.02% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000024s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000182s : 0.02% optimize.cse_after_recomputation.cse : 0.000066s : 0.01% optimize.environ_conv : 0.000029s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000020s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000008s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000040s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000012s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000011s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000056s : 0.00% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000309s : 0.03% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000041s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000057s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000037s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000062s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000004s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000070s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000902s : 0.08% validate : 0.000118s : 0.01% Time group info: ------[substitution.] 0.004159 569 0.44% : 0.000018s : 8: substitution.depend_value_elim 0.35% : 0.000015s : 14: substitution.elim_not_effective 0.46% : 0.000019s : 12: substitution.float_tuple_getitem_switch 0.71% : 0.000029s : 14: substitution.fold_const_symbol 0.42% : 0.000018s : 17: substitution.graph_param_transform 69.27% : 0.002881s : 71: substitution.inline 0.36% : 0.000015s : 29: substitution.j_node_and_user_rematch 0.85% : 0.000035s : 6: substitution.less_batch_normalization 1.50% : 0.000062s : 2: substitution.list_to_tuple_eliminator_ 0.16% : 0.000007s : 6: substitution.load_eliminater 1.21% : 0.000050s : 21: substitution.minmaximum_grad 0.26% : 0.000011s : 2: substitution.opt_reshape 0.50% : 0.000021s : 29: substitution.remove_not_recompute_node 0.25% : 0.000010s : 4: substitution.replace_old_param 1.75% : 0.000073s : 18: substitution.reshape_eliminate 1.57% : 0.000065s : 24: substitution.switch_simplify 4.16% : 0.000173s : 51: substitution.tuple_list_convert_item_index_to_positive 2.41% : 0.000100s : 51: substitution.tuple_list_get_item_const_eliminator 3.14% : 0.000131s : 51: substitution.tuple_list_get_item_depend_reorder 6.00% : 0.000250s : 73: substitution.tuple_list_get_item_eliminator 3.01% : 0.000125s : 51: substitution.tuple_list_get_set_item_eliminator 0.72% : 0.000030s : 6: substitution.updatestate_pure_node_eliminater 0.50% : 0.000021s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.959050 2 93.77% : 0.899317s : 1: type_inference.infer 6.23% : 0.059733s : 1: type_inference.specialize ------[replace.] 0.001332 115 5.49% : 0.000073s : 5: replace.depend_value_elim 55.41% : 0.000738s : 71: replace.inline 3.45% : 0.000046s : 2: replace.list_to_tuple_eliminator_ 26.03% : 0.000347s : 24: replace.switch_simplify 9.63% : 0.000128s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002983 115 0.08% : 0.000002s : 5: match.depend_value_elim 94.94% : 0.002832s : 71: match.inline 1.96% : 0.000059s : 2: match.list_to_tuple_eliminator_ 1.62% : 0.000048s : 24: match.switch_simplify 1.40% : 0.000042s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.056457 9701 0.04% : 0.000021s : 123: predicate.accumulaten_eliminater 0.01% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.01% : 0.000008s : 58: predicate.addn_check_dump 0.03% : 0.000019s : 123: predicate.addn_zero_filter 0.03% : 0.000017s : 123: predicate.adjust_all_reduce_mul_add 0.07% : 0.000040s : 181: predicate.arithmetic_simplify 0.04% : 0.000021s : 123: predicate.cast_eliminate 0.01% : 0.000005s : 34: predicate.check_bprop_eliminate 0.01% : 0.000008s : 58: predicate.compare_switch_simplify 0.00% : 0.000001s : 17: predicate.const_output_eliminate 0.02% : 0.000010s : 61: predicate.depend_value_elim 0.03% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 0.04% : 0.000021s : 123: predicate.dict_get_item_eliminator 0.03% : 0.000018s : 123: predicate.dict_set_item_eliminator 0.02% : 0.000011s : 34: predicate.dumpgradient_eliminate 0.00% : 0.000002s : 17: predicate.elim_not_effective 0.01% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 0.04% : 0.000021s : 140: predicate.environ_add_const_eliminate 0.05% : 0.000028s : 140: predicate.environ_get_add_eliminate 0.04% : 0.000020s : 140: predicate.environ_get_depend_swap 0.05% : 0.000030s : 198: predicate.environ_get_eliminate 0.04% : 0.000020s : 140: predicate.environ_get_set_eliminate 0.06% : 0.000031s : 209: predicate.exchange_switch_depend_value 0.08% : 0.000046s : 209: predicate.float_depend_g_call 0.01% : 0.000008s : 58: predicate.float_environ_get_switch 0.02% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.00% : 0.000001s : 17: predicate.fold_const_symbol 0.01% : 0.000006s : 35: predicate.get_grad_eliminate 0.00% : 0.000002s : 17: predicate.graph_param_transform 0.01% : 0.000008s : 58: predicate.incorporate_call 0.01% : 0.000008s : 58: predicate.incorporate_call_switch 0.18% : 0.000104s : 459: predicate.inline 0.01% : 0.000007s : 35: predicate.inline_without_move 0.00% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.01% : 0.000008s : 36: predicate.less_batch_normalization 0.05% : 0.000030s : 172: predicate.list_to_tuple_eliminator_ 0.08% : 0.000044s : 295: predicate.load_eliminater 0.01% : 0.000005s : 17: predicate.loop_unroll_after_grad 0.13% : 0.000072s : 314: predicate.loop_unroll_before_grad 0.05% : 0.000030s : 157: predicate.make_slice_get_slice_eliminator 0.01% : 0.000008s : 58: predicate.merge_addn 0.01% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.01% : 0.000005s : 34: predicate.mini_step_allgather_replace 0.03% : 0.000020s : 123: predicate.minmaximum_grad 0.01% : 0.000007s : 17: predicate.mutable_eliminate 0.01% : 0.000003s : 17: predicate.opt_reshape 0.01% : 0.000003s : 17: predicate.parallel_virtual_node 0.11% : 0.000063s : 209: predicate.partial_defer_inline 96.99% : 0.054758s : 155: predicate.partial_eliminate 0.04% : 0.000020s : 123: predicate.print_const_string_wrapper 0.01% : 0.000008s : 53: predicate.reduce_all_const_elim 0.05% : 0.000028s : 123: predicate.reduce_eliminate 0.09% : 0.000048s : 295: predicate.redundant_stop_gradient_eliminater 0.01% : 0.000003s : 35: predicate.remove_not_recompute_node 0.03% : 0.000018s : 172: predicate.replace_applicator 0.01% : 0.000004s : 35: predicate.replace_old_param 0.00% : 0.000002s : 17: predicate.reset_defer_inline 0.04% : 0.000022s : 123: predicate.reshape_eliminate 0.01% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.01% : 0.000003s : 17: predicate.row_tensor_eliminate 0.02% : 0.000009s : 34: predicate.same_eliminate 0.01% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.01% : 0.000007s : 35: predicate.shard_identity_eliminate 0.01% : 0.000005s : 34: predicate.special_op_eliminate 0.02% : 0.000010s : 58: predicate.specialize_transform 0.01% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.01% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.01% : 0.000003s : 17: predicate.switch_call_monad_eliminater 0.07% : 0.000037s : 209: predicate.switch_defer_inline 0.07% : 0.000039s : 243: predicate.switch_layer_defer_inline 0.25% : 0.000140s : 646: predicate.switch_simplify 0.03% : 0.000018s : 123: predicate.tile_eliminate 0.03% : 0.000018s : 123: predicate.transpose_eliminate 0.05% : 0.000030s : 157: predicate.tuple_list_convert_item_index_to_positive 0.05% : 0.000029s : 157: predicate.tuple_list_get_item_const_eliminator 0.05% : 0.000026s : 157: predicate.tuple_list_get_item_depend_reorder 0.09% : 0.000052s : 228: predicate.tuple_list_get_item_eliminator 0.05% : 0.000027s : 157: predicate.tuple_list_get_set_item_eliminator 0.07% : 0.000039s : 215: predicate.tuple_list_set_item_eliminator 0.04% : 0.000025s : 170: predicate.tuple_to_list_eliminator_ 0.08% : 0.000042s : 295: predicate.updatestate_pure_node_eliminater 0.10% : 0.000054s : 353: predicate.updatestate_useless_node_eliminater 0.00% : 0.000003s : 17: predicate.value_based_eliminate 0.01% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.01% : 0.000006s : 35: predicate.virtual_output_eliminate 0.00% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.01% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.149696 135 95.65% : 0.143177s : 55: func_graph_cloner_run.FuncGraphClonerGraph 4.35% : 0.006519s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.605451 209 0.00% : 0.000004s : 1: ForceFp32Comm 0.32% : 0.005202s : 1: add_attr 0.32% : 0.005181s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000188s : 1: add_recomputation 0.00% : 0.000006s : 1: assign_add_opt 0.04% : 0.000690s : 1: auto_monad 0.00% : 0.000078s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.06% : 0.000937s : 1: bootstrap 0.00% : 0.000049s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000045s : 1: control_data_broadcast_order 0.00% : 0.000041s : 1: convert_after_rewriter 0.01% : 0.000086s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.00% : 0.000033s : 1: environ_conv 0.01% : 0.000212s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000052s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000012s : 1: label_micro_interleaved_index 0.03% : 0.000546s : 1: loop_unroll 0.00% : 0.000007s : 1: merge_cast_opt 0.00% : 0.000008s : 1: micro_interleaved_order_control 0.06% : 0.000884s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.00% : 0.000034s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000045s : 1: opt.transform.mutable_eliminate 8.41% : 0.134942s : 95: opt.transform.opt_a 3.42% : 0.054924s : 1: opt.transform.opt_after_cconv 0.00% : 0.000079s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000542s : 28: opt.transform.opt_b 0.02% : 0.000260s : 2: opt.transform.opt_trans_graph 0.01% : 0.000189s : 4: opt.transform.symbol_engine_opt 9.50% : 0.152464s : 1: opt_a 3.44% : 0.055269s : 1: opt_after_cconv 0.06% : 0.000918s : 1: opt_after_jit_grad 0.05% : 0.000777s : 1: opt_b 13.33% : 0.214075s : 1: optimize 0.00% : 0.000048s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000019s : 1: order_py_execute_after_rewriter 0.00% : 0.000061s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000014s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000007s : 1: overlap_recompute_comm 0.00% : 0.000010s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.02% : 0.000277s : 1: pre_auto_parallel 0.02% : 0.000273s : 1: py_interpret_to_execute 0.00% : 0.000043s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000123s : 1: remove_dup_value 0.43% : 0.006876s : 1: renormalize.infer 0.40% : 0.006367s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.06% : 0.000912s : 1: rewriter_after_opt_a 0.06% : 0.000943s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000069s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000024s : 1: swap_dp_allreduce_reducescatter 0.04% : 0.000580s : 1: symbol_engine_optimizer 0.02% : 0.000314s : 1: tuple_transform 59.77% : 0.959524s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape1-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape1-1],max_mem:140.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:44:00.809.116 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 0.793063, [21] [bootstrap]: 0.00074545 [type_inference]: 0.674041 [event_method]: 0.00022407 [auto_monad]: 0.00078312 [graph_reusing]: 5.23e-05 [inline]: 3.26001e-06 [add_attr]: 0.00540684, [1] [add_attr_with_inline]: 0.00538948, [1] [Cycle 1]: 0.00031374, [2] [tag_attr]: 0.00020962 [meta_addattr_fg_expand]: 5.046e-05 [parallel-infer-symbol]: 4.55001e-06 [pre_auto_parallel]: 0.00024289 [insert-virtual-dataset]: 2.42001e-06 [parallel-infer-symbol-second]: 1.10999e-06 [dataset_repeat_opt]: 2.36998e-06 [pipeline_split]: 1.57999e-06 [optimize]: 0.110307, [53] [py_interpret_to_execute]: 0.00022416 [rewriter_before_opt_a]: 0.00082981 [opt_a]: 0.0789024, [2] [Cycle 1]: 0.0768071, [45] [expand_dump_flag]: 1.696e-05 [switch_simplify]: 0.00080289 [loop_unroll]: 0.00032081 [a_1]: 0.0460677 [with_stream_mark]: 0.00013494 [recompute_prepare]: 4.775e-05 [updatestate_depend_eliminate]: 2.415e-05 [updatestate_assign_eliminate]: 1.607e-05 [updatestate_loads_eliminate]: 2.121e-05 [parameter_eliminate]: 3.91001e-06 [a_2]: 0.00083886 [accelerated_algorithm]: 7.919e-05 [shard]: 3.26001e-06 [meta_shard_fg_expand]: 2.563e-05 [shard_inline]: 2.655e-05 [merge_send_recv]: 2.237e-05 [auto_parallel]: 2.184e-05 [parallel]: 3.635e-05 [flash_sp]: 1.666e-05 [merge_comm]: 1.51e-05 [allreduce_fusion]: 1.253e-05 [matmul_add_comm_reduction]: 2.313e-05 [allreduce_slice_to_reducescatter]: 9.79984e-07 [virtual_shard_identity]: 3.221e-05 [virtual_dataset]: 2.452e-05 [get_grad_eliminate_]: 2.146e-05 [virtual_output]: 2.201e-05 [merge_forward]: 1.508e-05 [cell_reuse_recompute_pass]: 2.69001e-06 [offload_activation]: 2.541e-05 [cell_reuse_handle_not_recompute_node_pass]: 5.569e-05 [merge_recompute_call_nodes]: 2.34001e-06 [before_grad]: 4.32e-05 [set_forward_comm_id_for_comm_node_pass]: 1.751e-05 [meta_fg_expand]: 1.787e-05 [flash_sp_send_recv_attached]: 6.89001e-06 [receive_attached]: 2.74001e-06 [after_resolve]: 3.338e-05 [a_after_grad]: 3.755e-05 [renormalize]: 0.0259962 [add_forward_monad_depend]: 1.482e-05 [auto_monad_grad]: 2.99001e-06 [auto_monad_eliminator]: 9.131e-05 [cse]: 0.00108149 [a_3]: 0.0001733 [Cycle 2]: 0.00207687, [45] [expand_dump_flag]: 3.66999e-06 [switch_simplify]: 2.562e-05 [loop_unroll]: 2.169e-05 [a_1]: 0.00065242 [with_stream_mark]: 3.143e-05 [recompute_prepare]: 2.326e-05 [updatestate_depend_eliminate]: 1.571e-05 [updatestate_assign_eliminate]: 1.231e-05 [updatestate_loads_eliminate]: 1.384e-05 [parameter_eliminate]: 2.34001e-06 [a_2]: 0.00031014 [accelerated_algorithm]: 3.166e-05 [shard]: 2.21998e-06 [meta_shard_fg_expand]: 6.68e-06 [shard_inline]: 2.001e-05 [merge_send_recv]: 1.996e-05 [auto_parallel]: 2.029e-05 [parallel]: 1.046e-05 [flash_sp]: 4.71002e-06 [merge_comm]: 1.318e-05 [allreduce_fusion]: 1.167e-05 [matmul_add_comm_reduction]: 2.245e-05 [allreduce_slice_to_reducescatter]: 1.08001e-06 [virtual_shard_identity]: 2.269e-05 [virtual_dataset]: 1.993e-05 [get_grad_eliminate_]: 2.162e-05 [virtual_output]: 1.93e-05 [merge_forward]: 1.262e-05 [cell_reuse_recompute_pass]: 3.26999e-06 [offload_activation]: 2.394e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.078e-05 [merge_recompute_call_nodes]: 1.56002e-06 [before_grad]: 3.754e-05 [set_forward_comm_id_for_comm_node_pass]: 1.372e-05 [meta_fg_expand]: 9.17999e-06 [flash_sp_send_recv_attached]: 1.82999e-06 [receive_attached]: 2.54001e-06 [after_resolve]: 2.662e-05 [a_after_grad]: 3.316e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 3.86999e-06 [auto_monad_grad]: 2.08002e-06 [auto_monad_eliminator]: 5.153e-05 [cse]: 7.578e-05 [a_3]: 0.00013627 [py_interpret_to_execute_after_opt_a]: 3.226e-05 [slice_cell_reuse_recomputed_activation]: 2.30002e-06 [rewriter_after_opt_a]: 0.0251447 [convert_after_rewriter]: 3.955e-05 [order_py_execute_after_rewriter]: 1.626e-05 [mutable_eliminate]: 0.0008777 [opt_b]: 0.0008102, [1] [Cycle 1]: 0.0007985, [7] [b_1]: 0.00056108 [b_2]: 2.446e-05 [updatestate_depend_eliminate]: 2.245e-05 [updatestate_assign_eliminate]: 1.204e-05 [updatestate_loads_eliminate]: 1.486e-05 [renormalize]: 9.89996e-07 [cse]: 0.00010928 [optimize_parallel_all_gather_comm]: 4.361e-05 [overlap_param_gather]: 2.16e-06 [cconv]: 4.442e-05 [loop_unroll]: 0.00065676 [opt_after_cconv]: 0.00032543, [1] [Cycle 1]: 0.00031459, [7] [c_1]: 0.00013256 [parameter_eliminate]: 5.96e-06 [updatestate_depend_eliminate]: 1.928e-05 [updatestate_assign_eliminate]: 1.059e-05 [updatestate_loads_eliminate]: 1.418e-05 [cse]: 8.18e-05 [renormalize]: 5.60016e-07 [remove_dup_value]: 0.00011105 [tuple_transform]: 0.0003215, [1] [Cycle 1]: 0.00031417, [4] [d_1]: 0.00025787 [none_parameter_eliminate]: 3.89002e-06 [renormalize]: 5.69999e-07 [switch_simplify]: 2.406e-05 [partial_unused_args_eliminate]: 2.48998e-06 [add_recomputation]: 0.00074054 [cse_after_recomputation]: 8.627e-05, [1] [Cycle 1]: 7.677e-05, [1] [cse]: 6.603e-05 [environ_conv]: 2.551e-05 [swap_dp_allreduce_reducescatter]: 1.865e-05 [bias_add_comm_swap]: 3.35e-06 [label_micro_interleaved_index]: 7.71999e-06 [label_fine_grained_interleaved_index]: 2.93e-06 [merge_cast_opt]: 1.52999e-06 [slice_recompute_activation]: 2.24999e-06 [micro_interleaved_order_control]: 2.34001e-06 [assign_add_opt]: 1.52999e-06 [ForceFp32Comm]: 8.50006e-07 [remove_cast_before_assign_add]: 1.17999e-06 [full_micro_interleaved_order_control]: 2.46e-06 [reorder_send_recv_between_fp_bp]: 2.89999e-06 [comm_op_add_attrs]: 1.25001e-06 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.22999e-06 [interleave_parallel_branches]: 1.09e-06 [overlap_opt_shard_in_pipeline]: 1.44998e-06 [overlap_opt_shard_grad_in_pipeline]: 2.06e-06 [control_data_broadcast_order]: 4.067e-05 [grouped_pairwise_exchange_alltoall]: 1.59998e-06 [offloading_packed_experts]: 1.177e-05 [overlap_recompute_and_grad_model_parallel]: 1.024e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.60001e-06 [overlap_recompute_comm]: 2.18998e-06 [overlap_grad_ring_attention]: 9.95002e-06 [overlap_grad_flash_sp]: 5.049e-05 [begin_end_overlap_inline]: 6.09987e-07 [split_matmul_comm_elemetwise]: 2.54999e-06 [split_layernorm_comm]: 2.01003e-06 [handle_group_info]: 9.50007e-07 [symbol_engine_optimizer]: 0.00053236, [1] [Cycle 1]: 0.0005267, [6] [build]: 0.00028966 [elim_shapecalc]: 4.179e-05 [elim_not_effective]: 5.309e-05 [opt_reshape]: 3.472e-05 [fold_const_symbol]: 6.057e-05 [renormalize]: 2.89991e-07 [detach_backward]: 2.91e-06 [pipeline_parallel_scheduler]: 1.55999e-06 [auto_monad_reorder]: 6.411e-05 [get_jit_bprop_graph]: 2.25002e-06 [rewriter_after_jit_bprop_graph]: 6.84001e-06 [opt_after_jit_grad]: 0.00080313 [validate]: 0.00010819 Sums bootstrap : 0.000745s : 0.09% type_inference : 0.674041s : 85.72% event_method : 0.000224s : 0.03% auto_monad : 0.000783s : 0.10% graph_reusing : 0.000052s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000210s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000050s : 0.01% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000243s : 0.03% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000224s : 0.03% optimize.rewriter_before_opt_a : 0.000830s : 0.11% optimize.opt_a.expand_dump_flag : 0.000021s : 0.00% optimize.opt_a.switch_simplify : 0.000829s : 0.11% optimize.opt_a.loop_unroll : 0.000342s : 0.04% optimize.opt_a.a_1 : 0.046720s : 5.94% optimize.opt_a.with_stream_mark : 0.000166s : 0.02% optimize.opt_a.recompute_prepare : 0.000071s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000040s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000028s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000035s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001149s : 0.15% optimize.opt_a.accelerated_algorithm : 0.000111s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000032s : 0.00% optimize.opt_a.shard_inline : 0.000047s : 0.01% optimize.opt_a.merge_send_recv : 0.000042s : 0.01% optimize.opt_a.auto_parallel : 0.000042s : 0.01% optimize.opt_a.parallel : 0.000047s : 0.01% optimize.opt_a.flash_sp : 0.000021s : 0.00% optimize.opt_a.merge_comm : 0.000028s : 0.00% optimize.opt_a.allreduce_fusion : 0.000024s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000046s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000055s : 0.01% optimize.opt_a.virtual_dataset : 0.000044s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000043s : 0.01% optimize.opt_a.virtual_output : 0.000041s : 0.01% optimize.opt_a.merge_forward : 0.000028s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000049s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000096s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000081s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000031s : 0.00% optimize.opt_a.meta_fg_expand : 0.000027s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000060s : 0.01% optimize.opt_a.a_after_grad : 0.000071s : 0.01% optimize.opt_a.renormalize : 0.025996s : 3.31% optimize.opt_a.add_forward_monad_depend : 0.000019s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000143s : 0.02% optimize.opt_a.cse : 0.001157s : 0.15% optimize.opt_a.a_3 : 0.000310s : 0.04% optimize.py_interpret_to_execute_after_opt_a : 0.000032s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.025145s : 3.20% optimize.convert_after_rewriter : 0.000040s : 0.01% optimize.order_py_execute_after_rewriter : 0.000016s : 0.00% optimize.mutable_eliminate : 0.000878s : 0.11% optimize.opt_b.b_1 : 0.000561s : 0.07% optimize.opt_b.b_2 : 0.000024s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000022s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000109s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000044s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000044s : 0.01% optimize.loop_unroll : 0.000657s : 0.08% optimize.opt_after_cconv.c_1 : 0.000133s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000019s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000014s : 0.00% optimize.opt_after_cconv.cse : 0.000082s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000111s : 0.01% optimize.tuple_transform.d_1 : 0.000258s : 0.03% optimize.tuple_transform.none_parameter_eliminate : 0.000004s : 0.00% optimize.tuple_transform.renormalize : 0.000001s : 0.00% optimize.tuple_transform.switch_simplify : 0.000024s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000741s : 0.09% optimize.cse_after_recomputation.cse : 0.000066s : 0.01% optimize.environ_conv : 0.000026s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000019s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000008s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000041s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000012s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000010s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000010s : 0.00% optimize.overlap_grad_flash_sp : 0.000050s : 0.01% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000290s : 0.04% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000042s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000053s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000035s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000061s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000064s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000803s : 0.10% validate : 0.000108s : 0.01% Time group info: ------[substitution.] 0.003870 569 0.46% : 0.000018s : 8: substitution.depend_value_elim 0.35% : 0.000014s : 14: substitution.elim_not_effective 0.49% : 0.000019s : 12: substitution.float_tuple_getitem_switch 0.73% : 0.000028s : 14: substitution.fold_const_symbol 0.39% : 0.000015s : 17: substitution.graph_param_transform 66.53% : 0.002575s : 71: substitution.inline 0.37% : 0.000014s : 29: substitution.j_node_and_user_rematch 0.99% : 0.000038s : 6: substitution.less_batch_normalization 1.58% : 0.000061s : 2: substitution.list_to_tuple_eliminator_ 0.18% : 0.000007s : 6: substitution.load_eliminater 0.95% : 0.000037s : 21: substitution.minmaximum_grad 0.24% : 0.000009s : 2: substitution.opt_reshape 0.51% : 0.000020s : 29: substitution.remove_not_recompute_node 0.22% : 0.000008s : 4: substitution.replace_old_param 1.62% : 0.000063s : 18: substitution.reshape_eliminate 1.24% : 0.000048s : 24: substitution.switch_simplify 4.80% : 0.000186s : 51: substitution.tuple_list_convert_item_index_to_positive 3.78% : 0.000146s : 51: substitution.tuple_list_get_item_const_eliminator 3.28% : 0.000127s : 51: substitution.tuple_list_get_item_depend_reorder 6.76% : 0.000262s : 73: substitution.tuple_list_get_item_eliminator 3.28% : 0.000127s : 51: substitution.tuple_list_get_set_item_eliminator 0.65% : 0.000025s : 6: substitution.updatestate_pure_node_eliminater 0.60% : 0.000023s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.673478 2 91.51% : 0.616297s : 1: type_inference.infer 8.49% : 0.057181s : 1: type_inference.specialize ------[replace.] 0.001246 115 4.96% : 0.000062s : 5: replace.depend_value_elim 57.63% : 0.000718s : 71: replace.inline 3.32% : 0.000041s : 2: replace.list_to_tuple_eliminator_ 23.17% : 0.000289s : 24: replace.switch_simplify 10.93% : 0.000136s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002667 115 0.09% : 0.000002s : 5: match.depend_value_elim 94.70% : 0.002526s : 71: match.inline 2.19% : 0.000058s : 2: match.list_to_tuple_eliminator_ 1.29% : 0.000035s : 24: match.switch_simplify 1.73% : 0.000046s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001660 9701 2.02% : 0.000034s : 123: predicate.accumulaten_eliminater 0.34% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.50% : 0.000008s : 58: predicate.addn_check_dump 1.16% : 0.000019s : 123: predicate.addn_zero_filter 1.16% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 2.20% : 0.000037s : 181: predicate.arithmetic_simplify 1.13% : 0.000019s : 123: predicate.cast_eliminate 0.30% : 0.000005s : 34: predicate.check_bprop_eliminate 0.47% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.59% : 0.000010s : 61: predicate.depend_value_elim 1.19% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 1.38% : 0.000023s : 123: predicate.dict_get_item_eliminator 1.14% : 0.000019s : 123: predicate.dict_set_item_eliminator 0.38% : 0.000006s : 34: predicate.dumpgradient_eliminate 0.11% : 0.000002s : 17: predicate.elim_not_effective 0.20% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.31% : 0.000022s : 140: predicate.environ_add_const_eliminate 1.19% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.25% : 0.000021s : 140: predicate.environ_get_depend_swap 1.79% : 0.000030s : 198: predicate.environ_get_eliminate 1.31% : 0.000022s : 140: predicate.environ_get_set_eliminate 1.85% : 0.000031s : 209: predicate.exchange_switch_depend_value 3.08% : 0.000051s : 209: predicate.float_depend_g_call 0.49% : 0.000008s : 58: predicate.float_environ_get_switch 0.66% : 0.000011s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.34% : 0.000006s : 35: predicate.get_grad_eliminate 0.10% : 0.000002s : 17: predicate.graph_param_transform 0.52% : 0.000009s : 58: predicate.incorporate_call 0.45% : 0.000008s : 58: predicate.incorporate_call_switch 6.12% : 0.000102s : 459: predicate.inline 0.44% : 0.000007s : 35: predicate.inline_without_move 0.14% : 0.000002s : 35: predicate.j_node_and_user_rematch 0.56% : 0.000009s : 36: predicate.less_batch_normalization 1.65% : 0.000027s : 172: predicate.list_to_tuple_eliminator_ 2.62% : 0.000044s : 295: predicate.load_eliminater 0.35% : 0.000006s : 17: predicate.loop_unroll_after_grad 3.14% : 0.000052s : 314: predicate.loop_unroll_before_grad 1.54% : 0.000026s : 157: predicate.make_slice_get_slice_eliminator 0.49% : 0.000008s : 58: predicate.merge_addn 0.28% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.30% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.13% : 0.000019s : 123: predicate.minmaximum_grad 0.50% : 0.000008s : 17: predicate.mutable_eliminate 0.17% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 3.29% : 0.000055s : 209: predicate.partial_defer_inline 1.48% : 0.000025s : 155: predicate.partial_eliminate 1.13% : 0.000019s : 123: predicate.print_const_string_wrapper 0.49% : 0.000008s : 53: predicate.reduce_all_const_elim 1.69% : 0.000028s : 123: predicate.reduce_eliminate 2.56% : 0.000042s : 295: predicate.redundant_stop_gradient_eliminater 0.19% : 0.000003s : 35: predicate.remove_not_recompute_node 1.05% : 0.000017s : 172: predicate.replace_applicator 0.20% : 0.000003s : 35: predicate.replace_old_param 0.12% : 0.000002s : 17: predicate.reset_defer_inline 1.39% : 0.000023s : 123: predicate.reshape_eliminate 0.29% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.19% : 0.000003s : 17: predicate.row_tensor_eliminate 0.42% : 0.000007s : 34: predicate.same_eliminate 0.25% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.36% : 0.000006s : 35: predicate.shard_identity_eliminate 0.35% : 0.000006s : 34: predicate.special_op_eliminate 0.61% : 0.000010s : 58: predicate.specialize_transform 0.37% : 0.000006s : 34: predicate.split_environ_get_set_with_tuple_value 0.38% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.17% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.46% : 0.000041s : 209: predicate.switch_defer_inline 2.34% : 0.000039s : 243: predicate.switch_layer_defer_inline 6.47% : 0.000107s : 646: predicate.switch_simplify 1.07% : 0.000018s : 123: predicate.tile_eliminate 1.19% : 0.000020s : 123: predicate.transpose_eliminate 1.93% : 0.000032s : 157: predicate.tuple_list_convert_item_index_to_positive 1.73% : 0.000029s : 157: predicate.tuple_list_get_item_const_eliminator 1.64% : 0.000027s : 157: predicate.tuple_list_get_item_depend_reorder 3.22% : 0.000054s : 228: predicate.tuple_list_get_item_eliminator 1.69% : 0.000028s : 157: predicate.tuple_list_get_set_item_eliminator 2.46% : 0.000041s : 215: predicate.tuple_list_set_item_eliminator 1.57% : 0.000026s : 170: predicate.tuple_to_list_eliminator_ 2.48% : 0.000041s : 295: predicate.updatestate_pure_node_eliminater 3.15% : 0.000052s : 353: predicate.updatestate_useless_node_eliminater 0.18% : 0.000003s : 17: predicate.value_based_eliminate 0.33% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.34% : 0.000006s : 35: predicate.virtual_output_eliminate 0.14% : 0.000002s : 17: predicate.virtual_view_grad_eliminate 0.20% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.042959 135 81.13% : 0.034851s : 55: func_graph_cloner_run.FuncGraphClonerGraph 18.87% : 0.008108s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.985725 209 0.00% : 0.000003s : 1: ForceFp32Comm 0.55% : 0.005415s : 1: add_attr 0.55% : 0.005395s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.08% : 0.000752s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.08% : 0.000800s : 1: auto_monad 0.01% : 0.000069s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.08% : 0.000802s : 1: bootstrap 0.01% : 0.000050s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000044s : 1: control_data_broadcast_order 0.00% : 0.000047s : 1: convert_after_rewriter 0.01% : 0.000091s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000029s : 1: environ_conv 0.02% : 0.000237s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.01% : 0.000060s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000011s : 1: label_micro_interleaved_index 0.07% : 0.000672s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000895s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.00% : 0.000041s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000056s : 1: opt.transform.mutable_eliminate 5.06% : 0.049909s : 95: opt.transform.opt_a 0.01% : 0.000131s : 1: opt.transform.opt_after_cconv 0.01% : 0.000079s : 1: opt.transform.opt_after_jit_grad 0.06% : 0.000545s : 28: opt.transform.opt_b 0.03% : 0.000278s : 2: opt.transform.opt_trans_graph 0.02% : 0.000182s : 4: opt.transform.symbol_engine_opt 8.00% : 0.078907s : 1: opt_a 0.03% : 0.000330s : 1: opt_after_cconv 0.08% : 0.000818s : 1: opt_after_jit_grad 0.08% : 0.000815s : 1: opt_b 11.19% : 0.110313s : 1: optimize 0.00% : 0.000048s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000020s : 1: order_py_execute_after_rewriter 0.01% : 0.000054s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000013s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000014s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000011s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.03% : 0.000252s : 1: pre_auto_parallel 0.02% : 0.000233s : 1: py_interpret_to_execute 0.00% : 0.000037s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000121s : 1: remove_dup_value 1.87% : 0.018480s : 1: renormalize.infer 0.76% : 0.007494s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 2.55% : 0.025169s : 1: rewriter_after_opt_a 0.09% : 0.000843s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000022s : 1: swap_dp_allreduce_reducescatter 0.05% : 0.000536s : 1: symbol_engine_optimizer 0.03% : 0.000326s : 1: tuple_transform 68.38% : 0.674074s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape1-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape1-16],max_mem:140.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:44:15.252.548 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 0.666591, [21] [bootstrap]: 0.00082587 [type_inference]: 0.615994 [event_method]: 0.00021198 [auto_monad]: 0.00076645 [graph_reusing]: 5.112e-05 [inline]: 7e-06 [add_attr]: 0.00547102, [1] [add_attr_with_inline]: 0.00545402, [1] [Cycle 1]: 0.00029756, [2] [tag_attr]: 0.00020257 [meta_addattr_fg_expand]: 4.973e-05 [parallel-infer-symbol]: 4.89e-06 [pre_auto_parallel]: 0.00024458 [insert-virtual-dataset]: 2.44001e-06 [parallel-infer-symbol-second]: 1.35001e-06 [dataset_repeat_opt]: 2.59999e-06 [pipeline_split]: 1.77999e-06 [optimize]: 0.0417255, [53] [py_interpret_to_execute]: 0.00023691 [rewriter_before_opt_a]: 0.00078342 [opt_a]: 0.0353492, [2] [Cycle 1]: 0.0327204, [45] [expand_dump_flag]: 1.877e-05 [switch_simplify]: 0.00085587 [loop_unroll]: 0.00032079 [a_1]: 0.0129302 [with_stream_mark]: 6.318e-05 [recompute_prepare]: 4.792e-05 [updatestate_depend_eliminate]: 2.198e-05 [updatestate_assign_eliminate]: 1.612e-05 [updatestate_loads_eliminate]: 1.884e-05 [parameter_eliminate]: 4.08999e-06 [a_2]: 0.00085073 [accelerated_algorithm]: 7.92e-05 [shard]: 2.76e-06 [meta_shard_fg_expand]: 2.541e-05 [shard_inline]: 2.673e-05 [merge_send_recv]: 2.238e-05 [auto_parallel]: 2.269e-05 [parallel]: 4.812e-05 [flash_sp]: 1.667e-05 [merge_comm]: 1.641e-05 [allreduce_fusion]: 1.249e-05 [matmul_add_comm_reduction]: 2.211e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 3.649e-05 [virtual_dataset]: 2.706e-05 [get_grad_eliminate_]: 2.377e-05 [virtual_output]: 2.382e-05 [merge_forward]: 1.608e-05 [cell_reuse_recompute_pass]: 2.64001e-06 [offload_activation]: 2.694e-05 [cell_reuse_handle_not_recompute_node_pass]: 6.246e-05 [merge_recompute_call_nodes]: 1.99999e-06 [before_grad]: 4.28e-05 [set_forward_comm_id_for_comm_node_pass]: 1.81e-05 [meta_fg_expand]: 1.576e-05 [flash_sp_send_recv_attached]: 8.17998e-06 [receive_attached]: 3.01001e-06 [after_resolve]: 3.684e-05 [a_after_grad]: 3.905e-05 [renormalize]: 0.014168 [add_forward_monad_depend]: 1.613e-05 [auto_monad_grad]: 3.4e-06 [auto_monad_eliminator]: 8.571e-05 [cse]: 0.00193923 [a_3]: 0.00017545 [Cycle 2]: 0.00261085, [45] [expand_dump_flag]: 3.71999e-06 [switch_simplify]: 2.59e-05 [loop_unroll]: 2.204e-05 [a_1]: 0.00064458 [with_stream_mark]: 3.201e-05 [recompute_prepare]: 2.093e-05 [updatestate_depend_eliminate]: 0.0004075 [updatestate_assign_eliminate]: 1.952e-05 [updatestate_loads_eliminate]: 1.487e-05 [parameter_eliminate]: 3.76001e-06 [a_2]: 0.00034674 [accelerated_algorithm]: 3.287e-05 [shard]: 3.14001e-06 [meta_shard_fg_expand]: 8.80999e-06 [shard_inline]: 2.18e-05 [merge_send_recv]: 2.071e-05 [auto_parallel]: 2.292e-05 [parallel]: 1.048e-05 [flash_sp]: 4.92e-06 [merge_comm]: 1.295e-05 [allreduce_fusion]: 1.197e-05 [matmul_add_comm_reduction]: 2.411e-05 [allreduce_slice_to_reducescatter]: 1.12e-06 [virtual_shard_identity]: 2.403e-05 [virtual_dataset]: 2.111e-05 [get_grad_eliminate_]: 2.259e-05 [virtual_output]: 2.037e-05 [merge_forward]: 1.316e-05 [cell_reuse_recompute_pass]: 3.14999e-06 [offload_activation]: 3.196e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.526e-05 [merge_recompute_call_nodes]: 2.48e-06 [before_grad]: 3.848e-05 [set_forward_comm_id_for_comm_node_pass]: 1.394e-05 [meta_fg_expand]: 9.58002e-06 [flash_sp_send_recv_attached]: 2.63e-06 [receive_attached]: 3.4e-06 [after_resolve]: 3.064e-05 [a_after_grad]: 3.27e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 4.22e-06 [auto_monad_grad]: 1.92001e-06 [auto_monad_eliminator]: 6.394e-05 [cse]: 8.907e-05 [a_3]: 0.0001414 [py_interpret_to_execute_after_opt_a]: 3.594e-05 [slice_cell_reuse_recomputed_activation]: 2.56e-06 [rewriter_after_opt_a]: 0.00093571 [convert_after_rewriter]: 2.808e-05 [order_py_execute_after_rewriter]: 1.616e-05 [mutable_eliminate]: 0.00086303 [opt_b]: 0.00081095, [1] [Cycle 1]: 0.00080075, [7] [b_1]: 0.0005756 [b_2]: 2.913e-05 [updatestate_depend_eliminate]: 1.818e-05 [updatestate_assign_eliminate]: 1.173e-05 [updatestate_loads_eliminate]: 1.357e-05 [renormalize]: 6.60017e-07 [cse]: 9.617e-05 [optimize_parallel_all_gather_comm]: 4.337e-05 [overlap_param_gather]: 3.38999e-06 [cconv]: 4.223e-05 [loop_unroll]: 0.00062 [opt_after_cconv]: 0.00030511, [1] [Cycle 1]: 0.00029663, [7] [c_1]: 0.0001307 [parameter_eliminate]: 5.74e-06 [updatestate_depend_eliminate]: 1.866e-05 [updatestate_assign_eliminate]: 1.095e-05 [updatestate_loads_eliminate]: 1.325e-05 [cse]: 7.849e-05 [renormalize]: 6.39993e-07 [remove_dup_value]: 9.919e-05 [tuple_transform]: 0.00026172, [1] [Cycle 1]: 0.00025556, [4] [d_1]: 0.00020688 [none_parameter_eliminate]: 2.37999e-06 [renormalize]: 2.00002e-07 [switch_simplify]: 2.358e-05 [partial_unused_args_eliminate]: 2.36998e-06 [add_recomputation]: 0.00016061 [cse_after_recomputation]: 7.229e-05, [1] [Cycle 1]: 6.584e-05, [1] [cse]: 5.697e-05 [environ_conv]: 2.299e-05 [swap_dp_allreduce_reducescatter]: 1.853e-05 [bias_add_comm_swap]: 3.52002e-06 [label_micro_interleaved_index]: 5.87999e-06 [label_fine_grained_interleaved_index]: 3.08e-06 [merge_cast_opt]: 1.47001e-06 [slice_recompute_activation]: 2.16e-06 [micro_interleaved_order_control]: 3.14999e-06 [assign_add_opt]: 1.35999e-06 [ForceFp32Comm]: 1.05999e-06 [remove_cast_before_assign_add]: 1.18001e-06 [full_micro_interleaved_order_control]: 2.35002e-06 [reorder_send_recv_between_fp_bp]: 3.30998e-06 [comm_op_add_attrs]: 1.05001e-06 [add_comm_op_reuse_tag]: 1.10001e-06 [interleave_split_concat_branches]: 1.55999e-06 [interleave_parallel_branches]: 1.34e-06 [overlap_opt_shard_in_pipeline]: 3.41999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.81e-06 [control_data_broadcast_order]: 3.942e-05 [grouped_pairwise_exchange_alltoall]: 1.95001e-06 [offloading_packed_experts]: 1.134e-05 [overlap_recompute_and_grad_model_parallel]: 1.048e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.24998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37999e-06 [overlap_recompute_comm]: 2.47001e-06 [overlap_grad_ring_attention]: 9.15999e-06 [overlap_grad_flash_sp]: 5.174e-05 [begin_end_overlap_inline]: 5.50004e-07 [split_matmul_comm_elemetwise]: 2.56e-06 [split_layernorm_comm]: 2.17999e-06 [handle_group_info]: 9.89996e-07 [symbol_engine_optimizer]: 0.00052402, [1] [Cycle 1]: 0.0005182, [6] [build]: 0.00029674 [elim_shapecalc]: 3.294e-05 [elim_not_effective]: 5.099e-05 [opt_reshape]: 3.422e-05 [fold_const_symbol]: 5.815e-05 [renormalize]: 6.59988e-07 [detach_backward]: 3.26001e-06 [pipeline_parallel_scheduler]: 1.63002e-06 [auto_monad_reorder]: 6.212e-05 [get_jit_bprop_graph]: 2.06e-06 [rewriter_after_jit_bprop_graph]: 8.22e-06 [opt_after_jit_grad]: 0.00081979 [validate]: 0.00011277 Sums bootstrap : 0.000826s : 0.13% type_inference : 0.615994s : 93.36% event_method : 0.000212s : 0.03% auto_monad : 0.000766s : 0.12% graph_reusing : 0.000051s : 0.01% inline : 0.000007s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000203s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000050s : 0.01% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000245s : 0.04% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000237s : 0.04% optimize.rewriter_before_opt_a : 0.000783s : 0.12% optimize.opt_a.expand_dump_flag : 0.000022s : 0.00% optimize.opt_a.switch_simplify : 0.000882s : 0.13% optimize.opt_a.loop_unroll : 0.000343s : 0.05% optimize.opt_a.a_1 : 0.013575s : 2.06% optimize.opt_a.with_stream_mark : 0.000095s : 0.01% optimize.opt_a.recompute_prepare : 0.000069s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000429s : 0.07% optimize.opt_a.updatestate_assign_eliminate : 0.000036s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000034s : 0.01% optimize.opt_a.parameter_eliminate : 0.000008s : 0.00% optimize.opt_a.a_2 : 0.001197s : 0.18% optimize.opt_a.accelerated_algorithm : 0.000112s : 0.02% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000034s : 0.01% optimize.opt_a.shard_inline : 0.000049s : 0.01% optimize.opt_a.merge_send_recv : 0.000043s : 0.01% optimize.opt_a.auto_parallel : 0.000046s : 0.01% optimize.opt_a.parallel : 0.000059s : 0.01% optimize.opt_a.flash_sp : 0.000022s : 0.00% optimize.opt_a.merge_comm : 0.000029s : 0.00% optimize.opt_a.allreduce_fusion : 0.000024s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000046s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000061s : 0.01% optimize.opt_a.virtual_dataset : 0.000048s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000046s : 0.01% optimize.opt_a.virtual_output : 0.000044s : 0.01% optimize.opt_a.merge_forward : 0.000029s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000059s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000108s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000081s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000032s : 0.00% optimize.opt_a.meta_fg_expand : 0.000025s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000011s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000067s : 0.01% optimize.opt_a.a_after_grad : 0.000072s : 0.01% optimize.opt_a.renormalize : 0.014168s : 2.15% optimize.opt_a.add_forward_monad_depend : 0.000020s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000150s : 0.02% optimize.opt_a.cse : 0.002028s : 0.31% optimize.opt_a.a_3 : 0.000317s : 0.05% optimize.py_interpret_to_execute_after_opt_a : 0.000036s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000936s : 0.14% optimize.convert_after_rewriter : 0.000028s : 0.00% optimize.order_py_execute_after_rewriter : 0.000016s : 0.00% optimize.mutable_eliminate : 0.000863s : 0.13% optimize.opt_b.b_1 : 0.000576s : 0.09% optimize.opt_b.b_2 : 0.000029s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000018s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000014s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000096s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000043s : 0.01% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000042s : 0.01% optimize.loop_unroll : 0.000620s : 0.09% optimize.opt_after_cconv.c_1 : 0.000131s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000019s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.cse : 0.000078s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000099s : 0.02% optimize.tuple_transform.d_1 : 0.000207s : 0.03% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000024s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000161s : 0.02% optimize.cse_after_recomputation.cse : 0.000057s : 0.01% optimize.environ_conv : 0.000023s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000019s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000006s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000002s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000003s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000039s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000011s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000010s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000052s : 0.01% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000297s : 0.04% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000033s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000051s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000034s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000058s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000001s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000062s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000820s : 0.12% validate : 0.000113s : 0.02% Time group info: ------[substitution.] 0.004501 569 0.41% : 0.000018s : 8: substitution.depend_value_elim 0.31% : 0.000014s : 14: substitution.elim_not_effective 0.46% : 0.000021s : 12: substitution.float_tuple_getitem_switch 0.58% : 0.000026s : 14: substitution.fold_const_symbol 0.36% : 0.000016s : 17: substitution.graph_param_transform 71.53% : 0.003219s : 71: substitution.inline 0.33% : 0.000015s : 29: substitution.j_node_and_user_rematch 0.81% : 0.000036s : 6: substitution.less_batch_normalization 1.22% : 0.000055s : 2: substitution.list_to_tuple_eliminator_ 0.14% : 0.000007s : 6: substitution.load_eliminater 1.01% : 0.000045s : 21: substitution.minmaximum_grad 0.19% : 0.000009s : 2: substitution.opt_reshape 0.48% : 0.000022s : 29: substitution.remove_not_recompute_node 0.22% : 0.000010s : 4: substitution.replace_old_param 1.48% : 0.000067s : 18: substitution.reshape_eliminate 1.17% : 0.000053s : 24: substitution.switch_simplify 4.21% : 0.000189s : 51: substitution.tuple_list_convert_item_index_to_positive 2.16% : 0.000097s : 51: substitution.tuple_list_get_item_const_eliminator 2.87% : 0.000129s : 51: substitution.tuple_list_get_item_depend_reorder 6.26% : 0.000282s : 73: substitution.tuple_list_get_item_eliminator 2.69% : 0.000121s : 51: substitution.tuple_list_get_set_item_eliminator 0.54% : 0.000024s : 6: substitution.updatestate_pure_node_eliminater 0.56% : 0.000025s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.615632 2 96.32% : 0.593001s : 1: type_inference.infer 3.68% : 0.022631s : 1: type_inference.specialize ------[replace.] 0.001340 115 4.76% : 0.000064s : 5: replace.depend_value_elim 56.62% : 0.000759s : 71: replace.inline 3.40% : 0.000046s : 2: replace.list_to_tuple_eliminator_ 23.39% : 0.000313s : 24: replace.switch_simplify 11.83% : 0.000159s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.003324 115 0.07% : 0.000002s : 5: match.depend_value_elim 95.23% : 0.003165s : 71: match.inline 1.57% : 0.000052s : 2: match.list_to_tuple_eliminator_ 1.15% : 0.000038s : 24: match.switch_simplify 1.98% : 0.000066s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.003530 9701 0.60% : 0.000021s : 123: predicate.accumulaten_eliminater 0.16% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.23% : 0.000008s : 58: predicate.addn_check_dump 0.59% : 0.000021s : 123: predicate.addn_zero_filter 0.53% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 1.13% : 0.000040s : 181: predicate.arithmetic_simplify 0.55% : 0.000020s : 123: predicate.cast_eliminate 0.14% : 0.000005s : 34: predicate.check_bprop_eliminate 0.23% : 0.000008s : 58: predicate.compare_switch_simplify 0.04% : 0.000001s : 17: predicate.const_output_eliminate 0.28% : 0.000010s : 61: predicate.depend_value_elim 0.57% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 0.63% : 0.000022s : 123: predicate.dict_get_item_eliminator 0.55% : 0.000019s : 123: predicate.dict_set_item_eliminator 0.26% : 0.000009s : 34: predicate.dumpgradient_eliminate 0.04% : 0.000001s : 17: predicate.elim_not_effective 0.10% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 0.63% : 0.000022s : 140: predicate.environ_add_const_eliminate 0.58% : 0.000021s : 140: predicate.environ_get_add_eliminate 0.59% : 0.000021s : 140: predicate.environ_get_depend_swap 0.87% : 0.000031s : 198: predicate.environ_get_eliminate 0.58% : 0.000021s : 140: predicate.environ_get_set_eliminate 0.95% : 0.000034s : 209: predicate.exchange_switch_depend_value 1.42% : 0.000050s : 209: predicate.float_depend_g_call 0.24% : 0.000008s : 58: predicate.float_environ_get_switch 0.34% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.04% : 0.000001s : 17: predicate.fold_const_symbol 0.18% : 0.000006s : 35: predicate.get_grad_eliminate 0.05% : 0.000002s : 17: predicate.graph_param_transform 0.25% : 0.000009s : 58: predicate.incorporate_call 0.21% : 0.000008s : 58: predicate.incorporate_call_switch 2.98% : 0.000105s : 459: predicate.inline 0.22% : 0.000008s : 35: predicate.inline_without_move 0.08% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.24% : 0.000009s : 36: predicate.less_batch_normalization 0.84% : 0.000030s : 172: predicate.list_to_tuple_eliminator_ 1.27% : 0.000045s : 295: predicate.load_eliminater 0.16% : 0.000006s : 17: predicate.loop_unroll_after_grad 1.50% : 0.000053s : 314: predicate.loop_unroll_before_grad 0.79% : 0.000028s : 157: predicate.make_slice_get_slice_eliminator 0.25% : 0.000009s : 58: predicate.merge_addn 0.13% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.13% : 0.000005s : 34: predicate.mini_step_allgather_replace 0.58% : 0.000021s : 123: predicate.minmaximum_grad 0.17% : 0.000006s : 17: predicate.mutable_eliminate 0.08% : 0.000003s : 17: predicate.opt_reshape 0.09% : 0.000003s : 17: predicate.parallel_virtual_node 1.89% : 0.000067s : 209: predicate.partial_defer_inline 0.76% : 0.000027s : 155: predicate.partial_eliminate 0.59% : 0.000021s : 123: predicate.print_const_string_wrapper 0.22% : 0.000008s : 53: predicate.reduce_all_const_elim 0.83% : 0.000029s : 123: predicate.reduce_eliminate 1.26% : 0.000045s : 295: predicate.redundant_stop_gradient_eliminater 0.09% : 0.000003s : 35: predicate.remove_not_recompute_node 0.55% : 0.000019s : 172: predicate.replace_applicator 0.08% : 0.000003s : 35: predicate.replace_old_param 0.04% : 0.000001s : 17: predicate.reset_defer_inline 0.65% : 0.000023s : 123: predicate.reshape_eliminate 0.14% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.10% : 0.000003s : 17: predicate.row_tensor_eliminate 0.22% : 0.000008s : 34: predicate.same_eliminate 0.11% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.19% : 0.000007s : 35: predicate.shard_identity_eliminate 0.16% : 0.000006s : 34: predicate.special_op_eliminate 0.29% : 0.000010s : 58: predicate.specialize_transform 0.19% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.18% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.08% : 0.000003s : 17: predicate.switch_call_monad_eliminater 1.13% : 0.000040s : 209: predicate.switch_defer_inline 1.17% : 0.000041s : 243: predicate.switch_layer_defer_inline 3.19% : 0.000113s : 646: predicate.switch_simplify 0.57% : 0.000020s : 123: predicate.tile_eliminate 0.58% : 0.000021s : 123: predicate.transpose_eliminate 0.87% : 0.000031s : 157: predicate.tuple_list_convert_item_index_to_positive 0.88% : 0.000031s : 157: predicate.tuple_list_get_item_const_eliminator 0.81% : 0.000029s : 157: predicate.tuple_list_get_item_depend_reorder 1.59% : 0.000056s : 228: predicate.tuple_list_get_item_eliminator 0.92% : 0.000032s : 157: predicate.tuple_list_get_set_item_eliminator 1.20% : 0.000042s : 215: predicate.tuple_list_set_item_eliminator 0.79% : 0.000028s : 170: predicate.tuple_to_list_eliminator_ 1.29% : 0.000045s : 295: predicate.updatestate_pure_node_eliminater 52.69% : 0.001860s : 353: predicate.updatestate_useless_node_eliminater 0.07% : 0.000003s : 17: predicate.value_based_eliminate 0.19% : 0.000007s : 35: predicate.virtual_dataset_eliminate 0.17% : 0.000006s : 35: predicate.virtual_output_eliminate 0.07% : 0.000002s : 17: predicate.virtual_view_grad_eliminate 0.10% : 0.000004s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.019438 135 61.11% : 0.011878s : 55: func_graph_cloner_run.FuncGraphClonerGraph 38.89% : 0.007559s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.745830 209 0.00% : 0.000004s : 1: ForceFp32Comm 0.73% : 0.005478s : 1: add_attr 0.73% : 0.005460s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.02% : 0.000167s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.11% : 0.000785s : 1: auto_monad 0.01% : 0.000067s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.12% : 0.000891s : 1: bootstrap 0.01% : 0.000047s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000043s : 1: control_data_broadcast_order 0.00% : 0.000035s : 1: convert_after_rewriter 0.01% : 0.000076s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000026s : 1: environ_conv 0.03% : 0.000226s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.01% : 0.000060s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 1: handle_group_info 0.00% : 0.000011s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.08% : 0.000632s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.12% : 0.000877s : 1: mutable_eliminate 0.00% : 0.000014s : 1: offloading_packed_experts 0.01% : 0.000038s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000043s : 1: opt.transform.mutable_eliminate 2.27% : 0.016901s : 95: opt.transform.opt_a 0.02% : 0.000129s : 1: opt.transform.opt_after_cconv 0.01% : 0.000075s : 1: opt.transform.opt_after_jit_grad 0.08% : 0.000563s : 28: opt.transform.opt_b 0.03% : 0.000228s : 2: opt.transform.opt_trans_graph 0.02% : 0.000171s : 4: opt.transform.symbol_engine_opt 4.74% : 0.035354s : 1: opt_a 0.04% : 0.000309s : 1: opt_after_cconv 0.11% : 0.000833s : 1: opt_after_jit_grad 0.11% : 0.000816s : 1: opt_b 5.60% : 0.041732s : 1: optimize 0.01% : 0.000048s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000020s : 1: order_py_execute_after_rewriter 0.01% : 0.000056s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000006s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000014s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000010s : 1: parallel-infer-symbol 0.00% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000007s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000253s : 1: pre_auto_parallel 0.03% : 0.000248s : 1: py_interpret_to_execute 0.01% : 0.000041s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000105s : 1: remove_dup_value 0.98% : 0.007313s : 1: renormalize.infer 0.92% : 0.006833s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.13% : 0.000951s : 1: rewriter_after_opt_a 0.11% : 0.000799s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000022s : 1: swap_dp_allreduce_reducescatter 0.07% : 0.000528s : 1: symbol_engine_optimizer 0.04% : 0.000267s : 1: tuple_transform 82.60% : 0.616024s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape2-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape2-1],max_mem:196.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:44:30.679.021 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 0.58966, [21] [bootstrap]: 0.00082988 [type_inference]: 0.459204 [event_method]: 0.00020768 [auto_monad]: 0.00072411 [graph_reusing]: 4.479e-05 [inline]: 3.31999e-06 [add_attr]: 0.00888174, [1] [add_attr_with_inline]: 0.00885929, [1] [Cycle 1]: 0.00035944, [2] [tag_attr]: 0.00023764 [meta_addattr_fg_expand]: 5.717e-05 [parallel-infer-symbol]: 4.16001e-06 [pre_auto_parallel]: 0.0002722 [insert-virtual-dataset]: 2.67001e-06 [parallel-infer-symbol-second]: 1.74e-06 [dataset_repeat_opt]: 3.23998e-06 [pipeline_split]: 1.93002e-06 [optimize]: 0.118206, [53] [py_interpret_to_execute]: 0.00027109 [rewriter_before_opt_a]: 0.00083044 [opt_a]: 0.11092, [2] [Cycle 1]: 0.108395, [45] [expand_dump_flag]: 2.693e-05 [switch_simplify]: 0.00091274 [loop_unroll]: 0.00035468 [a_1]: 0.0427554 [with_stream_mark]: 5.362e-05 [recompute_prepare]: 4.231e-05 [updatestate_depend_eliminate]: 1.977e-05 [updatestate_assign_eliminate]: 1.598e-05 [updatestate_loads_eliminate]: 1.849e-05 [parameter_eliminate]: 3.08998e-06 [a_2]: 0.00082211 [accelerated_algorithm]: 7.023e-05 [shard]: 2.34999e-06 [meta_shard_fg_expand]: 2.307e-05 [shard_inline]: 2.552e-05 [merge_send_recv]: 2.122e-05 [auto_parallel]: 2.077e-05 [parallel]: 3.634e-05 [flash_sp]: 1.807e-05 [merge_comm]: 1.541e-05 [allreduce_fusion]: 1.314e-05 [matmul_add_comm_reduction]: 2.433e-05 [allreduce_slice_to_reducescatter]: 7.2e-07 [virtual_shard_identity]: 3.16e-05 [virtual_dataset]: 2.602e-05 [get_grad_eliminate_]: 2.13e-05 [virtual_output]: 2.133e-05 [merge_forward]: 1.362e-05 [cell_reuse_recompute_pass]: 1.92001e-06 [offload_activation]: 2.51e-05 [cell_reuse_handle_not_recompute_node_pass]: 5.131e-05 [merge_recompute_call_nodes]: 1.79e-06 [before_grad]: 0.0243473 [set_forward_comm_id_for_comm_node_pass]: 3.854e-05 [meta_fg_expand]: 4.091e-05 [flash_sp_send_recv_attached]: 1.323e-05 [receive_attached]: 2.53998e-06 [after_resolve]: 4.707e-05 [a_after_grad]: 3.957e-05 [renormalize]: 0.0128974 [add_forward_monad_depend]: 1.394e-05 [auto_monad_grad]: 4.60001e-06 [auto_monad_eliminator]: 9.5e-05 [cse]: 0.00213717 [a_3]: 0.0226339 [Cycle 2]: 0.00249943, [45] [expand_dump_flag]: 5.54e-06 [switch_simplify]: 3.815e-05 [loop_unroll]: 2.627e-05 [a_1]: 0.00069445 [with_stream_mark]: 4.374e-05 [recompute_prepare]: 2.974e-05 [updatestate_depend_eliminate]: 1.965e-05 [updatestate_assign_eliminate]: 1.241e-05 [updatestate_loads_eliminate]: 1.481e-05 [parameter_eliminate]: 3.35e-06 [a_2]: 0.00032473 [accelerated_algorithm]: 3.437e-05 [shard]: 3.28998e-06 [meta_shard_fg_expand]: 1.052e-05 [shard_inline]: 5.039e-05 [merge_send_recv]: 2.215e-05 [auto_parallel]: 2.305e-05 [parallel]: 1.176e-05 [flash_sp]: 5.30999e-06 [merge_comm]: 1.465e-05 [allreduce_fusion]: 1.34e-05 [matmul_add_comm_reduction]: 2.79e-05 [allreduce_slice_to_reducescatter]: 7.30011e-07 [virtual_shard_identity]: 2.838e-05 [virtual_dataset]: 2.3e-05 [get_grad_eliminate_]: 2.377e-05 [virtual_output]: 2.309e-05 [merge_forward]: 1.414e-05 [cell_reuse_recompute_pass]: 3.35003e-06 [offload_activation]: 3.14e-05 [cell_reuse_handle_not_recompute_node_pass]: 5.237e-05 [merge_recompute_call_nodes]: 1.67001e-06 [before_grad]: 4.517e-05 [set_forward_comm_id_for_comm_node_pass]: 1.704e-05 [meta_fg_expand]: 1.248e-05 [flash_sp_send_recv_attached]: 1.96e-06 [receive_attached]: 3.45e-06 [after_resolve]: 3.473e-05 [a_after_grad]: 3.636e-05 [renormalize]: 6.00121e-08 [add_forward_monad_depend]: 6.71e-06 [auto_monad_grad]: 4.70001e-06 [auto_monad_eliminator]: 8.314e-05 [cse]: 0.00011726 [a_3]: 0.0001638 [py_interpret_to_execute_after_opt_a]: 4.036e-05 [slice_cell_reuse_recomputed_activation]: 2.73998e-06 [rewriter_after_opt_a]: 0.00096035 [convert_after_rewriter]: 3.35e-05 [order_py_execute_after_rewriter]: 1.686e-05 [mutable_eliminate]: 0.00092712 [opt_b]: 0.00085001, [1] [Cycle 1]: 0.00083945, [7] [b_1]: 0.00056781 [b_2]: 2.756e-05 [updatestate_depend_eliminate]: 2.49e-05 [updatestate_assign_eliminate]: 1.176e-05 [updatestate_loads_eliminate]: 1.594e-05 [renormalize]: 1.27999e-06 [cse]: 0.00012901 [optimize_parallel_all_gather_comm]: 4.829e-05 [overlap_param_gather]: 2.19999e-06 [cconv]: 4.865e-05 [loop_unroll]: 0.00111534 [opt_after_cconv]: 0.00033491, [1] [Cycle 1]: 0.00032291, [7] [c_1]: 0.00013486 [parameter_eliminate]: 6.66e-06 [updatestate_depend_eliminate]: 2.305e-05 [updatestate_assign_eliminate]: 1.173e-05 [updatestate_loads_eliminate]: 1.472e-05 [cse]: 8.734e-05 [renormalize]: 8.79983e-07 [remove_dup_value]: 0.00010482 [tuple_transform]: 0.00027672, [1] [Cycle 1]: 0.00026936, [4] [d_1]: 0.00021251 [none_parameter_eliminate]: 2.86e-06 [renormalize]: 3.7998e-07 [switch_simplify]: 2.327e-05 [partial_unused_args_eliminate]: 2.15002e-06 [add_recomputation]: 0.00016424 [cse_after_recomputation]: 8.161e-05, [1] [Cycle 1]: 7.379e-05, [1] [cse]: 6.429e-05 [environ_conv]: 2.609e-05 [swap_dp_allreduce_reducescatter]: 1.969e-05 [bias_add_comm_swap]: 3.53999e-06 [label_micro_interleaved_index]: 7.16001e-06 [label_fine_grained_interleaved_index]: 3.06001e-06 [merge_cast_opt]: 1.60999e-06 [slice_recompute_activation]: 2.13002e-06 [micro_interleaved_order_control]: 2.28002e-06 [assign_add_opt]: 1.26002e-06 [ForceFp32Comm]: 8.79983e-07 [remove_cast_before_assign_add]: 1.12e-06 [full_micro_interleaved_order_control]: 2.12999e-06 [reorder_send_recv_between_fp_bp]: 3.41001e-06 [comm_op_add_attrs]: 1.05999e-06 [add_comm_op_reuse_tag]: 1.30001e-06 [interleave_split_concat_branches]: 1.52999e-06 [interleave_parallel_branches]: 1.07e-06 [overlap_opt_shard_in_pipeline]: 1.29e-06 [overlap_opt_shard_grad_in_pipeline]: 1.73002e-06 [control_data_broadcast_order]: 3.965e-05 [grouped_pairwise_exchange_alltoall]: 1.67001e-06 [offloading_packed_experts]: 1.243e-05 [overlap_recompute_and_grad_model_parallel]: 1.039e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.55001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.45001e-06 [overlap_recompute_comm]: 2.31998e-06 [overlap_grad_ring_attention]: 1.014e-05 [overlap_grad_flash_sp]: 5.342e-05 [begin_end_overlap_inline]: 5.50004e-07 [split_matmul_comm_elemetwise]: 2.32001e-06 [split_layernorm_comm]: 1.83002e-06 [handle_group_info]: 1.20001e-06 [symbol_engine_optimizer]: 0.00057487, [1] [Cycle 1]: 0.00056661, [6] [build]: 0.00030376 [elim_shapecalc]: 4.299e-05 [elim_not_effective]: 5.556e-05 [opt_reshape]: 3.483e-05 [fold_const_symbol]: 6.003e-05 [renormalize]: 7.89994e-07 [detach_backward]: 3.25e-06 [pipeline_parallel_scheduler]: 1.54e-06 [auto_monad_reorder]: 6.842e-05 [get_jit_bprop_graph]: 2.37999e-06 [rewriter_after_jit_bprop_graph]: 7.45998e-06 [opt_after_jit_grad]: 0.00080234 [validate]: 0.00012116 Sums bootstrap : 0.000830s : 0.14% type_inference : 0.459204s : 79.28% event_method : 0.000208s : 0.04% auto_monad : 0.000724s : 0.13% graph_reusing : 0.000045s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000238s : 0.04% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000057s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000272s : 0.05% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000271s : 0.05% optimize.rewriter_before_opt_a : 0.000830s : 0.14% optimize.opt_a.expand_dump_flag : 0.000032s : 0.01% optimize.opt_a.switch_simplify : 0.000951s : 0.16% optimize.opt_a.loop_unroll : 0.000381s : 0.07% optimize.opt_a.a_1 : 0.043450s : 7.50% optimize.opt_a.with_stream_mark : 0.000097s : 0.02% optimize.opt_a.recompute_prepare : 0.000072s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000039s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000028s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000033s : 0.01% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001147s : 0.20% optimize.opt_a.accelerated_algorithm : 0.000105s : 0.02% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000034s : 0.01% optimize.opt_a.shard_inline : 0.000076s : 0.01% optimize.opt_a.merge_send_recv : 0.000043s : 0.01% optimize.opt_a.auto_parallel : 0.000044s : 0.01% optimize.opt_a.parallel : 0.000048s : 0.01% optimize.opt_a.flash_sp : 0.000023s : 0.00% optimize.opt_a.merge_comm : 0.000030s : 0.01% optimize.opt_a.allreduce_fusion : 0.000027s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000052s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000060s : 0.01% optimize.opt_a.virtual_dataset : 0.000049s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000045s : 0.01% optimize.opt_a.virtual_output : 0.000044s : 0.01% optimize.opt_a.merge_forward : 0.000028s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000056s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000104s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.024392s : 4.21% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000056s : 0.01% optimize.opt_a.meta_fg_expand : 0.000053s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000015s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000082s : 0.01% optimize.opt_a.a_after_grad : 0.000076s : 0.01% optimize.opt_a.renormalize : 0.012897s : 2.23% optimize.opt_a.add_forward_monad_depend : 0.000021s : 0.00% optimize.opt_a.auto_monad_grad : 0.000009s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000178s : 0.03% optimize.opt_a.cse : 0.002254s : 0.39% optimize.opt_a.a_3 : 0.022798s : 3.94% optimize.py_interpret_to_execute_after_opt_a : 0.000040s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000960s : 0.17% optimize.convert_after_rewriter : 0.000033s : 0.01% optimize.order_py_execute_after_rewriter : 0.000017s : 0.00% optimize.mutable_eliminate : 0.000927s : 0.16% optimize.opt_b.b_1 : 0.000568s : 0.10% optimize.opt_b.b_2 : 0.000028s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000025s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000016s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000129s : 0.02% optimize.optimize_parallel_all_gather_comm : 0.000048s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000049s : 0.01% optimize.loop_unroll : 0.001115s : 0.19% optimize.opt_after_cconv.c_1 : 0.000135s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000023s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_after_cconv.cse : 0.000087s : 0.02% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000105s : 0.02% optimize.tuple_transform.d_1 : 0.000213s : 0.04% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000023s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000164s : 0.03% optimize.cse_after_recomputation.cse : 0.000064s : 0.01% optimize.environ_conv : 0.000026s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000020s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000007s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000002s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000040s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000012s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000010s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000002s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000010s : 0.00% optimize.overlap_grad_flash_sp : 0.000053s : 0.01% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000304s : 0.05% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000043s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000056s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000035s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000060s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000001s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000068s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000802s : 0.14% validate : 0.000121s : 0.02% Time group info: ------[substitution.] 0.004069 569 0.45% : 0.000018s : 8: substitution.depend_value_elim 0.38% : 0.000015s : 14: substitution.elim_not_effective 0.46% : 0.000019s : 12: substitution.float_tuple_getitem_switch 0.70% : 0.000028s : 14: substitution.fold_const_symbol 0.40% : 0.000016s : 17: substitution.graph_param_transform 70.10% : 0.002852s : 71: substitution.inline 0.59% : 0.000024s : 29: substitution.j_node_and_user_rematch 0.93% : 0.000038s : 6: substitution.less_batch_normalization 1.58% : 0.000064s : 2: substitution.list_to_tuple_eliminator_ 0.20% : 0.000008s : 6: substitution.load_eliminater 0.79% : 0.000032s : 21: substitution.minmaximum_grad 0.21% : 0.000008s : 2: substitution.opt_reshape 0.53% : 0.000021s : 29: substitution.remove_not_recompute_node 0.30% : 0.000012s : 4: substitution.replace_old_param 1.98% : 0.000081s : 18: substitution.reshape_eliminate 1.43% : 0.000058s : 24: substitution.switch_simplify 3.99% : 0.000162s : 51: substitution.tuple_list_convert_item_index_to_positive 2.24% : 0.000091s : 51: substitution.tuple_list_get_item_const_eliminator 2.80% : 0.000114s : 51: substitution.tuple_list_get_item_depend_reorder 5.56% : 0.000226s : 73: substitution.tuple_list_get_item_eliminator 2.81% : 0.000114s : 51: substitution.tuple_list_get_set_item_eliminator 0.94% : 0.000038s : 6: substitution.updatestate_pure_node_eliminater 0.66% : 0.000027s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.458855 2 92.42% : 0.424083s : 1: type_inference.infer 7.58% : 0.034772s : 1: type_inference.specialize ------[replace.] 0.001318 115 3.77% : 0.000050s : 5: replace.depend_value_elim 56.85% : 0.000749s : 71: replace.inline 3.69% : 0.000049s : 2: replace.list_to_tuple_eliminator_ 27.50% : 0.000362s : 24: replace.switch_simplify 8.19% : 0.000108s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002941 115 0.08% : 0.000002s : 5: match.depend_value_elim 95.24% : 0.002801s : 71: match.inline 2.07% : 0.000061s : 2: match.list_to_tuple_eliminator_ 1.45% : 0.000043s : 24: match.switch_simplify 1.17% : 0.000035s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001733 9701 1.19% : 0.000021s : 123: predicate.accumulaten_eliminater 0.37% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.51% : 0.000009s : 58: predicate.addn_check_dump 1.31% : 0.000023s : 123: predicate.addn_zero_filter 1.07% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 2.45% : 0.000042s : 181: predicate.arithmetic_simplify 1.25% : 0.000022s : 123: predicate.cast_eliminate 0.32% : 0.000006s : 34: predicate.check_bprop_eliminate 0.47% : 0.000008s : 58: predicate.compare_switch_simplify 0.07% : 0.000001s : 17: predicate.const_output_eliminate 0.61% : 0.000011s : 61: predicate.depend_value_elim 1.12% : 0.000019s : 123: predicate.dict_get_item_const_eliminator 1.29% : 0.000022s : 123: predicate.dict_get_item_eliminator 1.17% : 0.000020s : 123: predicate.dict_set_item_eliminator 0.40% : 0.000007s : 34: predicate.dumpgradient_eliminate 0.10% : 0.000002s : 17: predicate.elim_not_effective 0.22% : 0.000004s : 17: predicate.elim_shapecalc_of_broadcastargs 1.20% : 0.000021s : 140: predicate.environ_add_const_eliminate 1.28% : 0.000022s : 140: predicate.environ_get_add_eliminate 1.25% : 0.000022s : 140: predicate.environ_get_depend_swap 1.84% : 0.000032s : 198: predicate.environ_get_eliminate 1.29% : 0.000022s : 140: predicate.environ_get_set_eliminate 2.25% : 0.000039s : 209: predicate.exchange_switch_depend_value 2.86% : 0.000050s : 209: predicate.float_depend_g_call 0.50% : 0.000009s : 58: predicate.float_environ_get_switch 0.70% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.34% : 0.000006s : 35: predicate.get_grad_eliminate 0.09% : 0.000002s : 17: predicate.graph_param_transform 0.48% : 0.000008s : 58: predicate.incorporate_call 0.44% : 0.000008s : 58: predicate.incorporate_call_switch 5.80% : 0.000101s : 459: predicate.inline 0.48% : 0.000008s : 35: predicate.inline_without_move 0.21% : 0.000004s : 35: predicate.j_node_and_user_rematch 0.46% : 0.000008s : 36: predicate.less_batch_normalization 1.69% : 0.000029s : 172: predicate.list_to_tuple_eliminator_ 2.58% : 0.000045s : 295: predicate.load_eliminater 0.50% : 0.000009s : 17: predicate.loop_unroll_after_grad 3.31% : 0.000057s : 314: predicate.loop_unroll_before_grad 1.50% : 0.000026s : 157: predicate.make_slice_get_slice_eliminator 0.48% : 0.000008s : 58: predicate.merge_addn 0.30% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.30% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.13% : 0.000020s : 123: predicate.minmaximum_grad 0.62% : 0.000011s : 17: predicate.mutable_eliminate 0.17% : 0.000003s : 17: predicate.opt_reshape 0.15% : 0.000003s : 17: predicate.parallel_virtual_node 3.64% : 0.000063s : 209: predicate.partial_defer_inline 1.43% : 0.000025s : 155: predicate.partial_eliminate 1.15% : 0.000020s : 123: predicate.print_const_string_wrapper 0.48% : 0.000008s : 53: predicate.reduce_all_const_elim 1.66% : 0.000029s : 123: predicate.reduce_eliminate 2.80% : 0.000048s : 295: predicate.redundant_stop_gradient_eliminater 0.18% : 0.000003s : 35: predicate.remove_not_recompute_node 1.02% : 0.000018s : 172: predicate.replace_applicator 0.20% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.20% : 0.000021s : 123: predicate.reshape_eliminate 0.33% : 0.000006s : 34: predicate.row_tensor_add_zeros_like 0.20% : 0.000004s : 17: predicate.row_tensor_eliminate 0.55% : 0.000010s : 34: predicate.same_eliminate 0.23% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.43% : 0.000007s : 35: predicate.shard_identity_eliminate 0.32% : 0.000006s : 34: predicate.special_op_eliminate 0.58% : 0.000010s : 58: predicate.specialize_transform 0.49% : 0.000009s : 34: predicate.split_environ_get_set_with_tuple_value 0.41% : 0.000007s : 35: predicate.stack_unstack_eliminate 0.16% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.07% : 0.000036s : 209: predicate.switch_defer_inline 3.00% : 0.000052s : 243: predicate.switch_layer_defer_inline 6.59% : 0.000114s : 646: predicate.switch_simplify 1.14% : 0.000020s : 123: predicate.tile_eliminate 1.13% : 0.000020s : 123: predicate.transpose_eliminate 1.64% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.79% : 0.000031s : 157: predicate.tuple_list_get_item_const_eliminator 1.54% : 0.000027s : 157: predicate.tuple_list_get_item_depend_reorder 3.08% : 0.000053s : 228: predicate.tuple_list_get_item_eliminator 1.57% : 0.000027s : 157: predicate.tuple_list_get_set_item_eliminator 2.32% : 0.000040s : 215: predicate.tuple_list_set_item_eliminator 1.62% : 0.000028s : 170: predicate.tuple_to_list_eliminator_ 2.45% : 0.000043s : 295: predicate.updatestate_pure_node_eliminater 3.02% : 0.000052s : 353: predicate.updatestate_useless_node_eliminater 0.18% : 0.000003s : 17: predicate.value_based_eliminate 0.37% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.37% : 0.000006s : 35: predicate.virtual_output_eliminate 0.15% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.19% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.017991 135 61.51% : 0.011066s : 55: func_graph_cloner_run.FuncGraphClonerGraph 38.49% : 0.006925s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.824298 209 0.00% : 0.000003s : 1: ForceFp32Comm 1.08% : 0.008890s : 1: add_attr 1.08% : 0.008866s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.02% : 0.000173s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.09% : 0.000746s : 1: auto_monad 0.01% : 0.000077s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.11% : 0.000869s : 1: bootstrap 0.01% : 0.000053s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000047s : 1: control_data_broadcast_order 0.00% : 0.000040s : 1: convert_after_rewriter 0.01% : 0.000086s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.00% : 0.000030s : 1: environ_conv 0.03% : 0.000220s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.01% : 0.000053s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000006s : 1: handle_group_info 0.00% : 0.000008s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000011s : 1: label_micro_interleaved_index 0.14% : 0.001135s : 1: loop_unroll 0.00% : 0.000006s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.11% : 0.000947s : 1: mutable_eliminate 0.00% : 0.000016s : 1: offloading_packed_experts 0.01% : 0.000049s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000060s : 1: opt.transform.mutable_eliminate 11.36% : 0.093610s : 95: opt.transform.opt_a 0.02% : 0.000134s : 1: opt.transform.opt_after_cconv 0.01% : 0.000119s : 1: opt.transform.opt_after_jit_grad 0.07% : 0.000554s : 28: opt.transform.opt_b 0.03% : 0.000232s : 2: opt.transform.opt_trans_graph 0.02% : 0.000187s : 4: opt.transform.symbol_engine_opt 13.46% : 0.110926s : 1: opt_a 0.04% : 0.000340s : 1: opt_after_cconv 0.10% : 0.000823s : 1: opt_after_jit_grad 0.10% : 0.000855s : 1: opt_b 14.34% : 0.118213s : 1: optimize 0.01% : 0.000055s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000021s : 1: order_py_execute_after_rewriter 0.01% : 0.000059s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000014s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000007s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000013s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000007s : 1: overlap_recompute_comm 0.00% : 0.000011s : 1: parallel-infer-symbol 0.00% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000283s : 1: pre_auto_parallel 0.03% : 0.000287s : 1: py_interpret_to_execute 0.01% : 0.000048s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000112s : 1: remove_dup_value 0.79% : 0.006523s : 1: renormalize.infer 0.77% : 0.006350s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.12% : 0.000981s : 1: rewriter_after_opt_a 0.10% : 0.000854s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000023s : 1: swap_dp_allreduce_reducescatter 0.07% : 0.000579s : 1: symbol_engine_optimizer 0.03% : 0.000281s : 1: tuple_transform 55.71% : 0.459244s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape2-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape2-16],max_mem:196.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:44:41.620.323 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 0.876373, [21] [bootstrap]: 0.00088574 [type_inference]: 0.758803 [event_method]: 0.00020718 [auto_monad]: 0.00075894 [graph_reusing]: 5.006e-05 [inline]: 2.93e-06 [add_attr]: 0.0378564, [1] [add_attr_with_inline]: 0.037837, [1] [Cycle 1]: 0.00031341, [2] [tag_attr]: 0.00020085 [meta_addattr_fg_expand]: 5.054e-05 [parallel-infer-symbol]: 3.81001e-06 [pre_auto_parallel]: 0.00025745 [insert-virtual-dataset]: 3.10998e-06 [parallel-infer-symbol-second]: 1.02998e-06 [dataset_repeat_opt]: 2.41e-06 [pipeline_split]: 1.94999e-06 [optimize]: 0.0764244, [53] [py_interpret_to_execute]: 0.00024335 [rewriter_before_opt_a]: 0.00078547 [opt_a]: 0.0697555, [2] [Cycle 1]: 0.0674427, [45] [expand_dump_flag]: 1.752e-05 [switch_simplify]: 0.00085041 [loop_unroll]: 0.017064 [a_1]: 0.0103277 [with_stream_mark]: 6.226e-05 [recompute_prepare]: 4.572e-05 [updatestate_depend_eliminate]: 2.319e-05 [updatestate_assign_eliminate]: 1.669e-05 [updatestate_loads_eliminate]: 1.871e-05 [parameter_eliminate]: 3.04999e-06 [a_2]: 0.00084006 [accelerated_algorithm]: 7.04e-05 [shard]: 3.35e-06 [meta_shard_fg_expand]: 2.067e-05 [shard_inline]: 2.54e-05 [merge_send_recv]: 1.771e-05 [auto_parallel]: 2.24e-05 [parallel]: 0.00014005 [flash_sp]: 1.816e-05 [merge_comm]: 1.647e-05 [allreduce_fusion]: 1.276e-05 [matmul_add_comm_reduction]: 1.984e-05 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 2.804e-05 [virtual_dataset]: 2.419e-05 [get_grad_eliminate_]: 2.352e-05 [virtual_output]: 2.442e-05 [merge_forward]: 1.429e-05 [cell_reuse_recompute_pass]: 1.81e-06 [offload_activation]: 2.276e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.785e-05 [merge_recompute_call_nodes]: 1.29998e-06 [before_grad]: 4.2e-05 [set_forward_comm_id_for_comm_node_pass]: 1.594e-05 [meta_fg_expand]: 1.694e-05 [flash_sp_send_recv_attached]: 6.23e-06 [receive_attached]: 2.29001e-06 [after_resolve]: 3.02e-05 [a_after_grad]: 3.75e-05 [renormalize]: 0.0347495 [add_forward_monad_depend]: 1.422e-05 [auto_monad_grad]: 3.60998e-06 [auto_monad_eliminator]: 9.241e-05 [cse]: 0.00187707 [a_3]: 0.00018215 [Cycle 2]: 0.00229137, [45] [expand_dump_flag]: 3.65e-06 [switch_simplify]: 2.827e-05 [loop_unroll]: 2.273e-05 [a_1]: 0.00068435 [with_stream_mark]: 4.028e-05 [recompute_prepare]: 2.659e-05 [updatestate_depend_eliminate]: 1.693e-05 [updatestate_assign_eliminate]: 1.195e-05 [updatestate_loads_eliminate]: 1.408e-05 [parameter_eliminate]: 2.86e-06 [a_2]: 0.00033835 [accelerated_algorithm]: 3.737e-05 [shard]: 2.71e-06 [meta_shard_fg_expand]: 9.86e-06 [shard_inline]: 2.396e-05 [merge_send_recv]: 2.222e-05 [auto_parallel]: 2.324e-05 [parallel]: 1.066e-05 [flash_sp]: 5.84e-06 [merge_comm]: 1.271e-05 [allreduce_fusion]: 1.249e-05 [matmul_add_comm_reduction]: 2.403e-05 [allreduce_slice_to_reducescatter]: 8.39995e-07 [virtual_shard_identity]: 2.56e-05 [virtual_dataset]: 2.045e-05 [get_grad_eliminate_]: 2.123e-05 [virtual_output]: 2.083e-05 [merge_forward]: 1.469e-05 [cell_reuse_recompute_pass]: 2.93e-06 [offload_activation]: 2.751e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.5e-05 [merge_recompute_call_nodes]: 1.62999e-06 [before_grad]: 3.782e-05 [set_forward_comm_id_for_comm_node_pass]: 1.536e-05 [meta_fg_expand]: 9.65002e-06 [flash_sp_send_recv_attached]: 1.83002e-06 [receive_attached]: 2.66e-06 [after_resolve]: 3.113e-05 [a_after_grad]: 3.534e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 4.25999e-06 [auto_monad_grad]: 2.51998e-06 [auto_monad_eliminator]: 6.996e-05 [cse]: 8.754e-05 [a_3]: 0.0001435 [py_interpret_to_execute_after_opt_a]: 3.701e-05 [slice_cell_reuse_recomputed_activation]: 2.89001e-06 [rewriter_after_opt_a]: 0.00092227 [convert_after_rewriter]: 3.122e-05 [order_py_execute_after_rewriter]: 1.495e-05 [mutable_eliminate]: 0.00090633 [opt_b]: 0.0008891, [1] [Cycle 1]: 0.00087708, [7] [b_1]: 0.00063636 [b_2]: 3.855e-05 [updatestate_depend_eliminate]: 2.09e-05 [updatestate_assign_eliminate]: 1.222e-05 [updatestate_loads_eliminate]: 1.509e-05 [renormalize]: 8.2e-07 [cse]: 9.615e-05 [optimize_parallel_all_gather_comm]: 4.705e-05 [overlap_param_gather]: 3.83001e-06 [cconv]: 4.412e-05 [loop_unroll]: 0.00058181 [opt_after_cconv]: 0.00031102, [1] [Cycle 1]: 0.00030062, [7] [c_1]: 0.00013067 [parameter_eliminate]: 5.24998e-06 [updatestate_depend_eliminate]: 1.931e-05 [updatestate_assign_eliminate]: 1.112e-05 [updatestate_loads_eliminate]: 1.278e-05 [cse]: 7.849e-05 [renormalize]: 4.89992e-07 [remove_dup_value]: 0.00010654 [tuple_transform]: 0.00026683, [1] [Cycle 1]: 0.00025963, [4] [d_1]: 0.00020447 [none_parameter_eliminate]: 2.73998e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 2.439e-05 [partial_unused_args_eliminate]: 1.83002e-06 [add_recomputation]: 0.0001596 [cse_after_recomputation]: 7.73e-05, [1] [Cycle 1]: 6.89e-05, [1] [cse]: 5.998e-05 [environ_conv]: 2.514e-05 [swap_dp_allreduce_reducescatter]: 2.02e-05 [bias_add_comm_swap]: 3.27997e-06 [label_micro_interleaved_index]: 4.94e-06 [label_fine_grained_interleaved_index]: 2.97002e-06 [merge_cast_opt]: 1.35999e-06 [slice_recompute_activation]: 2.21e-06 [micro_interleaved_order_control]: 2.69999e-06 [assign_add_opt]: 1.87001e-06 [ForceFp32Comm]: 1.19998e-06 [remove_cast_before_assign_add]: 9.09989e-07 [full_micro_interleaved_order_control]: 2.77002e-06 [reorder_send_recv_between_fp_bp]: 2.82002e-06 [comm_op_add_attrs]: 1.25999e-06 [add_comm_op_reuse_tag]: 9.50007e-07 [interleave_split_concat_branches]: 1.20001e-06 [interleave_parallel_branches]: 1.00999e-06 [overlap_opt_shard_in_pipeline]: 3.78001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.03002e-06 [control_data_broadcast_order]: 4.084e-05 [grouped_pairwise_exchange_alltoall]: 1.59e-06 [offloading_packed_experts]: 1.158e-05 [overlap_recompute_and_grad_model_parallel]: 1.055e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.41998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.76e-06 [overlap_recompute_comm]: 2.02001e-06 [overlap_grad_ring_attention]: 9.42999e-06 [overlap_grad_flash_sp]: 0.00012231 [begin_end_overlap_inline]: 7.7e-07 [split_matmul_comm_elemetwise]: 2.25002e-06 [split_layernorm_comm]: 1.65001e-06 [handle_group_info]: 9.79984e-07 [symbol_engine_optimizer]: 0.00060967, [1] [Cycle 1]: 0.00060273, [6] [build]: 0.0003584 [elim_shapecalc]: 3.713e-05 [elim_not_effective]: 5.596e-05 [opt_reshape]: 3.468e-05 [fold_const_symbol]: 6.058e-05 [renormalize]: 3.00002e-07 [detach_backward]: 2.83e-06 [pipeline_parallel_scheduler]: 1.45999e-06 [auto_monad_reorder]: 6.5e-05 [get_jit_bprop_graph]: 2.32001e-06 [rewriter_after_jit_bprop_graph]: 5.45001e-06 [opt_after_jit_grad]: 0.00066458 [validate]: 0.00011006 Sums bootstrap : 0.000886s : 0.11% type_inference : 0.758803s : 90.65% event_method : 0.000207s : 0.02% auto_monad : 0.000759s : 0.09% graph_reusing : 0.000050s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000201s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000051s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000257s : 0.03% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000243s : 0.03% optimize.rewriter_before_opt_a : 0.000785s : 0.09% optimize.opt_a.expand_dump_flag : 0.000021s : 0.00% optimize.opt_a.switch_simplify : 0.000879s : 0.10% optimize.opt_a.loop_unroll : 0.017087s : 2.04% optimize.opt_a.a_1 : 0.011012s : 1.32% optimize.opt_a.with_stream_mark : 0.000103s : 0.01% optimize.opt_a.recompute_prepare : 0.000072s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000040s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000029s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000033s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001178s : 0.14% optimize.opt_a.accelerated_algorithm : 0.000108s : 0.01% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000031s : 0.00% optimize.opt_a.shard_inline : 0.000049s : 0.01% optimize.opt_a.merge_send_recv : 0.000040s : 0.00% optimize.opt_a.auto_parallel : 0.000046s : 0.01% optimize.opt_a.parallel : 0.000151s : 0.02% optimize.opt_a.flash_sp : 0.000024s : 0.00% optimize.opt_a.merge_comm : 0.000029s : 0.00% optimize.opt_a.allreduce_fusion : 0.000025s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000044s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000054s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000045s : 0.01% optimize.opt_a.virtual_output : 0.000045s : 0.01% optimize.opt_a.merge_forward : 0.000029s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000050s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000093s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000080s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000031s : 0.00% optimize.opt_a.meta_fg_expand : 0.000027s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000061s : 0.01% optimize.opt_a.a_after_grad : 0.000073s : 0.01% optimize.opt_a.renormalize : 0.034750s : 4.15% optimize.opt_a.add_forward_monad_depend : 0.000018s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000162s : 0.02% optimize.opt_a.cse : 0.001965s : 0.23% optimize.opt_a.a_3 : 0.000326s : 0.04% optimize.py_interpret_to_execute_after_opt_a : 0.000037s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000922s : 0.11% optimize.convert_after_rewriter : 0.000031s : 0.00% optimize.order_py_execute_after_rewriter : 0.000015s : 0.00% optimize.mutable_eliminate : 0.000906s : 0.11% optimize.opt_b.b_1 : 0.000636s : 0.08% optimize.opt_b.b_2 : 0.000039s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000021s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000096s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000047s : 0.01% optimize.overlap_param_gather : 0.000004s : 0.00% optimize.cconv : 0.000044s : 0.01% optimize.loop_unroll : 0.000582s : 0.07% optimize.opt_after_cconv.c_1 : 0.000131s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000019s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.cse : 0.000078s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000107s : 0.01% optimize.tuple_transform.d_1 : 0.000204s : 0.02% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000024s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000160s : 0.02% optimize.cse_after_recomputation.cse : 0.000060s : 0.01% optimize.environ_conv : 0.000025s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000020s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000041s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000012s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000011s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000122s : 0.01% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000358s : 0.04% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000037s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000056s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000035s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000061s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000065s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000665s : 0.08% validate : 0.000110s : 0.01% Time group info: ------[substitution.] 0.004143 569 0.41% : 0.000017s : 8: substitution.depend_value_elim 0.40% : 0.000017s : 14: substitution.elim_not_effective 0.45% : 0.000019s : 12: substitution.float_tuple_getitem_switch 0.68% : 0.000028s : 14: substitution.fold_const_symbol 0.38% : 0.000016s : 17: substitution.graph_param_transform 70.89% : 0.002937s : 71: substitution.inline 0.34% : 0.000014s : 29: substitution.j_node_and_user_rematch 0.90% : 0.000037s : 6: substitution.less_batch_normalization 1.58% : 0.000065s : 2: substitution.list_to_tuple_eliminator_ 0.15% : 0.000006s : 6: substitution.load_eliminater 0.87% : 0.000036s : 21: substitution.minmaximum_grad 0.21% : 0.000009s : 2: substitution.opt_reshape 0.49% : 0.000020s : 29: substitution.remove_not_recompute_node 0.22% : 0.000009s : 4: substitution.replace_old_param 1.75% : 0.000072s : 18: substitution.reshape_eliminate 1.29% : 0.000053s : 24: substitution.switch_simplify 3.60% : 0.000149s : 51: substitution.tuple_list_convert_item_index_to_positive 1.97% : 0.000081s : 51: substitution.tuple_list_get_item_const_eliminator 2.61% : 0.000108s : 51: substitution.tuple_list_get_item_depend_reorder 5.84% : 0.000242s : 73: substitution.tuple_list_get_item_eliminator 2.52% : 0.000105s : 51: substitution.tuple_list_get_set_item_eliminator 1.89% : 0.000078s : 6: substitution.updatestate_pure_node_eliminater 0.55% : 0.000023s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.758484 2 87.58% : 0.664287s : 1: type_inference.infer 12.42% : 0.094197s : 1: type_inference.specialize ------[replace.] 0.001304 115 4.06% : 0.000053s : 5: replace.depend_value_elim 60.70% : 0.000792s : 71: replace.inline 3.58% : 0.000047s : 2: replace.list_to_tuple_eliminator_ 23.39% : 0.000305s : 24: replace.switch_simplify 8.28% : 0.000108s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.003027 115 0.08% : 0.000002s : 5: match.depend_value_elim 95.37% : 0.002887s : 71: match.inline 2.03% : 0.000062s : 2: match.list_to_tuple_eliminator_ 1.28% : 0.000039s : 24: match.switch_simplify 1.23% : 0.000037s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001829 9701 1.01% : 0.000018s : 123: predicate.accumulaten_eliminater 0.32% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.46% : 0.000008s : 58: predicate.addn_check_dump 1.20% : 0.000022s : 123: predicate.addn_zero_filter 1.04% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 2.10% : 0.000038s : 181: predicate.arithmetic_simplify 1.17% : 0.000021s : 123: predicate.cast_eliminate 0.30% : 0.000005s : 34: predicate.check_bprop_eliminate 0.47% : 0.000009s : 58: predicate.compare_switch_simplify 0.09% : 0.000002s : 17: predicate.const_output_eliminate 0.54% : 0.000010s : 61: predicate.depend_value_elim 1.09% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 1.19% : 0.000022s : 123: predicate.dict_get_item_eliminator 0.99% : 0.000018s : 123: predicate.dict_set_item_eliminator 4.48% : 0.000082s : 34: predicate.dumpgradient_eliminate 0.09% : 0.000002s : 17: predicate.elim_not_effective 0.18% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.16% : 0.000021s : 140: predicate.environ_add_const_eliminate 1.12% : 0.000021s : 140: predicate.environ_get_add_eliminate 1.30% : 0.000024s : 140: predicate.environ_get_depend_swap 1.72% : 0.000031s : 198: predicate.environ_get_eliminate 1.13% : 0.000021s : 140: predicate.environ_get_set_eliminate 1.71% : 0.000031s : 209: predicate.exchange_switch_depend_value 2.88% : 0.000053s : 209: predicate.float_depend_g_call 0.47% : 0.000009s : 58: predicate.float_environ_get_switch 0.64% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.07% : 0.000001s : 17: predicate.fold_const_symbol 0.34% : 0.000006s : 35: predicate.get_grad_eliminate 0.10% : 0.000002s : 17: predicate.graph_param_transform 0.53% : 0.000010s : 58: predicate.incorporate_call 0.42% : 0.000008s : 58: predicate.incorporate_call_switch 5.79% : 0.000106s : 459: predicate.inline 0.43% : 0.000008s : 35: predicate.inline_without_move 0.13% : 0.000002s : 35: predicate.j_node_and_user_rematch 0.50% : 0.000009s : 36: predicate.less_batch_normalization 1.45% : 0.000027s : 172: predicate.list_to_tuple_eliminator_ 2.42% : 0.000044s : 295: predicate.load_eliminater 0.35% : 0.000006s : 17: predicate.loop_unroll_after_grad 6.04% : 0.000110s : 314: predicate.loop_unroll_before_grad 1.38% : 0.000025s : 157: predicate.make_slice_get_slice_eliminator 0.49% : 0.000009s : 58: predicate.merge_addn 0.26% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.26% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.04% : 0.000019s : 123: predicate.minmaximum_grad 0.42% : 0.000008s : 17: predicate.mutable_eliminate 0.17% : 0.000003s : 17: predicate.opt_reshape 0.16% : 0.000003s : 17: predicate.parallel_virtual_node 3.21% : 0.000059s : 209: predicate.partial_defer_inline 1.39% : 0.000025s : 155: predicate.partial_eliminate 1.02% : 0.000019s : 123: predicate.print_const_string_wrapper 0.46% : 0.000008s : 53: predicate.reduce_all_const_elim 1.39% : 0.000025s : 123: predicate.reduce_eliminate 2.39% : 0.000044s : 295: predicate.redundant_stop_gradient_eliminater 0.17% : 0.000003s : 35: predicate.remove_not_recompute_node 1.00% : 0.000018s : 172: predicate.replace_applicator 0.18% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.43% : 0.000026s : 123: predicate.reshape_eliminate 0.28% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.18% : 0.000003s : 17: predicate.row_tensor_eliminate 0.41% : 0.000007s : 34: predicate.same_eliminate 0.22% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.37% : 0.000007s : 35: predicate.shard_identity_eliminate 0.31% : 0.000006s : 34: predicate.special_op_eliminate 1.26% : 0.000023s : 58: predicate.specialize_transform 0.40% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.38% : 0.000007s : 35: predicate.stack_unstack_eliminate 0.15% : 0.000003s : 17: predicate.switch_call_monad_eliminater 1.93% : 0.000035s : 209: predicate.switch_defer_inline 2.10% : 0.000038s : 243: predicate.switch_layer_defer_inline 6.07% : 0.000111s : 646: predicate.switch_simplify 1.01% : 0.000018s : 123: predicate.tile_eliminate 1.07% : 0.000020s : 123: predicate.transpose_eliminate 1.63% : 0.000030s : 157: predicate.tuple_list_convert_item_index_to_positive 1.68% : 0.000031s : 157: predicate.tuple_list_get_item_const_eliminator 1.63% : 0.000030s : 157: predicate.tuple_list_get_item_depend_reorder 3.07% : 0.000056s : 228: predicate.tuple_list_get_item_eliminator 1.55% : 0.000028s : 157: predicate.tuple_list_get_set_item_eliminator 2.06% : 0.000038s : 215: predicate.tuple_list_set_item_eliminator 1.51% : 0.000028s : 170: predicate.tuple_to_list_eliminator_ 2.41% : 0.000044s : 295: predicate.updatestate_pure_node_eliminater 2.87% : 0.000053s : 353: predicate.updatestate_useless_node_eliminater 0.16% : 0.000003s : 17: predicate.value_based_eliminate 0.31% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.32% : 0.000006s : 35: predicate.virtual_output_eliminate 0.14% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.17% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.018961 135 56.88% : 0.010785s : 55: func_graph_cloner_run.FuncGraphClonerGraph 43.12% : 0.008176s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.057499 209 0.00% : 0.000004s : 1: ForceFp32Comm 3.58% : 0.037865s : 1: add_attr 3.58% : 0.037843s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.02% : 0.000168s : 1: add_recomputation 0.00% : 0.000006s : 1: assign_add_opt 0.07% : 0.000776s : 1: auto_monad 0.01% : 0.000073s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.09% : 0.000950s : 1: bootstrap 0.00% : 0.000048s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000046s : 1: control_data_broadcast_order 0.00% : 0.000038s : 1: convert_after_rewriter 0.01% : 0.000081s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000029s : 1: environ_conv 0.02% : 0.000219s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.01% : 0.000056s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000007s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.06% : 0.000596s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000923s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.00% : 0.000039s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000049s : 1: opt.transform.mutable_eliminate 2.93% : 0.031032s : 95: opt.transform.opt_a 0.01% : 0.000129s : 1: opt.transform.opt_after_cconv 0.01% : 0.000077s : 1: opt.transform.opt_after_jit_grad 0.06% : 0.000630s : 28: opt.transform.opt_b 0.02% : 0.000225s : 2: opt.transform.opt_trans_graph 0.02% : 0.000181s : 4: opt.transform.symbol_engine_opt 6.60% : 0.069761s : 1: opt_a 0.03% : 0.000316s : 1: opt_after_cconv 0.06% : 0.000680s : 1: opt_after_jit_grad 0.08% : 0.000894s : 1: opt_b 7.23% : 0.076431s : 1: optimize 0.00% : 0.000052s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000019s : 1: order_py_execute_after_rewriter 0.01% : 0.000131s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000007s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000014s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000269s : 1: pre_auto_parallel 0.02% : 0.000253s : 1: py_interpret_to_execute 0.00% : 0.000042s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.01% : 0.000114s : 1: remove_dup_value 2.52% : 0.026651s : 1: renormalize.infer 0.76% : 0.008077s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.09% : 0.000937s : 1: rewriter_after_opt_a 0.08% : 0.000799s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000024s : 1: swap_dp_allreduce_reducescatter 0.06% : 0.000613s : 1: symbol_engine_optimizer 0.03% : 0.000272s : 1: tuple_transform 71.76% : 0.758835s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape3-1] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape3-1],max_mem:208.0M [WARNING] ME(167695:281473322254128,MainProcess):2026-01-29-17:44:54.720.776 [mindspore/context.py:1334] For 'context.set_context', the parameter 'device_target' will be deprecated and removed in a future version. Please use the api mindspore.set_device() instead. TotalTime = 0.316385, [21] [bootstrap]: 0.00085227 [type_inference]: 0.275 [event_method]: 0.00019952 [auto_monad]: 0.00102096 [graph_reusing]: 4.892e-05 [inline]: 2.56998e-06 [add_attr]: 0.00503694, [1] [add_attr_with_inline]: 0.00501998, [1] [Cycle 1]: 0.00031, [2] [tag_attr]: 0.00019588 [meta_addattr_fg_expand]: 5.913e-05 [parallel-infer-symbol]: 4.34997e-06 [pre_auto_parallel]: 0.00024146 [insert-virtual-dataset]: 2.79001e-06 [parallel-infer-symbol-second]: 1.50999e-06 [dataset_repeat_opt]: 2.10002e-06 [pipeline_split]: 1.54e-06 [optimize]: 0.0329572, [53] [py_interpret_to_execute]: 0.00024854 [rewriter_before_opt_a]: 0.00077698 [opt_a]: 0.0269534, [2] [Cycle 1]: 0.0246342, [45] [expand_dump_flag]: 1.513e-05 [switch_simplify]: 0.00078529 [loop_unroll]: 0.00035827 [a_1]: 0.00877767 [with_stream_mark]: 4.268e-05 [recompute_prepare]: 3.551e-05 [updatestate_depend_eliminate]: 1.988e-05 [updatestate_assign_eliminate]: 1.517e-05 [updatestate_loads_eliminate]: 1.784e-05 [parameter_eliminate]: 2.91e-06 [a_2]: 0.00075206 [accelerated_algorithm]: 5.628e-05 [shard]: 2.34999e-06 [meta_shard_fg_expand]: 1.729e-05 [shard_inline]: 2.243e-05 [merge_send_recv]: 1.986e-05 [auto_parallel]: 1.957e-05 [parallel]: 0.00010679 [flash_sp]: 1.444e-05 [merge_comm]: 1.421e-05 [allreduce_fusion]: 1.271e-05 [matmul_add_comm_reduction]: 2.057e-05 [allreduce_slice_to_reducescatter]: 6.59988e-07 [virtual_shard_identity]: 2.972e-05 [virtual_dataset]: 2.653e-05 [get_grad_eliminate_]: 2.271e-05 [virtual_output]: 2.315e-05 [merge_forward]: 1.343e-05 [cell_reuse_recompute_pass]: 2.68998e-06 [offload_activation]: 2.426e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.816e-05 [merge_recompute_call_nodes]: 1.40999e-06 [before_grad]: 3.817e-05 [set_forward_comm_id_for_comm_node_pass]: 1.346e-05 [meta_fg_expand]: 1.767e-05 [flash_sp_send_recv_attached]: 4.95001e-06 [receive_attached]: 2.22999e-06 [after_resolve]: 2.85e-05 [a_after_grad]: 3.464e-05 [renormalize]: 0.0111279 [add_forward_monad_depend]: 1.233e-05 [auto_monad_grad]: 2.66999e-06 [auto_monad_eliminator]: 7.424e-05 [cse]: 0.00139025 [a_3]: 0.0001707 [Cycle 2]: 0.00230313, [45] [expand_dump_flag]: 3.09999e-06 [switch_simplify]: 2.509e-05 [loop_unroll]: 2.101e-05 [a_1]: 0.00087307 [with_stream_mark]: 3.014e-05 [recompute_prepare]: 2.409e-05 [updatestate_depend_eliminate]: 1.602e-05 [updatestate_assign_eliminate]: 1.145e-05 [updatestate_loads_eliminate]: 1.388e-05 [parameter_eliminate]: 2.54999e-06 [a_2]: 0.00031055 [accelerated_algorithm]: 2.993e-05 [shard]: 2.49001e-06 [meta_shard_fg_expand]: 7.1e-06 [shard_inline]: 1.993e-05 [merge_send_recv]: 1.865e-05 [auto_parallel]: 2.025e-05 [parallel]: 9.39998e-06 [flash_sp]: 4.08999e-06 [merge_comm]: 1.233e-05 [allreduce_fusion]: 1.221e-05 [matmul_add_comm_reduction]: 2.256e-05 [allreduce_slice_to_reducescatter]: 6.60017e-07 [virtual_shard_identity]: 2.234e-05 [virtual_dataset]: 2.012e-05 [get_grad_eliminate_]: 2.135e-05 [virtual_output]: 2.019e-05 [merge_forward]: 1.236e-05 [cell_reuse_recompute_pass]: 3.03998e-06 [offload_activation]: 2.377e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.85e-05 [merge_recompute_call_nodes]: 1.45999e-06 [before_grad]: 3.533e-05 [set_forward_comm_id_for_comm_node_pass]: 1.3e-05 [meta_fg_expand]: 9.49e-06 [flash_sp_send_recv_attached]: 1.66e-06 [receive_attached]: 2.49999e-06 [after_resolve]: 2.598e-05 [a_after_grad]: 3.146e-05 [renormalize]: 1.59984e-07 [add_forward_monad_depend]: 2.62001e-06 [auto_monad_grad]: 2.09999e-06 [auto_monad_eliminator]: 4.969e-05 [cse]: 0.00010603 [a_3]: 0.00013721 [py_interpret_to_execute_after_opt_a]: 2.832e-05 [slice_cell_reuse_recomputed_activation]: 2.31e-06 [rewriter_after_opt_a]: 0.00081604 [convert_after_rewriter]: 2.49e-05 [order_py_execute_after_rewriter]: 1.475e-05 [mutable_eliminate]: 0.0007305 [opt_b]: 0.00080931, [1] [Cycle 1]: 0.00080121, [7] [b_1]: 0.00061179 [b_2]: 2.306e-05 [updatestate_depend_eliminate]: 1.654e-05 [updatestate_assign_eliminate]: 1.112e-05 [updatestate_loads_eliminate]: 1.291e-05 [renormalize]: 5.99975e-07 [cse]: 7.921e-05 [optimize_parallel_all_gather_comm]: 4.073e-05 [overlap_param_gather]: 3.39001e-06 [cconv]: 3.811e-05 [loop_unroll]: 0.00053392 [opt_after_cconv]: 0.00028769, [1] [Cycle 1]: 0.00027958, [7] [c_1]: 0.00012416 [parameter_eliminate]: 4.71002e-06 [updatestate_depend_eliminate]: 1.743e-05 [updatestate_assign_eliminate]: 1.055e-05 [updatestate_loads_eliminate]: 1.295e-05 [cse]: 7.077e-05 [renormalize]: 4.40021e-07 [remove_dup_value]: 9.117e-05 [tuple_transform]: 0.00024945, [1] [Cycle 1]: 0.00024314, [4] [d_1]: 0.00019581 [none_parameter_eliminate]: 2.36e-06 [renormalize]: 3.30008e-07 [switch_simplify]: 2.146e-05 [partial_unused_args_eliminate]: 1.99999e-06 [add_recomputation]: 0.00015524 [cse_after_recomputation]: 6.869e-05, [1] [Cycle 1]: 6.244e-05, [1] [cse]: 5.551e-05 [environ_conv]: 2.116e-05 [swap_dp_allreduce_reducescatter]: 1.833e-05 [bias_add_comm_swap]: 2.41998e-06 [label_micro_interleaved_index]: 4.47998e-06 [label_fine_grained_interleaved_index]: 3.06001e-06 [merge_cast_opt]: 1.32e-06 [slice_recompute_activation]: 2.43002e-06 [micro_interleaved_order_control]: 2.48e-06 [assign_add_opt]: 1.49e-06 [ForceFp32Comm]: 7.39994e-07 [remove_cast_before_assign_add]: 1.05001e-06 [full_micro_interleaved_order_control]: 2.37999e-06 [reorder_send_recv_between_fp_bp]: 3.25e-06 [comm_op_add_attrs]: 1.09e-06 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.19e-06 [interleave_parallel_branches]: 1.20999e-06 [overlap_opt_shard_in_pipeline]: 3.91999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.97999e-06 [control_data_broadcast_order]: 3.81e-05 [grouped_pairwise_exchange_alltoall]: 1.58002e-06 [offloading_packed_experts]: 1.033e-05 [overlap_recompute_and_grad_model_parallel]: 1.013e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.23002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.44998e-06 [overlap_recompute_comm]: 2.28002e-06 [overlap_grad_ring_attention]: 9.10001e-06 [overlap_grad_flash_sp]: 9.122e-05 [begin_end_overlap_inline]: 5.69999e-07 [split_matmul_comm_elemetwise]: 2.80002e-06 [split_layernorm_comm]: 1.71e-06 [handle_group_info]: 1.20999e-06 [symbol_engine_optimizer]: 0.00054894, [1] [Cycle 1]: 0.00054319, [6] [build]: 0.00031936 [elim_shapecalc]: 3.285e-05 [elim_not_effective]: 5.351e-05 [opt_reshape]: 3.27e-05 [fold_const_symbol]: 5.594e-05 [renormalize]: 2.89991e-07 [detach_backward]: 2.36e-06 [pipeline_parallel_scheduler]: 1.84e-06 [auto_monad_reorder]: 6.103e-05 [get_jit_bprop_graph]: 2.14999e-06 [rewriter_after_jit_bprop_graph]: 4.40999e-06 [opt_after_jit_grad]: 0.00059706 [validate]: 0.00010136 Sums bootstrap : 0.000852s : 0.27% type_inference : 0.275000s : 88.64% event_method : 0.000200s : 0.06% auto_monad : 0.001021s : 0.33% graph_reusing : 0.000049s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000196s : 0.06% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000059s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000241s : 0.08% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000249s : 0.08% optimize.rewriter_before_opt_a : 0.000777s : 0.25% optimize.opt_a.expand_dump_flag : 0.000018s : 0.01% optimize.opt_a.switch_simplify : 0.000810s : 0.26% optimize.opt_a.loop_unroll : 0.000379s : 0.12% optimize.opt_a.a_1 : 0.009651s : 3.11% optimize.opt_a.with_stream_mark : 0.000073s : 0.02% optimize.opt_a.recompute_prepare : 0.000060s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000036s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000027s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000032s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.001063s : 0.34% optimize.opt_a.accelerated_algorithm : 0.000086s : 0.03% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000024s : 0.01% optimize.opt_a.shard_inline : 0.000042s : 0.01% optimize.opt_a.merge_send_recv : 0.000039s : 0.01% optimize.opt_a.auto_parallel : 0.000040s : 0.01% optimize.opt_a.parallel : 0.000116s : 0.04% optimize.opt_a.flash_sp : 0.000019s : 0.01% optimize.opt_a.merge_comm : 0.000027s : 0.01% optimize.opt_a.allreduce_fusion : 0.000025s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000043s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000052s : 0.02% optimize.opt_a.virtual_dataset : 0.000047s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.01% optimize.opt_a.virtual_output : 0.000043s : 0.01% optimize.opt_a.merge_forward : 0.000026s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000048s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000087s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000074s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000026s : 0.01% optimize.opt_a.meta_fg_expand : 0.000027s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000054s : 0.02% optimize.opt_a.a_after_grad : 0.000066s : 0.02% optimize.opt_a.renormalize : 0.011128s : 3.59% optimize.opt_a.add_forward_monad_depend : 0.000015s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000124s : 0.04% optimize.opt_a.cse : 0.001496s : 0.48% optimize.opt_a.a_3 : 0.000308s : 0.10% optimize.py_interpret_to_execute_after_opt_a : 0.000028s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000816s : 0.26% optimize.convert_after_rewriter : 0.000025s : 0.01% optimize.order_py_execute_after_rewriter : 0.000015s : 0.00% optimize.mutable_eliminate : 0.000730s : 0.24% optimize.opt_b.b_1 : 0.000612s : 0.20% optimize.opt_b.b_2 : 0.000023s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000017s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000079s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000041s : 0.01% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000038s : 0.01% optimize.loop_unroll : 0.000534s : 0.17% optimize.opt_after_cconv.c_1 : 0.000124s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000017s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.cse : 0.000071s : 0.02% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000091s : 0.03% optimize.tuple_transform.d_1 : 0.000196s : 0.06% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000021s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000155s : 0.05% optimize.cse_after_recomputation.cse : 0.000056s : 0.02% optimize.environ_conv : 0.000021s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000018s : 0.01% optimize.bias_add_comm_swap : 0.000002s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000038s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000010s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000010s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000091s : 0.03% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000319s : 0.10% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000033s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000054s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000033s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000056s : 0.02% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000061s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000597s : 0.19% validate : 0.000101s : 0.03% Time group info: ------[substitution.] 0.003164 569 0.49% : 0.000016s : 8: substitution.depend_value_elim 0.45% : 0.000014s : 14: substitution.elim_not_effective 0.51% : 0.000016s : 12: substitution.float_tuple_getitem_switch 0.79% : 0.000025s : 14: substitution.fold_const_symbol 0.48% : 0.000015s : 17: substitution.graph_param_transform 66.43% : 0.002102s : 71: substitution.inline 0.40% : 0.000013s : 29: substitution.j_node_and_user_rematch 0.95% : 0.000030s : 6: substitution.less_batch_normalization 1.37% : 0.000043s : 2: substitution.list_to_tuple_eliminator_ 0.21% : 0.000007s : 6: substitution.load_eliminater 1.15% : 0.000036s : 21: substitution.minmaximum_grad 0.24% : 0.000008s : 2: substitution.opt_reshape 0.64% : 0.000020s : 29: substitution.remove_not_recompute_node 0.22% : 0.000007s : 4: substitution.replace_old_param 1.99% : 0.000063s : 18: substitution.reshape_eliminate 1.48% : 0.000047s : 24: substitution.switch_simplify 4.52% : 0.000143s : 51: substitution.tuple_list_convert_item_index_to_positive 2.43% : 0.000077s : 51: substitution.tuple_list_get_item_const_eliminator 3.12% : 0.000099s : 51: substitution.tuple_list_get_item_depend_reorder 6.02% : 0.000190s : 73: substitution.tuple_list_get_item_eliminator 3.12% : 0.000099s : 51: substitution.tuple_list_get_set_item_eliminator 2.37% : 0.000075s : 6: substitution.updatestate_pure_node_eliminater 0.60% : 0.000019s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.274672 2 93.50% : 0.256821s : 1: type_inference.infer 6.50% : 0.017851s : 1: type_inference.specialize ------[replace.] 0.001055 115 3.26% : 0.000034s : 5: replace.depend_value_elim 57.45% : 0.000606s : 71: replace.inline 3.60% : 0.000038s : 2: replace.list_to_tuple_eliminator_ 26.45% : 0.000279s : 24: replace.switch_simplify 9.25% : 0.000098s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002163 115 0.10% : 0.000002s : 5: match.depend_value_elim 95.09% : 0.002057s : 71: match.inline 1.90% : 0.000041s : 2: match.list_to_tuple_eliminator_ 1.52% : 0.000033s : 24: match.switch_simplify 1.38% : 0.000030s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001633 9701 1.25% : 0.000020s : 123: predicate.accumulaten_eliminater 0.32% : 0.000005s : 17: predicate.ad_related_special_op_eliminate 0.50% : 0.000008s : 58: predicate.addn_check_dump 1.23% : 0.000020s : 123: predicate.addn_zero_filter 1.19% : 0.000019s : 123: predicate.adjust_all_reduce_mul_add 2.36% : 0.000038s : 181: predicate.arithmetic_simplify 1.16% : 0.000019s : 123: predicate.cast_eliminate 0.31% : 0.000005s : 34: predicate.check_bprop_eliminate 0.48% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.58% : 0.000009s : 61: predicate.depend_value_elim 1.22% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 1.28% : 0.000021s : 123: predicate.dict_get_item_eliminator 1.11% : 0.000018s : 123: predicate.dict_set_item_eliminator 3.69% : 0.000060s : 34: predicate.dumpgradient_eliminate 0.10% : 0.000002s : 17: predicate.elim_not_effective 0.21% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.22% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.17% : 0.000019s : 140: predicate.environ_get_add_eliminate 1.21% : 0.000020s : 140: predicate.environ_get_depend_swap 1.75% : 0.000029s : 198: predicate.environ_get_eliminate 1.19% : 0.000019s : 140: predicate.environ_get_set_eliminate 1.89% : 0.000031s : 209: predicate.exchange_switch_depend_value 2.70% : 0.000044s : 209: predicate.float_depend_g_call 0.49% : 0.000008s : 58: predicate.float_environ_get_switch 0.66% : 0.000011s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.35% : 0.000006s : 35: predicate.get_grad_eliminate 0.10% : 0.000002s : 17: predicate.graph_param_transform 0.49% : 0.000008s : 58: predicate.incorporate_call 0.45% : 0.000007s : 58: predicate.incorporate_call_switch 5.71% : 0.000093s : 459: predicate.inline 0.40% : 0.000007s : 35: predicate.inline_without_move 0.15% : 0.000002s : 35: predicate.j_node_and_user_rematch 0.48% : 0.000008s : 36: predicate.less_batch_normalization 1.68% : 0.000027s : 172: predicate.list_to_tuple_eliminator_ 2.79% : 0.000046s : 295: predicate.load_eliminater 0.36% : 0.000006s : 17: predicate.loop_unroll_after_grad 3.43% : 0.000056s : 314: predicate.loop_unroll_before_grad 1.61% : 0.000026s : 157: predicate.make_slice_get_slice_eliminator 0.49% : 0.000008s : 58: predicate.merge_addn 0.29% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.29% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.16% : 0.000019s : 123: predicate.minmaximum_grad 0.35% : 0.000006s : 17: predicate.mutable_eliminate 0.18% : 0.000003s : 17: predicate.opt_reshape 0.18% : 0.000003s : 17: predicate.parallel_virtual_node 2.99% : 0.000049s : 209: predicate.partial_defer_inline 1.48% : 0.000024s : 155: predicate.partial_eliminate 1.17% : 0.000019s : 123: predicate.print_const_string_wrapper 0.46% : 0.000007s : 53: predicate.reduce_all_const_elim 1.57% : 0.000026s : 123: predicate.reduce_eliminate 2.52% : 0.000041s : 295: predicate.redundant_stop_gradient_eliminater 0.16% : 0.000003s : 35: predicate.remove_not_recompute_node 1.05% : 0.000017s : 172: predicate.replace_applicator 0.18% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.24% : 0.000020s : 123: predicate.reshape_eliminate 0.32% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.17% : 0.000003s : 17: predicate.row_tensor_eliminate 0.43% : 0.000007s : 34: predicate.same_eliminate 0.21% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.38% : 0.000006s : 35: predicate.shard_identity_eliminate 0.34% : 0.000006s : 34: predicate.special_op_eliminate 0.57% : 0.000009s : 58: predicate.specialize_transform 0.38% : 0.000006s : 34: predicate.split_environ_get_set_with_tuple_value 0.35% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.17% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.11% : 0.000034s : 209: predicate.switch_defer_inline 2.29% : 0.000037s : 243: predicate.switch_layer_defer_inline 6.66% : 0.000109s : 646: predicate.switch_simplify 1.09% : 0.000018s : 123: predicate.tile_eliminate 1.21% : 0.000020s : 123: predicate.transpose_eliminate 1.67% : 0.000027s : 157: predicate.tuple_list_convert_item_index_to_positive 1.65% : 0.000027s : 157: predicate.tuple_list_get_item_const_eliminator 1.53% : 0.000025s : 157: predicate.tuple_list_get_item_depend_reorder 2.92% : 0.000048s : 228: predicate.tuple_list_get_item_eliminator 1.63% : 0.000027s : 157: predicate.tuple_list_get_set_item_eliminator 2.22% : 0.000036s : 215: predicate.tuple_list_set_item_eliminator 1.49% : 0.000024s : 170: predicate.tuple_to_list_eliminator_ 2.57% : 0.000042s : 295: predicate.updatestate_pure_node_eliminater 3.11% : 0.000051s : 353: predicate.updatestate_useless_node_eliminater 0.18% : 0.000003s : 17: predicate.value_based_eliminate 0.35% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.37% : 0.000006s : 35: predicate.virtual_output_eliminate 0.15% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.20% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.014971 135 60.17% : 0.009008s : 55: func_graph_cloner_run.FuncGraphClonerGraph 39.83% : 0.005964s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.379239 209 0.00% : 0.000003s : 1: ForceFp32Comm 1.33% : 0.005044s : 1: add_attr 1.33% : 0.005025s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.04% : 0.000161s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.27% : 0.001039s : 1: auto_monad 0.02% : 0.000066s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.24% : 0.000911s : 1: bootstrap 0.01% : 0.000042s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000041s : 1: control_data_broadcast_order 0.01% : 0.000030s : 1: convert_after_rewriter 0.02% : 0.000072s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000024s : 1: environ_conv 0.06% : 0.000212s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.02% : 0.000058s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.14% : 0.000544s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.20% : 0.000740s : 1: mutable_eliminate 0.00% : 0.000013s : 1: offloading_packed_experts 0.01% : 0.000036s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000037s : 1: opt.transform.mutable_eliminate 3.36% : 0.012726s : 95: opt.transform.opt_a 0.03% : 0.000123s : 1: opt.transform.opt_after_cconv 0.02% : 0.000072s : 1: opt.transform.opt_after_jit_grad 0.16% : 0.000599s : 28: opt.transform.opt_b 0.06% : 0.000214s : 2: opt.transform.opt_trans_graph 0.04% : 0.000169s : 4: opt.transform.symbol_engine_opt 7.11% : 0.026958s : 1: opt_a 0.08% : 0.000291s : 1: opt_after_cconv 0.16% : 0.000608s : 1: opt_after_jit_grad 0.21% : 0.000813s : 1: opt_b 8.69% : 0.032964s : 1: optimize 0.01% : 0.000045s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000019s : 1: order_py_execute_after_rewriter 0.03% : 0.000096s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000012s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000007s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000013s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.07% : 0.000250s : 1: pre_auto_parallel 0.07% : 0.000258s : 1: py_interpret_to_execute 0.01% : 0.000032s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.03% : 0.000097s : 1: remove_dup_value 1.53% : 0.005810s : 1: renormalize.infer 1.40% : 0.005301s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.22% : 0.000824s : 1: rewriter_after_opt_a 0.21% : 0.000789s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.01% : 0.000022s : 1: swap_dp_allreduce_reducescatter 0.15% : 0.000553s : 1: symbol_engine_optimizer 0.07% : 0.000254s : 1: tuple_transform 72.52% : 0.275027s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_0[True-mstype0-k_n_shape3-16] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_0[True-mstype0-k_n_shape3-16],max_mem:208.0M =============================== warnings summary =============================== ../../../../../../../../../../usr/local/Ascend/cann-8.5.0/python/site-packages/tbe/dsl/classifier/transdata/transdata_classifier.py:222 /usr/local/Ascend/cann-8.5.0/python/site-packages/tbe/dsl/classifier/transdata/transdata_classifier.py:222: DeprecationWarning: invalid escape sequence \B """ ../../../../../../../../../../usr/local/Ascend/cann-8.5.0/python/site-packages/tbe/dsl/unify_schedule/vector/transdata/common/graph/transdata_graph_info.py:143 /usr/local/Ascend/cann-8.5.0/python/site-packages/tbe/dsl/unify_schedule/vector/transdata/common/graph/transdata_graph_info.py:143: DeprecationWarning: invalid escape sequence \c """ ../../../../../../../../../../usr/local/Ascend/cann-8.5.0/python/site-packages/tbe/dsl/unify_schedule/vector/transdata/common/graph/transdata_graph_info.py:170 /usr/local/Ascend/cann-8.5.0/python/site-packages/tbe/dsl/unify_schedule/vector/transdata/common/graph/transdata_graph_info.py:170: DeprecationWarning: invalid escape sequence \c """ ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:549 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:89 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:549 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:89 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/batchnorm_fold2.py:57 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/batchnorm_fold2.py:57: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("batchnorm_fold2") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/batchnorm_fold2_grad.py:56 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/batchnorm_fold2_grad.py:56: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("batchnorm_fold2_grad") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/batchnorm_fold2_grad_reduce.py:48 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/batchnorm_fold2_grad_reduce.py:48: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("batchnorm_fold2_grad_reduce") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/correction_mul.py:51 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/correction_mul.py:51: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("correction_mul") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/correction_mul_grad.py:51 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/correction_mul_grad.py:51: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("correction_mul_grad") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/correction_mul_grad.py:143 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/correction_mul_grad.py:143: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("correction_mul_grad_reduce") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perlayer.py:50 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perlayer.py:50: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_learned_scale_quant_perlayer") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perlayer_grad.py:92 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perlayer_grad.py:92: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_learned_scale_quant_perlayer_grad_d") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perlayer_grad_reduce.py:49 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perlayer_grad_reduce.py:49: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_learned_scale_quant_perlayer_grad_d_reduce") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perchannel.py:50 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perchannel.py:50: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_learned_scale_quant_perchannel") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perchannel_grad.py:91 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perchannel_grad.py:91: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_learned_scale_quant_perchannel_grad_d") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perchannel_grad_reduce.py:48 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_learned_scale_quant_perchannel_grad_reduce.py:48: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_learned_scale_quant_perchannel_grad_d_reduce") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perchannel.py:52 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perchannel.py:52: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_quant_perchannel") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perchannel_grad.py:81 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perchannel_grad.py:81: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_quant_perchannel_grad") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perlayer.py:54 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perlayer.py:54: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_quant_per_layer") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perlayer_grad.py:81 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/fake_quant_perlayer_grad.py:81: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("fake_quant_per_layer_grad") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/minmax_update_perchannel.py:50 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/minmax_update_perchannel.py:50: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("minmax_update_perchannel") ../../../../../../../../anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/minmax_update_perlayer.py:50 /home/jenkins/anaconda3/envs/ci39/lib/python3.9/site-packages/mindspore/ops/_op_impl/_custom_op/minmax_update_perlayer.py:50: DeprecationWarning: te_fusion.fusion_manager.fusion_manager.register is deprecated,please replace it with tbe.common.register.register_op_compute @fusion_manager.register("minmax_update_perlayer") -- Docs: https://docs.pytest.org/en/stable/warnings.html ================= 16 passed, 25 warnings in 511.49s (0:08:31) ==================