==================================================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_006/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(163223:281473160007472,MainProcess):2026-01-29-17:37:39.117.731 [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 = 2.26794, [21] [bootstrap]: 0.0006126 [type_inference]: 2.05921 [event_method]: 1.941e-05 [auto_monad]: 0.0716568 [graph_reusing]: 7.08e-06 [inline]: 2.78e-06 [add_attr]: 0.00847951, [1] [add_attr_with_inline]: 0.0084638, [1] [Cycle 1]: 9.582e-05, [2] [tag_attr]: 2.078e-05 [meta_addattr_fg_expand]: 4.68001e-06 [parallel-infer-symbol]: 3.65998e-06 [pre_auto_parallel]: 3.771e-05 [insert-virtual-dataset]: 2.56e-06 [parallel-infer-symbol-second]: 8.89995e-07 [dataset_repeat_opt]: 2.22999e-06 [pipeline_split]: 1.96e-06 [optimize]: 0.126965, [53] [py_interpret_to_execute]: 3.13e-05 [rewriter_before_opt_a]: 9.215e-05 [opt_a]: 0.123209, [2] [Cycle 1]: 0.121619, [45] [expand_dump_flag]: 2.88e-06 [switch_simplify]: 3.625e-05 [loop_unroll]: 1.975e-05 [a_1]: 0.00066982 [with_stream_mark]: 2.357e-05 [recompute_prepare]: 1.718e-05 [updatestate_depend_eliminate]: 9.04998e-06 [updatestate_assign_eliminate]: 8.02998e-06 [updatestate_loads_eliminate]: 1.05e-05 [parameter_eliminate]: 2.05002e-06 [a_2]: 0.00022504 [accelerated_algorithm]: 3.524e-05 [shard]: 2.34999e-06 [meta_shard_fg_expand]: 3.33e-06 [shard_inline]: 1.51e-05 [merge_send_recv]: 1.376e-05 [auto_parallel]: 1.051e-05 [parallel]: 5.274e-05 [flash_sp]: 2.064e-05 [merge_comm]: 9.05001e-06 [allreduce_fusion]: 8.02e-06 [matmul_add_comm_reduction]: 1.523e-05 [allreduce_slice_to_reducescatter]: 6.30011e-07 [virtual_shard_identity]: 1.823e-05 [virtual_dataset]: 1.573e-05 [get_grad_eliminate_]: 1.507e-05 [virtual_output]: 1.547e-05 [merge_forward]: 8.55999e-06 [cell_reuse_recompute_pass]: 1.80001e-06 [offload_activation]: 1.785e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.758e-05 [merge_recompute_call_nodes]: 1.47999e-06 [before_grad]: 4.774e-05 [set_forward_comm_id_for_comm_node_pass]: 9.55001e-06 [meta_fg_expand]: 5.82001e-06 [flash_sp_send_recv_attached]: 5.37001e-06 [receive_attached]: 2.58e-06 [after_resolve]: 2.278e-05 [a_after_grad]: 2.593e-05 [renormalize]: 0.119324 [add_forward_monad_depend]: 1.956e-05 [auto_monad_grad]: 2.33002e-06 [auto_monad_eliminator]: 6.03e-05 [cse]: 0.0003016 [a_3]: 0.00013418 [Cycle 2]: 0.00157553, [45] [expand_dump_flag]: 2.66e-06 [switch_simplify]: 1.825e-05 [loop_unroll]: 1.545e-05 [a_1]: 0.00044017 [with_stream_mark]: 2.574e-05 [recompute_prepare]: 1.675e-05 [updatestate_depend_eliminate]: 9.23002e-06 [updatestate_assign_eliminate]: 7.83999e-06 [updatestate_loads_eliminate]: 1.062e-05 [parameter_eliminate]: 2.06998e-06 [a_2]: 0.00021303 [accelerated_algorithm]: 2.129e-05 [shard]: 2.13002e-06 [meta_shard_fg_expand]: 3.63e-06 [shard_inline]: 1.472e-05 [merge_send_recv]: 1.377e-05 [auto_parallel]: 1.43e-05 [parallel]: 9.17001e-06 [flash_sp]: 3.81001e-06 [merge_comm]: 8.28999e-06 [allreduce_fusion]: 9.55001e-06 [matmul_add_comm_reduction]: 1.745e-05 [allreduce_slice_to_reducescatter]: 9.09989e-07 [virtual_shard_identity]: 1.778e-05 [virtual_dataset]: 1.436e-05 [get_grad_eliminate_]: 1.452e-05 [virtual_output]: 1.488e-05 [merge_forward]: 8.76002e-06 [cell_reuse_recompute_pass]: 2.83003e-06 [offload_activation]: 1.727e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.758e-05 [merge_recompute_call_nodes]: 1.52999e-06 [before_grad]: 2.485e-05 [set_forward_comm_id_for_comm_node_pass]: 8.28999e-06 [meta_fg_expand]: 6.24999e-06 [flash_sp_send_recv_attached]: 2.06e-06 [receive_attached]: 2.64001e-06 [after_resolve]: 2.222e-05 [a_after_grad]: 2.415e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.99e-06 [auto_monad_grad]: 2.18998e-06 [auto_monad_eliminator]: 3.993e-05 [cse]: 0.0001121 [a_3]: 9.918e-05 [py_interpret_to_execute_after_opt_a]: 2.745e-05 [slice_cell_reuse_recomputed_activation]: 2.24001e-06 [rewriter_after_opt_a]: 0.00026718 [convert_after_rewriter]: 1.677e-05 [order_py_execute_after_rewriter]: 9.81e-06 [mutable_eliminate]: 0.00086899 [opt_b]: 0.00058773, [1] [Cycle 1]: 0.00057844, [7] [b_1]: 0.00039597 [b_2]: 2.122e-05 [updatestate_depend_eliminate]: 1.601e-05 [updatestate_assign_eliminate]: 9.91998e-06 [updatestate_loads_eliminate]: 1.114e-05 [renormalize]: 6.69999e-07 [cse]: 7.817e-05 [optimize_parallel_all_gather_comm]: 3.24e-05 [overlap_param_gather]: 5.76e-06 [cconv]: 3.654e-05 [loop_unroll]: 0.00053625 [opt_after_cconv]: 0.00023035, [1] [Cycle 1]: 0.00022349, [7] [c_1]: 0.00010284 [parameter_eliminate]: 4.35e-06 [updatestate_depend_eliminate]: 1.191e-05 [updatestate_assign_eliminate]: 7.43e-06 [updatestate_loads_eliminate]: 9.17999e-06 [cse]: 4.954e-05 [renormalize]: 3.89991e-07 [remove_dup_value]: 6.951e-05 [tuple_transform]: 0.00016537, [1] [Cycle 1]: 0.00015982, [4] [d_1]: 0.00011538 [none_parameter_eliminate]: 1.88002e-06 [renormalize]: 4.30009e-07 [switch_simplify]: 2.016e-05 [partial_unused_args_eliminate]: 1.92001e-06 [add_recomputation]: 0.00010686 [cse_after_recomputation]: 5.267e-05, [1] [Cycle 1]: 4.741e-05, [1] [cse]: 4.037e-05 [environ_conv]: 3.163e-05 [swap_dp_allreduce_reducescatter]: 1.258e-05 [bias_add_comm_swap]: 3.28e-06 [label_micro_interleaved_index]: 4.85999e-06 [label_fine_grained_interleaved_index]: 2.74999e-06 [merge_cast_opt]: 1.59e-06 [slice_recompute_activation]: 2.07001e-06 [micro_interleaved_order_control]: 2.79999e-06 [assign_add_opt]: 1.28002e-06 [ForceFp32Comm]: 7.89994e-07 [remove_cast_before_assign_add]: 1.12e-06 [full_micro_interleaved_order_control]: 2.76e-06 [reorder_send_recv_between_fp_bp]: 2.94999e-06 [comm_op_add_attrs]: 1.10999e-06 [add_comm_op_reuse_tag]: 1.09e-06 [interleave_split_concat_branches]: 1.39e-06 [interleave_parallel_branches]: 1.10999e-06 [overlap_opt_shard_in_pipeline]: 2.257e-05 [overlap_opt_shard_grad_in_pipeline]: 2.16998e-06 [control_data_broadcast_order]: 2.608e-05 [grouped_pairwise_exchange_alltoall]: 1.92999e-06 [offloading_packed_experts]: 7.11001e-06 [overlap_recompute_and_grad_model_parallel]: 7.61001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.45999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.38002e-06 [overlap_recompute_comm]: 2.52001e-06 [overlap_grad_ring_attention]: 6.67002e-06 [overlap_grad_flash_sp]: 5.327e-05 [begin_end_overlap_inline]: 5.69999e-07 [split_matmul_comm_elemetwise]: 2.31e-06 [split_layernorm_comm]: 1.74e-06 [handle_group_info]: 1.02e-06 [symbol_engine_optimizer]: 0.00014477, [1] [Cycle 1]: 0.00013902, [6] [build]: 1.558e-05 [elim_shapecalc]: 2.04e-05 [elim_not_effective]: 2.792e-05 [opt_reshape]: 2.109e-05 [fold_const_symbol]: 2.385e-05 [renormalize]: 1.90019e-07 [detach_backward]: 2.49001e-06 [pipeline_parallel_scheduler]: 1.86e-06 [auto_monad_reorder]: 4.86e-05 [get_jit_bprop_graph]: 2.07999e-06 [rewriter_after_jit_bprop_graph]: 5.72999e-06 [opt_after_jit_grad]: 0.00054539 [validate]: 9.831e-05 Sums bootstrap : 0.000613s : 0.03% type_inference : 2.059213s : 91.19% event_method : 0.000019s : 0.00% auto_monad : 0.071657s : 3.17% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000021s : 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 : 0.000038s : 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 : 0.000031s : 0.00% optimize.rewriter_before_opt_a : 0.000092s : 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 : 0.000035s : 0.00% optimize.opt_a.a_1 : 0.001110s : 0.05% optimize.opt_a.with_stream_mark : 0.000049s : 0.00% optimize.opt_a.recompute_prepare : 0.000034s : 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 : 0.000438s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000057s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000030s : 0.00% optimize.opt_a.merge_send_recv : 0.000028s : 0.00% optimize.opt_a.auto_parallel : 0.000025s : 0.00% optimize.opt_a.parallel : 0.000062s : 0.00% optimize.opt_a.flash_sp : 0.000024s : 0.00% optimize.opt_a.merge_comm : 0.000017s : 0.00% optimize.opt_a.allreduce_fusion : 0.000018s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000033s : 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 : 0.000030s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000030s : 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 : 0.000035s : 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 : 0.000073s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000018s : 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 : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000045s : 0.00% optimize.opt_a.a_after_grad : 0.000050s : 0.00% optimize.opt_a.renormalize : 0.119325s : 5.28% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000100s : 0.00% optimize.opt_a.cse : 0.000414s : 0.02% optimize.opt_a.a_3 : 0.000233s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000027s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000267s : 0.01% optimize.convert_after_rewriter : 0.000017s : 0.00% optimize.order_py_execute_after_rewriter : 0.000010s : 0.00% optimize.mutable_eliminate : 0.000869s : 0.04% optimize.opt_b.b_1 : 0.000396s : 0.02% optimize.opt_b.b_2 : 0.000021s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000016s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000078s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000032s : 0.00% optimize.overlap_param_gather : 0.000006s : 0.00% optimize.cconv : 0.000037s : 0.00% optimize.loop_unroll : 0.000536s : 0.02% optimize.opt_after_cconv.c_1 : 0.000103s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 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 : 0.000009s : 0.00% optimize.opt_after_cconv.cse : 0.000050s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000070s : 0.00% optimize.tuple_transform.d_1 : 0.000115s : 0.01% 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 : 0.000002s : 0.00% optimize.add_recomputation : 0.000107s : 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 : 0.000002s : 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 : 0.000023s : 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 : 0.000007s : 0.00% optimize.overlap_grad_flash_sp : 0.000053s : 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 : 0.000016s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000020s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000028s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000021s : 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 : 0.000049s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000545s : 0.02% validate : 0.000098s : 0.00% Time group info: ------[substitution.] 0.000292 96 3.40% : 0.000010s : 2: substitution.depend_value_elim 1.44% : 0.000004s : 8: substitution.elim_not_effective 1.16% : 0.000003s : 8: substitution.fold_const_symbol 4.28% : 0.000013s : 13: substitution.graph_param_transform 37.72% : 0.000110s : 1: substitution.inline 3.42% : 0.000010s : 16: substitution.j_node_and_user_rematch 5.97% : 0.000017s : 2: substitution.less_batch_normalization 2.00% : 0.000006s : 6: substitution.load_eliminater 0.86% : 0.000003s : 2: substitution.opt_reshape 7.41% : 0.000022s : 16: substitution.remove_not_recompute_node 3.05% : 0.000009s : 4: substitution.replace_old_param 16.19% : 0.000047s : 4: substitution.reshape_eliminate 6.90% : 0.000020s : 6: substitution.updatestate_pure_node_eliminater 6.18% : 0.000018s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.059102 2 99.80% : 2.054987s : 1: type_inference.infer 0.20% : 0.004114s : 1: type_inference.specialize ------[replace.] 0.000018 1 100.00% : 0.000018s : 1: replace.inline ------[match.] 0.000109 1 100.00% : 0.000109s : 1: match.inline ------[predicate.] 0.000427 3002 0.80% : 0.000003s : 27: predicate.accumulaten_eliminater 0.71% : 0.000003s : 13: predicate.ad_related_special_op_eliminate 0.76% : 0.000003s : 26: predicate.addn_check_dump 0.82% : 0.000004s : 27: predicate.addn_zero_filter 0.70% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.07% : 0.000009s : 53: predicate.arithmetic_simplify 0.83% : 0.000004s : 27: predicate.cast_eliminate 0.79% : 0.000003s : 26: predicate.check_bprop_eliminate 0.75% : 0.000003s : 26: predicate.compare_switch_simplify 0.27% : 0.000001s : 13: predicate.const_output_eliminate 0.82% : 0.000004s : 26: predicate.depend_value_elim 0.80% : 0.000003s : 27: predicate.dict_get_item_const_eliminator 0.88% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.77% : 0.000003s : 27: predicate.dict_set_item_eliminator 2.00% : 0.000009s : 26: predicate.dumpgradient_eliminate 0.33% : 0.000001s : 13: predicate.elim_not_effective 0.44% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.21% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.23% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.14% : 0.000005s : 40: predicate.environ_get_depend_swap 2.01% : 0.000009s : 66: predicate.environ_get_eliminate 1.15% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.83% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.29% : 0.000005s : 28: predicate.float_depend_g_call 0.77% : 0.000003s : 26: predicate.float_environ_get_switch 1.15% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.24% : 0.000001s : 13: predicate.fold_const_symbol 0.84% : 0.000004s : 26: predicate.get_grad_eliminate 0.27% : 0.000001s : 13: predicate.graph_param_transform 0.79% : 0.000003s : 26: predicate.incorporate_call 0.71% : 0.000003s : 26: predicate.incorporate_call_switch 5.55% : 0.000024s : 133: predicate.inline 1.03% : 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 1.80% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.48% : 0.000011s : 80: predicate.load_eliminater 0.92% : 0.000004s : 13: predicate.loop_unroll_after_grad 1.05% : 0.000005s : 31: predicate.loop_unroll_before_grad 1.84% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.93% : 0.000004s : 26: predicate.merge_addn 0.76% : 0.000003s : 26: predicate.micro_step_allgather_replace 0.82% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.74% : 0.000003s : 27: predicate.minmaximum_grad 1.26% : 0.000005s : 13: predicate.mutable_eliminate 0.51% : 0.000002s : 13: predicate.opt_reshape 0.49% : 0.000002s : 13: predicate.parallel_virtual_node 1.08% : 0.000005s : 28: predicate.partial_defer_inline 1.29% : 0.000006s : 40: predicate.partial_eliminate 0.75% : 0.000003s : 27: predicate.print_const_string_wrapper 0.78% : 0.000003s : 26: predicate.reduce_all_const_elim 1.07% : 0.000005s : 27: predicate.reduce_eliminate 2.35% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.52% : 0.000002s : 26: predicate.remove_not_recompute_node 1.40% : 0.000006s : 53: predicate.replace_applicator 0.56% : 0.000002s : 26: predicate.replace_old_param 0.46% : 0.000002s : 13: predicate.reset_defer_inline 0.99% : 0.000004s : 27: predicate.reshape_eliminate 0.82% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.51% : 0.000002s : 13: predicate.row_tensor_eliminate 1.20% : 0.000005s : 26: predicate.same_eliminate 0.66% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.92% : 0.000004s : 26: predicate.shard_identity_eliminate 0.90% : 0.000004s : 26: predicate.special_op_eliminate 0.89% : 0.000004s : 26: predicate.specialize_transform 1.14% : 0.000005s : 26: predicate.split_environ_get_set_with_tuple_value 0.94% : 0.000004s : 26: predicate.stack_unstack_eliminate 0.50% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.86% : 0.000004s : 28: predicate.switch_defer_inline 1.75% : 0.000007s : 54: predicate.switch_layer_defer_inline 3.41% : 0.000015s : 98: predicate.switch_simplify 0.83% : 0.000004s : 27: predicate.tile_eliminate 0.83% : 0.000004s : 27: predicate.transpose_eliminate 1.70% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.67% : 0.000007s : 53: predicate.tuple_list_get_item_const_eliminator 1.56% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.97% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.63% : 0.000007s : 53: predicate.tuple_list_get_set_item_eliminator 2.58% : 0.000011s : 79: predicate.tuple_list_set_item_eliminator 1.64% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.34% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.35% : 0.000014s : 106: predicate.updatestate_useless_node_eliminater 0.50% : 0.000002s : 13: predicate.value_based_eliminate 0.86% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.91% : 0.000004s : 26: predicate.virtual_output_eliminate 0.40% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.53% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.132638 58 14.51% : 0.019245s : 55: func_graph_cloner_run.FuncGraphClonerGraph 85.49% : 0.113393s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.525550 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.34% : 0.008486s : 1: add_attr 0.34% : 0.008468s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000112s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 2.84% : 0.071691s : 1: auto_monad 0.00% : 0.000054s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.03% : 0.000644s : 1: bootstrap 0.00% : 0.000041s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000030s : 1: control_data_broadcast_order 0.00% : 0.000021s : 1: convert_after_rewriter 0.00% : 0.000056s : 1: cse_after_recomputation 0.00% : 0.000020s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000036s : 1: environ_conv 0.00% : 0.000027s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000014s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 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% : 0.000005s : 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% : 0.000546s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.03% : 0.000883s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000026s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000036s : 1: opt.transform.mutable_eliminate 0.09% : 0.002240s : 78: opt.transform.opt_a 0.00% : 0.000101s : 1: opt.transform.opt_after_cconv 0.00% : 0.000053s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000384s : 28: opt.transform.opt_b 0.01% : 0.000133s : 2: opt.transform.opt_trans_graph 0.00% : 0.000089s : 4: opt.transform.symbol_engine_opt 4.88% : 0.123212s : 1: opt_a 0.01% : 0.000234s : 1: opt_after_cconv 0.02% : 0.000556s : 1: opt_after_jit_grad 0.02% : 0.000592s : 1: opt_b 5.03% : 0.126970s : 1: optimize 0.00% : 0.000037s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000013s : 1: order_py_execute_after_rewriter 0.00% : 0.000057s : 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% : 0.000026s : 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% : 0.000011s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000007s : 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% : 0.000042s : 1: pre_auto_parallel 0.00% : 0.000035s : 1: py_interpret_to_execute 0.00% : 0.000031s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000075s : 1: remove_dup_value 0.22% : 0.005566s : 1: renormalize.infer 4.50% : 0.113741s : 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% : 0.000274s : 1: rewriter_after_opt_a 0.00% : 0.000097s : 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% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000016s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000148s : 1: symbol_engine_optimizer 0.01% : 0.000169s : 1: tuple_transform 81.54% : 2.059245s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape0-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape0-256],max_mem:124.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:38:27.237.495 [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 = 4.28466, [21] [bootstrap]: 0.00064901 [type_inference]: 3.87076 [event_method]: 1.519e-05 [auto_monad]: 0.0003714 [graph_reusing]: 6.32001e-06 [inline]: 3.06001e-06 [add_attr]: 0.188902, [1] [add_attr_with_inline]: 0.188886, [1] [Cycle 1]: 7.298e-05, [2] [tag_attr]: 2.406e-05 [meta_addattr_fg_expand]: 4.63999e-06 [parallel-infer-symbol]: 6.29999e-06 [pre_auto_parallel]: 4.053e-05 [insert-virtual-dataset]: 2.54001e-06 [parallel-infer-symbol-second]: 6.79982e-07 [dataset_repeat_opt]: 1.96e-06 [pipeline_split]: 1.72999e-06 [optimize]: 0.22265, [53] [py_interpret_to_execute]: 3.386e-05 [rewriter_before_opt_a]: 9.038e-05 [opt_a]: 0.217717, [2] [Cycle 1]: 0.215594, [45] [expand_dump_flag]: 3.16001e-06 [switch_simplify]: 3.43e-05 [loop_unroll]: 2.032e-05 [a_1]: 0.00096211 [with_stream_mark]: 2.854e-05 [recompute_prepare]: 1.962e-05 [updatestate_depend_eliminate]: 1.162e-05 [updatestate_assign_eliminate]: 8.52e-06 [updatestate_loads_eliminate]: 1.039e-05 [parameter_eliminate]: 2.03997e-06 [a_2]: 0.0002279 [accelerated_algorithm]: 4.06e-05 [shard]: 2.34999e-06 [meta_shard_fg_expand]: 5.44e-06 [shard_inline]: 1.708e-05 [merge_send_recv]: 1.378e-05 [auto_parallel]: 1.128e-05 [parallel]: 3.142e-05 [flash_sp]: 1.393e-05 [merge_comm]: 8.50001e-06 [allreduce_fusion]: 7.73001e-06 [matmul_add_comm_reduction]: 1.808e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 1.81e-05 [virtual_dataset]: 6.004e-05 [get_grad_eliminate_]: 1.966e-05 [virtual_output]: 1.567e-05 [merge_forward]: 9.39998e-06 [cell_reuse_recompute_pass]: 2.38002e-06 [offload_activation]: 0.0001333 [cell_reuse_handle_not_recompute_node_pass]: 3.406e-05 [merge_recompute_call_nodes]: 1.67999e-06 [before_grad]: 2.802e-05 [set_forward_comm_id_for_comm_node_pass]: 9.72999e-06 [meta_fg_expand]: 7.09001e-06 [flash_sp_send_recv_attached]: 7.23e-06 [receive_attached]: 2.79999e-06 [after_resolve]: 2.39e-05 [a_after_grad]: 2.741e-05 [renormalize]: 0.212835 [add_forward_monad_depend]: 2.107e-05 [auto_monad_grad]: 3.23e-06 [auto_monad_eliminator]: 5.797e-05 [cse]: 0.00024179 [a_3]: 0.000144 [Cycle 2]: 0.00210666, [45] [expand_dump_flag]: 3.65e-06 [switch_simplify]: 2.081e-05 [loop_unroll]: 1.792e-05 [a_1]: 0.00046969 [with_stream_mark]: 2.431e-05 [recompute_prepare]: 1.885e-05 [updatestate_depend_eliminate]: 1.028e-05 [updatestate_assign_eliminate]: 9.05999e-06 [updatestate_loads_eliminate]: 1.075e-05 [parameter_eliminate]: 2.59001e-06 [a_2]: 0.00027718 [accelerated_algorithm]: 2.573e-05 [shard]: 2.61999e-06 [meta_shard_fg_expand]: 5.78002e-06 [shard_inline]: 1.661e-05 [merge_send_recv]: 1.606e-05 [auto_parallel]: 1.747e-05 [parallel]: 1.04e-05 [flash_sp]: 5.81998e-06 [merge_comm]: 8.79e-06 [allreduce_fusion]: 3.836e-05 [matmul_add_comm_reduction]: 1.823e-05 [allreduce_slice_to_reducescatter]: 7.30011e-07 [virtual_shard_identity]: 2.205e-05 [virtual_dataset]: 1.624e-05 [get_grad_eliminate_]: 1.62e-05 [virtual_output]: 1.524e-05 [merge_forward]: 1.028e-05 [cell_reuse_recompute_pass]: 3.5e-06 [offload_activation]: 1.743e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.123e-05 [merge_recompute_call_nodes]: 1.80001e-06 [before_grad]: 2.613e-05 [set_forward_comm_id_for_comm_node_pass]: 8.97999e-06 [meta_fg_expand]: 7.77e-06 [flash_sp_send_recv_attached]: 2.10002e-06 [receive_attached]: 2.75002e-06 [after_resolve]: 2.438e-05 [a_after_grad]: 2.839e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 3.26001e-06 [auto_monad_grad]: 2.76e-06 [auto_monad_eliminator]: 0.00014389 [cse]: 0.00017777 [a_3]: 0.00020972 [py_interpret_to_execute_after_opt_a]: 2.852e-05 [slice_cell_reuse_recomputed_activation]: 2.41e-06 [rewriter_after_opt_a]: 0.00020635 [convert_after_rewriter]: 1.667e-05 [order_py_execute_after_rewriter]: 1.208e-05 [mutable_eliminate]: 0.00111076 [opt_b]: 0.00107216, [1] [Cycle 1]: 0.0010603, [7] [b_1]: 0.00084531 [b_2]: 3.189e-05 [updatestate_depend_eliminate]: 1.399e-05 [updatestate_assign_eliminate]: 8.06001e-06 [updatestate_loads_eliminate]: 1.059e-05 [renormalize]: 8.09989e-07 [cse]: 8.273e-05 [optimize_parallel_all_gather_comm]: 3.341e-05 [overlap_param_gather]: 2.29001e-06 [cconv]: 3.165e-05 [loop_unroll]: 0.00070979 [opt_after_cconv]: 0.00025285, [1] [Cycle 1]: 0.00024185, [7] [c_1]: 0.00010624 [parameter_eliminate]: 5.25999e-06 [updatestate_depend_eliminate]: 1.493e-05 [updatestate_assign_eliminate]: 6.88e-06 [updatestate_loads_eliminate]: 1.015e-05 [cse]: 5.986e-05 [renormalize]: 4.39992e-07 [remove_dup_value]: 7.512e-05 [tuple_transform]: 0.00016978, [1] [Cycle 1]: 0.00016323, [4] [d_1]: 0.00011748 [none_parameter_eliminate]: 1.94e-06 [renormalize]: 3.59985e-07 [switch_simplify]: 1.718e-05 [partial_unused_args_eliminate]: 2.29999e-06 [add_recomputation]: 0.00013275 [cse_after_recomputation]: 6.811e-05, [1] [Cycle 1]: 6.199e-05, [1] [cse]: 5.297e-05 [environ_conv]: 2.547e-05 [swap_dp_allreduce_reducescatter]: 1.237e-05 [bias_add_comm_swap]: 4.11001e-06 [label_micro_interleaved_index]: 6.59999e-06 [label_fine_grained_interleaved_index]: 2.81e-06 [merge_cast_opt]: 1.62999e-06 [slice_recompute_activation]: 2.46e-06 [micro_interleaved_order_control]: 3.06001e-06 [assign_add_opt]: 1.34e-06 [ForceFp32Comm]: 9.09989e-07 [remove_cast_before_assign_add]: 1.37e-06 [full_micro_interleaved_order_control]: 2.27999e-06 [reorder_send_recv_between_fp_bp]: 2.86e-06 [comm_op_add_attrs]: 1.10999e-06 [add_comm_op_reuse_tag]: 1.03001e-06 [interleave_split_concat_branches]: 1.25999e-06 [interleave_parallel_branches]: 1.45001e-06 [overlap_opt_shard_in_pipeline]: 1.28002e-06 [overlap_opt_shard_grad_in_pipeline]: 2.32001e-06 [control_data_broadcast_order]: 8.26e-05 [grouped_pairwise_exchange_alltoall]: 1.66998e-06 [offloading_packed_experts]: 8.06001e-06 [overlap_recompute_and_grad_model_parallel]: 7.82e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.47001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37999e-06 [overlap_recompute_comm]: 2.67001e-06 [overlap_grad_ring_attention]: 7.12997e-06 [overlap_grad_flash_sp]: 4.659e-05 [begin_end_overlap_inline]: 5.99975e-07 [split_matmul_comm_elemetwise]: 2.25002e-06 [split_layernorm_comm]: 2.07001e-06 [handle_group_info]: 1.22e-06 [symbol_engine_optimizer]: 0.0001988, [1] [Cycle 1]: 0.0001886, [6] [build]: 2.378e-05 [elim_shapecalc]: 3.53e-05 [elim_not_effective]: 3.353e-05 [opt_reshape]: 2.529e-05 [fold_const_symbol]: 2.5e-05 [renormalize]: 2.89991e-07 [detach_backward]: 2.42001e-06 [pipeline_parallel_scheduler]: 1.45999e-06 [auto_monad_reorder]: 5.461e-05 [get_jit_bprop_graph]: 1.97001e-06 [rewriter_after_jit_bprop_graph]: 4.47e-06 [opt_after_jit_grad]: 0.00086196 [validate]: 9.736e-05 Sums bootstrap : 0.000649s : 0.02% type_inference : 3.870763s : 94.54% event_method : 0.000015s : 0.00% auto_monad : 0.000371s : 0.01% graph_reusing : 0.000006s : 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 : 0.000005s : 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 : 0.000034s : 0.00% optimize.rewriter_before_opt_a : 0.000090s : 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 : 0.001432s : 0.03% optimize.opt_a.with_stream_mark : 0.000053s : 0.00% optimize.opt_a.recompute_prepare : 0.000038s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000022s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000018s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000021s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000505s : 0.01% optimize.opt_a.accelerated_algorithm : 0.000066s : 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 : 0.000034s : 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 : 0.000042s : 0.00% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000017s : 0.00% optimize.opt_a.allreduce_fusion : 0.000046s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000036s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000040s : 0.00% optimize.opt_a.virtual_dataset : 0.000076s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000036s : 0.00% optimize.opt_a.virtual_output : 0.000031s : 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 : 0.000151s : 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 : 0.000054s : 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 : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000048s : 0.00% optimize.opt_a.a_after_grad : 0.000056s : 0.00% optimize.opt_a.renormalize : 0.212835s : 5.20% optimize.opt_a.add_forward_monad_depend : 0.000024s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000202s : 0.00% optimize.opt_a.cse : 0.000420s : 0.01% optimize.opt_a.a_3 : 0.000354s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000029s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000206s : 0.01% optimize.convert_after_rewriter : 0.000017s : 0.00% optimize.order_py_execute_after_rewriter : 0.000012s : 0.00% optimize.mutable_eliminate : 0.001111s : 0.03% optimize.opt_b.b_1 : 0.000845s : 0.02% optimize.opt_b.b_2 : 0.000032s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000014s : 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 : 0.000083s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000033s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000032s : 0.00% optimize.loop_unroll : 0.000710s : 0.02% optimize.opt_after_cconv.c_1 : 0.000106s : 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 : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.cse : 0.000060s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000075s : 0.00% optimize.tuple_transform.d_1 : 0.000117s : 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 : 0.000133s : 0.00% optimize.cse_after_recomputation.cse : 0.000053s : 0.00% optimize.environ_conv : 0.000025s : 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 : 0.000007s : 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 : 0.000083s : 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 : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000007s : 0.00% optimize.overlap_grad_flash_sp : 0.000047s : 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 : 0.000024s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000035s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000034s : 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 : 0.000055s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000862s : 0.02% validate : 0.000097s : 0.00% Time group info: ------[substitution.] 0.000304 96 3.62% : 0.000011s : 2: substitution.depend_value_elim 1.32% : 0.000004s : 8: substitution.elim_not_effective 1.12% : 0.000003s : 8: substitution.fold_const_symbol 4.35% : 0.000013s : 13: substitution.graph_param_transform 45.55% : 0.000138s : 1: substitution.inline 3.17% : 0.000010s : 16: substitution.j_node_and_user_rematch 7.68% : 0.000023s : 2: substitution.less_batch_normalization 1.92% : 0.000006s : 6: substitution.load_eliminater 1.24% : 0.000004s : 2: substitution.opt_reshape 4.32% : 0.000013s : 16: substitution.remove_not_recompute_node 2.74% : 0.000008s : 4: substitution.replace_old_param 10.63% : 0.000032s : 4: substitution.reshape_eliminate 6.33% : 0.000019s : 6: substitution.updatestate_pure_node_eliminater 6.00% : 0.000018s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 3.870657 2 99.75% : 3.861036s : 1: type_inference.infer 0.25% : 0.009621s : 1: type_inference.specialize ------[replace.] 0.000021 1 100.00% : 0.000021s : 1: replace.inline ------[match.] 0.000138 1 100.00% : 0.000138s : 1: match.inline ------[predicate.] 0.000457 3002 0.81% : 0.000004s : 27: predicate.accumulaten_eliminater 1.14% : 0.000005s : 13: predicate.ad_related_special_op_eliminate 0.70% : 0.000003s : 26: predicate.addn_check_dump 0.87% : 0.000004s : 27: predicate.addn_zero_filter 0.69% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.22% : 0.000010s : 53: predicate.arithmetic_simplify 0.78% : 0.000004s : 27: predicate.cast_eliminate 0.99% : 0.000005s : 26: predicate.check_bprop_eliminate 0.71% : 0.000003s : 26: predicate.compare_switch_simplify 0.26% : 0.000001s : 13: predicate.const_output_eliminate 1.04% : 0.000005s : 26: predicate.depend_value_elim 0.78% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.88% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.82% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.17% : 0.000005s : 26: predicate.dumpgradient_eliminate 0.32% : 0.000001s : 13: predicate.elim_not_effective 0.80% : 0.000004s : 13: predicate.elim_shapecalc_of_broadcastargs 1.12% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.04% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.09% : 0.000005s : 40: predicate.environ_get_depend_swap 1.91% : 0.000009s : 66: predicate.environ_get_eliminate 1.22% : 0.000006s : 40: predicate.environ_get_set_eliminate 0.78% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.21% : 0.000006s : 28: predicate.float_depend_g_call 0.85% : 0.000004s : 26: predicate.float_environ_get_switch 1.22% : 0.000006s : 39: predicate.float_tuple_getitem_switch 0.22% : 0.000001s : 13: predicate.fold_const_symbol 0.89% : 0.000004s : 26: predicate.get_grad_eliminate 0.30% : 0.000001s : 13: predicate.graph_param_transform 0.79% : 0.000004s : 26: predicate.incorporate_call 0.72% : 0.000003s : 26: predicate.incorporate_call_switch 5.48% : 0.000025s : 133: predicate.inline 1.05% : 0.000005s : 26: predicate.inline_without_move 0.42% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.51% : 0.000007s : 27: predicate.less_batch_normalization 1.70% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.28% : 0.000010s : 80: predicate.load_eliminater 1.23% : 0.000006s : 13: predicate.loop_unroll_after_grad 1.01% : 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 0.78% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.75% : 0.000003s : 26: predicate.mini_step_allgather_replace 0.71% : 0.000003s : 27: predicate.minmaximum_grad 1.35% : 0.000006s : 13: predicate.mutable_eliminate 0.58% : 0.000003s : 13: predicate.opt_reshape 0.49% : 0.000002s : 13: predicate.parallel_virtual_node 1.21% : 0.000006s : 28: predicate.partial_defer_inline 1.27% : 0.000006s : 40: predicate.partial_eliminate 0.80% : 0.000004s : 27: predicate.print_const_string_wrapper 0.80% : 0.000004s : 26: predicate.reduce_all_const_elim 1.09% : 0.000005s : 27: predicate.reduce_eliminate 2.24% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.54% : 0.000002s : 26: predicate.remove_not_recompute_node 1.48% : 0.000007s : 53: predicate.replace_applicator 0.53% : 0.000002s : 26: predicate.replace_old_param 0.32% : 0.000001s : 13: predicate.reset_defer_inline 0.93% : 0.000004s : 27: predicate.reshape_eliminate 0.78% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.99% : 0.000005s : 13: predicate.row_tensor_eliminate 1.30% : 0.000006s : 26: predicate.same_eliminate 0.51% : 0.000002s : 26: predicate.set_cell_output_no_recompute 0.99% : 0.000005s : 26: predicate.shard_identity_eliminate 1.30% : 0.000006s : 26: predicate.special_op_eliminate 0.84% : 0.000004s : 26: predicate.specialize_transform 1.25% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.93% : 0.000004s : 26: predicate.stack_unstack_eliminate 0.49% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.84% : 0.000004s : 28: predicate.switch_defer_inline 1.66% : 0.000008s : 54: predicate.switch_layer_defer_inline 3.28% : 0.000015s : 98: predicate.switch_simplify 0.84% : 0.000004s : 27: predicate.tile_eliminate 0.74% : 0.000003s : 27: predicate.transpose_eliminate 1.68% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 1.61% : 0.000007s : 53: predicate.tuple_list_get_item_const_eliminator 1.58% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.76% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.58% : 0.000007s : 53: predicate.tuple_list_get_set_item_eliminator 2.42% : 0.000011s : 79: predicate.tuple_list_set_item_eliminator 1.62% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.17% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.13% : 0.000014s : 106: predicate.updatestate_useless_node_eliminater 0.53% : 0.000002s : 13: predicate.value_based_eliminate 1.06% : 0.000005s : 26: predicate.virtual_dataset_eliminate 0.81% : 0.000004s : 26: predicate.virtual_output_eliminate 0.36% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.54% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.020656 58 65.26% : 0.013479s : 55: func_graph_cloner_run.FuncGraphClonerGraph 34.74% : 0.007176s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 4.912598 192 0.00% : 0.000004s : 1: ForceFp32Comm 3.85% : 0.188910s : 1: add_attr 3.85% : 0.188890s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 0.00% : 0.000140s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000384s : 1: auto_monad 0.00% : 0.000060s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000012s : 1: bias_add_comm_swap 0.01% : 0.000688s : 1: bootstrap 0.00% : 0.000036s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000088s : 1: control_data_broadcast_order 0.00% : 0.000020s : 1: convert_after_rewriter 0.00% : 0.000072s : 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 0.00% : 0.000022s : 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% : 0.000007s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.01% : 0.000722s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.02% : 0.001128s : 1: mutable_eliminate 0.00% : 0.000011s : 1: offloading_packed_experts 0.00% : 0.000032s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000155s : 1: opt.transform.mutable_eliminate 0.06% : 0.002794s : 78: opt.transform.opt_a 0.00% : 0.000104s : 1: opt.transform.opt_after_cconv 0.00% : 0.000082s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000404s : 28: opt.transform.opt_b 0.00% : 0.000132s : 2: opt.transform.opt_trans_graph 0.00% : 0.000114s : 4: opt.transform.symbol_engine_opt 4.43% : 0.217722s : 1: opt_a 0.01% : 0.000257s : 1: opt_after_cconv 0.02% : 0.000878s : 1: opt_after_jit_grad 0.02% : 0.001077s : 1: opt_b 4.53% : 0.222658s : 1: optimize 0.00% : 0.000037s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000016s : 1: order_py_execute_after_rewriter 0.00% : 0.000051s : 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% : 0.000006s : 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% : 0.000038s : 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% : 0.000081s : 1: remove_dup_value 4.17% : 0.204864s : 1: renormalize.infer 0.16% : 0.007947s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000212s : 1: rewriter_after_opt_a 0.00% : 0.000095s : 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% : 0.000007s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000151s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000202s : 1: symbol_engine_optimizer 0.00% : 0.000173s : 1: tuple_transform 78.79% : 3.870788s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape0-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape0-1024],max_mem:142.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:39:12.958.743 [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 = 4.70554, [21] [bootstrap]: 0.0106705 [type_inference]: 4.2374 [event_method]: 2.138e-05 [auto_monad]: 0.00039118 [graph_reusing]: 6.79001e-06 [inline]: 3.2e-06 [add_attr]: 0.00762916, [1] [add_attr_with_inline]: 0.00761183, [1] [Cycle 1]: 7.614e-05, [2] [tag_attr]: 2.337e-05 [meta_addattr_fg_expand]: 4.20999e-06 [parallel-infer-symbol]: 6.23998e-06 [pre_auto_parallel]: 3.958e-05 [insert-virtual-dataset]: 2.66999e-06 [parallel-infer-symbol-second]: 6.99976e-07 [dataset_repeat_opt]: 1.94e-06 [pipeline_split]: 1.75001e-06 [optimize]: 0.448001, [53] [py_interpret_to_execute]: 3.246e-05 [rewriter_before_opt_a]: 8.763e-05 [opt_a]: 0.313533, [2] [Cycle 1]: 0.311499, [45] [expand_dump_flag]: 2.86999e-06 [switch_simplify]: 3.415e-05 [loop_unroll]: 1.926e-05 [a_1]: 0.0709411 [with_stream_mark]: 7.533e-05 [recompute_prepare]: 6.102e-05 [updatestate_depend_eliminate]: 1.608e-05 [updatestate_assign_eliminate]: 9.36e-06 [updatestate_loads_eliminate]: 1.392e-05 [parameter_eliminate]: 3.04999e-06 [a_2]: 0.00032968 [accelerated_algorithm]: 5.879e-05 [shard]: 3.51001e-06 [meta_shard_fg_expand]: 7.36999e-06 [shard_inline]: 2.252e-05 [merge_send_recv]: 1.694e-05 [auto_parallel]: 1.829e-05 [parallel]: 4.642e-05 [flash_sp]: 1.949e-05 [merge_comm]: 9.88002e-06 [allreduce_fusion]: 9.44e-06 [matmul_add_comm_reduction]: 2.08e-05 [allreduce_slice_to_reducescatter]: 1.38002e-06 [virtual_shard_identity]: 2.666e-05 [virtual_dataset]: 3.116e-05 [get_grad_eliminate_]: 3.12e-05 [virtual_output]: 2.271e-05 [merge_forward]: 1.1e-05 [cell_reuse_recompute_pass]: 3.69002e-06 [offload_activation]: 2.168e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.977e-05 [merge_recompute_call_nodes]: 2.26e-06 [before_grad]: 3.846e-05 [set_forward_comm_id_for_comm_node_pass]: 1.209e-05 [meta_fg_expand]: 7.83001e-06 [flash_sp_send_recv_attached]: 8.95999e-06 [receive_attached]: 2.76e-06 [after_resolve]: 3.603e-05 [a_after_grad]: 4.224e-05 [renormalize]: 0.0305958 [add_forward_monad_depend]: 2.248e-05 [auto_monad_grad]: 3.01999e-06 [auto_monad_eliminator]: 6.784e-05 [cse]: 0.208048 [a_3]: 0.0001676 [Cycle 2]: 0.00201659, [45] [expand_dump_flag]: 4.69998e-06 [switch_simplify]: 1.925e-05 [loop_unroll]: 1.667e-05 [a_1]: 0.00053092 [with_stream_mark]: 4.364e-05 [recompute_prepare]: 2.033e-05 [updatestate_depend_eliminate]: 1.047e-05 [updatestate_assign_eliminate]: 8.12e-06 [updatestate_loads_eliminate]: 1.073e-05 [parameter_eliminate]: 2.93998e-06 [a_2]: 0.00023113 [accelerated_algorithm]: 2.567e-05 [shard]: 2.21e-06 [meta_shard_fg_expand]: 7.41001e-06 [shard_inline]: 1.758e-05 [merge_send_recv]: 1.583e-05 [auto_parallel]: 1.68e-05 [parallel]: 1.11e-05 [flash_sp]: 6.01998e-06 [merge_comm]: 8.64998e-06 [allreduce_fusion]: 1.031e-05 [matmul_add_comm_reduction]: 2.12e-05 [allreduce_slice_to_reducescatter]: 9.70002e-07 [virtual_shard_identity]: 2.123e-05 [virtual_dataset]: 1.758e-05 [get_grad_eliminate_]: 1.588e-05 [virtual_output]: 1.539e-05 [merge_forward]: 9.10001e-06 [cell_reuse_recompute_pass]: 3.86999e-06 [offload_activation]: 1.949e-05 [cell_reuse_handle_not_recompute_node_pass]: 5.784e-05 [merge_recompute_call_nodes]: 1.88997e-06 [before_grad]: 3.239e-05 [set_forward_comm_id_for_comm_node_pass]: 1.028e-05 [meta_fg_expand]: 7.13998e-06 [flash_sp_send_recv_attached]: 1.92999e-06 [receive_attached]: 2.54999e-06 [after_resolve]: 2.658e-05 [a_after_grad]: 3.038e-05 [renormalize]: 5.00004e-08 [add_forward_monad_depend]: 4.28999e-06 [auto_monad_grad]: 3.06999e-06 [auto_monad_eliminator]: 6.942e-05 [cse]: 0.00018406 [a_3]: 0.00011157 [py_interpret_to_execute_after_opt_a]: 3.093e-05 [slice_cell_reuse_recomputed_activation]: 2.89999e-06 [rewriter_after_opt_a]: 0.00022703 [convert_after_rewriter]: 1.728e-05 [order_py_execute_after_rewriter]: 1.291e-05 [mutable_eliminate]: 0.00138056 [opt_b]: 0.000625, [1] [Cycle 1]: 0.00061284, [7] [b_1]: 0.00040674 [b_2]: 2.02e-05 [updatestate_depend_eliminate]: 2.056e-05 [updatestate_assign_eliminate]: 8.52e-06 [updatestate_loads_eliminate]: 1.06e-05 [renormalize]: 1.14e-06 [cse]: 9.536e-05 [optimize_parallel_all_gather_comm]: 3.896e-05 [overlap_param_gather]: 3.03e-06 [cconv]: 4.555e-05 [loop_unroll]: 0.00080204 [opt_after_cconv]: 0.00030703, [1] [Cycle 1]: 0.00029537, [7] [c_1]: 0.00011714 [parameter_eliminate]: 6.41e-06 [updatestate_depend_eliminate]: 1.793e-05 [updatestate_assign_eliminate]: 7.28e-06 [updatestate_loads_eliminate]: 1.043e-05 [cse]: 9.55e-05 [renormalize]: 5.39992e-07 [remove_dup_value]: 8.434e-05 [tuple_transform]: 0.00031175, [1] [Cycle 1]: 0.00028589, [4] [d_1]: 0.00019401 [none_parameter_eliminate]: 5.72999e-06 [renormalize]: 8.59989e-07 [switch_simplify]: 2.276e-05 [partial_unused_args_eliminate]: 2.74999e-06 [add_recomputation]: 0.00014478 [cse_after_recomputation]: 0.000106, [1] [Cycle 1]: 9.97e-05, [1] [cse]: 9.259e-05 [environ_conv]: 2.733e-05 [swap_dp_allreduce_reducescatter]: 1.271e-05 [bias_add_comm_swap]: 3.94002e-06 [label_micro_interleaved_index]: 8.69e-06 [label_fine_grained_interleaved_index]: 2.86999e-06 [merge_cast_opt]: 1.87001e-06 [slice_recompute_activation]: 2.77002e-06 [micro_interleaved_order_control]: 2.84999e-06 [assign_add_opt]: 2.11998e-06 [ForceFp32Comm]: 1.09e-06 [remove_cast_before_assign_add]: 1.19998e-06 [full_micro_interleaved_order_control]: 2.63e-06 [reorder_send_recv_between_fp_bp]: 2.96001e-06 [comm_op_add_attrs]: 1.15999e-06 [add_comm_op_reuse_tag]: 1.37e-06 [interleave_split_concat_branches]: 1.30001e-06 [interleave_parallel_branches]: 1.32e-06 [overlap_opt_shard_in_pipeline]: 3.03e-06 [overlap_opt_shard_grad_in_pipeline]: 1.92001e-06 [control_data_broadcast_order]: 3.081e-05 [grouped_pairwise_exchange_alltoall]: 1.60999e-06 [offloading_packed_experts]: 6.88998e-06 [overlap_recompute_and_grad_model_parallel]: 7.8e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.29e-06 [overlap_recompute_allgather_and_fa_grad]: 1.49e-06 [overlap_recompute_comm]: 2.91999e-06 [overlap_grad_ring_attention]: 8.28001e-06 [overlap_grad_flash_sp]: 4.095e-05 [begin_end_overlap_inline]: 7.30011e-07 [split_matmul_comm_elemetwise]: 2.91999e-06 [split_layernorm_comm]: 1.92001e-06 [handle_group_info]: 1.46998e-06 [symbol_engine_optimizer]: 0.00018175, [1] [Cycle 1]: 0.00017457, [6] [build]: 1.797e-05 [elim_shapecalc]: 2.36e-05 [elim_not_effective]: 4.287e-05 [opt_reshape]: 2.699e-05 [fold_const_symbol]: 2.475e-05 [renormalize]: 3.69997e-07 [detach_backward]: 2.76999e-06 [pipeline_parallel_scheduler]: 1.40999e-06 [auto_monad_reorder]: 5.429e-05 [get_jit_bprop_graph]: 2.32999e-06 [rewriter_after_jit_bprop_graph]: 7.38e-06 [opt_after_jit_grad]: 0.00088811 [validate]: 0.00011188 Sums bootstrap : 0.010670s : 0.23% type_inference : 4.237396s : 92.79% event_method : 0.000021s : 0.00% auto_monad : 0.000391s : 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 : 0.000006s : 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 : 0.000032s : 0.00% optimize.rewriter_before_opt_a : 0.000088s : 0.00% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000053s : 0.00% optimize.opt_a.loop_unroll : 0.000036s : 0.00% optimize.opt_a.a_1 : 0.071472s : 1.57% optimize.opt_a.with_stream_mark : 0.000119s : 0.00% optimize.opt_a.recompute_prepare : 0.000081s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000027s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000017s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000025s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000561s : 0.01% optimize.opt_a.accelerated_algorithm : 0.000084s : 0.00% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000015s : 0.00% optimize.opt_a.shard_inline : 0.000040s : 0.00% optimize.opt_a.merge_send_recv : 0.000033s : 0.00% optimize.opt_a.auto_parallel : 0.000035s : 0.00% optimize.opt_a.parallel : 0.000058s : 0.00% optimize.opt_a.flash_sp : 0.000026s : 0.00% optimize.opt_a.merge_comm : 0.000019s : 0.00% optimize.opt_a.allreduce_fusion : 0.000020s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000042s : 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_ : 0.000047s : 0.00% optimize.opt_a.virtual_output : 0.000038s : 0.00% optimize.opt_a.merge_forward : 0.000020s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000008s : 0.00% optimize.opt_a.offload_activation : 0.000041s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000108s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000071s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000022s : 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 : 0.000063s : 0.00% optimize.opt_a.a_after_grad : 0.000073s : 0.00% optimize.opt_a.renormalize : 0.030596s : 0.67% optimize.opt_a.add_forward_monad_depend : 0.000027s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000137s : 0.00% optimize.opt_a.cse : 0.208232s : 4.56% optimize.opt_a.a_3 : 0.000279s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000031s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000227s : 0.00% optimize.convert_after_rewriter : 0.000017s : 0.00% optimize.order_py_execute_after_rewriter : 0.000013s : 0.00% optimize.mutable_eliminate : 0.001381s : 0.03% optimize.opt_b.b_1 : 0.000407s : 0.01% optimize.opt_b.b_2 : 0.000020s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000021s : 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 : 0.000095s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000039s : 0.00% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000046s : 0.00% optimize.loop_unroll : 0.000802s : 0.02% optimize.opt_after_cconv.c_1 : 0.000117s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000018s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.cse : 0.000095s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000084s : 0.00% optimize.tuple_transform.d_1 : 0.000194s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000006s : 0.00% optimize.tuple_transform.renormalize : 0.000001s : 0.00% optimize.tuple_transform.switch_simplify : 0.000023s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000145s : 0.00% optimize.cse_after_recomputation.cse : 0.000093s : 0.00% optimize.environ_conv : 0.000027s : 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 : 0.000009s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000003s : 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 : 0.000003s : 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 : 0.000007s : 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 : 0.000041s : 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 : 0.000018s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000024s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000043s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000027s : 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 : 0.000001s : 0.00% auto_monad_reorder : 0.000054s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000888s : 0.02% validate : 0.000112s : 0.00% Time group info: ------[substitution.] 0.000358 96 3.25% : 0.000012s : 2: substitution.depend_value_elim 1.32% : 0.000005s : 8: substitution.elim_not_effective 0.91% : 0.000003s : 8: substitution.fold_const_symbol 6.46% : 0.000023s : 13: substitution.graph_param_transform 35.08% : 0.000126s : 1: substitution.inline 3.03% : 0.000011s : 16: substitution.j_node_and_user_rematch 8.14% : 0.000029s : 2: substitution.less_batch_normalization 1.78% : 0.000006s : 6: substitution.load_eliminater 1.20% : 0.000004s : 2: substitution.opt_reshape 9.18% : 0.000033s : 16: substitution.remove_not_recompute_node 2.95% : 0.000011s : 4: substitution.replace_old_param 10.71% : 0.000038s : 4: substitution.reshape_eliminate 9.30% : 0.000033s : 6: substitution.updatestate_pure_node_eliminater 6.68% : 0.000024s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 4.237278 2 94.43% : 4.001237s : 1: type_inference.infer 5.57% : 0.236041s : 1: type_inference.specialize ------[replace.] 0.000019 1 100.00% : 0.000019s : 1: replace.inline ------[match.] 0.000125 1 100.00% : 0.000125s : 1: match.inline ------[predicate.] 0.000494 3002 0.79% : 0.000004s : 27: predicate.accumulaten_eliminater 0.94% : 0.000005s : 13: predicate.ad_related_special_op_eliminate 0.74% : 0.000004s : 26: predicate.addn_check_dump 0.75% : 0.000004s : 27: predicate.addn_zero_filter 0.70% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.38% : 0.000012s : 53: predicate.arithmetic_simplify 0.79% : 0.000004s : 27: predicate.cast_eliminate 0.84% : 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 0.79% : 0.000004s : 26: predicate.depend_value_elim 0.78% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.84% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.74% : 0.000004s : 27: predicate.dict_set_item_eliminator 2.13% : 0.000011s : 26: predicate.dumpgradient_eliminate 0.27% : 0.000001s : 13: predicate.elim_not_effective 0.48% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.16% : 0.000006s : 40: predicate.environ_add_const_eliminate 1.14% : 0.000006s : 40: predicate.environ_get_add_eliminate 1.08% : 0.000005s : 40: predicate.environ_get_depend_swap 1.98% : 0.000010s : 66: predicate.environ_get_eliminate 1.18% : 0.000006s : 40: predicate.environ_get_set_eliminate 0.72% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.15% : 0.000006s : 28: predicate.float_depend_g_call 0.79% : 0.000004s : 26: predicate.float_environ_get_switch 1.25% : 0.000006s : 39: predicate.float_tuple_getitem_switch 0.21% : 0.000001s : 13: predicate.fold_const_symbol 1.04% : 0.000005s : 26: predicate.get_grad_eliminate 0.36% : 0.000002s : 13: predicate.graph_param_transform 0.78% : 0.000004s : 26: predicate.incorporate_call 0.64% : 0.000003s : 26: predicate.incorporate_call_switch 4.80% : 0.000024s : 133: predicate.inline 1.14% : 0.000006s : 26: predicate.inline_without_move 0.50% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.20% : 0.000006s : 27: predicate.less_batch_normalization 1.58% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.21% : 0.000011s : 80: predicate.load_eliminater 1.42% : 0.000007s : 13: predicate.loop_unroll_after_grad 0.91% : 0.000004s : 31: predicate.loop_unroll_before_grad 1.77% : 0.000009s : 53: predicate.make_slice_get_slice_eliminator 0.81% : 0.000004s : 26: predicate.merge_addn 0.73% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.83% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.77% : 0.000004s : 27: predicate.minmaximum_grad 2.12% : 0.000010s : 13: predicate.mutable_eliminate 0.49% : 0.000002s : 13: predicate.opt_reshape 0.60% : 0.000003s : 13: predicate.parallel_virtual_node 1.08% : 0.000005s : 28: predicate.partial_defer_inline 1.12% : 0.000006s : 40: predicate.partial_eliminate 0.74% : 0.000004s : 27: predicate.print_const_string_wrapper 0.80% : 0.000004s : 26: predicate.reduce_all_const_elim 1.07% : 0.000005s : 27: predicate.reduce_eliminate 2.66% : 0.000013s : 80: predicate.redundant_stop_gradient_eliminater 0.57% : 0.000003s : 26: predicate.remove_not_recompute_node 1.37% : 0.000007s : 53: predicate.replace_applicator 0.53% : 0.000003s : 26: predicate.replace_old_param 0.32% : 0.000002s : 13: predicate.reset_defer_inline 1.03% : 0.000005s : 27: predicate.reshape_eliminate 0.81% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.48% : 0.000002s : 13: predicate.row_tensor_eliminate 1.31% : 0.000006s : 26: predicate.same_eliminate 0.79% : 0.000004s : 26: predicate.set_cell_output_no_recompute 1.02% : 0.000005s : 26: predicate.shard_identity_eliminate 1.00% : 0.000005s : 26: predicate.special_op_eliminate 0.86% : 0.000004s : 26: predicate.specialize_transform 1.18% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.99% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.43% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.79% : 0.000004s : 28: predicate.switch_defer_inline 1.52% : 0.000008s : 54: predicate.switch_layer_defer_inline 3.11% : 0.000015s : 98: predicate.switch_simplify 0.87% : 0.000004s : 27: predicate.tile_eliminate 0.71% : 0.000004s : 27: predicate.transpose_eliminate 1.72% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 1.74% : 0.000009s : 53: predicate.tuple_list_get_item_const_eliminator 1.53% : 0.000008s : 53: predicate.tuple_list_get_item_depend_reorder 3.17% : 0.000016s : 79: predicate.tuple_list_get_item_eliminator 1.66% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.83% : 0.000014s : 79: predicate.tuple_list_set_item_eliminator 1.48% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.27% : 0.000011s : 80: predicate.updatestate_pure_node_eliminater 3.01% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.47% : 0.000002s : 13: predicate.value_based_eliminate 0.94% : 0.000005s : 26: predicate.virtual_dataset_eliminate 0.80% : 0.000004s : 26: predicate.virtual_output_eliminate 0.33% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.54% : 0.000003s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.206489 58 90.47% : 0.186817s : 55: func_graph_cloner_run.FuncGraphClonerGraph 9.53% : 0.019672s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 5.265358 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.15% : 0.007637s : 1: add_attr 0.14% : 0.007616s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000150s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.01% : 0.000402s : 1: auto_monad 0.00% : 0.000059s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.20% : 0.010753s : 1: bootstrap 0.00% : 0.000051s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000034s : 1: control_data_broadcast_order 0.00% : 0.000023s : 1: convert_after_rewriter 0.00% : 0.000109s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000033s : 1: environ_conv 0.00% : 0.000030s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 1: graph_reusing 0.00% : 0.000004s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000012s : 1: label_micro_interleaved_index 0.02% : 0.000821s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.03% : 0.001408s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000040s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000058s : 1: opt.transform.mutable_eliminate 1.39% : 0.072929s : 78: opt.transform.opt_a 0.00% : 0.000115s : 1: opt.transform.opt_after_cconv 0.00% : 0.000067s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000381s : 28: opt.transform.opt_b 0.00% : 0.000205s : 2: opt.transform.opt_trans_graph 0.00% : 0.000114s : 4: opt.transform.symbol_engine_opt 5.95% : 0.313538s : 1: opt_a 0.01% : 0.000311s : 1: opt_after_cconv 0.02% : 0.000907s : 1: opt_after_jit_grad 0.01% : 0.000630s : 1: opt_b 8.51% : 0.448009s : 1: optimize 0.00% : 0.000044s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000016s : 1: order_py_execute_after_rewriter 0.00% : 0.000045s : 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% : 0.000006s : 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% : 0.000011s : 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% : 0.000009s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000007s : 1: pipeline_split 0.00% : 0.000044s : 1: pre_auto_parallel 0.00% : 0.000036s : 1: py_interpret_to_execute 0.00% : 0.000036s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 2.46% : 0.129555s : 1: remove_dup_value 0.20% : 0.010369s : 1: renormalize.infer 0.38% : 0.020200s : 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% : 0.000238s : 1: rewriter_after_opt_a 0.00% : 0.000092s : 1: rewriter_before_opt_a 0.00% : 0.000007s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000006s : 1: slice_recompute_activation 0.00% : 0.000007s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000016s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000185s : 1: symbol_engine_optimizer 0.01% : 0.000323s : 1: tuple_transform 80.48% : 4.237452s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape1-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape1-256],max_mem:146.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:40:07.818.522 [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 = 3.69477, [21] [bootstrap]: 0.0136878 [type_inference]: 3.46552 [event_method]: 2.132e-05 [auto_monad]: 0.00061692 [graph_reusing]: 7.58001e-06 [inline]: 2.50002e-06 [add_attr]: 0.00801563, [1] [add_attr_with_inline]: 0.00799759, [1] [Cycle 1]: 7.922e-05, [2] [tag_attr]: 2.627e-05 [meta_addattr_fg_expand]: 4.34002e-06 [parallel-infer-symbol]: 5.87999e-06 [pre_auto_parallel]: 4.082e-05 [insert-virtual-dataset]: 2.63998e-06 [parallel-infer-symbol-second]: 7.60017e-07 [dataset_repeat_opt]: 2.05002e-06 [pipeline_split]: 1.85001e-06 [optimize]: 0.205564, [53] [py_interpret_to_execute]: 3.595e-05 [rewriter_before_opt_a]: 8.776e-05 [opt_a]: 0.201381, [2] [Cycle 1]: 0.191658, [45] [expand_dump_flag]: 3.03998e-06 [switch_simplify]: 3.573e-05 [loop_unroll]: 2.065e-05 [a_1]: 0.00121141 [with_stream_mark]: 3.893e-05 [recompute_prepare]: 2.985e-05 [updatestate_depend_eliminate]: 1.4e-05 [updatestate_assign_eliminate]: 9.04e-06 [updatestate_loads_eliminate]: 1.335e-05 [parameter_eliminate]: 2.56e-06 [a_2]: 0.00031354 [accelerated_algorithm]: 6.008e-05 [shard]: 4.07998e-06 [meta_shard_fg_expand]: 5.80002e-06 [shard_inline]: 2.162e-05 [merge_send_recv]: 1.638e-05 [auto_parallel]: 1.608e-05 [parallel]: 4.05e-05 [flash_sp]: 1.961e-05 [merge_comm]: 9.40001e-06 [allreduce_fusion]: 9.84999e-06 [matmul_add_comm_reduction]: 2.043e-05 [allreduce_slice_to_reducescatter]: 9.5999e-07 [virtual_shard_identity]: 2.71e-05 [virtual_dataset]: 2.172e-05 [get_grad_eliminate_]: 2.593e-05 [virtual_output]: 2.345e-05 [merge_forward]: 1.117e-05 [cell_reuse_recompute_pass]: 3.73001e-06 [offload_activation]: 2.144e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.505e-05 [merge_recompute_call_nodes]: 1.91e-06 [before_grad]: 4.186e-05 [set_forward_comm_id_for_comm_node_pass]: 1.007e-05 [meta_fg_expand]: 7.05002e-06 [flash_sp_send_recv_attached]: 8.65999e-06 [receive_attached]: 3.16001e-06 [after_resolve]: 3.32e-05 [a_after_grad]: 3.97e-05 [renormalize]: 0.187751 [add_forward_monad_depend]: 2.028e-05 [auto_monad_grad]: 3.5e-06 [auto_monad_eliminator]: 5.749e-05 [cse]: 0.00103778 [a_3]: 0.00014736 [Cycle 2]: 0.00970564, [45] [expand_dump_flag]: 3.79002e-06 [switch_simplify]: 1.91e-05 [loop_unroll]: 1.667e-05 [a_1]: 0.00047935 [with_stream_mark]: 2.86e-05 [recompute_prepare]: 1.753e-05 [updatestate_depend_eliminate]: 9.77001e-06 [updatestate_assign_eliminate]: 7.82998e-06 [updatestate_loads_eliminate]: 1.076e-05 [parameter_eliminate]: 2.48e-06 [a_2]: 0.00027126 [accelerated_algorithm]: 2.407e-05 [shard]: 2.62001e-06 [meta_shard_fg_expand]: 4.19002e-06 [shard_inline]: 1.555e-05 [merge_send_recv]: 1.349e-05 [auto_parallel]: 1.649e-05 [parallel]: 1.12e-05 [flash_sp]: 0.00756048 [merge_comm]: 4.937e-05 [allreduce_fusion]: 1.299e-05 [matmul_add_comm_reduction]: 2.383e-05 [allreduce_slice_to_reducescatter]: 9.20001e-07 [virtual_shard_identity]: 6.901e-05 [virtual_dataset]: 2.068e-05 [get_grad_eliminate_]: 2.027e-05 [virtual_output]: 1.718e-05 [merge_forward]: 1.185e-05 [cell_reuse_recompute_pass]: 3.9e-06 [offload_activation]: 2.01e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.799e-05 [merge_recompute_call_nodes]: 1.87999e-06 [before_grad]: 2.894e-05 [set_forward_comm_id_for_comm_node_pass]: 1.066e-05 [meta_fg_expand]: 1.11e-05 [flash_sp_send_recv_attached]: 2.67001e-06 [receive_attached]: 2.88e-06 [after_resolve]: 2.735e-05 [a_after_grad]: 3.114e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 7.8e-06 [auto_monad_grad]: 3.61001e-06 [auto_monad_eliminator]: 8.157e-05 [cse]: 0.00020231 [a_3]: 0.00012014 [py_interpret_to_execute_after_opt_a]: 3.57e-05 [slice_cell_reuse_recomputed_activation]: 2.46998e-06 [rewriter_after_opt_a]: 0.00022137 [convert_after_rewriter]: 1.651e-05 [order_py_execute_after_rewriter]: 1.274e-05 [mutable_eliminate]: 0.00090633 [opt_b]: 0.00059804, [1] [Cycle 1]: 0.00058543, [7] [b_1]: 0.00038755 [b_2]: 2.234e-05 [updatestate_depend_eliminate]: 1.819e-05 [updatestate_assign_eliminate]: 7.66999e-06 [updatestate_loads_eliminate]: 1.168e-05 [renormalize]: 9.49978e-07 [cse]: 8.128e-05 [optimize_parallel_all_gather_comm]: 3.616e-05 [overlap_param_gather]: 2.46e-06 [cconv]: 4.238e-05 [loop_unroll]: 0.00066794 [opt_after_cconv]: 0.00027585, [1] [Cycle 1]: 0.00026351, [7] [c_1]: 0.00011474 [parameter_eliminate]: 5.86003e-06 [updatestate_depend_eliminate]: 1.882e-05 [updatestate_assign_eliminate]: 6.72002e-06 [updatestate_loads_eliminate]: 1.141e-05 [cse]: 6.381e-05 [renormalize]: 8.59989e-07 [remove_dup_value]: 8.132e-05 [tuple_transform]: 0.00021773, [1] [Cycle 1]: 0.00021005, [4] [d_1]: 0.00015274 [none_parameter_eliminate]: 3.34001e-06 [renormalize]: 1.8999e-07 [switch_simplify]: 2.143e-05 [partial_unused_args_eliminate]: 2.81e-06 [add_recomputation]: 0.00012657 [cse_after_recomputation]: 7.744e-05, [1] [Cycle 1]: 7.044e-05, [1] [cse]: 6.053e-05 [environ_conv]: 2.832e-05 [swap_dp_allreduce_reducescatter]: 1.472e-05 [bias_add_comm_swap]: 3.93001e-06 [label_micro_interleaved_index]: 7.17002e-06 [label_fine_grained_interleaved_index]: 2.68e-06 [merge_cast_opt]: 1.59998e-06 [slice_recompute_activation]: 2.06e-06 [micro_interleaved_order_control]: 2.55002e-06 [assign_add_opt]: 1.44e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.15001e-06 [full_micro_interleaved_order_control]: 2.41e-06 [reorder_send_recv_between_fp_bp]: 3.18e-06 [comm_op_add_attrs]: 1.20001e-06 [add_comm_op_reuse_tag]: 1.30999e-06 [interleave_split_concat_branches]: 1.60001e-06 [interleave_parallel_branches]: 1.37e-06 [overlap_opt_shard_in_pipeline]: 1.35001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.29999e-06 [control_data_broadcast_order]: 3.181e-05 [grouped_pairwise_exchange_alltoall]: 1.98002e-06 [offloading_packed_experts]: 7.05e-06 [overlap_recompute_and_grad_model_parallel]: 8.07e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.49998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.52999e-06 [overlap_recompute_comm]: 2.89001e-06 [overlap_grad_ring_attention]: 8.50001e-06 [overlap_grad_flash_sp]: 4.211e-05 [begin_end_overlap_inline]: 5.90022e-07 [split_matmul_comm_elemetwise]: 2.66e-06 [split_layernorm_comm]: 1.84e-06 [handle_group_info]: 1.23002e-06 [symbol_engine_optimizer]: 0.00020321, [1] [Cycle 1]: 0.00019423, [6] [build]: 2.321e-05 [elim_shapecalc]: 3.269e-05 [elim_not_effective]: 3.622e-05 [opt_reshape]: 2.585e-05 [fold_const_symbol]: 2.702e-05 [renormalize]: 1.8999e-07 [detach_backward]: 3.97e-06 [pipeline_parallel_scheduler]: 1.44998e-06 [auto_monad_reorder]: 5.935e-05 [get_jit_bprop_graph]: 2.36998e-06 [rewriter_after_jit_bprop_graph]: 8.82e-06 [opt_after_jit_grad]: 0.00080344 [validate]: 0.00011511 Sums bootstrap : 0.013688s : 0.37% type_inference : 3.465522s : 94.04% event_method : 0.000021s : 0.00% auto_monad : 0.000617s : 0.02% graph_reusing : 0.000008s : 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 : 0.000004s : 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 : 0.000036s : 0.00% optimize.rewriter_before_opt_a : 0.000088s : 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 : 0.000037s : 0.00% optimize.opt_a.a_1 : 0.001691s : 0.05% optimize.opt_a.with_stream_mark : 0.000068s : 0.00% optimize.opt_a.recompute_prepare : 0.000047s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000024s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000017s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000024s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000585s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000084s : 0.00% optimize.opt_a.shard : 0.000007s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000037s : 0.00% optimize.opt_a.merge_send_recv : 0.000030s : 0.00% optimize.opt_a.auto_parallel : 0.000033s : 0.00% optimize.opt_a.parallel : 0.000052s : 0.00% optimize.opt_a.flash_sp : 0.007580s : 0.21% optimize.opt_a.merge_comm : 0.000059s : 0.00% optimize.opt_a.allreduce_fusion : 0.000023s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000044s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000096s : 0.00% optimize.opt_a.virtual_dataset : 0.000042s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000046s : 0.00% optimize.opt_a.virtual_output : 0.000041s : 0.00% optimize.opt_a.merge_forward : 0.000023s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000008s : 0.00% optimize.opt_a.offload_activation : 0.000042s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000093s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000071s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000021s : 0.00% optimize.opt_a.meta_fg_expand : 0.000018s : 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 : 0.000061s : 0.00% optimize.opt_a.a_after_grad : 0.000071s : 0.00% optimize.opt_a.renormalize : 0.187751s : 5.09% optimize.opt_a.add_forward_monad_depend : 0.000028s : 0.00% optimize.opt_a.auto_monad_grad : 0.000007s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000139s : 0.00% optimize.opt_a.cse : 0.001240s : 0.03% optimize.opt_a.a_3 : 0.000267s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000036s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000221s : 0.01% optimize.convert_after_rewriter : 0.000017s : 0.00% optimize.order_py_execute_after_rewriter : 0.000013s : 0.00% optimize.mutable_eliminate : 0.000906s : 0.02% optimize.opt_b.b_1 : 0.000388s : 0.01% optimize.opt_b.b_2 : 0.000022s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000018s : 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 : 0.000081s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000036s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000042s : 0.00% optimize.loop_unroll : 0.000668s : 0.02% optimize.opt_after_cconv.c_1 : 0.000115s : 0.00% 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 : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.cse : 0.000064s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000081s : 0.00% optimize.tuple_transform.d_1 : 0.000153s : 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 : 0.000021s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000127s : 0.00% optimize.cse_after_recomputation.cse : 0.000061s : 0.00% optimize.environ_conv : 0.000028s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000015s : 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 : 0.000002s : 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 : 0.000032s : 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 : 0.000009s : 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 : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000023s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000033s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000036s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000026s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000027s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000004s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000059s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000009s : 0.00% opt_after_jit_grad : 0.000803s : 0.02% validate : 0.000115s : 0.00% Time group info: ------[substitution.] 0.000327 96 3.51% : 0.000011s : 2: substitution.depend_value_elim 1.40% : 0.000005s : 8: substitution.elim_not_effective 1.02% : 0.000003s : 8: substitution.fold_const_symbol 3.79% : 0.000012s : 13: substitution.graph_param_transform 41.29% : 0.000135s : 1: substitution.inline 3.22% : 0.000011s : 16: substitution.j_node_and_user_rematch 8.78% : 0.000029s : 2: substitution.less_batch_normalization 2.07% : 0.000007s : 6: substitution.load_eliminater 1.34% : 0.000004s : 2: substitution.opt_reshape 4.23% : 0.000014s : 16: substitution.remove_not_recompute_node 3.58% : 0.000012s : 4: substitution.replace_old_param 10.77% : 0.000035s : 4: substitution.reshape_eliminate 7.84% : 0.000026s : 6: substitution.updatestate_pure_node_eliminater 7.16% : 0.000023s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 3.465394 2 99.81% : 3.458788s : 1: type_inference.infer 0.19% : 0.006606s : 1: type_inference.specialize ------[replace.] 0.000020 1 100.00% : 0.000020s : 1: replace.inline ------[match.] 0.000134 1 100.00% : 0.000134s : 1: match.inline ------[predicate.] 0.000482 3002 0.78% : 0.000004s : 27: predicate.accumulaten_eliminater 1.33% : 0.000006s : 13: predicate.ad_related_special_op_eliminate 0.72% : 0.000003s : 26: predicate.addn_check_dump 0.84% : 0.000004s : 27: predicate.addn_zero_filter 0.78% : 0.000004s : 27: predicate.adjust_all_reduce_mul_add 2.19% : 0.000011s : 53: predicate.arithmetic_simplify 0.83% : 0.000004s : 27: predicate.cast_eliminate 0.81% : 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 0.81% : 0.000004s : 26: predicate.depend_value_elim 0.84% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.87% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.85% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.13% : 0.000005s : 26: predicate.dumpgradient_eliminate 0.28% : 0.000001s : 13: predicate.elim_not_effective 0.75% : 0.000004s : 13: predicate.elim_shapecalc_of_broadcastargs 1.44% : 0.000007s : 40: predicate.environ_add_const_eliminate 1.11% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.24% : 0.000006s : 40: predicate.environ_get_depend_swap 1.89% : 0.000009s : 66: predicate.environ_get_eliminate 1.15% : 0.000006s : 40: predicate.environ_get_set_eliminate 0.76% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.15% : 0.000006s : 28: predicate.float_depend_g_call 0.78% : 0.000004s : 26: predicate.float_environ_get_switch 1.18% : 0.000006s : 39: predicate.float_tuple_getitem_switch 0.21% : 0.000001s : 13: predicate.fold_const_symbol 1.09% : 0.000005s : 26: predicate.get_grad_eliminate 0.23% : 0.000001s : 13: predicate.graph_param_transform 0.85% : 0.000004s : 26: predicate.incorporate_call 0.67% : 0.000003s : 26: predicate.incorporate_call_switch 5.42% : 0.000026s : 133: predicate.inline 1.14% : 0.000005s : 26: predicate.inline_without_move 0.40% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.33% : 0.000006s : 27: predicate.less_batch_normalization 1.84% : 0.000009s : 53: predicate.list_to_tuple_eliminator_ 2.20% : 0.000011s : 80: predicate.load_eliminater 1.18% : 0.000006s : 13: predicate.loop_unroll_after_grad 0.90% : 0.000004s : 31: predicate.loop_unroll_before_grad 1.75% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.79% : 0.000004s : 26: predicate.merge_addn 0.73% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.75% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.69% : 0.000003s : 27: predicate.minmaximum_grad 1.54% : 0.000007s : 13: predicate.mutable_eliminate 0.51% : 0.000002s : 13: predicate.opt_reshape 0.46% : 0.000002s : 13: predicate.parallel_virtual_node 1.37% : 0.000007s : 28: predicate.partial_defer_inline 1.18% : 0.000006s : 40: predicate.partial_eliminate 0.81% : 0.000004s : 27: predicate.print_const_string_wrapper 0.79% : 0.000004s : 26: predicate.reduce_all_const_elim 1.07% : 0.000005s : 27: predicate.reduce_eliminate 2.21% : 0.000011s : 80: predicate.redundant_stop_gradient_eliminater 0.64% : 0.000003s : 26: predicate.remove_not_recompute_node 1.33% : 0.000006s : 53: predicate.replace_applicator 0.50% : 0.000002s : 26: predicate.replace_old_param 0.25% : 0.000001s : 13: predicate.reset_defer_inline 0.91% : 0.000004s : 27: predicate.reshape_eliminate 0.81% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.47% : 0.000002s : 13: predicate.row_tensor_eliminate 1.21% : 0.000006s : 26: predicate.same_eliminate 0.55% : 0.000003s : 26: predicate.set_cell_output_no_recompute 1.47% : 0.000007s : 26: predicate.shard_identity_eliminate 0.89% : 0.000004s : 26: predicate.special_op_eliminate 0.92% : 0.000004s : 26: predicate.specialize_transform 1.22% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 1.05% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.46% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.87% : 0.000004s : 28: predicate.switch_defer_inline 1.63% : 0.000008s : 54: predicate.switch_layer_defer_inline 3.27% : 0.000016s : 98: predicate.switch_simplify 0.83% : 0.000004s : 27: predicate.tile_eliminate 0.89% : 0.000004s : 27: predicate.transpose_eliminate 1.50% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.62% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.56% : 0.000008s : 53: predicate.tuple_list_get_item_depend_reorder 2.82% : 0.000014s : 79: predicate.tuple_list_get_item_eliminator 1.58% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.70% : 0.000013s : 79: predicate.tuple_list_set_item_eliminator 1.73% : 0.000008s : 53: predicate.tuple_to_list_eliminator_ 2.12% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.18% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.53% : 0.000003s : 13: predicate.value_based_eliminate 0.93% : 0.000005s : 26: predicate.virtual_dataset_eliminate 1.04% : 0.000005s : 26: predicate.virtual_output_eliminate 0.36% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.50% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.014988 58 82.20% : 0.012321s : 55: func_graph_cloner_run.FuncGraphClonerGraph 17.80% : 0.002667s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 4.099894 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.20% : 0.008024s : 1: add_attr 0.20% : 0.008003s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000135s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.02% : 0.000636s : 1: auto_monad 0.00% : 0.000069s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.34% : 0.013750s : 1: bootstrap 0.00% : 0.000048s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000038s : 1: control_data_broadcast_order 0.00% : 0.000021s : 1: convert_after_rewriter 0.00% : 0.000081s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000009s : 1: detach_backward 0.00% : 0.000034s : 1: environ_conv 0.00% : 0.000030s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000013s : 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% : 0.000005s : 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% : 0.000688s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.02% : 0.000925s : 1: mutable_eliminate 0.00% : 0.000011s : 1: offloading_packed_experts 0.00% : 0.000041s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000045s : 1: opt.transform.mutable_eliminate 0.08% : 0.003167s : 78: opt.transform.opt_a 0.00% : 0.000113s : 1: opt.transform.opt_after_cconv 0.00% : 0.000074s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000367s : 28: opt.transform.opt_b 0.00% : 0.000169s : 2: opt.transform.opt_trans_graph 0.00% : 0.000115s : 4: opt.transform.symbol_engine_opt 4.91% : 0.201386s : 1: opt_a 0.01% : 0.000281s : 1: opt_after_cconv 0.02% : 0.000826s : 1: opt_after_jit_grad 0.01% : 0.000602s : 1: opt_b 5.01% : 0.205572s : 1: optimize 0.00% : 0.000041s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000016s : 1: order_py_execute_after_rewriter 0.00% : 0.000049s : 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% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000007s : 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% : 0.000010s : 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% : 0.000007s : 1: pipeline_split 0.00% : 0.000046s : 1: pre_auto_parallel 0.00% : 0.000040s : 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% : 0.000089s : 1: remove_dup_value 4.50% : 0.184676s : 1: renormalize.infer 0.07% : 0.003053s : 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% : 0.000232s : 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% : 0.000007s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000018s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000206s : 1: symbol_engine_optimizer 0.01% : 0.000221s : 1: tuple_transform 84.53% : 3.465560s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape1-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape1-1024],max_mem:170.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:40:56.528.288 [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 = 3.04718, [21] [bootstrap]: 0.00082575 [type_inference]: 2.76524 [event_method]: 1.85e-05 [auto_monad]: 0.00037429 [graph_reusing]: 6.77002e-06 [inline]: 3.44001e-06 [add_attr]: 0.131014, [1] [add_attr_with_inline]: 0.130997, [1] [Cycle 1]: 7.872e-05, [2] [tag_attr]: 2.506e-05 [meta_addattr_fg_expand]: 5.05001e-06 [parallel-infer-symbol]: 6.23998e-06 [pre_auto_parallel]: 4.169e-05 [insert-virtual-dataset]: 2.63e-06 [parallel-infer-symbol-second]: 6.80011e-07 [dataset_repeat_opt]: 2.24001e-06 [pipeline_split]: 2.00002e-06 [optimize]: 0.14849, [53] [py_interpret_to_execute]: 3.376e-05 [rewriter_before_opt_a]: 9.424e-05 [opt_a]: 0.144477, [2] [Cycle 1]: 0.117093, [45] [expand_dump_flag]: 2.91e-06 [switch_simplify]: 3.766e-05 [loop_unroll]: 2.007e-05 [a_1]: 0.00117232 [with_stream_mark]: 3.433e-05 [recompute_prepare]: 2.524e-05 [updatestate_depend_eliminate]: 1.175e-05 [updatestate_assign_eliminate]: 9.19e-06 [updatestate_loads_eliminate]: 1.218e-05 [parameter_eliminate]: 2.39001e-06 [a_2]: 0.00028699 [accelerated_algorithm]: 4.863e-05 [shard]: 2.47001e-06 [meta_shard_fg_expand]: 4.89e-06 [shard_inline]: 2.096e-05 [merge_send_recv]: 1.395e-05 [auto_parallel]: 1.601e-05 [parallel]: 3.933e-05 [flash_sp]: 1.633e-05 [merge_comm]: 1.013e-05 [allreduce_fusion]: 8.28001e-06 [matmul_add_comm_reduction]: 2.023e-05 [allreduce_slice_to_reducescatter]: 9.00007e-07 [virtual_shard_identity]: 2.279e-05 [virtual_dataset]: 1.961e-05 [get_grad_eliminate_]: 2.367e-05 [virtual_output]: 1.955e-05 [merge_forward]: 1.051e-05 [cell_reuse_recompute_pass]: 2.36e-06 [offload_activation]: 2.059e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.823e-05 [merge_recompute_call_nodes]: 1.92001e-06 [before_grad]: 3.275e-05 [set_forward_comm_id_for_comm_node_pass]: 1.098e-05 [meta_fg_expand]: 6.43e-06 [flash_sp_send_recv_attached]: 7.6e-06 [receive_attached]: 2.75002e-06 [after_resolve]: 3.064e-05 [a_after_grad]: 3.449e-05 [renormalize]: 0.114219 [add_forward_monad_depend]: 2.341e-05 [auto_monad_grad]: 2.73998e-06 [auto_monad_eliminator]: 6.508e-05 [cse]: 0.0001541 [a_3]: 0.0001655 [Cycle 2]: 0.0273678, [45] [expand_dump_flag]: 3.6e-06 [switch_simplify]: 1.904e-05 [loop_unroll]: 1.617e-05 [a_1]: 0.0004785 [with_stream_mark]: 2.661e-05 [recompute_prepare]: 1.664e-05 [updatestate_depend_eliminate]: 9.44e-06 [updatestate_assign_eliminate]: 8.35001e-06 [updatestate_loads_eliminate]: 1.046e-05 [parameter_eliminate]: 7.56999e-06 [a_2]: 0.0003171 [accelerated_algorithm]: 2.859e-05 [shard]: 3.48999e-06 [meta_shard_fg_expand]: 7.28e-06 [shard_inline]: 1.54e-05 [merge_send_recv]: 2.35e-05 [auto_parallel]: 1.775e-05 [parallel]: 1.12e-05 [flash_sp]: 5.72999e-06 [merge_comm]: 8.17e-06 [allreduce_fusion]: 1.075e-05 [matmul_add_comm_reduction]: 1.952e-05 [allreduce_slice_to_reducescatter]: 8.70001e-07 [virtual_shard_identity]: 1.969e-05 [virtual_dataset]: 1.657e-05 [get_grad_eliminate_]: 1.543e-05 [virtual_output]: 1.592e-05 [merge_forward]: 8.74e-06 [cell_reuse_recompute_pass]: 3.08998e-06 [offload_activation]: 1.68e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.087e-05 [merge_recompute_call_nodes]: 1.39998e-06 [before_grad]: 2.438e-05 [set_forward_comm_id_for_comm_node_pass]: 9.46e-06 [meta_fg_expand]: 6.47001e-06 [flash_sp_send_recv_attached]: 1.99999e-06 [receive_attached]: 2.76999e-06 [after_resolve]: 2.348e-05 [a_after_grad]: 2.787e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 4.67e-06 [auto_monad_grad]: 3.06001e-06 [auto_monad_eliminator]: 6.312e-05 [cse]: 0.00215647 [a_3]: 0.00014736 [py_interpret_to_execute_after_opt_a]: 0.00011535 [slice_cell_reuse_recomputed_activation]: 3.45e-06 [rewriter_after_opt_a]: 0.00031834 [convert_after_rewriter]: 1.58e-05 [order_py_execute_after_rewriter]: 1.214e-05 [mutable_eliminate]: 0.00090289 [opt_b]: 0.00058575, [1] [Cycle 1]: 0.00057492, [7] [b_1]: 0.00038889 [b_2]: 1.865e-05 [updatestate_depend_eliminate]: 1.598e-05 [updatestate_assign_eliminate]: 7.51001e-06 [updatestate_loads_eliminate]: 1.243e-05 [renormalize]: 6.09987e-07 [cse]: 8.628e-05 [optimize_parallel_all_gather_comm]: 3.468e-05 [overlap_param_gather]: 6.61999e-06 [cconv]: 4.516e-05 [loop_unroll]: 0.00056993 [opt_after_cconv]: 0.00023528, [1] [Cycle 1]: 0.00022529, [7] [c_1]: 0.00010005 [parameter_eliminate]: 5.87999e-06 [updatestate_depend_eliminate]: 1.524e-05 [updatestate_assign_eliminate]: 6.82002e-06 [updatestate_loads_eliminate]: 8.80001e-06 [cse]: 5.165e-05 [renormalize]: 3.19997e-07 [remove_dup_value]: 7.212e-05 [tuple_transform]: 0.0001633, [1] [Cycle 1]: 0.00015794, [4] [d_1]: 0.00011449 [none_parameter_eliminate]: 2.12001e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 1.745e-05 [partial_unused_args_eliminate]: 1.71998e-06 [add_recomputation]: 0.00010596 [cse_after_recomputation]: 5.655e-05, [1] [Cycle 1]: 5.141e-05, [1] [cse]: 4.516e-05 [environ_conv]: 2.646e-05 [swap_dp_allreduce_reducescatter]: 1.123e-05 [bias_add_comm_swap]: 3.55e-06 [label_micro_interleaved_index]: 6.26e-06 [label_fine_grained_interleaved_index]: 2.59001e-06 [merge_cast_opt]: 1.42e-06 [slice_recompute_activation]: 2.02001e-06 [micro_interleaved_order_control]: 3.02002e-06 [assign_add_opt]: 1.24998e-06 [ForceFp32Comm]: 1.24e-06 [remove_cast_before_assign_add]: 1.15001e-06 [full_micro_interleaved_order_control]: 2.53998e-06 [reorder_send_recv_between_fp_bp]: 2.92002e-06 [comm_op_add_attrs]: 1.32e-06 [add_comm_op_reuse_tag]: 1.05999e-06 [interleave_split_concat_branches]: 1.24e-06 [interleave_parallel_branches]: 1.29998e-06 [overlap_opt_shard_in_pipeline]: 3.06001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.04e-06 [control_data_broadcast_order]: 2.661e-05 [grouped_pairwise_exchange_alltoall]: 1.71e-06 [offloading_packed_experts]: 6.51999e-06 [overlap_recompute_and_grad_model_parallel]: 8.05999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.22999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.67001e-06 [overlap_recompute_comm]: 2.41e-06 [overlap_grad_ring_attention]: 8.04002e-06 [overlap_grad_flash_sp]: 4.045e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.26e-06 [split_layernorm_comm]: 1.98002e-06 [handle_group_info]: 1.10999e-06 [symbol_engine_optimizer]: 0.00016212, [1] [Cycle 1]: 0.00015478, [6] [build]: 1.95e-05 [elim_shapecalc]: 2.158e-05 [elim_not_effective]: 3.006e-05 [opt_reshape]: 2.289e-05 [fold_const_symbol]: 2.416e-05 [renormalize]: 1.19995e-07 [detach_backward]: 2.55002e-06 [pipeline_parallel_scheduler]: 1.45001e-06 [auto_monad_reorder]: 4.595e-05 [get_jit_bprop_graph]: 2.10002e-06 [rewriter_after_jit_bprop_graph]: 7.13998e-06 [opt_after_jit_grad]: 0.0006987 [validate]: 0.00010163 Sums bootstrap : 0.000826s : 0.03% type_inference : 2.765243s : 95.64% event_method : 0.000018s : 0.00% auto_monad : 0.000374s : 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 : 0.000005s : 0.00% parallel-infer-symbol : 0.000006s : 0.00% pre_auto_parallel : 0.000042s : 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 : 0.000094s : 0.00% optimize.opt_a.expand_dump_flag : 0.000007s : 0.00% optimize.opt_a.switch_simplify : 0.000057s : 0.00% optimize.opt_a.loop_unroll : 0.000036s : 0.00% optimize.opt_a.a_1 : 0.001651s : 0.06% optimize.opt_a.with_stream_mark : 0.000061s : 0.00% optimize.opt_a.recompute_prepare : 0.000042s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000021s : 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 : 0.000010s : 0.00% optimize.opt_a.a_2 : 0.000604s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000077s : 0.00% optimize.opt_a.shard : 0.000006s : 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 : 0.000037s : 0.00% optimize.opt_a.auto_parallel : 0.000034s : 0.00% optimize.opt_a.parallel : 0.000051s : 0.00% optimize.opt_a.flash_sp : 0.000022s : 0.00% optimize.opt_a.merge_comm : 0.000018s : 0.00% optimize.opt_a.allreduce_fusion : 0.000019s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000040s : 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 : 0.000036s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000039s : 0.00% optimize.opt_a.virtual_output : 0.000035s : 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 : 0.000037s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000069s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000057s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000020s : 0.00% optimize.opt_a.meta_fg_expand : 0.000013s : 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 : 0.000054s : 0.00% optimize.opt_a.a_after_grad : 0.000062s : 0.00% optimize.opt_a.renormalize : 0.114219s : 3.95% optimize.opt_a.add_forward_monad_depend : 0.000028s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000128s : 0.00% optimize.opt_a.cse : 0.002311s : 0.08% optimize.opt_a.a_3 : 0.000313s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000115s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000318s : 0.01% optimize.convert_after_rewriter : 0.000016s : 0.00% optimize.order_py_execute_after_rewriter : 0.000012s : 0.00% optimize.mutable_eliminate : 0.000903s : 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 : 0.000012s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000086s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000035s : 0.00% optimize.overlap_param_gather : 0.000007s : 0.00% optimize.cconv : 0.000045s : 0.00% optimize.loop_unroll : 0.000570s : 0.02% optimize.opt_after_cconv.c_1 : 0.000100s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000015s : 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 : 0.000052s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000072s : 0.00% optimize.tuple_transform.d_1 : 0.000114s : 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 : 0.000106s : 0.00% optimize.cse_after_recomputation.cse : 0.000045s : 0.00% optimize.environ_conv : 0.000026s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000011s : 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 : 0.000003s : 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 : 0.000003s : 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 : 0.000007s : 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 : 0.000002s : 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 : 0.000020s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000022s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000030s : 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 : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000046s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000699s : 0.02% validate : 0.000102s : 0.00% Time group info: ------[substitution.] 0.000317 96 3.56% : 0.000011s : 2: substitution.depend_value_elim 1.27% : 0.000004s : 8: substitution.elim_not_effective 1.05% : 0.000003s : 8: substitution.fold_const_symbol 3.95% : 0.000013s : 13: substitution.graph_param_transform 44.59% : 0.000141s : 1: substitution.inline 2.73% : 0.000009s : 16: substitution.j_node_and_user_rematch 8.25% : 0.000026s : 2: substitution.less_batch_normalization 1.77% : 0.000006s : 6: substitution.load_eliminater 1.08% : 0.000003s : 2: substitution.opt_reshape 3.78% : 0.000012s : 16: substitution.remove_not_recompute_node 2.81% : 0.000009s : 4: substitution.replace_old_param 10.36% : 0.000033s : 4: substitution.reshape_eliminate 7.41% : 0.000024s : 6: substitution.updatestate_pure_node_eliminater 7.39% : 0.000023s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.765132 2 99.78% : 2.759013s : 1: type_inference.infer 0.22% : 0.006119s : 1: type_inference.specialize ------[replace.] 0.000020 1 100.00% : 0.000020s : 1: replace.inline ------[match.] 0.000140 1 100.00% : 0.000140s : 1: match.inline ------[predicate.] 0.000472 3002 0.94% : 0.000004s : 27: predicate.accumulaten_eliminater 0.92% : 0.000004s : 13: predicate.ad_related_special_op_eliminate 0.77% : 0.000004s : 26: predicate.addn_check_dump 0.92% : 0.000004s : 27: predicate.addn_zero_filter 0.73% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.69% : 0.000013s : 53: predicate.arithmetic_simplify 0.85% : 0.000004s : 27: predicate.cast_eliminate 0.80% : 0.000004s : 26: predicate.check_bprop_eliminate 0.75% : 0.000004s : 26: predicate.compare_switch_simplify 0.24% : 0.000001s : 13: predicate.const_output_eliminate 0.98% : 0.000005s : 26: predicate.depend_value_elim 0.82% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.90% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.76% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.12% : 0.000005s : 26: predicate.dumpgradient_eliminate 0.24% : 0.000001s : 13: predicate.elim_not_effective 0.50% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.19% : 0.000006s : 40: predicate.environ_add_const_eliminate 1.10% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.08% : 0.000005s : 40: predicate.environ_get_depend_swap 1.82% : 0.000009s : 66: predicate.environ_get_eliminate 1.19% : 0.000006s : 40: predicate.environ_get_set_eliminate 0.74% : 0.000003s : 28: predicate.exchange_switch_depend_value 1.25% : 0.000006s : 28: predicate.float_depend_g_call 0.91% : 0.000004s : 26: predicate.float_environ_get_switch 1.08% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.21% : 0.000001s : 13: predicate.fold_const_symbol 0.97% : 0.000005s : 26: predicate.get_grad_eliminate 0.23% : 0.000001s : 13: predicate.graph_param_transform 0.78% : 0.000004s : 26: predicate.incorporate_call 0.67% : 0.000003s : 26: predicate.incorporate_call_switch 4.95% : 0.000023s : 133: predicate.inline 0.94% : 0.000004s : 26: predicate.inline_without_move 0.40% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.21% : 0.000006s : 27: predicate.less_batch_normalization 1.61% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.28% : 0.000011s : 80: predicate.load_eliminater 0.97% : 0.000005s : 13: predicate.loop_unroll_after_grad 0.97% : 0.000005s : 31: predicate.loop_unroll_before_grad 1.72% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.83% : 0.000004s : 26: predicate.merge_addn 0.78% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.78% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.72% : 0.000003s : 27: predicate.minmaximum_grad 1.17% : 0.000006s : 13: predicate.mutable_eliminate 0.50% : 0.000002s : 13: predicate.opt_reshape 0.49% : 0.000002s : 13: predicate.parallel_virtual_node 1.04% : 0.000005s : 28: predicate.partial_defer_inline 1.21% : 0.000006s : 40: predicate.partial_eliminate 0.80% : 0.000004s : 27: predicate.print_const_string_wrapper 0.78% : 0.000004s : 26: predicate.reduce_all_const_elim 1.05% : 0.000005s : 27: predicate.reduce_eliminate 2.40% : 0.000011s : 80: predicate.redundant_stop_gradient_eliminater 0.61% : 0.000003s : 26: predicate.remove_not_recompute_node 1.48% : 0.000007s : 53: predicate.replace_applicator 0.53% : 0.000003s : 26: predicate.replace_old_param 0.39% : 0.000002s : 13: predicate.reset_defer_inline 0.89% : 0.000004s : 27: predicate.reshape_eliminate 1.13% : 0.000005s : 26: predicate.row_tensor_add_zeros_like 0.44% : 0.000002s : 13: predicate.row_tensor_eliminate 1.67% : 0.000008s : 26: predicate.same_eliminate 0.49% : 0.000002s : 26: predicate.set_cell_output_no_recompute 0.93% : 0.000004s : 26: predicate.shard_identity_eliminate 0.86% : 0.000004s : 26: predicate.special_op_eliminate 1.28% : 0.000006s : 26: predicate.specialize_transform 1.29% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.99% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.45% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.84% : 0.000004s : 28: predicate.switch_defer_inline 1.54% : 0.000007s : 54: predicate.switch_layer_defer_inline 4.89% : 0.000023s : 98: predicate.switch_simplify 0.78% : 0.000004s : 27: predicate.tile_eliminate 0.78% : 0.000004s : 27: predicate.transpose_eliminate 1.56% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.61% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.46% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.95% : 0.000014s : 79: predicate.tuple_list_get_item_eliminator 1.69% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.54% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.54% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.22% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.39% : 0.000016s : 106: predicate.updatestate_useless_node_eliminater 0.44% : 0.000002s : 13: predicate.value_based_eliminate 0.86% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.85% : 0.000004s : 26: predicate.virtual_output_eliminate 0.36% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.51% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.196149 58 97.22% : 0.190687s : 55: func_graph_cloner_run.FuncGraphClonerGraph 2.78% : 0.005462s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 3.444542 192 0.00% : 0.000004s : 1: ForceFp32Comm 3.80% : 0.131022s : 1: add_attr 3.80% : 0.131001s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000111s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000386s : 1: auto_monad 0.00% : 0.000052s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.03% : 0.000922s : 1: bootstrap 0.00% : 0.000049s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000030s : 1: control_data_broadcast_order 0.00% : 0.000020s : 1: convert_after_rewriter 0.00% : 0.000059s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000032s : 1: environ_conv 0.00% : 0.000026s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 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% : 0.000006s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000005s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 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 0.03% : 0.000923s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000029s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000038s : 1: opt.transform.mutable_eliminate 0.09% : 0.003020s : 78: opt.transform.opt_a 0.00% : 0.000099s : 1: opt.transform.opt_after_cconv 0.00% : 0.000132s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000366s : 28: opt.transform.opt_b 0.00% : 0.000129s : 2: opt.transform.opt_trans_graph 0.00% : 0.000094s : 4: opt.transform.symbol_engine_opt 4.19% : 0.144482s : 1: opt_a 0.01% : 0.000239s : 1: opt_after_cconv 0.02% : 0.000716s : 1: opt_after_jit_grad 0.02% : 0.000590s : 1: opt_b 4.31% : 0.148498s : 1: optimize 0.00% : 0.000038s : 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% : 0.000013s : 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% : 0.000012s : 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% : 0.000006s : 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% : 0.000007s : 1: pipeline_split 0.00% : 0.000046s : 1: pre_auto_parallel 0.00% : 0.000039s : 1: py_interpret_to_execute 0.00% : 0.000126s : 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 3.14% : 0.108117s : 1: renormalize.infer 0.18% : 0.006074s : 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% : 0.000326s : 1: rewriter_after_opt_a 0.00% : 0.000099s : 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% : 0.000007s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000014s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000165s : 1: symbol_engine_optimizer 0.00% : 0.000166s : 1: tuple_transform 80.28% : 2.765281s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape2-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape2-256],max_mem:212.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:42:03.870.014 [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 = 2.86301, [21] [bootstrap]: 0.00187771 [type_inference]: 2.73047 [event_method]: 2.02e-05 [auto_monad]: 0.00039832 [graph_reusing]: 7.62002e-06 [inline]: 2.70002e-06 [add_attr]: 0.0924924, [1] [add_attr_with_inline]: 0.0924744, [1] [Cycle 1]: 7.318e-05, [2] [tag_attr]: 2.316e-05 [meta_addattr_fg_expand]: 4.22e-06 [parallel-infer-symbol]: 5.01002e-06 [pre_auto_parallel]: 4.08e-05 [insert-virtual-dataset]: 2.91999e-06 [parallel-infer-symbol-second]: 7.10017e-07 [dataset_repeat_opt]: 2.14e-06 [pipeline_split]: 1.61998e-06 [optimize]: 0.0362491, [53] [py_interpret_to_execute]: 3.534e-05 [rewriter_before_opt_a]: 8.728e-05 [opt_a]: 0.0322796, [2] [Cycle 1]: 0.0129089, [45] [expand_dump_flag]: 2.82002e-06 [switch_simplify]: 3.235e-05 [loop_unroll]: 1.902e-05 [a_1]: 0.00089997 [with_stream_mark]: 2.983e-05 [recompute_prepare]: 2.413e-05 [updatestate_depend_eliminate]: 1.133e-05 [updatestate_assign_eliminate]: 7.93999e-06 [updatestate_loads_eliminate]: 1.018e-05 [parameter_eliminate]: 2.58e-06 [a_2]: 0.00025988 [accelerated_algorithm]: 5.033e-05 [shard]: 2.11e-06 [meta_shard_fg_expand]: 3.49001e-06 [shard_inline]: 1.733e-05 [merge_send_recv]: 1.352e-05 [auto_parallel]: 1.405e-05 [parallel]: 3.966e-05 [flash_sp]: 1.416e-05 [merge_comm]: 8.69e-06 [allreduce_fusion]: 7.98001e-06 [matmul_add_comm_reduction]: 1.735e-05 [allreduce_slice_to_reducescatter]: 8.29983e-07 [virtual_shard_identity]: 2.406e-05 [virtual_dataset]: 1.83e-05 [get_grad_eliminate_]: 2.28e-05 [virtual_output]: 1.828e-05 [merge_forward]: 9.64e-06 [cell_reuse_recompute_pass]: 1.97001e-06 [offload_activation]: 1.804e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.527e-05 [merge_recompute_call_nodes]: 1.84e-06 [before_grad]: 4.045e-05 [set_forward_comm_id_for_comm_node_pass]: 8.85001e-06 [meta_fg_expand]: 6.22001e-06 [flash_sp_send_recv_attached]: 6.96001e-06 [receive_attached]: 2.49001e-06 [after_resolve]: 2.659e-05 [a_after_grad]: 3.225e-05 [renormalize]: 0.0104323 [add_forward_monad_depend]: 2.061e-05 [auto_monad_grad]: 2.74001e-06 [auto_monad_eliminator]: 6.006e-05 [cse]: 0.00014299 [a_3]: 0.00015251 [Cycle 2]: 0.0193524, [45] [expand_dump_flag]: 2.63003e-06 [switch_simplify]: 1.867e-05 [loop_unroll]: 0.00014999 [a_1]: 0.0162663 [with_stream_mark]: 4.567e-05 [recompute_prepare]: 3.087e-05 [updatestate_depend_eliminate]: 1.063e-05 [updatestate_assign_eliminate]: 8.23001e-06 [updatestate_loads_eliminate]: 1.185e-05 [parameter_eliminate]: 2.36998e-06 [a_2]: 0.00023112 [accelerated_algorithm]: 2.457e-05 [shard]: 3.61999e-06 [meta_shard_fg_expand]: 7.01001e-06 [shard_inline]: 1.49e-05 [merge_send_recv]: 1.687e-05 [auto_parallel]: 1.513e-05 [parallel]: 1.183e-05 [flash_sp]: 4.68999e-06 [merge_comm]: 8.82e-06 [allreduce_fusion]: 1e-05 [matmul_add_comm_reduction]: 1.894e-05 [allreduce_slice_to_reducescatter]: 9.09989e-07 [virtual_shard_identity]: 2.162e-05 [virtual_dataset]: 1.643e-05 [get_grad_eliminate_]: 1.609e-05 [virtual_output]: 1.6e-05 [merge_forward]: 9.84001e-06 [cell_reuse_recompute_pass]: 3.21001e-06 [offload_activation]: 1.826e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.42e-05 [merge_recompute_call_nodes]: 1.84e-06 [before_grad]: 2.828e-05 [set_forward_comm_id_for_comm_node_pass]: 9.54e-06 [meta_fg_expand]: 6.89001e-06 [flash_sp_send_recv_attached]: 1.97001e-06 [receive_attached]: 2.76999e-06 [after_resolve]: 2.506e-05 [a_after_grad]: 2.847e-05 [renormalize]: 6.00121e-08 [add_forward_monad_depend]: 4.75999e-06 [auto_monad_grad]: 3.36001e-06 [auto_monad_eliminator]: 6.651e-05 [cse]: 0.00135603 [a_3]: 0.00013906 [py_interpret_to_execute_after_opt_a]: 3.548e-05 [slice_cell_reuse_recomputed_activation]: 3.03998e-06 [rewriter_after_opt_a]: 0.00026935 [convert_after_rewriter]: 1.778e-05 [order_py_execute_after_rewriter]: 1.099e-05 [mutable_eliminate]: 0.00088257 [opt_b]: 0.00058716, [1] [Cycle 1]: 0.00057598, [7] [b_1]: 0.00038374 [b_2]: 1.889e-05 [updatestate_depend_eliminate]: 1.842e-05 [updatestate_assign_eliminate]: 8.27e-06 [updatestate_loads_eliminate]: 1.143e-05 [renormalize]: 1.23002e-06 [cse]: 8.484e-05 [optimize_parallel_all_gather_comm]: 3.915e-05 [overlap_param_gather]: 2.34999e-06 [cconv]: 4.406e-05 [loop_unroll]: 0.0006413 [opt_after_cconv]: 0.00024587, [1] [Cycle 1]: 0.00023658, [7] [c_1]: 0.00010374 [parameter_eliminate]: 6.56e-06 [updatestate_depend_eliminate]: 1.553e-05 [updatestate_assign_eliminate]: 7e-06 [updatestate_loads_eliminate]: 8.95999e-06 [cse]: 5.676e-05 [renormalize]: 6.00005e-07 [remove_dup_value]: 7.448e-05 [tuple_transform]: 0.00016279, [1] [Cycle 1]: 0.00015667, [4] [d_1]: 0.00011544 [none_parameter_eliminate]: 2.04999e-06 [renormalize]: 1.70025e-07 [switch_simplify]: 1.673e-05 [partial_unused_args_eliminate]: 1.90001e-06 [add_recomputation]: 0.00011266 [cse_after_recomputation]: 6.016e-05, [1] [Cycle 1]: 5.523e-05, [1] [cse]: 4.833e-05 [environ_conv]: 2.753e-05 [swap_dp_allreduce_reducescatter]: 1.251e-05 [bias_add_comm_swap]: 4.10998e-06 [label_micro_interleaved_index]: 7.28999e-06 [label_fine_grained_interleaved_index]: 2.92002e-06 [merge_cast_opt]: 1.49998e-06 [slice_recompute_activation]: 2.32999e-06 [micro_interleaved_order_control]: 2.51e-06 [assign_add_opt]: 1.25999e-06 [ForceFp32Comm]: 8.89995e-07 [remove_cast_before_assign_add]: 1.43002e-06 [full_micro_interleaved_order_control]: 2.19999e-06 [reorder_send_recv_between_fp_bp]: 2.78e-06 [comm_op_add_attrs]: 1.07998e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.25001e-06 [interleave_parallel_branches]: 1.39e-06 [overlap_opt_shard_in_pipeline]: 1.62001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.13998e-06 [control_data_broadcast_order]: 2.71e-05 [grouped_pairwise_exchange_alltoall]: 1.64998e-06 [offloading_packed_experts]: 6.71e-06 [overlap_recompute_and_grad_model_parallel]: 7.61999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.16002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.44e-06 [overlap_recompute_comm]: 2.36998e-06 [overlap_grad_ring_attention]: 7.74002e-06 [overlap_grad_flash_sp]: 4.238e-05 [begin_end_overlap_inline]: 7.30011e-07 [split_matmul_comm_elemetwise]: 2.29001e-06 [split_layernorm_comm]: 1.77001e-06 [handle_group_info]: 1.86e-06 [symbol_engine_optimizer]: 0.00016493, [1] [Cycle 1]: 0.00015904, [6] [build]: 1.842e-05 [elim_shapecalc]: 2.35e-05 [elim_not_effective]: 3.193e-05 [opt_reshape]: 2.385e-05 [fold_const_symbol]: 2.418e-05 [renormalize]: 2.09984e-07 [detach_backward]: 2.34999e-06 [pipeline_parallel_scheduler]: 1.46998e-06 [auto_monad_reorder]: 4.693e-05 [get_jit_bprop_graph]: 2.19001e-06 [rewriter_after_jit_bprop_graph]: 7.17002e-06 [opt_after_jit_grad]: 0.0007478 [validate]: 0.00010885 Sums bootstrap : 0.001878s : 0.07% type_inference : 2.730466s : 98.62% event_method : 0.000020s : 0.00% auto_monad : 0.000398s : 0.01% graph_reusing : 0.000008s : 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 : 0.000041s : 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 : 0.000035s : 0.00% optimize.rewriter_before_opt_a : 0.000087s : 0.00% optimize.opt_a.expand_dump_flag : 0.000005s : 0.00% optimize.opt_a.switch_simplify : 0.000051s : 0.00% optimize.opt_a.loop_unroll : 0.000169s : 0.01% optimize.opt_a.a_1 : 0.017166s : 0.62% optimize.opt_a.with_stream_mark : 0.000076s : 0.00% optimize.opt_a.recompute_prepare : 0.000055s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000022s : 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 : 0.000491s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000075s : 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 : 0.000051s : 0.00% 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 : 0.000036s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000046s : 0.00% optimize.opt_a.virtual_dataset : 0.000035s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000039s : 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 : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000036s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000069s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000069s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000018s : 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 : 0.000052s : 0.00% optimize.opt_a.a_after_grad : 0.000061s : 0.00% optimize.opt_a.renormalize : 0.010432s : 0.38% optimize.opt_a.add_forward_monad_depend : 0.000025s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000127s : 0.00% optimize.opt_a.cse : 0.001499s : 0.05% optimize.opt_a.a_3 : 0.000292s : 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 : 0.000269s : 0.01% optimize.convert_after_rewriter : 0.000018s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.000883s : 0.03% optimize.opt_b.b_1 : 0.000384s : 0.01% optimize.opt_b.b_2 : 0.000019s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000018s : 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 : 0.000085s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000039s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000044s : 0.00% optimize.loop_unroll : 0.000641s : 0.02% optimize.opt_after_cconv.c_1 : 0.000104s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000016s : 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 : 0.000057s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000074s : 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 : 0.000113s : 0.00% optimize.cse_after_recomputation.cse : 0.000048s : 0.00% optimize.environ_conv : 0.000028s : 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 : 0.000001s : 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 : 0.000002s : 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 : 0.000007s : 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 : 0.000002s : 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 : 0.000032s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000024s : 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 : 0.000001s : 0.00% auto_monad_reorder : 0.000047s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000748s : 0.03% validate : 0.000109s : 0.00% Time group info: ------[substitution.] 0.000313 96 3.50% : 0.000011s : 2: substitution.depend_value_elim 1.18% : 0.000004s : 8: substitution.elim_not_effective 1.07% : 0.000003s : 8: substitution.fold_const_symbol 3.96% : 0.000012s : 13: substitution.graph_param_transform 41.41% : 0.000130s : 1: substitution.inline 2.98% : 0.000009s : 16: substitution.j_node_and_user_rematch 7.82% : 0.000024s : 2: substitution.less_batch_normalization 2.18% : 0.000007s : 6: substitution.load_eliminater 1.09% : 0.000003s : 2: substitution.opt_reshape 3.82% : 0.000012s : 16: substitution.remove_not_recompute_node 2.81% : 0.000009s : 4: substitution.replace_old_param 10.38% : 0.000033s : 4: substitution.reshape_eliminate 10.80% : 0.000034s : 6: substitution.updatestate_pure_node_eliminater 7.00% : 0.000022s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.730347 2 99.81% : 2.725284s : 1: type_inference.infer 0.19% : 0.005063s : 1: type_inference.specialize ------[replace.] 0.000019 1 100.00% : 0.000019s : 1: replace.inline ------[match.] 0.000129 1 100.00% : 0.000129s : 1: match.inline ------[predicate.] 0.000557 3002 1.95% : 0.000011s : 27: predicate.accumulaten_eliminater 1.01% : 0.000006s : 13: predicate.ad_related_special_op_eliminate 0.60% : 0.000003s : 26: predicate.addn_check_dump 0.80% : 0.000004s : 27: predicate.addn_zero_filter 0.61% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.47% : 0.000014s : 53: predicate.arithmetic_simplify 1.42% : 0.000008s : 27: predicate.cast_eliminate 0.69% : 0.000004s : 26: predicate.check_bprop_eliminate 0.59% : 0.000003s : 26: predicate.compare_switch_simplify 0.20% : 0.000001s : 13: predicate.const_output_eliminate 0.66% : 0.000004s : 26: predicate.depend_value_elim 0.72% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.83% : 0.000005s : 27: predicate.dict_get_item_eliminator 1.35% : 0.000008s : 27: predicate.dict_set_item_eliminator 2.09% : 0.000012s : 26: predicate.dumpgradient_eliminate 0.23% : 0.000001s : 13: predicate.elim_not_effective 0.40% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 0.97% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.47% : 0.000008s : 40: predicate.environ_get_add_eliminate 0.91% : 0.000005s : 40: predicate.environ_get_depend_swap 1.97% : 0.000011s : 66: predicate.environ_get_eliminate 0.94% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.68% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.15% : 0.000006s : 28: predicate.float_depend_g_call 0.62% : 0.000003s : 26: predicate.float_environ_get_switch 0.92% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.18% : 0.000001s : 13: predicate.fold_const_symbol 0.78% : 0.000004s : 26: predicate.get_grad_eliminate 0.21% : 0.000001s : 13: predicate.graph_param_transform 0.64% : 0.000004s : 26: predicate.incorporate_call 0.55% : 0.000003s : 26: predicate.incorporate_call_switch 5.34% : 0.000030s : 133: predicate.inline 0.87% : 0.000005s : 26: predicate.inline_without_move 0.33% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.13% : 0.000006s : 27: predicate.less_batch_normalization 1.52% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.03% : 0.000011s : 80: predicate.load_eliminater 0.83% : 0.000005s : 13: predicate.loop_unroll_after_grad 6.06% : 0.000034s : 31: predicate.loop_unroll_before_grad 1.43% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.63% : 0.000003s : 26: predicate.merge_addn 0.63% : 0.000004s : 26: predicate.micro_step_allgather_replace 0.64% : 0.000004s : 26: predicate.mini_step_allgather_replace 1.17% : 0.000007s : 27: predicate.minmaximum_grad 1.23% : 0.000007s : 13: predicate.mutable_eliminate 0.53% : 0.000003s : 13: predicate.opt_reshape 0.36% : 0.000002s : 13: predicate.parallel_virtual_node 1.54% : 0.000009s : 28: predicate.partial_defer_inline 1.00% : 0.000006s : 40: predicate.partial_eliminate 0.68% : 0.000004s : 27: predicate.print_const_string_wrapper 0.64% : 0.000004s : 26: predicate.reduce_all_const_elim 0.94% : 0.000005s : 27: predicate.reduce_eliminate 2.41% : 0.000013s : 80: predicate.redundant_stop_gradient_eliminater 0.42% : 0.000002s : 26: predicate.remove_not_recompute_node 1.14% : 0.000006s : 53: predicate.replace_applicator 0.46% : 0.000003s : 26: predicate.replace_old_param 0.26% : 0.000001s : 13: predicate.reset_defer_inline 0.86% : 0.000005s : 27: predicate.reshape_eliminate 0.69% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.38% : 0.000002s : 13: predicate.row_tensor_eliminate 1.49% : 0.000008s : 26: predicate.same_eliminate 0.48% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.89% : 0.000005s : 26: predicate.shard_identity_eliminate 0.81% : 0.000005s : 26: predicate.special_op_eliminate 0.73% : 0.000004s : 26: predicate.specialize_transform 1.02% : 0.000006s : 26: predicate.split_environ_get_set_with_tuple_value 0.96% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.38% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.72% : 0.000004s : 28: predicate.switch_defer_inline 1.55% : 0.000009s : 54: predicate.switch_layer_defer_inline 3.26% : 0.000018s : 98: predicate.switch_simplify 1.40% : 0.000008s : 27: predicate.tile_eliminate 0.66% : 0.000004s : 27: predicate.transpose_eliminate 1.36% : 0.000008s : 53: predicate.tuple_list_convert_item_index_to_positive 2.03% : 0.000011s : 53: predicate.tuple_list_get_item_const_eliminator 1.46% : 0.000008s : 53: predicate.tuple_list_get_item_depend_reorder 2.73% : 0.000015s : 79: predicate.tuple_list_get_item_eliminator 1.37% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.31% : 0.000013s : 79: predicate.tuple_list_set_item_eliminator 2.01% : 0.000011s : 53: predicate.tuple_to_list_eliminator_ 2.49% : 0.000014s : 80: predicate.updatestate_pure_node_eliminater 2.67% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.34% : 0.000002s : 13: predicate.value_based_eliminate 0.68% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.70% : 0.000004s : 26: predicate.virtual_output_eliminate 0.34% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.41% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.084726 58 96.23% : 0.081536s : 55: func_graph_cloner_run.FuncGraphClonerGraph 3.77% : 0.003190s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 3.021324 192 0.00% : 0.000003s : 1: ForceFp32Comm 3.06% : 0.092500s : 1: add_attr 3.06% : 0.092479s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000120s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000420s : 1: auto_monad 0.00% : 0.000053s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.07% : 0.002198s : 1: bootstrap 0.00% : 0.000049s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000031s : 1: control_data_broadcast_order 0.00% : 0.000024s : 1: convert_after_rewriter 0.00% : 0.000063s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000033s : 1: environ_conv 0.00% : 0.000029s : 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% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 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% : 0.000658s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.03% : 0.000903s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000034s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000043s : 1: opt.transform.mutable_eliminate 0.61% : 0.018563s : 78: opt.transform.opt_a 0.00% : 0.000102s : 1: opt.transform.opt_after_cconv 0.00% : 0.000068s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000366s : 28: opt.transform.opt_b 0.00% : 0.000129s : 2: opt.transform.opt_trans_graph 0.00% : 0.000099s : 4: opt.transform.symbol_engine_opt 1.07% : 0.032284s : 1: opt_a 0.01% : 0.000250s : 1: opt_after_cconv 0.03% : 0.000769s : 1: opt_after_jit_grad 0.02% : 0.000591s : 1: opt_b 1.20% : 0.036257s : 1: optimize 0.00% : 0.000044s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000015s : 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% : 0.000013s : 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% : 0.000005s : 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% : 0.000005s : 1: pipeline_split 0.00% : 0.000045s : 1: pre_auto_parallel 0.00% : 0.000040s : 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% : 0.000080s : 1: remove_dup_value 0.23% : 0.006840s : 1: renormalize.infer 0.12% : 0.003571s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000284s : 1: rewriter_after_opt_a 0.00% : 0.000092s : 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% : 0.000016s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000168s : 1: symbol_engine_optimizer 0.01% : 0.000166s : 1: tuple_transform 90.37% : 2.730507s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape2-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape2-1024],max_mem:260.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:42:57.129.745 [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 = 1.45665, [21] [bootstrap]: 0.00065377 [type_inference]: 1.3252 [event_method]: 1.923e-05 [auto_monad]: 0.00043674 [graph_reusing]: 7.01001e-06 [inline]: 2.53998e-06 [add_attr]: 0.00695073, [1] [add_attr_with_inline]: 0.0069347, [1] [Cycle 1]: 7.091e-05, [2] [tag_attr]: 2.276e-05 [meta_addattr_fg_expand]: 4.32e-06 [parallel-infer-symbol]: 4.81002e-06 [pre_auto_parallel]: 4.076e-05 [insert-virtual-dataset]: 2.59001e-06 [parallel-infer-symbol-second]: 6.59988e-07 [dataset_repeat_opt]: 1.86e-06 [pipeline_split]: 1.74e-06 [optimize]: 0.122216, [53] [py_interpret_to_execute]: 3.267e-05 [rewriter_before_opt_a]: 8.174e-05 [opt_a]: 0.118847, [2] [Cycle 1]: 0.117284, [45] [expand_dump_flag]: 2.71e-06 [switch_simplify]: 3.217e-05 [loop_unroll]: 1.903e-05 [a_1]: 0.00088527 [with_stream_mark]: 2.871e-05 [recompute_prepare]: 2.47e-05 [updatestate_depend_eliminate]: 1.238e-05 [updatestate_assign_eliminate]: 8.05e-06 [updatestate_loads_eliminate]: 1.085e-05 [parameter_eliminate]: 1.85001e-06 [a_2]: 0.00025717 [accelerated_algorithm]: 4.285e-05 [shard]: 2.17999e-06 [meta_shard_fg_expand]: 3.33e-06 [shard_inline]: 1.773e-05 [merge_send_recv]: 1.344e-05 [auto_parallel]: 1.23e-05 [parallel]: 3.487e-05 [flash_sp]: 1.284e-05 [merge_comm]: 8.02998e-06 [allreduce_fusion]: 8.13001e-06 [matmul_add_comm_reduction]: 1.787e-05 [allreduce_slice_to_reducescatter]: 6.60017e-07 [virtual_shard_identity]: 2.129e-05 [virtual_dataset]: 1.688e-05 [get_grad_eliminate_]: 1.87e-05 [virtual_output]: 1.778e-05 [merge_forward]: 8.65999e-06 [cell_reuse_recompute_pass]: 2.22001e-06 [offload_activation]: 1.797e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.228e-05 [merge_recompute_call_nodes]: 1.52999e-06 [before_grad]: 2.971e-05 [set_forward_comm_id_for_comm_node_pass]: 9.61e-06 [meta_fg_expand]: 5.59998e-06 [flash_sp_send_recv_attached]: 7.15e-06 [receive_attached]: 2.49001e-06 [after_resolve]: 2.425e-05 [a_after_grad]: 3.368e-05 [renormalize]: 0.113826 [add_forward_monad_depend]: 2.137e-05 [auto_monad_grad]: 2.69999e-06 [auto_monad_eliminator]: 6.028e-05 [cse]: 0.00118601 [a_3]: 0.00013654 [Cycle 2]: 0.00154847, [45] [expand_dump_flag]: 3.39001e-06 [switch_simplify]: 1.79e-05 [loop_unroll]: 1.626e-05 [a_1]: 0.0004527 [with_stream_mark]: 2.753e-05 [recompute_prepare]: 1.608e-05 [updatestate_depend_eliminate]: 8.95999e-06 [updatestate_assign_eliminate]: 7.85e-06 [updatestate_loads_eliminate]: 1.008e-05 [parameter_eliminate]: 1.82999e-06 [a_2]: 0.0002166 [accelerated_algorithm]: 2.101e-05 [shard]: 2.06998e-06 [meta_shard_fg_expand]: 4.05998e-06 [shard_inline]: 1.481e-05 [merge_send_recv]: 1.361e-05 [auto_parallel]: 1.377e-05 [parallel]: 9.79999e-06 [flash_sp]: 4.24002e-06 [merge_comm]: 8.01001e-06 [allreduce_fusion]: 8.38999e-06 [matmul_add_comm_reduction]: 1.641e-05 [allreduce_slice_to_reducescatter]: 9.79984e-07 [virtual_shard_identity]: 1.634e-05 [virtual_dataset]: 1.491e-05 [get_grad_eliminate_]: 1.525e-05 [virtual_output]: 1.473e-05 [merge_forward]: 8.21002e-06 [cell_reuse_recompute_pass]: 3.76999e-06 [offload_activation]: 1.674e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.619e-05 [merge_recompute_call_nodes]: 1.41002e-06 [before_grad]: 2.435e-05 [set_forward_comm_id_for_comm_node_pass]: 8.24002e-06 [meta_fg_expand]: 5.96e-06 [flash_sp_send_recv_attached]: 1.81998e-06 [receive_attached]: 2.34999e-06 [after_resolve]: 2.122e-05 [a_after_grad]: 2.542e-05 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 2.35002e-06 [auto_monad_grad]: 2.68e-06 [auto_monad_eliminator]: 3.927e-05 [cse]: 8.184e-05 [a_3]: 0.00010096 [py_interpret_to_execute_after_opt_a]: 2.695e-05 [slice_cell_reuse_recomputed_activation]: 2.16e-06 [rewriter_after_opt_a]: 0.00018051 [convert_after_rewriter]: 1.257e-05 [order_py_execute_after_rewriter]: 1.055e-05 [mutable_eliminate]: 0.00076024 [opt_b]: 0.00049868, [1] [Cycle 1]: 0.00049106, [7] [b_1]: 0.00035289 [b_2]: 1.8e-05 [updatestate_depend_eliminate]: 1.218e-05 [updatestate_assign_eliminate]: 6.59001e-06 [updatestate_loads_eliminate]: 8.70999e-06 [renormalize]: 7.30011e-07 [cse]: 5.487e-05 [optimize_parallel_all_gather_comm]: 2.767e-05 [overlap_param_gather]: 2.02001e-06 [cconv]: 3.633e-05 [loop_unroll]: 0.00052395 [opt_after_cconv]: 0.0002219, [1] [Cycle 1]: 0.00021376, [7] [c_1]: 9.654e-05 [parameter_eliminate]: 5.00999e-06 [updatestate_depend_eliminate]: 1.407e-05 [updatestate_assign_eliminate]: 7.16001e-06 [updatestate_loads_eliminate]: 8.75001e-06 [cse]: 4.542e-05 [renormalize]: 4.80009e-07 [remove_dup_value]: 7.082e-05 [tuple_transform]: 0.00015403, [1] [Cycle 1]: 0.00014861, [4] [d_1]: 0.00010984 [none_parameter_eliminate]: 2.17999e-06 [renormalize]: 1.69995e-07 [switch_simplify]: 1.653e-05 [partial_unused_args_eliminate]: 1.98997e-06 [add_recomputation]: 0.00010094 [cse_after_recomputation]: 5.012e-05, [1] [Cycle 1]: 4.566e-05, [1] [cse]: 3.968e-05 [environ_conv]: 2.604e-05 [swap_dp_allreduce_reducescatter]: 1.191e-05 [bias_add_comm_swap]: 2.95002e-06 [label_micro_interleaved_index]: 5.67001e-06 [label_fine_grained_interleaved_index]: 2.84999e-06 [merge_cast_opt]: 1.59e-06 [slice_recompute_activation]: 2.19001e-06 [micro_interleaved_order_control]: 2.47001e-06 [assign_add_opt]: 1.25999e-06 [ForceFp32Comm]: 9.10019e-07 [remove_cast_before_assign_add]: 1.15999e-06 [full_micro_interleaved_order_control]: 2.58998e-06 [reorder_send_recv_between_fp_bp]: 3.31999e-06 [comm_op_add_attrs]: 1.31002e-06 [add_comm_op_reuse_tag]: 1.17999e-06 [interleave_split_concat_branches]: 1.26002e-06 [interleave_parallel_branches]: 1.21002e-06 [overlap_opt_shard_in_pipeline]: 1.34e-06 [overlap_opt_shard_grad_in_pipeline]: 1.97001e-06 [control_data_broadcast_order]: 2.546e-05 [grouped_pairwise_exchange_alltoall]: 2.52001e-06 [offloading_packed_experts]: 6.57002e-06 [overlap_recompute_and_grad_model_parallel]: 8.08999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.44e-06 [overlap_recompute_comm]: 2.68e-06 [overlap_grad_ring_attention]: 7.43999e-06 [overlap_grad_flash_sp]: 3.781e-05 [begin_end_overlap_inline]: 5.29981e-07 [split_matmul_comm_elemetwise]: 2.65997e-06 [split_layernorm_comm]: 1.72001e-06 [handle_group_info]: 1.11002e-06 [symbol_engine_optimizer]: 0.00015293, [1] [Cycle 1]: 0.00014726, [6] [build]: 1.676e-05 [elim_shapecalc]: 2.164e-05 [elim_not_effective]: 2.849e-05 [opt_reshape]: 2.232e-05 [fold_const_symbol]: 2.528e-05 [renormalize]: 2.79979e-07 [detach_backward]: 2.58e-06 [pipeline_parallel_scheduler]: 1.69e-06 [auto_monad_reorder]: 4.502e-05 [get_jit_bprop_graph]: 2.19001e-06 [rewriter_after_jit_bprop_graph]: 6.56e-06 [opt_after_jit_grad]: 0.00071822 [validate]: 9.452e-05 Sums bootstrap : 0.000654s : 0.05% type_inference : 1.325203s : 91.49% event_method : 0.000019s : 0.00% auto_monad : 0.000437s : 0.03% 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 : 0.000041s : 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 : 0.000033s : 0.00% optimize.rewriter_before_opt_a : 0.000082s : 0.01% optimize.opt_a.expand_dump_flag : 0.000006s : 0.00% optimize.opt_a.switch_simplify : 0.000050s : 0.00% optimize.opt_a.loop_unroll : 0.000035s : 0.00% optimize.opt_a.a_1 : 0.001338s : 0.09% optimize.opt_a.with_stream_mark : 0.000056s : 0.00% optimize.opt_a.recompute_prepare : 0.000041s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000021s : 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 : 0.000474s : 0.03% optimize.opt_a.accelerated_algorithm : 0.000064s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000033s : 0.00% optimize.opt_a.merge_send_recv : 0.000027s : 0.00% optimize.opt_a.auto_parallel : 0.000026s : 0.00% optimize.opt_a.parallel : 0.000045s : 0.00% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000016s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000034s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000038s : 0.00% optimize.opt_a.virtual_dataset : 0.000032s : 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 : 0.000017s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000035s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000058s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000054s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000018s : 0.00% optimize.opt_a.meta_fg_expand : 0.000012s : 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 : 0.000045s : 0.00% optimize.opt_a.a_after_grad : 0.000059s : 0.00% optimize.opt_a.renormalize : 0.113826s : 7.86% optimize.opt_a.add_forward_monad_depend : 0.000024s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000100s : 0.01% optimize.opt_a.cse : 0.001268s : 0.09% optimize.opt_a.a_3 : 0.000238s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000027s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000181s : 0.01% optimize.convert_after_rewriter : 0.000013s : 0.00% optimize.order_py_execute_after_rewriter : 0.000011s : 0.00% optimize.mutable_eliminate : 0.000760s : 0.05% optimize.opt_b.b_1 : 0.000353s : 0.02% 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 : 0.000009s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000055s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000028s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.00% optimize.loop_unroll : 0.000524s : 0.04% optimize.opt_after_cconv.c_1 : 0.000097s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000014s : 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 : 0.000045s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000071s : 0.00% optimize.tuple_transform.d_1 : 0.000110s : 0.01% 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 : 0.000101s : 0.01% optimize.cse_after_recomputation.cse : 0.000040s : 0.00% optimize.environ_conv : 0.000026s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000012s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 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 : 0.000002s : 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 : 0.000003s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000025s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000003s : 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 : 0.000007s : 0.00% optimize.overlap_grad_flash_sp : 0.000038s : 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 : 0.000022s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000028s : 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 : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000045s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000718s : 0.05% validate : 0.000095s : 0.01% Time group info: ------[substitution.] 0.000278 96 3.55% : 0.000010s : 2: substitution.depend_value_elim 1.35% : 0.000004s : 8: substitution.elim_not_effective 1.20% : 0.000003s : 8: substitution.fold_const_symbol 4.34% : 0.000012s : 13: substitution.graph_param_transform 43.84% : 0.000122s : 1: substitution.inline 3.23% : 0.000009s : 16: substitution.j_node_and_user_rematch 8.02% : 0.000022s : 2: substitution.less_batch_normalization 2.06% : 0.000006s : 6: substitution.load_eliminater 1.11% : 0.000003s : 2: substitution.opt_reshape 4.31% : 0.000012s : 16: substitution.remove_not_recompute_node 2.66% : 0.000007s : 4: substitution.replace_old_param 12.18% : 0.000034s : 4: substitution.reshape_eliminate 6.47% : 0.000018s : 6: substitution.updatestate_pure_node_eliminater 5.70% : 0.000016s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 1.325092 2 99.60% : 1.319815s : 1: type_inference.infer 0.40% : 0.005277s : 1: type_inference.specialize ------[replace.] 0.000018 1 100.00% : 0.000018s : 1: replace.inline ------[match.] 0.000121 1 100.00% : 0.000121s : 1: match.inline ------[predicate.] 0.000436 3002 0.83% : 0.000004s : 27: predicate.accumulaten_eliminater 1.04% : 0.000005s : 13: predicate.ad_related_special_op_eliminate 0.85% : 0.000004s : 26: predicate.addn_check_dump 0.84% : 0.000004s : 27: predicate.addn_zero_filter 0.79% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.23% : 0.000010s : 53: predicate.arithmetic_simplify 0.82% : 0.000004s : 27: predicate.cast_eliminate 0.80% : 0.000004s : 26: predicate.check_bprop_eliminate 0.83% : 0.000004s : 26: predicate.compare_switch_simplify 0.26% : 0.000001s : 13: predicate.const_output_eliminate 0.83% : 0.000004s : 26: predicate.depend_value_elim 0.80% : 0.000003s : 27: predicate.dict_get_item_const_eliminator 0.95% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.84% : 0.000004s : 27: predicate.dict_set_item_eliminator 1.18% : 0.000005s : 26: predicate.dumpgradient_eliminate 0.25% : 0.000001s : 13: predicate.elim_not_effective 0.49% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.17% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.21% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.15% : 0.000005s : 40: predicate.environ_get_depend_swap 1.96% : 0.000009s : 66: predicate.environ_get_eliminate 1.14% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.82% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.28% : 0.000006s : 28: predicate.float_depend_g_call 0.85% : 0.000004s : 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 0.91% : 0.000004s : 26: predicate.get_grad_eliminate 0.26% : 0.000001s : 13: predicate.graph_param_transform 0.79% : 0.000003s : 26: predicate.incorporate_call 0.71% : 0.000003s : 26: predicate.incorporate_call_switch 5.39% : 0.000024s : 133: predicate.inline 1.12% : 0.000005s : 26: predicate.inline_without_move 0.43% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.24% : 0.000005s : 27: predicate.less_batch_normalization 1.76% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.50% : 0.000011s : 80: predicate.load_eliminater 0.90% : 0.000004s : 13: predicate.loop_unroll_after_grad 1.01% : 0.000004s : 31: predicate.loop_unroll_before_grad 1.77% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.87% : 0.000004s : 26: predicate.merge_addn 0.80% : 0.000003s : 26: predicate.micro_step_allgather_replace 0.77% : 0.000003s : 26: predicate.mini_step_allgather_replace 0.76% : 0.000003s : 27: predicate.minmaximum_grad 0.97% : 0.000004s : 13: predicate.mutable_eliminate 0.51% : 0.000002s : 13: predicate.opt_reshape 0.50% : 0.000002s : 13: predicate.parallel_virtual_node 1.13% : 0.000005s : 28: predicate.partial_defer_inline 1.28% : 0.000006s : 40: predicate.partial_eliminate 0.83% : 0.000004s : 27: predicate.print_const_string_wrapper 0.82% : 0.000004s : 26: predicate.reduce_all_const_elim 1.08% : 0.000005s : 27: predicate.reduce_eliminate 2.39% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.47% : 0.000002s : 26: predicate.remove_not_recompute_node 1.34% : 0.000006s : 53: predicate.replace_applicator 0.48% : 0.000002s : 26: predicate.replace_old_param 0.29% : 0.000001s : 13: predicate.reset_defer_inline 0.99% : 0.000004s : 27: predicate.reshape_eliminate 0.80% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.54% : 0.000002s : 13: predicate.row_tensor_eliminate 1.13% : 0.000005s : 26: predicate.same_eliminate 0.53% : 0.000002s : 26: predicate.set_cell_output_no_recompute 0.99% : 0.000004s : 26: predicate.shard_identity_eliminate 0.90% : 0.000004s : 26: predicate.special_op_eliminate 0.89% : 0.000004s : 26: predicate.specialize_transform 1.08% : 0.000005s : 26: predicate.split_environ_get_set_with_tuple_value 1.04% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.56% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.92% : 0.000004s : 28: predicate.switch_defer_inline 1.72% : 0.000008s : 54: predicate.switch_layer_defer_inline 3.45% : 0.000015s : 98: predicate.switch_simplify 0.86% : 0.000004s : 27: predicate.tile_eliminate 0.80% : 0.000003s : 27: predicate.transpose_eliminate 1.71% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.61% : 0.000007s : 53: predicate.tuple_list_get_item_const_eliminator 1.65% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.91% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.82% : 0.000008s : 53: predicate.tuple_list_get_set_item_eliminator 2.82% : 0.000012s : 79: predicate.tuple_list_set_item_eliminator 1.65% : 0.000007s : 53: predicate.tuple_to_list_eliminator_ 2.43% : 0.000011s : 80: predicate.updatestate_pure_node_eliminater 3.31% : 0.000014s : 106: predicate.updatestate_useless_node_eliminater 0.46% : 0.000002s : 13: predicate.value_based_eliminate 0.86% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.91% : 0.000004s : 26: predicate.virtual_output_eliminate 0.40% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.48% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.088347 58 74.12% : 0.065484s : 55: func_graph_cloner_run.FuncGraphClonerGraph 25.88% : 0.022863s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.702683 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.41% : 0.006958s : 1: add_attr 0.41% : 0.006939s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000105s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000451s : 1: auto_monad 0.00% : 0.000049s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.04% : 0.000707s : 1: bootstrap 0.00% : 0.000040s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000029s : 1: control_data_broadcast_order 0.00% : 0.000016s : 1: convert_after_rewriter 0.00% : 0.000053s : 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 0.00% : 0.000027s : 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% : 0.000006s : 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% : 0.000005s : 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% : 0.000008s : 1: label_micro_interleaved_index 0.03% : 0.000534s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.05% : 0.000770s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000026s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000032s : 1: opt.transform.mutable_eliminate 0.15% : 0.002511s : 78: opt.transform.opt_a 0.01% : 0.000095s : 1: opt.transform.opt_after_cconv 0.00% : 0.000062s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000338s : 28: opt.transform.opt_b 0.01% : 0.000124s : 2: opt.transform.opt_trans_graph 0.01% : 0.000093s : 4: opt.transform.symbol_engine_opt 6.98% : 0.118851s : 1: opt_a 0.01% : 0.000225s : 1: opt_after_cconv 0.04% : 0.000731s : 1: opt_after_jit_grad 0.03% : 0.000502s : 1: opt_b 7.18% : 0.122223s : 1: optimize 0.00% : 0.000031s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000014s : 1: order_py_execute_after_rewriter 0.00% : 0.000041s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000011s : 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% : 0.000006s : 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% : 0.000005s : 1: pipeline_split 0.00% : 0.000045s : 1: pre_auto_parallel 0.00% : 0.000037s : 1: py_interpret_to_execute 0.00% : 0.000030s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000076s : 1: remove_dup_value 5.32% : 0.090528s : 1: renormalize.infer 1.37% : 0.023271s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000187s : 1: rewriter_after_opt_a 0.01% : 0.000087s : 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% : 0.000015s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000156s : 1: symbol_engine_optimizer 0.01% : 0.000157s : 1: tuple_transform 77.83% : 1.325242s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape3-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape3-256],max_mem:260.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:43:40.331.421 [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 = 0.553307, [21] [bootstrap]: 0.00081996 [type_inference]: 0.471388 [event_method]: 1.381e-05 [auto_monad]: 0.0404061 [graph_reusing]: 8.94e-06 [inline]: 2.81e-06 [add_attr]: 0.00728037, [1] [add_attr_with_inline]: 0.00726511, [1] [Cycle 1]: 8.046e-05, [2] [tag_attr]: 2.278e-05 [meta_addattr_fg_expand]: 4.10998e-06 [parallel-infer-symbol]: 5.12e-06 [pre_auto_parallel]: 3.793e-05 [insert-virtual-dataset]: 3.23e-06 [parallel-infer-symbol-second]: 6.19999e-07 [dataset_repeat_opt]: 2.56e-06 [pipeline_split]: 1.61998e-06 [optimize]: 0.0324235, [53] [py_interpret_to_execute]: 3.451e-05 [rewriter_before_opt_a]: 8.493e-05 [opt_a]: 0.0291622, [2] [Cycle 1]: 0.0266483, [45] [expand_dump_flag]: 2.74999e-06 [switch_simplify]: 0.0121807 [loop_unroll]: 3.619e-05 [a_1]: 0.00096784 [with_stream_mark]: 4.242e-05 [recompute_prepare]: 2.935e-05 [updatestate_depend_eliminate]: 1.331e-05 [updatestate_assign_eliminate]: 8.37e-06 [updatestate_loads_eliminate]: 1.583e-05 [parameter_eliminate]: 3.01001e-06 [a_2]: 0.00027693 [accelerated_algorithm]: 5.326e-05 [shard]: 2.66999e-06 [meta_shard_fg_expand]: 6.74001e-06 [shard_inline]: 1.924e-05 [merge_send_recv]: 1.691e-05 [auto_parallel]: 1.592e-05 [parallel]: 4.339e-05 [flash_sp]: 1.732e-05 [merge_comm]: 9.46e-06 [allreduce_fusion]: 7.93999e-06 [matmul_add_comm_reduction]: 1.847e-05 [allreduce_slice_to_reducescatter]: 1.02e-06 [virtual_shard_identity]: 2.896e-05 [virtual_dataset]: 1.905e-05 [get_grad_eliminate_]: 2.027e-05 [virtual_output]: 1.987e-05 [merge_forward]: 1.01e-05 [cell_reuse_recompute_pass]: 2.87002e-06 [offload_activation]: 1.97e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.937e-05 [merge_recompute_call_nodes]: 1.91998e-06 [before_grad]: 3.408e-05 [set_forward_comm_id_for_comm_node_pass]: 1.065e-05 [meta_fg_expand]: 6.61e-06 [flash_sp_send_recv_attached]: 6.19001e-06 [receive_attached]: 2.96999e-06 [after_resolve]: 2.859e-05 [a_after_grad]: 3.837e-05 [renormalize]: 0.011805 [add_forward_monad_depend]: 1.519e-05 [auto_monad_grad]: 3.16001e-06 [auto_monad_eliminator]: 5.147e-05 [cse]: 0.0001629 [a_3]: 0.00012851 [Cycle 2]: 0.00249746, [45] [expand_dump_flag]: 2.24999e-06 [switch_simplify]: 1.723e-05 [loop_unroll]: 1.571e-05 [a_1]: 0.00044567 [with_stream_mark]: 2.177e-05 [recompute_prepare]: 1.537e-05 [updatestate_depend_eliminate]: 9.10001e-06 [updatestate_assign_eliminate]: 7.76001e-06 [updatestate_loads_eliminate]: 1.033e-05 [parameter_eliminate]: 2.06998e-06 [a_2]: 0.00021382 [accelerated_algorithm]: 2.004e-05 [shard]: 2.72001e-06 [meta_shard_fg_expand]: 3.91999e-06 [shard_inline]: 1.464e-05 [merge_send_recv]: 1.356e-05 [auto_parallel]: 1.402e-05 [parallel]: 8.63001e-06 [flash_sp]: 4.15999e-06 [merge_comm]: 7.88999e-06 [allreduce_fusion]: 8.74e-06 [matmul_add_comm_reduction]: 1.452e-05 [allreduce_slice_to_reducescatter]: 8.49977e-07 [virtual_shard_identity]: 1.699e-05 [virtual_dataset]: 1.642e-05 [get_grad_eliminate_]: 1.503e-05 [virtual_output]: 1.503e-05 [merge_forward]: 8.87999e-06 [cell_reuse_recompute_pass]: 3.07002e-06 [offload_activation]: 1.518e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.604e-05 [merge_recompute_call_nodes]: 1.62001e-06 [before_grad]: 2.411e-05 [set_forward_comm_id_for_comm_node_pass]: 7.99002e-06 [meta_fg_expand]: 5.71003e-06 [flash_sp_send_recv_attached]: 2.01e-06 [receive_attached]: 2.75002e-06 [after_resolve]: 2e-05 [a_after_grad]: 2.51e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 1.60999e-06 [auto_monad_grad]: 1.20999e-06 [auto_monad_eliminator]: 2.797e-05 [cse]: 0.0010473 [a_3]: 0.00011133 [py_interpret_to_execute_after_opt_a]: 2.319e-05 [slice_cell_reuse_recomputed_activation]: 2.17999e-06 [rewriter_after_opt_a]: 0.00021408 [convert_after_rewriter]: 1.402e-05 [order_py_execute_after_rewriter]: 1.012e-05 [mutable_eliminate]: 0.00071781 [opt_b]: 0.00048699, [1] [Cycle 1]: 0.00047915, [7] [b_1]: 0.0003445 [b_2]: 1.649e-05 [updatestate_depend_eliminate]: 1.081e-05 [updatestate_assign_eliminate]: 7.03e-06 [updatestate_loads_eliminate]: 8.53001e-06 [renormalize]: 5.89993e-07 [cse]: 5.386e-05 [optimize_parallel_all_gather_comm]: 2.632e-05 [overlap_param_gather]: 5.99e-06 [cconv]: 3.292e-05 [loop_unroll]: 0.00045976 [opt_after_cconv]: 0.00020668, [1] [Cycle 1]: 0.00019975, [7] [c_1]: 9.41e-05 [parameter_eliminate]: 2.84999e-06 [updatestate_depend_eliminate]: 1.113e-05 [updatestate_assign_eliminate]: 7e-06 [updatestate_loads_eliminate]: 7.98999e-06 [cse]: 4.199e-05 [renormalize]: 5.39992e-07 [remove_dup_value]: 6.377e-05 [tuple_transform]: 0.00015041, [1] [Cycle 1]: 0.00014546, [4] [d_1]: 0.00010761 [none_parameter_eliminate]: 1.85001e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 1.591e-05 [partial_unused_args_eliminate]: 2.18002e-06 [add_recomputation]: 0.00010146 [cse_after_recomputation]: 4.88e-05, [1] [Cycle 1]: 4.403e-05, [1] [cse]: 3.823e-05 [environ_conv]: 1.956e-05 [swap_dp_allreduce_reducescatter]: 1.1e-05 [bias_add_comm_swap]: 3.83999e-06 [label_micro_interleaved_index]: 4.71002e-06 [label_fine_grained_interleaved_index]: 2.81999e-06 [merge_cast_opt]: 1.37999e-06 [slice_recompute_activation]: 2.10002e-06 [micro_interleaved_order_control]: 2.31e-06 [assign_add_opt]: 1.48002e-06 [ForceFp32Comm]: 1.00001e-06 [remove_cast_before_assign_add]: 1.07998e-06 [full_micro_interleaved_order_control]: 2.99999e-06 [reorder_send_recv_between_fp_bp]: 3.16001e-06 [comm_op_add_attrs]: 1.32e-06 [add_comm_op_reuse_tag]: 1.07e-06 [interleave_split_concat_branches]: 1.10999e-06 [interleave_parallel_branches]: 1.48002e-06 [overlap_opt_shard_in_pipeline]: 3.24001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.87999e-06 [control_data_broadcast_order]: 2.408e-05 [grouped_pairwise_exchange_alltoall]: 1.43002e-06 [offloading_packed_experts]: 6.49999e-06 [overlap_recompute_and_grad_model_parallel]: 7.39002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.29e-06 [overlap_recompute_comm]: 2.26e-06 [overlap_grad_ring_attention]: 9.03002e-06 [overlap_grad_flash_sp]: 3.543e-05 [begin_end_overlap_inline]: 5.19998e-07 [split_matmul_comm_elemetwise]: 2.22001e-06 [split_layernorm_comm]: 2.02001e-06 [handle_group_info]: 1.10001e-06 [symbol_engine_optimizer]: 0.00014893, [1] [Cycle 1]: 0.00014307, [6] [build]: 1.613e-05 [elim_shapecalc]: 1.947e-05 [elim_not_effective]: 2.911e-05 [opt_reshape]: 2.161e-05 [fold_const_symbol]: 2.495e-05 [renormalize]: 1.69995e-07 [detach_backward]: 2.60997e-06 [pipeline_parallel_scheduler]: 1.44998e-06 [auto_monad_reorder]: 4.131e-05 [get_jit_bprop_graph]: 1.91003e-06 [rewriter_after_jit_bprop_graph]: 4.16001e-06 [opt_after_jit_grad]: 0.00053308 [validate]: 7.659e-05 Sums bootstrap : 0.000820s : 0.15% type_inference : 0.471388s : 86.54% event_method : 0.000014s : 0.00% auto_monad : 0.040406s : 7.42% graph_reusing : 0.000009s : 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 : 0.000038s : 0.01% 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 : 0.000035s : 0.01% optimize.rewriter_before_opt_a : 0.000085s : 0.02% optimize.opt_a.expand_dump_flag : 0.000005s : 0.00% optimize.opt_a.switch_simplify : 0.012198s : 2.24% optimize.opt_a.loop_unroll : 0.000052s : 0.01% optimize.opt_a.a_1 : 0.001414s : 0.26% optimize.opt_a.with_stream_mark : 0.000064s : 0.01% optimize.opt_a.recompute_prepare : 0.000045s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000022s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000016s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000026s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000491s : 0.09% optimize.opt_a.accelerated_algorithm : 0.000073s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000011s : 0.00% optimize.opt_a.shard_inline : 0.000034s : 0.01% optimize.opt_a.merge_send_recv : 0.000030s : 0.01% optimize.opt_a.auto_parallel : 0.000030s : 0.01% optimize.opt_a.parallel : 0.000052s : 0.01% optimize.opt_a.flash_sp : 0.000021s : 0.00% optimize.opt_a.merge_comm : 0.000017s : 0.00% optimize.opt_a.allreduce_fusion : 0.000017s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000033s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000046s : 0.01% optimize.opt_a.virtual_dataset : 0.000035s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000035s : 0.01% optimize.opt_a.virtual_output : 0.000035s : 0.01% 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 : 0.000035s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000065s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000058s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000019s : 0.00% optimize.opt_a.meta_fg_expand : 0.000012s : 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 : 0.000049s : 0.01% optimize.opt_a.a_after_grad : 0.000063s : 0.01% optimize.opt_a.renormalize : 0.011805s : 2.17% optimize.opt_a.add_forward_monad_depend : 0.000017s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000079s : 0.01% optimize.opt_a.cse : 0.001210s : 0.22% optimize.opt_a.a_3 : 0.000240s : 0.04% optimize.py_interpret_to_execute_after_opt_a : 0.000023s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000214s : 0.04% optimize.convert_after_rewriter : 0.000014s : 0.00% optimize.order_py_execute_after_rewriter : 0.000010s : 0.00% optimize.mutable_eliminate : 0.000718s : 0.13% optimize.opt_b.b_1 : 0.000345s : 0.06% optimize.opt_b.b_2 : 0.000016s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000054s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000026s : 0.00% optimize.overlap_param_gather : 0.000006s : 0.00% optimize.cconv : 0.000033s : 0.01% optimize.loop_unroll : 0.000460s : 0.08% optimize.opt_after_cconv.c_1 : 0.000094s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000011s : 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 : 0.000042s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000064s : 0.01% optimize.tuple_transform.d_1 : 0.000108s : 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 : 0.000016s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000101s : 0.02% optimize.cse_after_recomputation.cse : 0.000038s : 0.01% optimize.environ_conv : 0.000020s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000011s : 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 : 0.000001s : 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 : 0.000003s : 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 : 0.000003s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000024s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000006s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 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 : 0.000035s : 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 : 0.000016s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000019s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000029s : 0.01% 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 : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000041s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000533s : 0.10% validate : 0.000077s : 0.01% Time group info: ------[substitution.] 0.000322 96 3.44% : 0.000011s : 2: substitution.depend_value_elim 1.23% : 0.000004s : 8: substitution.elim_not_effective 1.10% : 0.000004s : 8: substitution.fold_const_symbol 3.45% : 0.000011s : 13: substitution.graph_param_transform 49.75% : 0.000160s : 1: substitution.inline 2.97% : 0.000010s : 16: substitution.j_node_and_user_rematch 8.45% : 0.000027s : 2: substitution.less_batch_normalization 1.77% : 0.000006s : 6: substitution.load_eliminater 0.85% : 0.000003s : 2: substitution.opt_reshape 3.73% : 0.000012s : 16: substitution.remove_not_recompute_node 2.47% : 0.000008s : 4: substitution.replace_old_param 9.56% : 0.000031s : 4: substitution.reshape_eliminate 5.43% : 0.000017s : 6: substitution.updatestate_pure_node_eliminater 5.79% : 0.000019s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.471299 2 98.99% : 0.466547s : 1: type_inference.infer 1.01% : 0.004752s : 1: type_inference.specialize ------[replace.] 0.000021 1 100.00% : 0.000021s : 1: replace.inline ------[match.] 0.000159 1 100.00% : 0.000159s : 1: match.inline ------[predicate.] 0.000452 3002 0.84% : 0.000004s : 27: predicate.accumulaten_eliminater 0.71% : 0.000003s : 13: predicate.ad_related_special_op_eliminate 0.73% : 0.000003s : 26: predicate.addn_check_dump 0.92% : 0.000004s : 27: predicate.addn_zero_filter 0.71% : 0.000003s : 27: predicate.adjust_all_reduce_mul_add 2.47% : 0.000011s : 53: predicate.arithmetic_simplify 0.92% : 0.000004s : 27: predicate.cast_eliminate 0.77% : 0.000003s : 26: predicate.check_bprop_eliminate 0.86% : 0.000004s : 26: predicate.compare_switch_simplify 0.25% : 0.000001s : 13: predicate.const_output_eliminate 0.81% : 0.000004s : 26: predicate.depend_value_elim 0.88% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 0.87% : 0.000004s : 27: predicate.dict_get_item_eliminator 0.78% : 0.000004s : 27: predicate.dict_set_item_eliminator 0.97% : 0.000004s : 26: predicate.dumpgradient_eliminate 0.26% : 0.000001s : 13: predicate.elim_not_effective 0.45% : 0.000002s : 13: predicate.elim_shapecalc_of_broadcastargs 1.21% : 0.000005s : 40: predicate.environ_add_const_eliminate 1.11% : 0.000005s : 40: predicate.environ_get_add_eliminate 1.14% : 0.000005s : 40: predicate.environ_get_depend_swap 1.87% : 0.000008s : 66: predicate.environ_get_eliminate 1.12% : 0.000005s : 40: predicate.environ_get_set_eliminate 0.81% : 0.000004s : 28: predicate.exchange_switch_depend_value 1.45% : 0.000007s : 28: predicate.float_depend_g_call 0.75% : 0.000003s : 26: predicate.float_environ_get_switch 1.09% : 0.000005s : 39: predicate.float_tuple_getitem_switch 0.23% : 0.000001s : 13: predicate.fold_const_symbol 0.83% : 0.000004s : 26: predicate.get_grad_eliminate 0.24% : 0.000001s : 13: predicate.graph_param_transform 0.76% : 0.000003s : 26: predicate.incorporate_call 0.67% : 0.000003s : 26: predicate.incorporate_call_switch 5.02% : 0.000023s : 133: predicate.inline 1.10% : 0.000005s : 26: predicate.inline_without_move 0.42% : 0.000002s : 26: predicate.j_node_and_user_rematch 1.46% : 0.000007s : 27: predicate.less_batch_normalization 1.67% : 0.000008s : 53: predicate.list_to_tuple_eliminator_ 2.26% : 0.000010s : 80: predicate.load_eliminater 0.74% : 0.000003s : 13: predicate.loop_unroll_after_grad 1.71% : 0.000008s : 31: predicate.loop_unroll_before_grad 1.75% : 0.000008s : 53: predicate.make_slice_get_slice_eliminator 0.83% : 0.000004s : 26: predicate.merge_addn 0.74% : 0.000003s : 26: predicate.micro_step_allgather_replace 0.84% : 0.000004s : 26: predicate.mini_step_allgather_replace 0.79% : 0.000004s : 27: predicate.minmaximum_grad 0.80% : 0.000004s : 13: predicate.mutable_eliminate 0.48% : 0.000002s : 13: predicate.opt_reshape 0.44% : 0.000002s : 13: predicate.parallel_virtual_node 1.17% : 0.000005s : 28: predicate.partial_defer_inline 1.23% : 0.000006s : 40: predicate.partial_eliminate 0.76% : 0.000003s : 27: predicate.print_const_string_wrapper 0.82% : 0.000004s : 26: predicate.reduce_all_const_elim 1.06% : 0.000005s : 27: predicate.reduce_eliminate 2.30% : 0.000010s : 80: predicate.redundant_stop_gradient_eliminater 0.50% : 0.000002s : 26: predicate.remove_not_recompute_node 1.34% : 0.000006s : 53: predicate.replace_applicator 0.53% : 0.000002s : 26: predicate.replace_old_param 0.29% : 0.000001s : 13: predicate.reset_defer_inline 1.19% : 0.000005s : 27: predicate.reshape_eliminate 0.82% : 0.000004s : 26: predicate.row_tensor_add_zeros_like 0.47% : 0.000002s : 13: predicate.row_tensor_eliminate 1.09% : 0.000005s : 26: predicate.same_eliminate 0.63% : 0.000003s : 26: predicate.set_cell_output_no_recompute 0.98% : 0.000004s : 26: predicate.shard_identity_eliminate 0.87% : 0.000004s : 26: predicate.special_op_eliminate 0.95% : 0.000004s : 26: predicate.specialize_transform 1.11% : 0.000005s : 26: predicate.split_environ_get_set_with_tuple_value 1.03% : 0.000005s : 26: predicate.stack_unstack_eliminate 0.54% : 0.000002s : 13: predicate.switch_call_monad_eliminater 0.89% : 0.000004s : 28: predicate.switch_defer_inline 1.60% : 0.000007s : 54: predicate.switch_layer_defer_inline 5.33% : 0.000024s : 98: predicate.switch_simplify 0.82% : 0.000004s : 27: predicate.tile_eliminate 0.89% : 0.000004s : 27: predicate.transpose_eliminate 1.60% : 0.000007s : 53: predicate.tuple_list_convert_item_index_to_positive 1.68% : 0.000008s : 53: predicate.tuple_list_get_item_const_eliminator 1.48% : 0.000007s : 53: predicate.tuple_list_get_item_depend_reorder 2.81% : 0.000013s : 79: predicate.tuple_list_get_item_eliminator 1.63% : 0.000007s : 53: predicate.tuple_list_get_set_item_eliminator 2.82% : 0.000013s : 79: predicate.tuple_list_set_item_eliminator 1.67% : 0.000008s : 53: predicate.tuple_to_list_eliminator_ 2.18% : 0.000010s : 80: predicate.updatestate_pure_node_eliminater 3.26% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.45% : 0.000002s : 13: predicate.value_based_eliminate 0.93% : 0.000004s : 26: predicate.virtual_dataset_eliminate 0.90% : 0.000004s : 26: predicate.virtual_output_eliminate 0.35% : 0.000002s : 13: predicate.virtual_view_grad_eliminate 0.54% : 0.000002s : 13: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.012775 58 71.87% : 0.009181s : 55: func_graph_cloner_run.FuncGraphClonerGraph 28.13% : 0.003594s : 3: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.620139 192 0.00% : 0.000004s : 1: ForceFp32Comm 1.18% : 0.007287s : 1: add_attr 1.17% : 0.007269s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000106s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 6.52% : 0.040447s : 1: auto_monad 0.01% : 0.000045s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.14% : 0.000879s : 1: bootstrap 0.01% : 0.000036s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000027s : 1: control_data_broadcast_order 0.00% : 0.000018s : 1: convert_after_rewriter 0.01% : 0.000052s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000024s : 1: environ_conv 0.00% : 0.000020s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000016s : 1: graph_reusing 0.00% : 0.000004s : 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 0.08% : 0.000469s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.12% : 0.000727s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000025s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000028s : 1: opt.transform.mutable_eliminate 2.39% : 0.014795s : 78: opt.transform.opt_a 0.01% : 0.000093s : 1: opt.transform.opt_after_cconv 0.01% : 0.000055s : 1: opt.transform.opt_after_jit_grad 0.05% : 0.000331s : 28: opt.transform.opt_b 0.02% : 0.000121s : 2: opt.transform.opt_trans_graph 0.01% : 0.000091s : 4: opt.transform.symbol_engine_opt 4.70% : 0.029166s : 1: opt_a 0.03% : 0.000210s : 1: opt_after_cconv 0.09% : 0.000544s : 1: opt_after_jit_grad 0.08% : 0.000490s : 1: opt_b 5.23% : 0.032430s : 1: optimize 0.00% : 0.000030s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000013s : 1: order_py_execute_after_rewriter 0.01% : 0.000039s : 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% : 0.000006s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000010s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000021s : 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 0.01% : 0.000042s : 1: pre_auto_parallel 0.01% : 0.000039s : 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 0.01% : 0.000068s : 1: remove_dup_value 1.26% : 0.007837s : 1: renormalize.infer 0.64% : 0.003947s : 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% : 0.000220s : 1: rewriter_after_opt_a 0.01% : 0.000091s : 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% : 0.000014s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000152s : 1: symbol_engine_optimizer 0.02% : 0.000153s : 1: tuple_transform 76.02% : 0.471414s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[False-mstype0-k_n_shape3-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[False-mstype0-k_n_shape3-1024],max_mem:252.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:44:13.865.662 [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 = 0.682109, [21] [bootstrap]: 0.00075155 [type_inference]: 0.637741 [event_method]: 0.00020421 [auto_monad]: 0.00070385 [graph_reusing]: 4.944e-05 [inline]: 3.37002e-06 [add_attr]: 0.00513797, [1] [add_attr_with_inline]: 0.0051182, [1] [Cycle 1]: 0.00031113, [2] [tag_attr]: 0.00020861 [meta_addattr_fg_expand]: 5.335e-05 [parallel-infer-symbol]: 3.59002e-06 [pre_auto_parallel]: 0.00025653 [insert-virtual-dataset]: 2.86e-06 [parallel-infer-symbol-second]: 1.24e-06 [dataset_repeat_opt]: 2.65002e-06 [pipeline_split]: 1.60999e-06 [optimize]: 0.0360334, [53] [py_interpret_to_execute]: 0.00025648 [rewriter_before_opt_a]: 0.00104203 [opt_a]: 0.0293283, [2] [Cycle 1]: 0.0271727, [45] [expand_dump_flag]: 2.37e-05 [switch_simplify]: 0.00082517 [loop_unroll]: 0.00032184 [a_1]: 0.00955377 [with_stream_mark]: 5.806e-05 [recompute_prepare]: 4.217e-05 [updatestate_depend_eliminate]: 2.054e-05 [updatestate_assign_eliminate]: 4.956e-05 [updatestate_loads_eliminate]: 1.926e-05 [parameter_eliminate]: 3.25e-06 [a_2]: 0.0007844 [accelerated_algorithm]: 6.205e-05 [shard]: 2.39001e-06 [meta_shard_fg_expand]: 2.363e-05 [shard_inline]: 2.131e-05 [merge_send_recv]: 1.916e-05 [auto_parallel]: 1.884e-05 [parallel]: 3.096e-05 [flash_sp]: 1.487e-05 [merge_comm]: 1.43e-05 [allreduce_fusion]: 1.287e-05 [matmul_add_comm_reduction]: 2.135e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 2.628e-05 [virtual_dataset]: 2.383e-05 [get_grad_eliminate_]: 2.106e-05 [virtual_output]: 2.122e-05 [merge_forward]: 1.228e-05 [cell_reuse_recompute_pass]: 2.40002e-06 [offload_activation]: 2.323e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.441e-05 [merge_recompute_call_nodes]: 1.43002e-06 [before_grad]: 3.8e-05 [set_forward_comm_id_for_comm_node_pass]: 1.292e-05 [meta_fg_expand]: 1.568e-05 [flash_sp_send_recv_attached]: 5.17e-06 [receive_attached]: 2.22999e-06 [after_resolve]: 2.749e-05 [a_after_grad]: 3.455e-05 [renormalize]: 0.0131041 [add_forward_monad_depend]: 1.6e-05 [auto_monad_grad]: 2.74999e-06 [auto_monad_eliminator]: 7.385e-05 [cse]: 0.00113987 [a_3]: 0.00017838 [Cycle 2]: 0.00213691, [45] [expand_dump_flag]: 2.71999e-06 [switch_simplify]: 2.567e-05 [loop_unroll]: 2.086e-05 [a_1]: 0.00071797 [with_stream_mark]: 3.138e-05 [recompute_prepare]: 2.23e-05 [updatestate_depend_eliminate]: 1.5e-05 [updatestate_assign_eliminate]: 1.134e-05 [updatestate_loads_eliminate]: 1.384e-05 [parameter_eliminate]: 2.31e-06 [a_2]: 0.00030542 [accelerated_algorithm]: 2.918e-05 [shard]: 2.71e-06 [meta_shard_fg_expand]: 7.88001e-06 [shard_inline]: 2.024e-05 [merge_send_recv]: 1.872e-05 [auto_parallel]: 2.053e-05 [parallel]: 1.055e-05 [flash_sp]: 5.05001e-06 [merge_comm]: 1.282e-05 [allreduce_fusion]: 1.176e-05 [matmul_add_comm_reduction]: 2.087e-05 [allreduce_slice_to_reducescatter]: 1.03001e-06 [virtual_shard_identity]: 2.328e-05 [virtual_dataset]: 2.072e-05 [get_grad_eliminate_]: 2.067e-05 [virtual_output]: 1.991e-05 [merge_forward]: 1.204e-05 [cell_reuse_recompute_pass]: 3.28e-06 [offload_activation]: 2.315e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.838e-05 [merge_recompute_call_nodes]: 1.70001e-06 [before_grad]: 3.483e-05 [set_forward_comm_id_for_comm_node_pass]: 1.285e-05 [meta_fg_expand]: 9.32001e-06 [flash_sp_send_recv_attached]: 2.14e-06 [receive_attached]: 2.13002e-06 [after_resolve]: 2.634e-05 [a_after_grad]: 3.205e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 3.51001e-06 [auto_monad_grad]: 2.41e-06 [auto_monad_eliminator]: 5.604e-05 [cse]: 8.26e-05 [a_3]: 0.00013773 [py_interpret_to_execute_after_opt_a]: 3.162e-05 [slice_cell_reuse_recomputed_activation]: 2.01998e-06 [rewriter_after_opt_a]: 0.00083508 [convert_after_rewriter]: 2.553e-05 [order_py_execute_after_rewriter]: 1.451e-05 [mutable_eliminate]: 0.00080853 [opt_b]: 0.00079041, [1] [Cycle 1]: 0.00078103, [7] [b_1]: 0.00058114 [b_2]: 2.31e-05 [updatestate_depend_eliminate]: 1.749e-05 [updatestate_assign_eliminate]: 1.099e-05 [updatestate_loads_eliminate]: 1.32e-05 [renormalize]: 1.30999e-06 [cse]: 8.989e-05 [optimize_parallel_all_gather_comm]: 4.193e-05 [overlap_param_gather]: 2.14999e-06 [cconv]: 3.669e-05 [loop_unroll]: 0.00057407 [opt_after_cconv]: 0.00029572, [1] [Cycle 1]: 0.00028615, [7] [c_1]: 0.00013501 [parameter_eliminate]: 4.13001e-06 [updatestate_depend_eliminate]: 1.757e-05 [updatestate_assign_eliminate]: 1.042e-05 [updatestate_loads_eliminate]: 1.216e-05 [cse]: 6.994e-05 [renormalize]: 5.40022e-07 [remove_dup_value]: 9.562e-05 [tuple_transform]: 0.00024272, [1] [Cycle 1]: 0.00023737, [4] [d_1]: 0.00019429 [none_parameter_eliminate]: 2.21e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 2.085e-05 [partial_unused_args_eliminate]: 2.19001e-06 [add_recomputation]: 0.00014807 [cse_after_recomputation]: 7.118e-05, [1] [Cycle 1]: 6.447e-05, [1] [cse]: 5.806e-05 [environ_conv]: 2.981e-05 [swap_dp_allreduce_reducescatter]: 1.962e-05 [bias_add_comm_swap]: 2.86e-06 [label_micro_interleaved_index]: 6.07999e-06 [label_fine_grained_interleaved_index]: 2.59001e-06 [merge_cast_opt]: 0.00027977 [slice_recompute_activation]: 3.26001e-06 [micro_interleaved_order_control]: 4.12e-06 [assign_add_opt]: 1.81e-06 [ForceFp32Comm]: 1.13001e-06 [remove_cast_before_assign_add]: 1.09998e-06 [full_micro_interleaved_order_control]: 2.43e-06 [reorder_send_recv_between_fp_bp]: 3.04001e-06 [comm_op_add_attrs]: 1.10001e-06 [add_comm_op_reuse_tag]: 1.29998e-06 [interleave_split_concat_branches]: 1.32999e-06 [interleave_parallel_branches]: 1.22e-06 [overlap_opt_shard_in_pipeline]: 1.60001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.94999e-06 [control_data_broadcast_order]: 4.166e-05 [grouped_pairwise_exchange_alltoall]: 1.50001e-06 [offloading_packed_experts]: 1.096e-05 [overlap_recompute_and_grad_model_parallel]: 9.96998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.49e-06 [overlap_recompute_allgather_and_fa_grad]: 1.27999e-06 [overlap_recompute_comm]: 2.44001e-06 [overlap_grad_ring_attention]: 8.97999e-06 [overlap_grad_flash_sp]: 5.609e-05 [begin_end_overlap_inline]: 5.8001e-07 [split_matmul_comm_elemetwise]: 2.38998e-06 [split_layernorm_comm]: 1.84e-06 [handle_group_info]: 1.19e-06 [symbol_engine_optimizer]: 0.00055189, [1] [Cycle 1]: 0.00054366, [6] [build]: 0.00031 [elim_shapecalc]: 3.576e-05 [elim_not_effective]: 6.176e-05 [opt_reshape]: 3.223e-05 [fold_const_symbol]: 5.87e-05 [renormalize]: 3.7998e-07 [detach_backward]: 2.99001e-06 [pipeline_parallel_scheduler]: 1.57001e-06 [auto_monad_reorder]: 6.172e-05 [get_jit_bprop_graph]: 1.80001e-06 [rewriter_after_jit_bprop_graph]: 5.92999e-06 [opt_after_jit_grad]: 0.00079002 [validate]: 0.00011233 Sums bootstrap : 0.000752s : 0.11% type_inference : 0.637741s : 94.37% event_method : 0.000204s : 0.03% auto_monad : 0.000704s : 0.10% graph_reusing : 0.000049s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000209s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000053s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000257s : 0.04% 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 : 0.000256s : 0.04% optimize.rewriter_before_opt_a : 0.001042s : 0.15% optimize.opt_a.expand_dump_flag : 0.000026s : 0.00% optimize.opt_a.switch_simplify : 0.000851s : 0.13% optimize.opt_a.loop_unroll : 0.000343s : 0.05% optimize.opt_a.a_1 : 0.010272s : 1.52% optimize.opt_a.with_stream_mark : 0.000089s : 0.01% optimize.opt_a.recompute_prepare : 0.000064s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000036s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000061s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000033s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001090s : 0.16% optimize.opt_a.accelerated_algorithm : 0.000091s : 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 : 0.000042s : 0.01% optimize.opt_a.merge_send_recv : 0.000038s : 0.01% optimize.opt_a.auto_parallel : 0.000039s : 0.01% optimize.opt_a.parallel : 0.000042s : 0.01% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000027s : 0.00% optimize.opt_a.allreduce_fusion : 0.000025s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000042s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000050s : 0.01% optimize.opt_a.virtual_dataset : 0.000045s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.01% optimize.opt_a.virtual_output : 0.000041s : 0.01% optimize.opt_a.merge_forward : 0.000024s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000046s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000083s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000073s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000026s : 0.00% optimize.opt_a.meta_fg_expand : 0.000025s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000054s : 0.01% optimize.opt_a.a_after_grad : 0.000067s : 0.01% optimize.opt_a.renormalize : 0.013104s : 1.94% 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 : 0.000130s : 0.02% optimize.opt_a.cse : 0.001222s : 0.18% optimize.opt_a.a_3 : 0.000316s : 0.05% 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 : 0.000835s : 0.12% optimize.convert_after_rewriter : 0.000026s : 0.00% optimize.order_py_execute_after_rewriter : 0.000015s : 0.00% optimize.mutable_eliminate : 0.000809s : 0.12% optimize.opt_b.b_1 : 0.000581s : 0.09% optimize.opt_b.b_2 : 0.000023s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000017s : 0.00% 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 : 0.000090s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000042s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000037s : 0.01% optimize.loop_unroll : 0.000574s : 0.08% optimize.opt_after_cconv.c_1 : 0.000135s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000018s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.cse : 0.000070s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000096s : 0.01% optimize.tuple_transform.d_1 : 0.000194s : 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 : 0.000021s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000148s : 0.02% optimize.cse_after_recomputation.cse : 0.000058s : 0.01% optimize.environ_conv : 0.000030s : 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 : 0.000006s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000280s : 0.04% optimize.slice_recompute_activation : 0.000003s : 0.00% optimize.micro_interleaved_order_control : 0.000004s : 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 : 0.000002s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000042s : 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 : 0.000056s : 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 : 0.000310s : 0.05% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000036s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000062s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000032s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000059s : 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 : 0.000062s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000790s : 0.12% validate : 0.000112s : 0.02% Time group info: ------[substitution.] 0.003559 569 0.43% : 0.000015s : 8: substitution.depend_value_elim 0.68% : 0.000024s : 14: substitution.elim_not_effective 0.49% : 0.000017s : 12: substitution.float_tuple_getitem_switch 0.77% : 0.000027s : 14: substitution.fold_const_symbol 0.43% : 0.000015s : 17: substitution.graph_param_transform 69.80% : 0.002484s : 71: substitution.inline 0.38% : 0.000014s : 29: substitution.j_node_and_user_rematch 0.93% : 0.000033s : 6: substitution.less_batch_normalization 1.31% : 0.000046s : 2: substitution.list_to_tuple_eliminator_ 0.18% : 0.000006s : 6: substitution.load_eliminater 1.00% : 0.000035s : 21: substitution.minmaximum_grad 0.25% : 0.000009s : 2: substitution.opt_reshape 0.55% : 0.000020s : 29: substitution.remove_not_recompute_node 0.24% : 0.000008s : 4: substitution.replace_old_param 1.87% : 0.000067s : 18: substitution.reshape_eliminate 1.39% : 0.000049s : 24: substitution.switch_simplify 4.18% : 0.000149s : 51: substitution.tuple_list_convert_item_index_to_positive 2.31% : 0.000082s : 51: substitution.tuple_list_get_item_const_eliminator 3.05% : 0.000108s : 51: substitution.tuple_list_get_item_depend_reorder 5.55% : 0.000198s : 73: substitution.tuple_list_get_item_eliminator 2.97% : 0.000106s : 51: substitution.tuple_list_get_set_item_eliminator 0.69% : 0.000024s : 6: substitution.updatestate_pure_node_eliminater 0.55% : 0.000020s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.637302 2 94.31% : 0.601035s : 1: type_inference.infer 5.69% : 0.036267s : 1: type_inference.specialize ------[replace.] 0.001152 115 4.13% : 0.000048s : 5: replace.depend_value_elim 59.15% : 0.000681s : 71: replace.inline 3.53% : 0.000041s : 2: replace.list_to_tuple_eliminator_ 24.59% : 0.000283s : 24: replace.switch_simplify 8.60% : 0.000099s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002551 115 0.08% : 0.000002s : 5: match.depend_value_elim 95.63% : 0.002439s : 71: match.inline 1.73% : 0.000044s : 2: match.list_to_tuple_eliminator_ 1.36% : 0.000035s : 24: match.switch_simplify 1.20% : 0.000031s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001631 9701 1.15% : 0.000019s : 123: predicate.accumulaten_eliminater 0.37% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.48% : 0.000008s : 58: predicate.addn_check_dump 1.14% : 0.000019s : 123: predicate.addn_zero_filter 1.07% : 0.000018s : 123: predicate.adjust_all_reduce_mul_add 2.27% : 0.000037s : 181: predicate.arithmetic_simplify 1.21% : 0.000020s : 123: predicate.cast_eliminate 0.30% : 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 0.58% : 0.000009s : 61: predicate.depend_value_elim 1.20% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 1.30% : 0.000021s : 123: predicate.dict_get_item_eliminator 1.13% : 0.000018s : 123: predicate.dict_set_item_eliminator 2.16% : 0.000035s : 34: predicate.dumpgradient_eliminate 0.11% : 0.000002s : 17: predicate.elim_not_effective 0.20% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.26% : 0.000021s : 140: predicate.environ_add_const_eliminate 1.24% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.25% : 0.000020s : 140: predicate.environ_get_depend_swap 1.82% : 0.000030s : 198: predicate.environ_get_eliminate 1.24% : 0.000020s : 140: predicate.environ_get_set_eliminate 1.91% : 0.000031s : 209: predicate.exchange_switch_depend_value 2.60% : 0.000042s : 209: predicate.float_depend_g_call 0.51% : 0.000008s : 58: predicate.float_environ_get_switch 0.73% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.34% : 0.000005s : 35: predicate.get_grad_eliminate 0.09% : 0.000001s : 17: predicate.graph_param_transform 0.48% : 0.000008s : 58: predicate.incorporate_call 0.46% : 0.000007s : 58: predicate.incorporate_call_switch 5.67% : 0.000093s : 459: predicate.inline 0.42% : 0.000007s : 35: predicate.inline_without_move 0.15% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.45% : 0.000007s : 36: predicate.less_batch_normalization 1.80% : 0.000029s : 172: predicate.list_to_tuple_eliminator_ 2.71% : 0.000044s : 295: predicate.load_eliminater 0.31% : 0.000005s : 17: predicate.loop_unroll_after_grad 3.31% : 0.000054s : 314: predicate.loop_unroll_before_grad 1.60% : 0.000026s : 157: predicate.make_slice_get_slice_eliminator 0.50% : 0.000008s : 58: predicate.merge_addn 0.33% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.31% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.18% : 0.000019s : 123: predicate.minmaximum_grad 0.41% : 0.000007s : 17: predicate.mutable_eliminate 0.17% : 0.000003s : 17: predicate.opt_reshape 0.18% : 0.000003s : 17: predicate.parallel_virtual_node 3.41% : 0.000056s : 209: predicate.partial_defer_inline 1.52% : 0.000025s : 155: predicate.partial_eliminate 1.19% : 0.000019s : 123: predicate.print_const_string_wrapper 0.48% : 0.000008s : 53: predicate.reduce_all_const_elim 1.59% : 0.000026s : 123: predicate.reduce_eliminate 2.76% : 0.000045s : 295: predicate.redundant_stop_gradient_eliminater 0.16% : 0.000003s : 35: predicate.remove_not_recompute_node 1.06% : 0.000017s : 172: predicate.replace_applicator 0.19% : 0.000003s : 35: predicate.replace_old_param 0.09% : 0.000002s : 17: predicate.reset_defer_inline 1.31% : 0.000021s : 123: predicate.reshape_eliminate 0.35% : 0.000006s : 34: predicate.row_tensor_add_zeros_like 0.20% : 0.000003s : 17: predicate.row_tensor_eliminate 0.45% : 0.000007s : 34: predicate.same_eliminate 0.21% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.33% : 0.000005s : 35: predicate.shard_identity_eliminate 0.33% : 0.000005s : 34: predicate.special_op_eliminate 0.60% : 0.000010s : 58: predicate.specialize_transform 0.39% : 0.000006s : 34: predicate.split_environ_get_set_with_tuple_value 0.35% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.19% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.12% : 0.000035s : 209: predicate.switch_defer_inline 2.39% : 0.000039s : 243: predicate.switch_layer_defer_inline 6.74% : 0.000110s : 646: predicate.switch_simplify 1.18% : 0.000019s : 123: predicate.tile_eliminate 1.10% : 0.000018s : 123: predicate.transpose_eliminate 1.81% : 0.000029s : 157: predicate.tuple_list_convert_item_index_to_positive 1.78% : 0.000029s : 157: predicate.tuple_list_get_item_const_eliminator 1.62% : 0.000026s : 157: predicate.tuple_list_get_item_depend_reorder 2.92% : 0.000048s : 228: predicate.tuple_list_get_item_eliminator 1.73% : 0.000028s : 157: predicate.tuple_list_get_set_item_eliminator 2.22% : 0.000036s : 215: predicate.tuple_list_set_item_eliminator 1.61% : 0.000026s : 170: predicate.tuple_to_list_eliminator_ 2.58% : 0.000042s : 295: predicate.updatestate_pure_node_eliminater 3.12% : 0.000051s : 353: predicate.updatestate_useless_node_eliminater 0.17% : 0.000003s : 17: predicate.value_based_eliminate 0.32% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.33% : 0.000005s : 35: predicate.virtual_output_eliminate 0.15% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.18% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.036579 135 79.12% : 0.028943s : 55: func_graph_cloner_run.FuncGraphClonerGraph 20.88% : 0.007637s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.750750 209 0.00% : 0.000004s : 1: ForceFp32Comm 0.69% : 0.005145s : 1: add_attr 0.68% : 0.005124s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.02% : 0.000154s : 1: add_recomputation 0.00% : 0.000006s : 1: assign_add_opt 0.10% : 0.000718s : 1: auto_monad 0.01% : 0.000066s : 1: auto_monad_reorder 0.00% : 0.000014s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.11% : 0.000808s : 1: bootstrap 0.01% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000045s : 1: control_data_broadcast_order 0.00% : 0.000031s : 1: convert_after_rewriter 0.01% : 0.000074s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000034s : 1: environ_conv 0.03% : 0.000215s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000055s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000006s : 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% : 0.000009s : 1: label_micro_interleaved_index 0.08% : 0.000585s : 1: loop_unroll 0.04% : 0.000293s : 1: merge_cast_opt 0.00% : 0.000007s : 1: micro_interleaved_order_control 0.11% : 0.000822s : 1: mutable_eliminate 0.00% : 0.000014s : 1: offloading_packed_experts 0.00% : 0.000034s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000045s : 1: opt.transform.mutable_eliminate 1.78% : 0.013379s : 95: opt.transform.opt_a 0.02% : 0.000134s : 1: opt.transform.opt_after_cconv 0.01% : 0.000074s : 1: opt.transform.opt_after_jit_grad 0.08% : 0.000566s : 28: opt.transform.opt_b 0.03% : 0.000213s : 2: opt.transform.opt_trans_graph 0.02% : 0.000183s : 4: opt.transform.symbol_engine_opt 3.91% : 0.029334s : 1: opt_a 0.04% : 0.000299s : 1: opt_after_cconv 0.11% : 0.000803s : 1: opt_after_jit_grad 0.11% : 0.000794s : 1: opt_b 4.80% : 0.036041s : 1: optimize 0.01% : 0.000046s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000019s : 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% : 0.000012s : 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% : 0.000007s : 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% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.04% : 0.000266s : 1: pre_auto_parallel 0.04% : 0.000266s : 1: py_interpret_to_execute 0.00% : 0.000036s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000100s : 1: remove_dup_value 0.80% : 0.006006s : 1: renormalize.infer 0.94% : 0.007079s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.11% : 0.000849s : 1: rewriter_after_opt_a 0.14% : 0.001061s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000007s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000023s : 1: swap_dp_allreduce_reducescatter 0.07% : 0.000556s : 1: symbol_engine_optimizer 0.03% : 0.000247s : 1: tuple_transform 84.95% : 0.637770s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape0-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape0-256],max_mem:252.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:44:21.434.381 [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 = 0.347993, [21] [bootstrap]: 0.00107273 [type_inference]: 0.3029 [event_method]: 0.00021041 [auto_monad]: 0.00092026 [graph_reusing]: 4.968e-05 [inline]: 3.43e-06 [add_attr]: 0.00547483, [1] [add_attr_with_inline]: 0.00545575, [1] [Cycle 1]: 0.00030536, [2] [tag_attr]: 0.0001948 [meta_addattr_fg_expand]: 5.486e-05 [parallel-infer-symbol]: 4.07998e-06 [pre_auto_parallel]: 0.00025677 [insert-virtual-dataset]: 3.71999e-06 [parallel-infer-symbol-second]: 1.25999e-06 [dataset_repeat_opt]: 2.41e-06 [pipeline_split]: 1.62001e-06 [optimize]: 0.0357417, [53] [py_interpret_to_execute]: 0.00023149 [rewriter_before_opt_a]: 0.0007769 [opt_a]: 0.0293449, [2] [Cycle 1]: 0.0270186, [45] [expand_dump_flag]: 1.726e-05 [switch_simplify]: 0.00082411 [loop_unroll]: 0.00033008 [a_1]: 0.010264 [with_stream_mark]: 5.814e-05 [recompute_prepare]: 4.246e-05 [updatestate_depend_eliminate]: 2.139e-05 [updatestate_assign_eliminate]: 1.576e-05 [updatestate_loads_eliminate]: 1.936e-05 [parameter_eliminate]: 2.89999e-06 [a_2]: 0.0008386 [accelerated_algorithm]: 6.921e-05 [shard]: 2.60002e-06 [meta_shard_fg_expand]: 2.485e-05 [shard_inline]: 2.549e-05 [merge_send_recv]: 1.989e-05 [auto_parallel]: 2.12e-05 [parallel]: 3.97e-05 [flash_sp]: 1.62e-05 [merge_comm]: 1.356e-05 [allreduce_fusion]: 1.241e-05 [matmul_add_comm_reduction]: 2.236e-05 [allreduce_slice_to_reducescatter]: 8.80013e-07 [virtual_shard_identity]: 3.008e-05 [virtual_dataset]: 2.56e-05 [get_grad_eliminate_]: 2.141e-05 [virtual_output]: 2.093e-05 [merge_forward]: 1.213e-05 [cell_reuse_recompute_pass]: 1.94999e-06 [offload_activation]: 2.311e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.403e-05 [merge_recompute_call_nodes]: 1.84e-06 [before_grad]: 3.888e-05 [set_forward_comm_id_for_comm_node_pass]: 1.328e-05 [meta_fg_expand]: 1.654e-05 [flash_sp_send_recv_attached]: 5.17999e-06 [receive_attached]: 2.17001e-06 [after_resolve]: 3.008e-05 [a_after_grad]: 3.935e-05 [renormalize]: 0.0122205 [add_forward_monad_depend]: 2.182e-05 [auto_monad_grad]: 3.45998e-06 [auto_monad_eliminator]: 8.323e-05 [cse]: 0.00104461 [a_3]: 0.00017527 [Cycle 2]: 0.00230901, [45] [expand_dump_flag]: 3.21999e-06 [switch_simplify]: 2.532e-05 [loop_unroll]: 2.115e-05 [a_1]: 0.00066394 [with_stream_mark]: 3.305e-05 [recompute_prepare]: 2.042e-05 [updatestate_depend_eliminate]: 1.436e-05 [updatestate_assign_eliminate]: 1.149e-05 [updatestate_loads_eliminate]: 1.407e-05 [parameter_eliminate]: 2.56998e-06 [a_2]: 0.00050365 [accelerated_algorithm]: 3.264e-05 [shard]: 2.36e-06 [meta_shard_fg_expand]: 6.68e-06 [shard_inline]: 2.118e-05 [merge_send_recv]: 2.105e-05 [auto_parallel]: 2.095e-05 [parallel]: 1.016e-05 [flash_sp]: 4.84e-06 [merge_comm]: 1.279e-05 [allreduce_fusion]: 1.922e-05 [matmul_add_comm_reduction]: 2.305e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 2.196e-05 [virtual_dataset]: 2.112e-05 [get_grad_eliminate_]: 2.105e-05 [virtual_output]: 1.92e-05 [merge_forward]: 1.267e-05 [cell_reuse_recompute_pass]: 3.46999e-06 [offload_activation]: 2.528e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.017e-05 [merge_recompute_call_nodes]: 1.44e-06 [before_grad]: 3.555e-05 [set_forward_comm_id_for_comm_node_pass]: 1.343e-05 [meta_fg_expand]: 9.39998e-06 [flash_sp_send_recv_attached]: 1.91e-06 [receive_attached]: 2.71e-06 [after_resolve]: 2.906e-05 [a_after_grad]: 3.325e-05 [renormalize]: 9.00181e-08 [add_forward_monad_depend]: 3.93001e-06 [auto_monad_grad]: 2.36e-06 [auto_monad_eliminator]: 5.759e-05 [cse]: 8.606e-05 [a_3]: 0.00013743 [py_interpret_to_execute_after_opt_a]: 3.53e-05 [slice_cell_reuse_recomputed_activation]: 3.03998e-06 [rewriter_after_opt_a]: 0.00090955 [convert_after_rewriter]: 2.577e-05 [order_py_execute_after_rewriter]: 1.5e-05 [mutable_eliminate]: 0.0008708 [opt_b]: 0.00077296, [1] [Cycle 1]: 0.00076192, [7] [b_1]: 0.00054061 [b_2]: 2.442e-05 [updatestate_depend_eliminate]: 2e-05 [updatestate_assign_eliminate]: 1.151e-05 [updatestate_loads_eliminate]: 1.511e-05 [renormalize]: 1.34e-06 [cse]: 0.00010372 [optimize_parallel_all_gather_comm]: 4.116e-05 [overlap_param_gather]: 2.21e-06 [cconv]: 4.156e-05 [loop_unroll]: 0.00065289 [opt_after_cconv]: 0.00031362, [1] [Cycle 1]: 0.00030384, [7] [c_1]: 0.00013079 [parameter_eliminate]: 6.01998e-06 [updatestate_depend_eliminate]: 1.804e-05 [updatestate_assign_eliminate]: 1.089e-05 [updatestate_loads_eliminate]: 1.29e-05 [cse]: 8.405e-05 [renormalize]: 8.00006e-07 [remove_dup_value]: 9.934e-05 [tuple_transform]: 0.00026865, [1] [Cycle 1]: 0.0002623, [4] [d_1]: 0.00020694 [none_parameter_eliminate]: 3.06999e-06 [renormalize]: 3.80009e-07 [switch_simplify]: 2.254e-05 [partial_unused_args_eliminate]: 2.13002e-06 [add_recomputation]: 0.00014429 [cse_after_recomputation]: 8.137e-05, [1] [Cycle 1]: 7.397e-05, [1] [cse]: 6.409e-05 [environ_conv]: 3.628e-05 [swap_dp_allreduce_reducescatter]: 1.993e-05 [bias_add_comm_swap]: 3.47997e-06 [label_micro_interleaved_index]: 7.19001e-06 [label_fine_grained_interleaved_index]: 3.05002e-06 [merge_cast_opt]: 1.54998e-06 [slice_recompute_activation]: 2.24999e-06 [micro_interleaved_order_control]: 2.45997e-06 [assign_add_opt]: 1.45999e-06 [ForceFp32Comm]: 1.12999e-06 [remove_cast_before_assign_add]: 1.37999e-06 [full_micro_interleaved_order_control]: 2.48998e-06 [reorder_send_recv_between_fp_bp]: 3.14999e-06 [comm_op_add_attrs]: 1.09003e-06 [add_comm_op_reuse_tag]: 1.14998e-06 [interleave_split_concat_branches]: 1.37e-06 [interleave_parallel_branches]: 1.22999e-06 [overlap_opt_shard_in_pipeline]: 1.54e-06 [overlap_opt_shard_grad_in_pipeline]: 1.93997e-06 [control_data_broadcast_order]: 3.906e-05 [grouped_pairwise_exchange_alltoall]: 1.82999e-06 [offloading_packed_experts]: 1.016e-05 [overlap_recompute_and_grad_model_parallel]: 1.083e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.51002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.49998e-06 [overlap_recompute_comm]: 2.44999e-06 [overlap_grad_ring_attention]: 9.90002e-06 [overlap_grad_flash_sp]: 5.226e-05 [begin_end_overlap_inline]: 7.09988e-07 [split_matmul_comm_elemetwise]: 2.66e-06 [split_layernorm_comm]: 1.69998e-06 [handle_group_info]: 1.27999e-06 [symbol_engine_optimizer]: 0.00054896, [1] [Cycle 1]: 0.00054206, [6] [build]: 0.00029927 [elim_shapecalc]: 3.979e-05 [elim_not_effective]: 5.301e-05 [opt_reshape]: 3.354e-05 [fold_const_symbol]: 6.033e-05 [renormalize]: 5.00004e-07 [detach_backward]: 3.21001e-06 [pipeline_parallel_scheduler]: 1.86e-06 [auto_monad_reorder]: 6.743e-05 [get_jit_bprop_graph]: 2.49999e-06 [rewriter_after_jit_bprop_graph]: 7.2e-06 [opt_after_jit_grad]: 0.00086178 [validate]: 0.00012816 Sums bootstrap : 0.001073s : 0.31% type_inference : 0.302900s : 88.76% event_method : 0.000210s : 0.06% auto_monad : 0.000920s : 0.27% graph_reusing : 0.000050s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000195s : 0.06% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000055s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000257s : 0.08% insert-virtual-dataset : 0.000004s : 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 : 0.000231s : 0.07% optimize.rewriter_before_opt_a : 0.000777s : 0.23% optimize.opt_a.expand_dump_flag : 0.000020s : 0.01% optimize.opt_a.switch_simplify : 0.000849s : 0.25% optimize.opt_a.loop_unroll : 0.000351s : 0.10% optimize.opt_a.a_1 : 0.010928s : 3.20% optimize.opt_a.with_stream_mark : 0.000091s : 0.03% optimize.opt_a.recompute_prepare : 0.000063s : 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 : 0.000033s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.001342s : 0.39% optimize.opt_a.accelerated_algorithm : 0.000102s : 0.03% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000032s : 0.01% optimize.opt_a.shard_inline : 0.000047s : 0.01% optimize.opt_a.merge_send_recv : 0.000041s : 0.01% optimize.opt_a.auto_parallel : 0.000042s : 0.01% optimize.opt_a.parallel : 0.000050s : 0.01% optimize.opt_a.flash_sp : 0.000021s : 0.01% optimize.opt_a.merge_comm : 0.000026s : 0.01% optimize.opt_a.allreduce_fusion : 0.000032s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000045s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000052s : 0.02% optimize.opt_a.virtual_dataset : 0.000047s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.01% optimize.opt_a.virtual_output : 0.000040s : 0.01% optimize.opt_a.merge_forward : 0.000025s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000048s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000084s : 0.02% 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 : 0.000027s : 0.01% optimize.opt_a.meta_fg_expand : 0.000026s : 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 : 0.000059s : 0.02% optimize.opt_a.a_after_grad : 0.000073s : 0.02% optimize.opt_a.renormalize : 0.012221s : 3.58% optimize.opt_a.add_forward_monad_depend : 0.000026s : 0.01% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000141s : 0.04% optimize.opt_a.cse : 0.001131s : 0.33% optimize.opt_a.a_3 : 0.000313s : 0.09% optimize.py_interpret_to_execute_after_opt_a : 0.000035s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000910s : 0.27% optimize.convert_after_rewriter : 0.000026s : 0.01% optimize.order_py_execute_after_rewriter : 0.000015s : 0.00% optimize.mutable_eliminate : 0.000871s : 0.26% optimize.opt_b.b_1 : 0.000541s : 0.16% optimize.opt_b.b_2 : 0.000024s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000020s : 0.01% 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 : 0.000104s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000041s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000042s : 0.01% optimize.loop_unroll : 0.000653s : 0.19% optimize.opt_after_cconv.c_1 : 0.000131s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000018s : 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 : 0.000084s : 0.02% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000099s : 0.03% optimize.tuple_transform.d_1 : 0.000207s : 0.06% 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 : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000144s : 0.04% optimize.cse_after_recomputation.cse : 0.000064s : 0.02% optimize.environ_conv : 0.000036s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000020s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 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 : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000002s : 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 : 0.000010s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000011s : 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 : 0.000052s : 0.02% 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 : 0.000299s : 0.09% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000040s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000053s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000034s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000060s : 0.02% 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 : 0.000067s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000862s : 0.25% validate : 0.000128s : 0.04% Time group info: ------[substitution.] 0.004119 569 0.42% : 0.000017s : 8: substitution.depend_value_elim 0.34% : 0.000014s : 14: substitution.elim_not_effective 0.43% : 0.000018s : 12: substitution.float_tuple_getitem_switch 0.72% : 0.000029s : 14: substitution.fold_const_symbol 0.36% : 0.000015s : 17: substitution.graph_param_transform 72.84% : 0.003000s : 71: substitution.inline 0.35% : 0.000015s : 29: substitution.j_node_and_user_rematch 0.88% : 0.000036s : 6: substitution.less_batch_normalization 1.23% : 0.000051s : 2: substitution.list_to_tuple_eliminator_ 0.16% : 0.000007s : 6: substitution.load_eliminater 0.79% : 0.000033s : 21: substitution.minmaximum_grad 0.21% : 0.000008s : 2: substitution.opt_reshape 0.48% : 0.000020s : 29: substitution.remove_not_recompute_node 0.22% : 0.000009s : 4: substitution.replace_old_param 1.65% : 0.000068s : 18: substitution.reshape_eliminate 1.14% : 0.000047s : 24: substitution.switch_simplify 3.74% : 0.000154s : 51: substitution.tuple_list_convert_item_index_to_positive 1.92% : 0.000079s : 51: substitution.tuple_list_get_item_const_eliminator 2.92% : 0.000120s : 51: substitution.tuple_list_get_item_depend_reorder 5.58% : 0.000230s : 73: substitution.tuple_list_get_item_eliminator 2.52% : 0.000104s : 51: substitution.tuple_list_get_set_item_eliminator 0.56% : 0.000023s : 6: substitution.updatestate_pure_node_eliminater 0.54% : 0.000022s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.302434 2 92.76% : 0.280524s : 1: type_inference.infer 7.24% : 0.021911s : 1: type_inference.specialize ------[replace.] 0.001189 115 5.79% : 0.000069s : 5: replace.depend_value_elim 57.98% : 0.000690s : 71: replace.inline 3.38% : 0.000040s : 2: replace.list_to_tuple_eliminator_ 24.00% : 0.000285s : 24: replace.switch_simplify 8.84% : 0.000105s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.003066 115 0.07% : 0.000002s : 5: match.depend_value_elim 96.15% : 0.002948s : 71: match.inline 1.58% : 0.000048s : 2: match.list_to_tuple_eliminator_ 1.08% : 0.000033s : 24: match.switch_simplify 1.12% : 0.000034s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001609 9701 1.31% : 0.000021s : 123: predicate.accumulaten_eliminater 0.41% : 0.000007s : 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 1.12% : 0.000018s : 123: predicate.adjust_all_reduce_mul_add 2.35% : 0.000038s : 181: predicate.arithmetic_simplify 1.19% : 0.000019s : 123: predicate.cast_eliminate 0.33% : 0.000005s : 34: predicate.check_bprop_eliminate 0.51% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.71% : 0.000011s : 61: predicate.depend_value_elim 1.28% : 0.000021s : 123: predicate.dict_get_item_const_eliminator 1.30% : 0.000021s : 123: predicate.dict_get_item_eliminator 1.18% : 0.000019s : 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 1.25% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.28% : 0.000021s : 140: predicate.environ_get_add_eliminate 1.24% : 0.000020s : 140: predicate.environ_get_depend_swap 1.96% : 0.000032s : 198: predicate.environ_get_eliminate 1.25% : 0.000020s : 140: predicate.environ_get_set_eliminate 1.94% : 0.000031s : 209: predicate.exchange_switch_depend_value 2.67% : 0.000043s : 209: predicate.float_depend_g_call 0.51% : 0.000008s : 58: predicate.float_environ_get_switch 0.71% : 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 0.11% : 0.000002s : 17: predicate.graph_param_transform 0.53% : 0.000009s : 58: predicate.incorporate_call 0.47% : 0.000008s : 58: predicate.incorporate_call_switch 5.78% : 0.000093s : 459: predicate.inline 0.43% : 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_ 2.68% : 0.000043s : 295: predicate.load_eliminater 0.39% : 0.000006s : 17: predicate.loop_unroll_after_grad 3.17% : 0.000051s : 314: predicate.loop_unroll_before_grad 1.59% : 0.000026s : 157: predicate.make_slice_get_slice_eliminator 0.52% : 0.000008s : 58: predicate.merge_addn 0.30% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.31% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.20% : 0.000019s : 123: predicate.minmaximum_grad 0.38% : 0.000006s : 17: predicate.mutable_eliminate 0.17% : 0.000003s : 17: predicate.opt_reshape 0.18% : 0.000003s : 17: predicate.parallel_virtual_node 3.25% : 0.000052s : 209: predicate.partial_defer_inline 1.54% : 0.000025s : 155: predicate.partial_eliminate 1.16% : 0.000019s : 123: predicate.print_const_string_wrapper 0.49% : 0.000008s : 53: predicate.reduce_all_const_elim 1.62% : 0.000026s : 123: predicate.reduce_eliminate 2.65% : 0.000043s : 295: predicate.redundant_stop_gradient_eliminater 0.19% : 0.000003s : 35: predicate.remove_not_recompute_node 1.10% : 0.000018s : 172: predicate.replace_applicator 0.19% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.32% : 0.000021s : 123: predicate.reshape_eliminate 0.30% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.22% : 0.000004s : 17: predicate.row_tensor_eliminate 0.43% : 0.000007s : 34: predicate.same_eliminate 0.22% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.42% : 0.000007s : 35: predicate.shard_identity_eliminate 0.39% : 0.000006s : 34: predicate.special_op_eliminate 0.59% : 0.000010s : 58: predicate.specialize_transform 0.44% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.37% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.19% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.15% : 0.000035s : 209: predicate.switch_defer_inline 2.55% : 0.000041s : 243: predicate.switch_layer_defer_inline 6.69% : 0.000108s : 646: predicate.switch_simplify 1.21% : 0.000020s : 123: predicate.tile_eliminate 1.20% : 0.000019s : 123: predicate.transpose_eliminate 1.73% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.71% : 0.000028s : 157: predicate.tuple_list_get_item_const_eliminator 1.71% : 0.000027s : 157: predicate.tuple_list_get_item_depend_reorder 3.05% : 0.000049s : 228: predicate.tuple_list_get_item_eliminator 1.65% : 0.000027s : 157: predicate.tuple_list_get_set_item_eliminator 2.32% : 0.000037s : 215: predicate.tuple_list_set_item_eliminator 1.78% : 0.000029s : 170: predicate.tuple_to_list_eliminator_ 2.64% : 0.000042s : 295: predicate.updatestate_pure_node_eliminater 3.22% : 0.000052s : 353: predicate.updatestate_useless_node_eliminater 0.17% : 0.000003s : 17: predicate.value_based_eliminate 0.35% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.34% : 0.000005s : 35: predicate.virtual_output_eliminate 0.14% : 0.000002s : 17: predicate.virtual_view_grad_eliminate 0.19% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.016798 135 58.83% : 0.009883s : 55: func_graph_cloner_run.FuncGraphClonerGraph 41.17% : 0.006915s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.416680 209 0.00% : 0.000004s : 1: ForceFp32Comm 1.32% : 0.005482s : 1: add_attr 1.31% : 0.005462s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.04% : 0.000152s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.23% : 0.000948s : 1: auto_monad 0.02% : 0.000076s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.27% : 0.001119s : 1: bootstrap 0.01% : 0.000045s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000044s : 1: control_data_broadcast_order 0.01% : 0.000032s : 1: convert_after_rewriter 0.02% : 0.000085s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.01% : 0.000041s : 1: environ_conv 0.05% : 0.000224s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.01% : 0.000058s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000006s : 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% : 0.000010s : 1: label_micro_interleaved_index 0.16% : 0.000667s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.21% : 0.000883s : 1: mutable_eliminate 0.00% : 0.000014s : 1: offloading_packed_experts 0.01% : 0.000041s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000047s : 1: opt.transform.mutable_eliminate 3.43% : 0.014310s : 95: opt.transform.opt_a 0.03% : 0.000129s : 1: opt.transform.opt_after_cconv 0.02% : 0.000082s : 1: opt.transform.opt_after_jit_grad 0.13% : 0.000527s : 28: opt.transform.opt_b 0.05% : 0.000226s : 2: opt.transform.opt_trans_graph 0.04% : 0.000180s : 4: opt.transform.symbol_engine_opt 7.04% : 0.029350s : 1: opt_a 0.08% : 0.000318s : 1: opt_after_cconv 0.21% : 0.000881s : 1: opt_after_jit_grad 0.19% : 0.000777s : 1: opt_b 8.58% : 0.035750s : 1: optimize 0.01% : 0.000045s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000018s : 1: order_py_execute_after_rewriter 0.01% : 0.000059s : 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% : 0.000015s : 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% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.06% : 0.000268s : 1: pre_auto_parallel 0.06% : 0.000243s : 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 0.03% : 0.000106s : 1: remove_dup_value 1.49% : 0.006219s : 1: renormalize.infer 1.43% : 0.005979s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.22% : 0.000925s : 1: rewriter_after_opt_a 0.19% : 0.000793s : 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 0.01% : 0.000024s : 1: swap_dp_allreduce_reducescatter 0.13% : 0.000553s : 1: symbol_engine_optimizer 0.07% : 0.000273s : 1: tuple_transform 72.70% : 0.302938s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape0-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape0-1024],max_mem:208.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:44:28.727.184 [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 = 0.69544, [21] [bootstrap]: 0.00074674 [type_inference]: 0.58981 [event_method]: 0.00040401 [auto_monad]: 0.00073267 [graph_reusing]: 4.843e-05 [inline]: 2.48e-06 [add_attr]: 0.00602847, [1] [add_attr_with_inline]: 0.00600938, [1] [Cycle 1]: 0.00033608, [2] [tag_attr]: 0.00022345 [meta_addattr_fg_expand]: 5.588e-05 [parallel-infer-symbol]: 3.81999e-06 [pre_auto_parallel]: 0.00027382 [insert-virtual-dataset]: 2.81e-06 [parallel-infer-symbol-second]: 1.16002e-06 [dataset_repeat_opt]: 2.37001e-06 [pipeline_split]: 1.72999e-06 [optimize]: 0.0962503, [53] [py_interpret_to_execute]: 0.00024541 [rewriter_before_opt_a]: 0.00085551 [opt_a]: 0.0780546, [2] [Cycle 1]: 0.0759615, [45] [expand_dump_flag]: 1.763e-05 [switch_simplify]: 0.00085932 [loop_unroll]: 0.00035506 [a_1]: 0.0380276 [with_stream_mark]: 5.141e-05 [recompute_prepare]: 3.819e-05 [updatestate_depend_eliminate]: 2.017e-05 [updatestate_assign_eliminate]: 1.535e-05 [updatestate_loads_eliminate]: 1.837e-05 [parameter_eliminate]: 3.26001e-06 [a_2]: 0.0007586 [accelerated_algorithm]: 5.711e-05 [shard]: 1.91e-06 [meta_shard_fg_expand]: 2.288e-05 [shard_inline]: 2.236e-05 [merge_send_recv]: 1.921e-05 [auto_parallel]: 2.063e-05 [parallel]: 3.15e-05 [flash_sp]: 1.525e-05 [merge_comm]: 1.297e-05 [allreduce_fusion]: 1.227e-05 [matmul_add_comm_reduction]: 2.023e-05 [allreduce_slice_to_reducescatter]: 8.39995e-07 [virtual_shard_identity]: 2.376e-05 [virtual_dataset]: 2.233e-05 [get_grad_eliminate_]: 1.964e-05 [virtual_output]: 2.14e-05 [merge_forward]: 1.236e-05 [cell_reuse_recompute_pass]: 2.26e-06 [offload_activation]: 2.148e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.22e-05 [merge_recompute_call_nodes]: 1.44e-06 [before_grad]: 3.61e-05 [set_forward_comm_id_for_comm_node_pass]: 1.294e-05 [meta_fg_expand]: 1.697e-05 [flash_sp_send_recv_attached]: 5.09e-06 [receive_attached]: 2.76999e-06 [after_resolve]: 2.775e-05 [a_after_grad]: 3.473e-05 [renormalize]: 0.0329417 [add_forward_monad_depend]: 1.851e-05 [auto_monad_grad]: 2.86e-06 [auto_monad_eliminator]: 8.183e-05 [cse]: 0.00162174 [a_3]: 0.0001745 [Cycle 2]: 0.00207386, [45] [expand_dump_flag]: 2.73e-06 [switch_simplify]: 2.636e-05 [loop_unroll]: 2.1e-05 [a_1]: 0.00065082 [with_stream_mark]: 3.092e-05 [recompute_prepare]: 2.322e-05 [updatestate_depend_eliminate]: 1.52e-05 [updatestate_assign_eliminate]: 1.19e-05 [updatestate_loads_eliminate]: 1.335e-05 [parameter_eliminate]: 2.21e-06 [a_2]: 0.00030695 [accelerated_algorithm]: 2.968e-05 [shard]: 2.32001e-06 [meta_shard_fg_expand]: 7.51999e-06 [shard_inline]: 2.03e-05 [merge_send_recv]: 1.868e-05 [auto_parallel]: 1.996e-05 [parallel]: 9.51998e-06 [flash_sp]: 4.36002e-06 [merge_comm]: 1.305e-05 [allreduce_fusion]: 1.195e-05 [matmul_add_comm_reduction]: 2.16e-05 [allreduce_slice_to_reducescatter]: 1.10001e-06 [virtual_shard_identity]: 2.26e-05 [virtual_dataset]: 1.999e-05 [get_grad_eliminate_]: 2.1e-05 [virtual_output]: 1.996e-05 [merge_forward]: 1.173e-05 [cell_reuse_recompute_pass]: 3.2e-06 [offload_activation]: 2.24e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.999e-05 [merge_recompute_call_nodes]: 1.74998e-06 [before_grad]: 3.553e-05 [set_forward_comm_id_for_comm_node_pass]: 1.212e-05 [meta_fg_expand]: 9.61003e-06 [flash_sp_send_recv_attached]: 2.04e-06 [receive_attached]: 3.11001e-06 [after_resolve]: 2.597e-05 [a_after_grad]: 3.178e-05 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 3.21999e-06 [auto_monad_grad]: 3.07002e-06 [auto_monad_eliminator]: 5.656e-05 [cse]: 8.7e-05 [a_3]: 0.0001351 [py_interpret_to_execute_after_opt_a]: 3.22e-05 [slice_cell_reuse_recomputed_activation]: 2.02001e-06 [rewriter_after_opt_a]: 0.0129825 [convert_after_rewriter]: 3.083e-05 [order_py_execute_after_rewriter]: 1.553e-05 [mutable_eliminate]: 0.00078624 [opt_b]: 0.00076256, [1] [Cycle 1]: 0.00075269, [7] [b_1]: 0.00054181 [b_2]: 2.302e-05 [updatestate_depend_eliminate]: 1.621e-05 [updatestate_assign_eliminate]: 1.173e-05 [updatestate_loads_eliminate]: 1.464e-05 [renormalize]: 1.12e-06 [cse]: 0.00010133 [optimize_parallel_all_gather_comm]: 3.791e-05 [overlap_param_gather]: 2.32999e-06 [cconv]: 3.886e-05 [loop_unroll]: 0.00052863 [opt_after_cconv]: 0.00028223, [1] [Cycle 1]: 0.00027432, [7] [c_1]: 0.00012238 [parameter_eliminate]: 4.50001e-06 [updatestate_depend_eliminate]: 1.62e-05 [updatestate_assign_eliminate]: 1.058e-05 [updatestate_loads_eliminate]: 1.224e-05 [cse]: 7.021e-05 [renormalize]: 7.99977e-07 [remove_dup_value]: 9.401e-05 [tuple_transform]: 0.00024133, [1] [Cycle 1]: 0.000236, [4] [d_1]: 0.00019134 [none_parameter_eliminate]: 2.30002e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 2.194e-05 [partial_unused_args_eliminate]: 2.15002e-06 [add_recomputation]: 0.00013506 [cse_after_recomputation]: 6.578e-05, [1] [Cycle 1]: 5.987e-05, [1] [cse]: 5.31e-05 [environ_conv]: 2.843e-05 [swap_dp_allreduce_reducescatter]: 1.776e-05 [bias_add_comm_swap]: 3.95998e-06 [label_micro_interleaved_index]: 5.51e-06 [label_fine_grained_interleaved_index]: 3.06999e-06 [merge_cast_opt]: 1.52999e-06 [slice_recompute_activation]: 2.89001e-06 [micro_interleaved_order_control]: 2.84001e-06 [assign_add_opt]: 1.25999e-06 [ForceFp32Comm]: 9.10019e-07 [remove_cast_before_assign_add]: 1.09003e-06 [full_micro_interleaved_order_control]: 2.55002e-06 [reorder_send_recv_between_fp_bp]: 2.99999e-06 [comm_op_add_attrs]: 1.10999e-06 [add_comm_op_reuse_tag]: 1.05001e-06 [interleave_split_concat_branches]: 1.41002e-06 [interleave_parallel_branches]: 1.24e-06 [overlap_opt_shard_in_pipeline]: 1.45999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.93002e-06 [control_data_broadcast_order]: 3.795e-05 [grouped_pairwise_exchange_alltoall]: 1.62999e-06 [offloading_packed_experts]: 1.11e-05 [overlap_recompute_and_grad_model_parallel]: 1.035e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.52001e-06 [overlap_recompute_comm]: 2.60002e-06 [overlap_grad_ring_attention]: 9.59e-06 [overlap_grad_flash_sp]: 5.009e-05 [begin_end_overlap_inline]: 6.19999e-07 [split_matmul_comm_elemetwise]: 2.43998e-06 [split_layernorm_comm]: 2.26e-06 [handle_group_info]: 1.04e-06 [symbol_engine_optimizer]: 0.00049079, [1] [Cycle 1]: 0.00048386, [6] [build]: 0.00026556 [elim_shapecalc]: 3.254e-05 [elim_not_effective]: 4.94e-05 [opt_reshape]: 3.222e-05 [fold_const_symbol]: 5.866e-05 [renormalize]: 4.60015e-07 [detach_backward]: 2.26e-06 [pipeline_parallel_scheduler]: 1.44e-06 [auto_monad_reorder]: 6.041e-05 [get_jit_bprop_graph]: 1.94e-06 [rewriter_after_jit_bprop_graph]: 4.99e-06 [opt_after_jit_grad]: 0.00069242 [validate]: 9.877e-05 Sums bootstrap : 0.000747s : 0.11% type_inference : 0.589810s : 85.70% event_method : 0.000404s : 0.06% auto_monad : 0.000733s : 0.11% graph_reusing : 0.000048s : 0.01% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000223s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000056s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000274s : 0.04% 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 : 0.000245s : 0.04% optimize.rewriter_before_opt_a : 0.000856s : 0.12% optimize.opt_a.expand_dump_flag : 0.000020s : 0.00% optimize.opt_a.switch_simplify : 0.000886s : 0.13% optimize.opt_a.loop_unroll : 0.000376s : 0.05% optimize.opt_a.a_1 : 0.038678s : 5.62% optimize.opt_a.with_stream_mark : 0.000082s : 0.01% optimize.opt_a.recompute_prepare : 0.000061s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000035s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000027s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000032s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.001066s : 0.15% optimize.opt_a.accelerated_algorithm : 0.000087s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000030s : 0.00% optimize.opt_a.shard_inline : 0.000043s : 0.01% optimize.opt_a.merge_send_recv : 0.000038s : 0.01% optimize.opt_a.auto_parallel : 0.000041s : 0.01% optimize.opt_a.parallel : 0.000041s : 0.01% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000026s : 0.00% optimize.opt_a.allreduce_fusion : 0.000024s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000042s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000046s : 0.01% optimize.opt_a.virtual_dataset : 0.000042s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.01% optimize.opt_a.virtual_output : 0.000041s : 0.01% optimize.opt_a.merge_forward : 0.000024s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000044s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000082s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000072s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000025s : 0.00% optimize.opt_a.meta_fg_expand : 0.000027s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000054s : 0.01% optimize.opt_a.a_after_grad : 0.000067s : 0.01% optimize.opt_a.renormalize : 0.032942s : 4.79% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000138s : 0.02% optimize.opt_a.cse : 0.001709s : 0.25% 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 : 0.012983s : 1.89% optimize.convert_after_rewriter : 0.000031s : 0.00% optimize.order_py_execute_after_rewriter : 0.000016s : 0.00% optimize.mutable_eliminate : 0.000786s : 0.11% optimize.opt_b.b_1 : 0.000542s : 0.08% optimize.opt_b.b_2 : 0.000023s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000016s : 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 : 0.000101s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000038s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000039s : 0.01% optimize.loop_unroll : 0.000529s : 0.08% optimize.opt_after_cconv.c_1 : 0.000122s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000016s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.cse : 0.000070s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000094s : 0.01% optimize.tuple_transform.d_1 : 0.000191s : 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 : 0.000022s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000135s : 0.02% optimize.cse_after_recomputation.cse : 0.000053s : 0.01% optimize.environ_conv : 0.000028s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000018s : 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 : 0.000003s : 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 : 0.000001s : 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 : 0.000011s : 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 : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000010s : 0.00% optimize.overlap_grad_flash_sp : 0.000050s : 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 : 0.000266s : 0.04% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000033s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000049s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000032s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000059s : 0.01% 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 : 0.000060s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000692s : 0.10% validate : 0.000099s : 0.01% Time group info: ------[substitution.] 0.031719 569 0.05% : 0.000017s : 8: substitution.depend_value_elim 0.04% : 0.000013s : 14: substitution.elim_not_effective 0.05% : 0.000016s : 12: substitution.float_tuple_getitem_switch 0.09% : 0.000027s : 14: substitution.fold_const_symbol 0.05% : 0.000015s : 17: substitution.graph_param_transform 8.06% : 0.002556s : 71: substitution.inline 0.04% : 0.000013s : 29: substitution.j_node_and_user_rematch 0.10% : 0.000031s : 6: substitution.less_batch_normalization 0.22% : 0.000069s : 2: substitution.list_to_tuple_eliminator_ 0.02% : 0.000006s : 6: substitution.load_eliminater 0.09% : 0.000030s : 21: substitution.minmaximum_grad 0.02% : 0.000007s : 2: substitution.opt_reshape 0.06% : 0.000019s : 29: substitution.remove_not_recompute_node 0.02% : 0.000008s : 4: substitution.replace_old_param 0.21% : 0.000067s : 18: substitution.reshape_eliminate 0.17% : 0.000054s : 24: substitution.switch_simplify 88.88% : 0.028192s : 51: substitution.tuple_list_convert_item_index_to_positive 0.29% : 0.000092s : 51: substitution.tuple_list_get_item_const_eliminator 0.37% : 0.000117s : 51: substitution.tuple_list_get_item_depend_reorder 0.68% : 0.000217s : 73: substitution.tuple_list_get_item_eliminator 0.36% : 0.000113s : 51: substitution.tuple_list_get_set_item_eliminator 0.07% : 0.000021s : 6: substitution.updatestate_pure_node_eliminater 0.06% : 0.000020s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.589434 2 93.42% : 0.550673s : 1: type_inference.infer 6.58% : 0.038761s : 1: type_inference.specialize ------[replace.] 0.001135 115 3.69% : 0.000042s : 5: replace.depend_value_elim 57.36% : 0.000651s : 71: replace.inline 3.93% : 0.000045s : 2: replace.list_to_tuple_eliminator_ 25.86% : 0.000293s : 24: replace.switch_simplify 9.16% : 0.000104s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002648 115 0.08% : 0.000002s : 5: match.depend_value_elim 94.76% : 0.002510s : 71: match.inline 2.47% : 0.000065s : 2: match.list_to_tuple_eliminator_ 1.40% : 0.000037s : 24: match.switch_simplify 1.29% : 0.000034s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001654 9701 1.31% : 0.000022s : 123: predicate.accumulaten_eliminater 0.31% : 0.000005s : 17: predicate.ad_related_special_op_eliminate 0.49% : 0.000008s : 58: predicate.addn_check_dump 1.25% : 0.000021s : 123: predicate.addn_zero_filter 1.10% : 0.000018s : 123: predicate.adjust_all_reduce_mul_add 2.50% : 0.000041s : 181: predicate.arithmetic_simplify 1.21% : 0.000020s : 123: predicate.cast_eliminate 0.30% : 0.000005s : 34: predicate.check_bprop_eliminate 0.49% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.57% : 0.000009s : 61: predicate.depend_value_elim 1.28% : 0.000021s : 123: predicate.dict_get_item_const_eliminator 1.45% : 0.000024s : 123: predicate.dict_get_item_eliminator 1.15% : 0.000019s : 123: predicate.dict_set_item_eliminator 0.34% : 0.000006s : 34: predicate.dumpgradient_eliminate 0.09% : 0.000002s : 17: predicate.elim_not_effective 0.18% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.27% : 0.000021s : 140: predicate.environ_add_const_eliminate 1.29% : 0.000021s : 140: predicate.environ_get_add_eliminate 1.29% : 0.000021s : 140: predicate.environ_get_depend_swap 1.77% : 0.000029s : 198: predicate.environ_get_eliminate 1.31% : 0.000022s : 140: predicate.environ_get_set_eliminate 1.90% : 0.000032s : 209: predicate.exchange_switch_depend_value 2.64% : 0.000044s : 209: predicate.float_depend_g_call 0.48% : 0.000008s : 58: predicate.float_environ_get_switch 0.65% : 0.000011s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.33% : 0.000005s : 35: predicate.get_grad_eliminate 0.09% : 0.000001s : 17: predicate.graph_param_transform 0.47% : 0.000008s : 58: predicate.incorporate_call 0.45% : 0.000008s : 58: predicate.incorporate_call_switch 5.95% : 0.000098s : 459: predicate.inline 0.35% : 0.000006s : 35: predicate.inline_without_move 0.16% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.44% : 0.000007s : 36: predicate.less_batch_normalization 1.80% : 0.000030s : 172: predicate.list_to_tuple_eliminator_ 2.72% : 0.000045s : 295: predicate.load_eliminater 0.30% : 0.000005s : 17: predicate.loop_unroll_after_grad 3.60% : 0.000060s : 314: predicate.loop_unroll_before_grad 1.58% : 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 1.19% : 0.000020s : 123: predicate.minmaximum_grad 0.29% : 0.000005s : 17: predicate.mutable_eliminate 0.19% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 3.37% : 0.000056s : 209: predicate.partial_defer_inline 1.53% : 0.000025s : 155: predicate.partial_eliminate 1.23% : 0.000020s : 123: predicate.print_const_string_wrapper 0.45% : 0.000007s : 53: predicate.reduce_all_const_elim 1.82% : 0.000030s : 123: predicate.reduce_eliminate 2.61% : 0.000043s : 295: predicate.redundant_stop_gradient_eliminater 0.17% : 0.000003s : 35: predicate.remove_not_recompute_node 1.08% : 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% : 0.000024s : 123: predicate.reshape_eliminate 0.31% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.19% : 0.000003s : 17: predicate.row_tensor_eliminate 0.46% : 0.000008s : 34: predicate.same_eliminate 0.21% : 0.000004s : 40: predicate.set_cell_output_no_recompute 0.37% : 0.000006s : 35: predicate.shard_identity_eliminate 0.32% : 0.000005s : 34: predicate.special_op_eliminate 0.56% : 0.000009s : 58: predicate.specialize_transform 0.40% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.36% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.17% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.14% : 0.000035s : 209: predicate.switch_defer_inline 2.55% : 0.000042s : 243: predicate.switch_layer_defer_inline 7.05% : 0.000117s : 646: predicate.switch_simplify 1.19% : 0.000020s : 123: predicate.tile_eliminate 1.18% : 0.000020s : 123: predicate.transpose_eliminate 1.72% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.82% : 0.000030s : 157: predicate.tuple_list_get_item_const_eliminator 1.70% : 0.000028s : 157: predicate.tuple_list_get_item_depend_reorder 2.85% : 0.000047s : 228: predicate.tuple_list_get_item_eliminator 1.75% : 0.000029s : 157: predicate.tuple_list_get_set_item_eliminator 2.24% : 0.000037s : 215: predicate.tuple_list_set_item_eliminator 1.62% : 0.000027s : 170: predicate.tuple_to_list_eliminator_ 2.65% : 0.000044s : 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 0.33% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.35% : 0.000006s : 35: predicate.virtual_output_eliminate 0.15% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.18% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.016541 135 57.89% : 0.009575s : 55: func_graph_cloner_run.FuncGraphClonerGraph 42.11% : 0.006966s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.873363 209 0.00% : 0.000004s : 1: ForceFp32Comm 0.69% : 0.006036s : 1: add_attr 0.69% : 0.006016s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.02% : 0.000141s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.09% : 0.000748s : 1: auto_monad 0.01% : 0.000067s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.09% : 0.000811s : 1: bootstrap 0.00% : 0.000043s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000042s : 1: control_data_broadcast_order 0.00% : 0.000038s : 1: convert_after_rewriter 0.01% : 0.000069s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000032s : 1: environ_conv 0.05% : 0.000425s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.01% : 0.000054s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000006s : 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% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.06% : 0.000539s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000797s : 1: mutable_eliminate 0.00% : 0.000014s : 1: offloading_packed_experts 0.00% : 0.000033s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000043s : 1: opt.transform.mutable_eliminate 4.79% : 0.041812s : 95: opt.transform.opt_a 0.01% : 0.000121s : 1: opt.transform.opt_after_cconv 0.01% : 0.000074s : 1: opt.transform.opt_after_jit_grad 0.06% : 0.000526s : 28: opt.transform.opt_b 0.02% : 0.000211s : 2: opt.transform.opt_trans_graph 0.02% : 0.000168s : 4: opt.transform.symbol_engine_opt 8.94% : 0.078060s : 1: opt_a 0.03% : 0.000286s : 1: opt_after_cconv 0.08% : 0.000706s : 1: opt_after_jit_grad 0.09% : 0.000767s : 1: opt_b 11.02% : 0.096258s : 1: optimize 0.00% : 0.000042s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000019s : 1: order_py_execute_after_rewriter 0.01% : 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% : 0.000007s : 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% : 0.000013s : 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% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000284s : 1: pre_auto_parallel 0.03% : 0.000255s : 1: py_interpret_to_execute 0.00% : 0.000036s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000099s : 1: remove_dup_value 2.93% : 0.025595s : 1: renormalize.infer 0.84% : 0.007327s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 1.49% : 0.013003s : 1: rewriter_after_opt_a 0.10% : 0.000869s : 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% : 0.000021s : 1: swap_dp_allreduce_reducescatter 0.06% : 0.000494s : 1: symbol_engine_optimizer 0.03% : 0.000246s : 1: tuple_transform 67.54% : 0.589845s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape1-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape1-256],max_mem:146.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:44:38.258.75 [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 = 1.07397, [21] [bootstrap]: 0.0249186 [type_inference]: 0.940748 [event_method]: 0.00022226 [auto_monad]: 0.00076514 [graph_reusing]: 5.061e-05 [inline]: 3.23998e-06 [add_attr]: 0.00527907, [1] [add_attr_with_inline]: 0.0052585, [1] [Cycle 1]: 0.00033503, [2] [tag_attr]: 0.00020439 [meta_addattr_fg_expand]: 5.85e-05 [parallel-infer-symbol]: 3.95998e-06 [pre_auto_parallel]: 0.00026943 [insert-virtual-dataset]: 2.96999e-06 [parallel-infer-symbol-second]: 1.50001e-06 [dataset_repeat_opt]: 2.89001e-06 [pipeline_split]: 1.64998e-06 [optimize]: 0.100429, [53] [py_interpret_to_execute]: 0.00025111 [rewriter_before_opt_a]: 0.00082763 [opt_a]: 0.0813432, [2] [Cycle 1]: 0.0791044, [45] [expand_dump_flag]: 2.781e-05 [switch_simplify]: 0.00085987 [loop_unroll]: 0.00032048 [a_1]: 0.046567 [with_stream_mark]: 5.864e-05 [recompute_prepare]: 4.036e-05 [updatestate_depend_eliminate]: 2.135e-05 [updatestate_assign_eliminate]: 1.525e-05 [updatestate_loads_eliminate]: 1.934e-05 [parameter_eliminate]: 2.79001e-06 [a_2]: 0.00079611 [accelerated_algorithm]: 6.678e-05 [shard]: 2.73998e-06 [meta_shard_fg_expand]: 2.454e-05 [shard_inline]: 2.249e-05 [merge_send_recv]: 2.028e-05 [auto_parallel]: 2.063e-05 [parallel]: 3.188e-05 [flash_sp]: 1.627e-05 [merge_comm]: 1.495e-05 [allreduce_fusion]: 1.275e-05 [matmul_add_comm_reduction]: 2.207e-05 [allreduce_slice_to_reducescatter]: 8.29983e-07 [virtual_shard_identity]: 2.717e-05 [virtual_dataset]: 2.252e-05 [get_grad_eliminate_]: 2.186e-05 [virtual_output]: 2.169e-05 [merge_forward]: 1.316e-05 [cell_reuse_recompute_pass]: 2.19001e-06 [offload_activation]: 2.432e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.68e-05 [merge_recompute_call_nodes]: 1.70001e-06 [before_grad]: 3.919e-05 [set_forward_comm_id_for_comm_node_pass]: 1.39e-05 [meta_fg_expand]: 1.665e-05 [flash_sp_send_recv_attached]: 5.35001e-06 [receive_attached]: 2.33998e-06 [after_resolve]: 2.98e-05 [a_after_grad]: 3.707e-05 [renormalize]: 0.026182 [add_forward_monad_depend]: 2.082e-05 [auto_monad_grad]: 3.85e-06 [auto_monad_eliminator]: 9.29e-05 [cse]: 0.00283217 [a_3]: 0.00017945 [Cycle 2]: 0.00221703, [45] [expand_dump_flag]: 3.91001e-06 [switch_simplify]: 2.748e-05 [loop_unroll]: 2.175e-05 [a_1]: 0.00066349 [with_stream_mark]: 3.932e-05 [recompute_prepare]: 2.33e-05 [updatestate_depend_eliminate]: 1.669e-05 [updatestate_assign_eliminate]: 1.153e-05 [updatestate_loads_eliminate]: 1.391e-05 [parameter_eliminate]: 2.79001e-06 [a_2]: 0.0003127 [accelerated_algorithm]: 3.35e-05 [shard]: 2.27001e-06 [meta_shard_fg_expand]: 1.028e-05 [shard_inline]: 2.226e-05 [merge_send_recv]: 2.105e-05 [auto_parallel]: 2.157e-05 [parallel]: 1.182e-05 [flash_sp]: 4.36002e-06 [merge_comm]: 1.339e-05 [allreduce_fusion]: 1.197e-05 [matmul_add_comm_reduction]: 2.366e-05 [allreduce_slice_to_reducescatter]: 1.15001e-06 [virtual_shard_identity]: 2.336e-05 [virtual_dataset]: 1.988e-05 [get_grad_eliminate_]: 2.198e-05 [virtual_output]: 1.963e-05 [merge_forward]: 1.24e-05 [cell_reuse_recompute_pass]: 3.36999e-06 [offload_activation]: 2.66e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.384e-05 [merge_recompute_call_nodes]: 1.75001e-06 [before_grad]: 3.821e-05 [set_forward_comm_id_for_comm_node_pass]: 1.52e-05 [meta_fg_expand]: 9.32999e-06 [flash_sp_send_recv_attached]: 2.09e-06 [receive_attached]: 2.67001e-06 [after_resolve]: 2.832e-05 [a_after_grad]: 3.204e-05 [renormalize]: 2.30008e-07 [add_forward_monad_depend]: 3.78999e-06 [auto_monad_grad]: 3.16001e-06 [auto_monad_eliminator]: 6.867e-05 [cse]: 9.414e-05 [a_3]: 0.00013934 [py_interpret_to_execute_after_opt_a]: 3.432e-05 [slice_cell_reuse_recomputed_activation]: 2.56e-06 [rewriter_after_opt_a]: 0.00088233 [convert_after_rewriter]: 2.854e-05 [order_py_execute_after_rewriter]: 1.631e-05 [mutable_eliminate]: 0.0134023 [opt_b]: 0.00081892, [1] [Cycle 1]: 0.00080589, [7] [b_1]: 0.00056605 [b_2]: 2.392e-05 [updatestate_depend_eliminate]: 2.07e-05 [updatestate_assign_eliminate]: 1.12e-05 [updatestate_loads_eliminate]: 1.542e-05 [renormalize]: 9.30013e-07 [cse]: 0.00011777 [optimize_parallel_all_gather_comm]: 4.246e-05 [overlap_param_gather]: 2.30002e-06 [cconv]: 3.632e-05 [loop_unroll]: 0.000725 [opt_after_cconv]: 0.00032385, [1] [Cycle 1]: 0.00031491, [7] [c_1]: 0.00013552 [parameter_eliminate]: 5.25001e-06 [updatestate_depend_eliminate]: 2.039e-05 [updatestate_assign_eliminate]: 1.084e-05 [updatestate_loads_eliminate]: 1.368e-05 [cse]: 9.042e-05 [renormalize]: 8.59989e-07 [remove_dup_value]: 0.00010202 [tuple_transform]: 0.00025819, [1] [Cycle 1]: 0.00025187, [4] [d_1]: 0.00020634 [none_parameter_eliminate]: 2.39001e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 2.11e-05 [partial_unused_args_eliminate]: 2.17001e-06 [add_recomputation]: 0.00014543 [cse_after_recomputation]: 7.748e-05, [1] [Cycle 1]: 7.042e-05, [1] [cse]: 6.331e-05 [environ_conv]: 5.538e-05 [swap_dp_allreduce_reducescatter]: 1.995e-05 [bias_add_comm_swap]: 3.78001e-06 [label_micro_interleaved_index]: 7.51001e-06 [label_fine_grained_interleaved_index]: 2.96001e-06 [merge_cast_opt]: 1.47999e-06 [slice_recompute_activation]: 2.71999e-06 [micro_interleaved_order_control]: 2.86999e-06 [assign_add_opt]: 1.74e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 1.05999e-06 [full_micro_interleaved_order_control]: 2.54001e-06 [reorder_send_recv_between_fp_bp]: 3.12002e-06 [comm_op_add_attrs]: 1.46002e-06 [add_comm_op_reuse_tag]: 1.00999e-06 [interleave_split_concat_branches]: 1.22e-06 [interleave_parallel_branches]: 1.30001e-06 [overlap_opt_shard_in_pipeline]: 1.30001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.78002e-06 [control_data_broadcast_order]: 3.68e-05 [grouped_pairwise_exchange_alltoall]: 1.81e-06 [offloading_packed_experts]: 1.137e-05 [overlap_recompute_and_grad_model_parallel]: 9.96e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.24e-06 [overlap_recompute_allgather_and_fa_grad]: 1.81e-06 [overlap_recompute_comm]: 2.32001e-06 [overlap_grad_ring_attention]: 9.17999e-06 [overlap_grad_flash_sp]: 5.076e-05 [begin_end_overlap_inline]: 5.50004e-07 [split_matmul_comm_elemetwise]: 2.59999e-06 [split_layernorm_comm]: 2.13002e-06 [handle_group_info]: 1.41002e-06 [symbol_engine_optimizer]: 0.00051673, [1] [Cycle 1]: 0.00051006, [6] [build]: 0.00028426 [elim_shapecalc]: 3.44e-05 [elim_not_effective]: 5.335e-05 [opt_reshape]: 3.401e-05 [fold_const_symbol]: 5.892e-05 [renormalize]: 7.49977e-07 [detach_backward]: 2.66999e-06 [pipeline_parallel_scheduler]: 1.49e-06 [auto_monad_reorder]: 6.637e-05 [get_jit_bprop_graph]: 1.74e-06 [rewriter_after_jit_bprop_graph]: 5.10001e-06 [opt_after_jit_grad]: 0.00083146 [validate]: 0.00010697 Sums bootstrap : 0.024919s : 2.33% type_inference : 0.940748s : 88.14% event_method : 0.000222s : 0.02% auto_monad : 0.000765s : 0.07% graph_reusing : 0.000051s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000204s : 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 : 0.000269s : 0.03% 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 : 0.000251s : 0.02% optimize.rewriter_before_opt_a : 0.000828s : 0.08% optimize.opt_a.expand_dump_flag : 0.000032s : 0.00% optimize.opt_a.switch_simplify : 0.000887s : 0.08% optimize.opt_a.loop_unroll : 0.000342s : 0.03% optimize.opt_a.a_1 : 0.047230s : 4.43% optimize.opt_a.with_stream_mark : 0.000098s : 0.01% optimize.opt_a.recompute_prepare : 0.000064s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000038s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000027s : 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 : 0.001109s : 0.10% optimize.opt_a.accelerated_algorithm : 0.000100s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000035s : 0.00% optimize.opt_a.shard_inline : 0.000045s : 0.00% optimize.opt_a.merge_send_recv : 0.000041s : 0.00% optimize.opt_a.auto_parallel : 0.000042s : 0.00% optimize.opt_a.parallel : 0.000044s : 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 : 0.000046s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000051s : 0.00% optimize.opt_a.virtual_dataset : 0.000042s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000044s : 0.00% optimize.opt_a.virtual_output : 0.000041s : 0.00% optimize.opt_a.merge_forward : 0.000026s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000051s : 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 : 0.000077s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000029s : 0.00% optimize.opt_a.meta_fg_expand : 0.000026s : 0.00% 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 : 0.000058s : 0.01% optimize.opt_a.a_after_grad : 0.000069s : 0.01% optimize.opt_a.renormalize : 0.026182s : 2.45% optimize.opt_a.add_forward_monad_depend : 0.000025s : 0.00% optimize.opt_a.auto_monad_grad : 0.000007s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000162s : 0.02% optimize.opt_a.cse : 0.002926s : 0.27% optimize.opt_a.a_3 : 0.000319s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000034s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000882s : 0.08% optimize.convert_after_rewriter : 0.000029s : 0.00% optimize.order_py_execute_after_rewriter : 0.000016s : 0.00% optimize.mutable_eliminate : 0.013402s : 1.26% optimize.opt_b.b_1 : 0.000566s : 0.05% optimize.opt_b.b_2 : 0.000024s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000021s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000118s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000042s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.00% optimize.loop_unroll : 0.000725s : 0.07% optimize.opt_after_cconv.c_1 : 0.000136s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 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 : 0.000014s : 0.00% optimize.opt_after_cconv.cse : 0.000090s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000102s : 0.01% optimize.tuple_transform.d_1 : 0.000206s : 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 : 0.000021s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000145s : 0.01% optimize.cse_after_recomputation.cse : 0.000063s : 0.01% optimize.environ_conv : 0.000055s : 0.01% 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 : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000003s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000037s : 0.00% 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 : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 0.00% optimize.overlap_grad_flash_sp : 0.000051s : 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 : 0.000284s : 0.03% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000034s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000053s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000034s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000059s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000001s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000001s : 0.00% auto_monad_reorder : 0.000066s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000831s : 0.08% validate : 0.000107s : 0.01% Time group info: ------[substitution.] 0.040251 569 0.04% : 0.000017s : 8: substitution.depend_value_elim 0.03% : 0.000014s : 14: substitution.elim_not_effective 0.04% : 0.000017s : 12: substitution.float_tuple_getitem_switch 0.07% : 0.000028s : 14: substitution.fold_const_symbol 0.04% : 0.000015s : 17: substitution.graph_param_transform 97.07% : 0.039070s : 71: substitution.inline 0.04% : 0.000015s : 29: substitution.j_node_and_user_rematch 0.08% : 0.000034s : 6: substitution.less_batch_normalization 0.16% : 0.000063s : 2: substitution.list_to_tuple_eliminator_ 0.02% : 0.000007s : 6: substitution.load_eliminater 0.08% : 0.000032s : 21: substitution.minmaximum_grad 0.02% : 0.000008s : 2: substitution.opt_reshape 0.05% : 0.000019s : 29: substitution.remove_not_recompute_node 0.02% : 0.000009s : 4: substitution.replace_old_param 0.18% : 0.000073s : 18: substitution.reshape_eliminate 0.13% : 0.000054s : 24: substitution.switch_simplify 0.41% : 0.000164s : 51: substitution.tuple_list_convert_item_index_to_positive 0.23% : 0.000091s : 51: substitution.tuple_list_get_item_const_eliminator 0.30% : 0.000121s : 51: substitution.tuple_list_get_item_depend_reorder 0.58% : 0.000233s : 73: substitution.tuple_list_get_item_eliminator 0.29% : 0.000115s : 51: substitution.tuple_list_get_set_item_eliminator 0.07% : 0.000028s : 6: substitution.updatestate_pure_node_eliminater 0.06% : 0.000022s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.940308 2 91.62% : 0.861542s : 1: type_inference.infer 8.38% : 0.078767s : 1: type_inference.specialize ------[replace.] 0.001289 115 3.72% : 0.000048s : 5: replace.depend_value_elim 59.40% : 0.000765s : 71: replace.inline 3.93% : 0.000051s : 2: replace.list_to_tuple_eliminator_ 24.62% : 0.000317s : 24: replace.switch_simplify 8.32% : 0.000107s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.039152 115 0.00% : 0.000002s : 5: match.depend_value_elim 99.65% : 0.039017s : 71: match.inline 0.15% : 0.000059s : 2: match.list_to_tuple_eliminator_ 0.10% : 0.000040s : 24: match.switch_simplify 0.09% : 0.000034s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001641 9701 1.36% : 0.000022s : 123: predicate.accumulaten_eliminater 0.37% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.48% : 0.000008s : 58: predicate.addn_check_dump 1.17% : 0.000019s : 123: predicate.addn_zero_filter 1.06% : 0.000017s : 123: predicate.adjust_all_reduce_mul_add 2.27% : 0.000037s : 181: predicate.arithmetic_simplify 1.20% : 0.000020s : 123: predicate.cast_eliminate 0.30% : 0.000005s : 34: predicate.check_bprop_eliminate 0.51% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.56% : 0.000009s : 61: predicate.depend_value_elim 1.21% : 0.000020s : 123: predicate.dict_get_item_const_eliminator 1.41% : 0.000023s : 123: predicate.dict_get_item_eliminator 1.11% : 0.000018s : 123: predicate.dict_set_item_eliminator 0.39% : 0.000006s : 34: predicate.dumpgradient_eliminate 0.09% : 0.000001s : 17: predicate.elim_not_effective 0.20% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.29% : 0.000021s : 140: predicate.environ_add_const_eliminate 1.22% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.27% : 0.000021s : 140: predicate.environ_get_depend_swap 1.72% : 0.000028s : 198: predicate.environ_get_eliminate 1.25% : 0.000020s : 140: predicate.environ_get_set_eliminate 1.93% : 0.000032s : 209: predicate.exchange_switch_depend_value 2.86% : 0.000047s : 209: predicate.float_depend_g_call 0.50% : 0.000008s : 58: predicate.float_environ_get_switch 0.70% : 0.000011s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.33% : 0.000005s : 35: predicate.get_grad_eliminate 0.09% : 0.000002s : 17: predicate.graph_param_transform 0.49% : 0.000008s : 58: predicate.incorporate_call 0.45% : 0.000007s : 58: predicate.incorporate_call_switch 5.77% : 0.000095s : 459: predicate.inline 0.40% : 0.000007s : 35: predicate.inline_without_move 0.15% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.55% : 0.000009s : 36: predicate.less_batch_normalization 1.84% : 0.000030s : 172: predicate.list_to_tuple_eliminator_ 2.68% : 0.000044s : 295: predicate.load_eliminater 0.43% : 0.000007s : 17: predicate.loop_unroll_after_grad 3.31% : 0.000054s : 314: predicate.loop_unroll_before_grad 1.56% : 0.000026s : 157: predicate.make_slice_get_slice_eliminator 0.51% : 0.000008s : 58: predicate.merge_addn 0.29% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.28% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.20% : 0.000020s : 123: predicate.minmaximum_grad 0.88% : 0.000014s : 17: predicate.mutable_eliminate 0.20% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 3.65% : 0.000060s : 209: predicate.partial_defer_inline 1.48% : 0.000024s : 155: predicate.partial_eliminate 1.15% : 0.000019s : 123: predicate.print_const_string_wrapper 0.46% : 0.000008s : 53: predicate.reduce_all_const_elim 1.74% : 0.000029s : 123: predicate.reduce_eliminate 2.59% : 0.000042s : 295: predicate.redundant_stop_gradient_eliminater 0.16% : 0.000003s : 35: predicate.remove_not_recompute_node 1.08% : 0.000018s : 172: predicate.replace_applicator 0.20% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.28% : 0.000021s : 123: predicate.reshape_eliminate 0.29% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.17% : 0.000003s : 17: predicate.row_tensor_eliminate 0.45% : 0.000007s : 34: predicate.same_eliminate 0.20% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.37% : 0.000006s : 35: predicate.shard_identity_eliminate 0.34% : 0.000006s : 34: predicate.special_op_eliminate 0.61% : 0.000010s : 58: predicate.specialize_transform 0.45% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.42% : 0.000007s : 35: predicate.stack_unstack_eliminate 0.18% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.13% : 0.000035s : 209: predicate.switch_defer_inline 2.40% : 0.000039s : 243: predicate.switch_layer_defer_inline 6.81% : 0.000112s : 646: predicate.switch_simplify 1.24% : 0.000020s : 123: predicate.tile_eliminate 1.10% : 0.000018s : 123: predicate.transpose_eliminate 1.74% : 0.000029s : 157: predicate.tuple_list_convert_item_index_to_positive 1.71% : 0.000028s : 157: predicate.tuple_list_get_item_const_eliminator 1.70% : 0.000028s : 157: predicate.tuple_list_get_item_depend_reorder 3.13% : 0.000051s : 228: predicate.tuple_list_get_item_eliminator 1.82% : 0.000030s : 157: predicate.tuple_list_get_set_item_eliminator 2.30% : 0.000038s : 215: predicate.tuple_list_set_item_eliminator 1.59% : 0.000026s : 170: predicate.tuple_to_list_eliminator_ 2.53% : 0.000041s : 295: predicate.updatestate_pure_node_eliminater 3.06% : 0.000050s : 353: predicate.updatestate_useless_node_eliminater 0.17% : 0.000003s : 17: predicate.value_based_eliminate 0.32% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.33% : 0.000005s : 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.] 0.086452 135 40.34% : 0.034878s : 55: func_graph_cloner_run.FuncGraphClonerGraph 59.66% : 0.051574s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.257274 209 0.00% : 0.000004s : 1: ForceFp32Comm 0.42% : 0.005287s : 1: add_attr 0.42% : 0.005265s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000150s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.06% : 0.000783s : 1: auto_monad 0.01% : 0.000071s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 1.99% : 0.024961s : 1: bootstrap 0.00% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000041s : 1: control_data_broadcast_order 0.00% : 0.000036s : 1: convert_after_rewriter 0.01% : 0.000080s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000060s : 1: environ_conv 0.02% : 0.000236s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000057s : 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 0.06% : 0.000740s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 1.07% : 0.013429s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.00% : 0.000043s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000071s : 1: opt.transform.mutable_eliminate 4.01% : 0.050412s : 95: opt.transform.opt_a 0.01% : 0.000134s : 1: opt.transform.opt_after_cconv 0.01% : 0.000081s : 1: opt.transform.opt_after_jit_grad 0.04% : 0.000547s : 28: opt.transform.opt_b 0.02% : 0.000225s : 2: opt.transform.opt_trans_graph 0.01% : 0.000175s : 4: opt.transform.symbol_engine_opt 6.47% : 0.081347s : 1: opt_a 0.03% : 0.000328s : 1: opt_after_cconv 0.07% : 0.000846s : 1: opt_after_jit_grad 0.07% : 0.000824s : 1: opt_b 7.99% : 0.100436s : 1: optimize 0.00% : 0.000047s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000020s : 1: order_py_execute_after_rewriter 0.00% : 0.000054s : 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% : 0.000007s : 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% : 0.000006s : 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% : 0.000004s : 1: pipeline_split 0.02% : 0.000278s : 1: pre_auto_parallel 0.02% : 0.000262s : 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 0.01% : 0.000107s : 1: remove_dup_value 1.51% : 0.018993s : 1: renormalize.infer 0.57% : 0.007165s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.07% : 0.000899s : 1: rewriter_after_opt_a 0.07% : 0.000845s : 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% : 0.000024s : 1: swap_dp_allreduce_reducescatter 0.04% : 0.000521s : 1: symbol_engine_optimizer 0.02% : 0.000263s : 1: tuple_transform 74.83% : 0.940793s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape1-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape1-1024],max_mem:170.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:44:51.778.316 [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 = 0.745947, [21] [bootstrap]: 0.00073268 [type_inference]: 0.64639 [event_method]: 0.00047209 [auto_monad]: 0.00074527 [graph_reusing]: 5.158e-05 [inline]: 3.2e-06 [add_attr]: 0.0146667, [1] [add_attr_with_inline]: 0.0146473, [1] [Cycle 1]: 0.00032997, [2] [tag_attr]: 0.00019991 [meta_addattr_fg_expand]: 5.562e-05 [parallel-infer-symbol]: 3.97e-06 [pre_auto_parallel]: 0.00025896 [insert-virtual-dataset]: 2.74001e-06 [parallel-infer-symbol-second]: 1.67001e-06 [dataset_repeat_opt]: 2.57001e-06 [pipeline_split]: 1.98002e-06 [optimize]: 0.0810724, [53] [py_interpret_to_execute]: 0.00023987 [rewriter_before_opt_a]: 0.0007874 [opt_a]: 0.0463364, [2] [Cycle 1]: 0.0441186, [45] [expand_dump_flag]: 1.825e-05 [switch_simplify]: 0.00322218 [loop_unroll]: 0.0003366 [a_1]: 0.00863733 [with_stream_mark]: 5.331e-05 [recompute_prepare]: 3.748e-05 [updatestate_depend_eliminate]: 2.021e-05 [updatestate_assign_eliminate]: 1.59e-05 [updatestate_loads_eliminate]: 1.857e-05 [parameter_eliminate]: 2.71e-06 [a_2]: 0.00078137 [accelerated_algorithm]: 6.231e-05 [shard]: 2.67001e-06 [meta_shard_fg_expand]: 2.015e-05 [shard_inline]: 2.191e-05 [merge_send_recv]: 1.977e-05 [auto_parallel]: 1.943e-05 [parallel]: 3.092e-05 [flash_sp]: 1.418e-05 [merge_comm]: 1.381e-05 [allreduce_fusion]: 1.247e-05 [matmul_add_comm_reduction]: 2.014e-05 [allreduce_slice_to_reducescatter]: 8.70001e-07 [virtual_shard_identity]: 2.471e-05 [virtual_dataset]: 2.215e-05 [get_grad_eliminate_]: 2.068e-05 [virtual_output]: 2.121e-05 [merge_forward]: 1.216e-05 [cell_reuse_recompute_pass]: 1.48002e-06 [offload_activation]: 2.183e-05 [cell_reuse_handle_not_recompute_node_pass]: 0.00692637 [merge_recompute_call_nodes]: 4.03999e-06 [before_grad]: 6.59e-05 [set_forward_comm_id_for_comm_node_pass]: 2.962e-05 [meta_fg_expand]: 3.953e-05 [flash_sp_send_recv_attached]: 1.255e-05 [receive_attached]: 2.91e-06 [after_resolve]: 3.274e-05 [a_after_grad]: 3.761e-05 [renormalize]: 0.0218013 [add_forward_monad_depend]: 1.954e-05 [auto_monad_grad]: 3.21999e-06 [auto_monad_eliminator]: 7.469e-05 [cse]: 0.00092977 [a_3]: 0.00017274 [Cycle 2]: 0.00220069, [45] [expand_dump_flag]: 2.71e-06 [switch_simplify]: 2.559e-05 [loop_unroll]: 2.11e-05 [a_1]: 0.00063767 [with_stream_mark]: 2.902e-05 [recompute_prepare]: 2.026e-05 [updatestate_depend_eliminate]: 1.429e-05 [updatestate_assign_eliminate]: 1.122e-05 [updatestate_loads_eliminate]: 1.368e-05 [parameter_eliminate]: 1.87999e-06 [a_2]: 0.00030227 [accelerated_algorithm]: 2.842e-05 [shard]: 2.41e-06 [meta_shard_fg_expand]: 6.64001e-06 [shard_inline]: 1.952e-05 [merge_send_recv]: 1.827e-05 [auto_parallel]: 1.876e-05 [parallel]: 9.99001e-06 [flash_sp]: 4.4e-06 [merge_comm]: 1.228e-05 [allreduce_fusion]: 1.173e-05 [matmul_add_comm_reduction]: 2.124e-05 [allreduce_slice_to_reducescatter]: 8.50006e-07 [virtual_shard_identity]: 2.126e-05 [virtual_dataset]: 2.019e-05 [get_grad_eliminate_]: 2.011e-05 [virtual_output]: 1.95e-05 [merge_forward]: 1.183e-05 [cell_reuse_recompute_pass]: 2.99999e-06 [offload_activation]: 2.302e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.715e-05 [merge_recompute_call_nodes]: 1.49e-06 [before_grad]: 3.507e-05 [set_forward_comm_id_for_comm_node_pass]: 1.218e-05 [meta_fg_expand]: 9.95002e-06 [flash_sp_send_recv_attached]: 2.19001e-06 [receive_attached]: 3.55e-06 [after_resolve]: 2.657e-05 [a_after_grad]: 3.125e-05 [renormalize]: 9.00181e-08 [add_forward_monad_depend]: 2.71e-06 [auto_monad_grad]: 2.54999e-06 [auto_monad_eliminator]: 5.175e-05 [cse]: 0.00026942 [a_3]: 0.00014174 [py_interpret_to_execute_after_opt_a]: 3.393e-05 [slice_cell_reuse_recomputed_activation]: 2.20002e-06 [rewriter_after_opt_a]: 0.00086994 [convert_after_rewriter]: 2.639e-05 [order_py_execute_after_rewriter]: 1.436e-05 [mutable_eliminate]: 0.00085026 [opt_b]: 0.00077327, [1] [Cycle 1]: 0.00076196, [7] [b_1]: 0.00054366 [b_2]: 2.393e-05 [updatestate_depend_eliminate]: 2.159e-05 [updatestate_assign_eliminate]: 1.128e-05 [updatestate_loads_eliminate]: 1.463e-05 [renormalize]: 1.25999e-06 [cse]: 9.72e-05 [optimize_parallel_all_gather_comm]: 4.081e-05 [overlap_param_gather]: 2.09999e-06 [cconv]: 4.245e-05 [loop_unroll]: 0.00059378 [opt_after_cconv]: 0.00029038, [1] [Cycle 1]: 0.00028183, [7] [c_1]: 0.00012663 [parameter_eliminate]: 4.77e-06 [updatestate_depend_eliminate]: 1.728e-05 [updatestate_assign_eliminate]: 1.011e-05 [updatestate_loads_eliminate]: 1.257e-05 [cse]: 7.002e-05 [renormalize]: 5.50004e-07 [remove_dup_value]: 9.943e-05 [tuple_transform]: 0.00024745, [1] [Cycle 1]: 0.00024194, [4] [d_1]: 0.00019474 [none_parameter_eliminate]: 2.03002e-06 [renormalize]: 2.30008e-07 [switch_simplify]: 2.217e-05 [partial_unused_args_eliminate]: 2.26e-06 [add_recomputation]: 0.00013857 [cse_after_recomputation]: 0.0284095, [1] [Cycle 1]: 0.0283986, [1] [cse]: 0.0283457 [environ_conv]: 4.401e-05 [swap_dp_allreduce_reducescatter]: 3.401e-05 [bias_add_comm_swap]: 4.38001e-06 [label_micro_interleaved_index]: 8.49002e-06 [label_fine_grained_interleaved_index]: 2.88e-06 [merge_cast_opt]: 1.65001e-06 [slice_recompute_activation]: 2.56e-06 [micro_interleaved_order_control]: 3.2e-06 [assign_add_opt]: 1.67999e-06 [ForceFp32Comm]: 9.5999e-07 [remove_cast_before_assign_add]: 1.22e-06 [full_micro_interleaved_order_control]: 3.07002e-06 [reorder_send_recv_between_fp_bp]: 3.04999e-06 [comm_op_add_attrs]: 1.29998e-06 [add_comm_op_reuse_tag]: 1.49e-06 [interleave_split_concat_branches]: 1.79e-06 [interleave_parallel_branches]: 1.17e-06 [overlap_opt_shard_in_pipeline]: 1.49e-06 [overlap_opt_shard_grad_in_pipeline]: 2.43e-06 [control_data_broadcast_order]: 4.446e-05 [grouped_pairwise_exchange_alltoall]: 1.63002e-06 [offloading_packed_experts]: 1.134e-05 [overlap_recompute_and_grad_model_parallel]: 1.142e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.29e-06 [overlap_recompute_allgather_and_fa_grad]: 1.77999e-06 [overlap_recompute_comm]: 2.74999e-06 [overlap_grad_ring_attention]: 9.32999e-06 [overlap_grad_flash_sp]: 5.818e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.69001e-06 [split_layernorm_comm]: 2.14e-06 [handle_group_info]: 1.17999e-06 [symbol_engine_optimizer]: 0.00064887, [1] [Cycle 1]: 0.00063574, [6] [build]: 0.0003267 [elim_shapecalc]: 5.055e-05 [elim_not_effective]: 6.31e-05 [opt_reshape]: 3.99e-05 [fold_const_symbol]: 7.049e-05 [renormalize]: 9.50007e-07 [detach_backward]: 3.71999e-06 [pipeline_parallel_scheduler]: 2.09999e-06 [auto_monad_reorder]: 7.951e-05 [get_jit_bprop_graph]: 2.53e-06 [rewriter_after_jit_bprop_graph]: 9.34e-06 [opt_after_jit_grad]: 0.00101373 [validate]: 0.00011766 Sums bootstrap : 0.000733s : 0.10% type_inference : 0.646390s : 88.56% event_method : 0.000472s : 0.06% auto_monad : 0.000745s : 0.10% graph_reusing : 0.000052s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000200s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000056s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000259s : 0.04% 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 : 0.000240s : 0.03% optimize.rewriter_before_opt_a : 0.000787s : 0.11% optimize.opt_a.expand_dump_flag : 0.000021s : 0.00% optimize.opt_a.switch_simplify : 0.003248s : 0.44% optimize.opt_a.loop_unroll : 0.000358s : 0.05% optimize.opt_a.a_1 : 0.009275s : 1.27% optimize.opt_a.with_stream_mark : 0.000082s : 0.01% optimize.opt_a.recompute_prepare : 0.000058s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000035s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000027s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000032s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.001084s : 0.15% optimize.opt_a.accelerated_algorithm : 0.000091s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000027s : 0.00% optimize.opt_a.shard_inline : 0.000041s : 0.01% optimize.opt_a.merge_send_recv : 0.000038s : 0.01% optimize.opt_a.auto_parallel : 0.000038s : 0.01% optimize.opt_a.parallel : 0.000041s : 0.01% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000026s : 0.00% optimize.opt_a.allreduce_fusion : 0.000024s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000041s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000046s : 0.01% optimize.opt_a.virtual_dataset : 0.000042s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000041s : 0.01% optimize.opt_a.virtual_output : 0.000041s : 0.01% optimize.opt_a.merge_forward : 0.000024s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000045s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.006964s : 0.95% optimize.opt_a.merge_recompute_call_nodes : 0.000006s : 0.00% optimize.opt_a.before_grad : 0.000101s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000042s : 0.01% optimize.opt_a.meta_fg_expand : 0.000049s : 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 : 0.000059s : 0.01% optimize.opt_a.a_after_grad : 0.000069s : 0.01% optimize.opt_a.renormalize : 0.021801s : 2.99% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000126s : 0.02% optimize.opt_a.cse : 0.001199s : 0.16% optimize.opt_a.a_3 : 0.000314s : 0.04% optimize.py_interpret_to_execute_after_opt_a : 0.000034s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000870s : 0.12% optimize.convert_after_rewriter : 0.000026s : 0.00% optimize.order_py_execute_after_rewriter : 0.000014s : 0.00% optimize.mutable_eliminate : 0.000850s : 0.12% optimize.opt_b.b_1 : 0.000544s : 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 : 0.000011s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000097s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000041s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000042s : 0.01% optimize.loop_unroll : 0.000594s : 0.08% optimize.opt_after_cconv.c_1 : 0.000127s : 0.02% 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 : 0.000010s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.cse : 0.000070s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000099s : 0.01% optimize.tuple_transform.d_1 : 0.000195s : 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 : 0.000022s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000139s : 0.02% optimize.cse_after_recomputation.cse : 0.028346s : 3.88% optimize.environ_conv : 0.000044s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000034s : 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 : 0.000003s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000044s : 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 : 0.000011s : 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 : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000009s : 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 : 0.000327s : 0.04% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000051s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000063s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000040s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000070s : 0.01% optimize.symbol_engine_optimizer.renormalize : 0.000001s : 0.00% detach_backward : 0.000004s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000080s : 0.01% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000009s : 0.00% opt_after_jit_grad : 0.001014s : 0.14% validate : 0.000118s : 0.02% Time group info: ------[substitution.] 0.003224 569 0.48% : 0.000015s : 8: substitution.depend_value_elim 0.48% : 0.000016s : 14: substitution.elim_not_effective 0.54% : 0.000018s : 12: substitution.float_tuple_getitem_switch 1.07% : 0.000035s : 14: substitution.fold_const_symbol 0.46% : 0.000015s : 17: substitution.graph_param_transform 67.04% : 0.002161s : 71: substitution.inline 0.57% : 0.000018s : 29: substitution.j_node_and_user_rematch 1.11% : 0.000036s : 6: substitution.less_batch_normalization 1.41% : 0.000045s : 2: substitution.list_to_tuple_eliminator_ 0.21% : 0.000007s : 6: substitution.load_eliminater 0.89% : 0.000029s : 21: substitution.minmaximum_grad 0.33% : 0.000011s : 2: substitution.opt_reshape 0.89% : 0.000029s : 29: substitution.remove_not_recompute_node 0.30% : 0.000010s : 4: substitution.replace_old_param 1.75% : 0.000056s : 18: substitution.reshape_eliminate 2.38% : 0.000077s : 24: substitution.switch_simplify 4.38% : 0.000141s : 51: substitution.tuple_list_convert_item_index_to_positive 2.31% : 0.000075s : 51: substitution.tuple_list_get_item_const_eliminator 2.96% : 0.000095s : 51: substitution.tuple_list_get_item_depend_reorder 6.15% : 0.000198s : 73: substitution.tuple_list_get_item_eliminator 2.93% : 0.000094s : 51: substitution.tuple_list_get_set_item_eliminator 0.68% : 0.000022s : 6: substitution.updatestate_pure_node_eliminater 0.67% : 0.000022s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.645997 2 91.77% : 0.592836s : 1: type_inference.infer 8.23% : 0.053161s : 1: type_inference.specialize ------[replace.] 0.001133 115 3.59% : 0.000041s : 5: replace.depend_value_elim 50.74% : 0.000575s : 71: replace.inline 3.37% : 0.000038s : 2: replace.list_to_tuple_eliminator_ 33.79% : 0.000383s : 24: replace.switch_simplify 8.50% : 0.000096s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002267 115 0.09% : 0.000002s : 5: match.depend_value_elim 93.57% : 0.002121s : 71: match.inline 1.88% : 0.000043s : 2: match.list_to_tuple_eliminator_ 2.48% : 0.000056s : 24: match.switch_simplify 1.98% : 0.000045s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001622 9701 1.22% : 0.000020s : 123: predicate.accumulaten_eliminater 0.50% : 0.000008s : 17: predicate.ad_related_special_op_eliminate 0.48% : 0.000008s : 58: predicate.addn_check_dump 1.16% : 0.000019s : 123: predicate.addn_zero_filter 1.03% : 0.000017s : 123: predicate.adjust_all_reduce_mul_add 2.45% : 0.000040s : 181: predicate.arithmetic_simplify 1.15% : 0.000019s : 123: predicate.cast_eliminate 0.30% : 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 0.61% : 0.000010s : 61: predicate.depend_value_elim 1.18% : 0.000019s : 123: predicate.dict_get_item_const_eliminator 1.24% : 0.000020s : 123: predicate.dict_get_item_eliminator 1.10% : 0.000018s : 123: predicate.dict_set_item_eliminator 0.48% : 0.000008s : 34: predicate.dumpgradient_eliminate 0.12% : 0.000002s : 17: predicate.elim_not_effective 0.27% : 0.000004s : 17: predicate.elim_shapecalc_of_broadcastargs 1.26% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.25% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.26% : 0.000020s : 140: predicate.environ_get_depend_swap 1.77% : 0.000029s : 198: predicate.environ_get_eliminate 1.20% : 0.000019s : 140: predicate.environ_get_set_eliminate 1.87% : 0.000030s : 209: predicate.exchange_switch_depend_value 2.51% : 0.000041s : 209: predicate.float_depend_g_call 0.51% : 0.000008s : 58: predicate.float_environ_get_switch 0.71% : 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% : 0.000001s : 17: predicate.graph_param_transform 0.50% : 0.000008s : 58: predicate.incorporate_call 0.45% : 0.000007s : 58: predicate.incorporate_call_switch 5.56% : 0.000090s : 459: predicate.inline 0.41% : 0.000007s : 35: predicate.inline_without_move 0.16% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.46% : 0.000007s : 36: predicate.less_batch_normalization 1.61% : 0.000026s : 172: predicate.list_to_tuple_eliminator_ 2.57% : 0.000042s : 295: predicate.load_eliminater 0.37% : 0.000006s : 17: predicate.loop_unroll_after_grad 3.74% : 0.000061s : 314: predicate.loop_unroll_before_grad 1.46% : 0.000024s : 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 1.15% : 0.000019s : 123: predicate.minmaximum_grad 0.49% : 0.000008s : 17: predicate.mutable_eliminate 0.21% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 2.97% : 0.000048s : 209: predicate.partial_defer_inline 1.48% : 0.000024s : 155: predicate.partial_eliminate 1.12% : 0.000018s : 123: predicate.print_const_string_wrapper 0.46% : 0.000007s : 53: predicate.reduce_all_const_elim 1.64% : 0.000027s : 123: predicate.reduce_eliminate 2.70% : 0.000044s : 295: predicate.redundant_stop_gradient_eliminater 0.20% : 0.000003s : 35: predicate.remove_not_recompute_node 1.16% : 0.000019s : 172: predicate.replace_applicator 0.18% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.27% : 0.000021s : 123: predicate.reshape_eliminate 0.32% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.19% : 0.000003s : 17: predicate.row_tensor_eliminate 0.47% : 0.000008s : 34: predicate.same_eliminate 0.20% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.35% : 0.000006s : 35: predicate.shard_identity_eliminate 0.35% : 0.000006s : 34: predicate.special_op_eliminate 0.58% : 0.000009s : 58: predicate.specialize_transform 0.40% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.38% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.18% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.06% : 0.000033s : 209: predicate.switch_defer_inline 2.79% : 0.000045s : 243: predicate.switch_layer_defer_inline 9.00% : 0.000146s : 646: predicate.switch_simplify 1.18% : 0.000019s : 123: predicate.tile_eliminate 1.11% : 0.000018s : 123: predicate.transpose_eliminate 1.63% : 0.000026s : 157: predicate.tuple_list_convert_item_index_to_positive 1.77% : 0.000029s : 157: predicate.tuple_list_get_item_const_eliminator 1.58% : 0.000026s : 157: predicate.tuple_list_get_item_depend_reorder 2.89% : 0.000047s : 228: predicate.tuple_list_get_item_eliminator 1.56% : 0.000025s : 157: predicate.tuple_list_get_set_item_eliminator 2.28% : 0.000037s : 215: predicate.tuple_list_set_item_eliminator 1.63% : 0.000026s : 170: predicate.tuple_to_list_eliminator_ 2.52% : 0.000041s : 295: predicate.updatestate_pure_node_eliminater 3.04% : 0.000049s : 353: predicate.updatestate_useless_node_eliminater 0.19% : 0.000003s : 17: predicate.value_based_eliminate 0.34% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.34% : 0.000006s : 35: predicate.virtual_output_eliminate 0.15% : 0.000002s : 17: predicate.virtual_view_grad_eliminate 0.19% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.030835 135 77.37% : 0.023857s : 55: func_graph_cloner_run.FuncGraphClonerGraph 22.63% : 0.006978s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.886124 209 0.00% : 0.000004s : 1: ForceFp32Comm 1.66% : 0.014674s : 1: add_attr 1.65% : 0.014654s : 1: add_attr_with_inline 0.00% : 0.000006s : 1: add_comm_op_reuse_tag 0.02% : 0.000145s : 1: add_recomputation 0.00% : 0.000007s : 1: assign_add_opt 0.09% : 0.000764s : 1: auto_monad 0.01% : 0.000091s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: begin_end_overlap_inline 0.00% : 0.000008s : 1: bias_add_comm_swap 0.09% : 0.000786s : 1: bootstrap 0.01% : 0.000047s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000050s : 1: control_data_broadcast_order 0.00% : 0.000033s : 1: convert_after_rewriter 3.21% : 0.028418s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.01% : 0.000052s : 1: environ_conv 0.06% : 0.000495s : 1: event_method 0.00% : 0.000008s : 1: full_micro_interleaved_order_control 0.00% : 0.000009s : 1: get_jit_bprop_graph 0.01% : 0.000059s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000006s : 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% : 0.000005s : 1: interleave_split_concat_branches 0.00% : 0.000007s : 1: label_fine_grained_interleaved_index 0.00% : 0.000012s : 1: label_micro_interleaved_index 0.07% : 0.000606s : 1: loop_unroll 0.00% : 0.000006s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.10% : 0.000868s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.00% : 0.000037s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000048s : 1: opt.transform.mutable_eliminate 2.45% : 0.021675s : 95: opt.transform.opt_a 0.01% : 0.000125s : 1: opt.transform.opt_after_cconv 0.01% : 0.000092s : 1: opt.transform.opt_after_jit_grad 0.06% : 0.000531s : 28: opt.transform.opt_b 0.02% : 0.000214s : 2: opt.transform.opt_trans_graph 0.02% : 0.000215s : 4: opt.transform.symbol_engine_opt 5.23% : 0.046341s : 1: opt_a 0.03% : 0.000294s : 1: opt_after_cconv 0.12% : 0.001038s : 1: opt_after_jit_grad 0.09% : 0.000778s : 1: opt_b 9.15% : 0.081081s : 1: optimize 0.01% : 0.000046s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000019s : 1: order_py_execute_after_rewriter 0.01% : 0.000064s : 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% : 0.000008s : 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% : 0.000015s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000008s : 1: overlap_recompute_comm 0.00% : 0.000010s : 1: parallel-infer-symbol 0.00% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000006s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000269s : 1: pre_auto_parallel 0.03% : 0.000249s : 1: py_interpret_to_execute 0.00% : 0.000038s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000005s : 1: remove_cast_before_assign_add 0.01% : 0.000104s : 1: remove_dup_value 1.72% : 0.015220s : 1: renormalize.infer 0.74% : 0.006563s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000014s : 1: rewriter_after_jit_bprop_graph 0.10% : 0.000882s : 1: rewriter_after_opt_a 0.09% : 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% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000039s : 1: swap_dp_allreduce_reducescatter 0.07% : 0.000653s : 1: symbol_engine_optimizer 0.03% : 0.000252s : 1: tuple_transform 72.95% : 0.646432s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape2-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape2-256],max_mem:212.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:45:07.920.210 [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 = 0.317763, [21] [bootstrap]: 0.0008394 [type_inference]: 0.272795 [event_method]: 0.00020116 [auto_monad]: 0.00072643 [graph_reusing]: 4.852e-05 [inline]: 2.86e-06 [add_attr]: 0.00516451, [1] [add_attr_with_inline]: 0.00514606, [1] [Cycle 1]: 0.00029547, [2] [tag_attr]: 0.00019752 [meta_addattr_fg_expand]: 5.183e-05 [parallel-infer-symbol]: 3.63e-06 [pre_auto_parallel]: 0.00031329 [insert-virtual-dataset]: 2.68e-06 [parallel-infer-symbol-second]: 1.35999e-06 [dataset_repeat_opt]: 2.36998e-06 [pipeline_split]: 1.67001e-06 [optimize]: 0.0364042, [53] [py_interpret_to_execute]: 0.00024304 [rewriter_before_opt_a]: 0.00077724 [opt_a]: 0.0297991, [2] [Cycle 1]: 0.0275268, [45] [expand_dump_flag]: 1.558e-05 [switch_simplify]: 0.00079787 [loop_unroll]: 0.00032346 [a_1]: 0.0092932 [with_stream_mark]: 5.371e-05 [recompute_prepare]: 4.199e-05 [updatestate_depend_eliminate]: 1.961e-05 [updatestate_assign_eliminate]: 1.534e-05 [updatestate_loads_eliminate]: 1.897e-05 [parameter_eliminate]: 4.00998e-06 [a_2]: 0.0008435 [accelerated_algorithm]: 6.94e-05 [shard]: 2.61999e-06 [meta_shard_fg_expand]: 1.903e-05 [shard_inline]: 2.372e-05 [merge_send_recv]: 1.931e-05 [auto_parallel]: 2.04e-05 [parallel]: 3.351e-05 [flash_sp]: 1.598e-05 [merge_comm]: 1.374e-05 [allreduce_fusion]: 1.21e-05 [matmul_add_comm_reduction]: 2.18e-05 [allreduce_slice_to_reducescatter]: 7.99977e-07 [virtual_shard_identity]: 2.82e-05 [virtual_dataset]: 2.215e-05 [get_grad_eliminate_]: 2.073e-05 [virtual_output]: 2.158e-05 [merge_forward]: 1.429e-05 [cell_reuse_recompute_pass]: 1.74998e-06 [offload_activation]: 2.428e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.61e-05 [merge_recompute_call_nodes]: 1.87001e-06 [before_grad]: 3.89e-05 [set_forward_comm_id_for_comm_node_pass]: 1.495e-05 [meta_fg_expand]: 1.69e-05 [flash_sp_send_recv_attached]: 5.42999e-06 [receive_attached]: 2.38998e-06 [after_resolve]: 2.981e-05 [a_after_grad]: 3.512e-05 [renormalize]: 0.0134857 [add_forward_monad_depend]: 2.205e-05 [auto_monad_grad]: 3.68999e-06 [auto_monad_eliminator]: 8.836e-05 [cse]: 0.00128048 [a_3]: 0.00018968 [Cycle 2]: 0.00225441, [45] [expand_dump_flag]: 2.88e-06 [switch_simplify]: 2.847e-05 [loop_unroll]: 2.247e-05 [a_1]: 0.0006663 [with_stream_mark]: 3.799e-05 [recompute_prepare]: 2.66e-05 [updatestate_depend_eliminate]: 1.69e-05 [updatestate_assign_eliminate]: 1.227e-05 [updatestate_loads_eliminate]: 1.426e-05 [parameter_eliminate]: 2.94999e-06 [a_2]: 0.00031901 [accelerated_algorithm]: 3.55e-05 [shard]: 2.28002e-06 [meta_shard_fg_expand]: 9.10001e-06 [shard_inline]: 2.298e-05 [merge_send_recv]: 2.272e-05 [auto_parallel]: 1.996e-05 [parallel]: 1.019e-05 [flash_sp]: 4.74e-06 [merge_comm]: 1.434e-05 [allreduce_fusion]: 1.164e-05 [matmul_add_comm_reduction]: 2.364e-05 [allreduce_slice_to_reducescatter]: 9.69972e-07 [virtual_shard_identity]: 2.74e-05 [virtual_dataset]: 2.215e-05 [get_grad_eliminate_]: 2.136e-05 [virtual_output]: 2.03e-05 [merge_forward]: 1.327e-05 [cell_reuse_recompute_pass]: 3.4e-06 [offload_activation]: 2.658e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.614e-05 [merge_recompute_call_nodes]: 1.50999e-06 [before_grad]: 3.696e-05 [set_forward_comm_id_for_comm_node_pass]: 1.655e-05 [meta_fg_expand]: 9.77999e-06 [flash_sp_send_recv_attached]: 2.49999e-06 [receive_attached]: 2.86e-06 [after_resolve]: 3.374e-05 [a_after_grad]: 3.366e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 4.70999e-06 [auto_monad_grad]: 3.11999e-06 [auto_monad_eliminator]: 7.254e-05 [cse]: 9.116e-05 [a_3]: 0.00014218 [py_interpret_to_execute_after_opt_a]: 3.688e-05 [slice_cell_reuse_recomputed_activation]: 2.67001e-06 [rewriter_after_opt_a]: 0.00092925 [convert_after_rewriter]: 5.218e-05 [order_py_execute_after_rewriter]: 1.744e-05 [mutable_eliminate]: 0.0009163 [opt_b]: 0.00077863, [1] [Cycle 1]: 0.00076823, [7] [b_1]: 0.00054085 [b_2]: 2.67e-05 [updatestate_depend_eliminate]: 2.058e-05 [updatestate_assign_eliminate]: 1.146e-05 [updatestate_loads_eliminate]: 1.469e-05 [renormalize]: 1.06997e-06 [cse]: 9.281e-05 [optimize_parallel_all_gather_comm]: 4.173e-05 [overlap_param_gather]: 1.99999e-06 [cconv]: 4.22e-05 [loop_unroll]: 0.00065359 [opt_after_cconv]: 0.00030032, [1] [Cycle 1]: 0.00029075, [7] [c_1]: 0.00012484 [parameter_eliminate]: 5.47001e-06 [updatestate_depend_eliminate]: 1.818e-05 [updatestate_assign_eliminate]: 1.08e-05 [updatestate_loads_eliminate]: 1.299e-05 [cse]: 7.542e-05 [renormalize]: 1.05999e-06 [remove_dup_value]: 0.000103 [tuple_transform]: 0.00028533, [1] [Cycle 1]: 0.00027849, [4] [d_1]: 0.00021804 [none_parameter_eliminate]: 3.21999e-06 [renormalize]: 5.60016e-07 [switch_simplify]: 2.488e-05 [partial_unused_args_eliminate]: 2.82002e-06 [add_recomputation]: 0.00015175 [cse_after_recomputation]: 7.646e-05, [1] [Cycle 1]: 6.886e-05, [1] [cse]: 6.079e-05 [environ_conv]: 3.184e-05 [swap_dp_allreduce_reducescatter]: 1.924e-05 [bias_add_comm_swap]: 4.05998e-06 [label_micro_interleaved_index]: 6.71999e-06 [label_fine_grained_interleaved_index]: 2.93e-06 [merge_cast_opt]: 1.39e-06 [slice_recompute_activation]: 2.66999e-06 [micro_interleaved_order_control]: 2.51e-06 [assign_add_opt]: 1.57001e-06 [ForceFp32Comm]: 8.89995e-07 [remove_cast_before_assign_add]: 1.22999e-06 [full_micro_interleaved_order_control]: 2.83e-06 [reorder_send_recv_between_fp_bp]: 2.98e-06 [comm_op_add_attrs]: 1.76e-06 [add_comm_op_reuse_tag]: 1.11002e-06 [interleave_split_concat_branches]: 1.20999e-06 [interleave_parallel_branches]: 1.35999e-06 [overlap_opt_shard_in_pipeline]: 1.50999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.22999e-06 [control_data_broadcast_order]: 4.04e-05 [grouped_pairwise_exchange_alltoall]: 1.86e-06 [offloading_packed_experts]: 1.177e-05 [overlap_recompute_and_grad_model_parallel]: 1.01e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.22999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40999e-06 [overlap_recompute_comm]: 2.49001e-06 [overlap_grad_ring_attention]: 9.41e-06 [overlap_grad_flash_sp]: 5.294e-05 [begin_end_overlap_inline]: 5.59987e-07 [split_matmul_comm_elemetwise]: 2.52001e-06 [split_layernorm_comm]: 2.29001e-06 [handle_group_info]: 1.37999e-06 [symbol_engine_optimizer]: 0.00061917, [1] [Cycle 1]: 0.00061043, [6] [build]: 0.00034528 [elim_shapecalc]: 4.204e-05 [elim_not_effective]: 5.733e-05 [opt_reshape]: 3.815e-05 [fold_const_symbol]: 6.279e-05 [renormalize]: 9.00007e-07 [detach_backward]: 3.06999e-06 [pipeline_parallel_scheduler]: 1.60001e-06 [auto_monad_reorder]: 7.011e-05 [get_jit_bprop_graph]: 2.66999e-06 [rewriter_after_jit_bprop_graph]: 7.16999e-06 [opt_after_jit_grad]: 0.00078351 [validate]: 0.00011231 Sums bootstrap : 0.000839s : 0.27% type_inference : 0.272795s : 87.66% event_method : 0.000201s : 0.06% auto_monad : 0.000726s : 0.23% graph_reusing : 0.000049s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000198s : 0.06% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000052s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000313s : 0.10% 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 : 0.08% optimize.rewriter_before_opt_a : 0.000777s : 0.25% optimize.opt_a.expand_dump_flag : 0.000018s : 0.01% optimize.opt_a.switch_simplify : 0.000826s : 0.27% optimize.opt_a.loop_unroll : 0.000346s : 0.11% optimize.opt_a.a_1 : 0.009959s : 3.20% optimize.opt_a.with_stream_mark : 0.000092s : 0.03% optimize.opt_a.recompute_prepare : 0.000069s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000037s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000028s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000033s : 0.01% optimize.opt_a.parameter_eliminate : 0.000007s : 0.00% optimize.opt_a.a_2 : 0.001163s : 0.37% optimize.opt_a.accelerated_algorithm : 0.000105s : 0.03% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000028s : 0.01% optimize.opt_a.shard_inline : 0.000047s : 0.02% optimize.opt_a.merge_send_recv : 0.000042s : 0.01% optimize.opt_a.auto_parallel : 0.000040s : 0.01% optimize.opt_a.parallel : 0.000044s : 0.01% optimize.opt_a.flash_sp : 0.000021s : 0.01% optimize.opt_a.merge_comm : 0.000028s : 0.01% optimize.opt_a.allreduce_fusion : 0.000024s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000045s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000056s : 0.02% optimize.opt_a.virtual_dataset : 0.000044s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.01% optimize.opt_a.virtual_output : 0.000042s : 0.01% optimize.opt_a.merge_forward : 0.000028s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000051s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000092s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000076s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000032s : 0.01% optimize.opt_a.meta_fg_expand : 0.000027s : 0.01% 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 : 0.000064s : 0.02% optimize.opt_a.a_after_grad : 0.000069s : 0.02% optimize.opt_a.renormalize : 0.013486s : 4.33% optimize.opt_a.add_forward_monad_depend : 0.000027s : 0.01% optimize.opt_a.auto_monad_grad : 0.000007s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000161s : 0.05% optimize.opt_a.cse : 0.001372s : 0.44% optimize.opt_a.a_3 : 0.000332s : 0.11% optimize.py_interpret_to_execute_after_opt_a : 0.000037s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000929s : 0.30% optimize.convert_after_rewriter : 0.000052s : 0.02% optimize.order_py_execute_after_rewriter : 0.000017s : 0.01% optimize.mutable_eliminate : 0.000916s : 0.29% optimize.opt_b.b_1 : 0.000541s : 0.17% optimize.opt_b.b_2 : 0.000027s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000021s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000015s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000093s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000042s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000042s : 0.01% optimize.loop_unroll : 0.000654s : 0.21% optimize.opt_after_cconv.c_1 : 0.000125s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000018s : 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 : 0.000075s : 0.02% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000103s : 0.03% optimize.tuple_transform.d_1 : 0.000218s : 0.07% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000001s : 0.00% optimize.tuple_transform.switch_simplify : 0.000025s : 0.01% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000152s : 0.05% optimize.cse_after_recomputation.cse : 0.000061s : 0.02% optimize.environ_conv : 0.000032s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000019s : 0.01% 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 : 0.000003s : 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 : 0.000002s : 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 : 0.000002s : 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 : 0.000001s : 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 : 0.000053s : 0.02% 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 : 0.000345s : 0.11% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000042s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000057s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000038s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000063s : 0.02% 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 : 0.000070s : 0.02% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000784s : 0.25% validate : 0.000112s : 0.04% Time group info: ------[substitution.] 0.003439 569 0.49% : 0.000017s : 8: substitution.depend_value_elim 0.43% : 0.000015s : 14: substitution.elim_not_effective 0.50% : 0.000017s : 12: substitution.float_tuple_getitem_switch 0.85% : 0.000029s : 14: substitution.fold_const_symbol 0.47% : 0.000016s : 17: substitution.graph_param_transform 69.03% : 0.002374s : 71: substitution.inline 0.42% : 0.000015s : 29: substitution.j_node_and_user_rematch 1.06% : 0.000036s : 6: substitution.less_batch_normalization 1.62% : 0.000056s : 2: substitution.list_to_tuple_eliminator_ 0.19% : 0.000007s : 6: substitution.load_eliminater 0.89% : 0.000031s : 21: substitution.minmaximum_grad 0.27% : 0.000009s : 2: substitution.opt_reshape 0.59% : 0.000020s : 29: substitution.remove_not_recompute_node 0.26% : 0.000009s : 4: substitution.replace_old_param 1.76% : 0.000061s : 18: substitution.reshape_eliminate 1.45% : 0.000050s : 24: substitution.switch_simplify 4.27% : 0.000147s : 51: substitution.tuple_list_convert_item_index_to_positive 2.25% : 0.000077s : 51: substitution.tuple_list_get_item_const_eliminator 2.96% : 0.000102s : 51: substitution.tuple_list_get_item_depend_reorder 6.07% : 0.000209s : 73: substitution.tuple_list_get_item_eliminator 2.88% : 0.000099s : 51: substitution.tuple_list_get_set_item_eliminator 0.65% : 0.000022s : 6: substitution.updatestate_pure_node_eliminater 0.62% : 0.000021s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.272476 2 93.35% : 0.254360s : 1: type_inference.infer 6.65% : 0.018116s : 1: type_inference.specialize ------[replace.] 0.001169 115 4.22% : 0.000049s : 5: replace.depend_value_elim 58.44% : 0.000683s : 71: replace.inline 3.47% : 0.000041s : 2: replace.list_to_tuple_eliminator_ 25.65% : 0.000300s : 24: replace.switch_simplify 8.22% : 0.000096s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002450 115 0.11% : 0.000003s : 5: match.depend_value_elim 94.95% : 0.002326s : 71: match.inline 2.17% : 0.000053s : 2: match.list_to_tuple_eliminator_ 1.45% : 0.000036s : 24: match.switch_simplify 1.31% : 0.000032s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001606 9701 1.24% : 0.000020s : 123: predicate.accumulaten_eliminater 0.40% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.50% : 0.000008s : 58: predicate.addn_check_dump 1.15% : 0.000018s : 123: predicate.addn_zero_filter 1.11% : 0.000018s : 123: predicate.adjust_all_reduce_mul_add 2.31% : 0.000037s : 181: predicate.arithmetic_simplify 1.18% : 0.000019s : 123: predicate.cast_eliminate 0.33% : 0.000005s : 34: predicate.check_bprop_eliminate 0.51% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.60% : 0.000010s : 61: predicate.depend_value_elim 1.29% : 0.000021s : 123: predicate.dict_get_item_const_eliminator 1.35% : 0.000022s : 123: predicate.dict_get_item_eliminator 1.16% : 0.000019s : 123: predicate.dict_set_item_eliminator 0.43% : 0.000007s : 34: predicate.dumpgradient_eliminate 0.11% : 0.000002s : 17: predicate.elim_not_effective 0.27% : 0.000004s : 17: predicate.elim_shapecalc_of_broadcastargs 1.29% : 0.000021s : 140: predicate.environ_add_const_eliminate 1.26% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.40% : 0.000023s : 140: predicate.environ_get_depend_swap 1.91% : 0.000031s : 198: predicate.environ_get_eliminate 1.22% : 0.000020s : 140: predicate.environ_get_set_eliminate 1.99% : 0.000032s : 209: predicate.exchange_switch_depend_value 3.00% : 0.000048s : 209: predicate.float_depend_g_call 0.50% : 0.000008s : 58: predicate.float_environ_get_switch 0.70% : 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 0.11% : 0.000002s : 17: predicate.graph_param_transform 0.51% : 0.000008s : 58: predicate.incorporate_call 0.47% : 0.000008s : 58: predicate.incorporate_call_switch 5.67% : 0.000091s : 459: predicate.inline 0.43% : 0.000007s : 35: predicate.inline_without_move 0.16% : 0.000003s : 35: predicate.j_node_and_user_rematch 0.48% : 0.000008s : 36: predicate.less_batch_normalization 1.80% : 0.000029s : 172: predicate.list_to_tuple_eliminator_ 2.88% : 0.000046s : 295: predicate.load_eliminater 0.43% : 0.000007s : 17: predicate.loop_unroll_after_grad 3.22% : 0.000052s : 314: predicate.loop_unroll_before_grad 1.49% : 0.000024s : 157: predicate.make_slice_get_slice_eliminator 0.54% : 0.000009s : 58: predicate.merge_addn 0.29% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.30% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.19% : 0.000019s : 123: predicate.minmaximum_grad 0.36% : 0.000006s : 17: predicate.mutable_eliminate 0.19% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 3.21% : 0.000052s : 209: predicate.partial_defer_inline 1.54% : 0.000025s : 155: predicate.partial_eliminate 1.19% : 0.000019s : 123: predicate.print_const_string_wrapper 0.51% : 0.000008s : 53: predicate.reduce_all_const_elim 1.63% : 0.000026s : 123: predicate.reduce_eliminate 2.63% : 0.000042s : 295: predicate.redundant_stop_gradient_eliminater 0.20% : 0.000003s : 35: predicate.remove_not_recompute_node 1.11% : 0.000018s : 172: predicate.replace_applicator 0.19% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000002s : 17: predicate.reset_defer_inline 1.41% : 0.000023s : 123: predicate.reshape_eliminate 0.32% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.22% : 0.000004s : 17: predicate.row_tensor_eliminate 0.59% : 0.000009s : 34: predicate.same_eliminate 0.27% : 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 0.59% : 0.000009s : 58: predicate.specialize_transform 0.41% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.41% : 0.000007s : 35: predicate.stack_unstack_eliminate 0.18% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.09% : 0.000034s : 209: predicate.switch_defer_inline 2.53% : 0.000041s : 243: predicate.switch_layer_defer_inline 6.53% : 0.000105s : 646: predicate.switch_simplify 1.18% : 0.000019s : 123: predicate.tile_eliminate 1.14% : 0.000018s : 123: predicate.transpose_eliminate 1.74% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.74% : 0.000028s : 157: predicate.tuple_list_get_item_const_eliminator 1.64% : 0.000026s : 157: predicate.tuple_list_get_item_depend_reorder 2.93% : 0.000047s : 228: predicate.tuple_list_get_item_eliminator 1.72% : 0.000028s : 157: predicate.tuple_list_get_set_item_eliminator 2.24% : 0.000036s : 215: predicate.tuple_list_set_item_eliminator 1.76% : 0.000028s : 170: predicate.tuple_to_list_eliminator_ 2.54% : 0.000041s : 295: predicate.updatestate_pure_node_eliminater 3.16% : 0.000051s : 353: predicate.updatestate_useless_node_eliminater 0.18% : 0.000003s : 17: predicate.value_based_eliminate 0.39% : 0.000006s : 35: predicate.virtual_dataset_eliminate 0.33% : 0.000005s : 35: predicate.virtual_output_eliminate 0.16% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.19% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.015118 135 55.54% : 0.008397s : 55: func_graph_cloner_run.FuncGraphClonerGraph 44.46% : 0.006721s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.386964 209 0.00% : 0.000004s : 1: ForceFp32Comm 1.34% : 0.005171s : 1: add_attr 1.33% : 0.005153s : 1: add_attr_with_inline 0.00% : 0.000006s : 1: add_comm_op_reuse_tag 0.04% : 0.000160s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.19% : 0.000746s : 1: auto_monad 0.02% : 0.000078s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.23% : 0.000901s : 1: bootstrap 0.01% : 0.000048s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.01% : 0.000046s : 1: control_data_broadcast_order 0.02% : 0.000060s : 1: convert_after_rewriter 0.02% : 0.000080s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.01% : 0.000036s : 1: environ_conv 0.06% : 0.000214s : 1: event_method 0.00% : 0.000008s : 1: full_micro_interleaved_order_control 0.00% : 0.000009s : 1: get_jit_bprop_graph 0.01% : 0.000057s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000007s : 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% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.17% : 0.000669s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.24% : 0.000933s : 1: mutable_eliminate 0.00% : 0.000015s : 1: offloading_packed_experts 0.01% : 0.000041s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000043s : 1: opt.transform.mutable_eliminate 3.40% : 0.013162s : 95: opt.transform.opt_a 0.03% : 0.000123s : 1: opt.transform.opt_after_cconv 0.02% : 0.000087s : 1: opt.transform.opt_after_jit_grad 0.14% : 0.000527s : 28: opt.transform.opt_b 0.06% : 0.000239s : 2: opt.transform.opt_trans_graph 0.05% : 0.000192s : 4: opt.transform.symbol_engine_opt 7.70% : 0.029805s : 1: opt_a 0.08% : 0.000305s : 1: opt_after_cconv 0.21% : 0.000802s : 1: opt_after_jit_grad 0.20% : 0.000783s : 1: opt_b 9.41% : 0.036412s : 1: optimize 0.01% : 0.000046s : 1: optimize_parallel_all_gather_comm 0.01% : 0.000022s : 1: order_py_execute_after_rewriter 0.01% : 0.000058s : 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% : 0.000005s : 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% : 0.000013s : 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% : 0.000007s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.08% : 0.000323s : 1: pre_auto_parallel 0.07% : 0.000252s : 1: py_interpret_to_execute 0.01% : 0.000042s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.03% : 0.000111s : 1: remove_dup_value 1.79% : 0.006925s : 1: renormalize.infer 1.69% : 0.006540s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.24% : 0.000945s : 1: rewriter_after_opt_a 0.20% : 0.000790s : 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% : 0.000006s : 1: split_matmul_comm_elemetwise 0.01% : 0.000024s : 1: swap_dp_allreduce_reducescatter 0.16% : 0.000624s : 1: symbol_engine_optimizer 0.08% : 0.000290s : 1: tuple_transform 70.51% : 0.272830s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape2-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape2-1024],max_mem:260.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:45:16.844.697 [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 = 0.303244, [21] [bootstrap]: 0.00063355 [type_inference]: 0.263013 [event_method]: 0.00020591 [auto_monad]: 0.00069205 [graph_reusing]: 4.686e-05 [inline]: 3.58999e-06 [add_attr]: 0.00499288, [1] [add_attr_with_inline]: 0.00497573, [1] [Cycle 1]: 0.00030292, [2] [tag_attr]: 0.00019716 [meta_addattr_fg_expand]: 5.653e-05 [parallel-infer-symbol]: 4.03999e-06 [pre_auto_parallel]: 0.00024979 [insert-virtual-dataset]: 2.39001e-06 [parallel-infer-symbol-second]: 1.10001e-06 [dataset_repeat_opt]: 2.04999e-06 [pipeline_split]: 1.66998e-06 [optimize]: 0.0324451, [53] [py_interpret_to_execute]: 0.00023006 [rewriter_before_opt_a]: 0.00076156 [opt_a]: 0.026832, [2] [Cycle 1]: 0.0247371, [45] [expand_dump_flag]: 1.838e-05 [switch_simplify]: 0.00078693 [loop_unroll]: 0.00032385 [a_1]: 0.00868346 [with_stream_mark]: 6.159e-05 [recompute_prepare]: 3.889e-05 [updatestate_depend_eliminate]: 2.148e-05 [updatestate_assign_eliminate]: 1.602e-05 [updatestate_loads_eliminate]: 1.812e-05 [parameter_eliminate]: 2.80002e-06 [a_2]: 0.00078768 [accelerated_algorithm]: 5.7e-05 [shard]: 1.99e-06 [meta_shard_fg_expand]: 1.791e-05 [shard_inline]: 2.157e-05 [merge_send_recv]: 1.858e-05 [auto_parallel]: 1.96e-05 [parallel]: 3.466e-05 [flash_sp]: 1.471e-05 [merge_comm]: 1.273e-05 [allreduce_fusion]: 1.208e-05 [matmul_add_comm_reduction]: 2.03e-05 [allreduce_slice_to_reducescatter]: 9.00007e-07 [virtual_shard_identity]: 2.56e-05 [virtual_dataset]: 2.14e-05 [get_grad_eliminate_]: 2.008e-05 [virtual_output]: 2.02e-05 [merge_forward]: 1.173e-05 [cell_reuse_recompute_pass]: 1.97001e-06 [offload_activation]: 2.194e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.194e-05 [merge_recompute_call_nodes]: 1.40999e-06 [before_grad]: 3.713e-05 [set_forward_comm_id_for_comm_node_pass]: 1.302e-05 [meta_fg_expand]: 1.545e-05 [flash_sp_send_recv_attached]: 5.05001e-06 [receive_attached]: 2.48998e-06 [after_resolve]: 2.651e-05 [a_after_grad]: 3.428e-05 [renormalize]: 0.0118181 [add_forward_monad_depend]: 1.866e-05 [auto_monad_grad]: 2.60997e-06 [auto_monad_eliminator]: 7.402e-05 [cse]: 0.00097741 [a_3]: 0.00017012 [Cycle 2]: 0.00207785, [45] [expand_dump_flag]: 2.70002e-06 [switch_simplify]: 2.538e-05 [loop_unroll]: 2.051e-05 [a_1]: 0.00064057 [with_stream_mark]: 2.77e-05 [recompute_prepare]: 2.032e-05 [updatestate_depend_eliminate]: 1.446e-05 [updatestate_assign_eliminate]: 1.088e-05 [updatestate_loads_eliminate]: 1.344e-05 [parameter_eliminate]: 2.48002e-06 [a_2]: 0.00030035 [accelerated_algorithm]: 2.97e-05 [shard]: 1.97999e-06 [meta_shard_fg_expand]: 6.07999e-06 [shard_inline]: 5.62e-05 [merge_send_recv]: 1.889e-05 [auto_parallel]: 4.745e-05 [parallel]: 9.96e-06 [flash_sp]: 4.34997e-06 [merge_comm]: 1.358e-05 [allreduce_fusion]: 1.234e-05 [matmul_add_comm_reduction]: 2.078e-05 [allreduce_slice_to_reducescatter]: 1.22e-06 [virtual_shard_identity]: 2.564e-05 [virtual_dataset]: 2.043e-05 [get_grad_eliminate_]: 2.151e-05 [virtual_output]: 1.976e-05 [merge_forward]: 1.168e-05 [cell_reuse_recompute_pass]: 2.86e-06 [offload_activation]: 2.426e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.972e-05 [merge_recompute_call_nodes]: 1.59e-06 [before_grad]: 3.485e-05 [set_forward_comm_id_for_comm_node_pass]: 1.338e-05 [meta_fg_expand]: 1.07e-05 [flash_sp_send_recv_attached]: 1.77001e-06 [receive_attached]: 2.44999e-06 [after_resolve]: 2.885e-05 [a_after_grad]: 3.149e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.76999e-06 [auto_monad_grad]: 2.24001e-06 [auto_monad_eliminator]: 4.399e-05 [cse]: 7.531e-05 [a_3]: 0.00013509 [py_interpret_to_execute_after_opt_a]: 2.938e-05 [slice_cell_reuse_recomputed_activation]: 2.34001e-06 [rewriter_after_opt_a]: 0.00079667 [convert_after_rewriter]: 1.987e-05 [order_py_execute_after_rewriter]: 1.378e-05 [mutable_eliminate]: 0.00074979 [opt_b]: 0.00069613, [1] [Cycle 1]: 0.00068808, [7] [b_1]: 0.00051673 [b_2]: 2.197e-05 [updatestate_depend_eliminate]: 1.54e-05 [updatestate_assign_eliminate]: 1.102e-05 [updatestate_loads_eliminate]: 1.264e-05 [renormalize]: 7.10017e-07 [cse]: 6.943e-05 [optimize_parallel_all_gather_comm]: 3.512e-05 [overlap_param_gather]: 5.72001e-06 [cconv]: 3.548e-05 [loop_unroll]: 0.00048373 [opt_after_cconv]: 0.00028071, [1] [Cycle 1]: 0.00027271, [7] [c_1]: 0.00011884 [parameter_eliminate]: 3.00998e-06 [updatestate_depend_eliminate]: 1.626e-05 [updatestate_assign_eliminate]: 1.05e-05 [updatestate_loads_eliminate]: 1.17e-05 [cse]: 6.656e-05 [renormalize]: 4.50003e-07 [remove_dup_value]: 8.493e-05 [tuple_transform]: 0.00023776, [1] [Cycle 1]: 0.00023205, [4] [d_1]: 0.00018991 [none_parameter_eliminate]: 2.43e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 2.078e-05 [partial_unused_args_eliminate]: 1.82001e-06 [add_recomputation]: 0.00013319 [cse_after_recomputation]: 6.376e-05, [1] [Cycle 1]: 5.772e-05, [1] [cse]: 5.139e-05 [environ_conv]: 2.328e-05 [swap_dp_allreduce_reducescatter]: 1.821e-05 [bias_add_comm_swap]: 3.44001e-06 [label_micro_interleaved_index]: 5.35001e-06 [label_fine_grained_interleaved_index]: 2.89999e-06 [merge_cast_opt]: 1.45001e-06 [slice_recompute_activation]: 2.01e-06 [micro_interleaved_order_control]: 2.59001e-06 [assign_add_opt]: 1.59e-06 [ForceFp32Comm]: 8.80013e-07 [remove_cast_before_assign_add]: 1.14e-06 [full_micro_interleaved_order_control]: 2.62001e-06 [reorder_send_recv_between_fp_bp]: 3.03998e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 1.06002e-06 [interleave_split_concat_branches]: 1.19003e-06 [interleave_parallel_branches]: 1.12e-06 [overlap_opt_shard_in_pipeline]: 3.79002e-06 [overlap_opt_shard_grad_in_pipeline]: 1.81e-06 [control_data_broadcast_order]: 3.458e-05 [grouped_pairwise_exchange_alltoall]: 1.55999e-06 [offloading_packed_experts]: 1.018e-05 [overlap_recompute_and_grad_model_parallel]: 9.71998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.21002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37999e-06 [overlap_recompute_comm]: 2.30002e-06 [overlap_grad_ring_attention]: 8.85001e-06 [overlap_grad_flash_sp]: 4.962e-05 [begin_end_overlap_inline]: 7.00005e-07 [split_matmul_comm_elemetwise]: 2.25002e-06 [split_layernorm_comm]: 1.96e-06 [handle_group_info]: 1.19e-06 [symbol_engine_optimizer]: 0.00047466, [1] [Cycle 1]: 0.00046888, [6] [build]: 0.00026049 [elim_shapecalc]: 2.994e-05 [elim_not_effective]: 4.944e-05 [opt_reshape]: 3.211e-05 [fold_const_symbol]: 5.546e-05 [renormalize]: 3.39991e-07 [detach_backward]: 2.41e-06 [pipeline_parallel_scheduler]: 1.60001e-06 [auto_monad_reorder]: 5.564e-05 [get_jit_bprop_graph]: 1.81e-06 [rewriter_after_jit_bprop_graph]: 3.78999e-06 [opt_after_jit_grad]: 0.00056437 [validate]: 9.016e-05 Sums bootstrap : 0.000634s : 0.21% type_inference : 0.263013s : 88.50% event_method : 0.000206s : 0.07% auto_monad : 0.000692s : 0.23% graph_reusing : 0.000047s : 0.02% inline : 0.000004s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000197s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000057s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000250s : 0.08% 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 : 0.000230s : 0.08% optimize.rewriter_before_opt_a : 0.000762s : 0.26% optimize.opt_a.expand_dump_flag : 0.000021s : 0.01% optimize.opt_a.switch_simplify : 0.000812s : 0.27% optimize.opt_a.loop_unroll : 0.000344s : 0.12% optimize.opt_a.a_1 : 0.009324s : 3.14% optimize.opt_a.with_stream_mark : 0.000089s : 0.03% optimize.opt_a.recompute_prepare : 0.000059s : 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 : 0.001088s : 0.37% optimize.opt_a.accelerated_algorithm : 0.000087s : 0.03% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000024s : 0.01% optimize.opt_a.shard_inline : 0.000078s : 0.03% optimize.opt_a.merge_send_recv : 0.000037s : 0.01% optimize.opt_a.auto_parallel : 0.000067s : 0.02% optimize.opt_a.parallel : 0.000045s : 0.02% optimize.opt_a.flash_sp : 0.000019s : 0.01% optimize.opt_a.merge_comm : 0.000026s : 0.01% optimize.opt_a.allreduce_fusion : 0.000024s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000041s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000051s : 0.02% optimize.opt_a.virtual_dataset : 0.000042s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.01% optimize.opt_a.virtual_output : 0.000040s : 0.01% optimize.opt_a.merge_forward : 0.000023s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000046s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000082s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000072s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000026s : 0.01% optimize.opt_a.meta_fg_expand : 0.000026s : 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 : 0.000055s : 0.02% optimize.opt_a.a_after_grad : 0.000066s : 0.02% optimize.opt_a.renormalize : 0.011818s : 3.98% optimize.opt_a.add_forward_monad_depend : 0.000021s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000118s : 0.04% optimize.opt_a.cse : 0.001053s : 0.35% optimize.opt_a.a_3 : 0.000305s : 0.10% optimize.py_interpret_to_execute_after_opt_a : 0.000029s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000797s : 0.27% optimize.convert_after_rewriter : 0.000020s : 0.01% optimize.order_py_execute_after_rewriter : 0.000014s : 0.00% optimize.mutable_eliminate : 0.000750s : 0.25% optimize.opt_b.b_1 : 0.000517s : 0.17% optimize.opt_b.b_2 : 0.000022s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000015s : 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 : 0.000069s : 0.02% optimize.optimize_parallel_all_gather_comm : 0.000035s : 0.01% optimize.overlap_param_gather : 0.000006s : 0.00% optimize.cconv : 0.000035s : 0.01% optimize.loop_unroll : 0.000484s : 0.16% optimize.opt_after_cconv.c_1 : 0.000119s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000016s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000012s : 0.00% optimize.opt_after_cconv.cse : 0.000067s : 0.02% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000085s : 0.03% optimize.tuple_transform.d_1 : 0.000190s : 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 : 0.000133s : 0.04% optimize.cse_after_recomputation.cse : 0.000051s : 0.02% optimize.environ_conv : 0.000023s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000018s : 0.01% 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 : 0.000035s : 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 : 0.000050s : 0.02% 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 : 0.000260s : 0.09% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000030s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000049s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000032s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000055s : 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 : 0.000056s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000564s : 0.19% validate : 0.000090s : 0.03% Time group info: ------[substitution.] 0.003145 569 0.52% : 0.000016s : 8: substitution.depend_value_elim 0.44% : 0.000014s : 14: substitution.elim_not_effective 0.52% : 0.000016s : 12: substitution.float_tuple_getitem_switch 0.78% : 0.000024s : 14: substitution.fold_const_symbol 0.47% : 0.000015s : 17: substitution.graph_param_transform 68.99% : 0.002170s : 71: substitution.inline 0.42% : 0.000013s : 29: substitution.j_node_and_user_rematch 0.98% : 0.000031s : 6: substitution.less_batch_normalization 1.48% : 0.000046s : 2: substitution.list_to_tuple_eliminator_ 0.17% : 0.000005s : 6: substitution.load_eliminater 0.89% : 0.000028s : 21: substitution.minmaximum_grad 0.25% : 0.000008s : 2: substitution.opt_reshape 0.60% : 0.000019s : 29: substitution.remove_not_recompute_node 0.21% : 0.000007s : 4: substitution.replace_old_param 1.79% : 0.000056s : 18: substitution.reshape_eliminate 1.50% : 0.000047s : 24: substitution.switch_simplify 4.40% : 0.000138s : 51: substitution.tuple_list_convert_item_index_to_positive 2.37% : 0.000074s : 51: substitution.tuple_list_get_item_const_eliminator 3.15% : 0.000099s : 51: substitution.tuple_list_get_item_depend_reorder 5.78% : 0.000182s : 73: substitution.tuple_list_get_item_eliminator 3.09% : 0.000097s : 51: substitution.tuple_list_get_set_item_eliminator 0.60% : 0.000019s : 6: substitution.updatestate_pure_node_eliminater 0.61% : 0.000019s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.262666 2 93.44% : 0.245441s : 1: type_inference.infer 6.56% : 0.017225s : 1: type_inference.specialize ------[replace.] 0.001012 115 4.04% : 0.000041s : 5: replace.depend_value_elim 55.07% : 0.000558s : 71: replace.inline 3.40% : 0.000034s : 2: replace.list_to_tuple_eliminator_ 28.29% : 0.000286s : 24: replace.switch_simplify 9.20% : 0.000093s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002238 115 0.10% : 0.000002s : 5: match.depend_value_elim 95.16% : 0.002130s : 71: match.inline 1.96% : 0.000044s : 2: match.list_to_tuple_eliminator_ 1.48% : 0.000033s : 24: match.switch_simplify 1.30% : 0.000029s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001554 9701 1.22% : 0.000019s : 123: predicate.accumulaten_eliminater 0.35% : 0.000006s : 17: predicate.ad_related_special_op_eliminate 0.53% : 0.000008s : 58: predicate.addn_check_dump 1.19% : 0.000019s : 123: predicate.addn_zero_filter 1.10% : 0.000017s : 123: predicate.adjust_all_reduce_mul_add 2.31% : 0.000036s : 181: predicate.arithmetic_simplify 1.20% : 0.000019s : 123: predicate.cast_eliminate 0.33% : 0.000005s : 34: predicate.check_bprop_eliminate 0.52% : 0.000008s : 58: predicate.compare_switch_simplify 0.09% : 0.000001s : 17: predicate.const_output_eliminate 0.60% : 0.000009s : 61: predicate.depend_value_elim 1.24% : 0.000019s : 123: predicate.dict_get_item_const_eliminator 1.27% : 0.000020s : 123: predicate.dict_get_item_eliminator 1.16% : 0.000018s : 123: predicate.dict_set_item_eliminator 0.37% : 0.000006s : 34: predicate.dumpgradient_eliminate 0.10% : 0.000002s : 17: predicate.elim_not_effective 0.19% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.32% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.26% : 0.000020s : 140: predicate.environ_get_add_eliminate 1.26% : 0.000020s : 140: predicate.environ_get_depend_swap 1.79% : 0.000028s : 198: predicate.environ_get_eliminate 1.43% : 0.000022s : 140: predicate.environ_get_set_eliminate 1.95% : 0.000030s : 209: predicate.exchange_switch_depend_value 2.61% : 0.000041s : 209: predicate.float_depend_g_call 0.51% : 0.000008s : 58: predicate.float_environ_get_switch 0.74% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.08% : 0.000001s : 17: predicate.fold_const_symbol 0.34% : 0.000005s : 35: predicate.get_grad_eliminate 0.11% : 0.000002s : 17: predicate.graph_param_transform 0.50% : 0.000008s : 58: predicate.incorporate_call 0.48% : 0.000007s : 58: predicate.incorporate_call_switch 5.89% : 0.000092s : 459: predicate.inline 0.41% : 0.000006s : 35: predicate.inline_without_move 0.16% : 0.000002s : 35: predicate.j_node_and_user_rematch 0.52% : 0.000008s : 36: predicate.less_batch_normalization 1.67% : 0.000026s : 172: predicate.list_to_tuple_eliminator_ 2.74% : 0.000043s : 295: predicate.load_eliminater 0.33% : 0.000005s : 17: predicate.loop_unroll_after_grad 3.47% : 0.000054s : 314: predicate.loop_unroll_before_grad 1.58% : 0.000024s : 157: predicate.make_slice_get_slice_eliminator 0.54% : 0.000008s : 58: predicate.merge_addn 0.30% : 0.000005s : 34: predicate.micro_step_allgather_replace 0.31% : 0.000005s : 34: predicate.mini_step_allgather_replace 1.19% : 0.000018s : 123: predicate.minmaximum_grad 0.38% : 0.000006s : 17: predicate.mutable_eliminate 0.18% : 0.000003s : 17: predicate.opt_reshape 0.19% : 0.000003s : 17: predicate.parallel_virtual_node 3.12% : 0.000049s : 209: predicate.partial_defer_inline 1.59% : 0.000025s : 155: predicate.partial_eliminate 1.16% : 0.000018s : 123: predicate.print_const_string_wrapper 0.47% : 0.000007s : 53: predicate.reduce_all_const_elim 1.65% : 0.000026s : 123: predicate.reduce_eliminate 2.72% : 0.000042s : 295: predicate.redundant_stop_gradient_eliminater 0.19% : 0.000003s : 35: predicate.remove_not_recompute_node 1.11% : 0.000017s : 172: predicate.replace_applicator 0.21% : 0.000003s : 35: predicate.replace_old_param 0.10% : 0.000001s : 17: predicate.reset_defer_inline 1.31% : 0.000020s : 123: predicate.reshape_eliminate 0.31% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.21% : 0.000003s : 17: predicate.row_tensor_eliminate 0.44% : 0.000007s : 34: predicate.same_eliminate 0.21% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.39% : 0.000006s : 35: predicate.shard_identity_eliminate 0.35% : 0.000005s : 34: predicate.special_op_eliminate 0.59% : 0.000009s : 58: predicate.specialize_transform 0.42% : 0.000006s : 34: predicate.split_environ_get_set_with_tuple_value 0.38% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.18% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.14% : 0.000033s : 209: predicate.switch_defer_inline 2.44% : 0.000038s : 243: predicate.switch_layer_defer_inline 6.94% : 0.000108s : 646: predicate.switch_simplify 1.23% : 0.000019s : 123: predicate.tile_eliminate 1.19% : 0.000019s : 123: predicate.transpose_eliminate 1.73% : 0.000027s : 157: predicate.tuple_list_convert_item_index_to_positive 1.75% : 0.000027s : 157: predicate.tuple_list_get_item_const_eliminator 1.74% : 0.000027s : 157: predicate.tuple_list_get_item_depend_reorder 2.98% : 0.000046s : 228: predicate.tuple_list_get_item_eliminator 1.65% : 0.000026s : 157: predicate.tuple_list_get_set_item_eliminator 2.35% : 0.000036s : 215: predicate.tuple_list_set_item_eliminator 1.61% : 0.000025s : 170: predicate.tuple_to_list_eliminator_ 2.65% : 0.000041s : 295: predicate.updatestate_pure_node_eliminater 3.30% : 0.000051s : 353: predicate.updatestate_useless_node_eliminater 0.17% : 0.000003s : 17: predicate.value_based_eliminate 0.34% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.34% : 0.000005s : 35: predicate.virtual_output_eliminate 0.16% : 0.000002s : 17: predicate.virtual_view_grad_eliminate 0.20% : 0.000003s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.014024 135 58.19% : 0.008160s : 55: func_graph_cloner_run.FuncGraphClonerGraph 41.81% : 0.005864s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.365821 209 0.00% : 0.000003s : 1: ForceFp32Comm 1.37% : 0.004999s : 1: add_attr 1.36% : 0.004981s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.04% : 0.000138s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.19% : 0.000708s : 1: auto_monad 0.02% : 0.000061s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.19% : 0.000679s : 1: bootstrap 0.01% : 0.000039s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000038s : 1: control_data_broadcast_order 0.01% : 0.000025s : 1: convert_after_rewriter 0.02% : 0.000067s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000027s : 1: environ_conv 0.06% : 0.000218s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000053s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000005s : 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% : 0.000008s : 1: label_micro_interleaved_index 0.13% : 0.000493s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.21% : 0.000759s : 1: mutable_eliminate 0.00% : 0.000013s : 1: offloading_packed_experts 0.01% : 0.000032s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000037s : 1: opt.transform.mutable_eliminate 3.39% : 0.012412s : 95: opt.transform.opt_a 0.03% : 0.000117s : 1: opt.transform.opt_after_cconv 0.02% : 0.000072s : 1: opt.transform.opt_after_jit_grad 0.14% : 0.000506s : 28: opt.transform.opt_b 0.06% : 0.000208s : 2: opt.transform.opt_trans_graph 0.04% : 0.000162s : 4: opt.transform.symbol_engine_opt 7.34% : 0.026836s : 1: opt_a 0.08% : 0.000284s : 1: opt_after_cconv 0.16% : 0.000574s : 1: opt_after_jit_grad 0.19% : 0.000700s : 1: opt_b 8.87% : 0.032452s : 1: optimize 0.01% : 0.000039s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000018s : 1: order_py_execute_after_rewriter 0.01% : 0.000053s : 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% : 0.000010s : 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% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.07% : 0.000258s : 1: pre_auto_parallel 0.07% : 0.000240s : 1: py_interpret_to_execute 0.01% : 0.000033s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000089s : 1: remove_dup_value 1.67% : 0.006118s : 1: renormalize.infer 1.55% : 0.005683s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.22% : 0.000804s : 1: rewriter_after_opt_a 0.21% : 0.000773s : 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 0.01% : 0.000021s : 1: swap_dp_allreduce_reducescatter 0.13% : 0.000478s : 1: symbol_engine_optimizer 0.07% : 0.000242s : 1: tuple_transform 71.91% : 0.263049s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape3-256] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape3-256],max_mem:260.0M [WARNING] ME(163223:281473160007472,MainProcess):2026-01-29-17:45:27.961.233 [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 = 0.324132, [21] [bootstrap]: 0.00066095 [type_inference]: 0.27832 [event_method]: 0.00020589 [auto_monad]: 0.00074692 [graph_reusing]: 4.776e-05 [inline]: 2.68e-06 [add_attr]: 0.00567164, [1] [add_attr_with_inline]: 0.00564823, [1] [Cycle 1]: 0.00031934, [2] [tag_attr]: 0.00021126 [meta_addattr_fg_expand]: 5.643e-05 [parallel-infer-symbol]: 3.63999e-06 [pre_auto_parallel]: 0.00025933 [insert-virtual-dataset]: 2.99999e-06 [parallel-infer-symbol-second]: 9.5999e-07 [dataset_repeat_opt]: 2.04e-06 [pipeline_split]: 1.78002e-06 [optimize]: 0.0363185, [53] [py_interpret_to_execute]: 0.00026011 [rewriter_before_opt_a]: 0.0008617 [opt_a]: 0.0299677, [2] [Cycle 1]: 0.0277647, [45] [expand_dump_flag]: 1.913e-05 [switch_simplify]: 0.00086732 [loop_unroll]: 0.00035226 [a_1]: 0.00973508 [with_stream_mark]: 6.406e-05 [recompute_prepare]: 4.367e-05 [updatestate_depend_eliminate]: 2.14e-05 [updatestate_assign_eliminate]: 1.558e-05 [updatestate_loads_eliminate]: 1.866e-05 [parameter_eliminate]: 2.86e-06 [a_2]: 0.00082356 [accelerated_algorithm]: 7.015e-05 [shard]: 2.54999e-06 [meta_shard_fg_expand]: 2.304e-05 [shard_inline]: 2.26e-05 [merge_send_recv]: 2.007e-05 [auto_parallel]: 2.17e-05 [parallel]: 3.592e-05 [flash_sp]: 1.686e-05 [merge_comm]: 1.354e-05 [allreduce_fusion]: 1.311e-05 [matmul_add_comm_reduction]: 2.19e-05 [allreduce_slice_to_reducescatter]: 1.14003e-06 [virtual_shard_identity]: 2.923e-05 [virtual_dataset]: 2.334e-05 [get_grad_eliminate_]: 2.073e-05 [virtual_output]: 2.189e-05 [merge_forward]: 1.295e-05 [cell_reuse_recompute_pass]: 2.17001e-06 [offload_activation]: 2.412e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.878e-05 [merge_recompute_call_nodes]: 1.91e-06 [before_grad]: 3.845e-05 [set_forward_comm_id_for_comm_node_pass]: 1.477e-05 [meta_fg_expand]: 1.69e-05 [flash_sp_send_recv_attached]: 5.79999e-06 [receive_attached]: 2.74999e-06 [after_resolve]: 3.152e-05 [a_after_grad]: 3.778e-05 [renormalize]: 0.0132538 [add_forward_monad_depend]: 1.949e-05 [auto_monad_grad]: 3.01001e-06 [auto_monad_eliminator]: 8.882e-05 [cse]: 0.00112201 [a_3]: 0.00024595 [Cycle 2]: 0.00218111, [45] [expand_dump_flag]: 3.73999e-06 [switch_simplify]: 2.803e-05 [loop_unroll]: 2.17e-05 [a_1]: 0.00065416 [with_stream_mark]: 3.722e-05 [recompute_prepare]: 2.416e-05 [updatestate_depend_eliminate]: 1.576e-05 [updatestate_assign_eliminate]: 1.169e-05 [updatestate_loads_eliminate]: 1.437e-05 [parameter_eliminate]: 3.01001e-06 [a_2]: 0.00031371 [accelerated_algorithm]: 3.287e-05 [shard]: 2.14e-06 [meta_shard_fg_expand]: 6.34001e-06 [shard_inline]: 2.028e-05 [merge_send_recv]: 1.926e-05 [auto_parallel]: 2.1e-05 [parallel]: 1.16e-05 [flash_sp]: 5.24998e-06 [merge_comm]: 1.337e-05 [allreduce_fusion]: 1.201e-05 [matmul_add_comm_reduction]: 2.28e-05 [allreduce_slice_to_reducescatter]: 9.39996e-07 [virtual_shard_identity]: 2.521e-05 [virtual_dataset]: 2.003e-05 [get_grad_eliminate_]: 2.129e-05 [virtual_output]: 1.95e-05 [merge_forward]: 1.305e-05 [cell_reuse_recompute_pass]: 3.44001e-06 [offload_activation]: 2.535e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.288e-05 [merge_recompute_call_nodes]: 1.49e-06 [before_grad]: 3.646e-05 [set_forward_comm_id_for_comm_node_pass]: 1.42e-05 [meta_fg_expand]: 8.90001e-06 [flash_sp_send_recv_attached]: 1.77001e-06 [receive_attached]: 2.52001e-06 [after_resolve]: 3.182e-05 [a_after_grad]: 3.307e-05 [renormalize]: 1.10012e-07 [add_forward_monad_depend]: 5.38002e-06 [auto_monad_grad]: 2.50002e-06 [auto_monad_eliminator]: 7.102e-05 [cse]: 9.352e-05 [a_3]: 0.00014144 [py_interpret_to_execute_after_opt_a]: 3.828e-05 [slice_cell_reuse_recomputed_activation]: 1.96e-06 [rewriter_after_opt_a]: 0.00088536 [convert_after_rewriter]: 2.951e-05 [order_py_execute_after_rewriter]: 1.606e-05 [mutable_eliminate]: 0.0008861 [opt_b]: 0.00078315, [1] [Cycle 1]: 0.00077182, [7] [b_1]: 0.00056093 [b_2]: 2.47e-05 [updatestate_depend_eliminate]: 1.93e-05 [updatestate_assign_eliminate]: 1.148e-05 [updatestate_loads_eliminate]: 1.397e-05 [renormalize]: 1.51998e-06 [cse]: 8.851e-05 [optimize_parallel_all_gather_comm]: 4.31e-05 [overlap_param_gather]: 2.29001e-06 [cconv]: 4.497e-05 [loop_unroll]: 0.00055855 [opt_after_cconv]: 0.00029165, [1] [Cycle 1]: 0.00028319, [7] [c_1]: 0.00012581 [parameter_eliminate]: 5.22e-06 [updatestate_depend_eliminate]: 1.79e-05 [updatestate_assign_eliminate]: 1.01e-05 [updatestate_loads_eliminate]: 1.267e-05 [cse]: 7.282e-05 [renormalize]: 8.40024e-07 [remove_dup_value]: 9.78e-05 [tuple_transform]: 0.00025024, [1] [Cycle 1]: 0.00024332, [4] [d_1]: 0.00019564 [none_parameter_eliminate]: 3.25998e-06 [renormalize]: 1.8999e-07 [switch_simplify]: 2.185e-05 [partial_unused_args_eliminate]: 1.92999e-06 [add_recomputation]: 0.00014653 [cse_after_recomputation]: 6.982e-05, [1] [Cycle 1]: 6.178e-05, [1] [cse]: 5.382e-05 [environ_conv]: 2.678e-05 [swap_dp_allreduce_reducescatter]: 1.982e-05 [bias_add_comm_swap]: 3.80998e-06 [label_micro_interleaved_index]: 5.59e-06 [label_fine_grained_interleaved_index]: 2.59999e-06 [merge_cast_opt]: 1.37e-06 [slice_recompute_activation]: 2.14999e-06 [micro_interleaved_order_control]: 2.91e-06 [assign_add_opt]: 1.26997e-06 [ForceFp32Comm]: 1.45999e-06 [remove_cast_before_assign_add]: 1.11002e-06 [full_micro_interleaved_order_control]: 2.39001e-06 [reorder_send_recv_between_fp_bp]: 2.79999e-06 [comm_op_add_attrs]: 1.10001e-06 [add_comm_op_reuse_tag]: 1.07998e-06 [interleave_split_concat_branches]: 1.30999e-06 [interleave_parallel_branches]: 1.19e-06 [overlap_opt_shard_in_pipeline]: 1.22999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.10002e-06 [control_data_broadcast_order]: 3.63e-05 [grouped_pairwise_exchange_alltoall]: 1.69998e-06 [offloading_packed_experts]: 1.1e-05 [overlap_recompute_and_grad_model_parallel]: 1.025e-05 [overlap_grad_matmul_and_grad_allreduce]: 1.15999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.19998e-06 [overlap_recompute_comm]: 2.34999e-06 [overlap_grad_ring_attention]: 9.91998e-06 [overlap_grad_flash_sp]: 5.498e-05 [begin_end_overlap_inline]: 6.19999e-07 [split_matmul_comm_elemetwise]: 2.59999e-06 [split_layernorm_comm]: 1.79e-06 [handle_group_info]: 1.24e-06 [symbol_engine_optimizer]: 0.00053032, [1] [Cycle 1]: 0.00052444, [6] [build]: 0.00029566 [elim_shapecalc]: 3.506e-05 [elim_not_effective]: 5.032e-05 [opt_reshape]: 3.174e-05 [fold_const_symbol]: 6.02e-05 [renormalize]: 6.00005e-07 [detach_backward]: 3.00998e-06 [pipeline_parallel_scheduler]: 1.92999e-06 [auto_monad_reorder]: 6.278e-05 [get_jit_bprop_graph]: 2.14999e-06 [rewriter_after_jit_bprop_graph]: 7.15e-06 [opt_after_jit_grad]: 0.00145229 [validate]: 0.00011807 Sums bootstrap : 0.000661s : 0.21% type_inference : 0.278320s : 87.75% event_method : 0.000206s : 0.06% auto_monad : 0.000747s : 0.24% graph_reusing : 0.000048s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000211s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000056s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000259s : 0.08% 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 : 0.000260s : 0.08% optimize.rewriter_before_opt_a : 0.000862s : 0.27% optimize.opt_a.expand_dump_flag : 0.000023s : 0.01% optimize.opt_a.switch_simplify : 0.000895s : 0.28% optimize.opt_a.loop_unroll : 0.000374s : 0.12% optimize.opt_a.a_1 : 0.010389s : 3.28% optimize.opt_a.with_stream_mark : 0.000101s : 0.03% optimize.opt_a.recompute_prepare : 0.000068s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000037s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000027s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000033s : 0.01% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.001137s : 0.36% optimize.opt_a.accelerated_algorithm : 0.000103s : 0.03% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000029s : 0.01% optimize.opt_a.shard_inline : 0.000043s : 0.01% optimize.opt_a.merge_send_recv : 0.000039s : 0.01% optimize.opt_a.auto_parallel : 0.000043s : 0.01% optimize.opt_a.parallel : 0.000048s : 0.01% optimize.opt_a.flash_sp : 0.000022s : 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 : 0.000045s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000054s : 0.02% optimize.opt_a.virtual_dataset : 0.000043s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000042s : 0.01% optimize.opt_a.virtual_output : 0.000041s : 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 : 0.000049s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000092s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000075s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000029s : 0.01% optimize.opt_a.meta_fg_expand : 0.000026s : 0.01% 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 : 0.000063s : 0.02% optimize.opt_a.a_after_grad : 0.000071s : 0.02% optimize.opt_a.renormalize : 0.013254s : 4.18% optimize.opt_a.add_forward_monad_depend : 0.000025s : 0.01% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000160s : 0.05% optimize.opt_a.cse : 0.001216s : 0.38% optimize.opt_a.a_3 : 0.000387s : 0.12% optimize.py_interpret_to_execute_after_opt_a : 0.000038s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000885s : 0.28% optimize.convert_after_rewriter : 0.000030s : 0.01% optimize.order_py_execute_after_rewriter : 0.000016s : 0.01% optimize.mutable_eliminate : 0.000886s : 0.28% optimize.opt_b.b_1 : 0.000561s : 0.18% optimize.opt_b.b_2 : 0.000025s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000019s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000014s : 0.00% optimize.opt_b.renormalize : 0.000002s : 0.00% optimize.opt_b.cse : 0.000089s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000043s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000045s : 0.01% optimize.loop_unroll : 0.000559s : 0.18% optimize.opt_after_cconv.c_1 : 0.000126s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000018s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000013s : 0.00% optimize.opt_after_cconv.cse : 0.000073s : 0.02% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000098s : 0.03% optimize.tuple_transform.d_1 : 0.000196s : 0.06% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000022s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000147s : 0.05% optimize.cse_after_recomputation.cse : 0.000054s : 0.02% optimize.environ_conv : 0.000027s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000020s : 0.01% 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 : 0.000001s : 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 : 0.000036s : 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 : 0.000010s : 0.00% optimize.overlap_grad_flash_sp : 0.000055s : 0.02% 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 : 0.000296s : 0.09% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000035s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000050s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000032s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000060s : 0.02% 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 : 0.000063s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.001452s : 0.46% validate : 0.000118s : 0.04% Time group info: ------[substitution.] 0.003720 569 0.46% : 0.000017s : 8: substitution.depend_value_elim 0.36% : 0.000013s : 14: substitution.elim_not_effective 0.50% : 0.000019s : 12: substitution.float_tuple_getitem_switch 0.75% : 0.000028s : 14: substitution.fold_const_symbol 0.41% : 0.000015s : 17: substitution.graph_param_transform 70.66% : 0.002628s : 71: substitution.inline 0.38% : 0.000014s : 29: substitution.j_node_and_user_rematch 0.96% : 0.000036s : 6: substitution.less_batch_normalization 1.62% : 0.000060s : 2: substitution.list_to_tuple_eliminator_ 0.17% : 0.000006s : 6: substitution.load_eliminater 0.92% : 0.000034s : 21: substitution.minmaximum_grad 0.23% : 0.000009s : 2: substitution.opt_reshape 0.54% : 0.000020s : 29: substitution.remove_not_recompute_node 0.25% : 0.000009s : 4: substitution.replace_old_param 1.68% : 0.000062s : 18: substitution.reshape_eliminate 1.44% : 0.000054s : 24: substitution.switch_simplify 3.95% : 0.000147s : 51: substitution.tuple_list_convert_item_index_to_positive 2.13% : 0.000079s : 51: substitution.tuple_list_get_item_const_eliminator 2.76% : 0.000103s : 51: substitution.tuple_list_get_item_depend_reorder 5.66% : 0.000211s : 73: substitution.tuple_list_get_item_eliminator 2.73% : 0.000102s : 51: substitution.tuple_list_get_set_item_eliminator 0.83% : 0.000031s : 6: substitution.updatestate_pure_node_eliminater 0.62% : 0.000023s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.277896 2 92.59% : 0.257300s : 1: type_inference.infer 7.41% : 0.020596s : 1: type_inference.specialize ------[replace.] 0.001232 115 4.45% : 0.000055s : 5: replace.depend_value_elim 58.38% : 0.000719s : 71: replace.inline 3.64% : 0.000045s : 2: replace.list_to_tuple_eliminator_ 24.91% : 0.000307s : 24: replace.switch_simplify 8.63% : 0.000106s : 13: replace.tuple_list_get_item_eliminator ------[match.] 0.002709 115 0.08% : 0.000002s : 5: match.depend_value_elim 95.17% : 0.002578s : 71: match.inline 2.11% : 0.000057s : 2: match.list_to_tuple_eliminator_ 1.42% : 0.000039s : 24: match.switch_simplify 1.21% : 0.000033s : 13: match.tuple_list_get_item_eliminator ------[predicate.] 0.001739 9701 1.16% : 0.000020s : 123: predicate.accumulaten_eliminater 0.48% : 0.000008s : 17: predicate.ad_related_special_op_eliminate 0.48% : 0.000008s : 58: predicate.addn_check_dump 1.15% : 0.000020s : 123: predicate.addn_zero_filter 1.03% : 0.000018s : 123: predicate.adjust_all_reduce_mul_add 2.16% : 0.000038s : 181: predicate.arithmetic_simplify 1.16% : 0.000020s : 123: predicate.cast_eliminate 0.30% : 0.000005s : 34: predicate.check_bprop_eliminate 0.44% : 0.000008s : 58: predicate.compare_switch_simplify 0.08% : 0.000001s : 17: predicate.const_output_eliminate 0.52% : 0.000009s : 61: predicate.depend_value_elim 1.21% : 0.000021s : 123: predicate.dict_get_item_const_eliminator 1.23% : 0.000021s : 123: predicate.dict_get_item_eliminator 1.08% : 0.000019s : 123: predicate.dict_set_item_eliminator 0.59% : 0.000010s : 34: predicate.dumpgradient_eliminate 0.10% : 0.000002s : 17: predicate.elim_not_effective 0.17% : 0.000003s : 17: predicate.elim_shapecalc_of_broadcastargs 1.15% : 0.000020s : 140: predicate.environ_add_const_eliminate 1.22% : 0.000021s : 140: predicate.environ_get_add_eliminate 1.23% : 0.000021s : 140: predicate.environ_get_depend_swap 1.67% : 0.000029s : 198: predicate.environ_get_eliminate 1.19% : 0.000021s : 140: predicate.environ_get_set_eliminate 1.99% : 0.000035s : 209: predicate.exchange_switch_depend_value 2.61% : 0.000045s : 209: predicate.float_depend_g_call 0.47% : 0.000008s : 58: predicate.float_environ_get_switch 0.68% : 0.000012s : 75: predicate.float_tuple_getitem_switch 0.07% : 0.000001s : 17: predicate.fold_const_symbol 0.31% : 0.000005s : 35: predicate.get_grad_eliminate 0.10% : 0.000002s : 17: predicate.graph_param_transform 0.47% : 0.000008s : 58: predicate.incorporate_call 0.43% : 0.000007s : 58: predicate.incorporate_call_switch 5.47% : 0.000095s : 459: predicate.inline 0.40% : 0.000007s : 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 1.65% : 0.000029s : 172: predicate.list_to_tuple_eliminator_ 2.49% : 0.000043s : 295: predicate.load_eliminater 0.31% : 0.000005s : 17: predicate.loop_unroll_after_grad 3.26% : 0.000057s : 314: predicate.loop_unroll_before_grad 1.41% : 0.000025s : 157: predicate.make_slice_get_slice_eliminator 0.47% : 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 1.14% : 0.000020s : 123: predicate.minmaximum_grad 0.42% : 0.000007s : 17: predicate.mutable_eliminate 0.16% : 0.000003s : 17: predicate.opt_reshape 0.17% : 0.000003s : 17: predicate.parallel_virtual_node 2.85% : 0.000050s : 209: predicate.partial_defer_inline 1.44% : 0.000025s : 155: predicate.partial_eliminate 1.13% : 0.000020s : 123: predicate.print_const_string_wrapper 0.42% : 0.000007s : 53: predicate.reduce_all_const_elim 1.57% : 0.000027s : 123: predicate.reduce_eliminate 2.56% : 0.000044s : 295: predicate.redundant_stop_gradient_eliminater 0.21% : 0.000004s : 35: predicate.remove_not_recompute_node 1.04% : 0.000018s : 172: predicate.replace_applicator 0.18% : 0.000003s : 35: predicate.replace_old_param 0.11% : 0.000002s : 17: predicate.reset_defer_inline 1.27% : 0.000022s : 123: predicate.reshape_eliminate 0.27% : 0.000005s : 34: predicate.row_tensor_add_zeros_like 0.19% : 0.000003s : 17: predicate.row_tensor_eliminate 0.52% : 0.000009s : 34: predicate.same_eliminate 0.19% : 0.000003s : 40: predicate.set_cell_output_no_recompute 0.35% : 0.000006s : 35: predicate.shard_identity_eliminate 0.34% : 0.000006s : 34: predicate.special_op_eliminate 0.54% : 0.000009s : 58: predicate.specialize_transform 0.40% : 0.000007s : 34: predicate.split_environ_get_set_with_tuple_value 0.35% : 0.000006s : 35: predicate.stack_unstack_eliminate 0.16% : 0.000003s : 17: predicate.switch_call_monad_eliminater 2.04% : 0.000035s : 209: predicate.switch_defer_inline 2.26% : 0.000039s : 243: predicate.switch_layer_defer_inline 11.22% : 0.000195s : 646: predicate.switch_simplify 1.16% : 0.000020s : 123: predicate.tile_eliminate 1.08% : 0.000019s : 123: predicate.transpose_eliminate 1.60% : 0.000028s : 157: predicate.tuple_list_convert_item_index_to_positive 1.69% : 0.000029s : 157: predicate.tuple_list_get_item_const_eliminator 1.60% : 0.000028s : 157: predicate.tuple_list_get_item_depend_reorder 2.93% : 0.000051s : 228: predicate.tuple_list_get_item_eliminator 1.52% : 0.000026s : 157: predicate.tuple_list_get_set_item_eliminator 2.11% : 0.000037s : 215: predicate.tuple_list_set_item_eliminator 1.77% : 0.000031s : 170: predicate.tuple_to_list_eliminator_ 2.41% : 0.000042s : 295: predicate.updatestate_pure_node_eliminater 3.89% : 0.000068s : 353: predicate.updatestate_useless_node_eliminater 0.16% : 0.000003s : 17: predicate.value_based_eliminate 0.31% : 0.000005s : 35: predicate.virtual_dataset_eliminate 0.33% : 0.000006s : 35: predicate.virtual_output_eliminate 0.14% : 0.000003s : 17: predicate.virtual_view_grad_eliminate 0.21% : 0.000004s : 17: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.013942 135 65.07% : 0.009072s : 55: func_graph_cloner_run.FuncGraphClonerGraph 34.93% : 0.004870s : 80: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.393973 209 0.00% : 0.000004s : 1: ForceFp32Comm 1.44% : 0.005679s : 1: add_attr 1.44% : 0.005654s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.04% : 0.000154s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.19% : 0.000766s : 1: auto_monad 0.02% : 0.000068s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.18% : 0.000694s : 1: bootstrap 0.01% : 0.000049s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000041s : 1: control_data_broadcast_order 0.01% : 0.000036s : 1: convert_after_rewriter 0.02% : 0.000073s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000031s : 1: environ_conv 0.06% : 0.000219s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.01% : 0.000054s : 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% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.14% : 0.000570s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.23% : 0.000904s : 1: mutable_eliminate 0.00% : 0.000014s : 1: offloading_packed_experts 0.01% : 0.000035s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000047s : 1: opt.transform.mutable_eliminate 3.47% : 0.013664s : 95: opt.transform.opt_a 0.03% : 0.000125s : 1: opt.transform.opt_after_cconv 0.02% : 0.000085s : 1: opt.transform.opt_after_jit_grad 0.14% : 0.000546s : 28: opt.transform.opt_b 0.05% : 0.000214s : 2: opt.transform.opt_trans_graph 0.04% : 0.000171s : 4: opt.transform.symbol_engine_opt 7.61% : 0.029973s : 1: opt_a 0.08% : 0.000296s : 1: opt_after_cconv 0.37% : 0.001474s : 1: opt_after_jit_grad 0.20% : 0.000788s : 1: opt_b 9.22% : 0.036325s : 1: optimize 0.01% : 0.000047s : 1: optimize_parallel_all_gather_comm 0.01% : 0.000020s : 1: order_py_execute_after_rewriter 0.02% : 0.000060s : 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% : 0.000007s : 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% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.07% : 0.000269s : 1: pre_auto_parallel 0.07% : 0.000272s : 1: py_interpret_to_execute 0.01% : 0.000044s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.03% : 0.000104s : 1: remove_dup_value 2.31% : 0.009086s : 1: renormalize.infer 1.05% : 0.004142s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.23% : 0.000902s : 1: rewriter_after_opt_a 0.22% : 0.000877s : 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% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000023s : 1: swap_dp_allreduce_reducescatter 0.14% : 0.000534s : 1: symbol_engine_optimizer 0.06% : 0.000254s : 1: tuple_transform 70.65% : 0.278354s : 1: type_inference . [hook] pytest_runtest_teardown:test_matmul_qkv_fp16_1[True-mstype0-k_n_shape3-1024] tests/st/infer/ops/test_internal_ops/test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape3-1024],max_mem:252.0M =============================== 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") test_matmul_split.py::test_matmul_qkv_fp16_1[True-mstype0-k_n_shape2-1024] /home/jenkins/mindspore/testcases/testcases/tests/st/infer/ops/test_internal_ops/test_matmul_split.py:84: RuntimeWarning: divide by zero encountered in divide err_cnt = np.sum(np.abs(out_flatten - expect_flatten) / -- Docs: https://docs.pytest.org/en/stable/warnings.html ================= 16 passed, 26 warnings in 532.34s (0:08:52) ==================