==================================================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/tensor, configfile: ../../../../../../sault/virtual_test/virtualenv_004/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 3 items test_copy.py [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:37:44.202.489 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:38:07.160.531 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:38:07.822.536 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. skip Bool... Tensor.copy (100, 100) Int8 ... Tensor.copy (100, 100) Int16 ... Tensor.copy (100, 100) Int32 ... [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:46.690.494 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:50.517.34 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:50.249.288 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:50.291.879 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. .[WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:53.807.584 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:57.683.975 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:40:58.369.28 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. Tensor.copy (100, 100) Int64 ... Tensor.copy (100, 100) UInt8 ... Tensor.copy (100, 100) UInt16 ... Tensor.copy (100, 100) UInt32 ... Tensor.copy (100, 100) UInt64 ... Tensor.copy (100, 100) Float16 ... Tensor.copy (100, 100) Float32 ... Tensor.copy (100, 100) Float64 ... skip BFloat16... Tensor.copy (100, 100) Complex64 ... Tensor.copy (100, 100) Complex128 ... skip Int4... skip Float8E4M3FN... skip Float8E5M2... skip HiFloat8... . [hook] pytest_runtest_teardown:test_tensor_copy_limits[pynative] tests/st/tensor/test_copy.py::test_tensor_copy_limits[pynative],max_mem:2.0M [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:41:19.708.974 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 1.27566, [21] [bootstrap]: 0.00085328 [type_inference]: 1.21965 [event_method]: 7.374e-05 [auto_monad]: 0.00032834 [graph_reusing]: 1.227e-05 [inline]: 2.94999e-06 [add_attr]: 0.0423023, [1] [add_attr_with_inline]: 0.0422577, [1] [Cycle 1]: 0.00017362, [2] [tag_attr]: 6.794e-05 [meta_addattr_fg_expand]: 1.438e-05 [parallel-infer-symbol]: 4.32e-06 [pre_auto_parallel]: 8.865e-05 [insert-virtual-dataset]: 2.64001e-06 [parallel-infer-symbol-second]: 1.37999e-06 [dataset_repeat_opt]: 2.61e-06 [pipeline_split]: 1.72999e-06 [optimize]: 0.0114007, [53] [py_interpret_to_execute]: 9.92999e-06 [rewriter_before_opt_a]: 0.00047985 [opt_a]: 0.00809194, [2] [Cycle 1]: 0.00711009, [45] [expand_dump_flag]: 8.13999e-06 [switch_simplify]: 0.00035492 [loop_unroll]: 9.02e-05 [a_1]: 0.00202577 [with_stream_mark]: 3.217e-05 [recompute_prepare]: 1.55e-05 [updatestate_depend_eliminate]: 6.59999e-06 [updatestate_assign_eliminate]: 4.25999e-06 [updatestate_loads_eliminate]: 4.07e-06 [parameter_eliminate]: 2.66e-06 [a_2]: 0.00011799 [accelerated_algorithm]: 1.174e-05 [shard]: 2.69999e-06 [meta_shard_fg_expand]: 4.79e-06 [shard_inline]: 8.77999e-06 [merge_send_recv]: 1.183e-05 [auto_parallel]: 1.173e-05 [parallel]: 0.00013434 [flash_sp]: 2.444e-05 [merge_comm]: 7.58999e-06 [allreduce_fusion]: 5.28002e-06 [matmul_add_comm_reduction]: 1.384e-05 [allreduce_slice_to_reducescatter]: 9.70002e-07 [virtual_shard_identity]: 1.806e-05 [virtual_dataset]: 9.39998e-06 [get_grad_eliminate_]: 8.13001e-06 [virtual_output]: 8.29002e-06 [merge_forward]: 5.92001e-06 [cell_reuse_recompute_pass]: 2.43998e-06 [offload_activation]: 1.227e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.252e-05 [merge_recompute_call_nodes]: 1.67001e-06 [before_grad]: 1.523e-05 [set_forward_comm_id_for_comm_node_pass]: 5.54e-06 [meta_fg_expand]: 5.82999e-06 [flash_sp_send_recv_attached]: 3.39001e-06 [receive_attached]: 2.44001e-06 [after_resolve]: 1.453e-05 [a_after_grad]: 1.239e-05 [renormalize]: 0.00343157 [add_forward_monad_depend]: 1.189e-05 [auto_monad_grad]: 3.80998e-06 [auto_monad_eliminator]: 3.15e-05 [cse]: 6.974e-05 [a_3]: 7.874e-05 [Cycle 2]: 0.00096247, [45] [expand_dump_flag]: 2.91e-06 [switch_simplify]: 1.171e-05 [loop_unroll]: 8.08999e-06 [a_1]: 0.0002007 [with_stream_mark]: 2.475e-05 [recompute_prepare]: 1.015e-05 [updatestate_depend_eliminate]: 6.31998e-06 [updatestate_assign_eliminate]: 4.85999e-06 [updatestate_loads_eliminate]: 4.28999e-06 [parameter_eliminate]: 2.62001e-06 [a_2]: 9.937e-05 [accelerated_algorithm]: 8.83001e-06 [shard]: 2.58e-06 [meta_shard_fg_expand]: 3.26001e-06 [shard_inline]: 8.25999e-06 [merge_send_recv]: 1.016e-05 [auto_parallel]: 1.1e-05 [parallel]: 9.89999e-06 [flash_sp]: 3.55e-06 [merge_comm]: 5.12999e-06 [allreduce_fusion]: 5.97001e-06 [matmul_add_comm_reduction]: 1.305e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 1.281e-05 [virtual_dataset]: 7.73999e-06 [get_grad_eliminate_]: 7.56999e-06 [virtual_output]: 7.42002e-06 [merge_forward]: 6.19001e-06 [cell_reuse_recompute_pass]: 2.97002e-06 [offload_activation]: 1.337e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.849e-05 [merge_recompute_call_nodes]: 1.40999e-06 [before_grad]: 1.321e-05 [set_forward_comm_id_for_comm_node_pass]: 6.08002e-06 [meta_fg_expand]: 4.11001e-06 [flash_sp_send_recv_attached]: 2.12001e-06 [receive_attached]: 2.99999e-06 [after_resolve]: 1.365e-05 [a_after_grad]: 1.168e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.74001e-06 [auto_monad_grad]: 1.89e-06 [auto_monad_eliminator]: 1.331e-05 [cse]: 3.292e-05 [a_3]: 4.928e-05 [py_interpret_to_execute_after_opt_a]: 1.082e-05 [slice_cell_reuse_recomputed_activation]: 2.16e-06 [rewriter_after_opt_a]: 4.647e-05 [convert_after_rewriter]: 1.25999e-06 [order_py_execute_after_rewriter]: 1.20001e-06 [mutable_eliminate]: 0.00080802 [opt_b]: 0.00027516, [1] [Cycle 1]: 0.00026443, [7] [b_1]: 0.00016179 [b_2]: 9.77001e-06 [updatestate_depend_eliminate]: 1.128e-05 [updatestate_assign_eliminate]: 4.33001e-06 [updatestate_loads_eliminate]: 3.78001e-06 [renormalize]: 6.89994e-07 [cse]: 3.803e-05 [optimize_parallel_all_gather_comm]: 2.309e-05 [overlap_param_gather]: 4.84998e-06 [cconv]: 3.952e-05 [loop_unroll]: 0.0004968 [opt_after_cconv]: 0.00013057, [1] [Cycle 1]: 0.00012346, [7] [c_1]: 3.97e-05 [parameter_eliminate]: 6.05002e-06 [updatestate_depend_eliminate]: 6.86999e-06 [updatestate_assign_eliminate]: 3.51999e-06 [updatestate_loads_eliminate]: 3.26001e-06 [cse]: 3.029e-05 [renormalize]: 5.50004e-07 [remove_dup_value]: 5.328e-05 [tuple_transform]: 9.193e-05, [1] [Cycle 1]: 8.709e-05, [4] [d_1]: 5.692e-05 [none_parameter_eliminate]: 1.98002e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 8.35999e-06 [partial_unused_args_eliminate]: 2.54999e-06 [add_recomputation]: 7.184e-05 [cse_after_recomputation]: 2.845e-05, [1] [Cycle 1]: 2.36e-05, [1] [cse]: 1.812e-05 [environ_conv]: 5.556e-05 [swap_dp_allreduce_reducescatter]: 6.64001e-06 [bias_add_comm_swap]: 3.46001e-06 [label_micro_interleaved_index]: 5.19e-06 [label_fine_grained_interleaved_index]: 2.53e-06 [merge_cast_opt]: 1.34e-06 [slice_recompute_activation]: 2.36e-06 [micro_interleaved_order_control]: 3.24001e-06 [assign_add_opt]: 1.59e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.20001e-06 [full_micro_interleaved_order_control]: 2.91e-06 [reorder_send_recv_between_fp_bp]: 2.93e-06 [comm_op_add_attrs]: 1.25001e-06 [add_comm_op_reuse_tag]: 1.08001e-06 [interleave_split_concat_branches]: 1.15999e-06 [interleave_parallel_branches]: 1.33002e-06 [overlap_opt_shard_in_pipeline]: 5.499e-05 [overlap_opt_shard_grad_in_pipeline]: 1.84998e-06 [control_data_broadcast_order]: 2.288e-05 [grouped_pairwise_exchange_alltoall]: 1.39998e-06 [offloading_packed_experts]: 5.10999e-06 [overlap_recompute_and_grad_model_parallel]: 5.79e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.18001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.27e-06 [overlap_recompute_comm]: 2.69001e-06 [overlap_grad_ring_attention]: 4.92999e-06 [overlap_grad_flash_sp]: 3.085e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.30002e-06 [split_layernorm_comm]: 1.74998e-06 [handle_group_info]: 1.17e-06 [symbol_engine_optimizer]: 0.00010146, [1] [Cycle 1]: 9.479e-05, [6] [build]: 5.15999e-06 [elim_shapecalc]: 1.544e-05 [elim_not_effective]: 1.852e-05 [opt_reshape]: 9.24e-06 [fold_const_symbol]: 1.436e-05 [renormalize]: 1.90019e-07 [detach_backward]: 2.56e-06 [pipeline_parallel_scheduler]: 1.49998e-06 [auto_monad_reorder]: 2.848e-05 [get_jit_bprop_graph]: 2.14e-06 [rewriter_after_jit_bprop_graph]: 5.76e-06 [opt_after_jit_grad]: 0.00055511 [validate]: 0.00013003 Sums bootstrap : 0.000853s : 0.07% type_inference : 1.219652s : 99.00% event_method : 0.000074s : 0.01% auto_monad : 0.000328s : 0.03% graph_reusing : 0.000012s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000068s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000014s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000089s : 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 : 0.000010s : 0.00% optimize.rewriter_before_opt_a : 0.000480s : 0.04% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000367s : 0.03% optimize.opt_a.loop_unroll : 0.000098s : 0.01% optimize.opt_a.a_1 : 0.002226s : 0.18% optimize.opt_a.with_stream_mark : 0.000057s : 0.00% optimize.opt_a.recompute_prepare : 0.000026s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000008s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000217s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000021s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.00% optimize.opt_a.shard_inline : 0.000017s : 0.00% optimize.opt_a.merge_send_recv : 0.000022s : 0.00% optimize.opt_a.auto_parallel : 0.000023s : 0.00% optimize.opt_a.parallel : 0.000144s : 0.01% optimize.opt_a.flash_sp : 0.000028s : 0.00% optimize.opt_a.merge_comm : 0.000013s : 0.00% optimize.opt_a.allreduce_fusion : 0.000011s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000027s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000031s : 0.00% optimize.opt_a.virtual_dataset : 0.000017s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.00% optimize.opt_a.virtual_output : 0.000016s : 0.00% optimize.opt_a.merge_forward : 0.000012s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000041s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000028s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.00% optimize.opt_a.meta_fg_expand : 0.000010s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000006s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000028s : 0.00% optimize.opt_a.a_after_grad : 0.000024s : 0.00% optimize.opt_a.renormalize : 0.003432s : 0.28% optimize.opt_a.add_forward_monad_depend : 0.000015s : 0.00% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000045s : 0.00% optimize.opt_a.cse : 0.000103s : 0.01% optimize.opt_a.a_3 : 0.000128s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000046s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000808s : 0.07% optimize.opt_b.b_1 : 0.000162s : 0.01% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000038s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000023s : 0.00% optimize.overlap_param_gather : 0.000005s : 0.00% optimize.cconv : 0.000040s : 0.00% optimize.loop_unroll : 0.000497s : 0.04% optimize.opt_after_cconv.c_1 : 0.000040s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000030s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000053s : 0.00% optimize.tuple_transform.d_1 : 0.000057s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000072s : 0.01% optimize.cse_after_recomputation.cse : 0.000018s : 0.00% optimize.environ_conv : 0.000056s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000007s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000055s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000023s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000031s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000015s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000019s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 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 : 0.000028s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000555s : 0.05% validate : 0.000130s : 0.01% Time group info: ------[substitution.] 0.000763 75 7.98% : 0.000061s : 11: substitution.cast_eliminate 0.35% : 0.000003s : 4: substitution.elim_not_effective 0.32% : 0.000002s : 4: substitution.fold_const_symbol 0.95% : 0.000007s : 5: substitution.graph_param_transform 82.97% : 0.000633s : 24: substitution.inline 0.82% : 0.000006s : 8: substitution.j_node_and_user_rematch 1.12% : 0.000009s : 8: substitution.remove_not_recompute_node 0.86% : 0.000007s : 2: substitution.replace_old_param 3.13% : 0.000024s : 7: substitution.switch_simplify 1.50% : 0.000011s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 1.219508 2 99.60% : 1.214651s : 1: type_inference.infer 0.40% : 0.004857s : 1: type_inference.specialize ------[replace.] 0.000319 33 60.60% : 0.000193s : 24: replace.inline 33.48% : 0.000107s : 7: replace.switch_simplify 5.92% : 0.000019s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000645 33 95.53% : 0.000617s : 24: match.inline 3.02% : 0.000020s : 7: match.switch_simplify 1.45% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000442 2544 1.17% : 0.000005s : 32: predicate.accumulaten_eliminater 0.42% : 0.000002s : 5: predicate.ad_related_special_op_eliminate 0.33% : 0.000001s : 10: predicate.addn_check_dump 1.06% : 0.000005s : 32: predicate.addn_zero_filter 0.99% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.74% : 0.000012s : 42: predicate.arithmetic_simplify 1.72% : 0.000008s : 32: predicate.cast_eliminate 0.40% : 0.000002s : 10: predicate.check_bprop_eliminate 0.39% : 0.000002s : 10: predicate.compare_switch_simplify 0.10% : 0.000000s : 5: predicate.const_output_eliminate 0.52% : 0.000002s : 10: predicate.depend_value_elim 1.30% : 0.000006s : 32: predicate.dict_get_item_const_eliminator 1.23% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.04% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.57% : 0.000003s : 10: predicate.dumpgradient_eliminate 0.18% : 0.000001s : 5: predicate.elim_not_effective 0.19% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.39% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.16% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.16% : 0.000005s : 37: predicate.environ_get_depend_swap 1.55% : 0.000007s : 47: predicate.environ_get_eliminate 1.39% : 0.000006s : 37: predicate.environ_get_set_eliminate 1.92% : 0.000008s : 58: predicate.exchange_switch_depend_value 2.95% : 0.000013s : 58: predicate.float_depend_g_call 0.34% : 0.000002s : 10: predicate.float_environ_get_switch 0.48% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.10% : 0.000000s : 5: predicate.fold_const_symbol 0.52% : 0.000002s : 10: predicate.get_grad_eliminate 0.13% : 0.000001s : 5: predicate.graph_param_transform 0.37% : 0.000002s : 10: predicate.incorporate_call 0.28% : 0.000001s : 10: predicate.incorporate_call_switch 6.03% : 0.000027s : 120: predicate.inline 0.50% : 0.000002s : 10: predicate.inline_without_move 0.17% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.80% : 0.000004s : 10: predicate.less_batch_normalization 1.53% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.54% : 0.000011s : 76: predicate.load_eliminater 0.49% : 0.000002s : 5: predicate.loop_unroll_after_grad 3.66% : 0.000016s : 104: predicate.loop_unroll_before_grad 1.39% : 0.000006s : 42: predicate.make_slice_get_slice_eliminator 0.36% : 0.000002s : 10: predicate.merge_addn 0.54% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.34% : 0.000002s : 10: predicate.mini_step_allgather_replace 1.03% : 0.000005s : 32: predicate.minmaximum_grad 0.75% : 0.000003s : 5: predicate.mutable_eliminate 0.23% : 0.000001s : 5: predicate.opt_reshape 0.28% : 0.000001s : 5: predicate.parallel_virtual_node 2.45% : 0.000011s : 58: predicate.partial_defer_inline 1.35% : 0.000006s : 39: predicate.partial_eliminate 1.34% : 0.000006s : 32: predicate.print_const_string_wrapper 0.35% : 0.000002s : 10: predicate.reduce_all_const_elim 1.46% : 0.000006s : 32: predicate.reduce_eliminate 2.43% : 0.000011s : 76: predicate.redundant_stop_gradient_eliminater 0.25% : 0.000001s : 10: predicate.remove_not_recompute_node 1.23% : 0.000005s : 44: predicate.replace_applicator 0.33% : 0.000001s : 10: predicate.replace_old_param 0.18% : 0.000001s : 5: predicate.reset_defer_inline 1.07% : 0.000005s : 32: predicate.reshape_eliminate 0.46% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.28% : 0.000001s : 5: predicate.row_tensor_eliminate 0.48% : 0.000002s : 10: predicate.same_eliminate 0.44% : 0.000002s : 10: predicate.set_cell_output_no_recompute 0.55% : 0.000002s : 10: predicate.shard_identity_eliminate 0.43% : 0.000002s : 10: predicate.special_op_eliminate 0.38% : 0.000002s : 10: predicate.specialize_transform 0.71% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.54% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.20% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.28% : 0.000010s : 58: predicate.switch_defer_inline 2.57% : 0.000011s : 68: predicate.switch_layer_defer_inline 7.82% : 0.000035s : 191: predicate.switch_simplify 1.17% : 0.000005s : 32: predicate.tile_eliminate 1.18% : 0.000005s : 32: predicate.transpose_eliminate 1.42% : 0.000006s : 42: predicate.tuple_list_convert_item_index_to_positive 1.37% : 0.000006s : 42: predicate.tuple_list_get_item_const_eliminator 1.46% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.52% : 0.000011s : 54: predicate.tuple_list_get_item_eliminator 1.47% : 0.000006s : 42: predicate.tuple_list_get_set_item_eliminator 2.00% : 0.000009s : 52: predicate.tuple_list_set_item_eliminator 2.07% : 0.000009s : 44: predicate.tuple_to_list_eliminator_ 2.61% : 0.000012s : 76: predicate.updatestate_pure_node_eliminater 2.82% : 0.000012s : 86: predicate.updatestate_useless_node_eliminater 0.21% : 0.000001s : 5: predicate.value_based_eliminate 0.54% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.46% : 0.000002s : 10: predicate.virtual_output_eliminate 0.14% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.25% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003388 42 57.02% : 0.001932s : 16: func_graph_cloner_run.FuncGraphClonerGraph 42.98% : 0.001456s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.336040 192 0.00% : 0.000004s : 1: ForceFp32Comm 3.17% : 0.042309s : 1: add_attr 3.16% : 0.042263s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000076s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000339s : 1: auto_monad 0.00% : 0.000033s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.07% : 0.000895s : 1: bootstrap 0.00% : 0.000043s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000026s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000032s : 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 0.01% : 0.000083s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000016s : 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% : 0.000006s : 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 0.04% : 0.000507s : 1: loop_unroll 0.01% : 0.000114s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.06% : 0.000826s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000026s : 1: opt.transform.mutable_eliminate 0.24% : 0.003179s : 78: opt.transform.opt_a 0.00% : 0.000039s : 1: opt.transform.opt_after_cconv 0.00% : 0.000030s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000144s : 28: opt.transform.opt_b 0.00% : 0.000063s : 2: opt.transform.opt_trans_graph 0.00% : 0.000053s : 4: opt.transform.symbol_engine_opt 0.61% : 0.008096s : 1: opt_a 0.01% : 0.000134s : 1: opt_after_cconv 0.04% : 0.000567s : 1: opt_after_jit_grad 0.02% : 0.000279s : 1: opt_b 0.85% : 0.011407s : 1: optimize 0.00% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000034s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000059s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.01% : 0.000094s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000014s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000058s : 1: remove_dup_value 0.14% : 0.001863s : 1: renormalize.infer 0.12% : 0.001549s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000051s : 1: rewriter_after_opt_a 0.04% : 0.000496s : 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% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000104s : 1: symbol_engine_optimizer 0.01% : 0.000095s : 1: tuple_transform 91.29% : 1.219689s : 1: type_inference TotalTime = 1.58238, [21] [bootstrap]: 0.197413 [type_inference]: 0.999015 [event_method]: 9.121e-05 [auto_monad]: 0.00032588 [graph_reusing]: 2.083e-05 [inline]: 3.46999e-06 [add_attr]: 0.16966, [1] [add_attr_with_inline]: 0.169644, [1] [Cycle 1]: 0.00015883, [2] [tag_attr]: 9.126e-05 [meta_addattr_fg_expand]: 1.969e-05 [parallel-infer-symbol]: 4.07e-06 [pre_auto_parallel]: 0.00011268 [insert-virtual-dataset]: 3.04001e-06 [parallel-infer-symbol-second]: 9.5999e-07 [dataset_repeat_opt]: 2.10002e-06 [pipeline_split]: 1.76003e-06 [optimize]: 0.214928, [53] [py_interpret_to_execute]: 1.01e-05 [rewriter_before_opt_a]: 0.00059407 [opt_a]: 0.212025, [2] [Cycle 1]: 0.210819, [45] [expand_dump_flag]: 8.45999e-06 [switch_simplify]: 0.00036473 [loop_unroll]: 0.00011687 [a_1]: 0.00279786 [with_stream_mark]: 4.236e-05 [recompute_prepare]: 3.088e-05 [updatestate_depend_eliminate]: 1.179e-05 [updatestate_assign_eliminate]: 8.23001e-06 [updatestate_loads_eliminate]: 7.24001e-06 [parameter_eliminate]: 3.66999e-06 [a_2]: 0.00023822 [accelerated_algorithm]: 1.81e-05 [shard]: 2.31e-06 [meta_shard_fg_expand]: 2.622e-05 [shard_inline]: 1.922e-05 [merge_send_recv]: 2.32e-05 [auto_parallel]: 1.672e-05 [parallel]: 4.455e-05 [flash_sp]: 1.497e-05 [merge_comm]: 1.117e-05 [allreduce_fusion]: 9.03002e-06 [matmul_add_comm_reduction]: 3.517e-05 [allreduce_slice_to_reducescatter]: 1.10999e-06 [virtual_shard_identity]: 2.058e-05 [virtual_dataset]: 1.459e-05 [get_grad_eliminate_]: 1.499e-05 [virtual_output]: 1.54e-05 [merge_forward]: 1.023e-05 [cell_reuse_recompute_pass]: 1.76e-06 [offload_activation]: 1.925e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.421e-05 [merge_recompute_call_nodes]: 1.92001e-06 [before_grad]: 2.863e-05 [set_forward_comm_id_for_comm_node_pass]: 1.075e-05 [meta_fg_expand]: 0.00367752 [flash_sp_send_recv_attached]: 5.39e-06 [receive_attached]: 2.74001e-06 [after_resolve]: 0.00014453 [a_after_grad]: 0.00023013 [renormalize]: 0.202233 [add_forward_monad_depend]: 7.23e-06 [auto_monad_grad]: 2.76999e-06 [auto_monad_eliminator]: 1.558e-05 [cse]: 2.944e-05 [a_3]: 4.932e-05 [Cycle 2]: 0.00118674, [45] [expand_dump_flag]: 2.65997e-06 [switch_simplify]: 7.06001e-06 [loop_unroll]: 4.77998e-06 [a_1]: 5.885e-05 [with_stream_mark]: 1.58e-05 [recompute_prepare]: 4.12e-06 [updatestate_depend_eliminate]: 3.29001e-06 [updatestate_assign_eliminate]: 2.99999e-06 [updatestate_loads_eliminate]: 2.37001e-06 [parameter_eliminate]: 1.81e-06 [a_2]: 3.915e-05 [accelerated_algorithm]: 4.71997e-06 [shard]: 1.88002e-06 [meta_shard_fg_expand]: 2.43e-06 [shard_inline]: 3.65e-06 [merge_send_recv]: 7.26001e-06 [auto_parallel]: 8.57998e-06 [parallel]: 8.37998e-06 [flash_sp]: 3.9e-06 [merge_comm]: 2.19999e-06 [allreduce_fusion]: 2.46e-06 [matmul_add_comm_reduction]: 8.2e-06 [allreduce_slice_to_reducescatter]: 6.50005e-07 [virtual_shard_identity]: 4.82998e-06 [virtual_dataset]: 3.87002e-06 [get_grad_eliminate_]: 4.12998e-06 [virtual_output]: 4.58001e-06 [merge_forward]: 3.33998e-06 [cell_reuse_recompute_pass]: 2.99999e-06 [offload_activation]: 7.82e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.275e-05 [merge_recompute_call_nodes]: 1.72999e-06 [before_grad]: 7.27997e-06 [set_forward_comm_id_for_comm_node_pass]: 2.73e-06 [meta_fg_expand]: 0.00064114 [flash_sp_send_recv_attached]: 2.83998e-06 [receive_attached]: 2.68e-06 [after_resolve]: 1.36e-05 [a_after_grad]: 7.3e-06 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 3.7e-06 [auto_monad_grad]: 1.81e-06 [auto_monad_eliminator]: 8.44998e-06 [cse]: 1.638e-05 [a_3]: 2.424e-05 [py_interpret_to_execute_after_opt_a]: 7.49002e-06 [slice_cell_reuse_recomputed_activation]: 2.21e-06 [rewriter_after_opt_a]: 2.033e-05 [convert_after_rewriter]: 1.77001e-06 [order_py_execute_after_rewriter]: 1.24e-06 [mutable_eliminate]: 0.00081154 [opt_b]: 0.00016303, [1] [Cycle 1]: 0.00015271, [7] [b_1]: 8.053e-05 [b_2]: 5.77001e-06 [updatestate_depend_eliminate]: 5.96003e-06 [updatestate_assign_eliminate]: 2.86999e-06 [updatestate_loads_eliminate]: 2.21e-06 [renormalize]: 6.99976e-07 [cse]: 2.12e-05 [optimize_parallel_all_gather_comm]: 1.841e-05 [overlap_param_gather]: 2.16e-06 [cconv]: 3.267e-05 [loop_unroll]: 0.00047577 [opt_after_cconv]: 8.4e-05, [1] [Cycle 1]: 7.729e-05, [7] [c_1]: 1.667e-05 [parameter_eliminate]: 3.09999e-06 [updatestate_depend_eliminate]: 4.27e-06 [updatestate_assign_eliminate]: 1.74e-06 [updatestate_loads_eliminate]: 1.49e-06 [cse]: 1.551e-05 [renormalize]: 6.59988e-07 [remove_dup_value]: 1.547e-05 [tuple_transform]: 5.279e-05, [1] [Cycle 1]: 4.783e-05, [4] [d_1]: 2.397e-05 [none_parameter_eliminate]: 1.68002e-06 [renormalize]: 2.00002e-07 [switch_simplify]: 4.28999e-06 [partial_unused_args_eliminate]: 1.71e-06 [add_recomputation]: 3.654e-05 [cse_after_recomputation]: 1.908e-05, [1] [Cycle 1]: 1.38e-05, [1] [cse]: 8.36002e-06 [environ_conv]: 5.59e-06 [swap_dp_allreduce_reducescatter]: 4.60001e-06 [bias_add_comm_swap]: 3.26001e-06 [label_micro_interleaved_index]: 5.07999e-06 [label_fine_grained_interleaved_index]: 3.11001e-06 [merge_cast_opt]: 1.32999e-06 [slice_recompute_activation]: 2.12999e-06 [micro_interleaved_order_control]: 2.79001e-06 [assign_add_opt]: 1.52999e-06 [ForceFp32Comm]: 1.29e-06 [remove_cast_before_assign_add]: 1.12999e-06 [full_micro_interleaved_order_control]: 2.32999e-06 [reorder_send_recv_between_fp_bp]: 3.76001e-06 [comm_op_add_attrs]: 1.14e-06 [add_comm_op_reuse_tag]: 1.01002e-06 [interleave_split_concat_branches]: 1.20001e-06 [interleave_parallel_branches]: 1.04998e-06 [overlap_opt_shard_in_pipeline]: 7.03e-06 [overlap_opt_shard_grad_in_pipeline]: 1.77999e-06 [control_data_broadcast_order]: 1.216e-05 [grouped_pairwise_exchange_alltoall]: 1.54e-06 [offloading_packed_experts]: 3.80998e-06 [overlap_recompute_and_grad_model_parallel]: 4.18999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.40001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.25001e-06 [overlap_recompute_comm]: 2.64001e-06 [overlap_grad_ring_attention]: 3.2e-06 [overlap_grad_flash_sp]: 1.853e-05 [begin_end_overlap_inline]: 5.69999e-07 [split_matmul_comm_elemetwise]: 2.19999e-06 [split_layernorm_comm]: 1.66e-06 [handle_group_info]: 1.07998e-06 [symbol_engine_optimizer]: 6.805e-05, [1] [Cycle 1]: 6.276e-05, [6] [build]: 3.58999e-06 [elim_shapecalc]: 7.58999e-06 [elim_not_effective]: 9.96998e-06 [opt_reshape]: 6.09001e-06 [fold_const_symbol]: 7.03e-06 [renormalize]: 2.50002e-07 [detach_backward]: 2.72001e-06 [pipeline_parallel_scheduler]: 1.50999e-06 [auto_monad_reorder]: 1.53e-05 [get_jit_bprop_graph]: 2.13002e-06 [rewriter_after_jit_bprop_graph]: 5.08002e-06 [opt_after_jit_grad]: 0.00052047 [validate]: 3.795e-05 Sums bootstrap : 0.197413s : 13.99% type_inference : 0.999015s : 70.78% event_method : 0.000091s : 0.01% auto_monad : 0.000326s : 0.02% graph_reusing : 0.000021s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000091s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000020s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000113s : 0.01% 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 : 0.000010s : 0.00% optimize.rewriter_before_opt_a : 0.000594s : 0.04% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000372s : 0.03% optimize.opt_a.loop_unroll : 0.000122s : 0.01% optimize.opt_a.a_1 : 0.002857s : 0.20% optimize.opt_a.with_stream_mark : 0.000058s : 0.00% optimize.opt_a.recompute_prepare : 0.000035s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000277s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000023s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000029s : 0.00% optimize.opt_a.shard_inline : 0.000023s : 0.00% optimize.opt_a.merge_send_recv : 0.000030s : 0.00% optimize.opt_a.auto_parallel : 0.000025s : 0.00% optimize.opt_a.parallel : 0.000053s : 0.00% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000013s : 0.00% optimize.opt_a.allreduce_fusion : 0.000011s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000043s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000025s : 0.00% optimize.opt_a.virtual_dataset : 0.000018s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000019s : 0.00% optimize.opt_a.virtual_output : 0.000020s : 0.00% optimize.opt_a.merge_forward : 0.000014s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000027s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000047s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000036s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000013s : 0.00% optimize.opt_a.meta_fg_expand : 0.004319s : 0.31% 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 : 0.000158s : 0.01% optimize.opt_a.a_after_grad : 0.000237s : 0.02% optimize.opt_a.renormalize : 0.202233s : 14.33% optimize.opt_a.add_forward_monad_depend : 0.000011s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000024s : 0.00% optimize.opt_a.cse : 0.000046s : 0.00% optimize.opt_a.a_3 : 0.000074s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000007s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000020s : 0.00% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000812s : 0.06% optimize.opt_b.b_1 : 0.000081s : 0.01% optimize.opt_b.b_2 : 0.000006s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000021s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000018s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000033s : 0.00% optimize.loop_unroll : 0.000476s : 0.03% optimize.opt_after_cconv.c_1 : 0.000017s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000024s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000037s : 0.00% optimize.cse_after_recomputation.cse : 0.000008s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000004s : 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 : 0.000007s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000012s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000019s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000008s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000010s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000006s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 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 : 0.000015s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000520s : 0.04% validate : 0.000038s : 0.00% Time group info: ------[substitution.] 0.000969 134 3.33% : 0.000032s : 8: substitution.cast_eliminate 0.18% : 0.000002s : 1: substitution.elim_not_effective 0.68% : 0.000007s : 9: substitution.float_depend_g_call 0.48% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.15% : 0.000001s : 1: substitution.fold_const_symbol 0.52% : 0.000005s : 1: substitution.graph_param_transform 0.36% : 0.000003s : 2: substitution.incorporate_call 0.24% : 0.000002s : 2: substitution.incorporate_call_switch 76.42% : 0.000741s : 33: substitution.inline 3.84% : 0.000037s : 4: substitution.inline_without_move 0.93% : 0.000009s : 11: substitution.j_node_and_user_rematch 0.42% : 0.000004s : 2: substitution.minmaximum_grad 2.21% : 0.000021s : 9: substitution.partial_eliminate 1.03% : 0.000010s : 11: substitution.remove_not_recompute_node 0.29% : 0.000003s : 1: substitution.replace_applicator 1.06% : 0.000010s : 12: substitution.replace_old_param 0.48% : 0.000005s : 1: substitution.set_cell_output_no_recompute 2.55% : 0.000025s : 10: substitution.switch_simplify 0.87% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.33% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.58% : 0.000006s : 2: substitution.tuple_list_get_item_depend_reorder 2.49% : 0.000024s : 6: substitution.tuple_list_get_item_eliminator 0.57% : 0.000006s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.998759 2 99.35% : 0.992235s : 1: type_inference.infer 0.65% : 0.006524s : 1: type_inference.specialize ------[replace.] 0.000404 45 57.28% : 0.000232s : 33: replace.inline 38.99% : 0.000158s : 10: replace.switch_simplify 3.73% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000754 45 96.02% : 0.000724s : 33: match.inline 2.47% : 0.000019s : 10: match.switch_simplify 1.51% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000547 3189 1.17% : 0.000006s : 43: predicate.accumulaten_eliminater 0.16% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.40% : 0.000002s : 14: predicate.addn_check_dump 1.20% : 0.000007s : 43: predicate.addn_zero_filter 1.05% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 1.96% : 0.000011s : 57: predicate.arithmetic_simplify 1.53% : 0.000008s : 43: predicate.cast_eliminate 0.12% : 0.000001s : 2: predicate.check_bprop_eliminate 0.41% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.45% : 0.000002s : 14: predicate.depend_value_elim 1.20% : 0.000007s : 43: predicate.dict_get_item_const_eliminator 1.19% : 0.000007s : 43: predicate.dict_get_item_eliminator 1.12% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.16% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.04% : 0.000000s : 1: predicate.elim_not_effective 0.09% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.40% : 0.000008s : 44: predicate.environ_add_const_eliminate 1.13% : 0.000006s : 44: predicate.environ_get_add_eliminate 1.30% : 0.000007s : 44: predicate.environ_get_depend_swap 1.57% : 0.000009s : 58: predicate.environ_get_eliminate 1.05% : 0.000006s : 44: predicate.environ_get_set_eliminate 6.24% : 0.000034s : 78: predicate.exchange_switch_depend_value 2.87% : 0.000016s : 78: predicate.float_depend_g_call 0.38% : 0.000002s : 14: predicate.float_environ_get_switch 0.48% : 0.000003s : 15: predicate.float_tuple_getitem_switch 0.06% : 0.000000s : 1: predicate.fold_const_symbol 0.50% : 0.000003s : 14: predicate.get_grad_eliminate 0.03% : 0.000000s : 1: predicate.graph_param_transform 0.40% : 0.000002s : 14: predicate.incorporate_call 0.32% : 0.000002s : 14: predicate.incorporate_call_switch 5.76% : 0.000031s : 151: predicate.inline 2.59% : 0.000014s : 70: predicate.inline_without_move 0.19% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.60% : 0.000003s : 14: predicate.less_batch_normalization 1.29% : 0.000007s : 47: predicate.list_to_tuple_eliminator_ 2.40% : 0.000013s : 90: predicate.load_eliminater 0.25% : 0.000001s : 1: predicate.loop_unroll_after_grad 3.71% : 0.000020s : 141: predicate.loop_unroll_before_grad 1.50% : 0.000008s : 45: predicate.make_slice_get_slice_eliminator 0.40% : 0.000002s : 14: predicate.merge_addn 0.11% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.12% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.05% : 0.000006s : 43: predicate.minmaximum_grad 0.39% : 0.000002s : 1: predicate.mutable_eliminate 0.08% : 0.000000s : 1: predicate.opt_reshape 0.06% : 0.000000s : 1: predicate.parallel_virtual_node 2.76% : 0.000015s : 78: predicate.partial_defer_inline 1.32% : 0.000007s : 46: predicate.partial_eliminate 1.15% : 0.000006s : 43: predicate.print_const_string_wrapper 0.39% : 0.000002s : 14: predicate.reduce_all_const_elim 1.41% : 0.000008s : 43: predicate.reduce_eliminate 2.24% : 0.000012s : 90: predicate.redundant_stop_gradient_eliminater 0.35% : 0.000002s : 14: predicate.remove_not_recompute_node 1.02% : 0.000006s : 47: predicate.replace_applicator 1.23% : 0.000007s : 70: predicate.replace_old_param 0.06% : 0.000000s : 1: predicate.reset_defer_inline 1.12% : 0.000006s : 43: predicate.reshape_eliminate 0.19% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.11% : 0.000001s : 1: predicate.row_tensor_eliminate 0.23% : 0.000001s : 2: predicate.same_eliminate 0.28% : 0.000002s : 14: predicate.set_cell_output_no_recompute 0.49% : 0.000003s : 14: predicate.shard_identity_eliminate 0.14% : 0.000001s : 2: predicate.special_op_eliminate 0.40% : 0.000002s : 14: predicate.specialize_transform 0.35% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 2.53% : 0.000014s : 70: predicate.stack_unstack_eliminate 0.07% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.31% : 0.000013s : 78: predicate.switch_defer_inline 2.40% : 0.000013s : 80: predicate.switch_layer_defer_inline 7.38% : 0.000040s : 254: predicate.switch_simplify 1.15% : 0.000006s : 43: predicate.tile_eliminate 1.14% : 0.000006s : 43: predicate.transpose_eliminate 1.35% : 0.000007s : 45: predicate.tuple_list_convert_item_index_to_positive 1.46% : 0.000008s : 45: predicate.tuple_list_get_item_const_eliminator 1.24% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.45% : 0.000013s : 61: predicate.tuple_list_get_item_eliminator 1.40% : 0.000008s : 45: predicate.tuple_list_get_set_item_eliminator 1.95% : 0.000011s : 59: predicate.tuple_list_set_item_eliminator 1.28% : 0.000007s : 47: predicate.tuple_to_list_eliminator_ 2.25% : 0.000012s : 90: predicate.updatestate_pure_node_eliminater 2.73% : 0.000015s : 104: predicate.updatestate_useless_node_eliminater 0.08% : 0.000000s : 1: predicate.value_based_eliminate 0.44% : 0.000002s : 14: predicate.virtual_dataset_eliminate 0.47% : 0.000003s : 14: predicate.virtual_output_eliminate 0.05% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.09% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.005008 69 73.71% : 0.003691s : 33: func_graph_cloner_run.FuncGraphClonerGraph 26.29% : 0.001317s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.173294 192 0.00% : 0.000004s : 1: ForceFp32Comm 7.81% : 0.169667s : 1: add_attr 7.81% : 0.169649s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000040s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.02% : 0.000336s : 1: auto_monad 0.00% : 0.000019s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 9.09% : 0.197460s : 1: bootstrap 0.00% : 0.000036s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000015s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.00% : 0.000103s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000025s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.02% : 0.000485s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.04% : 0.000824s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000011s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000016s : 1: opt.transform.mutable_eliminate 0.19% : 0.004226s : 78: opt.transform.opt_a 0.00% : 0.000015s : 1: opt.transform.opt_after_cconv 0.00% : 0.000015s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000051s : 28: opt.transform.opt_b 0.00% : 0.000026s : 2: opt.transform.opt_trans_graph 0.00% : 0.000026s : 4: opt.transform.symbol_engine_opt 9.76% : 0.212028s : 1: opt_a 0.00% : 0.000087s : 1: opt_after_cconv 0.02% : 0.000532s : 1: opt_after_jit_grad 0.01% : 0.000167s : 1: opt_b 9.89% : 0.214934s : 1: optimize 0.00% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000022s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000010s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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 0.01% : 0.000119s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000011s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000020s : 1: remove_dup_value 9.27% : 0.201379s : 1: renormalize.infer 0.04% : 0.000840s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000024s : 1: rewriter_after_opt_a 0.03% : 0.000606s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000071s : 1: symbol_engine_optimizer 0.00% : 0.000056s : 1: tuple_transform 45.97% : 0.999052s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:41:38.858.339 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.929618, [21] [bootstrap]: 0.00049372 [type_inference]: 0.865786 [event_method]: 7.57e-05 [auto_monad]: 0.00018183 [graph_reusing]: 1.142e-05 [inline]: 3.09001e-06 [add_attr]: 0.00388001, [1] [add_attr_with_inline]: 0.00386724, [1] [Cycle 1]: 0.00011775, [2] [tag_attr]: 6.311e-05 [meta_addattr_fg_expand]: 1.346e-05 [parallel-infer-symbol]: 3.97e-06 [pre_auto_parallel]: 8.415e-05 [insert-virtual-dataset]: 2.49999e-06 [parallel-infer-symbol-second]: 8.39995e-07 [dataset_repeat_opt]: 2.26e-06 [pipeline_split]: 1.78002e-06 [optimize]: 0.0582138, [53] [py_interpret_to_execute]: 6.61e-06 [rewriter_before_opt_a]: 0.00042386 [opt_a]: 0.0552389, [2] [Cycle 1]: 0.0544068, [45] [expand_dump_flag]: 6.63998e-06 [switch_simplify]: 0.00023983 [loop_unroll]: 8.813e-05 [a_1]: 0.00170989 [with_stream_mark]: 2.485e-05 [recompute_prepare]: 1.261e-05 [updatestate_depend_eliminate]: 5.45001e-06 [updatestate_assign_eliminate]: 4.4e-06 [updatestate_loads_eliminate]: 2.926e-05 [parameter_eliminate]: 8.14997e-06 [a_2]: 0.00015466 [accelerated_algorithm]: 1.139e-05 [shard]: 3.95e-06 [meta_shard_fg_expand]: 1.071e-05 [shard_inline]: 8.70001e-06 [merge_send_recv]: 1.368e-05 [auto_parallel]: 1.644e-05 [parallel]: 2.357e-05 [flash_sp]: 1.533e-05 [merge_comm]: 5.08002e-06 [allreduce_fusion]: 4.68001e-06 [matmul_add_comm_reduction]: 1.469e-05 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 9.92999e-06 [virtual_dataset]: 8.99998e-06 [get_grad_eliminate_]: 8.87e-06 [virtual_output]: 8.85001e-06 [merge_forward]: 6.98e-06 [cell_reuse_recompute_pass]: 3.85e-06 [offload_activation]: 1.262e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.762e-05 [merge_recompute_call_nodes]: 1.48002e-06 [before_grad]: 1.56e-05 [set_forward_comm_id_for_comm_node_pass]: 5.10999e-06 [meta_fg_expand]: 6.61e-06 [flash_sp_send_recv_attached]: 2.79999e-06 [receive_attached]: 1.96e-06 [after_resolve]: 1.285e-05 [a_after_grad]: 1.224e-05 [renormalize]: 0.0029736 [add_forward_monad_depend]: 7.587e-05 [auto_monad_grad]: 2.81e-06 [auto_monad_eliminator]: 2.507e-05 [cse]: 4.961e-05 [a_3]: 7.294e-05 [Cycle 2]: 0.00081661, [45] [expand_dump_flag]: 2.17999e-06 [switch_simplify]: 1.027e-05 [loop_unroll]: 7.90998e-06 [a_1]: 0.0001898 [with_stream_mark]: 2.069e-05 [recompute_prepare]: 9.31e-06 [updatestate_depend_eliminate]: 5.05001e-06 [updatestate_assign_eliminate]: 4.1e-06 [updatestate_loads_eliminate]: 4.32003e-06 [parameter_eliminate]: 2.19999e-06 [a_2]: 9.374e-05 [accelerated_algorithm]: 7.52998e-06 [shard]: 2.46e-06 [meta_shard_fg_expand]: 2.79001e-06 [shard_inline]: 7.8e-06 [merge_send_recv]: 9.90002e-06 [auto_parallel]: 1.035e-05 [parallel]: 8.62998e-06 [flash_sp]: 4.53001e-06 [merge_comm]: 5.20999e-06 [allreduce_fusion]: 4.72e-06 [matmul_add_comm_reduction]: 1.129e-05 [allreduce_slice_to_reducescatter]: 6.80011e-07 [virtual_shard_identity]: 9.54e-06 [virtual_dataset]: 7.51001e-06 [get_grad_eliminate_]: 7.01999e-06 [virtual_output]: 6.86001e-06 [merge_forward]: 4.70001e-06 [cell_reuse_recompute_pass]: 2.59999e-06 [offload_activation]: 1.218e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.69e-05 [merge_recompute_call_nodes]: 1.16002e-06 [before_grad]: 1.289e-05 [set_forward_comm_id_for_comm_node_pass]: 5.31002e-06 [meta_fg_expand]: 3.12002e-06 [flash_sp_send_recv_attached]: 1.61998e-06 [receive_attached]: 2.32001e-06 [after_resolve]: 1.155e-05 [a_after_grad]: 1.127e-05 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 1.17e-06 [auto_monad_grad]: 1.54998e-06 [auto_monad_eliminator]: 9.49e-06 [cse]: 2.326e-05 [a_3]: 4.483e-05 [py_interpret_to_execute_after_opt_a]: 8.74e-06 [slice_cell_reuse_recomputed_activation]: 1.90001e-06 [rewriter_after_opt_a]: 2.737e-05 [convert_after_rewriter]: 1.14998e-06 [order_py_execute_after_rewriter]: 1.44998e-06 [mutable_eliminate]: 0.00076404 [opt_b]: 0.00027291, [1] [Cycle 1]: 0.00026366, [7] [b_1]: 0.00016925 [b_2]: 9.22001e-06 [updatestate_depend_eliminate]: 9.45001e-06 [updatestate_assign_eliminate]: 3.4e-06 [updatestate_loads_eliminate]: 3.43e-06 [renormalize]: 6.60017e-07 [cse]: 3.278e-05 [optimize_parallel_all_gather_comm]: 2.25e-05 [overlap_param_gather]: 2.19999e-06 [cconv]: 3.39e-05 [loop_unroll]: 0.00049367 [opt_after_cconv]: 0.00012278, [1] [Cycle 1]: 0.00011599, [7] [c_1]: 3.639e-05 [parameter_eliminate]: 4.62998e-06 [updatestate_depend_eliminate]: 7.40998e-06 [updatestate_assign_eliminate]: 3.35e-06 [updatestate_loads_eliminate]: 3.22002e-06 [cse]: 2.844e-05 [renormalize]: 3.59985e-07 [remove_dup_value]: 5.142e-05 [tuple_transform]: 8.832e-05, [1] [Cycle 1]: 8.268e-05, [4] [d_1]: 5.444e-05 [none_parameter_eliminate]: 1.82001e-06 [renormalize]: 1.80007e-07 [switch_simplify]: 8.25e-06 [partial_unused_args_eliminate]: 1.87001e-06 [add_recomputation]: 6.8e-05 [cse_after_recomputation]: 2.863e-05, [1] [Cycle 1]: 2.329e-05, [1] [cse]: 1.754e-05 [environ_conv]: 7.05e-06 [swap_dp_allreduce_reducescatter]: 6.49999e-06 [bias_add_comm_swap]: 2.85002e-06 [label_micro_interleaved_index]: 5.09e-06 [label_fine_grained_interleaved_index]: 2.39999e-06 [merge_cast_opt]: 1.49e-06 [slice_recompute_activation]: 1.91e-06 [micro_interleaved_order_control]: 2.07999e-06 [assign_add_opt]: 1.45001e-06 [ForceFp32Comm]: 8.60018e-07 [remove_cast_before_assign_add]: 1.25999e-06 [full_micro_interleaved_order_control]: 2.27001e-06 [reorder_send_recv_between_fp_bp]: 3.29001e-06 [comm_op_add_attrs]: 1.05999e-06 [add_comm_op_reuse_tag]: 1.02998e-06 [interleave_split_concat_branches]: 1.27e-06 [interleave_parallel_branches]: 1.24e-06 [overlap_opt_shard_in_pipeline]: 1.33002e-06 [overlap_opt_shard_grad_in_pipeline]: 1.72999e-06 [control_data_broadcast_order]: 1.735e-05 [grouped_pairwise_exchange_alltoall]: 1.39e-06 [offloading_packed_experts]: 4.87e-06 [overlap_recompute_and_grad_model_parallel]: 6.54999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.15999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.29e-06 [overlap_recompute_comm]: 2.31e-06 [overlap_grad_ring_attention]: 5.56e-06 [overlap_grad_flash_sp]: 2.598e-05 [begin_end_overlap_inline]: 4.90021e-07 [split_matmul_comm_elemetwise]: 2.34001e-06 [split_layernorm_comm]: 1.63002e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 9.983e-05, [1] [Cycle 1]: 9.359e-05, [6] [build]: 4.99998e-06 [elim_shapecalc]: 1.527e-05 [elim_not_effective]: 1.848e-05 [opt_reshape]: 8.99e-06 [fold_const_symbol]: 1.477e-05 [renormalize]: 1.80007e-07 [detach_backward]: 2.54001e-06 [pipeline_parallel_scheduler]: 1.54e-06 [auto_monad_reorder]: 2.234e-05 [get_jit_bprop_graph]: 1.96003e-06 [rewriter_after_jit_bprop_graph]: 6.19001e-06 [opt_after_jit_grad]: 0.00058713 [validate]: 5.62e-05 Sums bootstrap : 0.000494s : 0.06% type_inference : 0.865786s : 98.80% event_method : 0.000076s : 0.01% auto_monad : 0.000182s : 0.02% graph_reusing : 0.000011s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000063s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000013s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000084s : 0.01% 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 : 0.000007s : 0.00% optimize.rewriter_before_opt_a : 0.000424s : 0.05% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000250s : 0.03% optimize.opt_a.loop_unroll : 0.000096s : 0.01% optimize.opt_a.a_1 : 0.001900s : 0.22% optimize.opt_a.with_stream_mark : 0.000046s : 0.01% optimize.opt_a.recompute_prepare : 0.000022s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000034s : 0.00% optimize.opt_a.parameter_eliminate : 0.000010s : 0.00% optimize.opt_a.a_2 : 0.000248s : 0.03% optimize.opt_a.accelerated_algorithm : 0.000019s : 0.00% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000014s : 0.00% optimize.opt_a.shard_inline : 0.000017s : 0.00% optimize.opt_a.merge_send_recv : 0.000024s : 0.00% optimize.opt_a.auto_parallel : 0.000027s : 0.00% optimize.opt_a.parallel : 0.000032s : 0.00% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000010s : 0.00% optimize.opt_a.allreduce_fusion : 0.000009s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000026s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000019s : 0.00% optimize.opt_a.virtual_dataset : 0.000017s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.00% optimize.opt_a.virtual_output : 0.000016s : 0.00% optimize.opt_a.merge_forward : 0.000012s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000025s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000045s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000028s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000010s : 0.00% optimize.opt_a.meta_fg_expand : 0.000010s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000024s : 0.00% optimize.opt_a.a_after_grad : 0.000024s : 0.00% optimize.opt_a.renormalize : 0.002974s : 0.34% optimize.opt_a.add_forward_monad_depend : 0.000077s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000035s : 0.00% optimize.opt_a.cse : 0.000073s : 0.01% optimize.opt_a.a_3 : 0.000118s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000009s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000027s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000764s : 0.09% optimize.opt_b.b_1 : 0.000169s : 0.02% optimize.opt_b.b_2 : 0.000009s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000033s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000034s : 0.00% optimize.loop_unroll : 0.000494s : 0.06% optimize.opt_after_cconv.c_1 : 0.000036s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000028s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000051s : 0.01% optimize.tuple_transform.d_1 : 0.000054s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000068s : 0.01% optimize.cse_after_recomputation.cse : 0.000018s : 0.00% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 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 : 0.000002s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000017s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 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 : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000026s : 0.00% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000015s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 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 : 0.000022s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000587s : 0.07% validate : 0.000056s : 0.01% Time group info: ------[substitution.] 0.000569 75 8.80% : 0.000050s : 11: substitution.cast_eliminate 0.51% : 0.000003s : 4: substitution.elim_not_effective 0.40% : 0.000002s : 4: substitution.fold_const_symbol 1.17% : 0.000007s : 5: substitution.graph_param_transform 80.15% : 0.000456s : 24: substitution.inline 1.20% : 0.000007s : 8: substitution.j_node_and_user_rematch 1.64% : 0.000009s : 8: substitution.remove_not_recompute_node 1.21% : 0.000007s : 2: substitution.replace_old_param 3.39% : 0.000019s : 7: substitution.switch_simplify 1.52% : 0.000009s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.865662 2 66.11% : 0.572254s : 1: type_inference.infer 33.89% : 0.293408s : 1: type_inference.specialize ------[replace.] 0.000245 33 58.10% : 0.000142s : 24: replace.inline 36.38% : 0.000089s : 7: replace.switch_simplify 5.52% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000467 33 95.31% : 0.000445s : 24: match.inline 3.08% : 0.000014s : 7: match.switch_simplify 1.61% : 0.000007s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000415 2544 1.24% : 0.000005s : 32: predicate.accumulaten_eliminater 0.61% : 0.000003s : 5: predicate.ad_related_special_op_eliminate 0.33% : 0.000001s : 10: predicate.addn_check_dump 1.21% : 0.000005s : 32: predicate.addn_zero_filter 1.05% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.38% : 0.000010s : 42: predicate.arithmetic_simplify 1.50% : 0.000006s : 32: predicate.cast_eliminate 0.35% : 0.000001s : 10: predicate.check_bprop_eliminate 0.36% : 0.000001s : 10: predicate.compare_switch_simplify 0.11% : 0.000000s : 5: predicate.const_output_eliminate 0.40% : 0.000002s : 10: predicate.depend_value_elim 1.14% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.20% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.08% : 0.000004s : 32: predicate.dict_set_item_eliminator 0.72% : 0.000003s : 10: predicate.dumpgradient_eliminate 0.12% : 0.000001s : 5: predicate.elim_not_effective 0.45% : 0.000002s : 5: predicate.elim_shapecalc_of_broadcastargs 1.35% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.22% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.19% : 0.000005s : 37: predicate.environ_get_depend_swap 1.79% : 0.000007s : 47: predicate.environ_get_eliminate 1.18% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.03% : 0.000008s : 58: predicate.exchange_switch_depend_value 2.72% : 0.000011s : 58: predicate.float_depend_g_call 0.47% : 0.000002s : 10: predicate.float_environ_get_switch 0.52% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.10% : 0.000000s : 5: predicate.fold_const_symbol 0.43% : 0.000002s : 10: predicate.get_grad_eliminate 0.18% : 0.000001s : 5: predicate.graph_param_transform 0.42% : 0.000002s : 10: predicate.incorporate_call 0.30% : 0.000001s : 10: predicate.incorporate_call_switch 5.97% : 0.000025s : 120: predicate.inline 0.50% : 0.000002s : 10: predicate.inline_without_move 0.18% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.59% : 0.000002s : 10: predicate.less_batch_normalization 1.63% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.43% : 0.000010s : 76: predicate.load_eliminater 0.59% : 0.000002s : 5: predicate.loop_unroll_after_grad 3.78% : 0.000016s : 104: predicate.loop_unroll_before_grad 1.54% : 0.000006s : 42: predicate.make_slice_get_slice_eliminator 0.57% : 0.000002s : 10: predicate.merge_addn 0.39% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.36% : 0.000001s : 10: predicate.mini_step_allgather_replace 1.05% : 0.000004s : 32: predicate.minmaximum_grad 0.67% : 0.000003s : 5: predicate.mutable_eliminate 0.22% : 0.000001s : 5: predicate.opt_reshape 0.27% : 0.000001s : 5: predicate.parallel_virtual_node 2.49% : 0.000010s : 58: predicate.partial_defer_inline 1.40% : 0.000006s : 39: predicate.partial_eliminate 1.07% : 0.000004s : 32: predicate.print_const_string_wrapper 0.48% : 0.000002s : 10: predicate.reduce_all_const_elim 1.43% : 0.000006s : 32: predicate.reduce_eliminate 2.55% : 0.000011s : 76: predicate.redundant_stop_gradient_eliminater 0.26% : 0.000001s : 10: predicate.remove_not_recompute_node 1.11% : 0.000005s : 44: predicate.replace_applicator 0.29% : 0.000001s : 10: predicate.replace_old_param 0.13% : 0.000001s : 5: predicate.reset_defer_inline 1.18% : 0.000005s : 32: predicate.reshape_eliminate 0.42% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.24% : 0.000001s : 5: predicate.row_tensor_eliminate 0.44% : 0.000002s : 10: predicate.same_eliminate 0.25% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.55% : 0.000002s : 10: predicate.shard_identity_eliminate 0.44% : 0.000002s : 10: predicate.special_op_eliminate 0.81% : 0.000003s : 10: predicate.specialize_transform 0.61% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.55% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.18% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.27% : 0.000009s : 58: predicate.switch_defer_inline 2.62% : 0.000011s : 68: predicate.switch_layer_defer_inline 7.44% : 0.000031s : 191: predicate.switch_simplify 1.26% : 0.000005s : 32: predicate.tile_eliminate 1.26% : 0.000005s : 32: predicate.transpose_eliminate 1.61% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.49% : 0.000006s : 42: predicate.tuple_list_get_item_const_eliminator 1.39% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.63% : 0.000011s : 54: predicate.tuple_list_get_item_eliminator 1.52% : 0.000006s : 42: predicate.tuple_list_get_set_item_eliminator 2.20% : 0.000009s : 52: predicate.tuple_list_set_item_eliminator 1.55% : 0.000006s : 44: predicate.tuple_to_list_eliminator_ 2.51% : 0.000010s : 76: predicate.updatestate_pure_node_eliminater 2.85% : 0.000012s : 86: predicate.updatestate_useless_node_eliminater 0.30% : 0.000001s : 5: predicate.value_based_eliminate 0.44% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.43% : 0.000002s : 10: predicate.virtual_output_eliminate 0.19% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.28% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.292143 42 0.86% : 0.002502s : 16: func_graph_cloner_run.FuncGraphClonerGraph 99.14% : 0.289641s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.997636 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.39% : 0.003886s : 1: add_attr 0.39% : 0.003872s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000073s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.02% : 0.000189s : 1: auto_monad 0.00% : 0.000027s : 1: auto_monad_reorder 0.01% : 0.000104s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.05% : 0.000535s : 1: bootstrap 0.00% : 0.000037s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000020s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000032s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.01% : 0.000084s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000015s : 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 0.05% : 0.000505s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.08% : 0.000777s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000018s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000023s : 1: opt.transform.mutable_eliminate 0.27% : 0.002743s : 78: opt.transform.opt_a 0.00% : 0.000035s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000149s : 28: opt.transform.opt_b 0.01% : 0.000060s : 2: opt.transform.opt_trans_graph 0.01% : 0.000053s : 4: opt.transform.symbol_engine_opt 5.54% : 0.055243s : 1: opt_a 0.01% : 0.000126s : 1: opt_after_cconv 0.06% : 0.000601s : 1: opt_after_jit_grad 0.03% : 0.000277s : 1: opt_b 5.84% : 0.058220s : 1: optimize 0.00% : 0.000026s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000029s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 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 0.01% : 0.000089s : 1: pre_auto_parallel 0.00% : 0.000010s : 1: py_interpret_to_execute 0.00% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000056s : 1: remove_dup_value 0.16% : 0.001621s : 1: renormalize.infer 0.13% : 0.001338s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000031s : 1: rewriter_after_opt_a 0.04% : 0.000432s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000004s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000103s : 1: symbol_engine_optimizer 0.01% : 0.000092s : 1: tuple_transform 86.79% : 0.865820s : 1: type_inference TotalTime = 0.937612, [21] [bootstrap]: 0.00048021 [type_inference]: 0.779755 [event_method]: 7.134e-05 [auto_monad]: 0.00031215 [graph_reusing]: 2.233e-05 [inline]: 3.06001e-06 [add_attr]: 0.00448001, [1] [add_attr_with_inline]: 0.00446759, [1] [Cycle 1]: 0.00015329, [2] [tag_attr]: 8.568e-05 [meta_addattr_fg_expand]: 1.884e-05 [parallel-infer-symbol]: 3.90998e-06 [pre_auto_parallel]: 0.00011149 [insert-virtual-dataset]: 2.81999e-06 [parallel-infer-symbol-second]: 9.00007e-07 [dataset_repeat_opt]: 2.27999e-06 [pipeline_split]: 1.89999e-06 [optimize]: 0.151431, [53] [py_interpret_to_execute]: 8.32e-06 [rewriter_before_opt_a]: 0.00057269 [opt_a]: 0.148238, [2] [Cycle 1]: 0.146848, [45] [expand_dump_flag]: 8.05e-06 [switch_simplify]: 0.00034864 [loop_unroll]: 0.00011472 [a_1]: 0.0142853 [with_stream_mark]: 4.602e-05 [recompute_prepare]: 3.269e-05 [updatestate_depend_eliminate]: 1.133e-05 [updatestate_assign_eliminate]: 8.27e-06 [updatestate_loads_eliminate]: 7.73999e-06 [parameter_eliminate]: 4.10998e-06 [a_2]: 0.00023975 [accelerated_algorithm]: 1.741e-05 [shard]: 2.44999e-06 [meta_shard_fg_expand]: 1.517e-05 [shard_inline]: 1.528e-05 [merge_send_recv]: 2.216e-05 [auto_parallel]: 1.727e-05 [parallel]: 2.244e-05 [flash_sp]: 1.45e-05 [merge_comm]: 1.016e-05 [allreduce_fusion]: 9.11002e-06 [matmul_add_comm_reduction]: 3.628e-05 [allreduce_slice_to_reducescatter]: 9.60019e-07 [virtual_shard_identity]: 1.953e-05 [virtual_dataset]: 1.595e-05 [get_grad_eliminate_]: 1.48e-05 [virtual_output]: 1.56e-05 [merge_forward]: 1.098e-05 [cell_reuse_recompute_pass]: 2.02999e-06 [offload_activation]: 1.965e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.233e-05 [merge_recompute_call_nodes]: 1.74e-06 [before_grad]: 2.793e-05 [set_forward_comm_id_for_comm_node_pass]: 1.047e-05 [meta_fg_expand]: 0.121799 [flash_sp_send_recv_attached]: 7.1e-06 [receive_attached]: 2.93e-06 [after_resolve]: 0.00011764 [a_after_grad]: 0.00015282 [renormalize]: 0.00874197 [add_forward_monad_depend]: 1.234e-05 [auto_monad_grad]: 2.88998e-06 [auto_monad_eliminator]: 2.191e-05 [cse]: 3.425e-05 [a_3]: 5.283e-05 [Cycle 2]: 0.00137073, [45] [expand_dump_flag]: 3.18e-06 [switch_simplify]: 7.2e-06 [loop_unroll]: 4.26001e-06 [a_1]: 5.815e-05 [with_stream_mark]: 2.148e-05 [recompute_prepare]: 5.12e-06 [updatestate_depend_eliminate]: 3.46999e-06 [updatestate_assign_eliminate]: 2.78e-06 [updatestate_loads_eliminate]: 2.88998e-06 [parameter_eliminate]: 2.76999e-06 [a_2]: 3.936e-05 [accelerated_algorithm]: 4.2e-06 [shard]: 2.15002e-06 [meta_shard_fg_expand]: 2.83e-06 [shard_inline]: 3.9e-06 [merge_send_recv]: 3.432e-05 [auto_parallel]: 1.037e-05 [parallel]: 1.047e-05 [flash_sp]: 5.17999e-06 [merge_comm]: 3.28e-06 [allreduce_fusion]: 2.92002e-06 [matmul_add_comm_reduction]: 9.59999e-06 [allreduce_slice_to_reducescatter]: 7.89994e-07 [virtual_shard_identity]: 7.16001e-06 [virtual_dataset]: 4.13999e-06 [get_grad_eliminate_]: 3.85e-06 [virtual_output]: 3.8e-06 [merge_forward]: 3.53e-06 [cell_reuse_recompute_pass]: 3.5e-06 [offload_activation]: 8.48001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.538e-05 [merge_recompute_call_nodes]: 1.37999e-06 [before_grad]: 7.5e-06 [set_forward_comm_id_for_comm_node_pass]: 3.5e-06 [meta_fg_expand]: 0.0006966 [flash_sp_send_recv_attached]: 3.06001e-06 [receive_attached]: 3.09001e-06 [after_resolve]: 1.681e-05 [a_after_grad]: 8.47e-06 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 5.85002e-06 [auto_monad_grad]: 2.22001e-06 [auto_monad_eliminator]: 1.405e-05 [cse]: 2.689e-05 [a_3]: 2.852e-05 [py_interpret_to_execute_after_opt_a]: 1.344e-05 [slice_cell_reuse_recomputed_activation]: 3.4e-06 [rewriter_after_opt_a]: 2.511e-05 [convert_after_rewriter]: 1.29998e-06 [order_py_execute_after_rewriter]: 1.42999e-06 [mutable_eliminate]: 0.00079936 [opt_b]: 0.00021967, [1] [Cycle 1]: 0.00020836, [7] [b_1]: 7.674e-05 [b_2]: 5.45001e-06 [updatestate_depend_eliminate]: 5.169e-05 [updatestate_assign_eliminate]: 2.99001e-06 [updatestate_loads_eliminate]: 1.85001e-06 [renormalize]: 8.89995e-07 [cse]: 2.614e-05 [optimize_parallel_all_gather_comm]: 2.082e-05 [overlap_param_gather]: 2.19001e-06 [cconv]: 3.758e-05 [loop_unroll]: 0.0005879 [opt_after_cconv]: 9.427e-05, [1] [Cycle 1]: 8.619e-05, [7] [c_1]: 1.678e-05 [parameter_eliminate]: 5.71998e-06 [updatestate_depend_eliminate]: 6.26998e-06 [updatestate_assign_eliminate]: 1.77999e-06 [updatestate_loads_eliminate]: 1.53002e-06 [cse]: 1.977e-05 [renormalize]: 6.00005e-07 [remove_dup_value]: 1.761e-05 [tuple_transform]: 6.236e-05, [1] [Cycle 1]: 5.58e-05, [4] [d_1]: 3.168e-05 [none_parameter_eliminate]: 1.59998e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 4.92999e-06 [partial_unused_args_eliminate]: 1.91e-06 [add_recomputation]: 4.149e-05 [cse_after_recomputation]: 1.937e-05, [1] [Cycle 1]: 1.437e-05, [1] [cse]: 8.83001e-06 [environ_conv]: 5.68002e-06 [swap_dp_allreduce_reducescatter]: 3.86001e-06 [bias_add_comm_swap]: 3.42002e-06 [label_micro_interleaved_index]: 6.19001e-06 [label_fine_grained_interleaved_index]: 3.71999e-06 [merge_cast_opt]: 1.37e-06 [slice_recompute_activation]: 2.01998e-06 [micro_interleaved_order_control]: 3.01001e-06 [assign_add_opt]: 1.40999e-06 [ForceFp32Comm]: 1.02e-06 [remove_cast_before_assign_add]: 1.00999e-06 [full_micro_interleaved_order_control]: 2.83998e-06 [reorder_send_recv_between_fp_bp]: 2.88e-06 [comm_op_add_attrs]: 1.22999e-06 [add_comm_op_reuse_tag]: 1.39e-06 [interleave_split_concat_branches]: 1.17e-06 [interleave_parallel_branches]: 1.17e-06 [overlap_opt_shard_in_pipeline]: 1.22e-06 [overlap_opt_shard_grad_in_pipeline]: 2.29001e-06 [control_data_broadcast_order]: 1.642e-05 [grouped_pairwise_exchange_alltoall]: 1.54998e-06 [offloading_packed_experts]: 3.98999e-06 [overlap_recompute_and_grad_model_parallel]: 4.88001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.45001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.22999e-06 [overlap_recompute_comm]: 2.56e-06 [overlap_grad_ring_attention]: 3.68e-06 [overlap_grad_flash_sp]: 1.974e-05 [begin_end_overlap_inline]: 5.3001e-07 [split_matmul_comm_elemetwise]: 2.41e-06 [split_layernorm_comm]: 1.89e-06 [handle_group_info]: 1.10001e-06 [symbol_engine_optimizer]: 8.409e-05, [1] [Cycle 1]: 7.634e-05, [6] [build]: 4.92e-06 [elim_shapecalc]: 1.275e-05 [elim_not_effective]: 1.279e-05 [opt_reshape]: 4.79998e-06 [fold_const_symbol]: 6.74001e-06 [renormalize]: 2.10013e-07 [detach_backward]: 2.43998e-06 [pipeline_parallel_scheduler]: 1.37e-06 [auto_monad_reorder]: 1.932e-05 [get_jit_bprop_graph]: 2.61e-06 [rewriter_after_jit_bprop_graph]: 7.70998e-06 [opt_after_jit_grad]: 0.00063897 [validate]: 4.454e-05 Sums bootstrap : 0.000480s : 0.05% type_inference : 0.779755s : 83.70% event_method : 0.000071s : 0.01% auto_monad : 0.000312s : 0.03% graph_reusing : 0.000022s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000086s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000019s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000111s : 0.01% 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 : 0.000008s : 0.00% optimize.rewriter_before_opt_a : 0.000573s : 0.06% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000356s : 0.04% optimize.opt_a.loop_unroll : 0.000119s : 0.01% optimize.opt_a.a_1 : 0.014343s : 1.54% optimize.opt_a.with_stream_mark : 0.000068s : 0.01% optimize.opt_a.recompute_prepare : 0.000038s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_a.parameter_eliminate : 0.000007s : 0.00% optimize.opt_a.a_2 : 0.000279s : 0.03% optimize.opt_a.accelerated_algorithm : 0.000022s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000018s : 0.00% optimize.opt_a.shard_inline : 0.000019s : 0.00% optimize.opt_a.merge_send_recv : 0.000056s : 0.01% optimize.opt_a.auto_parallel : 0.000028s : 0.00% optimize.opt_a.parallel : 0.000033s : 0.00% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000013s : 0.00% optimize.opt_a.allreduce_fusion : 0.000012s : 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 : 0.000027s : 0.00% optimize.opt_a.virtual_dataset : 0.000020s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000019s : 0.00% optimize.opt_a.virtual_output : 0.000019s : 0.00% optimize.opt_a.merge_forward : 0.000015s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000028s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000048s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000035s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000014s : 0.00% optimize.opt_a.meta_fg_expand : 0.122495s : 13.15% 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 : 0.000134s : 0.01% optimize.opt_a.a_after_grad : 0.000161s : 0.02% optimize.opt_a.renormalize : 0.008742s : 0.94% optimize.opt_a.add_forward_monad_depend : 0.000018s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000036s : 0.00% optimize.opt_a.cse : 0.000061s : 0.01% optimize.opt_a.a_3 : 0.000081s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000013s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000025s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000799s : 0.09% optimize.opt_b.b_1 : 0.000077s : 0.01% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000052s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000026s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000038s : 0.00% optimize.loop_unroll : 0.000588s : 0.06% optimize.opt_after_cconv.c_1 : 0.000017s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000020s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000018s : 0.00% optimize.tuple_transform.d_1 : 0.000032s : 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 : 0.000005s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000041s : 0.00% optimize.cse_after_recomputation.cse : 0.000009s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 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 : 0.000004s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000016s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000020s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000013s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000013s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 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 : 0.000019s : 0.00% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000639s : 0.07% validate : 0.000045s : 0.00% Time group info: ------[substitution.] 0.001084 134 4.38% : 0.000047s : 8: substitution.cast_eliminate 0.17% : 0.000002s : 1: substitution.elim_not_effective 0.60% : 0.000006s : 9: substitution.float_depend_g_call 0.42% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.10% : 0.000001s : 1: substitution.fold_const_symbol 0.64% : 0.000007s : 1: substitution.graph_param_transform 0.34% : 0.000004s : 2: substitution.incorporate_call 0.23% : 0.000003s : 2: substitution.incorporate_call_switch 78.91% : 0.000855s : 33: substitution.inline 3.17% : 0.000034s : 4: substitution.inline_without_move 0.82% : 0.000009s : 11: substitution.j_node_and_user_rematch 0.34% : 0.000004s : 2: substitution.minmaximum_grad 0.63% : 0.000007s : 9: substitution.partial_eliminate 0.93% : 0.000010s : 11: substitution.remove_not_recompute_node 0.30% : 0.000003s : 1: substitution.replace_applicator 0.82% : 0.000009s : 12: substitution.replace_old_param 0.44% : 0.000005s : 1: substitution.set_cell_output_no_recompute 2.27% : 0.000025s : 10: substitution.switch_simplify 0.69% : 0.000007s : 2: substitution.tuple_list_convert_item_index_to_positive 0.30% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.45% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.63% : 0.000028s : 6: substitution.tuple_list_get_item_eliminator 0.44% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.779481 2 99.07% : 0.772222s : 1: type_inference.infer 0.93% : 0.007259s : 1: type_inference.specialize ------[replace.] 0.000414 45 60.96% : 0.000252s : 33: replace.inline 30.30% : 0.000125s : 10: replace.switch_simplify 8.74% : 0.000036s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000869 45 96.37% : 0.000837s : 33: match.inline 2.18% : 0.000019s : 10: match.switch_simplify 1.45% : 0.000013s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000560 3189 1.33% : 0.000007s : 43: predicate.accumulaten_eliminater 0.30% : 0.000002s : 1: predicate.ad_related_special_op_eliminate 0.36% : 0.000002s : 14: predicate.addn_check_dump 1.27% : 0.000007s : 43: predicate.addn_zero_filter 1.01% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 2.16% : 0.000012s : 57: predicate.arithmetic_simplify 1.54% : 0.000009s : 43: predicate.cast_eliminate 0.14% : 0.000001s : 2: predicate.check_bprop_eliminate 0.36% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.44% : 0.000002s : 14: predicate.depend_value_elim 1.24% : 0.000007s : 43: predicate.dict_get_item_const_eliminator 1.20% : 0.000007s : 43: predicate.dict_get_item_eliminator 1.11% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.30% : 0.000002s : 2: predicate.dumpgradient_eliminate 0.11% : 0.000001s : 1: predicate.elim_not_effective 0.14% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.37% : 0.000008s : 44: predicate.environ_add_const_eliminate 1.33% : 0.000007s : 44: predicate.environ_get_add_eliminate 1.06% : 0.000006s : 44: predicate.environ_get_depend_swap 2.57% : 0.000014s : 58: predicate.environ_get_eliminate 1.18% : 0.000007s : 44: predicate.environ_get_set_eliminate 2.01% : 0.000011s : 78: predicate.exchange_switch_depend_value 3.23% : 0.000018s : 78: predicate.float_depend_g_call 0.38% : 0.000002s : 14: predicate.float_environ_get_switch 0.49% : 0.000003s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.44% : 0.000002s : 14: predicate.get_grad_eliminate 0.07% : 0.000000s : 1: predicate.graph_param_transform 0.35% : 0.000002s : 14: predicate.incorporate_call 0.31% : 0.000002s : 14: predicate.incorporate_call_switch 6.28% : 0.000035s : 151: predicate.inline 2.64% : 0.000015s : 70: predicate.inline_without_move 0.19% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.60% : 0.000003s : 14: predicate.less_batch_normalization 1.60% : 0.000009s : 47: predicate.list_to_tuple_eliminator_ 2.17% : 0.000012s : 90: predicate.load_eliminater 0.38% : 0.000002s : 1: predicate.loop_unroll_after_grad 3.52% : 0.000020s : 141: predicate.loop_unroll_before_grad 1.26% : 0.000007s : 45: predicate.make_slice_get_slice_eliminator 0.39% : 0.000002s : 14: predicate.merge_addn 0.13% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.11% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.09% : 0.000006s : 43: predicate.minmaximum_grad 0.54% : 0.000003s : 1: predicate.mutable_eliminate 0.08% : 0.000000s : 1: predicate.opt_reshape 0.10% : 0.000001s : 1: predicate.parallel_virtual_node 2.89% : 0.000016s : 78: predicate.partial_defer_inline 1.27% : 0.000007s : 46: predicate.partial_eliminate 1.07% : 0.000006s : 43: predicate.print_const_string_wrapper 0.36% : 0.000002s : 14: predicate.reduce_all_const_elim 1.71% : 0.000010s : 43: predicate.reduce_eliminate 2.34% : 0.000013s : 90: predicate.redundant_stop_gradient_eliminater 0.25% : 0.000001s : 14: predicate.remove_not_recompute_node 1.21% : 0.000007s : 47: predicate.replace_applicator 1.26% : 0.000007s : 70: predicate.replace_old_param 0.19% : 0.000001s : 1: predicate.reset_defer_inline 1.46% : 0.000008s : 43: predicate.reshape_eliminate 0.23% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.16% : 0.000001s : 1: predicate.row_tensor_eliminate 0.27% : 0.000002s : 2: predicate.same_eliminate 0.24% : 0.000001s : 14: predicate.set_cell_output_no_recompute 0.54% : 0.000003s : 14: predicate.shard_identity_eliminate 0.14% : 0.000001s : 2: predicate.special_op_eliminate 0.40% : 0.000002s : 14: predicate.specialize_transform 0.39% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 2.37% : 0.000013s : 70: predicate.stack_unstack_eliminate 0.07% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.32% : 0.000013s : 78: predicate.switch_defer_inline 2.60% : 0.000015s : 80: predicate.switch_layer_defer_inline 7.31% : 0.000041s : 254: predicate.switch_simplify 1.30% : 0.000007s : 43: predicate.tile_eliminate 1.19% : 0.000007s : 43: predicate.transpose_eliminate 1.39% : 0.000008s : 45: predicate.tuple_list_convert_item_index_to_positive 1.53% : 0.000009s : 45: predicate.tuple_list_get_item_const_eliminator 1.22% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.46% : 0.000014s : 61: predicate.tuple_list_get_item_eliminator 1.27% : 0.000007s : 45: predicate.tuple_list_get_set_item_eliminator 1.99% : 0.000011s : 59: predicate.tuple_list_set_item_eliminator 1.38% : 0.000008s : 47: predicate.tuple_to_list_eliminator_ 2.23% : 0.000012s : 90: predicate.updatestate_pure_node_eliminater 2.91% : 0.000016s : 104: predicate.updatestate_useless_node_eliminater 0.10% : 0.000001s : 1: predicate.value_based_eliminate 0.52% : 0.000003s : 14: predicate.virtual_dataset_eliminate 0.44% : 0.000002s : 14: predicate.virtual_output_eliminate 0.03% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.09% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.006142 69 78.19% : 0.004802s : 33: func_graph_cloner_run.FuncGraphClonerGraph 21.81% : 0.001339s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.117893 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.40% : 0.004487s : 1: add_attr 0.40% : 0.004472s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.00% : 0.000046s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.03% : 0.000326s : 1: auto_monad 0.00% : 0.000024s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.05% : 0.000518s : 1: bootstrap 0.00% : 0.000041s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.01% : 0.000080s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000027s : 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% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000011s : 1: label_fine_grained_interleaved_index 0.00% : 0.000033s : 1: label_micro_interleaved_index 0.05% : 0.000599s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.07% : 0.000813s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000013s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 1.40% : 0.015601s : 78: opt.transform.opt_a 0.00% : 0.000016s : 1: opt.transform.opt_after_cconv 0.00% : 0.000022s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000052s : 28: opt.transform.opt_b 0.00% : 0.000034s : 2: opt.transform.opt_trans_graph 0.00% : 0.000032s : 4: opt.transform.symbol_engine_opt 13.26% : 0.148243s : 1: opt_a 0.01% : 0.000098s : 1: opt_after_cconv 0.06% : 0.000654s : 1: opt_after_jit_grad 0.02% : 0.000224s : 1: opt_b 13.55% : 0.151438s : 1: optimize 0.00% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000024s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.02% : 0.000178s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000008s : 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% : 0.000005s : 1: pipeline_split 0.01% : 0.000118s : 1: pre_auto_parallel 0.00% : 0.000012s : 1: py_interpret_to_execute 0.00% : 0.000017s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.00% : 0.000022s : 1: remove_dup_value 0.70% : 0.007771s : 1: renormalize.infer 0.09% : 0.000954s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000012s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000028s : 1: rewriter_after_opt_a 0.05% : 0.000601s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000087s : 1: symbol_engine_optimizer 0.01% : 0.000065s : 1: tuple_transform 69.76% : 0.779795s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:41:46.612.691 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.672788, [21] [bootstrap]: 0.00262019 [type_inference]: 0.657052 [event_method]: 0.00039052 [auto_monad]: 0.00016074 [graph_reusing]: 1.092e-05 [inline]: 3.15002e-06 [add_attr]: 0.0040528, [1] [add_attr_with_inline]: 0.0040401, [1] [Cycle 1]: 0.00010317, [2] [tag_attr]: 4.951e-05 [meta_addattr_fg_expand]: 9.94001e-06 [parallel-infer-symbol]: 3.42002e-06 [pre_auto_parallel]: 6.191e-05 [insert-virtual-dataset]: 2.69001e-06 [parallel-infer-symbol-second]: 8.80013e-07 [dataset_repeat_opt]: 2.27999e-06 [pipeline_split]: 1.59998e-06 [optimize]: 0.00756665, [53] [py_interpret_to_execute]: 5.74999e-06 [rewriter_before_opt_a]: 0.00030291 [opt_a]: 0.00479939, [2] [Cycle 1]: 0.0041741, [45] [expand_dump_flag]: 5.54998e-06 [switch_simplify]: 0.00016322 [loop_unroll]: 6.145e-05 [a_1]: 0.00123572 [with_stream_mark]: 1.775e-05 [recompute_prepare]: 1.004e-05 [updatestate_depend_eliminate]: 4.58001e-06 [updatestate_assign_eliminate]: 2.91e-06 [updatestate_loads_eliminate]: 2.71999e-06 [parameter_eliminate]: 2.14e-06 [a_2]: 7.608e-05 [accelerated_algorithm]: 7.53e-06 [shard]: 2.07001e-06 [meta_shard_fg_expand]: 2.81e-06 [shard_inline]: 5.71e-06 [merge_send_recv]: 8.62998e-06 [auto_parallel]: 8.81002e-06 [parallel]: 1.839e-05 [flash_sp]: 1.064e-05 [merge_comm]: 3.41999e-06 [allreduce_fusion]: 3.13998e-06 [matmul_add_comm_reduction]: 9.40001e-06 [allreduce_slice_to_reducescatter]: 7.80012e-07 [virtual_shard_identity]: 7.96001e-06 [virtual_dataset]: 6.21998e-06 [get_grad_eliminate_]: 5.74e-06 [virtual_output]: 5.67999e-06 [merge_forward]: 3.58999e-06 [cell_reuse_recompute_pass]: 1.02998e-06 [offload_activation]: 9.25999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.31e-05 [merge_recompute_call_nodes]: 1.42e-06 [before_grad]: 9.72001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.24001e-06 [meta_fg_expand]: 4.45999e-06 [flash_sp_send_recv_attached]: 2.40002e-06 [receive_attached]: 2.59999e-06 [after_resolve]: 9.02999e-06 [a_after_grad]: 8.45001e-06 [renormalize]: 0.00197976 [add_forward_monad_depend]: 1.015e-05 [auto_monad_grad]: 3.17002e-06 [auto_monad_eliminator]: 1.905e-05 [cse]: 3.228e-05 [a_3]: 5.25e-05 [Cycle 2]: 0.00060977, [45] [expand_dump_flag]: 2.17001e-06 [switch_simplify]: 8.00999e-06 [loop_unroll]: 5.14998e-06 [a_1]: 0.00010528 [with_stream_mark]: 1.693e-05 [recompute_prepare]: 5.71e-06 [updatestate_depend_eliminate]: 3.45998e-06 [updatestate_assign_eliminate]: 2.82002e-06 [updatestate_loads_eliminate]: 3.07002e-06 [parameter_eliminate]: 2.64999e-06 [a_2]: 5.856e-05 [accelerated_algorithm]: 5.90002e-06 [shard]: 2.13998e-06 [meta_shard_fg_expand]: 2.16e-06 [shard_inline]: 4.84003e-06 [merge_send_recv]: 8.13999e-06 [auto_parallel]: 8.73001e-06 [parallel]: 9.68002e-06 [flash_sp]: 3.85998e-06 [merge_comm]: 3.14999e-06 [allreduce_fusion]: 3.14001e-06 [matmul_add_comm_reduction]: 8.90001e-06 [allreduce_slice_to_reducescatter]: 8.60018e-07 [virtual_shard_identity]: 6.93e-06 [virtual_dataset]: 5.47999e-06 [get_grad_eliminate_]: 5.20001e-06 [virtual_output]: 4.87e-06 [merge_forward]: 4.46002e-06 [cell_reuse_recompute_pass]: 2.27999e-06 [offload_activation]: 9.52999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.521e-05 [merge_recompute_call_nodes]: 1.55001e-06 [before_grad]: 8.23999e-06 [set_forward_comm_id_for_comm_node_pass]: 3.79002e-06 [meta_fg_expand]: 2.27999e-06 [flash_sp_send_recv_attached]: 1.76003e-06 [receive_attached]: 2.34001e-06 [after_resolve]: 1.081e-05 [a_after_grad]: 7.68001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.52001e-06 [auto_monad_grad]: 1.67999e-06 [auto_monad_eliminator]: 5.73002e-06 [cse]: 1.347e-05 [a_3]: 3.038e-05 [py_interpret_to_execute_after_opt_a]: 3.507e-05 [slice_cell_reuse_recomputed_activation]: 2.67001e-06 [rewriter_after_opt_a]: 2.115e-05 [convert_after_rewriter]: 1.60001e-06 [order_py_execute_after_rewriter]: 1.18001e-06 [mutable_eliminate]: 0.00076588 [opt_b]: 0.00019038, [1] [Cycle 1]: 0.00018138, [7] [b_1]: 0.000104 [b_2]: 7.43e-06 [updatestate_depend_eliminate]: 7.80998e-06 [updatestate_assign_eliminate]: 2.46e-06 [updatestate_loads_eliminate]: 2.51e-06 [renormalize]: 1.10999e-06 [cse]: 2.117e-05 [optimize_parallel_all_gather_comm]: 4.8e-05 [overlap_param_gather]: 3.09999e-06 [cconv]: 3.519e-05 [loop_unroll]: 0.0005227 [opt_after_cconv]: 0.0001455, [1] [Cycle 1]: 0.00013828, [7] [c_1]: 2.588e-05 [parameter_eliminate]: 5.22e-06 [updatestate_depend_eliminate]: 5.59e-06 [updatestate_assign_eliminate]: 2.21e-06 [updatestate_loads_eliminate]: 1.99999e-06 [cse]: 2.14e-05 [renormalize]: 3.60014e-07 [remove_dup_value]: 1.579e-05 [tuple_transform]: 6.977e-05, [1] [Cycle 1]: 6.504e-05, [4] [d_1]: 3.93e-05 [none_parameter_eliminate]: 2.04e-06 [renormalize]: 1.60013e-07 [switch_simplify]: 5.59998e-06 [partial_unused_args_eliminate]: 1.79998e-06 [add_recomputation]: 5.35e-05 [cse_after_recomputation]: 2.068e-05, [1] [Cycle 1]: 1.597e-05, [1] [cse]: 1.068e-05 [environ_conv]: 6.21e-06 [swap_dp_allreduce_reducescatter]: 4.63001e-06 [bias_add_comm_swap]: 2.81999e-06 [label_micro_interleaved_index]: 5.72001e-06 [label_fine_grained_interleaved_index]: 2.96999e-06 [merge_cast_opt]: 1.39e-06 [slice_recompute_activation]: 2.18998e-06 [micro_interleaved_order_control]: 4.17e-06 [assign_add_opt]: 1.42999e-06 [ForceFp32Comm]: 1.12999e-06 [remove_cast_before_assign_add]: 9.80013e-07 [full_micro_interleaved_order_control]: 2.31998e-06 [reorder_send_recv_between_fp_bp]: 2.78998e-06 [comm_op_add_attrs]: 1.17e-06 [add_comm_op_reuse_tag]: 1.27e-06 [interleave_split_concat_branches]: 1.09e-06 [interleave_parallel_branches]: 1.12999e-06 [overlap_opt_shard_in_pipeline]: 1.23002e-06 [overlap_opt_shard_grad_in_pipeline]: 2.15002e-06 [control_data_broadcast_order]: 1.602e-05 [grouped_pairwise_exchange_alltoall]: 1.37e-06 [offloading_packed_experts]: 3.94002e-06 [overlap_recompute_and_grad_model_parallel]: 4.91997e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12e-06 [overlap_recompute_allgather_and_fa_grad]: 1.22e-06 [overlap_recompute_comm]: 2.93998e-06 [overlap_grad_ring_attention]: 3.66999e-06 [overlap_grad_flash_sp]: 2.351e-05 [begin_end_overlap_inline]: 5.19998e-07 [split_matmul_comm_elemetwise]: 2.27001e-06 [split_layernorm_comm]: 2.02001e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 8.207e-05, [1] [Cycle 1]: 7.608e-05, [6] [build]: 4.38001e-06 [elim_shapecalc]: 1.142e-05 [elim_not_effective]: 1.391e-05 [opt_reshape]: 6.69999e-06 [fold_const_symbol]: 9.29e-06 [renormalize]: 2.79979e-07 [detach_backward]: 2.51e-06 [pipeline_parallel_scheduler]: 1.81998e-06 [auto_monad_reorder]: 1.818e-05 [get_jit_bprop_graph]: 2.22001e-06 [rewriter_after_jit_bprop_graph]: 6.53e-06 [opt_after_jit_grad]: 0.00055928 [validate]: 4.307e-05 Sums bootstrap : 0.002620s : 0.39% type_inference : 0.657052s : 98.43% event_method : 0.000391s : 0.06% auto_monad : 0.000161s : 0.02% graph_reusing : 0.000011s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000050s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000010s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000062s : 0.01% 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 : 0.000006s : 0.00% optimize.rewriter_before_opt_a : 0.000303s : 0.05% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000171s : 0.03% optimize.opt_a.loop_unroll : 0.000067s : 0.01% optimize.opt_a.a_1 : 0.001341s : 0.20% optimize.opt_a.with_stream_mark : 0.000035s : 0.01% optimize.opt_a.recompute_prepare : 0.000016s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000135s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000013s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.00% optimize.opt_a.shard_inline : 0.000011s : 0.00% optimize.opt_a.merge_send_recv : 0.000017s : 0.00% optimize.opt_a.auto_parallel : 0.000018s : 0.00% optimize.opt_a.parallel : 0.000028s : 0.00% optimize.opt_a.flash_sp : 0.000014s : 0.00% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000006s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000018s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000015s : 0.00% optimize.opt_a.virtual_dataset : 0.000012s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.00% optimize.opt_a.virtual_output : 0.000011s : 0.00% optimize.opt_a.merge_forward : 0.000008s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000019s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000028s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000018s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.00% optimize.opt_a.meta_fg_expand : 0.000007s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000020s : 0.00% optimize.opt_a.a_after_grad : 0.000016s : 0.00% optimize.opt_a.renormalize : 0.001980s : 0.30% optimize.opt_a.add_forward_monad_depend : 0.000012s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000025s : 0.00% optimize.opt_a.cse : 0.000046s : 0.01% optimize.opt_a.a_3 : 0.000083s : 0.01% 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 : 0.000021s : 0.00% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000766s : 0.11% optimize.opt_b.b_1 : 0.000104s : 0.02% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000021s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000048s : 0.01% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000035s : 0.01% optimize.loop_unroll : 0.000523s : 0.08% optimize.opt_after_cconv.c_1 : 0.000026s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000021s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.00% optimize.tuple_transform.d_1 : 0.000039s : 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 : 0.000006s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000053s : 0.01% optimize.cse_after_recomputation.cse : 0.000011s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000004s : 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 : 0.000016s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000024s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000018s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000559s : 0.08% validate : 0.000043s : 0.01% Time group info: ------[substitution.] 0.000358 39 3.29% : 0.000012s : 1: substitution.cast_eliminate 0.66% : 0.000002s : 2: substitution.elim_not_effective 0.40% : 0.000001s : 2: substitution.fold_const_symbol 1.48% : 0.000005s : 3: substitution.graph_param_transform 83.52% : 0.000299s : 15: substitution.inline 0.99% : 0.000004s : 4: substitution.j_node_and_user_rematch 1.50% : 0.000005s : 4: substitution.remove_not_recompute_node 1.53% : 0.000006s : 2: substitution.replace_old_param 3.88% : 0.000014s : 4: substitution.switch_simplify 2.74% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.656917 2 78.75% : 0.517318s : 1: type_inference.infer 21.25% : 0.139600s : 1: type_inference.specialize ------[replace.] 0.000168 22 3.67% : 0.000006s : 1: replace.cast_eliminate 54.74% : 0.000092s : 15: replace.inline 32.72% : 0.000055s : 4: replace.switch_simplify 8.87% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000323 22 3.40% : 0.000011s : 1: match.cast_eliminate 90.31% : 0.000292s : 15: match.inline 3.52% : 0.000011s : 4: match.switch_simplify 2.76% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000303 1716 1.17% : 0.000004s : 22: predicate.accumulaten_eliminater 0.76% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.34% : 0.000001s : 6: predicate.addn_check_dump 1.28% : 0.000004s : 22: predicate.addn_zero_filter 1.02% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 1.98% : 0.000006s : 28: predicate.arithmetic_simplify 1.66% : 0.000005s : 23: predicate.cast_eliminate 0.32% : 0.000001s : 6: predicate.check_bprop_eliminate 0.28% : 0.000001s : 6: predicate.compare_switch_simplify 0.09% : 0.000000s : 3: predicate.const_output_eliminate 0.39% : 0.000001s : 6: predicate.depend_value_elim 1.20% : 0.000004s : 23: predicate.dict_get_item_const_eliminator 1.28% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.10% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.61% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.12% : 0.000000s : 3: predicate.elim_not_effective 0.30% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.36% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.28% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.19% : 0.000004s : 26: predicate.environ_get_depend_swap 1.59% : 0.000005s : 32: predicate.environ_get_eliminate 1.18% : 0.000004s : 26: predicate.environ_get_set_eliminate 1.96% : 0.000006s : 40: predicate.exchange_switch_depend_value 2.85% : 0.000009s : 40: predicate.float_depend_g_call 0.31% : 0.000001s : 6: predicate.float_environ_get_switch 0.40% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.09% : 0.000000s : 3: predicate.fold_const_symbol 0.38% : 0.000001s : 6: predicate.get_grad_eliminate 0.17% : 0.000001s : 3: predicate.graph_param_transform 0.30% : 0.000001s : 6: predicate.incorporate_call 0.26% : 0.000001s : 6: predicate.incorporate_call_switch 5.54% : 0.000017s : 80: predicate.inline 0.42% : 0.000001s : 6: predicate.inline_without_move 0.18% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.78% : 0.000002s : 6: predicate.less_batch_normalization 1.76% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.45% : 0.000007s : 53: predicate.load_eliminater 0.75% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.53% : 0.000011s : 69: predicate.loop_unroll_before_grad 1.60% : 0.000005s : 29: predicate.make_slice_get_slice_eliminator 0.43% : 0.000001s : 6: predicate.merge_addn 0.31% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.32% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.04% : 0.000003s : 22: predicate.minmaximum_grad 0.99% : 0.000003s : 3: predicate.mutable_eliminate 0.24% : 0.000001s : 3: predicate.opt_reshape 0.34% : 0.000001s : 3: predicate.parallel_virtual_node 2.56% : 0.000008s : 40: predicate.partial_defer_inline 1.42% : 0.000004s : 28: predicate.partial_eliminate 1.07% : 0.000003s : 22: predicate.print_const_string_wrapper 0.32% : 0.000001s : 6: predicate.reduce_all_const_elim 1.34% : 0.000004s : 22: predicate.reduce_eliminate 2.47% : 0.000007s : 53: predicate.redundant_stop_gradient_eliminater 0.36% : 0.000001s : 6: predicate.remove_not_recompute_node 1.45% : 0.000004s : 31: predicate.replace_applicator 0.30% : 0.000001s : 6: predicate.replace_old_param 0.23% : 0.000001s : 3: predicate.reset_defer_inline 1.22% : 0.000004s : 22: predicate.reshape_eliminate 0.34% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.25% : 0.000001s : 3: predicate.row_tensor_eliminate 0.60% : 0.000002s : 6: predicate.same_eliminate 0.23% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.51% : 0.000002s : 6: predicate.shard_identity_eliminate 0.50% : 0.000002s : 6: predicate.special_op_eliminate 0.45% : 0.000001s : 6: predicate.specialize_transform 0.64% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.53% : 0.000002s : 6: predicate.stack_unstack_eliminate 0.22% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.22% : 0.000007s : 40: predicate.switch_defer_inline 2.47% : 0.000007s : 46: predicate.switch_layer_defer_inline 7.19% : 0.000022s : 126: predicate.switch_simplify 1.16% : 0.000003s : 22: predicate.tile_eliminate 1.13% : 0.000003s : 22: predicate.transpose_eliminate 1.74% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.83% : 0.000006s : 29: predicate.tuple_list_get_item_const_eliminator 1.48% : 0.000004s : 29: predicate.tuple_list_get_item_depend_reorder 3.14% : 0.000010s : 37: predicate.tuple_list_get_item_eliminator 1.56% : 0.000005s : 29: predicate.tuple_list_get_set_item_eliminator 2.45% : 0.000007s : 35: predicate.tuple_list_set_item_eliminator 1.64% : 0.000005s : 31: predicate.tuple_to_list_eliminator_ 2.39% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 2.92% : 0.000009s : 59: predicate.updatestate_useless_node_eliminater 0.28% : 0.000001s : 3: predicate.value_based_eliminate 0.46% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.45% : 0.000001s : 6: predicate.virtual_output_eliminate 0.13% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.43% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002268 27 58.49% : 0.001326s : 10: func_graph_cloner_run.FuncGraphClonerGraph 41.51% : 0.000941s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.688351 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.59% : 0.004059s : 1: add_attr 0.59% : 0.004045s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000058s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000173s : 1: auto_monad 0.00% : 0.000022s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.39% : 0.002666s : 1: bootstrap 0.01% : 0.000039s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000019s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000023s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.06% : 0.000408s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000016s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000067s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.08% : 0.000534s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000011s : 1: micro_interleaved_order_control 0.11% : 0.000779s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 0.27% : 0.001872s : 78: opt.transform.opt_a 0.00% : 0.000025s : 1: opt.transform.opt_after_cconv 0.00% : 0.000024s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000083s : 28: opt.transform.opt_b 0.01% : 0.000043s : 2: opt.transform.opt_trans_graph 0.01% : 0.000038s : 4: opt.transform.symbol_engine_opt 0.70% : 0.004803s : 1: opt_a 0.02% : 0.000149s : 1: opt_after_cconv 0.08% : 0.000572s : 1: opt_after_jit_grad 0.03% : 0.000194s : 1: opt_b 1.10% : 0.007572s : 1: optimize 0.01% : 0.000055s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000027s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000008s : 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% : 0.000004s : 1: pipeline_split 0.01% : 0.000067s : 1: pre_auto_parallel 0.00% : 0.000009s : 1: py_interpret_to_execute 0.01% : 0.000040s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000020s : 1: remove_dup_value 0.15% : 0.001058s : 1: renormalize.infer 0.13% : 0.000910s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000025s : 1: rewriter_after_opt_a 0.05% : 0.000311s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000023s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000085s : 1: symbol_engine_optimizer 0.01% : 0.000073s : 1: tuple_transform 95.46% : 0.657093s : 1: type_inference TotalTime = 0.449605, [21] [bootstrap]: 0.00049554 [type_inference]: 0.40976 [event_method]: 0.00012092 [auto_monad]: 0.00021056 [graph_reusing]: 1.371e-05 [inline]: 3.41001e-06 [add_attr]: 0.00398554, [1] [add_attr_with_inline]: 0.00397267, [1] [Cycle 1]: 0.00013425, [2] [tag_attr]: 7.43e-05 [meta_addattr_fg_expand]: 1.503e-05 [parallel-infer-symbol]: 3.84002e-06 [pre_auto_parallel]: 9.07e-05 [insert-virtual-dataset]: 2.52001e-06 [parallel-infer-symbol-second]: 1.01002e-06 [dataset_repeat_opt]: 2.17999e-06 [pipeline_split]: 1.71e-06 [optimize]: 0.0340932, [53] [py_interpret_to_execute]: 8.26002e-06 [rewriter_before_opt_a]: 0.00046941 [opt_a]: 0.0312136, [2] [Cycle 1]: 0.0302811, [45] [expand_dump_flag]: 8.70001e-06 [switch_simplify]: 0.00026564 [loop_unroll]: 8.92e-05 [a_1]: 0.0227559 [with_stream_mark]: 4.041e-05 [recompute_prepare]: 3.27e-05 [updatestate_depend_eliminate]: 1.175e-05 [updatestate_assign_eliminate]: 7.60998e-06 [updatestate_loads_eliminate]: 6.45002e-06 [parameter_eliminate]: 4.90999e-06 [a_2]: 0.00020806 [accelerated_algorithm]: 1.544e-05 [shard]: 1.66e-06 [meta_shard_fg_expand]: 9.12001e-06 [shard_inline]: 1.315e-05 [merge_send_recv]: 2.073e-05 [auto_parallel]: 1.571e-05 [parallel]: 2.149e-05 [flash_sp]: 1.332e-05 [merge_comm]: 9.25001e-06 [allreduce_fusion]: 8.32e-06 [matmul_add_comm_reduction]: 3.493e-05 [allreduce_slice_to_reducescatter]: 8.80013e-07 [virtual_shard_identity]: 1.921e-05 [virtual_dataset]: 1.255e-05 [get_grad_eliminate_]: 1.262e-05 [virtual_output]: 1.284e-05 [merge_forward]: 9.91e-06 [cell_reuse_recompute_pass]: 1.64998e-06 [offload_activation]: 1.773e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.794e-05 [merge_recompute_call_nodes]: 1.51002e-06 [before_grad]: 2.528e-05 [set_forward_comm_id_for_comm_node_pass]: 9.54999e-06 [meta_fg_expand]: 0.00223585 [flash_sp_send_recv_attached]: 4.53999e-06 [receive_attached]: 2.46e-06 [after_resolve]: 0.00022103 [a_after_grad]: 9.107e-05 [renormalize]: 0.00346026 [add_forward_monad_depend]: 8.77e-06 [auto_monad_grad]: 2.73998e-06 [auto_monad_eliminator]: 2.061e-05 [cse]: 3.284e-05 [a_3]: 4.937e-05 [Cycle 2]: 0.0009137, [45] [expand_dump_flag]: 2.76e-06 [switch_simplify]: 7.29001e-06 [loop_unroll]: 4.46002e-06 [a_1]: 5.48e-05 [with_stream_mark]: 1.792e-05 [recompute_prepare]: 5.29e-06 [updatestate_depend_eliminate]: 3.37002e-06 [updatestate_assign_eliminate]: 2.68e-06 [updatestate_loads_eliminate]: 2.48998e-06 [parameter_eliminate]: 2.41e-06 [a_2]: 3.972e-05 [accelerated_algorithm]: 4.58999e-06 [shard]: 2.16e-06 [meta_shard_fg_expand]: 2.16e-06 [shard_inline]: 3.18e-06 [merge_send_recv]: 7.68001e-06 [auto_parallel]: 8.65999e-06 [parallel]: 9.44e-06 [flash_sp]: 4.08001e-06 [merge_comm]: 2.53003e-06 [allreduce_fusion]: 2.40002e-06 [matmul_add_comm_reduction]: 8.33999e-06 [allreduce_slice_to_reducescatter]: 8.2e-07 [virtual_shard_identity]: 4.94e-06 [virtual_dataset]: 3.61999e-06 [get_grad_eliminate_]: 3.71001e-06 [virtual_output]: 3.79002e-06 [merge_forward]: 3.86999e-06 [cell_reuse_recompute_pass]: 2.71999e-06 [offload_activation]: 8.33999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.705e-05 [merge_recompute_call_nodes]: 1.40001e-06 [before_grad]: 6.84001e-06 [set_forward_comm_id_for_comm_node_pass]: 2.97002e-06 [meta_fg_expand]: 0.00035794 [flash_sp_send_recv_attached]: 2.17001e-06 [receive_attached]: 2.99999e-06 [after_resolve]: 1.317e-05 [a_after_grad]: 5.55001e-06 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 3.4e-06 [auto_monad_grad]: 2.02999e-06 [auto_monad_eliminator]: 1.156e-05 [cse]: 1.907e-05 [a_3]: 2.274e-05 [py_interpret_to_execute_after_opt_a]: 1.096e-05 [slice_cell_reuse_recomputed_activation]: 3.5e-06 [rewriter_after_opt_a]: 2.06e-05 [convert_after_rewriter]: 1.20999e-06 [order_py_execute_after_rewriter]: 1.07998e-06 [mutable_eliminate]: 0.00079367 [opt_b]: 0.00016293, [1] [Cycle 1]: 0.00015126, [7] [b_1]: 7.404e-05 [b_2]: 6.20002e-06 [updatestate_depend_eliminate]: 7.70998e-06 [updatestate_assign_eliminate]: 2.81e-06 [updatestate_loads_eliminate]: 2.01e-06 [renormalize]: 6.79982e-07 [cse]: 2.165e-05 [optimize_parallel_all_gather_comm]: 1.985e-05 [overlap_param_gather]: 2.26e-06 [cconv]: 3.686e-05 [loop_unroll]: 0.00050212 [opt_after_cconv]: 8.881e-05, [1] [Cycle 1]: 8.214e-05, [7] [c_1]: 1.475e-05 [parameter_eliminate]: 5.17e-06 [updatestate_depend_eliminate]: 6.03002e-06 [updatestate_assign_eliminate]: 1.63002e-06 [updatestate_loads_eliminate]: 1.55999e-06 [cse]: 1.808e-05 [renormalize]: 7.00005e-07 [remove_dup_value]: 1.509e-05 [tuple_transform]: 5.513e-05, [1] [Cycle 1]: 4.968e-05, [4] [d_1]: 2.645e-05 [none_parameter_eliminate]: 1.80001e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 4.17e-06 [partial_unused_args_eliminate]: 1.97999e-06 [add_recomputation]: 3.893e-05 [cse_after_recomputation]: 1.829e-05, [1] [Cycle 1]: 1.339e-05, [1] [cse]: 8.08999e-06 [environ_conv]: 5.39e-06 [swap_dp_allreduce_reducescatter]: 4.26001e-06 [bias_add_comm_swap]: 3.11001e-06 [label_micro_interleaved_index]: 5.29e-06 [label_fine_grained_interleaved_index]: 2.83e-06 [merge_cast_opt]: 1.34e-06 [slice_recompute_activation]: 1.99e-06 [micro_interleaved_order_control]: 2.14e-06 [assign_add_opt]: 1.30999e-06 [ForceFp32Comm]: 8.89995e-07 [remove_cast_before_assign_add]: 1.06997e-06 [full_micro_interleaved_order_control]: 2.36e-06 [reorder_send_recv_between_fp_bp]: 3.01001e-06 [comm_op_add_attrs]: 1.00999e-06 [add_comm_op_reuse_tag]: 1.15001e-06 [interleave_split_concat_branches]: 1.11002e-06 [interleave_parallel_branches]: 1.72001e-06 [overlap_opt_shard_in_pipeline]: 1.45999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.86998e-06 [control_data_broadcast_order]: 1.233e-05 [grouped_pairwise_exchange_alltoall]: 1.52001e-06 [offloading_packed_experts]: 2.96001e-06 [overlap_recompute_and_grad_model_parallel]: 4.25e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.32e-06 [overlap_recompute_allgather_and_fa_grad]: 1.19e-06 [overlap_recompute_comm]: 2.74999e-06 [overlap_grad_ring_attention]: 3.73001e-06 [overlap_grad_flash_sp]: 1.889e-05 [begin_end_overlap_inline]: 5.10016e-07 [split_matmul_comm_elemetwise]: 2.16e-06 [split_layernorm_comm]: 1.81e-06 [handle_group_info]: 1.03001e-06 [symbol_engine_optimizer]: 0.00018092, [1] [Cycle 1]: 0.0001753, [6] [build]: 3.68e-06 [elim_shapecalc]: 8.60999e-06 [elim_not_effective]: 0.00011555 [opt_reshape]: 6.07001e-06 [fold_const_symbol]: 7.58001e-06 [renormalize]: 2.50002e-07 [detach_backward]: 2.21e-06 [pipeline_parallel_scheduler]: 1.55999e-06 [auto_monad_reorder]: 1.73e-05 [get_jit_bprop_graph]: 2.25002e-06 [rewriter_after_jit_bprop_graph]: 7.21999e-06 [opt_after_jit_grad]: 0.00054687 [validate]: 4.072e-05 Sums bootstrap : 0.000496s : 0.11% type_inference : 0.409760s : 92.21% event_method : 0.000121s : 0.03% auto_monad : 0.000211s : 0.05% graph_reusing : 0.000014s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000074s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000015s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000091s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000008s : 0.00% optimize.rewriter_before_opt_a : 0.000469s : 0.11% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000273s : 0.06% optimize.opt_a.loop_unroll : 0.000094s : 0.02% optimize.opt_a.a_1 : 0.022811s : 5.13% optimize.opt_a.with_stream_mark : 0.000058s : 0.01% optimize.opt_a.recompute_prepare : 0.000038s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000007s : 0.00% optimize.opt_a.a_2 : 0.000248s : 0.06% optimize.opt_a.accelerated_algorithm : 0.000020s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000011s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.00% optimize.opt_a.merge_send_recv : 0.000028s : 0.01% optimize.opt_a.auto_parallel : 0.000024s : 0.01% optimize.opt_a.parallel : 0.000031s : 0.01% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000012s : 0.00% optimize.opt_a.allreduce_fusion : 0.000011s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000043s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000024s : 0.01% optimize.opt_a.virtual_dataset : 0.000016s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.00% optimize.opt_a.virtual_output : 0.000017s : 0.00% optimize.opt_a.merge_forward : 0.000014s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000045s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000032s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000013s : 0.00% optimize.opt_a.meta_fg_expand : 0.002594s : 0.58% 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 : 0.000234s : 0.05% optimize.opt_a.a_after_grad : 0.000097s : 0.02% optimize.opt_a.renormalize : 0.003460s : 0.78% optimize.opt_a.add_forward_monad_depend : 0.000012s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000032s : 0.01% optimize.opt_a.cse : 0.000052s : 0.01% optimize.opt_a.a_3 : 0.000072s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000794s : 0.18% optimize.opt_b.b_1 : 0.000074s : 0.02% optimize.opt_b.b_2 : 0.000006s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000022s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000037s : 0.01% optimize.loop_unroll : 0.000502s : 0.11% optimize.opt_after_cconv.c_1 : 0.000015s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000018s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000026s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000039s : 0.01% optimize.cse_after_recomputation.cse : 0.000008s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000002s : 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 : 0.000012s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000019s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000116s : 0.03% optimize.symbol_engine_optimizer.opt_reshape : 0.000006s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000008s : 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 : 0.000017s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000547s : 0.12% validate : 0.000041s : 0.01% Time group info: ------[substitution.] 0.000859 102 1.56% : 0.000013s : 1: substitution.cast_eliminate 0.20% : 0.000002s : 1: substitution.elim_not_effective 1.33% : 0.000011s : 9: substitution.float_depend_g_call 0.50% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.13% : 0.000001s : 1: substitution.fold_const_symbol 0.52% : 0.000004s : 1: substitution.graph_param_transform 0.39% : 0.000003s : 2: substitution.incorporate_call 0.28% : 0.000002s : 2: substitution.incorporate_call_switch 77.64% : 0.000667s : 24: substitution.inline 3.24% : 0.000028s : 2: substitution.inline_without_move 0.95% : 0.000008s : 9: substitution.j_node_and_user_rematch 0.49% : 0.000004s : 2: substitution.minmaximum_grad 1.00% : 0.000009s : 9: substitution.partial_eliminate 0.97% : 0.000008s : 9: substitution.remove_not_recompute_node 0.30% : 0.000003s : 1: substitution.replace_applicator 0.81% : 0.000007s : 5: substitution.replace_old_param 0.65% : 0.000006s : 1: substitution.set_cell_output_no_recompute 2.38% : 0.000020s : 7: substitution.switch_simplify 1.10% : 0.000009s : 2: substitution.tuple_list_convert_item_index_to_positive 0.37% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.72% : 0.000006s : 2: substitution.tuple_list_get_item_depend_reorder 3.87% : 0.000033s : 6: substitution.tuple_list_get_item_eliminator 0.61% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.409318 2 98.80% : 0.404395s : 1: type_inference.infer 1.20% : 0.004923s : 1: type_inference.specialize ------[replace.] 0.000318 34 2.49% : 0.000008s : 1: replace.cast_eliminate 64.22% : 0.000204s : 24: replace.inline 28.63% : 0.000091s : 7: replace.switch_simplify 4.66% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000694 34 1.75% : 0.000012s : 1: match.cast_eliminate 94.07% : 0.000653s : 24: match.inline 2.34% : 0.000016s : 7: match.switch_simplify 1.84% : 0.000013s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000448 2554 1.34% : 0.000006s : 35: predicate.accumulaten_eliminater 0.30% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.44% : 0.000002s : 12: predicate.addn_check_dump 1.18% : 0.000005s : 35: predicate.addn_zero_filter 1.15% : 0.000005s : 35: predicate.adjust_all_reduce_mul_add 2.39% : 0.000011s : 47: predicate.arithmetic_simplify 1.58% : 0.000007s : 36: predicate.cast_eliminate 0.14% : 0.000001s : 2: predicate.check_bprop_eliminate 0.44% : 0.000002s : 12: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.41% : 0.000002s : 12: predicate.depend_value_elim 1.20% : 0.000005s : 36: predicate.dict_get_item_const_eliminator 1.29% : 0.000006s : 36: predicate.dict_get_item_eliminator 1.24% : 0.000006s : 36: predicate.dict_set_item_eliminator 0.36% : 0.000002s : 2: predicate.dumpgradient_eliminate 0.07% : 0.000000s : 1: predicate.elim_not_effective 0.11% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.44% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.16% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.19% : 0.000005s : 37: predicate.environ_get_depend_swap 1.83% : 0.000008s : 49: predicate.environ_get_eliminate 1.16% : 0.000005s : 37: predicate.environ_get_set_eliminate 3.07% : 0.000014s : 62: predicate.exchange_switch_depend_value 3.64% : 0.000016s : 62: predicate.float_depend_g_call 0.38% : 0.000002s : 12: predicate.float_environ_get_switch 0.44% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.46% : 0.000002s : 12: predicate.get_grad_eliminate 0.03% : 0.000000s : 1: predicate.graph_param_transform 0.42% : 0.000002s : 12: predicate.incorporate_call 0.34% : 0.000002s : 12: predicate.incorporate_call_switch 5.76% : 0.000026s : 123: predicate.inline 1.68% : 0.000008s : 33: predicate.inline_without_move 0.20% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.59% : 0.000003s : 12: predicate.less_batch_normalization 1.54% : 0.000007s : 40: predicate.list_to_tuple_eliminator_ 2.46% : 0.000011s : 75: predicate.load_eliminater 0.23% : 0.000001s : 1: predicate.loop_unroll_after_grad 3.49% : 0.000016s : 108: predicate.loop_unroll_before_grad 1.31% : 0.000006s : 38: predicate.make_slice_get_slice_eliminator 0.45% : 0.000002s : 12: predicate.merge_addn 0.14% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.08% : 0.000000s : 2: predicate.mini_step_allgather_replace 1.15% : 0.000005s : 35: predicate.minmaximum_grad 0.42% : 0.000002s : 1: predicate.mutable_eliminate 0.19% : 0.000001s : 1: predicate.opt_reshape 0.10% : 0.000000s : 1: predicate.parallel_virtual_node 3.18% : 0.000014s : 62: predicate.partial_defer_inline 1.39% : 0.000006s : 39: predicate.partial_eliminate 1.07% : 0.000005s : 35: predicate.print_const_string_wrapper 0.45% : 0.000002s : 12: predicate.reduce_all_const_elim 1.52% : 0.000007s : 35: predicate.reduce_eliminate 2.37% : 0.000011s : 75: predicate.redundant_stop_gradient_eliminater 0.34% : 0.000001s : 12: predicate.remove_not_recompute_node 1.09% : 0.000005s : 40: predicate.replace_applicator 1.03% : 0.000005s : 33: predicate.replace_old_param 0.11% : 0.000001s : 1: predicate.reset_defer_inline 1.28% : 0.000006s : 35: predicate.reshape_eliminate 0.22% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.22% : 0.000001s : 1: predicate.row_tensor_eliminate 0.20% : 0.000001s : 2: predicate.same_eliminate 0.25% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.57% : 0.000003s : 12: predicate.shard_identity_eliminate 0.18% : 0.000001s : 2: predicate.special_op_eliminate 0.48% : 0.000002s : 12: predicate.specialize_transform 0.37% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 1.44% : 0.000006s : 33: predicate.stack_unstack_eliminate 0.06% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.26% : 0.000010s : 62: predicate.switch_defer_inline 2.50% : 0.000011s : 64: predicate.switch_layer_defer_inline 7.29% : 0.000033s : 197: predicate.switch_simplify 1.22% : 0.000005s : 35: predicate.tile_eliminate 1.28% : 0.000006s : 35: predicate.transpose_eliminate 1.27% : 0.000006s : 38: predicate.tuple_list_convert_item_index_to_positive 1.47% : 0.000007s : 38: predicate.tuple_list_get_item_const_eliminator 1.32% : 0.000006s : 38: predicate.tuple_list_get_item_depend_reorder 2.84% : 0.000013s : 52: predicate.tuple_list_get_item_eliminator 1.46% : 0.000007s : 38: predicate.tuple_list_get_set_item_eliminator 2.49% : 0.000011s : 50: predicate.tuple_list_set_item_eliminator 1.58% : 0.000007s : 40: predicate.tuple_to_list_eliminator_ 2.29% : 0.000010s : 75: predicate.updatestate_pure_node_eliminater 2.73% : 0.000012s : 87: predicate.updatestate_useless_node_eliminater 0.11% : 0.000001s : 1: predicate.value_based_eliminate 0.46% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.44% : 0.000002s : 12: predicate.virtual_output_eliminate 0.04% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.12% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003380 46 70.36% : 0.002378s : 19: func_graph_cloner_run.FuncGraphClonerGraph 29.64% : 0.001002s : 27: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.515095 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.78% : 0.003992s : 1: add_attr 0.77% : 0.003978s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000044s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000222s : 1: auto_monad 0.00% : 0.000021s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.10% : 0.000532s : 1: bootstrap 0.01% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000016s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.03% : 0.000134s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000018s : 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% : 0.000005s : 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 0.10% : 0.000514s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.16% : 0.000809s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.00% : 0.000012s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 4.65% : 0.023940s : 78: opt.transform.opt_a 0.00% : 0.000014s : 1: opt.transform.opt_after_cconv 0.00% : 0.000018s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000050s : 28: opt.transform.opt_b 0.01% : 0.000029s : 2: opt.transform.opt_trans_graph 0.03% : 0.000133s : 4: opt.transform.symbol_engine_opt 6.06% : 0.031218s : 1: opt_a 0.02% : 0.000092s : 1: opt_after_cconv 0.11% : 0.000559s : 1: opt_after_jit_grad 0.03% : 0.000167s : 1: opt_b 6.62% : 0.034099s : 1: optimize 0.00% : 0.000023s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000023s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000005s : 1: pipeline_split 0.02% : 0.000096s : 1: pre_auto_parallel 0.00% : 0.000013s : 1: py_interpret_to_execute 0.00% : 0.000014s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000018s : 1: remove_dup_value 0.52% : 0.002666s : 1: renormalize.infer 0.15% : 0.000780s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000024s : 1: rewriter_after_opt_a 0.09% : 0.000479s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000004s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.04% : 0.000184s : 1: symbol_engine_optimizer 0.01% : 0.000058s : 1: tuple_transform 79.56% : 0.409791s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:41:53.215.593 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.458172, [21] [bootstrap]: 0.00047687 [type_inference]: 0.443776 [event_method]: 0.00032129 [auto_monad]: 0.00015682 [graph_reusing]: 9.75002e-06 [inline]: 3.43999e-06 [add_attr]: 0.00421109, [1] [add_attr_with_inline]: 0.00419741, [1] [Cycle 1]: 0.00010672, [2] [tag_attr]: 4.87e-05 [meta_addattr_fg_expand]: 9.93998e-06 [parallel-infer-symbol]: 3.97e-06 [pre_auto_parallel]: 6.695e-05 [insert-virtual-dataset]: 2.79999e-06 [parallel-infer-symbol-second]: 1.08001e-06 [dataset_repeat_opt]: 2.01998e-06 [pipeline_split]: 1.74e-06 [optimize]: 0.00820663, [53] [py_interpret_to_execute]: 9.59999e-06 [rewriter_before_opt_a]: 0.00034312 [opt_a]: 0.0053568, [2] [Cycle 1]: 0.00468273, [45] [expand_dump_flag]: 4.85999e-06 [switch_simplify]: 0.00016668 [loop_unroll]: 5.585e-05 [a_1]: 0.00153143 [with_stream_mark]: 2.58e-05 [recompute_prepare]: 9.89999e-06 [updatestate_depend_eliminate]: 4.18999e-06 [updatestate_assign_eliminate]: 3.29001e-06 [updatestate_loads_eliminate]: 3.11001e-06 [parameter_eliminate]: 2.41e-06 [a_2]: 7.226e-05 [accelerated_algorithm]: 6.44001e-06 [shard]: 2.63e-06 [meta_shard_fg_expand]: 3.15002e-06 [shard_inline]: 5.92999e-06 [merge_send_recv]: 8.63001e-06 [auto_parallel]: 9.32001e-06 [parallel]: 2.026e-05 [flash_sp]: 1.14e-05 [merge_comm]: 3.33e-06 [allreduce_fusion]: 3.01001e-06 [matmul_add_comm_reduction]: 9.65002e-06 [allreduce_slice_to_reducescatter]: 9.30013e-07 [virtual_shard_identity]: 9.09e-06 [virtual_dataset]: 5.81003e-06 [get_grad_eliminate_]: 3.425e-05 [virtual_output]: 7.01999e-06 [merge_forward]: 4.35e-06 [cell_reuse_recompute_pass]: 1.42999e-06 [offload_activation]: 1.124e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.446e-05 [merge_recompute_call_nodes]: 1.60999e-06 [before_grad]: 1.082e-05 [set_forward_comm_id_for_comm_node_pass]: 3.94002e-06 [meta_fg_expand]: 3.68999e-06 [flash_sp_send_recv_attached]: 3.48e-06 [receive_attached]: 2.39999e-06 [after_resolve]: 1.035e-05 [a_after_grad]: 8.70999e-06 [renormalize]: 0.00215679 [add_forward_monad_depend]: 1.176e-05 [auto_monad_grad]: 3.11001e-06 [auto_monad_eliminator]: 2.16e-05 [cse]: 3.499e-05 [a_3]: 5.439e-05 [Cycle 2]: 0.00065482, [45] [expand_dump_flag]: 1.91e-06 [switch_simplify]: 9.09998e-06 [loop_unroll]: 5.84e-06 [a_1]: 0.0001059 [with_stream_mark]: 1.827e-05 [recompute_prepare]: 5.86998e-06 [updatestate_depend_eliminate]: 3.39001e-06 [updatestate_assign_eliminate]: 3.36999e-06 [updatestate_loads_eliminate]: 3.23e-06 [parameter_eliminate]: 1.71e-06 [a_2]: 6.002e-05 [accelerated_algorithm]: 6.73e-06 [shard]: 2.29999e-06 [meta_shard_fg_expand]: 2.16998e-06 [shard_inline]: 5.17e-06 [merge_send_recv]: 9.14e-06 [auto_parallel]: 9.39e-06 [parallel]: 8.82e-06 [flash_sp]: 4.35999e-06 [merge_comm]: 3.28998e-06 [allreduce_fusion]: 3.15002e-06 [matmul_add_comm_reduction]: 1.059e-05 [allreduce_slice_to_reducescatter]: 6.80011e-07 [virtual_shard_identity]: 8.87e-06 [virtual_dataset]: 5.17999e-06 [get_grad_eliminate_]: 5.37001e-06 [virtual_output]: 5.44e-06 [merge_forward]: 4.60001e-06 [cell_reuse_recompute_pass]: 2.63998e-06 [offload_activation]: 9.49e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.779e-05 [merge_recompute_call_nodes]: 2.19001e-06 [before_grad]: 9.83998e-06 [set_forward_comm_id_for_comm_node_pass]: 3.8e-06 [meta_fg_expand]: 2.74001e-06 [flash_sp_send_recv_attached]: 2.01e-06 [receive_attached]: 2.06e-06 [after_resolve]: 1.159e-05 [a_after_grad]: 8.70001e-06 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 1.89e-06 [auto_monad_grad]: 1.67999e-06 [auto_monad_eliminator]: 8.80001e-06 [cse]: 1.594e-05 [a_3]: 3.245e-05 [py_interpret_to_execute_after_opt_a]: 1.066e-05 [slice_cell_reuse_recomputed_activation]: 2.39001e-06 [rewriter_after_opt_a]: 2.07e-05 [convert_after_rewriter]: 1.22e-06 [order_py_execute_after_rewriter]: 1.40001e-06 [mutable_eliminate]: 0.00082967 [opt_b]: 0.00020801, [1] [Cycle 1]: 0.00019753, [7] [b_1]: 0.00010889 [b_2]: 7.00998e-06 [updatestate_depend_eliminate]: 1.018e-05 [updatestate_assign_eliminate]: 3.02002e-06 [updatestate_loads_eliminate]: 2.19999e-06 [renormalize]: 9.09989e-07 [cse]: 2.589e-05 [optimize_parallel_all_gather_comm]: 2.46e-05 [overlap_param_gather]: 2.14999e-06 [cconv]: 4.074e-05 [loop_unroll]: 0.00055255 [opt_after_cconv]: 0.00010858, [1] [Cycle 1]: 9.992e-05, [7] [c_1]: 2.519e-05 [parameter_eliminate]: 5.65001e-06 [updatestate_depend_eliminate]: 7.92e-06 [updatestate_assign_eliminate]: 2.40002e-06 [updatestate_loads_eliminate]: 2.26e-06 [cse]: 2.122e-05 [renormalize]: 7.89994e-07 [remove_dup_value]: 1.469e-05 [tuple_transform]: 7.675e-05, [1] [Cycle 1]: 7.182e-05, [4] [d_1]: 4.287e-05 [none_parameter_eliminate]: 1.77001e-06 [renormalize]: 1.69995e-07 [switch_simplify]: 6.01998e-06 [partial_unused_args_eliminate]: 2.08998e-06 [add_recomputation]: 5.589e-05 [cse_after_recomputation]: 2.197e-05, [1] [Cycle 1]: 1.644e-05, [1] [cse]: 1.104e-05 [environ_conv]: 6.06998e-06 [swap_dp_allreduce_reducescatter]: 5.39e-06 [bias_add_comm_swap]: 3.42002e-06 [label_micro_interleaved_index]: 6.26e-06 [label_fine_grained_interleaved_index]: 3.07002e-06 [merge_cast_opt]: 1.35001e-06 [slice_recompute_activation]: 2.64999e-06 [micro_interleaved_order_control]: 2.91999e-06 [assign_add_opt]: 1.37e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.24e-06 [full_micro_interleaved_order_control]: 2.59999e-06 [reorder_send_recv_between_fp_bp]: 2.68e-06 [comm_op_add_attrs]: 1.14003e-06 [add_comm_op_reuse_tag]: 9.99979e-07 [interleave_split_concat_branches]: 1.25001e-06 [interleave_parallel_branches]: 1.01002e-06 [overlap_opt_shard_in_pipeline]: 1.27e-06 [overlap_opt_shard_grad_in_pipeline]: 1.81998e-06 [control_data_broadcast_order]: 1.552e-05 [grouped_pairwise_exchange_alltoall]: 1.78002e-06 [offloading_packed_experts]: 3.86001e-06 [overlap_recompute_and_grad_model_parallel]: 4.99e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.45999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.23002e-06 [overlap_recompute_comm]: 2.63e-06 [overlap_grad_ring_attention]: 4.32e-06 [overlap_grad_flash_sp]: 2.385e-05 [begin_end_overlap_inline]: 4.89992e-07 [split_matmul_comm_elemetwise]: 2.27999e-06 [split_layernorm_comm]: 2.03002e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 7.955e-05, [1] [Cycle 1]: 7.487e-05, [6] [build]: 3.97002e-06 [elim_shapecalc]: 1.04e-05 [elim_not_effective]: 1.357e-05 [opt_reshape]: 6.63998e-06 [fold_const_symbol]: 9.20999e-06 [renormalize]: 1.79978e-07 [detach_backward]: 2.71999e-06 [pipeline_parallel_scheduler]: 1.51998e-06 [auto_monad_reorder]: 7.054e-05 [get_jit_bprop_graph]: 2.28002e-06 [rewriter_after_jit_bprop_graph]: 7.85998e-06 [opt_after_jit_grad]: 0.00058841 [validate]: 4.848e-05 Sums bootstrap : 0.000477s : 0.11% type_inference : 0.443776s : 98.01% event_method : 0.000321s : 0.07% auto_monad : 0.000157s : 0.03% graph_reusing : 0.000010s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000049s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000010s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000067s : 0.01% 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 : 0.000010s : 0.00% optimize.rewriter_before_opt_a : 0.000343s : 0.08% optimize.opt_a.expand_dump_flag : 0.000007s : 0.00% optimize.opt_a.switch_simplify : 0.000176s : 0.04% optimize.opt_a.loop_unroll : 0.000062s : 0.01% optimize.opt_a.a_1 : 0.001637s : 0.36% optimize.opt_a.with_stream_mark : 0.000044s : 0.01% optimize.opt_a.recompute_prepare : 0.000016s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000132s : 0.03% optimize.opt_a.accelerated_algorithm : 0.000013s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.00% optimize.opt_a.shard_inline : 0.000011s : 0.00% optimize.opt_a.merge_send_recv : 0.000018s : 0.00% optimize.opt_a.auto_parallel : 0.000019s : 0.00% optimize.opt_a.parallel : 0.000029s : 0.01% optimize.opt_a.flash_sp : 0.000016s : 0.00% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000006s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000018s : 0.00% optimize.opt_a.virtual_dataset : 0.000011s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000040s : 0.01% optimize.opt_a.virtual_output : 0.000012s : 0.00% optimize.opt_a.merge_forward : 0.000009s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000021s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000032s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000021s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.00% optimize.opt_a.meta_fg_expand : 0.000006s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000022s : 0.00% optimize.opt_a.a_after_grad : 0.000017s : 0.00% optimize.opt_a.renormalize : 0.002157s : 0.48% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000030s : 0.01% optimize.opt_a.cse : 0.000051s : 0.01% optimize.opt_a.a_3 : 0.000087s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000830s : 0.18% optimize.opt_b.b_1 : 0.000109s : 0.02% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000026s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000025s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000041s : 0.01% optimize.loop_unroll : 0.000553s : 0.12% optimize.opt_after_cconv.c_1 : 0.000025s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000021s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000043s : 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 : 0.000006s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000056s : 0.01% optimize.cse_after_recomputation.cse : 0.000011s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000003s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 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 : 0.000016s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000024s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000071s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000588s : 0.13% validate : 0.000048s : 0.01% Time group info: ------[substitution.] 0.000395 39 3.56% : 0.000014s : 1: substitution.cast_eliminate 0.49% : 0.000002s : 2: substitution.elim_not_effective 0.40% : 0.000002s : 2: substitution.fold_const_symbol 1.48% : 0.000006s : 3: substitution.graph_param_transform 83.27% : 0.000329s : 15: substitution.inline 1.16% : 0.000005s : 4: substitution.j_node_and_user_rematch 1.38% : 0.000005s : 4: substitution.remove_not_recompute_node 1.40% : 0.000006s : 2: substitution.replace_old_param 3.75% : 0.000015s : 4: substitution.switch_simplify 3.11% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.443664 2 99.30% : 0.440551s : 1: type_inference.infer 0.70% : 0.003113s : 1: type_inference.specialize ------[replace.] 0.000197 22 3.14% : 0.000006s : 1: replace.cast_eliminate 54.43% : 0.000107s : 15: replace.inline 30.16% : 0.000059s : 4: replace.switch_simplify 12.27% : 0.000024s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000357 22 3.65% : 0.000013s : 1: match.cast_eliminate 89.90% : 0.000321s : 15: match.inline 3.38% : 0.000012s : 4: match.switch_simplify 3.08% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000301 1716 1.16% : 0.000003s : 22: predicate.accumulaten_eliminater 0.74% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.31% : 0.000001s : 6: predicate.addn_check_dump 1.10% : 0.000003s : 22: predicate.addn_zero_filter 0.95% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.23% : 0.000007s : 28: predicate.arithmetic_simplify 1.50% : 0.000005s : 23: predicate.cast_eliminate 0.36% : 0.000001s : 6: predicate.check_bprop_eliminate 0.41% : 0.000001s : 6: predicate.compare_switch_simplify 0.09% : 0.000000s : 3: predicate.const_output_eliminate 0.41% : 0.000001s : 6: predicate.depend_value_elim 1.22% : 0.000004s : 23: predicate.dict_get_item_const_eliminator 1.31% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.13% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.75% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.21% : 0.000001s : 3: predicate.elim_not_effective 0.26% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.38% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.18% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.36% : 0.000004s : 26: predicate.environ_get_depend_swap 1.57% : 0.000005s : 32: predicate.environ_get_eliminate 1.13% : 0.000003s : 26: predicate.environ_get_set_eliminate 1.90% : 0.000006s : 40: predicate.exchange_switch_depend_value 3.08% : 0.000009s : 40: predicate.float_depend_g_call 0.42% : 0.000001s : 6: predicate.float_environ_get_switch 0.43% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.09% : 0.000000s : 3: predicate.fold_const_symbol 0.66% : 0.000002s : 6: predicate.get_grad_eliminate 0.12% : 0.000000s : 3: predicate.graph_param_transform 0.30% : 0.000001s : 6: predicate.incorporate_call 0.26% : 0.000001s : 6: predicate.incorporate_call_switch 5.90% : 0.000018s : 80: predicate.inline 0.45% : 0.000001s : 6: predicate.inline_without_move 0.18% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.70% : 0.000002s : 6: predicate.less_batch_normalization 1.56% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.60% : 0.000008s : 53: predicate.load_eliminater 0.75% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.23% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.70% : 0.000005s : 29: predicate.make_slice_get_slice_eliminator 0.31% : 0.000001s : 6: predicate.merge_addn 0.39% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.32% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.04% : 0.000003s : 22: predicate.minmaximum_grad 1.16% : 0.000003s : 3: predicate.mutable_eliminate 0.34% : 0.000001s : 3: predicate.opt_reshape 0.27% : 0.000001s : 3: predicate.parallel_virtual_node 2.72% : 0.000008s : 40: predicate.partial_defer_inline 1.39% : 0.000004s : 28: predicate.partial_eliminate 1.01% : 0.000003s : 22: predicate.print_const_string_wrapper 0.35% : 0.000001s : 6: predicate.reduce_all_const_elim 1.48% : 0.000004s : 22: predicate.reduce_eliminate 2.86% : 0.000009s : 53: predicate.redundant_stop_gradient_eliminater 0.36% : 0.000001s : 6: predicate.remove_not_recompute_node 1.22% : 0.000004s : 31: predicate.replace_applicator 0.32% : 0.000001s : 6: predicate.replace_old_param 0.21% : 0.000001s : 3: predicate.reset_defer_inline 1.11% : 0.000003s : 22: predicate.reshape_eliminate 0.36% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.34% : 0.000001s : 3: predicate.row_tensor_eliminate 0.46% : 0.000001s : 6: predicate.same_eliminate 0.21% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.74% : 0.000002s : 6: predicate.shard_identity_eliminate 0.42% : 0.000001s : 6: predicate.special_op_eliminate 0.48% : 0.000001s : 6: predicate.specialize_transform 0.67% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.42% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.17% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.13% : 0.000006s : 40: predicate.switch_defer_inline 2.45% : 0.000007s : 46: predicate.switch_layer_defer_inline 6.81% : 0.000021s : 126: predicate.switch_simplify 1.16% : 0.000003s : 22: predicate.tile_eliminate 1.01% : 0.000003s : 22: predicate.transpose_eliminate 1.72% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.58% : 0.000005s : 29: predicate.tuple_list_get_item_const_eliminator 1.50% : 0.000005s : 29: predicate.tuple_list_get_item_depend_reorder 3.11% : 0.000009s : 37: predicate.tuple_list_get_item_eliminator 1.41% : 0.000004s : 29: predicate.tuple_list_get_set_item_eliminator 2.37% : 0.000007s : 35: predicate.tuple_list_set_item_eliminator 1.45% : 0.000004s : 31: predicate.tuple_to_list_eliminator_ 2.60% : 0.000008s : 53: predicate.updatestate_pure_node_eliminater 2.95% : 0.000009s : 59: predicate.updatestate_useless_node_eliminater 0.29% : 0.000001s : 3: predicate.value_based_eliminate 0.45% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.46% : 0.000001s : 6: predicate.virtual_output_eliminate 0.14% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.23% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.001877 27 50.55% : 0.000949s : 10: func_graph_cloner_run.FuncGraphClonerGraph 49.45% : 0.000928s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.475052 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.89% : 0.004217s : 1: add_attr 0.88% : 0.004202s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000061s : 1: add_recomputation 0.02% : 0.000076s : 1: assign_add_opt 0.04% : 0.000170s : 1: auto_monad 0.02% : 0.000077s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.11% : 0.000508s : 1: bootstrap 0.01% : 0.000044s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000020s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000025s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.07% : 0.000339s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000015s : 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% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000007s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.12% : 0.000565s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.18% : 0.000852s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000025s : 1: opt.transform.mutable_eliminate 0.46% : 0.002206s : 78: opt.transform.opt_a 0.01% : 0.000024s : 1: opt.transform.opt_after_cconv 0.01% : 0.000026s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000085s : 28: opt.transform.opt_b 0.01% : 0.000046s : 2: opt.transform.opt_trans_graph 0.01% : 0.000036s : 4: opt.transform.symbol_engine_opt 1.13% : 0.005362s : 1: opt_a 0.02% : 0.000112s : 1: opt_after_cconv 0.13% : 0.000601s : 1: opt_after_jit_grad 0.04% : 0.000212s : 1: opt_b 1.73% : 0.008212s : 1: optimize 0.01% : 0.000028s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000028s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.02% : 0.000071s : 1: pre_auto_parallel 0.00% : 0.000013s : 1: py_interpret_to_execute 0.00% : 0.000014s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000018s : 1: remove_dup_value 0.24% : 0.001147s : 1: renormalize.infer 0.21% : 0.000996s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000025s : 1: rewriter_after_opt_a 0.07% : 0.000353s : 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% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000082s : 1: symbol_engine_optimizer 0.02% : 0.000079s : 1: tuple_transform 93.42% : 0.443814s : 1: type_inference TotalTime = 1.33373, [21] [bootstrap]: 0.00048258 [type_inference]: 1.12617 [event_method]: 8.025e-05 [auto_monad]: 0.00021691 [graph_reusing]: 1.338e-05 [inline]: 2.95002e-06 [add_attr]: 0.00406372, [1] [add_attr_with_inline]: 0.00404937, [1] [Cycle 1]: 0.00012887, [2] [tag_attr]: 7.04e-05 [meta_addattr_fg_expand]: 1.456e-05 [parallel-infer-symbol]: 3.61001e-06 [pre_auto_parallel]: 9.323e-05 [insert-virtual-dataset]: 2.34999e-06 [parallel-infer-symbol-second]: 2.01e-06 [dataset_repeat_opt]: 2.25002e-06 [pipeline_split]: 1.54e-06 [optimize]: 0.201762, [53] [py_interpret_to_execute]: 6.33e-06 [rewriter_before_opt_a]: 0.0004595 [opt_a]: 0.198995, [2] [Cycle 1]: 0.198077, [45] [expand_dump_flag]: 6.88e-06 [switch_simplify]: 0.00026794 [loop_unroll]: 9.062e-05 [a_1]: 0.00213315 [with_stream_mark]: 4.193e-05 [recompute_prepare]: 3.202e-05 [updatestate_depend_eliminate]: 9.48002e-06 [updatestate_assign_eliminate]: 6.91001e-06 [updatestate_loads_eliminate]: 6.50002e-06 [parameter_eliminate]: 3.86001e-06 [a_2]: 0.00020613 [accelerated_algorithm]: 1.796e-05 [shard]: 2.99001e-06 [meta_shard_fg_expand]: 8.12e-06 [shard_inline]: 1.391e-05 [merge_send_recv]: 1.987e-05 [auto_parallel]: 1.564e-05 [parallel]: 2.406e-05 [flash_sp]: 1.474e-05 [merge_comm]: 8.40999e-06 [allreduce_fusion]: 7.83999e-06 [matmul_add_comm_reduction]: 3.487e-05 [allreduce_slice_to_reducescatter]: 9.00007e-07 [virtual_shard_identity]: 1.994e-05 [virtual_dataset]: 1.367e-05 [get_grad_eliminate_]: 1.396e-05 [virtual_output]: 1.357e-05 [merge_forward]: 1.025e-05 [cell_reuse_recompute_pass]: 2.27999e-06 [offload_activation]: 1.92e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.296e-05 [merge_recompute_call_nodes]: 1.57001e-06 [before_grad]: 2.676e-05 [set_forward_comm_id_for_comm_node_pass]: 9.19e-06 [meta_fg_expand]: 0.00220228 [flash_sp_send_recv_attached]: 5.11002e-06 [receive_attached]: 2.67001e-06 [after_resolve]: 7.683e-05 [a_after_grad]: 8.938e-05 [renormalize]: 0.192009 [add_forward_monad_depend]: 8.90001e-06 [auto_monad_grad]: 2.73e-06 [auto_monad_eliminator]: 1.957e-05 [cse]: 3.182e-05 [a_3]: 4.965e-05 [Cycle 2]: 0.0008999, [45] [expand_dump_flag]: 3.16001e-06 [switch_simplify]: 7.97e-06 [loop_unroll]: 4.25e-06 [a_1]: 5.574e-05 [with_stream_mark]: 1.629e-05 [recompute_prepare]: 4.74002e-06 [updatestate_depend_eliminate]: 3.29001e-06 [updatestate_assign_eliminate]: 2.71e-06 [updatestate_loads_eliminate]: 2.94001e-06 [parameter_eliminate]: 1.85001e-06 [a_2]: 3.872e-05 [accelerated_algorithm]: 4.62e-06 [shard]: 1.86e-06 [meta_shard_fg_expand]: 2.17001e-06 [shard_inline]: 3.71999e-06 [merge_send_recv]: 7.83001e-06 [auto_parallel]: 8.48999e-06 [parallel]: 1.18e-05 [flash_sp]: 3.80998e-06 [merge_comm]: 2.29001e-06 [allreduce_fusion]: 2.51e-06 [matmul_add_comm_reduction]: 7.94997e-06 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 5.01997e-06 [virtual_dataset]: 4.26001e-06 [get_grad_eliminate_]: 3.68999e-06 [virtual_output]: 3.97e-06 [merge_forward]: 3.31001e-06 [cell_reuse_recompute_pass]: 3.06001e-06 [offload_activation]: 8.12e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.455e-05 [merge_recompute_call_nodes]: 1.52001e-06 [before_grad]: 8.15e-06 [set_forward_comm_id_for_comm_node_pass]: 2.41e-06 [meta_fg_expand]: 0.00032208 [flash_sp_send_recv_attached]: 2.07001e-06 [receive_attached]: 2.41e-06 [after_resolve]: 1.024e-05 [a_after_grad]: 5.43002e-06 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 2.55002e-06 [auto_monad_grad]: 2.27001e-06 [auto_monad_eliminator]: 9.72001e-06 [cse]: 1.728e-05 [a_3]: 6.04e-05 [py_interpret_to_execute_after_opt_a]: 1.077e-05 [slice_cell_reuse_recomputed_activation]: 2.47001e-06 [rewriter_after_opt_a]: 1.935e-05 [convert_after_rewriter]: 1.20999e-06 [order_py_execute_after_rewriter]: 1.12999e-06 [mutable_eliminate]: 0.00080166 [opt_b]: 0.00015918, [1] [Cycle 1]: 0.00014962, [7] [b_1]: 7.525e-05 [b_2]: 4.97e-06 [updatestate_depend_eliminate]: 7.48e-06 [updatestate_assign_eliminate]: 2.26998e-06 [updatestate_loads_eliminate]: 1.94999e-06 [renormalize]: 1.02e-06 [cse]: 2.079e-05 [optimize_parallel_all_gather_comm]: 1.906e-05 [overlap_param_gather]: 1.99999e-06 [cconv]: 3.961e-05 [loop_unroll]: 0.00050499 [opt_after_cconv]: 8.527e-05, [1] [Cycle 1]: 7.844e-05, [7] [c_1]: 1.575e-05 [parameter_eliminate]: 4.95001e-06 [updatestate_depend_eliminate]: 4.68999e-06 [updatestate_assign_eliminate]: 1.71998e-06 [updatestate_loads_eliminate]: 1.37999e-06 [cse]: 1.644e-05 [renormalize]: 4.80009e-07 [remove_dup_value]: 1.446e-05 [tuple_transform]: 5.486e-05, [1] [Cycle 1]: 4.965e-05, [4] [d_1]: 2.629e-05 [none_parameter_eliminate]: 1.57001e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 4.15e-06 [partial_unused_args_eliminate]: 1.89e-06 [add_recomputation]: 3.825e-05 [cse_after_recomputation]: 1.951e-05, [1] [Cycle 1]: 1.52e-05, [1] [cse]: 9.65002e-06 [environ_conv]: 5.59e-06 [swap_dp_allreduce_reducescatter]: 3.66999e-06 [bias_add_comm_swap]: 3.47997e-06 [label_micro_interleaved_index]: 5.19e-06 [label_fine_grained_interleaved_index]: 3.09999e-06 [merge_cast_opt]: 1.30001e-06 [slice_recompute_activation]: 2.25002e-06 [micro_interleaved_order_control]: 2.49999e-06 [assign_add_opt]: 1.29998e-06 [ForceFp32Comm]: 8.80013e-07 [remove_cast_before_assign_add]: 1.00001e-06 [full_micro_interleaved_order_control]: 2.68e-06 [reorder_send_recv_between_fp_bp]: 2.69001e-06 [comm_op_add_attrs]: 1.36002e-06 [add_comm_op_reuse_tag]: 1.18001e-06 [interleave_split_concat_branches]: 1.19e-06 [interleave_parallel_branches]: 1.03001e-06 [overlap_opt_shard_in_pipeline]: 1.17999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.16e-06 [control_data_broadcast_order]: 1.129e-05 [grouped_pairwise_exchange_alltoall]: 1.30999e-06 [offloading_packed_experts]: 3.29001e-06 [overlap_recompute_and_grad_model_parallel]: 4.45999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.13001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.22999e-06 [overlap_recompute_comm]: 2.59999e-06 [overlap_grad_ring_attention]: 3.48999e-06 [overlap_grad_flash_sp]: 2.04e-05 [begin_end_overlap_inline]: 5.10016e-07 [split_matmul_comm_elemetwise]: 2.32999e-06 [split_layernorm_comm]: 1.79e-06 [handle_group_info]: 1.58002e-06 [symbol_engine_optimizer]: 7.015e-05, [1] [Cycle 1]: 6.491e-05, [6] [build]: 3.9e-06 [elim_shapecalc]: 7.80998e-06 [elim_not_effective]: 1.123e-05 [opt_reshape]: 4.53999e-06 [fold_const_symbol]: 7.25998e-06 [renormalize]: 1.50001e-07 [detach_backward]: 2.41e-06 [pipeline_parallel_scheduler]: 1.56002e-06 [auto_monad_reorder]: 1.627e-05 [get_jit_bprop_graph]: 1.94999e-06 [rewriter_after_jit_bprop_graph]: 6.48e-06 [opt_after_jit_grad]: 0.00053777 [validate]: 4.151e-05 Sums bootstrap : 0.000483s : 0.04% type_inference : 1.126171s : 84.78% event_method : 0.000080s : 0.01% auto_monad : 0.000217s : 0.02% graph_reusing : 0.000013s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000070s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000015s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000093s : 0.01% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000006s : 0.00% optimize.rewriter_before_opt_a : 0.000460s : 0.03% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000276s : 0.02% optimize.opt_a.loop_unroll : 0.000095s : 0.01% optimize.opt_a.a_1 : 0.002189s : 0.16% optimize.opt_a.with_stream_mark : 0.000058s : 0.00% optimize.opt_a.recompute_prepare : 0.000037s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000245s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000023s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000018s : 0.00% optimize.opt_a.merge_send_recv : 0.000028s : 0.00% optimize.opt_a.auto_parallel : 0.000024s : 0.00% optimize.opt_a.parallel : 0.000036s : 0.00% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000011s : 0.00% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000043s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000025s : 0.00% optimize.opt_a.virtual_dataset : 0.000018s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000018s : 0.00% optimize.opt_a.virtual_output : 0.000018s : 0.00% optimize.opt_a.merge_forward : 0.000014s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000027s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000048s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000035s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.00% optimize.opt_a.meta_fg_expand : 0.002524s : 0.19% 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 : 0.000087s : 0.01% optimize.opt_a.a_after_grad : 0.000095s : 0.01% optimize.opt_a.renormalize : 0.192009s : 14.45% optimize.opt_a.add_forward_monad_depend : 0.000011s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000029s : 0.00% optimize.opt_a.cse : 0.000049s : 0.00% optimize.opt_a.a_3 : 0.000110s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000019s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000802s : 0.06% optimize.opt_b.b_1 : 0.000075s : 0.01% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000021s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000019s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000040s : 0.00% optimize.loop_unroll : 0.000505s : 0.04% optimize.opt_after_cconv.c_1 : 0.000016s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.00% optimize.tuple_transform.d_1 : 0.000026s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000038s : 0.00% optimize.cse_after_recomputation.cse : 0.000010s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000011s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000020s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000008s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 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 : 0.000016s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000538s : 0.04% validate : 0.000042s : 0.00% Time group info: ------[substitution.] 0.000703 102 1.82% : 0.000013s : 1: substitution.cast_eliminate 0.26% : 0.000002s : 1: substitution.elim_not_effective 0.93% : 0.000007s : 9: substitution.float_depend_g_call 0.66% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.16% : 0.000001s : 1: substitution.fold_const_symbol 0.63% : 0.000004s : 1: substitution.graph_param_transform 0.44% : 0.000003s : 2: substitution.incorporate_call 0.30% : 0.000002s : 2: substitution.incorporate_call_switch 75.83% : 0.000533s : 24: substitution.inline 3.55% : 0.000025s : 2: substitution.inline_without_move 1.26% : 0.000009s : 9: substitution.j_node_and_user_rematch 0.54% : 0.000004s : 2: substitution.minmaximum_grad 0.90% : 0.000006s : 9: substitution.partial_eliminate 1.28% : 0.000009s : 9: substitution.remove_not_recompute_node 0.33% : 0.000002s : 1: substitution.replace_applicator 0.82% : 0.000006s : 5: substitution.replace_old_param 0.68% : 0.000005s : 1: substitution.set_cell_output_no_recompute 2.72% : 0.000019s : 7: substitution.switch_simplify 1.19% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.47% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.79% : 0.000006s : 2: substitution.tuple_list_get_item_depend_reorder 3.73% : 0.000026s : 6: substitution.tuple_list_get_item_eliminator 0.72% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 1.125236 2 99.59% : 1.120660s : 1: type_inference.infer 0.41% : 0.004576s : 1: type_inference.specialize ------[replace.] 0.000288 34 2.45% : 0.000007s : 1: replace.cast_eliminate 61.12% : 0.000176s : 24: replace.inline 31.17% : 0.000090s : 7: replace.switch_simplify 5.27% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000560 34 2.11% : 0.000012s : 1: match.cast_eliminate 92.93% : 0.000520s : 24: match.inline 2.71% : 0.000015s : 7: match.switch_simplify 2.25% : 0.000013s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000423 2554 1.26% : 0.000005s : 35: predicate.accumulaten_eliminater 0.53% : 0.000002s : 1: predicate.ad_related_special_op_eliminate 0.45% : 0.000002s : 12: predicate.addn_check_dump 1.22% : 0.000005s : 35: predicate.addn_zero_filter 1.09% : 0.000005s : 35: predicate.adjust_all_reduce_mul_add 2.30% : 0.000010s : 47: predicate.arithmetic_simplify 1.51% : 0.000006s : 36: predicate.cast_eliminate 0.14% : 0.000001s : 2: predicate.check_bprop_eliminate 0.44% : 0.000002s : 12: predicate.compare_switch_simplify 0.03% : 0.000000s : 1: predicate.const_output_eliminate 0.52% : 0.000002s : 12: predicate.depend_value_elim 1.19% : 0.000005s : 36: predicate.dict_get_item_const_eliminator 1.29% : 0.000005s : 36: predicate.dict_get_item_eliminator 1.17% : 0.000005s : 36: predicate.dict_set_item_eliminator 0.42% : 0.000002s : 2: predicate.dumpgradient_eliminate 0.06% : 0.000000s : 1: predicate.elim_not_effective 0.12% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.23% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.13% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.14% : 0.000005s : 37: predicate.environ_get_depend_swap 1.67% : 0.000007s : 49: predicate.environ_get_eliminate 1.16% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.10% : 0.000009s : 62: predicate.exchange_switch_depend_value 2.92% : 0.000012s : 62: predicate.float_depend_g_call 0.44% : 0.000002s : 12: predicate.float_environ_get_switch 0.50% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.51% : 0.000002s : 12: predicate.get_grad_eliminate 0.03% : 0.000000s : 1: predicate.graph_param_transform 0.40% : 0.000002s : 12: predicate.incorporate_call 0.40% : 0.000002s : 12: predicate.incorporate_call_switch 6.13% : 0.000026s : 123: predicate.inline 1.74% : 0.000007s : 33: predicate.inline_without_move 0.21% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.81% : 0.000003s : 12: predicate.less_batch_normalization 1.62% : 0.000007s : 40: predicate.list_to_tuple_eliminator_ 2.42% : 0.000010s : 75: predicate.load_eliminater 0.41% : 0.000002s : 1: predicate.loop_unroll_after_grad 3.77% : 0.000016s : 108: predicate.loop_unroll_before_grad 1.41% : 0.000006s : 38: predicate.make_slice_get_slice_eliminator 0.48% : 0.000002s : 12: predicate.merge_addn 0.13% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.10% : 0.000000s : 2: predicate.mini_step_allgather_replace 1.17% : 0.000005s : 35: predicate.minmaximum_grad 0.95% : 0.000004s : 1: predicate.mutable_eliminate 0.07% : 0.000000s : 1: predicate.opt_reshape 0.17% : 0.000001s : 1: predicate.parallel_virtual_node 3.11% : 0.000013s : 62: predicate.partial_defer_inline 1.41% : 0.000006s : 39: predicate.partial_eliminate 1.24% : 0.000005s : 35: predicate.print_const_string_wrapper 0.45% : 0.000002s : 12: predicate.reduce_all_const_elim 1.43% : 0.000006s : 35: predicate.reduce_eliminate 2.46% : 0.000010s : 75: predicate.redundant_stop_gradient_eliminater 0.39% : 0.000002s : 12: predicate.remove_not_recompute_node 1.26% : 0.000005s : 40: predicate.replace_applicator 0.88% : 0.000004s : 33: predicate.replace_old_param 0.12% : 0.000000s : 1: predicate.reset_defer_inline 1.19% : 0.000005s : 35: predicate.reshape_eliminate 0.18% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.11% : 0.000000s : 1: predicate.row_tensor_eliminate 0.21% : 0.000001s : 2: predicate.same_eliminate 0.26% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.67% : 0.000003s : 12: predicate.shard_identity_eliminate 0.16% : 0.000001s : 2: predicate.special_op_eliminate 0.52% : 0.000002s : 12: predicate.specialize_transform 0.39% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 1.59% : 0.000007s : 33: predicate.stack_unstack_eliminate 0.09% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.33% : 0.000010s : 62: predicate.switch_defer_inline 2.53% : 0.000011s : 64: predicate.switch_layer_defer_inline 7.38% : 0.000031s : 197: predicate.switch_simplify 1.14% : 0.000005s : 35: predicate.tile_eliminate 1.31% : 0.000006s : 35: predicate.transpose_eliminate 1.27% : 0.000005s : 38: predicate.tuple_list_convert_item_index_to_positive 1.51% : 0.000006s : 38: predicate.tuple_list_get_item_const_eliminator 1.29% : 0.000005s : 38: predicate.tuple_list_get_item_depend_reorder 2.68% : 0.000011s : 52: predicate.tuple_list_get_item_eliminator 1.33% : 0.000006s : 38: predicate.tuple_list_get_set_item_eliminator 2.09% : 0.000009s : 50: predicate.tuple_list_set_item_eliminator 1.57% : 0.000007s : 40: predicate.tuple_to_list_eliminator_ 2.34% : 0.000010s : 75: predicate.updatestate_pure_node_eliminater 2.75% : 0.000012s : 87: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 1: predicate.value_based_eliminate 0.52% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.59% : 0.000003s : 12: predicate.virtual_output_eliminate 0.05% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.12% : 0.000000s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003231 46 70.73% : 0.002285s : 19: func_graph_cloner_run.FuncGraphClonerGraph 29.27% : 0.000946s : 27: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.734789 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.23% : 0.004070s : 1: add_attr 0.23% : 0.004056s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000042s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000225s : 1: auto_monad 0.00% : 0.000021s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.03% : 0.000520s : 1: bootstrap 0.00% : 0.000043s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000015s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.01% : 0.000092s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000017s : 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% : 0.000005s : 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% : 0.000009s : 1: label_micro_interleaved_index 0.03% : 0.000516s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.05% : 0.000816s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.00% : 0.000011s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000017s : 1: opt.transform.mutable_eliminate 0.19% : 0.003220s : 78: opt.transform.opt_a 0.00% : 0.000014s : 1: opt.transform.opt_after_cconv 0.00% : 0.000017s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000048s : 28: opt.transform.opt_b 0.00% : 0.000028s : 2: opt.transform.opt_trans_graph 0.00% : 0.000026s : 4: opt.transform.symbol_engine_opt 11.47% : 0.198999s : 1: opt_a 0.01% : 0.000089s : 1: opt_after_cconv 0.03% : 0.000551s : 1: opt_after_jit_grad 0.01% : 0.000163s : 1: opt_b 11.63% : 0.201769s : 1: optimize 0.00% : 0.000023s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000025s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000122s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.01% : 0.000098s : 1: pre_auto_parallel 0.00% : 0.000010s : 1: py_interpret_to_execute 0.00% : 0.000014s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.00% : 0.000019s : 1: remove_dup_value 11.02% : 0.191232s : 1: renormalize.infer 0.04% : 0.000762s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000023s : 1: rewriter_after_opt_a 0.03% : 0.000468s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000073s : 1: symbol_engine_optimizer 0.00% : 0.000058s : 1: tuple_transform 64.92% : 1.126223s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:41:59.286.187 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 1.14709, [21] [bootstrap]: 0.00047267 [type_inference]: 0.96819 [event_method]: 8.274e-05 [auto_monad]: 0.00019419 [graph_reusing]: 1.346e-05 [inline]: 4.12e-06 [add_attr]: 0.00590305, [1] [add_attr_with_inline]: 0.00588848, [1] [Cycle 1]: 0.0001348, [2] [tag_attr]: 6.867e-05 [meta_addattr_fg_expand]: 1.388e-05 [parallel-infer-symbol]: 3.92002e-06 [pre_auto_parallel]: 9.264e-05 [insert-virtual-dataset]: 3.29001e-06 [parallel-infer-symbol-second]: 1.07e-06 [dataset_repeat_opt]: 2.29999e-06 [pipeline_split]: 2.22999e-06 [optimize]: 0.0103805, [53] [py_interpret_to_execute]: 9.51003e-06 [rewriter_before_opt_a]: 0.00044348 [opt_a]: 0.0074876, [2] [Cycle 1]: 0.00663037, [45] [expand_dump_flag]: 7.93999e-06 [switch_simplify]: 0.00025793 [loop_unroll]: 0.0001244 [a_1]: 0.00210443 [with_stream_mark]: 0.00015534 [recompute_prepare]: 2.262e-05 [updatestate_depend_eliminate]: 8.96998e-06 [updatestate_assign_eliminate]: 5.36998e-06 [updatestate_loads_eliminate]: 4.22998e-06 [parameter_eliminate]: 2.63e-06 [a_2]: 0.00010824 [accelerated_algorithm]: 9.37001e-06 [shard]: 2.03002e-06 [meta_shard_fg_expand]: 5.59998e-06 [shard_inline]: 8.79e-06 [merge_send_recv]: 1.106e-05 [auto_parallel]: 1.189e-05 [parallel]: 2.116e-05 [flash_sp]: 1.316e-05 [merge_comm]: 4.96002e-06 [allreduce_fusion]: 4.97999e-06 [matmul_add_comm_reduction]: 1.269e-05 [allreduce_slice_to_reducescatter]: 1.19e-06 [virtual_shard_identity]: 1.118e-05 [virtual_dataset]: 8.85999e-06 [get_grad_eliminate_]: 8.10999e-06 [virtual_output]: 8.27e-06 [merge_forward]: 6.78998e-06 [cell_reuse_recompute_pass]: 2.46998e-06 [offload_activation]: 1.312e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.327e-05 [merge_recompute_call_nodes]: 1.39e-06 [before_grad]: 1.55e-05 [set_forward_comm_id_for_comm_node_pass]: 6.18002e-06 [meta_fg_expand]: 5.32001e-06 [flash_sp_send_recv_attached]: 2.63e-06 [receive_attached]: 2.60002e-06 [after_resolve]: 1.273e-05 [a_after_grad]: 1.27e-05 [renormalize]: 0.00300288 [add_forward_monad_depend]: 1.149e-05 [auto_monad_grad]: 2.93e-06 [auto_monad_eliminator]: 2.889e-05 [cse]: 5.321e-05 [a_3]: 0.00010392 [Cycle 2]: 0.0008399, [45] [expand_dump_flag]: 2.68998e-06 [switch_simplify]: 1.053e-05 [loop_unroll]: 7.81001e-06 [a_1]: 0.00019373 [with_stream_mark]: 2.082e-05 [recompute_prepare]: 8.45001e-06 [updatestate_depend_eliminate]: 5.37999e-06 [updatestate_assign_eliminate]: 4.1e-06 [updatestate_loads_eliminate]: 4.18999e-06 [parameter_eliminate]: 2.88e-06 [a_2]: 9.632e-05 [accelerated_algorithm]: 8.11002e-06 [shard]: 2.19001e-06 [meta_shard_fg_expand]: 2.51e-06 [shard_inline]: 7.35998e-06 [merge_send_recv]: 9.82001e-06 [auto_parallel]: 1.098e-05 [parallel]: 8.64e-06 [flash_sp]: 4.14002e-06 [merge_comm]: 4.53001e-06 [allreduce_fusion]: 4.54002e-06 [matmul_add_comm_reduction]: 1.104e-05 [allreduce_slice_to_reducescatter]: 7.30011e-07 [virtual_shard_identity]: 9.89001e-06 [virtual_dataset]: 7.74002e-06 [get_grad_eliminate_]: 7.1e-06 [virtual_output]: 7.33e-06 [merge_forward]: 5.59e-06 [cell_reuse_recompute_pass]: 2.78e-06 [offload_activation]: 1.141e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.601e-05 [merge_recompute_call_nodes]: 1.35999e-06 [before_grad]: 1.371e-05 [set_forward_comm_id_for_comm_node_pass]: 5.86998e-06 [meta_fg_expand]: 3.26999e-06 [flash_sp_send_recv_attached]: 1.54998e-06 [receive_attached]: 2.14e-06 [after_resolve]: 1.181e-05 [a_after_grad]: 1.136e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.55999e-06 [auto_monad_grad]: 1.94e-06 [auto_monad_eliminator]: 1.162e-05 [cse]: 2.752e-05 [a_3]: 4.795e-05 [py_interpret_to_execute_after_opt_a]: 1.136e-05 [slice_cell_reuse_recomputed_activation]: 2.19001e-06 [rewriter_after_opt_a]: 2.925e-05 [convert_after_rewriter]: 1.29e-06 [order_py_execute_after_rewriter]: 1.24e-06 [mutable_eliminate]: 0.00076343 [opt_b]: 0.00026729, [1] [Cycle 1]: 0.00025802, [7] [b_1]: 0.00016151 [b_2]: 1.018e-05 [updatestate_depend_eliminate]: 1.023e-05 [updatestate_assign_eliminate]: 3.41999e-06 [updatestate_loads_eliminate]: 3.56001e-06 [renormalize]: 6.50005e-07 [cse]: 3.236e-05 [optimize_parallel_all_gather_comm]: 2.033e-05 [overlap_param_gather]: 2.26003e-06 [cconv]: 3.644e-05 [loop_unroll]: 0.00047791 [opt_after_cconv]: 0.00012798, [1] [Cycle 1]: 0.00012066, [7] [c_1]: 3.745e-05 [parameter_eliminate]: 4.08001e-06 [updatestate_depend_eliminate]: 7.16001e-06 [updatestate_assign_eliminate]: 3.56001e-06 [updatestate_loads_eliminate]: 3.38e-06 [cse]: 3.074e-05 [renormalize]: 3.50003e-07 [remove_dup_value]: 5.203e-05 [tuple_transform]: 9.318e-05, [1] [Cycle 1]: 8.664e-05, [4] [d_1]: 5.725e-05 [none_parameter_eliminate]: 1.90001e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 8.35999e-06 [partial_unused_args_eliminate]: 1.99999e-06 [add_recomputation]: 6.695e-05 [cse_after_recomputation]: 2.778e-05, [1] [Cycle 1]: 2.259e-05, [1] [cse]: 1.732e-05 [environ_conv]: 7.46001e-06 [swap_dp_allreduce_reducescatter]: 6.92002e-06 [bias_add_comm_swap]: 3.02002e-06 [label_micro_interleaved_index]: 5.27001e-06 [label_fine_grained_interleaved_index]: 2.56e-06 [merge_cast_opt]: 1.60999e-06 [slice_recompute_activation]: 2.34999e-06 [micro_interleaved_order_control]: 2.80002e-06 [assign_add_opt]: 1.40999e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 1.18001e-06 [full_micro_interleaved_order_control]: 2.17001e-06 [reorder_send_recv_between_fp_bp]: 3.09999e-06 [comm_op_add_attrs]: 1.50999e-06 [add_comm_op_reuse_tag]: 1.30999e-06 [interleave_split_concat_branches]: 1.23002e-06 [interleave_parallel_branches]: 1.05999e-06 [overlap_opt_shard_in_pipeline]: 1.22e-06 [overlap_opt_shard_grad_in_pipeline]: 1.90001e-06 [control_data_broadcast_order]: 1.759e-05 [grouped_pairwise_exchange_alltoall]: 1.80001e-06 [offloading_packed_experts]: 4.82e-06 [overlap_recompute_and_grad_model_parallel]: 5.84999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.24e-06 [overlap_recompute_comm]: 2.19001e-06 [overlap_grad_ring_attention]: 4.92e-06 [overlap_grad_flash_sp]: 2.508e-05 [begin_end_overlap_inline]: 4.60015e-07 [split_matmul_comm_elemetwise]: 2.51e-06 [split_layernorm_comm]: 1.82001e-06 [handle_group_info]: 1.44998e-06 [symbol_engine_optimizer]: 9.285e-05, [1] [Cycle 1]: 8.783e-05, [6] [build]: 4.50999e-06 [elim_shapecalc]: 1.379e-05 [elim_not_effective]: 1.718e-05 [opt_reshape]: 9.12001e-06 [fold_const_symbol]: 1.449e-05 [renormalize]: 2.69996e-07 [detach_backward]: 2.54999e-06 [pipeline_parallel_scheduler]: 1.54998e-06 [auto_monad_reorder]: 2.141e-05 [get_jit_bprop_graph]: 1.93002e-06 [rewriter_after_jit_bprop_graph]: 6.31998e-06 [opt_after_jit_grad]: 0.161415 [validate]: 6.544e-05 Sums bootstrap : 0.000473s : 0.04% type_inference : 0.968190s : 84.93% event_method : 0.000083s : 0.01% auto_monad : 0.000194s : 0.02% graph_reusing : 0.000013s : 0.00% inline : 0.000004s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000069s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000014s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000093s : 0.01% 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 : 0.000010s : 0.00% optimize.rewriter_before_opt_a : 0.000443s : 0.04% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000268s : 0.02% optimize.opt_a.loop_unroll : 0.000132s : 0.01% optimize.opt_a.a_1 : 0.002298s : 0.20% optimize.opt_a.with_stream_mark : 0.000176s : 0.02% optimize.opt_a.recompute_prepare : 0.000031s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000014s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000008s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000205s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000017s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.00% optimize.opt_a.merge_send_recv : 0.000021s : 0.00% optimize.opt_a.auto_parallel : 0.000023s : 0.00% optimize.opt_a.parallel : 0.000030s : 0.00% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000009s : 0.00% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000024s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000021s : 0.00% optimize.opt_a.virtual_dataset : 0.000017s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000015s : 0.00% optimize.opt_a.virtual_output : 0.000016s : 0.00% optimize.opt_a.merge_forward : 0.000012s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000025s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000039s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000029s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.00% optimize.opt_a.meta_fg_expand : 0.000009s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000025s : 0.00% optimize.opt_a.a_after_grad : 0.000024s : 0.00% optimize.opt_a.renormalize : 0.003003s : 0.26% optimize.opt_a.add_forward_monad_depend : 0.000013s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000041s : 0.00% optimize.opt_a.cse : 0.000081s : 0.01% optimize.opt_a.a_3 : 0.000152s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000029s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000763s : 0.07% optimize.opt_b.b_1 : 0.000162s : 0.01% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000032s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.00% optimize.loop_unroll : 0.000478s : 0.04% optimize.opt_after_cconv.c_1 : 0.000037s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000031s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000052s : 0.00% optimize.tuple_transform.d_1 : 0.000057s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000067s : 0.01% optimize.cse_after_recomputation.cse : 0.000017s : 0.00% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000007s : 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 : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000018s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000025s : 0.00% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000017s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 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 : 0.000021s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.161415s : 14.16% validate : 0.000065s : 0.01% Time group info: ------[substitution.] 0.000749 75 8.31% : 0.000062s : 11: substitution.cast_eliminate 0.33% : 0.000002s : 4: substitution.elim_not_effective 0.37% : 0.000003s : 4: substitution.fold_const_symbol 1.08% : 0.000008s : 5: substitution.graph_param_transform 83.17% : 0.000623s : 24: substitution.inline 0.87% : 0.000007s : 8: substitution.j_node_and_user_rematch 1.09% : 0.000008s : 8: substitution.remove_not_recompute_node 0.71% : 0.000005s : 2: substitution.replace_old_param 2.47% : 0.000019s : 7: substitution.switch_simplify 1.60% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.968052 2 99.49% : 0.963113s : 1: type_inference.infer 0.51% : 0.004939s : 1: type_inference.specialize ------[replace.] 0.000311 33 62.45% : 0.000194s : 24: replace.inline 30.87% : 0.000096s : 7: replace.switch_simplify 6.68% : 0.000021s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000634 33 96.23% : 0.000610s : 24: match.inline 2.25% : 0.000014s : 7: match.switch_simplify 1.52% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000464 2544 1.04% : 0.000005s : 32: predicate.accumulaten_eliminater 1.05% : 0.000005s : 5: predicate.ad_related_special_op_eliminate 0.32% : 0.000001s : 10: predicate.addn_check_dump 1.09% : 0.000005s : 32: predicate.addn_zero_filter 0.92% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.09% : 0.000010s : 42: predicate.arithmetic_simplify 1.63% : 0.000008s : 32: predicate.cast_eliminate 0.33% : 0.000002s : 10: predicate.check_bprop_eliminate 0.31% : 0.000001s : 10: predicate.compare_switch_simplify 0.09% : 0.000000s : 5: predicate.const_output_eliminate 0.37% : 0.000002s : 10: predicate.depend_value_elim 1.02% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.09% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.03% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.95% : 0.000004s : 10: predicate.dumpgradient_eliminate 0.13% : 0.000001s : 5: predicate.elim_not_effective 0.24% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.25% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.19% : 0.000006s : 37: predicate.environ_get_add_eliminate 1.31% : 0.000006s : 37: predicate.environ_get_depend_swap 1.45% : 0.000007s : 47: predicate.environ_get_eliminate 1.11% : 0.000005s : 37: predicate.environ_get_set_eliminate 1.81% : 0.000008s : 58: predicate.exchange_switch_depend_value 3.09% : 0.000014s : 58: predicate.float_depend_g_call 0.30% : 0.000001s : 10: predicate.float_environ_get_switch 0.47% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.10% : 0.000000s : 5: predicate.fold_const_symbol 0.36% : 0.000002s : 10: predicate.get_grad_eliminate 0.11% : 0.000001s : 5: predicate.graph_param_transform 0.31% : 0.000001s : 10: predicate.incorporate_call 0.28% : 0.000001s : 10: predicate.incorporate_call_switch 5.80% : 0.000027s : 120: predicate.inline 0.45% : 0.000002s : 10: predicate.inline_without_move 0.21% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.53% : 0.000002s : 10: predicate.less_batch_normalization 1.53% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.23% : 0.000010s : 76: predicate.load_eliminater 0.48% : 0.000002s : 5: predicate.loop_unroll_after_grad 3.93% : 0.000018s : 104: predicate.loop_unroll_before_grad 1.57% : 0.000007s : 42: predicate.make_slice_get_slice_eliminator 0.34% : 0.000002s : 10: predicate.merge_addn 0.35% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.32% : 0.000001s : 10: predicate.mini_step_allgather_replace 0.93% : 0.000004s : 32: predicate.minmaximum_grad 0.65% : 0.000003s : 5: predicate.mutable_eliminate 0.23% : 0.000001s : 5: predicate.opt_reshape 0.21% : 0.000001s : 5: predicate.parallel_virtual_node 2.68% : 0.000012s : 58: predicate.partial_defer_inline 1.31% : 0.000006s : 39: predicate.partial_eliminate 1.33% : 0.000006s : 32: predicate.print_const_string_wrapper 0.36% : 0.000002s : 10: predicate.reduce_all_const_elim 1.74% : 0.000008s : 32: predicate.reduce_eliminate 2.29% : 0.000011s : 76: predicate.redundant_stop_gradient_eliminater 0.22% : 0.000001s : 10: predicate.remove_not_recompute_node 7.06% : 0.000033s : 44: predicate.replace_applicator 0.22% : 0.000001s : 10: predicate.replace_old_param 0.18% : 0.000001s : 5: predicate.reset_defer_inline 1.04% : 0.000005s : 32: predicate.reshape_eliminate 0.43% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.24% : 0.000001s : 5: predicate.row_tensor_eliminate 0.68% : 0.000003s : 10: predicate.same_eliminate 0.34% : 0.000002s : 10: predicate.set_cell_output_no_recompute 0.48% : 0.000002s : 10: predicate.shard_identity_eliminate 0.41% : 0.000002s : 10: predicate.special_op_eliminate 0.47% : 0.000002s : 10: predicate.specialize_transform 0.55% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.40% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.18% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.15% : 0.000010s : 58: predicate.switch_defer_inline 2.26% : 0.000010s : 68: predicate.switch_layer_defer_inline 6.59% : 0.000031s : 191: predicate.switch_simplify 1.18% : 0.000006s : 32: predicate.tile_eliminate 0.99% : 0.000005s : 32: predicate.transpose_eliminate 1.49% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.43% : 0.000007s : 42: predicate.tuple_list_get_item_const_eliminator 1.33% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.60% : 0.000012s : 54: predicate.tuple_list_get_item_eliminator 1.31% : 0.000006s : 42: predicate.tuple_list_get_set_item_eliminator 1.94% : 0.000009s : 52: predicate.tuple_list_set_item_eliminator 1.38% : 0.000006s : 44: predicate.tuple_to_list_eliminator_ 2.17% : 0.000010s : 76: predicate.updatestate_pure_node_eliminater 2.63% : 0.000012s : 86: predicate.updatestate_useless_node_eliminater 0.22% : 0.000001s : 5: predicate.value_based_eliminate 0.37% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.37% : 0.000002s : 10: predicate.virtual_output_eliminate 0.14% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.25% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.005341 42 70.54% : 0.003768s : 16: func_graph_cloner_run.FuncGraphClonerGraph 29.46% : 0.001573s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.169785 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.51% : 0.005910s : 1: add_attr 0.50% : 0.005893s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000071s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.02% : 0.000206s : 1: auto_monad 0.00% : 0.000025s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.04% : 0.000510s : 1: bootstrap 0.00% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000031s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000011s : 1: environ_conv 0.01% : 0.000094s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000018s : 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% : 0.000008s : 1: label_micro_interleaved_index 0.04% : 0.000489s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.07% : 0.000776s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000023s : 1: opt.transform.mutable_eliminate 0.27% : 0.003204s : 78: opt.transform.opt_a 0.00% : 0.000036s : 1: opt.transform.opt_after_cconv 0.00% : 0.000057s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000143s : 28: opt.transform.opt_b 0.01% : 0.000063s : 2: opt.transform.opt_trans_graph 0.00% : 0.000051s : 4: opt.transform.symbol_engine_opt 0.64% : 0.007492s : 1: opt_a 0.01% : 0.000131s : 1: opt_after_cconv 13.80% : 0.161436s : 1: opt_after_jit_grad 0.02% : 0.000271s : 1: opt_b 0.89% : 0.010387s : 1: optimize 0.00% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000028s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 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 0.01% : 0.000099s : 1: pre_auto_parallel 0.00% : 0.000014s : 1: py_interpret_to_execute 0.00% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000057s : 1: remove_dup_value 0.14% : 0.001607s : 1: renormalize.infer 0.12% : 0.001381s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000033s : 1: rewriter_after_opt_a 0.04% : 0.000457s : 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% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000096s : 1: symbol_engine_optimizer 0.01% : 0.000096s : 1: tuple_transform 82.77% : 0.968230s : 1: type_inference TotalTime = 0.791364, [21] [bootstrap]: 0.00051284 [type_inference]: 0.673894 [event_method]: 6.777e-05 [auto_monad]: 0.00026281 [graph_reusing]: 1.92e-05 [inline]: 3.38e-06 [add_attr]: 0.00507687, [1] [add_attr_with_inline]: 0.00506089, [1] [Cycle 1]: 0.00017273, [2] [tag_attr]: 9.518e-05 [meta_addattr_fg_expand]: 2.025e-05 [parallel-infer-symbol]: 4.38001e-06 [pre_auto_parallel]: 0.00012772 [insert-virtual-dataset]: 2.61e-06 [parallel-infer-symbol-second]: 8.89995e-07 [dataset_repeat_opt]: 2.73998e-06 [pipeline_split]: 1.84998e-06 [optimize]: 0.110566, [53] [py_interpret_to_execute]: 1.233e-05 [rewriter_before_opt_a]: 0.00060521 [opt_a]: 0.107586, [2] [Cycle 1]: 0.106336, [45] [expand_dump_flag]: 1.408e-05 [switch_simplify]: 0.00054933 [loop_unroll]: 0.00012929 [a_1]: 0.00305487 [with_stream_mark]: 4.405e-05 [recompute_prepare]: 3.516e-05 [updatestate_depend_eliminate]: 1.23e-05 [updatestate_assign_eliminate]: 7.47998e-06 [updatestate_loads_eliminate]: 8.2e-06 [parameter_eliminate]: 3.95998e-06 [a_2]: 0.00024381 [accelerated_algorithm]: 1.828e-05 [shard]: 1.74e-06 [meta_shard_fg_expand]: 0.00015288 [shard_inline]: 2.523e-05 [merge_send_recv]: 2.577e-05 [auto_parallel]: 1.834e-05 [parallel]: 2.38e-05 [flash_sp]: 1.738e-05 [merge_comm]: 1.011e-05 [allreduce_fusion]: 8.84e-06 [matmul_add_comm_reduction]: 3.864e-05 [allreduce_slice_to_reducescatter]: 9.50007e-07 [virtual_shard_identity]: 2.385e-05 [virtual_dataset]: 1.653e-05 [get_grad_eliminate_]: 1.632e-05 [virtual_output]: 1.616e-05 [merge_forward]: 1.078e-05 [cell_reuse_recompute_pass]: 2.54999e-06 [offload_activation]: 5.179e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.741e-05 [merge_recompute_call_nodes]: 2.11998e-06 [before_grad]: 2.932e-05 [set_forward_comm_id_for_comm_node_pass]: 1.242e-05 [meta_fg_expand]: 0.00393009 [flash_sp_send_recv_attached]: 1.211e-05 [receive_attached]: 2.62001e-06 [after_resolve]: 0.00017385 [a_after_grad]: 0.0001688 [renormalize]: 0.00827 [add_forward_monad_depend]: 8.3e-06 [auto_monad_grad]: 2.73998e-06 [auto_monad_eliminator]: 1.905e-05 [cse]: 3.114e-05 [a_3]: 5.226e-05 [Cycle 2]: 0.00122912, [45] [expand_dump_flag]: 2.89001e-06 [switch_simplify]: 7.88001e-06 [loop_unroll]: 4.50001e-06 [a_1]: 5.411e-05 [with_stream_mark]: 1.722e-05 [recompute_prepare]: 4.40999e-06 [updatestate_depend_eliminate]: 3.15998e-06 [updatestate_assign_eliminate]: 2.46998e-06 [updatestate_loads_eliminate]: 2.51e-06 [parameter_eliminate]: 1.87001e-06 [a_2]: 3.906e-05 [accelerated_algorithm]: 4.13001e-06 [shard]: 2.24999e-06 [meta_shard_fg_expand]: 1.76e-06 [shard_inline]: 3.22997e-06 [merge_send_recv]: 7.41001e-06 [auto_parallel]: 8.35999e-06 [parallel]: 9.12999e-06 [flash_sp]: 3.96001e-06 [merge_comm]: 2.58e-06 [allreduce_fusion]: 2.78e-06 [matmul_add_comm_reduction]: 8.06001e-06 [allreduce_slice_to_reducescatter]: 9.70002e-07 [virtual_shard_identity]: 5.23002e-06 [virtual_dataset]: 4.09002e-06 [get_grad_eliminate_]: 3.73001e-06 [virtual_output]: 3.43999e-06 [merge_forward]: 3.6e-06 [cell_reuse_recompute_pass]: 3.16001e-06 [offload_activation]: 7.83001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.338e-05 [merge_recompute_call_nodes]: 1.81998e-06 [before_grad]: 7.20998e-06 [set_forward_comm_id_for_comm_node_pass]: 2.78e-06 [meta_fg_expand]: 0.00067179 [flash_sp_send_recv_attached]: 3.39001e-06 [receive_attached]: 2.78e-06 [after_resolve]: 1.684e-05 [a_after_grad]: 6.07999e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 4.43999e-06 [auto_monad_grad]: 2.12999e-06 [auto_monad_eliminator]: 1.126e-05 [cse]: 2.034e-05 [a_3]: 2.434e-05 [py_interpret_to_execute_after_opt_a]: 8.80999e-06 [slice_cell_reuse_recomputed_activation]: 2.36998e-06 [rewriter_after_opt_a]: 2.149e-05 [convert_after_rewriter]: 1.41998e-06 [order_py_execute_after_rewriter]: 1.42999e-06 [mutable_eliminate]: 0.0007459 [opt_b]: 0.00014723, [1] [Cycle 1]: 0.00013761, [7] [b_1]: 7.085e-05 [b_2]: 4.99e-06 [updatestate_depend_eliminate]: 6.05002e-06 [updatestate_assign_eliminate]: 2.07001e-06 [updatestate_loads_eliminate]: 1.84998e-06 [renormalize]: 6.50005e-07 [cse]: 1.829e-05 [optimize_parallel_all_gather_comm]: 1.897e-05 [overlap_param_gather]: 1.88002e-06 [cconv]: 3.557e-05 [loop_unroll]: 0.00062341 [opt_after_cconv]: 8.722e-05, [1] [Cycle 1]: 7.984e-05, [7] [c_1]: 1.514e-05 [parameter_eliminate]: 4.97e-06 [updatestate_depend_eliminate]: 6.01e-06 [updatestate_assign_eliminate]: 1.57999e-06 [updatestate_loads_eliminate]: 1.39e-06 [cse]: 1.935e-05 [renormalize]: 2.80008e-07 [remove_dup_value]: 1.475e-05 [tuple_transform]: 5.555e-05, [1] [Cycle 1]: 4.997e-05, [4] [d_1]: 2.711e-05 [none_parameter_eliminate]: 1.63002e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 4.03999e-06 [partial_unused_args_eliminate]: 2.09999e-06 [add_recomputation]: 3.97e-05 [cse_after_recomputation]: 1.895e-05, [1] [Cycle 1]: 1.336e-05, [1] [cse]: 8.25999e-06 [environ_conv]: 6.41998e-06 [swap_dp_allreduce_reducescatter]: 4.06001e-06 [bias_add_comm_swap]: 3.35e-06 [label_micro_interleaved_index]: 5.78997e-06 [label_fine_grained_interleaved_index]: 2.54999e-06 [merge_cast_opt]: 1.84998e-06 [slice_recompute_activation]: 2.14999e-06 [micro_interleaved_order_control]: 2.68e-06 [assign_add_opt]: 1.35999e-06 [ForceFp32Comm]: 7.89994e-07 [remove_cast_before_assign_add]: 1.07998e-06 [full_micro_interleaved_order_control]: 2.32999e-06 [reorder_send_recv_between_fp_bp]: 2.78e-06 [comm_op_add_attrs]: 1.19998e-06 [add_comm_op_reuse_tag]: 1.66998e-06 [interleave_split_concat_branches]: 1.10999e-06 [interleave_parallel_branches]: 1.04e-06 [overlap_opt_shard_in_pipeline]: 1.24e-06 [overlap_opt_shard_grad_in_pipeline]: 2.00002e-06 [control_data_broadcast_order]: 1.089e-05 [grouped_pairwise_exchange_alltoall]: 1.54e-06 [offloading_packed_experts]: 3.06001e-06 [overlap_recompute_and_grad_model_parallel]: 3.8e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.15001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.34998e-06 [overlap_recompute_comm]: 2.46e-06 [overlap_grad_ring_attention]: 3.88999e-06 [overlap_grad_flash_sp]: 1.834e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.50002e-06 [split_layernorm_comm]: 1.84e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 0.00019133, [1] [Cycle 1]: 0.00018609, [6] [build]: 3.33e-06 [elim_shapecalc]: 0.00013017 [elim_not_effective]: 1.103e-05 [opt_reshape]: 5.27001e-06 [fold_const_symbol]: 6.21998e-06 [renormalize]: 5.29981e-07 [detach_backward]: 2.69999e-06 [pipeline_parallel_scheduler]: 1.52999e-06 [auto_monad_reorder]: 1.806e-05 [get_jit_bprop_graph]: 2.25002e-06 [rewriter_after_jit_bprop_graph]: 6.41e-06 [opt_after_jit_grad]: 0.0005331 [validate]: 4.125e-05 Sums bootstrap : 0.000513s : 0.07% type_inference : 0.673894s : 96.75% event_method : 0.000068s : 0.01% auto_monad : 0.000263s : 0.04% graph_reusing : 0.000019s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000095s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000020s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000128s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000012s : 0.00% optimize.rewriter_before_opt_a : 0.000605s : 0.09% optimize.opt_a.expand_dump_flag : 0.000017s : 0.00% optimize.opt_a.switch_simplify : 0.000557s : 0.08% optimize.opt_a.loop_unroll : 0.000134s : 0.02% optimize.opt_a.a_1 : 0.003109s : 0.45% optimize.opt_a.with_stream_mark : 0.000061s : 0.01% optimize.opt_a.recompute_prepare : 0.000040s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000283s : 0.04% optimize.opt_a.accelerated_algorithm : 0.000022s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000155s : 0.02% optimize.opt_a.shard_inline : 0.000028s : 0.00% optimize.opt_a.merge_send_recv : 0.000033s : 0.00% optimize.opt_a.auto_parallel : 0.000027s : 0.00% optimize.opt_a.parallel : 0.000033s : 0.00% optimize.opt_a.flash_sp : 0.000021s : 0.00% optimize.opt_a.merge_comm : 0.000013s : 0.00% optimize.opt_a.allreduce_fusion : 0.000012s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000047s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000029s : 0.00% optimize.opt_a.virtual_dataset : 0.000021s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000020s : 0.00% optimize.opt_a.virtual_output : 0.000020s : 0.00% optimize.opt_a.merge_forward : 0.000014s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000060s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000051s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000037s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000015s : 0.00% optimize.opt_a.meta_fg_expand : 0.004602s : 0.66% optimize.opt_a.flash_sp_send_recv_attached : 0.000016s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000191s : 0.03% optimize.opt_a.a_after_grad : 0.000175s : 0.03% optimize.opt_a.renormalize : 0.008270s : 1.19% optimize.opt_a.add_forward_monad_depend : 0.000013s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000030s : 0.00% optimize.opt_a.cse : 0.000051s : 0.01% optimize.opt_a.a_3 : 0.000077s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000009s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000746s : 0.11% optimize.opt_b.b_1 : 0.000071s : 0.01% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000018s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000019s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.01% optimize.loop_unroll : 0.000623s : 0.09% optimize.opt_after_cconv.c_1 : 0.000015s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000019s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000027s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000040s : 0.01% optimize.cse_after_recomputation.cse : 0.000008s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 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 : 0.000003s : 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 : 0.000002s : 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 : 0.000011s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000018s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000130s : 0.02% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000006s : 0.00% 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 : 0.000018s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000533s : 0.08% validate : 0.000041s : 0.01% Time group info: ------[substitution.] 0.001188 134 3.10% : 0.000037s : 8: substitution.cast_eliminate 0.16% : 0.000002s : 1: substitution.elim_not_effective 0.75% : 0.000009s : 9: substitution.float_depend_g_call 0.38% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.09% : 0.000001s : 1: substitution.fold_const_symbol 0.42% : 0.000005s : 1: substitution.graph_param_transform 0.32% : 0.000004s : 2: substitution.incorporate_call 0.19% : 0.000002s : 2: substitution.incorporate_call_switch 79.43% : 0.000944s : 33: substitution.inline 3.82% : 0.000045s : 4: substitution.inline_without_move 0.72% : 0.000009s : 11: substitution.j_node_and_user_rematch 0.33% : 0.000004s : 2: substitution.minmaximum_grad 0.61% : 0.000007s : 9: substitution.partial_eliminate 0.85% : 0.000010s : 11: substitution.remove_not_recompute_node 0.28% : 0.000003s : 1: substitution.replace_applicator 1.04% : 0.000012s : 12: substitution.replace_old_param 0.36% : 0.000004s : 1: substitution.set_cell_output_no_recompute 2.36% : 0.000028s : 10: substitution.switch_simplify 0.78% : 0.000009s : 2: substitution.tuple_list_convert_item_index_to_positive 0.27% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.44% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.84% : 0.000034s : 6: substitution.tuple_list_get_item_eliminator 0.44% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.673656 2 98.96% : 0.666660s : 1: type_inference.infer 1.04% : 0.006996s : 1: type_inference.specialize ------[replace.] 0.000563 45 46.86% : 0.000264s : 33: replace.inline 49.90% : 0.000281s : 10: replace.switch_simplify 3.24% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000964 45 96.00% : 0.000926s : 33: match.inline 2.33% : 0.000022s : 10: match.switch_simplify 1.67% : 0.000016s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000549 3189 1.16% : 0.000006s : 43: predicate.accumulaten_eliminater 0.26% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.36% : 0.000002s : 14: predicate.addn_check_dump 1.11% : 0.000006s : 43: predicate.addn_zero_filter 1.07% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 2.35% : 0.000013s : 57: predicate.arithmetic_simplify 1.47% : 0.000008s : 43: predicate.cast_eliminate 0.11% : 0.000001s : 2: predicate.check_bprop_eliminate 0.42% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.43% : 0.000002s : 14: predicate.depend_value_elim 1.26% : 0.000007s : 43: predicate.dict_get_item_const_eliminator 1.30% : 0.000007s : 43: predicate.dict_get_item_eliminator 1.17% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.25% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.07% : 0.000000s : 1: predicate.elim_not_effective 0.10% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.13% : 0.000006s : 44: predicate.environ_add_const_eliminate 1.09% : 0.000006s : 44: predicate.environ_get_add_eliminate 1.19% : 0.000007s : 44: predicate.environ_get_depend_swap 1.60% : 0.000009s : 58: predicate.environ_get_eliminate 1.17% : 0.000006s : 44: predicate.environ_get_set_eliminate 2.02% : 0.000011s : 78: predicate.exchange_switch_depend_value 3.14% : 0.000017s : 78: predicate.float_depend_g_call 0.37% : 0.000002s : 14: predicate.float_environ_get_switch 0.42% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.49% : 0.000003s : 14: predicate.get_grad_eliminate 0.04% : 0.000000s : 1: predicate.graph_param_transform 0.36% : 0.000002s : 14: predicate.incorporate_call 0.32% : 0.000002s : 14: predicate.incorporate_call_switch 5.87% : 0.000032s : 151: predicate.inline 2.66% : 0.000015s : 70: predicate.inline_without_move 0.19% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.68% : 0.000004s : 14: predicate.less_batch_normalization 1.31% : 0.000007s : 47: predicate.list_to_tuple_eliminator_ 2.13% : 0.000012s : 90: predicate.load_eliminater 0.44% : 0.000002s : 1: predicate.loop_unroll_after_grad 4.76% : 0.000026s : 141: predicate.loop_unroll_before_grad 1.28% : 0.000007s : 45: predicate.make_slice_get_slice_eliminator 0.37% : 0.000002s : 14: predicate.merge_addn 0.12% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.16% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.05% : 0.000006s : 43: predicate.minmaximum_grad 0.38% : 0.000002s : 1: predicate.mutable_eliminate 0.10% : 0.000001s : 1: predicate.opt_reshape 0.09% : 0.000000s : 1: predicate.parallel_virtual_node 3.31% : 0.000018s : 78: predicate.partial_defer_inline 1.30% : 0.000007s : 46: predicate.partial_eliminate 1.09% : 0.000006s : 43: predicate.print_const_string_wrapper 0.44% : 0.000002s : 14: predicate.reduce_all_const_elim 1.43% : 0.000008s : 43: predicate.reduce_eliminate 2.27% : 0.000012s : 90: predicate.redundant_stop_gradient_eliminater 0.30% : 0.000002s : 14: predicate.remove_not_recompute_node 0.98% : 0.000005s : 47: predicate.replace_applicator 1.19% : 0.000007s : 70: predicate.replace_old_param 0.04% : 0.000000s : 1: predicate.reset_defer_inline 1.24% : 0.000007s : 43: predicate.reshape_eliminate 0.26% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.10% : 0.000001s : 1: predicate.row_tensor_eliminate 0.16% : 0.000001s : 2: predicate.same_eliminate 0.23% : 0.000001s : 14: predicate.set_cell_output_no_recompute 0.55% : 0.000003s : 14: predicate.shard_identity_eliminate 0.11% : 0.000001s : 2: predicate.special_op_eliminate 0.43% : 0.000002s : 14: predicate.specialize_transform 0.40% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 2.36% : 0.000013s : 70: predicate.stack_unstack_eliminate 0.08% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.25% : 0.000012s : 78: predicate.switch_defer_inline 2.54% : 0.000014s : 80: predicate.switch_layer_defer_inline 9.20% : 0.000051s : 254: predicate.switch_simplify 1.11% : 0.000006s : 43: predicate.tile_eliminate 1.17% : 0.000006s : 43: predicate.transpose_eliminate 1.44% : 0.000008s : 45: predicate.tuple_list_convert_item_index_to_positive 1.37% : 0.000008s : 45: predicate.tuple_list_get_item_const_eliminator 1.29% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.73% : 0.000015s : 61: predicate.tuple_list_get_item_eliminator 1.38% : 0.000008s : 45: predicate.tuple_list_get_set_item_eliminator 2.06% : 0.000011s : 59: predicate.tuple_list_set_item_eliminator 1.30% : 0.000007s : 47: predicate.tuple_to_list_eliminator_ 2.17% : 0.000012s : 90: predicate.updatestate_pure_node_eliminater 2.62% : 0.000014s : 104: predicate.updatestate_useless_node_eliminater 0.07% : 0.000000s : 1: predicate.value_based_eliminate 0.54% : 0.000003s : 14: predicate.virtual_dataset_eliminate 0.43% : 0.000002s : 14: predicate.virtual_output_eliminate 0.03% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.17% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.005480 69 73.34% : 0.004019s : 33: func_graph_cloner_run.FuncGraphClonerGraph 26.66% : 0.001461s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.919948 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.55% : 0.005084s : 1: add_attr 0.55% : 0.005065s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.00% : 0.000044s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.03% : 0.000274s : 1: auto_monad 0.00% : 0.000022s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.06% : 0.000552s : 1: bootstrap 0.00% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000014s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.01% : 0.000077s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000024s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.07% : 0.000637s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.08% : 0.000759s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.00% : 0.000016s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000014s : 1: opt.transform.mutable_eliminate 0.51% : 0.004666s : 78: opt.transform.opt_a 0.00% : 0.000014s : 1: opt.transform.opt_after_cconv 0.00% : 0.000018s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000046s : 28: opt.transform.opt_b 0.00% : 0.000029s : 2: opt.transform.opt_trans_graph 0.00% : 0.000028s : 4: opt.transform.symbol_engine_opt 11.70% : 0.107590s : 1: opt_a 0.01% : 0.000091s : 1: opt_after_cconv 0.06% : 0.000546s : 1: opt_after_jit_grad 0.02% : 0.000151s : 1: opt_b 12.02% : 0.110572s : 1: optimize 0.00% : 0.000023s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000022s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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 0.01% : 0.000135s : 1: pre_auto_parallel 0.00% : 0.000016s : 1: py_interpret_to_execute 0.00% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000019s : 1: remove_dup_value 0.79% : 0.007307s : 1: renormalize.infer 0.10% : 0.000948s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000025s : 1: rewriter_after_opt_a 0.07% : 0.000621s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000194s : 1: symbol_engine_optimizer 0.01% : 0.000058s : 1: tuple_transform 73.26% : 0.673935s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:42:09.357.307 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.69612, [21] [bootstrap]: 0.00214164 [type_inference]: 0.562637 [event_method]: 7.996e-05 [auto_monad]: 0.00018744 [graph_reusing]: 1.305e-05 [inline]: 2.97002e-06 [add_attr]: 0.00436709, [1] [add_attr_with_inline]: 0.00435404, [1] [Cycle 1]: 0.00013528, [2] [tag_attr]: 7.103e-05 [meta_addattr_fg_expand]: 1.316e-05 [parallel-infer-symbol]: 3.61999e-06 [pre_auto_parallel]: 9.152e-05 [insert-virtual-dataset]: 2.46e-06 [parallel-infer-symbol-second]: 6.59988e-07 [dataset_repeat_opt]: 1.97999e-06 [pipeline_split]: 1.67999e-06 [optimize]: 0.125458, [53] [py_interpret_to_execute]: 1.061e-05 [rewriter_before_opt_a]: 0.00042442 [opt_a]: 0.107147, [2] [Cycle 1]: 0.106221, [45] [expand_dump_flag]: 5.88998e-06 [switch_simplify]: 0.00024916 [loop_unroll]: 8.045e-05 [a_1]: 0.101655 [with_stream_mark]: 3.466e-05 [recompute_prepare]: 1.638e-05 [updatestate_depend_eliminate]: 6.14001e-06 [updatestate_assign_eliminate]: 4.65001e-06 [updatestate_loads_eliminate]: 3.85998e-06 [parameter_eliminate]: 3.24001e-06 [a_2]: 0.00011405 [accelerated_algorithm]: 1.05e-05 [shard]: 2.78e-06 [meta_shard_fg_expand]: 7.33e-06 [shard_inline]: 9.09e-06 [merge_send_recv]: 1.184e-05 [auto_parallel]: 1.225e-05 [parallel]: 2.183e-05 [flash_sp]: 1.134e-05 [merge_comm]: 5.40999e-06 [allreduce_fusion]: 5.15999e-06 [matmul_add_comm_reduction]: 1.282e-05 [allreduce_slice_to_reducescatter]: 1.21002e-06 [virtual_shard_identity]: 1.184e-05 [virtual_dataset]: 8.82999e-06 [get_grad_eliminate_]: 8.58001e-06 [virtual_output]: 9.10001e-06 [merge_forward]: 5.74e-06 [cell_reuse_recompute_pass]: 1.94e-06 [offload_activation]: 1.221e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.043e-05 [merge_recompute_call_nodes]: 1.42e-06 [before_grad]: 1.539e-05 [set_forward_comm_id_for_comm_node_pass]: 5.46e-06 [meta_fg_expand]: 5.30001e-06 [flash_sp_send_recv_attached]: 2.86e-06 [receive_attached]: 2.48002e-06 [after_resolve]: 1.514e-05 [a_after_grad]: 1.312e-05 [renormalize]: 0.00325642 [add_forward_monad_depend]: 1.205e-05 [auto_monad_grad]: 2.97002e-06 [auto_monad_eliminator]: 3.013e-05 [cse]: 5.117e-05 [a_3]: 7.881e-05 [Cycle 2]: 0.00090704, [45] [expand_dump_flag]: 2.71999e-06 [switch_simplify]: 1.214e-05 [loop_unroll]: 8.42e-06 [a_1]: 0.00021043 [with_stream_mark]: 2.551e-05 [recompute_prepare]: 8.97999e-06 [updatestate_depend_eliminate]: 5.57001e-06 [updatestate_assign_eliminate]: 4.32e-06 [updatestate_loads_eliminate]: 4.28999e-06 [parameter_eliminate]: 2.27999e-06 [a_2]: 0.0001004 [accelerated_algorithm]: 8.67998e-06 [shard]: 2.36e-06 [meta_shard_fg_expand]: 3.3e-06 [shard_inline]: 7.78999e-06 [merge_send_recv]: 1.125e-05 [auto_parallel]: 1.151e-05 [parallel]: 1.192e-05 [flash_sp]: 4.45e-06 [merge_comm]: 5.35999e-06 [allreduce_fusion]: 4.82998e-06 [matmul_add_comm_reduction]: 1.237e-05 [allreduce_slice_to_reducescatter]: 7.49977e-07 [virtual_shard_identity]: 1.16e-05 [virtual_dataset]: 8.40001e-06 [get_grad_eliminate_]: 8.01001e-06 [virtual_output]: 8.10999e-06 [merge_forward]: 5.02e-06 [cell_reuse_recompute_pass]: 3.39001e-06 [offload_activation]: 1.292e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.786e-05 [merge_recompute_call_nodes]: 1.44998e-06 [before_grad]: 1.465e-05 [set_forward_comm_id_for_comm_node_pass]: 5.87999e-06 [meta_fg_expand]: 4.05998e-06 [flash_sp_send_recv_attached]: 1.83002e-06 [receive_attached]: 2.48e-06 [after_resolve]: 1.345e-05 [a_after_grad]: 1.31e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 2.09e-06 [auto_monad_grad]: 1.40001e-06 [auto_monad_eliminator]: 1.168e-05 [cse]: 2.704e-05 [a_3]: 5.051e-05 [py_interpret_to_execute_after_opt_a]: 1.121e-05 [slice_cell_reuse_recomputed_activation]: 2.13002e-06 [rewriter_after_opt_a]: 3.436e-05 [convert_after_rewriter]: 1.25999e-06 [order_py_execute_after_rewriter]: 1.20001e-06 [mutable_eliminate]: 0.00085599 [opt_b]: 0.00030135, [1] [Cycle 1]: 0.00029134, [7] [b_1]: 0.00017671 [b_2]: 1.092e-05 [updatestate_depend_eliminate]: 1.137e-05 [updatestate_assign_eliminate]: 3.8e-06 [updatestate_loads_eliminate]: 3.55e-06 [renormalize]: 1.10999e-06 [cse]: 4.552e-05 [optimize_parallel_all_gather_comm]: 2.366e-05 [overlap_param_gather]: 2.61999e-06 [cconv]: 4.129e-05 [loop_unroll]: 0.0156391 [opt_after_cconv]: 0.00019074, [1] [Cycle 1]: 0.00017764, [7] [c_1]: 4.852e-05 [parameter_eliminate]: 7.15998e-06 [updatestate_depend_eliminate]: 1.427e-05 [updatestate_assign_eliminate]: 4.56002e-06 [updatestate_loads_eliminate]: 5.10001e-06 [cse]: 6.091e-05 [renormalize]: 1.11002e-06 [remove_dup_value]: 5.826e-05 [tuple_transform]: 9.996e-05, [1] [Cycle 1]: 9.36e-05, [4] [d_1]: 6.48e-05 [none_parameter_eliminate]: 1.79998e-06 [renormalize]: 1.60013e-07 [switch_simplify]: 8.48001e-06 [partial_unused_args_eliminate]: 1.93002e-06 [add_recomputation]: 7.45e-05 [cse_after_recomputation]: 3.334e-05, [1] [Cycle 1]: 2.823e-05, [1] [cse]: 2.133e-05 [environ_conv]: 7.82002e-06 [swap_dp_allreduce_reducescatter]: 8.36002e-06 [bias_add_comm_swap]: 3.21001e-06 [label_micro_interleaved_index]: 9.46998e-06 [label_fine_grained_interleaved_index]: 2.79001e-06 [merge_cast_opt]: 1.50001e-06 [slice_recompute_activation]: 2.31e-06 [micro_interleaved_order_control]: 2.26e-06 [assign_add_opt]: 1.42e-06 [ForceFp32Comm]: 9.30013e-07 [remove_cast_before_assign_add]: 7.30011e-07 [full_micro_interleaved_order_control]: 2.50002e-06 [reorder_send_recv_between_fp_bp]: 2.58e-06 [comm_op_add_attrs]: 1.32e-06 [add_comm_op_reuse_tag]: 1.05001e-06 [interleave_split_concat_branches]: 1.10999e-06 [interleave_parallel_branches]: 1.10001e-06 [overlap_opt_shard_in_pipeline]: 1.39e-06 [overlap_opt_shard_grad_in_pipeline]: 2.08998e-06 [control_data_broadcast_order]: 1.802e-05 [grouped_pairwise_exchange_alltoall]: 1.45001e-06 [offloading_packed_experts]: 6.54001e-06 [overlap_recompute_and_grad_model_parallel]: 5.20999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.08001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.19e-06 [overlap_recompute_comm]: 2.39001e-06 [overlap_grad_ring_attention]: 5.58002e-06 [overlap_grad_flash_sp]: 2.64e-05 [begin_end_overlap_inline]: 5.10016e-07 [split_matmul_comm_elemetwise]: 1.99e-06 [split_layernorm_comm]: 1.66e-06 [handle_group_info]: 1.07e-06 [symbol_engine_optimizer]: 9.687e-05, [1] [Cycle 1]: 9.128e-05, [6] [build]: 4.70999e-06 [elim_shapecalc]: 1.562e-05 [elim_not_effective]: 1.692e-05 [opt_reshape]: 9.00001e-06 [fold_const_symbol]: 1.395e-05 [renormalize]: 1.30007e-07 [detach_backward]: 2.05002e-06 [pipeline_parallel_scheduler]: 1.32e-06 [auto_monad_reorder]: 2.201e-05 [get_jit_bprop_graph]: 2.07999e-06 [rewriter_after_jit_bprop_graph]: 7.06001e-06 [opt_after_jit_grad]: 0.00077412 [validate]: 5.892e-05 Sums bootstrap : 0.002142s : 0.31% type_inference : 0.562637s : 81.48% event_method : 0.000080s : 0.01% auto_monad : 0.000187s : 0.03% graph_reusing : 0.000013s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000071s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000013s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000092s : 0.01% 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 : 0.000011s : 0.00% optimize.rewriter_before_opt_a : 0.000424s : 0.06% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000261s : 0.04% optimize.opt_a.loop_unroll : 0.000089s : 0.01% optimize.opt_a.a_1 : 0.101866s : 14.75% optimize.opt_a.with_stream_mark : 0.000060s : 0.01% optimize.opt_a.recompute_prepare : 0.000025s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000008s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000214s : 0.03% optimize.opt_a.accelerated_algorithm : 0.000019s : 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 : 0.000017s : 0.00% optimize.opt_a.merge_send_recv : 0.000023s : 0.00% optimize.opt_a.auto_parallel : 0.000024s : 0.00% optimize.opt_a.parallel : 0.000034s : 0.00% optimize.opt_a.flash_sp : 0.000016s : 0.00% optimize.opt_a.merge_comm : 0.000011s : 0.00% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000025s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000023s : 0.00% optimize.opt_a.virtual_dataset : 0.000017s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000017s : 0.00% optimize.opt_a.virtual_output : 0.000017s : 0.00% optimize.opt_a.merge_forward : 0.000011s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000025s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000038s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000030s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.00% optimize.opt_a.meta_fg_expand : 0.000009s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000029s : 0.00% optimize.opt_a.a_after_grad : 0.000026s : 0.00% optimize.opt_a.renormalize : 0.003256s : 0.47% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000042s : 0.01% optimize.opt_a.cse : 0.000078s : 0.01% optimize.opt_a.a_3 : 0.000129s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000034s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000856s : 0.12% optimize.opt_b.b_1 : 0.000177s : 0.03% optimize.opt_b.b_2 : 0.000011s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000046s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000024s : 0.00% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000041s : 0.01% optimize.loop_unroll : 0.015639s : 2.26% optimize.opt_after_cconv.c_1 : 0.000049s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000014s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.cse : 0.000061s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000058s : 0.01% optimize.tuple_transform.d_1 : 0.000065s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000075s : 0.01% optimize.cse_after_recomputation.cse : 0.000021s : 0.00% optimize.environ_conv : 0.000008s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000008s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 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 : 0.000002s : 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 : 0.000018s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000007s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000026s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000016s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000017s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 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 : 0.000022s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000774s : 0.11% validate : 0.000059s : 0.01% Time group info: ------[substitution.] 0.000814 75 10.13% : 0.000082s : 11: substitution.cast_eliminate 0.47% : 0.000004s : 4: substitution.elim_not_effective 0.24% : 0.000002s : 4: substitution.fold_const_symbol 0.88% : 0.000007s : 5: substitution.graph_param_transform 81.93% : 0.000667s : 24: substitution.inline 0.78% : 0.000006s : 8: substitution.j_node_and_user_rematch 0.99% : 0.000008s : 8: substitution.remove_not_recompute_node 0.67% : 0.000005s : 2: substitution.replace_old_param 2.30% : 0.000019s : 7: substitution.switch_simplify 1.60% : 0.000013s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.562501 2 99.02% : 0.557012s : 1: type_inference.infer 0.98% : 0.005488s : 1: type_inference.specialize ------[replace.] 0.000305 33 63.92% : 0.000195s : 24: replace.inline 30.14% : 0.000092s : 7: replace.switch_simplify 5.94% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000678 33 96.33% : 0.000653s : 24: match.inline 2.09% : 0.000014s : 7: match.switch_simplify 1.59% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000529 2544 1.12% : 0.000006s : 32: predicate.accumulaten_eliminater 0.66% : 0.000003s : 5: predicate.ad_related_special_op_eliminate 0.40% : 0.000002s : 10: predicate.addn_check_dump 1.10% : 0.000006s : 32: predicate.addn_zero_filter 0.86% : 0.000005s : 32: predicate.adjust_all_reduce_mul_add 2.26% : 0.000012s : 42: predicate.arithmetic_simplify 1.54% : 0.000008s : 32: predicate.cast_eliminate 0.33% : 0.000002s : 10: predicate.check_bprop_eliminate 0.29% : 0.000002s : 10: predicate.compare_switch_simplify 0.09% : 0.000000s : 5: predicate.const_output_eliminate 0.36% : 0.000002s : 10: predicate.depend_value_elim 1.00% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.02% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.06% : 0.000006s : 32: predicate.dict_set_item_eliminator 0.67% : 0.000004s : 10: predicate.dumpgradient_eliminate 0.13% : 0.000001s : 5: predicate.elim_not_effective 0.23% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.22% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.12% : 0.000006s : 37: predicate.environ_get_add_eliminate 1.02% : 0.000005s : 37: predicate.environ_get_depend_swap 2.67% : 0.000014s : 47: predicate.environ_get_eliminate 1.03% : 0.000005s : 37: predicate.environ_get_set_eliminate 7.00% : 0.000037s : 58: predicate.exchange_switch_depend_value 2.44% : 0.000013s : 58: predicate.float_depend_g_call 0.28% : 0.000002s : 10: predicate.float_environ_get_switch 0.44% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.07% : 0.000000s : 5: predicate.fold_const_symbol 0.39% : 0.000002s : 10: predicate.get_grad_eliminate 0.09% : 0.000001s : 5: predicate.graph_param_transform 0.27% : 0.000001s : 10: predicate.incorporate_call 0.25% : 0.000001s : 10: predicate.incorporate_call_switch 5.39% : 0.000029s : 120: predicate.inline 0.47% : 0.000002s : 10: predicate.inline_without_move 0.16% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.54% : 0.000003s : 10: predicate.less_batch_normalization 1.51% : 0.000008s : 44: predicate.list_to_tuple_eliminator_ 2.12% : 0.000011s : 76: predicate.load_eliminater 1.45% : 0.000008s : 5: predicate.loop_unroll_after_grad 2.78% : 0.000015s : 104: predicate.loop_unroll_before_grad 1.57% : 0.000008s : 42: predicate.make_slice_get_slice_eliminator 0.30% : 0.000002s : 10: predicate.merge_addn 0.28% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.32% : 0.000002s : 10: predicate.mini_step_allgather_replace 0.94% : 0.000005s : 32: predicate.minmaximum_grad 0.84% : 0.000004s : 5: predicate.mutable_eliminate 0.28% : 0.000001s : 5: predicate.opt_reshape 0.29% : 0.000002s : 5: predicate.parallel_virtual_node 2.37% : 0.000013s : 58: predicate.partial_defer_inline 1.15% : 0.000006s : 39: predicate.partial_eliminate 0.97% : 0.000005s : 32: predicate.print_const_string_wrapper 0.37% : 0.000002s : 10: predicate.reduce_all_const_elim 1.66% : 0.000009s : 32: predicate.reduce_eliminate 2.42% : 0.000013s : 76: predicate.redundant_stop_gradient_eliminater 0.17% : 0.000001s : 10: predicate.remove_not_recompute_node 0.94% : 0.000005s : 44: predicate.replace_applicator 0.20% : 0.000001s : 10: predicate.replace_old_param 0.20% : 0.000001s : 5: predicate.reset_defer_inline 5.75% : 0.000030s : 32: predicate.reshape_eliminate 0.38% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.22% : 0.000001s : 5: predicate.row_tensor_eliminate 0.46% : 0.000002s : 10: predicate.same_eliminate 0.19% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.47% : 0.000003s : 10: predicate.shard_identity_eliminate 0.41% : 0.000002s : 10: predicate.special_op_eliminate 0.34% : 0.000002s : 10: predicate.specialize_transform 0.48% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.47% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.18% : 0.000001s : 5: predicate.switch_call_monad_eliminater 1.81% : 0.000010s : 58: predicate.switch_defer_inline 2.25% : 0.000012s : 68: predicate.switch_layer_defer_inline 5.90% : 0.000031s : 191: predicate.switch_simplify 0.98% : 0.000005s : 32: predicate.tile_eliminate 1.01% : 0.000005s : 32: predicate.transpose_eliminate 1.51% : 0.000008s : 42: predicate.tuple_list_convert_item_index_to_positive 1.43% : 0.000008s : 42: predicate.tuple_list_get_item_const_eliminator 1.28% : 0.000007s : 42: predicate.tuple_list_get_item_depend_reorder 2.45% : 0.000013s : 54: predicate.tuple_list_get_item_eliminator 1.36% : 0.000007s : 42: predicate.tuple_list_get_set_item_eliminator 2.08% : 0.000011s : 52: predicate.tuple_list_set_item_eliminator 1.47% : 0.000008s : 44: predicate.tuple_to_list_eliminator_ 2.10% : 0.000011s : 76: predicate.updatestate_pure_node_eliminater 2.57% : 0.000014s : 86: predicate.updatestate_useless_node_eliminater 0.18% : 0.000001s : 5: predicate.value_based_eliminate 0.36% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.41% : 0.000002s : 10: predicate.virtual_output_eliminate 0.12% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.25% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003478 42 56.01% : 0.001948s : 16: func_graph_cloner_run.FuncGraphClonerGraph 43.99% : 0.001530s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.932171 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.47% : 0.004374s : 1: add_attr 0.47% : 0.004358s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000079s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.02% : 0.000196s : 1: auto_monad 0.00% : 0.000026s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.24% : 0.002231s : 1: bootstrap 0.00% : 0.000045s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000036s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000011s : 1: environ_conv 0.01% : 0.000089s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000017s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000012s : 1: label_micro_interleaved_index 1.68% : 0.015670s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000874s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000043s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000030s : 1: opt.transform.mutable_eliminate 11.02% : 0.102711s : 78: opt.transform.opt_a 0.01% : 0.000047s : 1: opt.transform.opt_after_cconv 0.00% : 0.000041s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000156s : 28: opt.transform.opt_b 0.01% : 0.000071s : 2: opt.transform.opt_trans_graph 0.01% : 0.000052s : 4: opt.transform.symbol_engine_opt 11.49% : 0.107151s : 1: opt_a 0.02% : 0.000195s : 1: opt_after_cconv 0.09% : 0.000793s : 1: opt_after_jit_grad 0.03% : 0.000305s : 1: opt_b 13.46% : 0.125465s : 1: optimize 0.00% : 0.000028s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000030s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000008s : 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% : 0.000003s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.01% : 0.000097s : 1: pre_auto_parallel 0.00% : 0.000014s : 1: py_interpret_to_execute 0.00% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.01% : 0.000064s : 1: remove_dup_value 0.18% : 0.001683s : 1: renormalize.infer 0.17% : 0.001557s : 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% : 0.000040s : 1: rewriter_after_opt_a 0.05% : 0.000434s : 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% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000099s : 1: symbol_engine_optimizer 0.01% : 0.000103s : 1: tuple_transform 60.36% : 0.562676s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:42:12.530.369 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.440467, [21] [bootstrap]: 0.00047692 [type_inference]: 0.332124 [event_method]: 0.00033372 [auto_monad]: 0.00015618 [graph_reusing]: 1.082e-05 [inline]: 3.09001e-06 [add_attr]: 0.00419608, [1] [add_attr_with_inline]: 0.00418294, [1] [Cycle 1]: 0.00011609, [2] [tag_attr]: 5.236e-05 [meta_addattr_fg_expand]: 9.59e-06 [parallel-infer-symbol]: 4.38001e-06 [pre_auto_parallel]: 6.798e-05 [insert-virtual-dataset]: 2.60997e-06 [parallel-infer-symbol-second]: 1.06002e-06 [dataset_repeat_opt]: 2.21e-06 [pipeline_split]: 1.74e-06 [optimize]: 0.101983, [53] [py_interpret_to_execute]: 9.65002e-06 [rewriter_before_opt_a]: 0.0003261 [opt_a]: 0.0992291, [2] [Cycle 1]: 0.0985616, [45] [expand_dump_flag]: 5.38002e-06 [switch_simplify]: 0.0001726 [loop_unroll]: 5.596e-05 [a_1]: 0.00138623 [with_stream_mark]: 3.065e-05 [recompute_prepare]: 1.502e-05 [updatestate_depend_eliminate]: 4e-06 [updatestate_assign_eliminate]: 3.4e-06 [updatestate_loads_eliminate]: 2.98998e-06 [parameter_eliminate]: 2.83e-06 [a_2]: 7.069e-05 [accelerated_algorithm]: 6.93e-06 [shard]: 1.91e-06 [meta_shard_fg_expand]: 3.74002e-06 [shard_inline]: 6.08002e-06 [merge_send_recv]: 8.62e-06 [auto_parallel]: 9.79999e-06 [parallel]: 2.075e-05 [flash_sp]: 1.037e-05 [merge_comm]: 3.45e-06 [allreduce_fusion]: 3.66999e-06 [matmul_add_comm_reduction]: 9.72999e-06 [allreduce_slice_to_reducescatter]: 1.00999e-06 [virtual_shard_identity]: 1.139e-05 [virtual_dataset]: 6.14999e-06 [get_grad_eliminate_]: 5.68997e-06 [virtual_output]: 6.31998e-06 [merge_forward]: 4.09002e-06 [cell_reuse_recompute_pass]: 2.30002e-06 [offload_activation]: 1.03e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.606e-05 [merge_recompute_call_nodes]: 1.36002e-06 [before_grad]: 1.153e-05 [set_forward_comm_id_for_comm_node_pass]: 3.9e-06 [meta_fg_expand]: 3.45e-06 [flash_sp_send_recv_attached]: 4.04002e-06 [receive_attached]: 3.61999e-06 [after_resolve]: 1.02e-05 [a_after_grad]: 9.39e-06 [renormalize]: 0.0961514 [add_forward_monad_depend]: 1.271e-05 [auto_monad_grad]: 2.83e-06 [auto_monad_eliminator]: 2.437e-05 [cse]: 3.472e-05 [a_3]: 6.05e-05 [Cycle 2]: 0.00065042, [45] [expand_dump_flag]: 2.03997e-06 [switch_simplify]: 8.87999e-06 [loop_unroll]: 5.74e-06 [a_1]: 0.00011019 [with_stream_mark]: 2.132e-05 [recompute_prepare]: 6.41998e-06 [updatestate_depend_eliminate]: 4e-06 [updatestate_assign_eliminate]: 3.18e-06 [updatestate_loads_eliminate]: 3.25e-06 [parameter_eliminate]: 2.56e-06 [a_2]: 5.938e-05 [accelerated_algorithm]: 5.57001e-06 [shard]: 2.59999e-06 [meta_shard_fg_expand]: 2.32001e-06 [shard_inline]: 4.94e-06 [merge_send_recv]: 8.52e-06 [auto_parallel]: 1.084e-05 [parallel]: 9.64e-06 [flash_sp]: 3.93001e-06 [merge_comm]: 3.81999e-06 [allreduce_fusion]: 3.11999e-06 [matmul_add_comm_reduction]: 1.029e-05 [allreduce_slice_to_reducescatter]: 1.20999e-06 [virtual_shard_identity]: 7.16999e-06 [virtual_dataset]: 5.22e-06 [get_grad_eliminate_]: 5.07e-06 [virtual_output]: 4.79e-06 [merge_forward]: 4.4e-06 [cell_reuse_recompute_pass]: 3.11999e-06 [offload_activation]: 9.99001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.815e-05 [merge_recompute_call_nodes]: 2.20002e-06 [before_grad]: 9.52001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.71999e-06 [meta_fg_expand]: 2.39001e-06 [flash_sp_send_recv_attached]: 1.66998e-06 [receive_attached]: 2.24999e-06 [after_resolve]: 1.039e-05 [a_after_grad]: 9.25001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.99999e-06 [auto_monad_grad]: 1.69e-06 [auto_monad_eliminator]: 7.36999e-06 [cse]: 1.727e-05 [a_3]: 3.055e-05 [py_interpret_to_execute_after_opt_a]: 1.289e-05 [slice_cell_reuse_recomputed_activation]: 1.97999e-06 [rewriter_after_opt_a]: 2.177e-05 [convert_after_rewriter]: 1.22e-06 [order_py_execute_after_rewriter]: 1.18001e-06 [mutable_eliminate]: 0.00080164 [opt_b]: 0.00019777, [1] [Cycle 1]: 0.0001884, [7] [b_1]: 0.00010596 [b_2]: 7.63001e-06 [updatestate_depend_eliminate]: 9.94001e-06 [updatestate_assign_eliminate]: 2.46998e-06 [updatestate_loads_eliminate]: 2.56998e-06 [renormalize]: 6.89994e-07 [cse]: 2.46e-05 [optimize_parallel_all_gather_comm]: 2.078e-05 [overlap_param_gather]: 1.99999e-06 [cconv]: 3.566e-05 [loop_unroll]: 0.00048493 [opt_after_cconv]: 9.521e-05, [1] [Cycle 1]: 8.864e-05, [7] [c_1]: 2.445e-05 [parameter_eliminate]: 4.90001e-06 [updatestate_depend_eliminate]: 5.82001e-06 [updatestate_assign_eliminate]: 2.11998e-06 [updatestate_loads_eliminate]: 2.09e-06 [cse]: 1.7e-05 [renormalize]: 4.2998e-07 [remove_dup_value]: 1.443e-05 [tuple_transform]: 7.545e-05, [1] [Cycle 1]: 6.981e-05, [4] [d_1]: 4.172e-05 [none_parameter_eliminate]: 1.55999e-06 [renormalize]: 1.70025e-07 [switch_simplify]: 7.23e-06 [partial_unused_args_eliminate]: 2.03002e-06 [add_recomputation]: 9.199e-05 [cse_after_recomputation]: 2.257e-05, [1] [Cycle 1]: 1.703e-05, [1] [cse]: 1.114e-05 [environ_conv]: 5.49e-06 [swap_dp_allreduce_reducescatter]: 5.63002e-06 [bias_add_comm_swap]: 2.64001e-06 [label_micro_interleaved_index]: 5.52001e-06 [label_fine_grained_interleaved_index]: 2.34001e-06 [merge_cast_opt]: 1.62999e-06 [slice_recompute_activation]: 2.04e-06 [micro_interleaved_order_control]: 2.59999e-06 [assign_add_opt]: 1.55999e-06 [ForceFp32Comm]: 7.89994e-07 [remove_cast_before_assign_add]: 1.15999e-06 [full_micro_interleaved_order_control]: 2.53e-06 [reorder_send_recv_between_fp_bp]: 2.77002e-06 [comm_op_add_attrs]: 1.00999e-06 [add_comm_op_reuse_tag]: 1.10001e-06 [interleave_split_concat_branches]: 1.30999e-06 [interleave_parallel_branches]: 1.09e-06 [overlap_opt_shard_in_pipeline]: 1.59e-06 [overlap_opt_shard_grad_in_pipeline]: 1.88002e-06 [control_data_broadcast_order]: 1.432e-05 [grouped_pairwise_exchange_alltoall]: 1.42e-06 [offloading_packed_experts]: 4.55001e-06 [overlap_recompute_and_grad_model_parallel]: 4.36002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.47999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.21002e-06 [overlap_recompute_comm]: 2.62001e-06 [overlap_grad_ring_attention]: 4.74e-06 [overlap_grad_flash_sp]: 2.543e-05 [begin_end_overlap_inline]: 7.2e-07 [split_matmul_comm_elemetwise]: 2.64001e-06 [split_layernorm_comm]: 1.94999e-06 [handle_group_info]: 1.09e-06 [symbol_engine_optimizer]: 9.273e-05, [1] [Cycle 1]: 8.706e-05, [6] [build]: 5.02999e-06 [elim_shapecalc]: 1.304e-05 [elim_not_effective]: 1.487e-05 [opt_reshape]: 8.02e-06 [fold_const_symbol]: 1.088e-05 [renormalize]: 1.50001e-07 [detach_backward]: 2.34001e-06 [pipeline_parallel_scheduler]: 1.49e-06 [auto_monad_reorder]: 2.102e-05 [get_jit_bprop_graph]: 3.10002e-06 [rewriter_after_jit_bprop_graph]: 8.12e-06 [opt_after_jit_grad]: 0.00078134 [validate]: 5.539e-05 Sums bootstrap : 0.000477s : 0.11% type_inference : 0.332124s : 76.34% event_method : 0.000334s : 0.08% auto_monad : 0.000156s : 0.04% graph_reusing : 0.000011s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000052s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000010s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000068s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000010s : 0.00% optimize.rewriter_before_opt_a : 0.000326s : 0.07% optimize.opt_a.expand_dump_flag : 0.000007s : 0.00% optimize.opt_a.switch_simplify : 0.000181s : 0.04% optimize.opt_a.loop_unroll : 0.000062s : 0.01% optimize.opt_a.a_1 : 0.001496s : 0.34% optimize.opt_a.with_stream_mark : 0.000052s : 0.01% optimize.opt_a.recompute_prepare : 0.000021s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000130s : 0.03% optimize.opt_a.accelerated_algorithm : 0.000013s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000006s : 0.00% optimize.opt_a.shard_inline : 0.000011s : 0.00% optimize.opt_a.merge_send_recv : 0.000017s : 0.00% optimize.opt_a.auto_parallel : 0.000021s : 0.00% optimize.opt_a.parallel : 0.000030s : 0.01% optimize.opt_a.flash_sp : 0.000014s : 0.00% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000007s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000019s : 0.00% optimize.opt_a.virtual_dataset : 0.000011s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.00% optimize.opt_a.virtual_output : 0.000011s : 0.00% optimize.opt_a.merge_forward : 0.000008s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000020s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000034s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000021s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.00% optimize.opt_a.meta_fg_expand : 0.000006s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000006s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000021s : 0.00% optimize.opt_a.a_after_grad : 0.000019s : 0.00% optimize.opt_a.renormalize : 0.096151s : 22.10% optimize.opt_a.add_forward_monad_depend : 0.000015s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000032s : 0.01% optimize.opt_a.cse : 0.000052s : 0.01% optimize.opt_a.a_3 : 0.000091s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000013s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000022s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000802s : 0.18% optimize.opt_b.b_1 : 0.000106s : 0.02% optimize.opt_b.b_2 : 0.000008s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000025s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.01% optimize.loop_unroll : 0.000485s : 0.11% optimize.opt_after_cconv.c_1 : 0.000024s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.00% optimize.tuple_transform.d_1 : 0.000042s : 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 : 0.000007s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000092s : 0.02% optimize.cse_after_recomputation.cse : 0.000011s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 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 : 0.000002s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000002s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000014s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000025s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000013s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000015s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000008s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000011s : 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 : 0.000021s : 0.00% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000781s : 0.18% validate : 0.000055s : 0.01% Time group info: ------[substitution.] 0.000428 39 3.54% : 0.000015s : 1: substitution.cast_eliminate 0.43% : 0.000002s : 2: substitution.elim_not_effective 0.41% : 0.000002s : 2: substitution.fold_const_symbol 1.43% : 0.000006s : 3: substitution.graph_param_transform 84.46% : 0.000361s : 15: substitution.inline 1.14% : 0.000005s : 4: substitution.j_node_and_user_rematch 1.32% : 0.000006s : 4: substitution.remove_not_recompute_node 1.31% : 0.000006s : 2: substitution.replace_old_param 3.79% : 0.000016s : 4: substitution.switch_simplify 2.17% : 0.000009s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.332008 2 98.88% : 0.328289s : 1: type_inference.infer 1.12% : 0.003719s : 1: type_inference.specialize ------[replace.] 0.000197 22 4.03% : 0.000008s : 1: replace.cast_eliminate 58.04% : 0.000114s : 15: replace.inline 29.97% : 0.000059s : 4: replace.switch_simplify 7.96% : 0.000016s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000387 22 3.46% : 0.000013s : 1: match.cast_eliminate 91.08% : 0.000352s : 15: match.inline 3.36% : 0.000013s : 4: match.switch_simplify 2.11% : 0.000008s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000302 1716 1.22% : 0.000004s : 22: predicate.accumulaten_eliminater 0.99% : 0.000003s : 3: predicate.ad_related_special_op_eliminate 0.35% : 0.000001s : 6: predicate.addn_check_dump 1.10% : 0.000003s : 22: predicate.addn_zero_filter 0.95% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.61% : 0.000008s : 28: predicate.arithmetic_simplify 1.59% : 0.000005s : 23: predicate.cast_eliminate 0.36% : 0.000001s : 6: predicate.check_bprop_eliminate 0.32% : 0.000001s : 6: predicate.compare_switch_simplify 0.10% : 0.000000s : 3: predicate.const_output_eliminate 0.36% : 0.000001s : 6: predicate.depend_value_elim 1.08% : 0.000003s : 23: predicate.dict_get_item_const_eliminator 1.21% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.13% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.80% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.19% : 0.000001s : 3: predicate.elim_not_effective 0.46% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.31% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.25% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.14% : 0.000003s : 26: predicate.environ_get_depend_swap 1.56% : 0.000005s : 32: predicate.environ_get_eliminate 1.16% : 0.000003s : 26: predicate.environ_get_set_eliminate 1.94% : 0.000006s : 40: predicate.exchange_switch_depend_value 2.86% : 0.000009s : 40: predicate.float_depend_g_call 0.47% : 0.000001s : 6: predicate.float_environ_get_switch 0.41% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.09% : 0.000000s : 3: predicate.fold_const_symbol 0.48% : 0.000001s : 6: predicate.get_grad_eliminate 0.17% : 0.000001s : 3: predicate.graph_param_transform 0.31% : 0.000001s : 6: predicate.incorporate_call 0.25% : 0.000001s : 6: predicate.incorporate_call_switch 5.94% : 0.000018s : 80: predicate.inline 0.47% : 0.000001s : 6: predicate.inline_without_move 0.17% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.62% : 0.000002s : 6: predicate.less_batch_normalization 1.73% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.37% : 0.000007s : 53: predicate.load_eliminater 0.64% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.15% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.44% : 0.000004s : 29: predicate.make_slice_get_slice_eliminator 0.42% : 0.000001s : 6: predicate.merge_addn 0.46% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.32% : 0.000001s : 6: predicate.mini_step_allgather_replace 0.99% : 0.000003s : 22: predicate.minmaximum_grad 1.29% : 0.000004s : 3: predicate.mutable_eliminate 0.25% : 0.000001s : 3: predicate.opt_reshape 0.25% : 0.000001s : 3: predicate.parallel_virtual_node 2.68% : 0.000008s : 40: predicate.partial_defer_inline 1.42% : 0.000004s : 28: predicate.partial_eliminate 1.07% : 0.000003s : 22: predicate.print_const_string_wrapper 0.38% : 0.000001s : 6: predicate.reduce_all_const_elim 1.44% : 0.000004s : 22: predicate.reduce_eliminate 2.41% : 0.000007s : 53: predicate.redundant_stop_gradient_eliminater 0.55% : 0.000002s : 6: predicate.remove_not_recompute_node 1.21% : 0.000004s : 31: predicate.replace_applicator 0.30% : 0.000001s : 6: predicate.replace_old_param 0.28% : 0.000001s : 3: predicate.reset_defer_inline 1.11% : 0.000003s : 22: predicate.reshape_eliminate 0.38% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.25% : 0.000001s : 3: predicate.row_tensor_eliminate 0.55% : 0.000002s : 6: predicate.same_eliminate 0.45% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.70% : 0.000002s : 6: predicate.shard_identity_eliminate 0.50% : 0.000001s : 6: predicate.special_op_eliminate 0.44% : 0.000001s : 6: predicate.specialize_transform 0.83% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.48% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.16% : 0.000000s : 3: predicate.switch_call_monad_eliminater 2.04% : 0.000006s : 40: predicate.switch_defer_inline 2.38% : 0.000007s : 46: predicate.switch_layer_defer_inline 6.87% : 0.000021s : 126: predicate.switch_simplify 1.15% : 0.000003s : 22: predicate.tile_eliminate 1.10% : 0.000003s : 22: predicate.transpose_eliminate 1.77% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.52% : 0.000005s : 29: predicate.tuple_list_get_item_const_eliminator 1.35% : 0.000004s : 29: predicate.tuple_list_get_item_depend_reorder 3.10% : 0.000009s : 37: predicate.tuple_list_get_item_eliminator 1.56% : 0.000005s : 29: predicate.tuple_list_get_set_item_eliminator 2.33% : 0.000007s : 35: predicate.tuple_list_set_item_eliminator 1.79% : 0.000005s : 31: predicate.tuple_to_list_eliminator_ 2.46% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 2.84% : 0.000009s : 59: predicate.updatestate_useless_node_eliminater 0.28% : 0.000001s : 3: predicate.value_based_eliminate 0.39% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.36% : 0.000001s : 6: predicate.virtual_output_eliminate 0.17% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.24% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002253 27 45.18% : 0.001018s : 10: func_graph_cloner_run.FuncGraphClonerGraph 54.82% : 0.001235s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.644865 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.65% : 0.004202s : 1: add_attr 0.65% : 0.004187s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000098s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.03% : 0.000168s : 1: auto_monad 0.00% : 0.000026s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.08% : 0.000519s : 1: bootstrap 0.01% : 0.000040s : 1: cconv 0.00% : 0.000003s : 1: comm_op_add_attrs 0.00% : 0.000018s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000025s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.05% : 0.000350s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000015s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.08% : 0.000494s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.13% : 0.000819s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000024s : 1: opt.transform.mutable_eliminate 0.32% : 0.002053s : 78: opt.transform.opt_a 0.00% : 0.000023s : 1: opt.transform.opt_after_cconv 0.01% : 0.000035s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000085s : 28: opt.transform.opt_b 0.01% : 0.000047s : 2: opt.transform.opt_trans_graph 0.01% : 0.000043s : 4: opt.transform.symbol_engine_opt 15.39% : 0.099233s : 1: opt_a 0.02% : 0.000098s : 1: opt_after_cconv 0.12% : 0.000803s : 1: opt_after_jit_grad 0.03% : 0.000201s : 1: opt_b 15.82% : 0.101989s : 1: optimize 0.00% : 0.000025s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000031s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.01% : 0.000073s : 1: pre_auto_parallel 0.00% : 0.000014s : 1: py_interpret_to_execute 0.00% : 0.000016s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000018s : 1: remove_dup_value 14.72% : 0.094955s : 1: renormalize.infer 0.18% : 0.001176s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000012s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000025s : 1: rewriter_after_opt_a 0.05% : 0.000338s : 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% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000097s : 1: symbol_engine_optimizer 0.01% : 0.000078s : 1: tuple_transform 51.51% : 0.332163s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:42:15.252.172 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.532168, [21] [bootstrap]: 0.00053234 [type_inference]: 0.422613 [event_method]: 0.0003022 [auto_monad]: 0.00014973 [graph_reusing]: 1.03e-05 [inline]: 2.76999e-06 [add_attr]: 0.00414173, [1] [add_attr_with_inline]: 0.00412529, [1] [Cycle 1]: 0.0001111, [2] [tag_attr]: 5.253e-05 [meta_addattr_fg_expand]: 9.46e-06 [parallel-infer-symbol]: 3.77002e-06 [pre_auto_parallel]: 6.826e-05 [insert-virtual-dataset]: 2.51e-06 [parallel-infer-symbol-second]: 6.19999e-07 [dataset_repeat_opt]: 1.67001e-06 [pipeline_split]: 1.94999e-06 [optimize]: 0.103401, [53] [py_interpret_to_execute]: 1.133e-05 [rewriter_before_opt_a]: 0.00032655 [opt_a]: 0.100614, [2] [Cycle 1]: 0.0999732, [45] [expand_dump_flag]: 4.87998e-06 [switch_simplify]: 0.00017477 [loop_unroll]: 5.539e-05 [a_1]: 0.00125624 [with_stream_mark]: 2.796e-05 [recompute_prepare]: 1.195e-05 [updatestate_depend_eliminate]: 4.59002e-06 [updatestate_assign_eliminate]: 3.13998e-06 [updatestate_loads_eliminate]: 3.2e-06 [parameter_eliminate]: 2.73e-06 [a_2]: 6.832e-05 [accelerated_algorithm]: 6.64001e-06 [shard]: 1.89e-06 [meta_shard_fg_expand]: 2.75997e-06 [shard_inline]: 5.27001e-06 [merge_send_recv]: 7.72998e-06 [auto_parallel]: 9.87001e-06 [parallel]: 2.005e-05 [flash_sp]: 1.188e-05 [merge_comm]: 3.66999e-06 [allreduce_fusion]: 3.16999e-06 [matmul_add_comm_reduction]: 1.042e-05 [allreduce_slice_to_reducescatter]: 1.25001e-06 [virtual_shard_identity]: 8.87999e-06 [virtual_dataset]: 5.94e-06 [get_grad_eliminate_]: 5.25001e-06 [virtual_output]: 5.55001e-06 [merge_forward]: 4.17e-06 [cell_reuse_recompute_pass]: 1.77001e-06 [offload_activation]: 1.093e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.598e-05 [merge_recompute_call_nodes]: 1.65001e-06 [before_grad]: 1.01e-05 [set_forward_comm_id_for_comm_node_pass]: 3.28e-06 [meta_fg_expand]: 3.54002e-06 [flash_sp_send_recv_attached]: 2.68e-06 [receive_attached]: 2.34999e-06 [after_resolve]: 9.31e-06 [a_after_grad]: 9.27999e-06 [renormalize]: 0.097745 [add_forward_monad_depend]: 1.115e-05 [auto_monad_grad]: 2.64001e-06 [auto_monad_eliminator]: 2.133e-05 [cse]: 3.139e-05 [a_3]: 5.867e-05 [Cycle 2]: 0.00062462, [45] [expand_dump_flag]: 3.21001e-06 [switch_simplify]: 8.43999e-06 [loop_unroll]: 5.30001e-06 [a_1]: 0.0001063 [with_stream_mark]: 1.858e-05 [recompute_prepare]: 5.94999e-06 [updatestate_depend_eliminate]: 3.55003e-06 [updatestate_assign_eliminate]: 3.11999e-06 [updatestate_loads_eliminate]: 3.01999e-06 [parameter_eliminate]: 1.94999e-06 [a_2]: 5.956e-05 [accelerated_algorithm]: 5.77999e-06 [shard]: 2.35002e-06 [meta_shard_fg_expand]: 2.22001e-06 [shard_inline]: 4.65001e-06 [merge_send_recv]: 8.23001e-06 [auto_parallel]: 8.89e-06 [parallel]: 9.81e-06 [flash_sp]: 3.61001e-06 [merge_comm]: 2.94999e-06 [allreduce_fusion]: 3.72002e-06 [matmul_add_comm_reduction]: 9.29e-06 [allreduce_slice_to_reducescatter]: 9.00007e-07 [virtual_shard_identity]: 7.31999e-06 [virtual_dataset]: 5.30999e-06 [get_grad_eliminate_]: 4.67e-06 [virtual_output]: 5.06002e-06 [merge_forward]: 4.55001e-06 [cell_reuse_recompute_pass]: 3.48999e-06 [offload_activation]: 1.101e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.647e-05 [merge_recompute_call_nodes]: 1.47001e-06 [before_grad]: 8.97e-06 [set_forward_comm_id_for_comm_node_pass]: 3.35998e-06 [meta_fg_expand]: 2.73e-06 [flash_sp_send_recv_attached]: 1.76e-06 [receive_attached]: 2.26998e-06 [after_resolve]: 1.048e-05 [a_after_grad]: 8.65001e-06 [renormalize]: 6.00121e-08 [add_forward_monad_depend]: 1.26002e-06 [auto_monad_grad]: 1.42e-06 [auto_monad_eliminator]: 6.46999e-06 [cse]: 1.381e-05 [a_3]: 2.899e-05 [py_interpret_to_execute_after_opt_a]: 1.022e-05 [slice_cell_reuse_recomputed_activation]: 2.31e-06 [rewriter_after_opt_a]: 2.065e-05 [convert_after_rewriter]: 1.17e-06 [order_py_execute_after_rewriter]: 1.10001e-06 [mutable_eliminate]: 0.00079697 [opt_b]: 0.00021155, [1] [Cycle 1]: 0.00020237, [7] [b_1]: 0.00011641 [b_2]: 6.59999e-06 [updatestate_depend_eliminate]: 1.025e-05 [updatestate_assign_eliminate]: 2.79001e-06 [updatestate_loads_eliminate]: 2.12999e-06 [renormalize]: 5.99975e-07 [cse]: 2.546e-05 [optimize_parallel_all_gather_comm]: 2.055e-05 [overlap_param_gather]: 2.17001e-06 [cconv]: 3.674e-05 [loop_unroll]: 0.0005264 [opt_after_cconv]: 0.00010478, [1] [Cycle 1]: 9.69e-05, [7] [c_1]: 2.57e-05 [parameter_eliminate]: 5.47001e-06 [updatestate_depend_eliminate]: 7.27002e-06 [updatestate_assign_eliminate]: 2.27001e-06 [updatestate_loads_eliminate]: 2.01998e-06 [cse]: 2.048e-05 [renormalize]: 8.00006e-07 [remove_dup_value]: 1.522e-05 [tuple_transform]: 0.000108, [1] [Cycle 1]: 0.00010321, [4] [d_1]: 4.056e-05 [none_parameter_eliminate]: 1.64e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 7.15998e-06 [partial_unused_args_eliminate]: 2.04e-06 [add_recomputation]: 5.18e-05 [cse_after_recomputation]: 2.251e-05, [1] [Cycle 1]: 1.688e-05, [1] [cse]: 1.062e-05 [environ_conv]: 5.54e-06 [swap_dp_allreduce_reducescatter]: 5.24e-06 [bias_add_comm_swap]: 3.5e-06 [label_micro_interleaved_index]: 5.25999e-06 [label_fine_grained_interleaved_index]: 2.71e-06 [merge_cast_opt]: 1.57999e-06 [slice_recompute_activation]: 2.44001e-06 [micro_interleaved_order_control]: 2.43998e-06 [assign_add_opt]: 1.64e-06 [ForceFp32Comm]: 9.29984e-07 [remove_cast_before_assign_add]: 1.06002e-06 [full_micro_interleaved_order_control]: 2.47001e-06 [reorder_send_recv_between_fp_bp]: 2.95998e-06 [comm_op_add_attrs]: 1.09998e-06 [add_comm_op_reuse_tag]: 9.50007e-07 [interleave_split_concat_branches]: 1.17999e-06 [interleave_parallel_branches]: 1.75001e-06 [overlap_opt_shard_in_pipeline]: 1.75001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.84998e-06 [control_data_broadcast_order]: 1.472e-05 [grouped_pairwise_exchange_alltoall]: 1.32e-06 [offloading_packed_experts]: 3.88999e-06 [overlap_recompute_and_grad_model_parallel]: 4.92e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.14e-06 [overlap_recompute_allgather_and_fa_grad]: 1.30999e-06 [overlap_recompute_comm]: 2.44999e-06 [overlap_grad_ring_attention]: 4.1e-06 [overlap_grad_flash_sp]: 2.259e-05 [begin_end_overlap_inline]: 4.90021e-07 [split_matmul_comm_elemetwise]: 1.92999e-06 [split_layernorm_comm]: 1.64998e-06 [handle_group_info]: 1.09e-06 [symbol_engine_optimizer]: 8.608e-05, [1] [Cycle 1]: 8.081e-05, [6] [build]: 4.03999e-06 [elim_shapecalc]: 1.442e-05 [elim_not_effective]: 1.561e-05 [opt_reshape]: 7.08e-06 [fold_const_symbol]: 9.47999e-06 [renormalize]: 1.80007e-07 [detach_backward]: 2.52001e-06 [pipeline_parallel_scheduler]: 1.64e-06 [auto_monad_reorder]: 1.892e-05 [get_jit_bprop_graph]: 2.22001e-06 [rewriter_after_jit_bprop_graph]: 7.25e-06 [opt_after_jit_grad]: 0.00064283 [validate]: 4.656e-05 Sums bootstrap : 0.000532s : 0.10% type_inference : 0.422613s : 80.22% event_method : 0.000302s : 0.06% auto_monad : 0.000150s : 0.03% graph_reusing : 0.000010s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000053s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000009s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000068s : 0.01% 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 : 0.000011s : 0.00% optimize.rewriter_before_opt_a : 0.000327s : 0.06% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000183s : 0.03% optimize.opt_a.loop_unroll : 0.000061s : 0.01% optimize.opt_a.a_1 : 0.001363s : 0.26% optimize.opt_a.with_stream_mark : 0.000047s : 0.01% optimize.opt_a.recompute_prepare : 0.000018s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000128s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000012s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.00% optimize.opt_a.shard_inline : 0.000010s : 0.00% optimize.opt_a.merge_send_recv : 0.000016s : 0.00% optimize.opt_a.auto_parallel : 0.000019s : 0.00% optimize.opt_a.parallel : 0.000030s : 0.01% optimize.opt_a.flash_sp : 0.000015s : 0.00% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000007s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000016s : 0.00% optimize.opt_a.virtual_dataset : 0.000011s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000010s : 0.00% optimize.opt_a.virtual_output : 0.000011s : 0.00% optimize.opt_a.merge_forward : 0.000009s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000022s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000032s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000019s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.00% optimize.opt_a.meta_fg_expand : 0.000006s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000020s : 0.00% optimize.opt_a.a_after_grad : 0.000018s : 0.00% optimize.opt_a.renormalize : 0.097745s : 18.55% optimize.opt_a.add_forward_monad_depend : 0.000012s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000028s : 0.01% optimize.opt_a.cse : 0.000045s : 0.01% optimize.opt_a.a_3 : 0.000088s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000010s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000797s : 0.15% optimize.opt_b.b_1 : 0.000116s : 0.02% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000025s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000037s : 0.01% optimize.loop_unroll : 0.000526s : 0.10% optimize.opt_after_cconv.c_1 : 0.000026s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000020s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000041s : 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 : 0.000007s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000052s : 0.01% optimize.cse_after_recomputation.cse : 0.000011s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000002s : 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 : 0.000015s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000023s : 0.00% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000016s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000019s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000643s : 0.12% validate : 0.000047s : 0.01% Time group info: ------[substitution.] 0.000389 39 3.50% : 0.000014s : 1: substitution.cast_eliminate 0.61% : 0.000002s : 2: substitution.elim_not_effective 0.38% : 0.000001s : 2: substitution.fold_const_symbol 1.40% : 0.000005s : 3: substitution.graph_param_transform 83.86% : 0.000326s : 15: substitution.inline 1.12% : 0.000004s : 4: substitution.j_node_and_user_rematch 1.49% : 0.000006s : 4: substitution.remove_not_recompute_node 1.27% : 0.000005s : 2: substitution.replace_old_param 3.77% : 0.000015s : 4: substitution.switch_simplify 2.61% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.422499 2 99.20% : 0.419135s : 1: type_inference.infer 0.80% : 0.003364s : 1: type_inference.specialize ------[replace.] 0.000187 22 3.72% : 0.000007s : 1: replace.cast_eliminate 55.26% : 0.000103s : 15: replace.inline 32.04% : 0.000060s : 4: replace.switch_simplify 8.99% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000351 22 3.57% : 0.000013s : 1: match.cast_eliminate 90.51% : 0.000318s : 15: match.inline 3.37% : 0.000012s : 4: match.switch_simplify 2.56% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000291 1716 1.27% : 0.000004s : 22: predicate.accumulaten_eliminater 0.79% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.32% : 0.000001s : 6: predicate.addn_check_dump 1.13% : 0.000003s : 22: predicate.addn_zero_filter 1.12% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.46% : 0.000007s : 28: predicate.arithmetic_simplify 1.41% : 0.000004s : 23: predicate.cast_eliminate 0.34% : 0.000001s : 6: predicate.check_bprop_eliminate 0.40% : 0.000001s : 6: predicate.compare_switch_simplify 0.11% : 0.000000s : 3: predicate.const_output_eliminate 0.36% : 0.000001s : 6: predicate.depend_value_elim 1.18% : 0.000003s : 23: predicate.dict_get_item_const_eliminator 1.28% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.18% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.94% : 0.000003s : 6: predicate.dumpgradient_eliminate 0.19% : 0.000001s : 3: predicate.elim_not_effective 0.25% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.34% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.25% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.21% : 0.000004s : 26: predicate.environ_get_depend_swap 1.61% : 0.000005s : 32: predicate.environ_get_eliminate 1.20% : 0.000003s : 26: predicate.environ_get_set_eliminate 1.97% : 0.000006s : 40: predicate.exchange_switch_depend_value 2.95% : 0.000009s : 40: predicate.float_depend_g_call 0.31% : 0.000001s : 6: predicate.float_environ_get_switch 0.46% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.10% : 0.000000s : 3: predicate.fold_const_symbol 0.40% : 0.000001s : 6: predicate.get_grad_eliminate 0.17% : 0.000000s : 3: predicate.graph_param_transform 0.37% : 0.000001s : 6: predicate.incorporate_call 0.27% : 0.000001s : 6: predicate.incorporate_call_switch 5.85% : 0.000017s : 80: predicate.inline 0.64% : 0.000002s : 6: predicate.inline_without_move 0.16% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.67% : 0.000002s : 6: predicate.less_batch_normalization 1.70% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.42% : 0.000007s : 53: predicate.load_eliminater 0.70% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.40% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.60% : 0.000005s : 29: predicate.make_slice_get_slice_eliminator 0.30% : 0.000001s : 6: predicate.merge_addn 0.29% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.29% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.04% : 0.000003s : 22: predicate.minmaximum_grad 1.14% : 0.000003s : 3: predicate.mutable_eliminate 0.25% : 0.000001s : 3: predicate.opt_reshape 0.28% : 0.000001s : 3: predicate.parallel_virtual_node 2.82% : 0.000008s : 40: predicate.partial_defer_inline 1.47% : 0.000004s : 28: predicate.partial_eliminate 1.08% : 0.000003s : 22: predicate.print_const_string_wrapper 0.34% : 0.000001s : 6: predicate.reduce_all_const_elim 1.34% : 0.000004s : 22: predicate.reduce_eliminate 2.62% : 0.000008s : 53: predicate.redundant_stop_gradient_eliminater 0.34% : 0.000001s : 6: predicate.remove_not_recompute_node 1.32% : 0.000004s : 31: predicate.replace_applicator 0.22% : 0.000001s : 6: predicate.replace_old_param 0.28% : 0.000001s : 3: predicate.reset_defer_inline 1.20% : 0.000003s : 22: predicate.reshape_eliminate 0.44% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.26% : 0.000001s : 3: predicate.row_tensor_eliminate 0.50% : 0.000001s : 6: predicate.same_eliminate 0.28% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.57% : 0.000002s : 6: predicate.shard_identity_eliminate 0.41% : 0.000001s : 6: predicate.special_op_eliminate 0.38% : 0.000001s : 6: predicate.specialize_transform 0.79% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.56% : 0.000002s : 6: predicate.stack_unstack_eliminate 0.23% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.15% : 0.000006s : 40: predicate.switch_defer_inline 2.47% : 0.000007s : 46: predicate.switch_layer_defer_inline 6.83% : 0.000020s : 126: predicate.switch_simplify 1.11% : 0.000003s : 22: predicate.tile_eliminate 1.14% : 0.000003s : 22: predicate.transpose_eliminate 1.64% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.55% : 0.000004s : 29: predicate.tuple_list_get_item_const_eliminator 1.69% : 0.000005s : 29: predicate.tuple_list_get_item_depend_reorder 2.74% : 0.000008s : 37: predicate.tuple_list_get_item_eliminator 1.56% : 0.000005s : 29: predicate.tuple_list_get_set_item_eliminator 2.34% : 0.000007s : 35: predicate.tuple_list_set_item_eliminator 1.56% : 0.000005s : 31: predicate.tuple_to_list_eliminator_ 2.35% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 2.79% : 0.000008s : 59: predicate.updatestate_useless_node_eliminater 0.21% : 0.000001s : 3: predicate.value_based_eliminate 0.61% : 0.000002s : 6: predicate.virtual_dataset_eliminate 0.42% : 0.000001s : 6: predicate.virtual_output_eliminate 0.14% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.21% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002143 27 51.20% : 0.001097s : 10: func_graph_cloner_run.FuncGraphClonerGraph 48.80% : 0.001046s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.739467 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.56% : 0.004149s : 1: add_attr 0.56% : 0.004129s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000057s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.02% : 0.000160s : 1: auto_monad 0.00% : 0.000023s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.08% : 0.000578s : 1: bootstrap 0.01% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000018s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000025s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.04% : 0.000317s : 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% : 0.000005s : 1: interleave_parallel_branches 0.01% : 0.000079s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.07% : 0.000538s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.11% : 0.000811s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000015s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000023s : 1: opt.transform.mutable_eliminate 0.26% : 0.001902s : 78: opt.transform.opt_a 0.00% : 0.000024s : 1: opt.transform.opt_after_cconv 0.00% : 0.000026s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000094s : 28: opt.transform.opt_b 0.01% : 0.000046s : 2: opt.transform.opt_trans_graph 0.01% : 0.000042s : 4: opt.transform.symbol_engine_opt 13.61% : 0.100618s : 1: opt_a 0.01% : 0.000109s : 1: opt_after_cconv 0.09% : 0.000660s : 1: opt_after_jit_grad 0.03% : 0.000215s : 1: opt_b 13.98% : 0.103407s : 1: optimize 0.00% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000026s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 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% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000003s : 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 0.01% : 0.000074s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000013s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000019s : 1: remove_dup_value 13.08% : 0.096696s : 1: renormalize.infer 0.14% : 0.001034s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000024s : 1: rewriter_after_opt_a 0.05% : 0.000336s : 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% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000089s : 1: symbol_engine_optimizer 0.01% : 0.000111s : 1: tuple_transform 57.16% : 0.422645s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:42:17.987.440 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.918699, [21] [bootstrap]: 0.00048339 [type_inference]: 0.746186 [event_method]: 9.657e-05 [auto_monad]: 0.00021346 [graph_reusing]: 1.383e-05 [inline]: 3.33e-06 [add_attr]: 0.00471272, [1] [add_attr_with_inline]: 0.00469741, [1] [Cycle 1]: 0.0001464, [2] [tag_attr]: 7.346e-05 [meta_addattr_fg_expand]: 1.561e-05 [parallel-infer-symbol]: 4.07e-06 [pre_auto_parallel]: 9.636e-05 [insert-virtual-dataset]: 2.63e-06 [parallel-infer-symbol-second]: 8.30012e-07 [dataset_repeat_opt]: 2.26e-06 [pipeline_split]: 1.94e-06 [optimize]: 0.166035, [53] [py_interpret_to_execute]: 9.36e-06 [rewriter_before_opt_a]: 0.00046138 [opt_a]: 0.162925, [2] [Cycle 1]: 0.162048, [45] [expand_dump_flag]: 1.122e-05 [switch_simplify]: 0.00028316 [loop_unroll]: 8.897e-05 [a_1]: 0.00201825 [with_stream_mark]: 3.236e-05 [recompute_prepare]: 1.847e-05 [updatestate_depend_eliminate]: 6.66999e-06 [updatestate_assign_eliminate]: 4.56002e-06 [updatestate_loads_eliminate]: 4.68001e-06 [parameter_eliminate]: 2.63998e-06 [a_2]: 0.00011508 [accelerated_algorithm]: 9.37001e-06 [shard]: 2.22001e-06 [meta_shard_fg_expand]: 6.67002e-06 [shard_inline]: 8.55001e-06 [merge_send_recv]: 1.184e-05 [auto_parallel]: 1.136e-05 [parallel]: 2.337e-05 [flash_sp]: 1.352e-05 [merge_comm]: 6.45002e-06 [allreduce_fusion]: 5.04e-06 [matmul_add_comm_reduction]: 1.286e-05 [allreduce_slice_to_reducescatter]: 9.20001e-07 [virtual_shard_identity]: 1.17e-05 [virtual_dataset]: 9.29e-06 [get_grad_eliminate_]: 8.90999e-06 [virtual_output]: 7.95e-06 [merge_forward]: 5.57001e-06 [cell_reuse_recompute_pass]: 2.69999e-06 [offload_activation]: 1.353e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.146e-05 [merge_recompute_call_nodes]: 1.71998e-06 [before_grad]: 1.482e-05 [set_forward_comm_id_for_comm_node_pass]: 5.42001e-06 [meta_fg_expand]: 5.86e-06 [flash_sp_send_recv_attached]: 2.98e-06 [receive_attached]: 2.66999e-06 [after_resolve]: 1.341e-05 [a_after_grad]: 1.409e-05 [renormalize]: 0.158676 [add_forward_monad_depend]: 1.304e-05 [auto_monad_grad]: 2.94999e-06 [auto_monad_eliminator]: 3.097e-05 [cse]: 5.312e-05 [a_3]: 7.749e-05 [Cycle 2]: 0.00085638, [45] [expand_dump_flag]: 2.74001e-06 [switch_simplify]: 1.129e-05 [loop_unroll]: 7.85998e-06 [a_1]: 0.00020306 [with_stream_mark]: 2.11e-05 [recompute_prepare]: 9.12001e-06 [updatestate_depend_eliminate]: 5.24e-06 [updatestate_assign_eliminate]: 4.23001e-06 [updatestate_loads_eliminate]: 4.34997e-06 [parameter_eliminate]: 1.94999e-06 [a_2]: 9.732e-05 [accelerated_algorithm]: 8.13999e-06 [shard]: 2.78e-06 [meta_shard_fg_expand]: 2.79001e-06 [shard_inline]: 7.61999e-06 [merge_send_recv]: 1.146e-05 [auto_parallel]: 1.064e-05 [parallel]: 9.62999e-06 [flash_sp]: 5.17e-06 [merge_comm]: 4.58999e-06 [allreduce_fusion]: 4.65001e-06 [matmul_add_comm_reduction]: 1.188e-05 [allreduce_slice_to_reducescatter]: 1.04e-06 [virtual_shard_identity]: 9.41e-06 [virtual_dataset]: 7.66999e-06 [get_grad_eliminate_]: 7.03998e-06 [virtual_output]: 6.92002e-06 [merge_forward]: 5.88998e-06 [cell_reuse_recompute_pass]: 3.70998e-06 [offload_activation]: 1.35e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.635e-05 [merge_recompute_call_nodes]: 1.32e-06 [before_grad]: 1.286e-05 [set_forward_comm_id_for_comm_node_pass]: 5.46998e-06 [meta_fg_expand]: 3.71999e-06 [flash_sp_send_recv_attached]: 1.62999e-06 [receive_attached]: 2.64001e-06 [after_resolve]: 1.284e-05 [a_after_grad]: 1.135e-05 [renormalize]: 7.00238e-08 [add_forward_monad_depend]: 1.86e-06 [auto_monad_grad]: 1.92999e-06 [auto_monad_eliminator]: 1.13e-05 [cse]: 2.767e-05 [a_3]: 4.666e-05 [py_interpret_to_execute_after_opt_a]: 1.359e-05 [slice_cell_reuse_recomputed_activation]: 2.24999e-06 [rewriter_after_opt_a]: 2.876e-05 [convert_after_rewriter]: 1.25999e-06 [order_py_execute_after_rewriter]: 1.45999e-06 [mutable_eliminate]: 0.00080221 [opt_b]: 0.00034541, [1] [Cycle 1]: 0.00033651, [7] [b_1]: 0.00022557 [b_2]: 1.1e-05 [updatestate_depend_eliminate]: 1.162e-05 [updatestate_assign_eliminate]: 4.54998e-06 [updatestate_loads_eliminate]: 3.39001e-06 [renormalize]: 7.7e-07 [cse]: 3.998e-05 [optimize_parallel_all_gather_comm]: 2.386e-05 [overlap_param_gather]: 1.91998e-06 [cconv]: 3.71e-05 [loop_unroll]: 0.00053824 [opt_after_cconv]: 0.00013417, [1] [Cycle 1]: 0.00012638, [7] [c_1]: 3.631e-05 [parameter_eliminate]: 5.47999e-06 [updatestate_depend_eliminate]: 8.87e-06 [updatestate_assign_eliminate]: 3.85e-06 [updatestate_loads_eliminate]: 3.52002e-06 [cse]: 3.389e-05 [renormalize]: 5.09986e-07 [remove_dup_value]: 5.631e-05 [tuple_transform]: 9.122e-05, [1] [Cycle 1]: 8.663e-05, [4] [d_1]: 5.895e-05 [none_parameter_eliminate]: 1.73002e-06 [renormalize]: 1.69995e-07 [switch_simplify]: 7.8e-06 [partial_unused_args_eliminate]: 1.71e-06 [add_recomputation]: 6.737e-05 [cse_after_recomputation]: 2.807e-05, [1] [Cycle 1]: 2.243e-05, [1] [cse]: 1.697e-05 [environ_conv]: 7.05e-06 [swap_dp_allreduce_reducescatter]: 6.99001e-06 [bias_add_comm_swap]: 2.78e-06 [label_micro_interleaved_index]: 6.51e-06 [label_fine_grained_interleaved_index]: 2.64001e-06 [merge_cast_opt]: 1.39e-06 [slice_recompute_activation]: 2.21e-06 [micro_interleaved_order_control]: 2.16998e-06 [assign_add_opt]: 1.30999e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 9.49978e-07 [full_micro_interleaved_order_control]: 2.04e-06 [reorder_send_recv_between_fp_bp]: 2.61e-06 [comm_op_add_attrs]: 1.02e-06 [add_comm_op_reuse_tag]: 1.22e-06 [interleave_split_concat_branches]: 1.13001e-06 [interleave_parallel_branches]: 1.05999e-06 [overlap_opt_shard_in_pipeline]: 1.19998e-06 [overlap_opt_shard_grad_in_pipeline]: 2.06998e-06 [control_data_broadcast_order]: 1.618e-05 [grouped_pairwise_exchange_alltoall]: 1.30001e-06 [offloading_packed_experts]: 4.57e-06 [overlap_recompute_and_grad_model_parallel]: 7.37002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.13001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.19998e-06 [overlap_recompute_comm]: 2.50002e-06 [overlap_grad_ring_attention]: 4.67e-06 [overlap_grad_flash_sp]: 2.546e-05 [begin_end_overlap_inline]: 5.09986e-07 [split_matmul_comm_elemetwise]: 2.07999e-06 [split_layernorm_comm]: 1.90001e-06 [handle_group_info]: 1.12999e-06 [symbol_engine_optimizer]: 9.48e-05, [1] [Cycle 1]: 8.959e-05, [6] [build]: 4.43001e-06 [elim_shapecalc]: 1.36e-05 [elim_not_effective]: 1.746e-05 [opt_reshape]: 8.59e-06 [fold_const_symbol]: 1.375e-05 [renormalize]: 1.80007e-07 [detach_backward]: 2.94999e-06 [pipeline_parallel_scheduler]: 1.44998e-06 [auto_monad_reorder]: 2.102e-05 [get_jit_bprop_graph]: 2.11e-06 [rewriter_after_jit_bprop_graph]: 5.47001e-06 [opt_after_jit_grad]: 0.00055189 [validate]: 5.147e-05 Sums bootstrap : 0.000483s : 0.05% type_inference : 0.746186s : 81.74% event_method : 0.000097s : 0.01% auto_monad : 0.000213s : 0.02% graph_reusing : 0.000014s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000073s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000016s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000096s : 0.01% 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 : 0.000009s : 0.00% optimize.rewriter_before_opt_a : 0.000461s : 0.05% optimize.opt_a.expand_dump_flag : 0.000014s : 0.00% optimize.opt_a.switch_simplify : 0.000294s : 0.03% optimize.opt_a.loop_unroll : 0.000097s : 0.01% optimize.opt_a.a_1 : 0.002221s : 0.24% optimize.opt_a.with_stream_mark : 0.000053s : 0.01% optimize.opt_a.recompute_prepare : 0.000028s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000212s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000009s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.00% optimize.opt_a.merge_send_recv : 0.000023s : 0.00% optimize.opt_a.auto_parallel : 0.000022s : 0.00% optimize.opt_a.parallel : 0.000033s : 0.00% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000011s : 0.00% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000025s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000021s : 0.00% optimize.opt_a.virtual_dataset : 0.000017s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.00% optimize.opt_a.virtual_output : 0.000015s : 0.00% optimize.opt_a.merge_forward : 0.000011s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000027s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000038s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000028s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.00% optimize.opt_a.meta_fg_expand : 0.000010s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000026s : 0.00% optimize.opt_a.a_after_grad : 0.000025s : 0.00% optimize.opt_a.renormalize : 0.158676s : 17.38% optimize.opt_a.add_forward_monad_depend : 0.000015s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000042s : 0.00% optimize.opt_a.cse : 0.000081s : 0.01% optimize.opt_a.a_3 : 0.000124s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000014s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000029s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000802s : 0.09% optimize.opt_b.b_1 : 0.000226s : 0.02% optimize.opt_b.b_2 : 0.000011s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000012s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000040s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000024s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000037s : 0.00% optimize.loop_unroll : 0.000538s : 0.06% optimize.opt_after_cconv.c_1 : 0.000036s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000034s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000056s : 0.01% optimize.tuple_transform.d_1 : 0.000059s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000067s : 0.01% optimize.cse_after_recomputation.cse : 0.000017s : 0.00% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000007s : 0.00% 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 : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000016s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 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 : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000025s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000017s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 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 : 0.000021s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000552s : 0.06% validate : 0.000051s : 0.01% Time group info: ------[substitution.] 0.000737 75 8.48% : 0.000062s : 11: substitution.cast_eliminate 0.44% : 0.000003s : 4: substitution.elim_not_effective 0.28% : 0.000002s : 4: substitution.fold_const_symbol 0.98% : 0.000007s : 5: substitution.graph_param_transform 82.19% : 0.000605s : 24: substitution.inline 0.87% : 0.000006s : 8: substitution.j_node_and_user_rematch 1.12% : 0.000008s : 8: substitution.remove_not_recompute_node 0.96% : 0.000007s : 2: substitution.replace_old_param 3.04% : 0.000022s : 7: substitution.switch_simplify 1.63% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.746020 2 99.23% : 0.740289s : 1: type_inference.infer 0.77% : 0.005730s : 1: type_inference.specialize ------[replace.] 0.000306 33 60.30% : 0.000184s : 24: replace.inline 32.43% : 0.000099s : 7: replace.switch_simplify 7.27% : 0.000022s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000620 33 95.55% : 0.000592s : 24: match.inline 2.82% : 0.000017s : 7: match.switch_simplify 1.63% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000499 2544 1.07% : 0.000005s : 32: predicate.accumulaten_eliminater 0.49% : 0.000002s : 5: predicate.ad_related_special_op_eliminate 0.25% : 0.000001s : 10: predicate.addn_check_dump 1.04% : 0.000005s : 32: predicate.addn_zero_filter 0.90% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.04% : 0.000010s : 42: predicate.arithmetic_simplify 1.45% : 0.000007s : 32: predicate.cast_eliminate 0.29% : 0.000001s : 10: predicate.check_bprop_eliminate 0.29% : 0.000001s : 10: predicate.compare_switch_simplify 0.09% : 0.000000s : 5: predicate.const_output_eliminate 0.40% : 0.000002s : 10: predicate.depend_value_elim 0.98% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.01% : 0.000005s : 32: predicate.dict_get_item_eliminator 0.99% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.57% : 0.000003s : 10: predicate.dumpgradient_eliminate 0.13% : 0.000001s : 5: predicate.elim_not_effective 0.35% : 0.000002s : 5: predicate.elim_shapecalc_of_broadcastargs 1.09% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.03% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.01% : 0.000005s : 37: predicate.environ_get_depend_swap 1.44% : 0.000007s : 47: predicate.environ_get_eliminate 1.09% : 0.000005s : 37: predicate.environ_get_set_eliminate 1.74% : 0.000009s : 58: predicate.exchange_switch_depend_value 2.63% : 0.000013s : 58: predicate.float_depend_g_call 0.26% : 0.000001s : 10: predicate.float_environ_get_switch 0.52% : 0.000003s : 15: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 5: predicate.fold_const_symbol 0.37% : 0.000002s : 10: predicate.get_grad_eliminate 0.15% : 0.000001s : 5: predicate.graph_param_transform 0.30% : 0.000001s : 10: predicate.incorporate_call 0.26% : 0.000001s : 10: predicate.incorporate_call_switch 5.18% : 0.000026s : 120: predicate.inline 0.45% : 0.000002s : 10: predicate.inline_without_move 0.14% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.51% : 0.000003s : 10: predicate.less_batch_normalization 1.58% : 0.000008s : 44: predicate.list_to_tuple_eliminator_ 2.12% : 0.000011s : 76: predicate.load_eliminater 0.62% : 0.000003s : 5: predicate.loop_unroll_after_grad 3.39% : 0.000017s : 104: predicate.loop_unroll_before_grad 1.43% : 0.000007s : 42: predicate.make_slice_get_slice_eliminator 0.34% : 0.000002s : 10: predicate.merge_addn 0.36% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.29% : 0.000001s : 10: predicate.mini_step_allgather_replace 0.88% : 0.000004s : 32: predicate.minmaximum_grad 0.90% : 0.000004s : 5: predicate.mutable_eliminate 0.18% : 0.000001s : 5: predicate.opt_reshape 0.31% : 0.000002s : 5: predicate.parallel_virtual_node 2.44% : 0.000012s : 58: predicate.partial_defer_inline 1.19% : 0.000006s : 39: predicate.partial_eliminate 0.96% : 0.000005s : 32: predicate.print_const_string_wrapper 0.29% : 0.000001s : 10: predicate.reduce_all_const_elim 1.22% : 0.000006s : 32: predicate.reduce_eliminate 2.27% : 0.000011s : 76: predicate.redundant_stop_gradient_eliminater 0.17% : 0.000001s : 10: predicate.remove_not_recompute_node 1.02% : 0.000005s : 44: predicate.replace_applicator 0.25% : 0.000001s : 10: predicate.replace_old_param 0.23% : 0.000001s : 5: predicate.reset_defer_inline 1.00% : 0.000005s : 32: predicate.reshape_eliminate 0.39% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.21% : 0.000001s : 5: predicate.row_tensor_eliminate 0.42% : 0.000002s : 10: predicate.same_eliminate 0.24% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.41% : 0.000002s : 10: predicate.shard_identity_eliminate 0.34% : 0.000002s : 10: predicate.special_op_eliminate 0.41% : 0.000002s : 10: predicate.specialize_transform 0.52% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.50% : 0.000003s : 10: predicate.stack_unstack_eliminate 0.20% : 0.000001s : 5: predicate.switch_call_monad_eliminater 1.83% : 0.000009s : 58: predicate.switch_defer_inline 2.12% : 0.000011s : 68: predicate.switch_layer_defer_inline 7.06% : 0.000035s : 191: predicate.switch_simplify 0.98% : 0.000005s : 32: predicate.tile_eliminate 1.18% : 0.000006s : 32: predicate.transpose_eliminate 1.40% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.34% : 0.000007s : 42: predicate.tuple_list_get_item_const_eliminator 1.21% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.45% : 0.000012s : 54: predicate.tuple_list_get_item_eliminator 1.32% : 0.000007s : 42: predicate.tuple_list_get_set_item_eliminator 1.89% : 0.000009s : 52: predicate.tuple_list_set_item_eliminator 1.44% : 0.000007s : 44: predicate.tuple_to_list_eliminator_ 2.05% : 0.000010s : 76: predicate.updatestate_pure_node_eliminater 14.79% : 0.000074s : 86: predicate.updatestate_useless_node_eliminater 0.19% : 0.000001s : 5: predicate.value_based_eliminate 0.34% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.36% : 0.000002s : 10: predicate.virtual_output_eliminate 0.13% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.20% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003270 42 50.79% : 0.001661s : 16: func_graph_cloner_run.FuncGraphClonerGraph 49.21% : 0.001609s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.251474 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.38% : 0.004720s : 1: add_attr 0.38% : 0.004702s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000073s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.02% : 0.000223s : 1: auto_monad 0.00% : 0.000025s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.04% : 0.000520s : 1: bootstrap 0.00% : 0.000041s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000020s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000031s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.01% : 0.000105s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000018s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.04% : 0.000551s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.07% : 0.000821s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000023s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000027s : 1: opt.transform.mutable_eliminate 0.25% : 0.003089s : 78: opt.transform.opt_a 0.00% : 0.000035s : 1: opt.transform.opt_after_cconv 0.00% : 0.000031s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000206s : 28: opt.transform.opt_b 0.01% : 0.000065s : 2: opt.transform.opt_trans_graph 0.00% : 0.000049s : 4: opt.transform.symbol_engine_opt 13.02% : 0.162930s : 1: opt_a 0.01% : 0.000138s : 1: opt_after_cconv 0.05% : 0.000565s : 1: opt_after_jit_grad 0.03% : 0.000349s : 1: opt_b 13.27% : 0.166041s : 1: optimize 0.00% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000029s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.01% : 0.000102s : 1: pre_auto_parallel 0.00% : 0.000013s : 1: py_interpret_to_execute 0.00% : 0.000017s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.00% : 0.000062s : 1: remove_dup_value 12.54% : 0.156880s : 1: renormalize.infer 0.14% : 0.001771s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000032s : 1: rewriter_after_opt_a 0.04% : 0.000475s : 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% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000097s : 1: symbol_engine_optimizer 0.01% : 0.000095s : 1: tuple_transform 59.63% : 0.746230s : 1: type_inference TotalTime = 0.805188, [21] [bootstrap]: 0.00049295 [type_inference]: 0.70787 [event_method]: 6.921e-05 [auto_monad]: 0.00026239 [graph_reusing]: 1.831e-05 [inline]: 3.36001e-06 [add_attr]: 0.00440393, [1] [add_attr_with_inline]: 0.00438886, [1] [Cycle 1]: 0.00015299, [2] [tag_attr]: 8.615e-05 [meta_addattr_fg_expand]: 1.885e-05 [parallel-infer-symbol]: 3.56001e-06 [pre_auto_parallel]: 0.0001133 [insert-virtual-dataset]: 3.14999e-06 [parallel-infer-symbol-second]: 8.30012e-07 [dataset_repeat_opt]: 2.22999e-06 [pipeline_split]: 1.66e-06 [optimize]: 0.0909667, [53] [py_interpret_to_execute]: 9.47001e-06 [rewriter_before_opt_a]: 0.00056215 [opt_a]: 0.0881057, [2] [Cycle 1]: 0.0867897, [45] [expand_dump_flag]: 9.81e-06 [switch_simplify]: 0.00031791 [loop_unroll]: 0.00011166 [a_1]: 0.00271696 [with_stream_mark]: 4.163e-05 [recompute_prepare]: 3.043e-05 [updatestate_depend_eliminate]: 1.043e-05 [updatestate_assign_eliminate]: 8e-06 [updatestate_loads_eliminate]: 7.35e-06 [parameter_eliminate]: 3.67002e-06 [a_2]: 0.00024397 [accelerated_algorithm]: 1.678e-05 [shard]: 2.17999e-06 [meta_shard_fg_expand]: 9.03002e-06 [shard_inline]: 1.579e-05 [merge_send_recv]: 1.936e-05 [auto_parallel]: 1.587e-05 [parallel]: 2.25e-05 [flash_sp]: 1.433e-05 [merge_comm]: 9.87999e-06 [allreduce_fusion]: 9.07001e-06 [matmul_add_comm_reduction]: 3.36e-05 [allreduce_slice_to_reducescatter]: 8.80013e-07 [virtual_shard_identity]: 1.93e-05 [virtual_dataset]: 1.536e-05 [get_grad_eliminate_]: 1.459e-05 [virtual_output]: 1.595e-05 [merge_forward]: 1.056e-05 [cell_reuse_recompute_pass]: 2.46998e-06 [offload_activation]: 1.81e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.286e-05 [merge_recompute_call_nodes]: 1.90001e-06 [before_grad]: 2.923e-05 [set_forward_comm_id_for_comm_node_pass]: 1.028e-05 [meta_fg_expand]: 0.0737566 [flash_sp_send_recv_attached]: 1.024e-05 [receive_attached]: 2.29001e-06 [after_resolve]: 0.00012836 [a_after_grad]: 0.00015818 [renormalize]: 0.00837404 [add_forward_monad_depend]: 1.012e-05 [auto_monad_grad]: 2.52001e-06 [auto_monad_eliminator]: 1.846e-05 [cse]: 3.034e-05 [a_3]: 4.91e-05 [Cycle 2]: 0.00129849, [45] [expand_dump_flag]: 2.53998e-06 [switch_simplify]: 6.77002e-06 [loop_unroll]: 4.60001e-06 [a_1]: 5.345e-05 [with_stream_mark]: 1.814e-05 [recompute_prepare]: 4.99e-06 [updatestate_depend_eliminate]: 3.83999e-06 [updatestate_assign_eliminate]: 2.51e-06 [updatestate_loads_eliminate]: 2.46998e-06 [parameter_eliminate]: 2.07001e-06 [a_2]: 3.845e-05 [accelerated_algorithm]: 4.08001e-06 [shard]: 2.027e-05 [meta_shard_fg_expand]: 2.41e-06 [shard_inline]: 4.30999e-06 [merge_send_recv]: 7.78001e-06 [auto_parallel]: 8.25e-06 [parallel]: 1.003e-05 [flash_sp]: 3.6e-06 [merge_comm]: 2.53e-06 [allreduce_fusion]: 2.37001e-06 [matmul_add_comm_reduction]: 8.88002e-06 [allreduce_slice_to_reducescatter]: 1.30999e-06 [virtual_shard_identity]: 6.13998e-06 [virtual_dataset]: 4.15e-06 [get_grad_eliminate_]: 3.53e-06 [virtual_output]: 3.93001e-06 [merge_forward]: 3.51001e-06 [cell_reuse_recompute_pass]: 3.83001e-06 [offload_activation]: 8.80001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.289e-05 [merge_recompute_call_nodes]: 1.74e-06 [before_grad]: 6.48998e-06 [set_forward_comm_id_for_comm_node_pass]: 2.63e-06 [meta_fg_expand]: 0.0006926 [flash_sp_send_recv_attached]: 3.37002e-06 [receive_attached]: 2.34001e-06 [after_resolve]: 1.59e-05 [a_after_grad]: 7.28e-06 [renormalize]: 7.00238e-08 [add_forward_monad_depend]: 5.30999e-06 [auto_monad_grad]: 2.54999e-06 [auto_monad_eliminator]: 1.448e-05 [cse]: 2.556e-05 [a_3]: 2.376e-05 [py_interpret_to_execute_after_opt_a]: 9.82999e-06 [slice_cell_reuse_recomputed_activation]: 2.29001e-06 [rewriter_after_opt_a]: 2.277e-05 [convert_after_rewriter]: 1.22999e-06 [order_py_execute_after_rewriter]: 1.20999e-06 [mutable_eliminate]: 0.00078781 [opt_b]: 0.00015198, [1] [Cycle 1]: 0.00014221, [7] [b_1]: 7.209e-05 [b_2]: 4.99003e-06 [updatestate_depend_eliminate]: 6.98e-06 [updatestate_assign_eliminate]: 2.21e-06 [updatestate_loads_eliminate]: 1.91e-06 [renormalize]: 9.09989e-07 [cse]: 1.966e-05 [optimize_parallel_all_gather_comm]: 2.052e-05 [overlap_param_gather]: 1.99e-06 [cconv]: 3.589e-05 [loop_unroll]: 0.00048495 [opt_after_cconv]: 8.705e-05, [1] [Cycle 1]: 8.021e-05, [7] [c_1]: 1.54e-05 [parameter_eliminate]: 5.67001e-06 [updatestate_depend_eliminate]: 5.25999e-06 [updatestate_assign_eliminate]: 1.96998e-06 [updatestate_loads_eliminate]: 1.39e-06 [cse]: 1.737e-05 [renormalize]: 4.69998e-07 [remove_dup_value]: 1.474e-05 [tuple_transform]: 5.448e-05, [1] [Cycle 1]: 4.937e-05, [4] [d_1]: 2.563e-05 [none_parameter_eliminate]: 1.59998e-06 [renormalize]: 1.60013e-07 [switch_simplify]: 4.03999e-06 [partial_unused_args_eliminate]: 1.94999e-06 [add_recomputation]: 3.82e-05 [cse_after_recomputation]: 1.99e-05, [1] [Cycle 1]: 1.386e-05, [1] [cse]: 8.67e-06 [environ_conv]: 5.00999e-06 [swap_dp_allreduce_reducescatter]: 4.33999e-06 [bias_add_comm_swap]: 3.08e-06 [label_micro_interleaved_index]: 5.81e-06 [label_fine_grained_interleaved_index]: 2.44999e-06 [merge_cast_opt]: 1.25999e-06 [slice_recompute_activation]: 2.08998e-06 [micro_interleaved_order_control]: 2.39999e-06 [assign_add_opt]: 1.45001e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 9.70002e-07 [full_micro_interleaved_order_control]: 2.35997e-06 [reorder_send_recv_between_fp_bp]: 2.68e-06 [comm_op_add_attrs]: 1.22e-06 [add_comm_op_reuse_tag]: 9.60019e-07 [interleave_split_concat_branches]: 1.29e-06 [interleave_parallel_branches]: 1.00999e-06 [overlap_opt_shard_in_pipeline]: 1.09998e-06 [overlap_opt_shard_grad_in_pipeline]: 1.64998e-06 [control_data_broadcast_order]: 1.225e-05 [grouped_pairwise_exchange_alltoall]: 1.69998e-06 [offloading_packed_experts]: 3.75e-06 [overlap_recompute_and_grad_model_parallel]: 3.83999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.47999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.17e-06 [overlap_recompute_comm]: 2.27999e-06 [overlap_grad_ring_attention]: 3.2e-06 [overlap_grad_flash_sp]: 1.797e-05 [begin_end_overlap_inline]: 5.09986e-07 [split_matmul_comm_elemetwise]: 2.14e-06 [split_layernorm_comm]: 1.69998e-06 [handle_group_info]: 1.33002e-06 [symbol_engine_optimizer]: 7.696e-05, [1] [Cycle 1]: 7.035e-05, [6] [build]: 4.1e-06 [elim_shapecalc]: 1.05e-05 [elim_not_effective]: 1.027e-05 [opt_reshape]: 4.74002e-06 [fold_const_symbol]: 6.76999e-06 [renormalize]: 1.69995e-07 [detach_backward]: 2.24999e-06 [pipeline_parallel_scheduler]: 1.77999e-06 [auto_monad_reorder]: 1.782e-05 [get_jit_bprop_graph]: 1.99e-06 [rewriter_after_jit_bprop_graph]: 6.62002e-06 [opt_after_jit_grad]: 0.00068757 [validate]: 4.362e-05 Sums bootstrap : 0.000493s : 0.06% type_inference : 0.707870s : 88.54% event_method : 0.000069s : 0.01% auto_monad : 0.000262s : 0.03% graph_reusing : 0.000018s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000086s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000019s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000113s : 0.01% 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 : 0.000009s : 0.00% optimize.rewriter_before_opt_a : 0.000562s : 0.07% optimize.opt_a.expand_dump_flag : 0.000012s : 0.00% optimize.opt_a.switch_simplify : 0.000325s : 0.04% optimize.opt_a.loop_unroll : 0.000116s : 0.01% optimize.opt_a.a_1 : 0.002770s : 0.35% optimize.opt_a.with_stream_mark : 0.000060s : 0.01% optimize.opt_a.recompute_prepare : 0.000035s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000014s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000282s : 0.04% optimize.opt_a.accelerated_algorithm : 0.000021s : 0.00% optimize.opt_a.shard : 0.000022s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000011s : 0.00% optimize.opt_a.shard_inline : 0.000020s : 0.00% optimize.opt_a.merge_send_recv : 0.000027s : 0.00% optimize.opt_a.auto_parallel : 0.000024s : 0.00% optimize.opt_a.parallel : 0.000033s : 0.00% optimize.opt_a.flash_sp : 0.000018s : 0.00% optimize.opt_a.merge_comm : 0.000012s : 0.00% optimize.opt_a.allreduce_fusion : 0.000011s : 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 : 0.000025s : 0.00% optimize.opt_a.virtual_dataset : 0.000020s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000018s : 0.00% optimize.opt_a.virtual_output : 0.000020s : 0.00% optimize.opt_a.merge_forward : 0.000014s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000027s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000046s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000036s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000013s : 0.00% optimize.opt_a.meta_fg_expand : 0.074449s : 9.31% optimize.opt_a.flash_sp_send_recv_attached : 0.000014s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000144s : 0.02% optimize.opt_a.a_after_grad : 0.000165s : 0.02% optimize.opt_a.renormalize : 0.008374s : 1.05% optimize.opt_a.add_forward_monad_depend : 0.000015s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000033s : 0.00% optimize.opt_a.cse : 0.000056s : 0.01% optimize.opt_a.a_3 : 0.000073s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000010s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000023s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000788s : 0.10% optimize.opt_b.b_1 : 0.000072s : 0.01% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000020s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.00% optimize.loop_unroll : 0.000485s : 0.06% optimize.opt_after_cconv.c_1 : 0.000015s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000017s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000026s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000038s : 0.00% optimize.cse_after_recomputation.cse : 0.000009s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 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 : 0.000002s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000012s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000018s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000010s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 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 : 0.000018s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000688s : 0.09% validate : 0.000044s : 0.01% Time group info: ------[substitution.] 0.000917 134 3.61% : 0.000033s : 8: substitution.cast_eliminate 0.17% : 0.000002s : 1: substitution.elim_not_effective 0.71% : 0.000007s : 9: substitution.float_depend_g_call 0.59% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.12% : 0.000001s : 1: substitution.fold_const_symbol 0.58% : 0.000005s : 1: substitution.graph_param_transform 0.34% : 0.000003s : 2: substitution.incorporate_call 0.27% : 0.000002s : 2: substitution.incorporate_call_switch 76.81% : 0.000705s : 33: substitution.inline 4.35% : 0.000040s : 4: substitution.inline_without_move 0.88% : 0.000008s : 11: substitution.j_node_and_user_rematch 0.45% : 0.000004s : 2: substitution.minmaximum_grad 0.73% : 0.000007s : 9: substitution.partial_eliminate 0.98% : 0.000009s : 11: substitution.remove_not_recompute_node 0.26% : 0.000002s : 1: substitution.replace_applicator 1.09% : 0.000010s : 12: substitution.replace_old_param 0.44% : 0.000004s : 1: substitution.set_cell_output_no_recompute 2.54% : 0.000023s : 10: substitution.switch_simplify 0.95% : 0.000009s : 2: substitution.tuple_list_convert_item_index_to_positive 0.34% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.54% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.77% : 0.000025s : 6: substitution.tuple_list_get_item_eliminator 0.49% : 0.000004s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.707625 2 99.06% : 0.701005s : 1: type_inference.infer 0.94% : 0.006619s : 1: type_inference.specialize ------[replace.] 0.000355 45 62.64% : 0.000223s : 33: replace.inline 33.07% : 0.000118s : 10: replace.switch_simplify 4.29% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000719 45 95.84% : 0.000689s : 33: match.inline 2.58% : 0.000019s : 10: match.switch_simplify 1.59% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000516 3189 1.27% : 0.000007s : 43: predicate.accumulaten_eliminater 0.34% : 0.000002s : 1: predicate.ad_related_special_op_eliminate 0.37% : 0.000002s : 14: predicate.addn_check_dump 1.14% : 0.000006s : 43: predicate.addn_zero_filter 1.08% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 2.28% : 0.000012s : 57: predicate.arithmetic_simplify 1.55% : 0.000008s : 43: predicate.cast_eliminate 0.11% : 0.000001s : 2: predicate.check_bprop_eliminate 0.39% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.43% : 0.000002s : 14: predicate.depend_value_elim 1.22% : 0.000006s : 43: predicate.dict_get_item_const_eliminator 1.25% : 0.000006s : 43: predicate.dict_get_item_eliminator 1.17% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.39% : 0.000002s : 2: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 1: predicate.elim_not_effective 0.19% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.41% : 0.000007s : 44: predicate.environ_add_const_eliminate 1.14% : 0.000006s : 44: predicate.environ_get_add_eliminate 1.20% : 0.000006s : 44: predicate.environ_get_depend_swap 1.60% : 0.000008s : 58: predicate.environ_get_eliminate 1.29% : 0.000007s : 44: predicate.environ_get_set_eliminate 2.17% : 0.000011s : 78: predicate.exchange_switch_depend_value 3.02% : 0.000016s : 78: predicate.float_depend_g_call 0.44% : 0.000002s : 14: predicate.float_environ_get_switch 0.49% : 0.000003s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.46% : 0.000002s : 14: predicate.get_grad_eliminate 0.07% : 0.000000s : 1: predicate.graph_param_transform 0.40% : 0.000002s : 14: predicate.incorporate_call 0.35% : 0.000002s : 14: predicate.incorporate_call_switch 6.06% : 0.000031s : 151: predicate.inline 2.74% : 0.000014s : 70: predicate.inline_without_move 0.21% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.67% : 0.000003s : 14: predicate.less_batch_normalization 1.47% : 0.000008s : 47: predicate.list_to_tuple_eliminator_ 2.34% : 0.000012s : 90: predicate.load_eliminater 0.33% : 0.000002s : 1: predicate.loop_unroll_after_grad 3.78% : 0.000019s : 141: predicate.loop_unroll_before_grad 1.27% : 0.000007s : 45: predicate.make_slice_get_slice_eliminator 0.39% : 0.000002s : 14: predicate.merge_addn 0.10% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.08% : 0.000000s : 2: predicate.mini_step_allgather_replace 1.10% : 0.000006s : 43: predicate.minmaximum_grad 0.58% : 0.000003s : 1: predicate.mutable_eliminate 0.10% : 0.000001s : 1: predicate.opt_reshape 0.07% : 0.000000s : 1: predicate.parallel_virtual_node 2.76% : 0.000014s : 78: predicate.partial_defer_inline 1.38% : 0.000007s : 46: predicate.partial_eliminate 1.25% : 0.000006s : 43: predicate.print_const_string_wrapper 0.41% : 0.000002s : 14: predicate.reduce_all_const_elim 1.51% : 0.000008s : 43: predicate.reduce_eliminate 2.36% : 0.000012s : 90: predicate.redundant_stop_gradient_eliminater 0.33% : 0.000002s : 14: predicate.remove_not_recompute_node 1.10% : 0.000006s : 47: predicate.replace_applicator 1.27% : 0.000007s : 70: predicate.replace_old_param 0.03% : 0.000000s : 1: predicate.reset_defer_inline 1.24% : 0.000006s : 43: predicate.reshape_eliminate 0.17% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.10% : 0.000001s : 1: predicate.row_tensor_eliminate 0.25% : 0.000001s : 2: predicate.same_eliminate 0.29% : 0.000002s : 14: predicate.set_cell_output_no_recompute 0.51% : 0.000003s : 14: predicate.shard_identity_eliminate 0.17% : 0.000001s : 2: predicate.special_op_eliminate 0.42% : 0.000002s : 14: predicate.specialize_transform 0.47% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 2.65% : 0.000014s : 70: predicate.stack_unstack_eliminate 0.16% : 0.000001s : 1: predicate.switch_call_monad_eliminater 2.36% : 0.000012s : 78: predicate.switch_defer_inline 2.48% : 0.000013s : 80: predicate.switch_layer_defer_inline 7.69% : 0.000040s : 254: predicate.switch_simplify 1.23% : 0.000006s : 43: predicate.tile_eliminate 1.14% : 0.000006s : 43: predicate.transpose_eliminate 1.31% : 0.000007s : 45: predicate.tuple_list_convert_item_index_to_positive 1.40% : 0.000007s : 45: predicate.tuple_list_get_item_const_eliminator 1.36% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.39% : 0.000012s : 61: predicate.tuple_list_get_item_eliminator 1.37% : 0.000007s : 45: predicate.tuple_list_get_set_item_eliminator 1.93% : 0.000010s : 59: predicate.tuple_list_set_item_eliminator 1.31% : 0.000007s : 47: predicate.tuple_to_list_eliminator_ 2.31% : 0.000012s : 90: predicate.updatestate_pure_node_eliminater 2.79% : 0.000014s : 104: predicate.updatestate_useless_node_eliminater 0.10% : 0.000001s : 1: predicate.value_based_eliminate 0.59% : 0.000003s : 14: predicate.virtual_dataset_eliminate 0.52% : 0.000003s : 14: predicate.virtual_output_eliminate 0.11% : 0.000001s : 1: predicate.virtual_view_grad_eliminate 0.16% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.005028 69 74.86% : 0.003764s : 33: func_graph_cloner_run.FuncGraphClonerGraph 25.14% : 0.001264s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.912812 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.48% : 0.004411s : 1: add_attr 0.48% : 0.004393s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000042s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.03% : 0.000274s : 1: auto_monad 0.00% : 0.000023s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.06% : 0.000528s : 1: bootstrap 0.00% : 0.000039s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000015s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000023s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.01% : 0.000078s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000023s : 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% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000003s : 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% : 0.000008s : 1: label_micro_interleaved_index 0.05% : 0.000498s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000802s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.00% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 0.44% : 0.004007s : 78: opt.transform.opt_a 0.00% : 0.000014s : 1: opt.transform.opt_after_cconv 0.00% : 0.000023s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000047s : 28: opt.transform.opt_b 0.00% : 0.000028s : 2: opt.transform.opt_trans_graph 0.00% : 0.000028s : 4: opt.transform.symbol_engine_opt 9.65% : 0.088109s : 1: opt_a 0.01% : 0.000091s : 1: opt_after_cconv 0.08% : 0.000706s : 1: opt_after_jit_grad 0.02% : 0.000156s : 1: opt_b 9.97% : 0.090973s : 1: optimize 0.00% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000021s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000004s : 1: pipeline_split 0.01% : 0.000119s : 1: pre_auto_parallel 0.00% : 0.000014s : 1: py_interpret_to_execute 0.00% : 0.000013s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.00% : 0.000019s : 1: remove_dup_value 0.82% : 0.007492s : 1: renormalize.infer 0.09% : 0.000866s : 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% : 0.000026s : 1: rewriter_after_opt_a 0.06% : 0.000572s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000080s : 1: symbol_engine_optimizer 0.01% : 0.000057s : 1: tuple_transform 77.55% : 0.707912s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:42:23.756.573 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.677938, [21] [bootstrap]: 0.00048489 [type_inference]: 0.435821 [event_method]: 8.197e-05 [auto_monad]: 0.00019151 [graph_reusing]: 1.154e-05 [inline]: 3.41999e-06 [add_attr]: 0.00464618, [1] [add_attr_with_inline]: 0.00463055, [1] [Cycle 1]: 0.00013903, [2] [tag_attr]: 7.149e-05 [meta_addattr_fg_expand]: 1.363e-05 [parallel-infer-symbol]: 4.06001e-06 [pre_auto_parallel]: 9.101e-05 [insert-virtual-dataset]: 2.83e-06 [parallel-infer-symbol-second]: 1.55001e-06 [dataset_repeat_opt]: 1.77999e-06 [pipeline_split]: 2.16e-06 [optimize]: 0.235781, [53] [py_interpret_to_execute]: 1.272e-05 [rewriter_before_opt_a]: 0.00043764 [opt_a]: 0.233026, [2] [Cycle 1]: 0.23236, [45] [expand_dump_flag]: 1.055e-05 [switch_simplify]: 0.00026351 [loop_unroll]: 9.784e-05 [a_1]: 0.228234 [with_stream_mark]: 3.247e-05 [recompute_prepare]: 1.799e-05 [updatestate_depend_eliminate]: 4.40999e-06 [updatestate_assign_eliminate]: 3.60998e-06 [updatestate_loads_eliminate]: 3.3e-06 [parameter_eliminate]: 2.83003e-06 [a_2]: 7.776e-05 [accelerated_algorithm]: 8.12e-06 [shard]: 2.65997e-06 [meta_shard_fg_expand]: 5.13002e-06 [shard_inline]: 6.20002e-06 [merge_send_recv]: 1.079e-05 [auto_parallel]: 9.84001e-06 [parallel]: 2.062e-05 [flash_sp]: 1.215e-05 [merge_comm]: 3.97002e-06 [allreduce_fusion]: 3.66999e-06 [matmul_add_comm_reduction]: 1.068e-05 [allreduce_slice_to_reducescatter]: 9.5999e-07 [virtual_shard_identity]: 9.75002e-06 [virtual_dataset]: 6.39001e-06 [get_grad_eliminate_]: 6.24999e-06 [virtual_output]: 6.19999e-06 [merge_forward]: 4.52998e-06 [cell_reuse_recompute_pass]: 2.81999e-06 [offload_activation]: 1.277e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.689e-05 [merge_recompute_call_nodes]: 1.60999e-06 [before_grad]: 1.22e-05 [set_forward_comm_id_for_comm_node_pass]: 4.16001e-06 [meta_fg_expand]: 3.93999e-06 [flash_sp_send_recv_attached]: 3.04999e-06 [receive_attached]: 2.43e-06 [after_resolve]: 1.118e-05 [a_after_grad]: 1.006e-05 [renormalize]: 0.00289625 [add_forward_monad_depend]: 1.243e-05 [auto_monad_grad]: 2.88e-06 [auto_monad_eliminator]: 2.72e-05 [cse]: 3.572e-05 [a_3]: 6.065e-05 [Cycle 2]: 0.00064698, [45] [expand_dump_flag]: 2.71e-06 [switch_simplify]: 8.27998e-06 [loop_unroll]: 5.25999e-06 [a_1]: 0.00011289 [with_stream_mark]: 1.99e-05 [recompute_prepare]: 6.07999e-06 [updatestate_depend_eliminate]: 4.05e-06 [updatestate_assign_eliminate]: 3.14999e-06 [updatestate_loads_eliminate]: 3.39001e-06 [parameter_eliminate]: 1.62001e-06 [a_2]: 6.577e-05 [accelerated_algorithm]: 6.98e-06 [shard]: 2.56e-06 [meta_shard_fg_expand]: 2.11e-06 [shard_inline]: 5.92001e-06 [merge_send_recv]: 9.19e-06 [auto_parallel]: 9.47999e-06 [parallel]: 1.008e-05 [flash_sp]: 3.97e-06 [merge_comm]: 3.07002e-06 [allreduce_fusion]: 3.21001e-06 [matmul_add_comm_reduction]: 1.018e-05 [allreduce_slice_to_reducescatter]: 6.59988e-07 [virtual_shard_identity]: 6.43e-06 [virtual_dataset]: 5.40999e-06 [get_grad_eliminate_]: 5.34e-06 [virtual_output]: 4.92e-06 [merge_forward]: 3.86999e-06 [cell_reuse_recompute_pass]: 2.51e-06 [offload_activation]: 9.83998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.651e-05 [merge_recompute_call_nodes]: 1.29e-06 [before_grad]: 9.07999e-06 [set_forward_comm_id_for_comm_node_pass]: 3.58999e-06 [meta_fg_expand]: 2.48e-06 [flash_sp_send_recv_attached]: 1.76998e-06 [receive_attached]: 2.00002e-06 [after_resolve]: 1.144e-05 [a_after_grad]: 8.08999e-06 [renormalize]: 1.10012e-07 [add_forward_monad_depend]: 2.09999e-06 [auto_monad_grad]: 1.37999e-06 [auto_monad_eliminator]: 8.05e-06 [cse]: 1.46e-05 [a_3]: 3.165e-05 [py_interpret_to_execute_after_opt_a]: 9.44e-06 [slice_cell_reuse_recomputed_activation]: 1.92999e-06 [rewriter_after_opt_a]: 2.276e-05 [convert_after_rewriter]: 1.34e-06 [order_py_execute_after_rewriter]: 1.17e-06 [mutable_eliminate]: 0.00078144 [opt_b]: 0.000198, [1] [Cycle 1]: 0.00018899, [7] [b_1]: 0.00010708 [b_2]: 7.25e-06 [updatestate_depend_eliminate]: 8.27e-06 [updatestate_assign_eliminate]: 2.58e-06 [updatestate_loads_eliminate]: 2.76e-06 [renormalize]: 5.60016e-07 [cse]: 2.536e-05 [optimize_parallel_all_gather_comm]: 1.97e-05 [overlap_param_gather]: 1.92001e-06 [cconv]: 3.527e-05 [loop_unroll]: 0.00049275 [opt_after_cconv]: 9.808e-05, [1] [Cycle 1]: 9.152e-05, [7] [c_1]: 2.541e-05 [parameter_eliminate]: 4.52998e-06 [updatestate_depend_eliminate]: 5.79999e-06 [updatestate_assign_eliminate]: 2.39001e-06 [updatestate_loads_eliminate]: 3.03e-06 [cse]: 1.863e-05 [renormalize]: 5.50004e-07 [remove_dup_value]: 1.448e-05 [tuple_transform]: 7.056e-05, [1] [Cycle 1]: 6.608e-05, [4] [d_1]: 3.973e-05 [none_parameter_eliminate]: 1.67001e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 5.92999e-06 [partial_unused_args_eliminate]: 2.09999e-06 [add_recomputation]: 4.967e-05 [cse_after_recomputation]: 2.121e-05, [1] [Cycle 1]: 1.572e-05, [1] [cse]: 1.057e-05 [environ_conv]: 6.10002e-06 [swap_dp_allreduce_reducescatter]: 4.76002e-06 [bias_add_comm_swap]: 2.48998e-06 [label_micro_interleaved_index]: 4.65999e-06 [label_fine_grained_interleaved_index]: 2.34001e-06 [merge_cast_opt]: 1.30999e-06 [slice_recompute_activation]: 2.14e-06 [micro_interleaved_order_control]: 2.20002e-06 [assign_add_opt]: 1.24e-06 [ForceFp32Comm]: 8.70001e-07 [remove_cast_before_assign_add]: 9.39996e-07 [full_micro_interleaved_order_control]: 2.58003e-06 [reorder_send_recv_between_fp_bp]: 2.38998e-06 [comm_op_add_attrs]: 9.99979e-07 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.12e-06 [interleave_parallel_branches]: 1.09003e-06 [overlap_opt_shard_in_pipeline]: 1.32999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.31e-06 [control_data_broadcast_order]: 1.272e-05 [grouped_pairwise_exchange_alltoall]: 1.35001e-06 [offloading_packed_experts]: 3.97002e-06 [overlap_recompute_and_grad_model_parallel]: 4.72e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.15999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.42999e-06 [overlap_recompute_comm]: 2.48e-06 [overlap_grad_ring_attention]: 3.7e-06 [overlap_grad_flash_sp]: 2.363e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.56e-06 [split_layernorm_comm]: 1.75001e-06 [handle_group_info]: 1.09e-06 [symbol_engine_optimizer]: 7.636e-05, [1] [Cycle 1]: 7.122e-05, [6] [build]: 4.15999e-06 [elim_shapecalc]: 1.043e-05 [elim_not_effective]: 1.357e-05 [opt_reshape]: 5.80002e-06 [fold_const_symbol]: 9.31e-06 [renormalize]: 1.8999e-07 [detach_backward]: 2.32001e-06 [pipeline_parallel_scheduler]: 1.40999e-06 [auto_monad_reorder]: 1.799e-05 [get_jit_bprop_graph]: 1.77001e-06 [rewriter_after_jit_bprop_graph]: 7.13e-06 [opt_after_jit_grad]: 0.00053 [validate]: 4.404e-05 Sums bootstrap : 0.000485s : 0.07% type_inference : 0.435821s : 64.84% event_method : 0.000082s : 0.01% auto_monad : 0.000192s : 0.03% graph_reusing : 0.000012s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000071s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000014s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000091s : 0.01% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000013s : 0.00% optimize.rewriter_before_opt_a : 0.000438s : 0.07% optimize.opt_a.expand_dump_flag : 0.000013s : 0.00% optimize.opt_a.switch_simplify : 0.000272s : 0.04% optimize.opt_a.loop_unroll : 0.000103s : 0.02% optimize.opt_a.a_1 : 0.228347s : 33.97% optimize.opt_a.with_stream_mark : 0.000052s : 0.01% optimize.opt_a.recompute_prepare : 0.000024s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000007s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000144s : 0.02% optimize.opt_a.accelerated_algorithm : 0.000015s : 0.00% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000012s : 0.00% optimize.opt_a.merge_send_recv : 0.000020s : 0.00% optimize.opt_a.auto_parallel : 0.000019s : 0.00% optimize.opt_a.parallel : 0.000031s : 0.00% optimize.opt_a.flash_sp : 0.000016s : 0.00% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000007s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000021s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000016s : 0.00% optimize.opt_a.virtual_dataset : 0.000012s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000012s : 0.00% optimize.opt_a.virtual_output : 0.000011s : 0.00% optimize.opt_a.merge_forward : 0.000008s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000023s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000033s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000021s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.00% optimize.opt_a.meta_fg_expand : 0.000006s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000023s : 0.00% optimize.opt_a.a_after_grad : 0.000018s : 0.00% optimize.opt_a.renormalize : 0.002896s : 0.43% optimize.opt_a.add_forward_monad_depend : 0.000015s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000035s : 0.01% optimize.opt_a.cse : 0.000050s : 0.01% optimize.opt_a.a_3 : 0.000092s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000009s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000023s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000781s : 0.12% optimize.opt_b.b_1 : 0.000107s : 0.02% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000025s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000035s : 0.01% optimize.loop_unroll : 0.000493s : 0.07% optimize.opt_after_cconv.c_1 : 0.000025s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000019s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000014s : 0.00% optimize.tuple_transform.d_1 : 0.000040s : 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 : 0.000006s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000050s : 0.01% optimize.cse_after_recomputation.cse : 0.000011s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.00% optimize.bias_add_comm_swap : 0.000002s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 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 : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000013s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000024s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000006s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000018s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000530s : 0.08% validate : 0.000044s : 0.01% Time group info: ------[substitution.] 0.000829 52 2.77% : 0.000023s : 2: substitution.cast_eliminate 0.20% : 0.000002s : 2: substitution.elim_not_effective 0.15% : 0.000001s : 2: substitution.fold_const_symbol 0.69% : 0.000006s : 3: substitution.graph_param_transform 90.14% : 0.000747s : 24: substitution.inline 0.63% : 0.000005s : 4: substitution.j_node_and_user_rematch 0.74% : 0.000006s : 4: substitution.remove_not_recompute_node 0.66% : 0.000005s : 2: substitution.replace_old_param 2.61% : 0.000022s : 7: substitution.switch_simplify 1.42% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.435672 2 98.87% : 0.430728s : 1: type_inference.infer 1.13% : 0.004944s : 1: type_inference.specialize ------[replace.] 0.000360 35 3.47% : 0.000013s : 2: replace.cast_eliminate 62.60% : 0.000225s : 24: replace.inline 28.52% : 0.000103s : 7: replace.switch_simplify 5.40% : 0.000019s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000780 35 2.73% : 0.000021s : 2: match.cast_eliminate 93.72% : 0.000731s : 24: match.inline 2.28% : 0.000018s : 7: match.switch_simplify 1.27% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.226300 2336 0.00% : 0.000006s : 32: predicate.accumulaten_eliminater 0.00% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.00% : 0.000001s : 6: predicate.addn_check_dump 0.00% : 0.000005s : 32: predicate.addn_zero_filter 0.00% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 0.01% : 0.000012s : 38: predicate.arithmetic_simplify 0.00% : 0.000007s : 34: predicate.cast_eliminate 0.00% : 0.000001s : 6: predicate.check_bprop_eliminate 0.00% : 0.000001s : 6: predicate.compare_switch_simplify 0.00% : 0.000000s : 3: predicate.const_output_eliminate 0.00% : 0.000001s : 6: predicate.depend_value_elim 0.00% : 0.000005s : 34: predicate.dict_get_item_const_eliminator 0.00% : 0.000005s : 34: predicate.dict_get_item_eliminator 0.00% : 0.000006s : 34: predicate.dict_set_item_eliminator 0.00% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.00% : 0.000001s : 3: predicate.elim_not_effective 0.00% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 0.00% : 0.000006s : 37: predicate.environ_add_const_eliminate 0.00% : 0.000005s : 37: predicate.environ_get_add_eliminate 0.00% : 0.000005s : 37: predicate.environ_get_depend_swap 0.00% : 0.000007s : 43: predicate.environ_get_eliminate 0.00% : 0.000005s : 37: predicate.environ_get_set_eliminate 0.00% : 0.000009s : 60: predicate.exchange_switch_depend_value 0.01% : 0.000013s : 60: predicate.float_depend_g_call 0.00% : 0.000001s : 6: predicate.float_environ_get_switch 0.00% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.00% : 0.000000s : 3: predicate.fold_const_symbol 0.00% : 0.000001s : 6: predicate.get_grad_eliminate 0.00% : 0.000000s : 3: predicate.graph_param_transform 0.00% : 0.000001s : 6: predicate.incorporate_call 0.00% : 0.000001s : 6: predicate.incorporate_call_switch 0.01% : 0.000027s : 110: predicate.inline 0.00% : 0.000002s : 6: predicate.inline_without_move 0.00% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.00% : 0.000002s : 6: predicate.less_batch_normalization 0.00% : 0.000007s : 42: predicate.list_to_tuple_eliminator_ 0.00% : 0.000010s : 74: predicate.load_eliminater 0.00% : 0.000002s : 3: predicate.loop_unroll_after_grad 0.01% : 0.000015s : 99: predicate.loop_unroll_before_grad 0.00% : 0.000006s : 40: predicate.make_slice_get_slice_eliminator 0.00% : 0.000001s : 6: predicate.merge_addn 0.00% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.00% : 0.000001s : 6: predicate.mini_step_allgather_replace 0.00% : 0.000005s : 32: predicate.minmaximum_grad 0.00% : 0.000002s : 3: predicate.mutable_eliminate 0.00% : 0.000001s : 3: predicate.opt_reshape 0.00% : 0.000001s : 3: predicate.parallel_virtual_node 0.01% : 0.000015s : 60: predicate.partial_defer_inline 0.00% : 0.000006s : 39: predicate.partial_eliminate 0.00% : 0.000005s : 32: predicate.print_const_string_wrapper 0.00% : 0.000001s : 6: predicate.reduce_all_const_elim 0.00% : 0.000007s : 32: predicate.reduce_eliminate 0.01% : 0.000011s : 74: predicate.redundant_stop_gradient_eliminater 0.00% : 0.000001s : 6: predicate.remove_not_recompute_node 0.00% : 0.000005s : 42: predicate.replace_applicator 0.00% : 0.000001s : 6: predicate.replace_old_param 0.00% : 0.000001s : 3: predicate.reset_defer_inline 0.00% : 0.000005s : 32: predicate.reshape_eliminate 0.00% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.00% : 0.000001s : 3: predicate.row_tensor_eliminate 0.00% : 0.000001s : 6: predicate.same_eliminate 0.00% : 0.000002s : 6: predicate.set_cell_output_no_recompute 0.00% : 0.000001s : 6: predicate.shard_identity_eliminate 0.00% : 0.000001s : 6: predicate.special_op_eliminate 0.00% : 0.000001s : 6: predicate.specialize_transform 0.00% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.00% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.00% : 0.000001s : 3: predicate.switch_call_monad_eliminater 0.00% : 0.000009s : 60: predicate.switch_defer_inline 0.00% : 0.000011s : 66: predicate.switch_layer_defer_inline 0.01% : 0.000031s : 182: predicate.switch_simplify 99.82% : 0.225890s : 32: predicate.tile_eliminate 0.00% : 0.000008s : 32: predicate.transpose_eliminate 0.00% : 0.000007s : 40: predicate.tuple_list_convert_item_index_to_positive 0.00% : 0.000006s : 40: predicate.tuple_list_get_item_const_eliminator 0.00% : 0.000006s : 40: predicate.tuple_list_get_item_depend_reorder 0.01% : 0.000012s : 48: predicate.tuple_list_get_item_eliminator 0.00% : 0.000006s : 40: predicate.tuple_list_get_set_item_eliminator 0.00% : 0.000008s : 46: predicate.tuple_list_set_item_eliminator 0.00% : 0.000006s : 42: predicate.tuple_to_list_eliminator_ 0.00% : 0.000011s : 74: predicate.updatestate_pure_node_eliminater 0.01% : 0.000012s : 80: predicate.updatestate_useless_node_eliminater 0.00% : 0.000001s : 3: predicate.value_based_eliminate 0.00% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.00% : 0.000001s : 6: predicate.virtual_output_eliminate 0.00% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.00% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003184 42 54.58% : 0.001738s : 16: func_graph_cloner_run.FuncGraphClonerGraph 2.20% : 0.000070s : 1: func_graph_cloner_run.FuncGraphClonerNode 43.22% : 0.001376s : 25: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.150398 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.40% : 0.004653s : 1: add_attr 0.40% : 0.004636s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000054s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.02% : 0.000202s : 1: auto_monad 0.00% : 0.000022s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.05% : 0.000522s : 1: bootstrap 0.00% : 0.000039s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000016s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000024s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.01% : 0.000092s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000016s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.04% : 0.000503s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.07% : 0.000796s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000020s : 1: opt.transform.mutable_eliminate 19.91% : 0.229048s : 78: opt.transform.opt_a 0.00% : 0.000024s : 1: opt.transform.opt_after_cconv 0.00% : 0.000024s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000086s : 28: opt.transform.opt_b 0.00% : 0.000044s : 2: opt.transform.opt_trans_graph 0.00% : 0.000036s : 4: opt.transform.symbol_engine_opt 20.26% : 0.233031s : 1: opt_a 0.01% : 0.000102s : 1: opt_after_cconv 0.05% : 0.000543s : 1: opt_after_jit_grad 0.02% : 0.000201s : 1: opt_b 20.50% : 0.235787s : 1: optimize 0.00% : 0.000023s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000027s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000008s : 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 0.01% : 0.000097s : 1: pre_auto_parallel 0.00% : 0.000017s : 1: py_interpret_to_execute 0.00% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.01% : 0.000062s : 1: remove_cast_before_assign_add 0.00% : 0.000018s : 1: remove_dup_value 0.13% : 0.001534s : 1: renormalize.infer 0.12% : 0.001347s : 1: renormalize.specialize 0.00% : 0.000005s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000026s : 1: rewriter_after_opt_a 0.04% : 0.000453s : 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% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000079s : 1: symbol_engine_optimizer 0.01% : 0.000073s : 1: tuple_transform 37.89% : 0.435862s : 1: type_inference TotalTime = 0.246621, [21] [bootstrap]: 0.0004593 [type_inference]: 0.199642 [event_method]: 6.525e-05 [auto_monad]: 0.00024899 [graph_reusing]: 1.92e-05 [inline]: 3.28998e-06 [add_attr]: 0.0173503, [1] [add_attr_with_inline]: 0.0173374, [1] [Cycle 1]: 0.00014567, [2] [tag_attr]: 7.98e-05 [meta_addattr_fg_expand]: 1.773e-05 [parallel-infer-symbol]: 3.61999e-06 [pre_auto_parallel]: 0.00010924 [insert-virtual-dataset]: 2.44001e-06 [parallel-infer-symbol-second]: 8.89995e-07 [dataset_repeat_opt]: 1.89e-06 [pipeline_split]: 1.53002e-06 [optimize]: 0.02774, [53] [py_interpret_to_execute]: 7.1e-06 [rewriter_before_opt_a]: 0.00056173 [opt_a]: 0.0251026, [2] [Cycle 1]: 0.0238736, [45] [expand_dump_flag]: 7.05e-06 [switch_simplify]: 0.00030464 [loop_unroll]: 0.00011058 [a_1]: 0.0027245 [with_stream_mark]: 3.359e-05 [recompute_prepare]: 2.378e-05 [updatestate_depend_eliminate]: 9.56998e-06 [updatestate_assign_eliminate]: 6.43003e-06 [updatestate_loads_eliminate]: 6.53e-06 [parameter_eliminate]: 3.25e-06 [a_2]: 0.0002037 [accelerated_algorithm]: 1.33e-05 [shard]: 2.07999e-06 [meta_shard_fg_expand]: 5.67001e-06 [shard_inline]: 1.319e-05 [merge_send_recv]: 1.688e-05 [auto_parallel]: 1.364e-05 [parallel]: 3.788e-05 [flash_sp]: 1.258e-05 [merge_comm]: 9.36e-06 [allreduce_fusion]: 7.46001e-06 [matmul_add_comm_reduction]: 3.187e-05 [allreduce_slice_to_reducescatter]: 1.16002e-06 [virtual_shard_identity]: 1.85e-05 [virtual_dataset]: 1.26e-05 [get_grad_eliminate_]: 1.343e-05 [virtual_output]: 1.312e-05 [merge_forward]: 8.90001e-06 [cell_reuse_recompute_pass]: 2.03002e-06 [offload_activation]: 1.799e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.742e-05 [merge_recompute_call_nodes]: 1.44e-06 [before_grad]: 2.542e-05 [set_forward_comm_id_for_comm_node_pass]: 8.50999e-06 [meta_fg_expand]: 0.00199848 [flash_sp_send_recv_attached]: 3.97e-06 [receive_attached]: 2.24999e-06 [after_resolve]: 6.732e-05 [a_after_grad]: 8.148e-05 [renormalize]: 0.0174427 [add_forward_monad_depend]: 8.47998e-06 [auto_monad_grad]: 3.03e-06 [auto_monad_eliminator]: 1.984e-05 [cse]: 3.405e-05 [a_3]: 5.562e-05 [Cycle 2]: 0.00121049, [45] [expand_dump_flag]: 2.70997e-06 [switch_simplify]: 8.57e-06 [loop_unroll]: 4.34002e-06 [a_1]: 5.851e-05 [with_stream_mark]: 1.712e-05 [recompute_prepare]: 4.68001e-06 [updatestate_depend_eliminate]: 3.97e-06 [updatestate_assign_eliminate]: 2.84001e-06 [updatestate_loads_eliminate]: 3.03e-06 [parameter_eliminate]: 2.05002e-06 [a_2]: 4.106e-05 [accelerated_algorithm]: 5.05999e-06 [shard]: 2.16e-06 [meta_shard_fg_expand]: 2.48e-06 [shard_inline]: 3.74002e-06 [merge_send_recv]: 7.87e-06 [auto_parallel]: 8.33999e-06 [parallel]: 1.053e-05 [flash_sp]: 4.44002e-06 [merge_comm]: 2.62001e-06 [allreduce_fusion]: 2.73e-06 [matmul_add_comm_reduction]: 8.34998e-06 [allreduce_slice_to_reducescatter]: 7.7e-07 [virtual_shard_identity]: 4.80001e-06 [virtual_dataset]: 3.81999e-06 [get_grad_eliminate_]: 3.8e-06 [virtual_output]: 3.81001e-06 [merge_forward]: 3.6e-06 [cell_reuse_recompute_pass]: 3.76001e-06 [offload_activation]: 7.61001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.373e-05 [merge_recompute_call_nodes]: 1.43002e-06 [before_grad]: 8.02e-06 [set_forward_comm_id_for_comm_node_pass]: 2.80997e-06 [meta_fg_expand]: 0.00064209 [flash_sp_send_recv_attached]: 3.56999e-06 [receive_attached]: 2.51e-06 [after_resolve]: 1.364e-05 [a_after_grad]: 6.49001e-06 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 3.78001e-06 [auto_monad_grad]: 1.82001e-06 [auto_monad_eliminator]: 1.045e-05 [cse]: 1.802e-05 [a_3]: 2.341e-05 [py_interpret_to_execute_after_opt_a]: 9.81e-06 [slice_cell_reuse_recomputed_activation]: 2.17999e-06 [rewriter_after_opt_a]: 2.108e-05 [convert_after_rewriter]: 1.29e-06 [order_py_execute_after_rewriter]: 1.49e-06 [mutable_eliminate]: 0.00075615 [opt_b]: 0.00015205, [1] [Cycle 1]: 0.00014178, [7] [b_1]: 7.284e-05 [b_2]: 5.52999e-06 [updatestate_depend_eliminate]: 5.84e-06 [updatestate_assign_eliminate]: 2.03002e-06 [updatestate_loads_eliminate]: 1.57001e-06 [renormalize]: 1.07e-06 [cse]: 1.862e-05 [optimize_parallel_all_gather_comm]: 1.906e-05 [overlap_param_gather]: 2.04e-06 [cconv]: 3.76e-05 [loop_unroll]: 0.00047173 [opt_after_cconv]: 8.105e-05, [1] [Cycle 1]: 7.447e-05, [7] [c_1]: 1.471e-05 [parameter_eliminate]: 4.04002e-06 [updatestate_depend_eliminate]: 5.29e-06 [updatestate_assign_eliminate]: 1.70001e-06 [updatestate_loads_eliminate]: 1.57001e-06 [cse]: 1.584e-05 [renormalize]: 4.59986e-07 [remove_dup_value]: 1.418e-05 [tuple_transform]: 5.225e-05, [1] [Cycle 1]: 4.734e-05, [4] [d_1]: 2.38e-05 [none_parameter_eliminate]: 1.64e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 4.26001e-06 [partial_unused_args_eliminate]: 1.67001e-06 [add_recomputation]: 3.827e-05 [cse_after_recomputation]: 1.774e-05, [1] [Cycle 1]: 1.257e-05, [1] [cse]: 7.36001e-06 [environ_conv]: 4.83001e-06 [swap_dp_allreduce_reducescatter]: 3.56999e-06 [bias_add_comm_swap]: 3.45e-06 [label_micro_interleaved_index]: 4.37e-06 [label_fine_grained_interleaved_index]: 2.42001e-06 [merge_cast_opt]: 1.92001e-06 [slice_recompute_activation]: 2.41e-06 [micro_interleaved_order_control]: 2.58998e-06 [assign_add_opt]: 1.27e-06 [ForceFp32Comm]: 9.39996e-07 [remove_cast_before_assign_add]: 1.25999e-06 [full_micro_interleaved_order_control]: 2.56e-06 [reorder_send_recv_between_fp_bp]: 2.22001e-06 [comm_op_add_attrs]: 1.24998e-06 [add_comm_op_reuse_tag]: 1.00001e-06 [interleave_split_concat_branches]: 1.05999e-06 [interleave_parallel_branches]: 1.00999e-06 [overlap_opt_shard_in_pipeline]: 1.20999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.87999e-06 [control_data_broadcast_order]: 1.051e-05 [grouped_pairwise_exchange_alltoall]: 1.37e-06 [offloading_packed_experts]: 3.89002e-06 [overlap_recompute_and_grad_model_parallel]: 3.87998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.09e-06 [overlap_recompute_allgather_and_fa_grad]: 1.23002e-06 [overlap_recompute_comm]: 2.25002e-06 [overlap_grad_ring_attention]: 3.53e-06 [overlap_grad_flash_sp]: 1.905e-05 [begin_end_overlap_inline]: 4.90021e-07 [split_matmul_comm_elemetwise]: 2.07001e-06 [split_layernorm_comm]: 1.98002e-06 [handle_group_info]: 1.41002e-06 [symbol_engine_optimizer]: 6.285e-05, [1] [Cycle 1]: 5.824e-05, [6] [build]: 2.76999e-06 [elim_shapecalc]: 8.51997e-06 [elim_not_effective]: 9.09e-06 [opt_reshape]: 4.48001e-06 [fold_const_symbol]: 7.21001e-06 [renormalize]: 3.19997e-07 [detach_backward]: 2.27999e-06 [pipeline_parallel_scheduler]: 1.30999e-06 [auto_monad_reorder]: 1.442e-05 [get_jit_bprop_graph]: 2.62001e-06 [rewriter_after_jit_bprop_graph]: 4.82e-06 [opt_after_jit_grad]: 0.00070986 [validate]: 4.147e-05 Sums bootstrap : 0.000459s : 0.20% type_inference : 0.199642s : 87.51% event_method : 0.000065s : 0.03% auto_monad : 0.000249s : 0.11% graph_reusing : 0.000019s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000080s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000018s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000109s : 0.05% 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 : 0.000007s : 0.00% optimize.rewriter_before_opt_a : 0.000562s : 0.25% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000313s : 0.14% optimize.opt_a.loop_unroll : 0.000115s : 0.05% optimize.opt_a.a_1 : 0.002783s : 1.22% optimize.opt_a.with_stream_mark : 0.000051s : 0.02% optimize.opt_a.recompute_prepare : 0.000028s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000014s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000245s : 0.11% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.00% optimize.opt_a.shard_inline : 0.000017s : 0.01% optimize.opt_a.merge_send_recv : 0.000025s : 0.01% optimize.opt_a.auto_parallel : 0.000022s : 0.01% optimize.opt_a.parallel : 0.000048s : 0.02% optimize.opt_a.flash_sp : 0.000017s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000040s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000023s : 0.01% optimize.opt_a.virtual_dataset : 0.000016s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000017s : 0.01% optimize.opt_a.virtual_output : 0.000017s : 0.01% optimize.opt_a.merge_forward : 0.000013s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000041s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000033s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.00% optimize.opt_a.meta_fg_expand : 0.002641s : 1.16% 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 : 0.000081s : 0.04% optimize.opt_a.a_after_grad : 0.000088s : 0.04% optimize.opt_a.renormalize : 0.017443s : 7.65% optimize.opt_a.add_forward_monad_depend : 0.000012s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000030s : 0.01% optimize.opt_a.cse : 0.000052s : 0.02% optimize.opt_a.a_3 : 0.000079s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000010s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000756s : 0.33% optimize.opt_b.b_1 : 0.000073s : 0.03% optimize.opt_b.b_2 : 0.000006s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000019s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000019s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000038s : 0.02% optimize.loop_unroll : 0.000472s : 0.21% optimize.opt_after_cconv.c_1 : 0.000015s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000024s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000038s : 0.02% optimize.cse_after_recomputation.cse : 0.000007s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 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 : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000011s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000019s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000009s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000004s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 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 : 0.000014s : 0.01% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000710s : 0.31% validate : 0.000041s : 0.02% Time group info: ------[substitution.] 0.000876 116 1.93% : 0.000017s : 2: substitution.cast_eliminate 0.17% : 0.000001s : 1: substitution.elim_not_effective 0.67% : 0.000006s : 9: substitution.float_depend_g_call 0.54% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.10% : 0.000001s : 1: substitution.fold_const_symbol 0.54% : 0.000005s : 1: substitution.graph_param_transform 0.33% : 0.000003s : 2: substitution.incorporate_call 0.28% : 0.000002s : 2: substitution.incorporate_call_switch 80.90% : 0.000709s : 33: substitution.inline 2.37% : 0.000021s : 2: substitution.inline_without_move 0.92% : 0.000008s : 9: substitution.j_node_and_user_rematch 0.44% : 0.000004s : 2: substitution.minmaximum_grad 0.76% : 0.000007s : 9: substitution.partial_eliminate 0.97% : 0.000008s : 9: substitution.remove_not_recompute_node 0.26% : 0.000002s : 1: substitution.replace_applicator 0.71% : 0.000006s : 6: substitution.replace_old_param 0.44% : 0.000004s : 1: substitution.set_cell_output_no_recompute 2.48% : 0.000022s : 10: substitution.switch_simplify 1.01% : 0.000009s : 2: substitution.tuple_list_convert_item_index_to_positive 0.36% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.60% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.74% : 0.000024s : 6: substitution.tuple_list_get_item_eliminator 0.48% : 0.000004s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.199418 2 96.86% : 0.193160s : 1: type_inference.infer 3.14% : 0.006257s : 1: type_inference.specialize ------[replace.] 0.000361 47 3.32% : 0.000012s : 2: replace.cast_eliminate 61.80% : 0.000223s : 33: replace.inline 31.12% : 0.000112s : 10: replace.switch_simplify 3.76% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000738 47 2.13% : 0.000016s : 2: match.cast_eliminate 94.05% : 0.000694s : 33: match.inline 2.31% : 0.000017s : 10: match.switch_simplify 1.51% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000494 3131 1.30% : 0.000006s : 44: predicate.accumulaten_eliminater 0.28% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.34% : 0.000002s : 12: predicate.addn_check_dump 1.22% : 0.000006s : 44: predicate.addn_zero_filter 1.13% : 0.000006s : 44: predicate.adjust_all_reduce_mul_add 2.05% : 0.000010s : 56: predicate.arithmetic_simplify 1.52% : 0.000008s : 46: predicate.cast_eliminate 0.13% : 0.000001s : 2: predicate.check_bprop_eliminate 0.43% : 0.000002s : 12: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.39% : 0.000002s : 12: predicate.depend_value_elim 1.35% : 0.000007s : 46: predicate.dict_get_item_const_eliminator 1.44% : 0.000007s : 46: predicate.dict_get_item_eliminator 1.31% : 0.000006s : 46: predicate.dict_set_item_eliminator 0.22% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.03% : 0.000000s : 1: predicate.elim_not_effective 0.20% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.38% : 0.000007s : 47: predicate.environ_add_const_eliminate 1.32% : 0.000006s : 47: predicate.environ_get_add_eliminate 1.24% : 0.000006s : 47: predicate.environ_get_depend_swap 1.66% : 0.000008s : 59: predicate.environ_get_eliminate 1.28% : 0.000006s : 47: predicate.environ_get_set_eliminate 2.35% : 0.000012s : 81: predicate.exchange_switch_depend_value 3.22% : 0.000016s : 81: predicate.float_depend_g_call 0.40% : 0.000002s : 12: predicate.float_environ_get_switch 0.47% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.50% : 0.000002s : 12: predicate.get_grad_eliminate 0.06% : 0.000000s : 1: predicate.graph_param_transform 0.33% : 0.000002s : 12: predicate.incorporate_call 0.31% : 0.000002s : 12: predicate.incorporate_call_switch 6.06% : 0.000030s : 151: predicate.inline 1.44% : 0.000007s : 34: predicate.inline_without_move 0.18% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.61% : 0.000003s : 12: predicate.less_batch_normalization 1.64% : 0.000008s : 50: predicate.list_to_tuple_eliminator_ 2.49% : 0.000012s : 94: predicate.load_eliminater 0.28% : 0.000001s : 1: predicate.loop_unroll_after_grad 3.86% : 0.000019s : 138: predicate.loop_unroll_before_grad 1.37% : 0.000007s : 48: predicate.make_slice_get_slice_eliminator 0.43% : 0.000002s : 12: predicate.merge_addn 0.13% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.11% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.15% : 0.000006s : 44: predicate.minmaximum_grad 0.35% : 0.000002s : 1: predicate.mutable_eliminate 0.13% : 0.000001s : 1: predicate.opt_reshape 0.11% : 0.000001s : 1: predicate.parallel_virtual_node 2.95% : 0.000015s : 81: predicate.partial_defer_inline 1.51% : 0.000007s : 49: predicate.partial_eliminate 1.23% : 0.000006s : 44: predicate.print_const_string_wrapper 0.36% : 0.000002s : 12: predicate.reduce_all_const_elim 1.56% : 0.000008s : 44: predicate.reduce_eliminate 2.68% : 0.000013s : 94: predicate.redundant_stop_gradient_eliminater 0.24% : 0.000001s : 12: predicate.remove_not_recompute_node 1.18% : 0.000006s : 50: predicate.replace_applicator 0.67% : 0.000003s : 34: predicate.replace_old_param 0.05% : 0.000000s : 1: predicate.reset_defer_inline 1.32% : 0.000007s : 44: predicate.reshape_eliminate 0.17% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.11% : 0.000001s : 1: predicate.row_tensor_eliminate 0.30% : 0.000001s : 2: predicate.same_eliminate 0.22% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.61% : 0.000003s : 12: predicate.shard_identity_eliminate 0.11% : 0.000001s : 2: predicate.special_op_eliminate 0.41% : 0.000002s : 12: predicate.specialize_transform 0.55% : 0.000003s : 2: predicate.split_environ_get_set_with_tuple_value 1.35% : 0.000007s : 34: predicate.stack_unstack_eliminate 0.08% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.61% : 0.000013s : 81: predicate.switch_defer_inline 2.77% : 0.000014s : 83: predicate.switch_layer_defer_inline 7.83% : 0.000039s : 252: predicate.switch_simplify 1.21% : 0.000006s : 44: predicate.tile_eliminate 1.25% : 0.000006s : 44: predicate.transpose_eliminate 1.49% : 0.000007s : 48: predicate.tuple_list_convert_item_index_to_positive 1.46% : 0.000007s : 48: predicate.tuple_list_get_item_const_eliminator 1.31% : 0.000006s : 48: predicate.tuple_list_get_item_depend_reorder 2.48% : 0.000012s : 62: predicate.tuple_list_get_item_eliminator 1.38% : 0.000007s : 48: predicate.tuple_list_get_set_item_eliminator 1.88% : 0.000009s : 60: predicate.tuple_list_set_item_eliminator 1.51% : 0.000007s : 50: predicate.tuple_to_list_eliminator_ 2.48% : 0.000012s : 94: predicate.updatestate_pure_node_eliminater 3.14% : 0.000016s : 106: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 1: predicate.value_based_eliminate 0.49% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.41% : 0.000002s : 12: predicate.virtual_output_eliminate 0.04% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.19% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.004316 61 70.18% : 0.003029s : 25: func_graph_cloner_run.FuncGraphClonerGraph 1.48% : 0.000064s : 1: func_graph_cloner_run.FuncGraphClonerNode 28.34% : 0.001223s : 35: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.312966 192 0.00% : 0.000003s : 1: ForceFp32Comm 5.55% : 0.017356s : 1: add_attr 5.54% : 0.017342s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000043s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.08% : 0.000258s : 1: auto_monad 0.01% : 0.000018s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.16% : 0.000495s : 1: bootstrap 0.01% : 0.000041s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000014s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000020s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.02% : 0.000073s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000023s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000005s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.15% : 0.000483s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.25% : 0.000767s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000011s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000014s : 1: opt.transform.mutable_eliminate 1.22% : 0.003804s : 78: opt.transform.opt_a 0.00% : 0.000013s : 1: opt.transform.opt_after_cconv 0.01% : 0.000020s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000048s : 28: opt.transform.opt_b 0.01% : 0.000026s : 2: opt.transform.opt_trans_graph 0.01% : 0.000025s : 4: opt.transform.symbol_engine_opt 8.02% : 0.025107s : 1: opt_a 0.03% : 0.000084s : 1: opt_after_cconv 0.23% : 0.000725s : 1: opt_after_jit_grad 0.05% : 0.000155s : 1: opt_b 8.87% : 0.027746s : 1: optimize 0.01% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000023s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000004s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.04% : 0.000115s : 1: pre_auto_parallel 0.00% : 0.000011s : 1: py_interpret_to_execute 0.00% : 0.000013s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000018s : 1: remove_dup_value 5.31% : 0.016622s : 1: renormalize.infer 0.26% : 0.000807s : 1: renormalize.specialize 0.00% : 0.000005s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000025s : 1: rewriter_after_opt_a 0.18% : 0.000571s : 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% : 0.000006s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000065s : 1: symbol_engine_optimizer 0.02% : 0.000055s : 1: tuple_transform 63.80% : 0.199677s : 1: type_inference TotalTime = 0.283973, [21] [bootstrap]: 0.00051523 [type_inference]: 0.269578 [event_method]: 8.196e-05 [auto_monad]: 0.00018558 [graph_reusing]: 1.116e-05 [inline]: 2.84999e-06 [add_attr]: 0.0042849, [1] [add_attr_with_inline]: 0.00427307, [1] [Cycle 1]: 0.00012167, [2] [tag_attr]: 5.949e-05 [meta_addattr_fg_expand]: 1.129e-05 [parallel-infer-symbol]: 3.56001e-06 [pre_auto_parallel]: 7.593e-05 [insert-virtual-dataset]: 2.89999e-06 [parallel-infer-symbol-second]: 7.39994e-07 [dataset_repeat_opt]: 2.07999e-06 [pipeline_split]: 1.81e-06 [optimize]: 0.00821849, [53] [py_interpret_to_execute]: 6.06003e-06 [rewriter_before_opt_a]: 0.00036344 [opt_a]: 0.00561149, [2] [Cycle 1]: 0.00493701, [45] [expand_dump_flag]: 4.79e-06 [switch_simplify]: 0.00020614 [loop_unroll]: 8.189e-05 [a_1]: 0.00153895 [with_stream_mark]: 2.552e-05 [recompute_prepare]: 1.008e-05 [updatestate_depend_eliminate]: 4.78001e-06 [updatestate_assign_eliminate]: 3.82998e-06 [updatestate_loads_eliminate]: 3.14001e-06 [parameter_eliminate]: 2.07001e-06 [a_2]: 7.277e-05 [accelerated_algorithm]: 7.18998e-06 [shard]: 2.00002e-06 [meta_shard_fg_expand]: 3.16001e-06 [shard_inline]: 6.14999e-06 [merge_send_recv]: 8.36002e-06 [auto_parallel]: 8.86002e-06 [parallel]: 2.089e-05 [flash_sp]: 9.50001e-06 [merge_comm]: 3.51001e-06 [allreduce_fusion]: 3.11001e-06 [matmul_add_comm_reduction]: 9.34998e-06 [allreduce_slice_to_reducescatter]: 8.89995e-07 [virtual_shard_identity]: 8.00999e-06 [virtual_dataset]: 6.08002e-06 [get_grad_eliminate_]: 6.19001e-06 [virtual_output]: 5.96e-06 [merge_forward]: 4.04002e-06 [cell_reuse_recompute_pass]: 1.66e-06 [offload_activation]: 9.81998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.415e-05 [merge_recompute_call_nodes]: 1.76e-06 [before_grad]: 1.015e-05 [set_forward_comm_id_for_comm_node_pass]: 3.31999e-06 [meta_fg_expand]: 4.18999e-06 [flash_sp_send_recv_attached]: 2.55002e-06 [receive_attached]: 2.24001e-06 [after_resolve]: 1.049e-05 [a_after_grad]: 8.45001e-06 [renormalize]: 0.00240256 [add_forward_monad_depend]: 1.046e-05 [auto_monad_grad]: 2.26998e-06 [auto_monad_eliminator]: 1.985e-05 [cse]: 3.377e-05 [a_3]: 5.4e-05 [Cycle 2]: 0.0006595, [45] [expand_dump_flag]: 2.82002e-06 [switch_simplify]: 7.54002e-06 [loop_unroll]: 5.29e-06 [a_1]: 9.967e-05 [with_stream_mark]: 1.653e-05 [recompute_prepare]: 5.53002e-06 [updatestate_depend_eliminate]: 3.63e-06 [updatestate_assign_eliminate]: 3.38999e-06 [updatestate_loads_eliminate]: 2.84999e-06 [parameter_eliminate]: 1.69e-06 [a_2]: 6.049e-05 [accelerated_algorithm]: 5.62999e-06 [shard]: 2.36998e-06 [meta_shard_fg_expand]: 2.16e-06 [shard_inline]: 5.72001e-06 [merge_send_recv]: 6.86999e-06 [auto_parallel]: 8.08001e-06 [parallel]: 8.22998e-06 [flash_sp]: 3.71999e-06 [merge_comm]: 2.99999e-06 [allreduce_fusion]: 3.01001e-06 [matmul_add_comm_reduction]: 8.82999e-06 [allreduce_slice_to_reducescatter]: 7.30011e-07 [virtual_shard_identity]: 6.57002e-06 [virtual_dataset]: 5.56002e-06 [get_grad_eliminate_]: 5.16002e-06 [virtual_output]: 5.14003e-06 [merge_forward]: 3.80998e-06 [cell_reuse_recompute_pass]: 2.24001e-06 [offload_activation]: 9.30001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.497e-05 [merge_recompute_call_nodes]: 1.45001e-06 [before_grad]: 9.35001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.71001e-06 [meta_fg_expand]: 2.37999e-06 [flash_sp_send_recv_attached]: 1.67001e-06 [receive_attached]: 2.08002e-06 [after_resolve]: 9.54999e-06 [a_after_grad]: 7.59002e-06 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 1.60001e-06 [auto_monad_grad]: 1.59e-06 [auto_monad_eliminator]: 6.98998e-06 [cse]: 1.346e-05 [a_3]: 3.024e-05 [py_interpret_to_execute_after_opt_a]: 8.84e-06 [slice_cell_reuse_recomputed_activation]: 2.24001e-06 [rewriter_after_opt_a]: 2.085e-05 [convert_after_rewriter]: 1.22e-06 [order_py_execute_after_rewriter]: 1.39e-06 [mutable_eliminate]: 0.00073653 [opt_b]: 0.00019155, [1] [Cycle 1]: 0.00018243, [7] [b_1]: 0.00010573 [b_2]: 6.93e-06 [updatestate_depend_eliminate]: 6.70002e-06 [updatestate_assign_eliminate]: 2.27999e-06 [updatestate_loads_eliminate]: 2.48e-06 [renormalize]: 9.89996e-07 [cse]: 2.124e-05 [optimize_parallel_all_gather_comm]: 1.967e-05 [overlap_param_gather]: 2.43e-06 [cconv]: 3.164e-05 [loop_unroll]: 0.00046092 [opt_after_cconv]: 9.781e-05, [1] [Cycle 1]: 9.099e-05, [7] [c_1]: 2.534e-05 [parameter_eliminate]: 4.57998e-06 [updatestate_depend_eliminate]: 6.31e-06 [updatestate_assign_eliminate]: 2.41e-06 [updatestate_loads_eliminate]: 2.01e-06 [cse]: 1.926e-05 [renormalize]: 4.60015e-07 [remove_dup_value]: 1.715e-05 [tuple_transform]: 7.135e-05, [1] [Cycle 1]: 6.613e-05, [4] [d_1]: 3.965e-05 [none_parameter_eliminate]: 1.66998e-06 [renormalize]: 1.79978e-07 [switch_simplify]: 7.11001e-06 [partial_unused_args_eliminate]: 1.67001e-06 [add_recomputation]: 5.58e-05 [cse_after_recomputation]: 2.268e-05, [1] [Cycle 1]: 1.716e-05, [1] [cse]: 1.206e-05 [environ_conv]: 5.76e-06 [swap_dp_allreduce_reducescatter]: 6.31e-06 [bias_add_comm_swap]: 3.13e-06 [label_micro_interleaved_index]: 7.31001e-06 [label_fine_grained_interleaved_index]: 2.86999e-06 [merge_cast_opt]: 1.55999e-06 [slice_recompute_activation]: 2.16e-06 [micro_interleaved_order_control]: 3.54002e-06 [assign_add_opt]: 1.32999e-06 [ForceFp32Comm]: 1.17e-06 [remove_cast_before_assign_add]: 9.80013e-07 [full_micro_interleaved_order_control]: 2.69999e-06 [reorder_send_recv_between_fp_bp]: 2.74001e-06 [comm_op_add_attrs]: 1.26002e-06 [add_comm_op_reuse_tag]: 9.89996e-07 [interleave_split_concat_branches]: 1.19e-06 [interleave_parallel_branches]: 1.04e-06 [overlap_opt_shard_in_pipeline]: 1.40001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.84998e-06 [control_data_broadcast_order]: 1.326e-05 [grouped_pairwise_exchange_alltoall]: 1.96998e-06 [offloading_packed_experts]: 4.47e-06 [overlap_recompute_and_grad_model_parallel]: 4.68001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.29e-06 [overlap_recompute_allgather_and_fa_grad]: 1.52999e-06 [overlap_recompute_comm]: 2.11998e-06 [overlap_grad_ring_attention]: 4.02998e-06 [overlap_grad_flash_sp]: 2.196e-05 [begin_end_overlap_inline]: 8.59989e-07 [split_matmul_comm_elemetwise]: 2.40002e-06 [split_layernorm_comm]: 1.68002e-06 [handle_group_info]: 1.44e-06 [symbol_engine_optimizer]: 7.469e-05, [1] [Cycle 1]: 7.016e-05, [6] [build]: 3.4e-06 [elim_shapecalc]: 1.08e-05 [elim_not_effective]: 1.202e-05 [opt_reshape]: 6.89001e-06 [fold_const_symbol]: 9.47001e-06 [renormalize]: 1.8999e-07 [detach_backward]: 2.23998e-06 [pipeline_parallel_scheduler]: 1.97999e-06 [auto_monad_reorder]: 1.807e-05 [get_jit_bprop_graph]: 2.21e-06 [rewriter_after_jit_bprop_graph]: 6.88e-06 [opt_after_jit_grad]: 0.000714 [validate]: 5.124e-05 Sums bootstrap : 0.000515s : 0.18% type_inference : 0.269578s : 96.78% event_method : 0.000082s : 0.03% auto_monad : 0.000186s : 0.07% graph_reusing : 0.000011s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000059s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000011s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000076s : 0.03% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000006s : 0.00% optimize.rewriter_before_opt_a : 0.000363s : 0.13% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000214s : 0.08% optimize.opt_a.loop_unroll : 0.000087s : 0.03% optimize.opt_a.a_1 : 0.001639s : 0.59% optimize.opt_a.with_stream_mark : 0.000042s : 0.02% optimize.opt_a.recompute_prepare : 0.000016s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000133s : 0.05% optimize.opt_a.accelerated_algorithm : 0.000013s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.00% optimize.opt_a.shard_inline : 0.000012s : 0.00% optimize.opt_a.merge_send_recv : 0.000015s : 0.01% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000029s : 0.01% optimize.opt_a.flash_sp : 0.000013s : 0.00% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000006s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000018s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000015s : 0.01% optimize.opt_a.virtual_dataset : 0.000012s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.00% optimize.opt_a.virtual_output : 0.000011s : 0.00% optimize.opt_a.merge_forward : 0.000008s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000019s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000029s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000020s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.00% optimize.opt_a.meta_fg_expand : 0.000007s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000020s : 0.01% optimize.opt_a.a_after_grad : 0.000016s : 0.01% optimize.opt_a.renormalize : 0.002403s : 0.86% optimize.opt_a.add_forward_monad_depend : 0.000012s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000027s : 0.01% optimize.opt_a.cse : 0.000047s : 0.02% optimize.opt_a.a_3 : 0.000084s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000009s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000737s : 0.26% optimize.opt_b.b_1 : 0.000106s : 0.04% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000021s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000032s : 0.01% optimize.loop_unroll : 0.000461s : 0.17% optimize.opt_after_cconv.c_1 : 0.000025s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000019s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000040s : 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 : 0.000007s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000056s : 0.02% optimize.cse_after_recomputation.cse : 0.000012s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 0.00% 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 : 0.000004s : 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 : 0.000013s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000022s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000012s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000018s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000714s : 0.26% validate : 0.000051s : 0.02% Time group info: ------[substitution.] 0.000486 46 2.64% : 0.000013s : 1: substitution.cast_eliminate 0.39% : 0.000002s : 2: substitution.elim_not_effective 0.44% : 0.000002s : 2: substitution.fold_const_symbol 1.22% : 0.000006s : 3: substitution.graph_param_transform 86.31% : 0.000420s : 20: substitution.inline 0.91% : 0.000004s : 4: substitution.j_node_and_user_rematch 1.18% : 0.000006s : 4: substitution.remove_not_recompute_node 0.99% : 0.000005s : 2: substitution.replace_old_param 3.52% : 0.000017s : 6: substitution.switch_simplify 2.39% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.269427 2 74.41% : 0.200482s : 1: type_inference.infer 25.59% : 0.068945s : 1: type_inference.specialize ------[replace.] 0.000216 29 2.81% : 0.000006s : 1: replace.cast_eliminate 55.29% : 0.000119s : 20: replace.inline 34.88% : 0.000075s : 6: replace.switch_simplify 7.02% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000446 29 2.68% : 0.000012s : 1: match.cast_eliminate 92.05% : 0.000410s : 20: match.inline 2.92% : 0.000013s : 6: match.switch_simplify 2.35% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000325 2021 1.37% : 0.000004s : 27: predicate.accumulaten_eliminater 0.46% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.25% : 0.000001s : 6: predicate.addn_check_dump 1.20% : 0.000004s : 27: predicate.addn_zero_filter 1.10% : 0.000004s : 27: predicate.adjust_all_reduce_mul_add 2.16% : 0.000007s : 33: predicate.arithmetic_simplify 1.55% : 0.000005s : 28: predicate.cast_eliminate 0.29% : 0.000001s : 6: predicate.check_bprop_eliminate 0.27% : 0.000001s : 6: predicate.compare_switch_simplify 0.08% : 0.000000s : 3: predicate.const_output_eliminate 0.40% : 0.000001s : 6: predicate.depend_value_elim 1.27% : 0.000004s : 28: predicate.dict_get_item_const_eliminator 1.35% : 0.000004s : 28: predicate.dict_get_item_eliminator 1.25% : 0.000004s : 28: predicate.dict_set_item_eliminator 0.68% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.12% : 0.000000s : 3: predicate.elim_not_effective 0.28% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.50% : 0.000005s : 31: predicate.environ_add_const_eliminate 1.27% : 0.000004s : 31: predicate.environ_get_add_eliminate 1.28% : 0.000004s : 31: predicate.environ_get_depend_swap 1.68% : 0.000005s : 37: predicate.environ_get_eliminate 1.29% : 0.000004s : 31: predicate.environ_get_set_eliminate 2.14% : 0.000007s : 50: predicate.exchange_switch_depend_value 3.30% : 0.000011s : 50: predicate.float_depend_g_call 0.25% : 0.000001s : 6: predicate.float_environ_get_switch 0.38% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 3: predicate.fold_const_symbol 0.37% : 0.000001s : 6: predicate.get_grad_eliminate 0.19% : 0.000001s : 3: predicate.graph_param_transform 0.30% : 0.000001s : 6: predicate.incorporate_call 0.24% : 0.000001s : 6: predicate.incorporate_call_switch 6.02% : 0.000020s : 95: predicate.inline 0.34% : 0.000001s : 6: predicate.inline_without_move 0.15% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.47% : 0.000002s : 6: predicate.less_batch_normalization 1.84% : 0.000006s : 36: predicate.list_to_tuple_eliminator_ 2.57% : 0.000008s : 63: predicate.load_eliminater 0.49% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.61% : 0.000012s : 87: predicate.loop_unroll_before_grad 1.60% : 0.000005s : 34: predicate.make_slice_get_slice_eliminator 0.26% : 0.000001s : 6: predicate.merge_addn 0.28% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.33% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.11% : 0.000004s : 27: predicate.minmaximum_grad 0.81% : 0.000003s : 3: predicate.mutable_eliminate 0.23% : 0.000001s : 3: predicate.opt_reshape 0.26% : 0.000001s : 3: predicate.parallel_virtual_node 2.63% : 0.000009s : 50: predicate.partial_defer_inline 1.52% : 0.000005s : 33: predicate.partial_eliminate 1.13% : 0.000004s : 27: predicate.print_const_string_wrapper 0.26% : 0.000001s : 6: predicate.reduce_all_const_elim 1.46% : 0.000005s : 27: predicate.reduce_eliminate 2.67% : 0.000009s : 63: predicate.redundant_stop_gradient_eliminater 0.31% : 0.000001s : 6: predicate.remove_not_recompute_node 1.28% : 0.000004s : 36: predicate.replace_applicator 0.27% : 0.000001s : 6: predicate.replace_old_param 0.20% : 0.000001s : 3: predicate.reset_defer_inline 1.23% : 0.000004s : 27: predicate.reshape_eliminate 0.34% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.26% : 0.000001s : 3: predicate.row_tensor_eliminate 0.39% : 0.000001s : 6: predicate.same_eliminate 0.20% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.53% : 0.000002s : 6: predicate.shard_identity_eliminate 0.36% : 0.000001s : 6: predicate.special_op_eliminate 0.39% : 0.000001s : 6: predicate.specialize_transform 0.57% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.39% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.15% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.36% : 0.000008s : 50: predicate.switch_defer_inline 2.75% : 0.000009s : 56: predicate.switch_layer_defer_inline 7.48% : 0.000024s : 158: predicate.switch_simplify 1.15% : 0.000004s : 27: predicate.tile_eliminate 1.13% : 0.000004s : 27: predicate.transpose_eliminate 1.62% : 0.000005s : 34: predicate.tuple_list_convert_item_index_to_positive 1.56% : 0.000005s : 34: predicate.tuple_list_get_item_const_eliminator 1.73% : 0.000006s : 34: predicate.tuple_list_get_item_depend_reorder 2.82% : 0.000009s : 42: predicate.tuple_list_get_item_eliminator 1.60% : 0.000005s : 34: predicate.tuple_list_get_set_item_eliminator 2.01% : 0.000007s : 40: predicate.tuple_list_set_item_eliminator 1.67% : 0.000005s : 36: predicate.tuple_to_list_eliminator_ 2.55% : 0.000008s : 63: predicate.updatestate_pure_node_eliminater 2.95% : 0.000010s : 69: predicate.updatestate_useless_node_eliminater 0.23% : 0.000001s : 3: predicate.value_based_eliminate 0.40% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.37% : 0.000001s : 6: predicate.virtual_output_eliminate 0.14% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.21% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002666 36 51.28% : 0.001367s : 14: func_graph_cloner_run.FuncGraphClonerGraph 48.72% : 0.001299s : 22: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.301132 192 0.00% : 0.000004s : 1: ForceFp32Comm 1.43% : 0.004291s : 1: add_attr 1.42% : 0.004277s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000060s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.06% : 0.000195s : 1: auto_monad 0.01% : 0.000022s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.18% : 0.000553s : 1: bootstrap 0.01% : 0.000035s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000017s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000026s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.03% : 0.000092s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000015s : 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% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000003s : 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% : 0.000471s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.25% : 0.000747s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000013s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000016s : 1: opt.transform.mutable_eliminate 0.74% : 0.002240s : 78: opt.transform.opt_a 0.01% : 0.000024s : 1: opt.transform.opt_after_cconv 0.01% : 0.000031s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000085s : 28: opt.transform.opt_b 0.01% : 0.000045s : 2: opt.transform.opt_trans_graph 0.01% : 0.000036s : 4: opt.transform.symbol_engine_opt 1.86% : 0.005616s : 1: opt_a 0.03% : 0.000102s : 1: opt_after_cconv 0.24% : 0.000730s : 1: opt_after_jit_grad 0.06% : 0.000195s : 1: opt_b 2.73% : 0.008226s : 1: optimize 0.01% : 0.000023s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000025s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 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% : 0.000005s : 1: pipeline_split 0.03% : 0.000081s : 1: pre_auto_parallel 0.00% : 0.000010s : 1: py_interpret_to_execute 0.00% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.01% : 0.000021s : 1: remove_dup_value 0.47% : 0.001427s : 1: renormalize.infer 0.32% : 0.000963s : 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% : 0.000024s : 1: rewriter_after_opt_a 0.12% : 0.000370s : 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% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.03% : 0.000077s : 1: symbol_engine_optimizer 0.02% : 0.000074s : 1: tuple_transform 89.53% : 0.269618s : 1: type_inference TotalTime = 0.234662, [21] [bootstrap]: 0.00051541 [type_inference]: 0.201924 [event_method]: 6.752e-05 [auto_monad]: 0.00026816 [graph_reusing]: 1.859e-05 [inline]: 3.18998e-06 [add_attr]: 0.0040949, [1] [add_attr_with_inline]: 0.00408272, [1] [Cycle 1]: 0.00014887, [2] [tag_attr]: 8.045e-05 [meta_addattr_fg_expand]: 1.63e-05 [parallel-infer-symbol]: 4.17e-06 [pre_auto_parallel]: 0.00010091 [insert-virtual-dataset]: 3.21999e-06 [parallel-infer-symbol-second]: 7.50006e-07 [dataset_repeat_opt]: 2.12999e-06 [pipeline_split]: 1.92999e-06 [optimize]: 0.0268328, [53] [py_interpret_to_execute]: 9.62999e-06 [rewriter_before_opt_a]: 0.00050691 [opt_a]: 0.0242234, [2] [Cycle 1]: 0.0233797, [45] [expand_dump_flag]: 6.28e-06 [switch_simplify]: 0.0126027 [loop_unroll]: 0.00012003 [a_1]: 0.00238243 [with_stream_mark]: 3.555e-05 [recompute_prepare]: 2.259e-05 [updatestate_depend_eliminate]: 8.89e-06 [updatestate_assign_eliminate]: 6.58998e-06 [updatestate_loads_eliminate]: 6.76999e-06 [parameter_eliminate]: 3.60998e-06 [a_2]: 0.00020031 [accelerated_algorithm]: 1.382e-05 [shard]: 1.95001e-06 [meta_shard_fg_expand]: 7.26001e-06 [shard_inline]: 1.27e-05 [merge_send_recv]: 1.732e-05 [auto_parallel]: 1.41e-05 [parallel]: 2.068e-05 [flash_sp]: 1.296e-05 [merge_comm]: 8.08001e-06 [allreduce_fusion]: 7.82e-06 [matmul_add_comm_reduction]: 3.39e-05 [allreduce_slice_to_reducescatter]: 9.39996e-07 [virtual_shard_identity]: 1.601e-05 [virtual_dataset]: 1.344e-05 [get_grad_eliminate_]: 1.271e-05 [virtual_output]: 1.266e-05 [merge_forward]: 8.52e-06 [cell_reuse_recompute_pass]: 1.11002e-06 [offload_activation]: 1.679e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.729e-05 [merge_recompute_call_nodes]: 1.66e-06 [before_grad]: 2.479e-05 [set_forward_comm_id_for_comm_node_pass]: 9.00999e-06 [meta_fg_expand]: 0.0021195 [flash_sp_send_recv_attached]: 4.17e-06 [receive_attached]: 1.99999e-06 [after_resolve]: 6.873e-05 [a_after_grad]: 8.022e-05 [renormalize]: 0.00493244 [add_forward_monad_depend]: 7.53e-06 [auto_monad_grad]: 2.51e-06 [auto_monad_eliminator]: 1.534e-05 [cse]: 2.917e-05 [a_3]: 4.53e-05 [Cycle 2]: 0.00082614, [45] [expand_dump_flag]: 2.17999e-06 [switch_simplify]: 6.08002e-06 [loop_unroll]: 4.39002e-06 [a_1]: 5.008e-05 [with_stream_mark]: 1.407e-05 [recompute_prepare]: 3.93999e-06 [updatestate_depend_eliminate]: 3.41001e-06 [updatestate_assign_eliminate]: 2.62001e-06 [updatestate_loads_eliminate]: 2.71e-06 [parameter_eliminate]: 1.90001e-06 [a_2]: 3.943e-05 [accelerated_algorithm]: 3.80998e-06 [shard]: 2.19999e-06 [meta_shard_fg_expand]: 2.39999e-06 [shard_inline]: 3.36999e-06 [merge_send_recv]: 6.94001e-06 [auto_parallel]: 8.03999e-06 [parallel]: 8.99998e-06 [flash_sp]: 3.93999e-06 [merge_comm]: 2.41998e-06 [allreduce_fusion]: 2.73e-06 [matmul_add_comm_reduction]: 7.38e-06 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 4.35e-06 [virtual_dataset]: 3.41001e-06 [get_grad_eliminate_]: 3.33e-06 [virtual_output]: 3.20998e-06 [merge_forward]: 3.21999e-06 [cell_reuse_recompute_pass]: 2.68003e-06 [offload_activation]: 8.11002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.123e-05 [merge_recompute_call_nodes]: 1.50999e-06 [before_grad]: 6.52001e-06 [set_forward_comm_id_for_comm_node_pass]: 2.61e-06 [meta_fg_expand]: 0.00030731 [flash_sp_send_recv_attached]: 2.14e-06 [receive_attached]: 2.59999e-06 [after_resolve]: 8.62e-06 [a_after_grad]: 6.10002e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.42001e-06 [auto_monad_grad]: 2.04e-06 [auto_monad_eliminator]: 8.05e-06 [cse]: 1.346e-05 [a_3]: 2.353e-05 [py_interpret_to_execute_after_opt_a]: 6.31e-06 [slice_cell_reuse_recomputed_activation]: 2.15002e-06 [rewriter_after_opt_a]: 1.868e-05 [convert_after_rewriter]: 1.13001e-06 [order_py_execute_after_rewriter]: 1.12999e-06 [mutable_eliminate]: 0.00072727 [opt_b]: 0.00013848, [1] [Cycle 1]: 0.00012984, [7] [b_1]: 6.722e-05 [b_2]: 4.82998e-06 [updatestate_depend_eliminate]: 5.47001e-06 [updatestate_assign_eliminate]: 1.64998e-06 [updatestate_loads_eliminate]: 1.66002e-06 [renormalize]: 7.30011e-07 [cse]: 1.589e-05 [optimize_parallel_all_gather_comm]: 1.536e-05 [overlap_param_gather]: 1.99e-06 [cconv]: 2.772e-05 [loop_unroll]: 0.00045018 [opt_after_cconv]: 8.002e-05, [1] [Cycle 1]: 7.381e-05, [7] [c_1]: 1.41e-05 [parameter_eliminate]: 2.87002e-06 [updatestate_depend_eliminate]: 4.47e-06 [updatestate_assign_eliminate]: 1.78002e-06 [updatestate_loads_eliminate]: 1.84e-06 [cse]: 1.64e-05 [renormalize]: 3.00002e-07 [remove_dup_value]: 1.434e-05 [tuple_transform]: 5.109e-05, [1] [Cycle 1]: 4.597e-05, [4] [d_1]: 2.292e-05 [none_parameter_eliminate]: 1.75001e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 4.4e-06 [partial_unused_args_eliminate]: 2.27999e-06 [add_recomputation]: 3.58e-05 [cse_after_recomputation]: 1.913e-05, [1] [Cycle 1]: 1.508e-05, [1] [cse]: 9.61e-06 [environ_conv]: 5.09e-06 [swap_dp_allreduce_reducescatter]: 3.83999e-06 [bias_add_comm_swap]: 3.43999e-06 [label_micro_interleaved_index]: 4.97e-06 [label_fine_grained_interleaved_index]: 2.66999e-06 [merge_cast_opt]: 1.40999e-06 [slice_recompute_activation]: 2.36e-06 [micro_interleaved_order_control]: 3.29001e-06 [assign_add_opt]: 1.50001e-06 [ForceFp32Comm]: 1.33002e-06 [remove_cast_before_assign_add]: 1.15001e-06 [full_micro_interleaved_order_control]: 2.11e-06 [reorder_send_recv_between_fp_bp]: 2.88003e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.15999e-06 [interleave_parallel_branches]: 1.05001e-06 [overlap_opt_shard_in_pipeline]: 1.35999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.96e-06 [control_data_broadcast_order]: 1.002e-05 [grouped_pairwise_exchange_alltoall]: 1.76e-06 [offloading_packed_experts]: 3.7e-06 [overlap_recompute_and_grad_model_parallel]: 4.33001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.25999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.20999e-06 [overlap_recompute_comm]: 2.48e-06 [overlap_grad_ring_attention]: 3.38e-06 [overlap_grad_flash_sp]: 1.813e-05 [begin_end_overlap_inline]: 4.60015e-07 [split_matmul_comm_elemetwise]: 2.46e-06 [split_layernorm_comm]: 1.78002e-06 [handle_group_info]: 1.05999e-06 [symbol_engine_optimizer]: 0.00017508, [1] [Cycle 1]: 0.00017007, [6] [build]: 3.23998e-06 [elim_shapecalc]: 8.99998e-06 [elim_not_effective]: 1.073e-05 [opt_reshape]: 0.0001078 [fold_const_symbol]: 8.80001e-06 [renormalize]: 1.50001e-07 [detach_backward]: 2.54001e-06 [pipeline_parallel_scheduler]: 1.57001e-06 [auto_monad_reorder]: 1.597e-05 [get_jit_bprop_graph]: 2.34001e-06 [rewriter_after_jit_bprop_graph]: 5.63002e-06 [opt_after_jit_grad]: 0.00056249 [validate]: 3.823e-05 Sums bootstrap : 0.000515s : 0.22% type_inference : 0.201924s : 87.99% event_method : 0.000068s : 0.03% auto_monad : 0.000268s : 0.12% graph_reusing : 0.000019s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000080s : 0.04% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000016s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000101s : 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 : 0.000010s : 0.00% optimize.rewriter_before_opt_a : 0.000507s : 0.22% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.012609s : 5.49% optimize.opt_a.loop_unroll : 0.000124s : 0.05% optimize.opt_a.a_1 : 0.002433s : 1.06% optimize.opt_a.with_stream_mark : 0.000050s : 0.02% optimize.opt_a.recompute_prepare : 0.000027s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000240s : 0.10% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.01% optimize.opt_a.merge_send_recv : 0.000024s : 0.01% optimize.opt_a.auto_parallel : 0.000022s : 0.01% optimize.opt_a.parallel : 0.000030s : 0.01% optimize.opt_a.flash_sp : 0.000017s : 0.01% optimize.opt_a.merge_comm : 0.000010s : 0.00% optimize.opt_a.allreduce_fusion : 0.000011s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000041s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000020s : 0.01% optimize.opt_a.virtual_dataset : 0.000017s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.01% optimize.opt_a.virtual_output : 0.000016s : 0.01% optimize.opt_a.merge_forward : 0.000012s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000025s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000039s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000031s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.01% optimize.opt_a.meta_fg_expand : 0.002427s : 1.06% optimize.opt_a.flash_sp_send_recv_attached : 0.000006s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000077s : 0.03% optimize.opt_a.a_after_grad : 0.000086s : 0.04% optimize.opt_a.renormalize : 0.004933s : 2.15% optimize.opt_a.add_forward_monad_depend : 0.000010s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000023s : 0.01% optimize.opt_a.cse : 0.000043s : 0.02% optimize.opt_a.a_3 : 0.000069s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000019s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000727s : 0.32% optimize.opt_b.b_1 : 0.000067s : 0.03% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000016s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000015s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000028s : 0.01% optimize.loop_unroll : 0.000450s : 0.20% optimize.opt_after_cconv.c_1 : 0.000014s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000023s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000036s : 0.02% optimize.cse_after_recomputation.cse : 0.000010s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000010s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000018s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000108s : 0.05% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000016s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000562s : 0.25% validate : 0.000038s : 0.02% Time group info: ------[substitution.] 0.000758 110 1.40% : 0.000011s : 1: substitution.cast_eliminate 0.20% : 0.000002s : 1: substitution.elim_not_effective 0.87% : 0.000007s : 9: substitution.float_depend_g_call 0.55% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.21% : 0.000002s : 1: substitution.fold_const_symbol 0.57% : 0.000004s : 1: substitution.graph_param_transform 0.45% : 0.000003s : 2: substitution.incorporate_call 0.31% : 0.000002s : 2: substitution.incorporate_call_switch 76.41% : 0.000580s : 29: substitution.inline 2.74% : 0.000021s : 2: substitution.inline_without_move 1.03% : 0.000008s : 9: substitution.j_node_and_user_rematch 0.58% : 0.000004s : 2: substitution.minmaximum_grad 0.84% : 0.000006s : 9: substitution.partial_eliminate 1.09% : 0.000008s : 9: substitution.remove_not_recompute_node 0.35% : 0.000003s : 1: substitution.replace_applicator 0.74% : 0.000006s : 6: substitution.replace_old_param 0.36% : 0.000003s : 1: substitution.set_cell_output_no_recompute 5.20% : 0.000039s : 9: substitution.switch_simplify 1.06% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.46% : 0.000004s : 2: substitution.tuple_list_get_item_const_eliminator 0.67% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 3.28% : 0.000025s : 6: substitution.tuple_list_get_item_eliminator 0.63% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.201681 2 97.07% : 0.195772s : 1: type_inference.infer 2.93% : 0.005909s : 1: type_inference.specialize ------[replace.] 0.000367 41 1.69% : 0.000006s : 1: replace.cast_eliminate 49.57% : 0.000182s : 29: replace.inline 44.89% : 0.000165s : 9: replace.switch_simplify 3.85% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000623 41 1.61% : 0.000010s : 1: match.cast_eliminate 90.83% : 0.000566s : 29: match.inline 5.31% : 0.000033s : 9: match.switch_simplify 2.25% : 0.000014s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000478 2862 1.27% : 0.000006s : 40: predicate.accumulaten_eliminater 0.27% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.33% : 0.000002s : 12: predicate.addn_check_dump 1.24% : 0.000006s : 40: predicate.addn_zero_filter 1.08% : 0.000005s : 40: predicate.adjust_all_reduce_mul_add 2.22% : 0.000011s : 52: predicate.arithmetic_simplify 1.36% : 0.000007s : 41: predicate.cast_eliminate 0.09% : 0.000000s : 2: predicate.check_bprop_eliminate 0.32% : 0.000002s : 12: predicate.compare_switch_simplify 0.03% : 0.000000s : 1: predicate.const_output_eliminate 0.35% : 0.000002s : 12: predicate.depend_value_elim 1.23% : 0.000006s : 41: predicate.dict_get_item_const_eliminator 1.37% : 0.000007s : 41: predicate.dict_get_item_eliminator 1.17% : 0.000006s : 41: predicate.dict_set_item_eliminator 0.27% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 1: predicate.elim_not_effective 0.09% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.25% : 0.000006s : 42: predicate.environ_add_const_eliminate 1.19% : 0.000006s : 42: predicate.environ_get_add_eliminate 1.17% : 0.000006s : 42: predicate.environ_get_depend_swap 1.61% : 0.000008s : 54: predicate.environ_get_eliminate 1.17% : 0.000006s : 42: predicate.environ_get_set_eliminate 2.15% : 0.000010s : 72: predicate.exchange_switch_depend_value 2.97% : 0.000014s : 72: predicate.float_depend_g_call 0.36% : 0.000002s : 12: predicate.float_environ_get_switch 0.42% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.44% : 0.000002s : 12: predicate.get_grad_eliminate 0.07% : 0.000000s : 1: predicate.graph_param_transform 0.35% : 0.000002s : 12: predicate.incorporate_call 0.32% : 0.000002s : 12: predicate.incorporate_call_switch 5.81% : 0.000028s : 138: predicate.inline 1.40% : 0.000007s : 34: predicate.inline_without_move 0.19% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.63% : 0.000003s : 12: predicate.less_batch_normalization 1.39% : 0.000007s : 45: predicate.list_to_tuple_eliminator_ 2.34% : 0.000011s : 85: predicate.load_eliminater 0.26% : 0.000001s : 1: predicate.loop_unroll_after_grad 5.01% : 0.000024s : 126: predicate.loop_unroll_before_grad 1.31% : 0.000006s : 43: predicate.make_slice_get_slice_eliminator 0.38% : 0.000002s : 12: predicate.merge_addn 0.10% : 0.000000s : 2: predicate.micro_step_allgather_replace 0.11% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.09% : 0.000005s : 40: predicate.minmaximum_grad 0.50% : 0.000002s : 1: predicate.mutable_eliminate 0.09% : 0.000000s : 1: predicate.opt_reshape 0.10% : 0.000000s : 1: predicate.parallel_virtual_node 2.77% : 0.000013s : 72: predicate.partial_defer_inline 1.48% : 0.000007s : 44: predicate.partial_eliminate 1.19% : 0.000006s : 40: predicate.print_const_string_wrapper 0.39% : 0.000002s : 12: predicate.reduce_all_const_elim 1.43% : 0.000007s : 40: predicate.reduce_eliminate 2.41% : 0.000012s : 85: predicate.redundant_stop_gradient_eliminater 0.27% : 0.000001s : 12: predicate.remove_not_recompute_node 1.19% : 0.000006s : 45: predicate.replace_applicator 0.72% : 0.000003s : 34: predicate.replace_old_param 0.06% : 0.000000s : 1: predicate.reset_defer_inline 1.24% : 0.000006s : 40: predicate.reshape_eliminate 0.19% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.10% : 0.000000s : 1: predicate.row_tensor_eliminate 0.24% : 0.000001s : 2: predicate.same_eliminate 0.24% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.48% : 0.000002s : 12: predicate.shard_identity_eliminate 0.09% : 0.000000s : 2: predicate.special_op_eliminate 0.40% : 0.000002s : 12: predicate.specialize_transform 0.37% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 1.29% : 0.000006s : 34: predicate.stack_unstack_eliminate 0.07% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.38% : 0.000011s : 72: predicate.switch_defer_inline 2.45% : 0.000012s : 74: predicate.switch_layer_defer_inline 12.04% : 0.000058s : 229: predicate.switch_simplify 1.23% : 0.000006s : 40: predicate.tile_eliminate 1.14% : 0.000005s : 40: predicate.transpose_eliminate 1.31% : 0.000006s : 43: predicate.tuple_list_convert_item_index_to_positive 1.40% : 0.000007s : 43: predicate.tuple_list_get_item_const_eliminator 1.23% : 0.000006s : 43: predicate.tuple_list_get_item_depend_reorder 2.34% : 0.000011s : 57: predicate.tuple_list_get_item_eliminator 1.34% : 0.000006s : 43: predicate.tuple_list_get_set_item_eliminator 1.90% : 0.000009s : 55: predicate.tuple_list_set_item_eliminator 1.44% : 0.000007s : 45: predicate.tuple_to_list_eliminator_ 2.30% : 0.000011s : 85: predicate.updatestate_pure_node_eliminater 2.86% : 0.000014s : 97: predicate.updatestate_useless_node_eliminater 0.06% : 0.000000s : 1: predicate.value_based_eliminate 0.49% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.41% : 0.000002s : 12: predicate.virtual_output_eliminate 0.03% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.12% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003997 55 71.17% : 0.002845s : 23: func_graph_cloner_run.FuncGraphClonerGraph 28.83% : 0.001152s : 32: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.286358 192 0.00% : 0.000004s : 1: ForceFp32Comm 1.43% : 0.004101s : 1: add_attr 1.43% : 0.004087s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000040s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.10% : 0.000278s : 1: auto_monad 0.01% : 0.000020s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.19% : 0.000549s : 1: bootstrap 0.01% : 0.000031s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000013s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.03% : 0.000077s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000022s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.16% : 0.000459s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000007s : 1: micro_interleaved_order_control 0.26% : 0.000737s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000010s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000013s : 1: opt.transform.mutable_eliminate 5.49% : 0.015730s : 78: opt.transform.opt_a 0.00% : 0.000013s : 1: opt.transform.opt_after_cconv 0.01% : 0.000016s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000044s : 28: opt.transform.opt_b 0.01% : 0.000025s : 2: opt.transform.opt_trans_graph 0.05% : 0.000131s : 4: opt.transform.symbol_engine_opt 8.46% : 0.024227s : 1: opt_a 0.03% : 0.000083s : 1: opt_after_cconv 0.20% : 0.000572s : 1: opt_after_jit_grad 0.05% : 0.000142s : 1: opt_b 9.37% : 0.026839s : 1: optimize 0.01% : 0.000019s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000022s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000005s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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 0.04% : 0.000106s : 1: pre_auto_parallel 0.00% : 0.000013s : 1: py_interpret_to_execute 0.00% : 0.000009s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000018s : 1: remove_dup_value 1.46% : 0.004188s : 1: renormalize.infer 0.26% : 0.000733s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000022s : 1: rewriter_after_opt_a 0.18% : 0.000516s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.06% : 0.000178s : 1: symbol_engine_optimizer 0.02% : 0.000054s : 1: tuple_transform 70.53% : 0.201964s : 1: type_inference TotalTime = 0.274794, [21] [bootstrap]: 0.00043681 [type_inference]: 0.236408 [event_method]: 0.00048863 [auto_monad]: 0.0124146 [graph_reusing]: 1.285e-05 [inline]: 3.71001e-06 [add_attr]: 0.00442014, [1] [add_attr_with_inline]: 0.00440393, [1] [Cycle 1]: 0.00013252, [2] [tag_attr]: 6.502e-05 [meta_addattr_fg_expand]: 1.149e-05 [parallel-infer-symbol]: 3.98001e-06 [pre_auto_parallel]: 8.155e-05 [insert-virtual-dataset]: 2.88998e-06 [parallel-infer-symbol-second]: 1.66e-06 [dataset_repeat_opt]: 2.12001e-06 [pipeline_split]: 1.57001e-06 [optimize]: 0.0196624, [53] [py_interpret_to_execute]: 1.257e-05 [rewriter_before_opt_a]: 0.00037716 [opt_a]: 0.0170308, [2] [Cycle 1]: 0.0163084, [45] [expand_dump_flag]: 7.17002e-06 [switch_simplify]: 0.00021423 [loop_unroll]: 6.942e-05 [a_1]: 0.00190198 [with_stream_mark]: 3.148e-05 [recompute_prepare]: 1.254e-05 [updatestate_depend_eliminate]: 5.27999e-06 [updatestate_assign_eliminate]: 3.85e-06 [updatestate_loads_eliminate]: 3.36999e-06 [parameter_eliminate]: 2.79999e-06 [a_2]: 9.563e-05 [accelerated_algorithm]: 2.203e-05 [shard]: 2.48002e-06 [meta_shard_fg_expand]: 5.05999e-06 [shard_inline]: 6.31e-06 [merge_send_recv]: 9.25999e-06 [auto_parallel]: 1.14e-05 [parallel]: 3.687e-05 [flash_sp]: 1.188e-05 [merge_comm]: 4.69002e-06 [allreduce_fusion]: 3.35e-06 [matmul_add_comm_reduction]: 1.259e-05 [allreduce_slice_to_reducescatter]: 8.49977e-07 [virtual_shard_identity]: 1.178e-05 [virtual_dataset]: 7.15998e-06 [get_grad_eliminate_]: 2.446e-05 [virtual_output]: 7.21001e-06 [merge_forward]: 4.47e-06 [cell_reuse_recompute_pass]: 3.34001e-06 [offload_activation]: 1.109e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.447e-05 [merge_recompute_call_nodes]: 1.45001e-06 [before_grad]: 1.131e-05 [set_forward_comm_id_for_comm_node_pass]: 3.86999e-06 [meta_fg_expand]: 4.23001e-06 [flash_sp_send_recv_attached]: 5.37001e-06 [receive_attached]: 2.44001e-06 [after_resolve]: 1.199e-05 [a_after_grad]: 9.31e-06 [renormalize]: 0.013201 [add_forward_monad_depend]: 1.296e-05 [auto_monad_grad]: 3.42002e-06 [auto_monad_eliminator]: 2.862e-05 [cse]: 3.803e-05 [a_3]: 5.863e-05 [Cycle 2]: 0.00070443, [45] [expand_dump_flag]: 2.68e-06 [switch_simplify]: 8.07e-06 [loop_unroll]: 5.57001e-06 [a_1]: 0.0001632 [with_stream_mark]: 1.616e-05 [recompute_prepare]: 5.79999e-06 [updatestate_depend_eliminate]: 3.71001e-06 [updatestate_assign_eliminate]: 3.35998e-06 [updatestate_loads_eliminate]: 3.55998e-06 [parameter_eliminate]: 2.16998e-06 [a_2]: 7.421e-05 [accelerated_algorithm]: 1.048e-05 [shard]: 2.44001e-06 [meta_shard_fg_expand]: 2.18002e-06 [shard_inline]: 5.87001e-06 [merge_send_recv]: 8.77999e-06 [auto_parallel]: 8.90999e-06 [parallel]: 1.052e-05 [flash_sp]: 4.07e-06 [merge_comm]: 3.56999e-06 [allreduce_fusion]: 3.23e-06 [matmul_add_comm_reduction]: 1.195e-05 [allreduce_slice_to_reducescatter]: 7.39994e-07 [virtual_shard_identity]: 6.88e-06 [virtual_dataset]: 5.60001e-06 [get_grad_eliminate_]: 4.95001e-06 [virtual_output]: 5.03002e-06 [merge_forward]: 4.03001e-06 [cell_reuse_recompute_pass]: 3.28998e-06 [offload_activation]: 1.088e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.102e-05 [merge_recompute_call_nodes]: 1.71e-06 [before_grad]: 9.77001e-06 [set_forward_comm_id_for_comm_node_pass]: 4.62998e-06 [meta_fg_expand]: 2.39001e-06 [flash_sp_send_recv_attached]: 1.77999e-06 [receive_attached]: 2.36e-06 [after_resolve]: 1.028e-05 [a_after_grad]: 7.56001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.37999e-06 [auto_monad_grad]: 1.13001e-06 [auto_monad_eliminator]: 7.83001e-06 [cse]: 1.71e-05 [a_3]: 3.047e-05 [py_interpret_to_execute_after_opt_a]: 1.103e-05 [slice_cell_reuse_recomputed_activation]: 1.93997e-06 [rewriter_after_opt_a]: 2.173e-05 [convert_after_rewriter]: 1.24e-06 [order_py_execute_after_rewriter]: 1.12999e-06 [mutable_eliminate]: 0.00075961 [opt_b]: 0.00019161, [1] [Cycle 1]: 0.00018325, [7] [b_1]: 0.00010935 [b_2]: 7.18e-06 [updatestate_depend_eliminate]: 7.5e-06 [updatestate_assign_eliminate]: 2.29001e-06 [updatestate_loads_eliminate]: 2.39001e-06 [renormalize]: 9.49978e-07 [cse]: 2.032e-05 [optimize_parallel_all_gather_comm]: 1.826e-05 [overlap_param_gather]: 2.06998e-06 [cconv]: 3.638e-05 [loop_unroll]: 0.00047782 [opt_after_cconv]: 0.00010085, [1] [Cycle 1]: 9.42e-05, [7] [c_1]: 2.541e-05 [parameter_eliminate]: 5.14e-06 [updatestate_depend_eliminate]: 6.08002e-06 [updatestate_assign_eliminate]: 2.36e-06 [updatestate_loads_eliminate]: 2.76e-06 [cse]: 1.901e-05 [renormalize]: 5.09986e-07 [remove_dup_value]: 1.804e-05 [tuple_transform]: 7.215e-05, [1] [Cycle 1]: 6.655e-05, [4] [d_1]: 4.068e-05 [none_parameter_eliminate]: 1.54e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 6.02999e-06 [partial_unused_args_eliminate]: 2.01e-06 [add_recomputation]: 5.143e-05 [cse_after_recomputation]: 2.048e-05, [1] [Cycle 1]: 1.589e-05, [1] [cse]: 1.072e-05 [environ_conv]: 5.94e-06 [swap_dp_allreduce_reducescatter]: 4.79998e-06 [bias_add_comm_swap]: 2.73e-06 [label_micro_interleaved_index]: 4.86002e-06 [label_fine_grained_interleaved_index]: 2.68003e-06 [merge_cast_opt]: 1.34003e-06 [slice_recompute_activation]: 2.02001e-06 [micro_interleaved_order_control]: 2.37999e-06 [assign_add_opt]: 1.40999e-06 [ForceFp32Comm]: 8.80013e-07 [remove_cast_before_assign_add]: 9.70002e-07 [full_micro_interleaved_order_control]: 2.06e-06 [reorder_send_recv_between_fp_bp]: 2.66999e-06 [comm_op_add_attrs]: 1.16002e-06 [add_comm_op_reuse_tag]: 1.25999e-06 [interleave_split_concat_branches]: 1.39998e-06 [interleave_parallel_branches]: 1.02e-06 [overlap_opt_shard_in_pipeline]: 1.24998e-06 [overlap_opt_shard_grad_in_pipeline]: 1.66002e-06 [control_data_broadcast_order]: 1.228e-05 [grouped_pairwise_exchange_alltoall]: 1.47001e-06 [offloading_packed_experts]: 3.38e-06 [overlap_recompute_and_grad_model_parallel]: 4.58999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12e-06 [overlap_recompute_allgather_and_fa_grad]: 1.17999e-06 [overlap_recompute_comm]: 2.46e-06 [overlap_grad_ring_attention]: 4.79e-06 [overlap_grad_flash_sp]: 2.017e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.62001e-06 [split_layernorm_comm]: 1.97001e-06 [handle_group_info]: 9.5999e-07 [symbol_engine_optimizer]: 7.192e-05, [1] [Cycle 1]: 6.713e-05, [6] [build]: 4.1e-06 [elim_shapecalc]: 9.17001e-06 [elim_not_effective]: 1.084e-05 [opt_reshape]: 7e-06 [fold_const_symbol]: 8.46997e-06 [renormalize]: 2.3999e-07 [detach_backward]: 2.33002e-06 [pipeline_parallel_scheduler]: 1.74e-06 [auto_monad_reorder]: 1.928e-05 [get_jit_bprop_graph]: 2.19001e-06 [rewriter_after_jit_bprop_graph]: 6.39001e-06 [opt_after_jit_grad]: 0.00052592 [validate]: 4.556e-05 Sums bootstrap : 0.000437s : 0.16% type_inference : 0.236408s : 87.82% event_method : 0.000489s : 0.18% auto_monad : 0.012415s : 4.61% graph_reusing : 0.000013s : 0.00% inline : 0.000004s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000065s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000011s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000082s : 0.03% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000013s : 0.00% optimize.rewriter_before_opt_a : 0.000377s : 0.14% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000222s : 0.08% optimize.opt_a.loop_unroll : 0.000075s : 0.03% optimize.opt_a.a_1 : 0.002065s : 0.77% optimize.opt_a.with_stream_mark : 0.000048s : 0.02% optimize.opt_a.recompute_prepare : 0.000018s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000007s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000170s : 0.06% optimize.opt_a.accelerated_algorithm : 0.000033s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000012s : 0.00% optimize.opt_a.merge_send_recv : 0.000018s : 0.01% optimize.opt_a.auto_parallel : 0.000020s : 0.01% optimize.opt_a.parallel : 0.000047s : 0.02% optimize.opt_a.flash_sp : 0.000016s : 0.01% optimize.opt_a.merge_comm : 0.000008s : 0.00% optimize.opt_a.allreduce_fusion : 0.000007s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000025s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000019s : 0.01% optimize.opt_a.virtual_dataset : 0.000013s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000029s : 0.01% optimize.opt_a.virtual_output : 0.000012s : 0.00% optimize.opt_a.merge_forward : 0.000009s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000007s : 0.00% optimize.opt_a.offload_activation : 0.000022s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000025s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000021s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.00% optimize.opt_a.meta_fg_expand : 0.000007s : 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 : 0.000022s : 0.01% optimize.opt_a.a_after_grad : 0.000017s : 0.01% optimize.opt_a.renormalize : 0.013201s : 4.90% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000036s : 0.01% optimize.opt_a.cse : 0.000055s : 0.02% optimize.opt_a.a_3 : 0.000089s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000022s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000760s : 0.28% optimize.opt_b.b_1 : 0.000109s : 0.04% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000020s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000018s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.01% optimize.loop_unroll : 0.000478s : 0.18% optimize.opt_after_cconv.c_1 : 0.000025s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000019s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000041s : 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 : 0.000006s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000051s : 0.02% optimize.cse_after_recomputation.cse : 0.000011s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000012s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000020s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000008s : 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 : 0.000019s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000526s : 0.20% validate : 0.000046s : 0.02% Time group info: ------[substitution.] 0.000728 51 17.66% : 0.000129s : 5: substitution.arithmetic_simplify 2.37% : 0.000017s : 1: substitution.cast_eliminate 0.23% : 0.000002s : 2: substitution.elim_not_effective 0.22% : 0.000002s : 2: substitution.fold_const_symbol 0.99% : 0.000007s : 3: substitution.graph_param_transform 70.21% : 0.000511s : 19: substitution.inline 0.72% : 0.000005s : 4: substitution.j_node_and_user_rematch 2.01% : 0.000015s : 2: substitution.less_batch_normalization 0.76% : 0.000006s : 4: substitution.remove_not_recompute_node 0.80% : 0.000006s : 2: substitution.replace_old_param 2.34% : 0.000017s : 5: substitution.switch_simplify 1.70% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.236262 2 92.06% : 0.217513s : 1: type_inference.infer 7.94% : 0.018748s : 1: type_inference.specialize ------[replace.] 0.000256 27 3.40% : 0.000009s : 1: replace.cast_eliminate 58.31% : 0.000149s : 19: replace.inline 29.77% : 0.000076s : 5: replace.switch_simplify 8.52% : 0.000022s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000541 27 2.86% : 0.000015s : 1: match.cast_eliminate 92.44% : 0.000500s : 19: match.inline 2.59% : 0.000014s : 5: match.switch_simplify 2.11% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000347 1960 1.25% : 0.000004s : 26: predicate.accumulaten_eliminater 0.40% : 0.000001s : 3: predicate.ad_related_special_op_eliminate 0.26% : 0.000001s : 6: predicate.addn_check_dump 1.28% : 0.000004s : 26: predicate.addn_zero_filter 1.02% : 0.000004s : 26: predicate.adjust_all_reduce_mul_add 3.14% : 0.000011s : 32: predicate.arithmetic_simplify 1.47% : 0.000005s : 27: predicate.cast_eliminate 0.26% : 0.000001s : 6: predicate.check_bprop_eliminate 0.30% : 0.000001s : 6: predicate.compare_switch_simplify 0.08% : 0.000000s : 3: predicate.const_output_eliminate 0.33% : 0.000001s : 6: predicate.depend_value_elim 1.23% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 1.40% : 0.000005s : 27: predicate.dict_get_item_eliminator 1.16% : 0.000004s : 27: predicate.dict_set_item_eliminator 0.42% : 0.000001s : 6: predicate.dumpgradient_eliminate 0.14% : 0.000000s : 3: predicate.elim_not_effective 0.18% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.53% : 0.000005s : 30: predicate.environ_add_const_eliminate 1.18% : 0.000004s : 30: predicate.environ_get_add_eliminate 1.23% : 0.000004s : 30: predicate.environ_get_depend_swap 1.64% : 0.000006s : 36: predicate.environ_get_eliminate 1.32% : 0.000005s : 30: predicate.environ_get_set_eliminate 2.09% : 0.000007s : 48: predicate.exchange_switch_depend_value 3.26% : 0.000011s : 48: predicate.float_depend_g_call 0.31% : 0.000001s : 6: predicate.float_environ_get_switch 0.38% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 3: predicate.fold_const_symbol 0.34% : 0.000001s : 6: predicate.get_grad_eliminate 0.27% : 0.000001s : 3: predicate.graph_param_transform 0.28% : 0.000001s : 6: predicate.incorporate_call 0.22% : 0.000001s : 6: predicate.incorporate_call_switch 6.50% : 0.000023s : 92: predicate.inline 0.46% : 0.000002s : 6: predicate.inline_without_move 0.14% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.57% : 0.000002s : 6: predicate.less_batch_normalization 1.60% : 0.000006s : 35: predicate.list_to_tuple_eliminator_ 2.51% : 0.000009s : 61: predicate.load_eliminater 0.51% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.57% : 0.000012s : 84: predicate.loop_unroll_before_grad 1.63% : 0.000006s : 33: predicate.make_slice_get_slice_eliminator 0.33% : 0.000001s : 6: predicate.merge_addn 0.26% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.28% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.05% : 0.000004s : 26: predicate.minmaximum_grad 0.65% : 0.000002s : 3: predicate.mutable_eliminate 0.28% : 0.000001s : 3: predicate.opt_reshape 0.19% : 0.000001s : 3: predicate.parallel_virtual_node 3.46% : 0.000012s : 48: predicate.partial_defer_inline 1.42% : 0.000005s : 32: predicate.partial_eliminate 1.17% : 0.000004s : 26: predicate.print_const_string_wrapper 0.31% : 0.000001s : 6: predicate.reduce_all_const_elim 1.50% : 0.000005s : 26: predicate.reduce_eliminate 2.49% : 0.000009s : 61: predicate.redundant_stop_gradient_eliminater 0.31% : 0.000001s : 6: predicate.remove_not_recompute_node 1.27% : 0.000004s : 35: predicate.replace_applicator 0.27% : 0.000001s : 6: predicate.replace_old_param 0.24% : 0.000001s : 3: predicate.reset_defer_inline 1.29% : 0.000004s : 26: predicate.reshape_eliminate 0.31% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.21% : 0.000001s : 3: predicate.row_tensor_eliminate 0.44% : 0.000002s : 6: predicate.same_eliminate 0.32% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.53% : 0.000002s : 6: predicate.shard_identity_eliminate 0.42% : 0.000001s : 6: predicate.special_op_eliminate 0.37% : 0.000001s : 6: predicate.specialize_transform 0.55% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.42% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.19% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.38% : 0.000008s : 48: predicate.switch_defer_inline 2.54% : 0.000009s : 54: predicate.switch_layer_defer_inline 7.15% : 0.000025s : 151: predicate.switch_simplify 1.13% : 0.000004s : 26: predicate.tile_eliminate 1.17% : 0.000004s : 26: predicate.transpose_eliminate 1.48% : 0.000005s : 33: predicate.tuple_list_convert_item_index_to_positive 1.43% : 0.000005s : 33: predicate.tuple_list_get_item_const_eliminator 1.33% : 0.000005s : 33: predicate.tuple_list_get_item_depend_reorder 2.55% : 0.000009s : 41: predicate.tuple_list_get_item_eliminator 1.40% : 0.000005s : 33: predicate.tuple_list_get_set_item_eliminator 2.16% : 0.000007s : 39: predicate.tuple_list_set_item_eliminator 1.95% : 0.000007s : 35: predicate.tuple_to_list_eliminator_ 2.38% : 0.000008s : 61: predicate.updatestate_pure_node_eliminater 3.05% : 0.000011s : 67: predicate.updatestate_useless_node_eliminater 0.19% : 0.000001s : 3: predicate.value_based_eliminate 0.51% : 0.000002s : 6: predicate.virtual_dataset_eliminate 0.37% : 0.000001s : 6: predicate.virtual_output_eliminate 0.12% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.24% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002550 33 41.74% : 0.001064s : 12: func_graph_cloner_run.FuncGraphClonerGraph 58.26% : 0.001486s : 21: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.314841 192 0.00% : 0.000003s : 1: ForceFp32Comm 1.41% : 0.004427s : 1: add_attr 1.40% : 0.004410s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000056s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 3.95% : 0.012450s : 1: auto_monad 0.01% : 0.000023s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.15% : 0.000471s : 1: bootstrap 0.01% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000016s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000024s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.16% : 0.000512s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000020s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000003s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000006s : 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 0.15% : 0.000487s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.25% : 0.000773s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.00% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000017s : 1: opt.transform.mutable_eliminate 0.87% : 0.002741s : 78: opt.transform.opt_a 0.01% : 0.000024s : 1: opt.transform.opt_after_cconv 0.01% : 0.000023s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000089s : 28: opt.transform.opt_b 0.01% : 0.000045s : 2: opt.transform.opt_trans_graph 0.01% : 0.000033s : 4: opt.transform.symbol_engine_opt 5.41% : 0.017035s : 1: opt_a 0.03% : 0.000104s : 1: opt_after_cconv 0.17% : 0.000538s : 1: opt_after_jit_grad 0.06% : 0.000195s : 1: opt_b 6.25% : 0.019669s : 1: optimize 0.01% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000023s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000005s : 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 0.03% : 0.000088s : 1: pre_auto_parallel 0.01% : 0.000016s : 1: py_interpret_to_execute 0.00% : 0.000014s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000023s : 1: remove_dup_value 3.77% : 0.011878s : 1: renormalize.infer 0.41% : 0.001299s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000025s : 1: rewriter_after_opt_a 0.12% : 0.000387s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000004s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000074s : 1: symbol_engine_optimizer 0.02% : 0.000075s : 1: tuple_transform 75.10% : 0.236448s : 1: type_inference TotalTime = 0.451994, [21] [bootstrap]: 0.00049512 [type_inference]: 0.304405 [event_method]: 7.254e-05 [auto_monad]: 0.00024533 [graph_reusing]: 1.892e-05 [inline]: 3.11001e-06 [add_attr]: 0.029058, [1] [add_attr_with_inline]: 0.0290431, [1] [Cycle 1]: 0.00015455, [2] [tag_attr]: 8.261e-05 [meta_addattr_fg_expand]: 1.808e-05 [parallel-infer-symbol]: 3.83001e-06 [pre_auto_parallel]: 0.00011144 [insert-virtual-dataset]: 2.71e-06 [parallel-infer-symbol-second]: 8.09989e-07 [dataset_repeat_opt]: 1.97001e-06 [pipeline_split]: 1.58002e-06 [optimize]: 0.116787, [53] [py_interpret_to_execute]: 1.2e-05 [rewriter_before_opt_a]: 0.00055954 [opt_a]: 0.11397, [3] [Cycle 1]: 0.109726, [45] [expand_dump_flag]: 1.645e-05 [switch_simplify]: 0.00033221 [loop_unroll]: 0.0001138 [a_1]: 0.00304319 [with_stream_mark]: 4.187e-05 [recompute_prepare]: 3.384e-05 [updatestate_depend_eliminate]: 1.03e-05 [updatestate_assign_eliminate]: 8.25e-06 [updatestate_loads_eliminate]: 8.08999e-06 [parameter_eliminate]: 4.1e-06 [a_2]: 0.0002694 [accelerated_algorithm]: 3.168e-05 [shard]: 2.36e-06 [meta_shard_fg_expand]: 1.248e-05 [shard_inline]: 1.588e-05 [merge_send_recv]: 2.03e-05 [auto_parallel]: 1.583e-05 [parallel]: 2.22e-05 [flash_sp]: 1.436e-05 [merge_comm]: 9.94001e-06 [allreduce_fusion]: 9.84999e-06 [matmul_add_comm_reduction]: 3.759e-05 [allreduce_slice_to_reducescatter]: 1.05001e-06 [virtual_shard_identity]: 1.985e-05 [virtual_dataset]: 1.693e-05 [get_grad_eliminate_]: 1.702e-05 [virtual_output]: 1.562e-05 [merge_forward]: 1.126e-05 [cell_reuse_recompute_pass]: 2.17999e-06 [offload_activation]: 2.032e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.581e-05 [merge_recompute_call_nodes]: 1.50999e-06 [before_grad]: 3.133e-05 [set_forward_comm_id_for_comm_node_pass]: 1.283e-05 [meta_fg_expand]: 0.00230761 [flash_sp_send_recv_attached]: 7.68001e-06 [receive_attached]: 2.58003e-06 [after_resolve]: 8.577e-05 [a_after_grad]: 9.809e-05 [renormalize]: 0.101737 [add_forward_monad_depend]: 1.443e-05 [auto_monad_grad]: 7.33999e-06 [auto_monad_eliminator]: 6.266e-05 [cse]: 0.00034109 [a_3]: 0.00035184 [Cycle 2]: 0.0034969, [45] [expand_dump_flag]: 4.06001e-06 [switch_simplify]: 4.466e-05 [loop_unroll]: 4.041e-05 [a_1]: 0.00137496 [with_stream_mark]: 2.146e-05 [recompute_prepare]: 9.52001e-06 [updatestate_depend_eliminate]: 5.17999e-06 [updatestate_assign_eliminate]: 3.67002e-06 [updatestate_loads_eliminate]: 3.56001e-06 [parameter_eliminate]: 2.39999e-06 [a_2]: 0.00010034 [accelerated_algorithm]: 2.072e-05 [shard]: 2.19001e-06 [meta_shard_fg_expand]: 2.74999e-06 [shard_inline]: 6.38998e-06 [merge_send_recv]: 9.09e-06 [auto_parallel]: 1.009e-05 [parallel]: 1.026e-05 [flash_sp]: 4.32e-06 [merge_comm]: 4e-06 [allreduce_fusion]: 3.50998e-06 [matmul_add_comm_reduction]: 1.226e-05 [allreduce_slice_to_reducescatter]: 1.05001e-06 [virtual_shard_identity]: 7.92e-06 [virtual_dataset]: 6.21e-06 [get_grad_eliminate_]: 6.26e-06 [virtual_output]: 6.02001e-06 [merge_forward]: 4.27e-06 [cell_reuse_recompute_pass]: 1.54e-06 [offload_activation]: 1.131e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.35e-05 [merge_recompute_call_nodes]: 1.81e-06 [before_grad]: 1.152e-05 [set_forward_comm_id_for_comm_node_pass]: 4.23001e-06 [meta_fg_expand]: 0.00034775 [flash_sp_send_recv_attached]: 1.94999e-06 [receive_attached]: 2.62001e-06 [after_resolve]: 1.698e-05 [a_after_grad]: 1.049e-05 [renormalize]: 0.00097471 [add_forward_monad_depend]: 6.09001e-06 [auto_monad_grad]: 2.49999e-06 [auto_monad_eliminator]: 1.594e-05 [cse]: 3.579e-05 [a_3]: 5.399e-05 [Cycle 3]: 0.00072153, [45] [expand_dump_flag]: 2.14e-06 [switch_simplify]: 8.17e-06 [loop_unroll]: 6.25002e-06 [a_1]: 0.00016587 [with_stream_mark]: 1.089e-05 [recompute_prepare]: 6.63e-06 [updatestate_depend_eliminate]: 4.08999e-06 [updatestate_assign_eliminate]: 2.70002e-06 [updatestate_loads_eliminate]: 2.74999e-06 [parameter_eliminate]: 1.25001e-06 [a_2]: 8.685e-05 [accelerated_algorithm]: 9.81998e-06 [shard]: 1.10999e-06 [meta_shard_fg_expand]: 1.57001e-06 [shard_inline]: 6.43e-06 [merge_send_recv]: 6.54999e-06 [auto_parallel]: 7.11999e-06 [parallel]: 6.21e-06 [flash_sp]: 1.12e-06 [merge_comm]: 3.65998e-06 [allreduce_fusion]: 1.993e-05 [matmul_add_comm_reduction]: 7.38e-06 [allreduce_slice_to_reducescatter]: 5.50004e-07 [virtual_shard_identity]: 7.87e-06 [virtual_dataset]: 6.38e-06 [get_grad_eliminate_]: 6.02999e-06 [virtual_output]: 6.30002e-06 [merge_forward]: 3.63999e-06 [cell_reuse_recompute_pass]: 2.56998e-06 [offload_activation]: 8.12e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.265e-05 [merge_recompute_call_nodes]: 1.27e-06 [before_grad]: 1.041e-05 [set_forward_comm_id_for_comm_node_pass]: 4.43999e-06 [meta_fg_expand]: 2.37999e-06 [flash_sp_send_recv_attached]: 1.39998e-06 [receive_attached]: 1.59e-06 [after_resolve]: 8.70001e-06 [a_after_grad]: 9.07001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.44e-06 [auto_monad_grad]: 1.04e-06 [auto_monad_eliminator]: 8.02e-06 [cse]: 1.849e-05 [a_3]: 3.736e-05 [py_interpret_to_execute_after_opt_a]: 6.61999e-06 [slice_cell_reuse_recomputed_activation]: 2.24001e-06 [rewriter_after_opt_a]: 2.315e-05 [convert_after_rewriter]: 1.62999e-06 [order_py_execute_after_rewriter]: 1.57001e-06 [mutable_eliminate]: 0.00075683 [opt_b]: 0.00021472, [1] [Cycle 1]: 0.00020661, [7] [b_1]: 0.00012817 [b_2]: 7.88001e-06 [updatestate_depend_eliminate]: 6.81999e-06 [updatestate_assign_eliminate]: 2.79999e-06 [updatestate_loads_eliminate]: 2.61e-06 [renormalize]: 5.59987e-07 [cse]: 2.516e-05 [optimize_parallel_all_gather_comm]: 1.804e-05 [overlap_param_gather]: 2.34999e-06 [cconv]: 2.915e-05 [loop_unroll]: 0.00046867 [opt_after_cconv]: 0.00010611, [1] [Cycle 1]: 9.936e-05, [7] [c_1]: 3.09e-05 [parameter_eliminate]: 3.28998e-06 [updatestate_depend_eliminate]: 5.74999e-06 [updatestate_assign_eliminate]: 2.89999e-06 [updatestate_loads_eliminate]: 2.60002e-06 [cse]: 2.188e-05 [renormalize]: 4.09986e-07 [remove_dup_value]: 1.627e-05 [tuple_transform]: 7.444e-05, [1] [Cycle 1]: 6.978e-05, [4] [d_1]: 4.367e-05 [none_parameter_eliminate]: 1.97999e-06 [renormalize]: 1.60013e-07 [switch_simplify]: 6.62002e-06 [partial_unused_args_eliminate]: 1.89999e-06 [add_recomputation]: 5.435e-05 [cse_after_recomputation]: 2.411e-05, [1] [Cycle 1]: 1.891e-05, [1] [cse]: 1.395e-05 [environ_conv]: 5.40999e-06 [swap_dp_allreduce_reducescatter]: 5.66e-06 [bias_add_comm_swap]: 2.59001e-06 [label_micro_interleaved_index]: 4.68999e-06 [label_fine_grained_interleaved_index]: 2.77002e-06 [merge_cast_opt]: 1.40999e-06 [slice_recompute_activation]: 1.97999e-06 [micro_interleaved_order_control]: 2.36998e-06 [assign_add_opt]: 1.30999e-06 [ForceFp32Comm]: 7.90023e-07 [remove_cast_before_assign_add]: 9.50007e-07 [full_micro_interleaved_order_control]: 2.64001e-06 [reorder_send_recv_between_fp_bp]: 2.60002e-06 [comm_op_add_attrs]: 1.00999e-06 [add_comm_op_reuse_tag]: 9.60019e-07 [interleave_split_concat_branches]: 1.29998e-06 [interleave_parallel_branches]: 1.38002e-06 [overlap_opt_shard_in_pipeline]: 1.26997e-06 [overlap_opt_shard_grad_in_pipeline]: 2.12999e-06 [control_data_broadcast_order]: 1.533e-05 [grouped_pairwise_exchange_alltoall]: 1.52999e-06 [offloading_packed_experts]: 4.44002e-06 [overlap_recompute_and_grad_model_parallel]: 5.57999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12e-06 [overlap_recompute_allgather_and_fa_grad]: 1.20001e-06 [overlap_recompute_comm]: 2.71999e-06 [overlap_grad_ring_attention]: 4.95999e-06 [overlap_grad_flash_sp]: 2.526e-05 [begin_end_overlap_inline]: 4.99975e-07 [split_matmul_comm_elemetwise]: 2.14e-06 [split_layernorm_comm]: 1.82001e-06 [handle_group_info]: 1.00999e-06 [symbol_engine_optimizer]: 8.063e-05, [1] [Cycle 1]: 7.554e-05, [6] [build]: 3.52997e-06 [elim_shapecalc]: 1.092e-05 [elim_not_effective]: 1.427e-05 [opt_reshape]: 8.1e-06 [fold_const_symbol]: 1.093e-05 [renormalize]: 1.80007e-07 [detach_backward]: 2.30002e-06 [pipeline_parallel_scheduler]: 1.67001e-06 [auto_monad_reorder]: 1.982e-05 [get_jit_bprop_graph]: 1.91998e-06 [rewriter_after_jit_bprop_graph]: 5.19e-06 [opt_after_jit_grad]: 0.00050465 [validate]: 4.512e-05 Sums bootstrap : 0.000495s : 0.12% type_inference : 0.304405s : 72.22% event_method : 0.000073s : 0.02% auto_monad : 0.000245s : 0.06% graph_reusing : 0.000019s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000083s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000018s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000111s : 0.03% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000012s : 0.00% optimize.rewriter_before_opt_a : 0.000560s : 0.13% optimize.opt_a.expand_dump_flag : 0.000023s : 0.01% optimize.opt_a.switch_simplify : 0.000385s : 0.09% optimize.opt_a.loop_unroll : 0.000160s : 0.04% optimize.opt_a.a_1 : 0.004584s : 1.09% optimize.opt_a.with_stream_mark : 0.000074s : 0.02% optimize.opt_a.recompute_prepare : 0.000050s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000020s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000014s : 0.00% optimize.opt_a.parameter_eliminate : 0.000008s : 0.00% optimize.opt_a.a_2 : 0.000457s : 0.11% optimize.opt_a.accelerated_algorithm : 0.000062s : 0.01% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000017s : 0.00% optimize.opt_a.shard_inline : 0.000029s : 0.01% optimize.opt_a.merge_send_recv : 0.000036s : 0.01% optimize.opt_a.auto_parallel : 0.000033s : 0.01% optimize.opt_a.parallel : 0.000039s : 0.01% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000018s : 0.00% optimize.opt_a.allreduce_fusion : 0.000033s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000057s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000036s : 0.01% optimize.opt_a.virtual_dataset : 0.000030s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000029s : 0.01% optimize.opt_a.virtual_output : 0.000028s : 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 : 0.000040s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000062s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000005s : 0.00% optimize.opt_a.before_grad : 0.000053s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000022s : 0.01% optimize.opt_a.meta_fg_expand : 0.002658s : 0.63% optimize.opt_a.flash_sp_send_recv_attached : 0.000011s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000111s : 0.03% optimize.opt_a.a_after_grad : 0.000118s : 0.03% optimize.opt_a.renormalize : 0.102712s : 24.37% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.01% optimize.opt_a.auto_monad_grad : 0.000011s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000087s : 0.02% optimize.opt_a.cse : 0.000395s : 0.09% optimize.opt_a.a_3 : 0.000443s : 0.11% optimize.py_interpret_to_execute_after_opt_a : 0.000007s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000023s : 0.01% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000002s : 0.00% optimize.mutable_eliminate : 0.000757s : 0.18% optimize.opt_b.b_1 : 0.000128s : 0.03% optimize.opt_b.b_2 : 0.000008s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000025s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000018s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000029s : 0.01% optimize.loop_unroll : 0.000469s : 0.11% optimize.opt_after_cconv.c_1 : 0.000031s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000022s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.00% optimize.tuple_transform.d_1 : 0.000044s : 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 : 0.000007s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000054s : 0.01% optimize.cse_after_recomputation.cse : 0.000014s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000015s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000025s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000008s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000011s : 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 : 0.000020s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000505s : 0.12% validate : 0.000045s : 0.01% Time group info: ------[substitution.] 0.001538 199 10.89% : 0.000167s : 6: substitution.arithmetic_simplify 0.86% : 0.000013s : 1: substitution.cast_eliminate 0.14% : 0.000002s : 3: substitution.elim_not_effective 0.66% : 0.000010s : 11: substitution.float_depend_g_call 0.33% : 0.000005s : 3: substitution.float_tuple_getitem_switch 0.11% : 0.000002s : 3: substitution.fold_const_symbol 0.38% : 0.000006s : 4: substitution.graph_param_transform 0.20% : 0.000003s : 2: substitution.incorporate_call 0.17% : 0.000003s : 2: substitution.incorporate_call_switch 66.44% : 0.001022s : 36: substitution.inline 1.71% : 0.000026s : 2: substitution.inline_without_move 0.77% : 0.000012s : 16: substitution.j_node_and_user_rematch 1.65% : 0.000025s : 3: substitution.less_batch_normalization 0.87% : 0.000013s : 7: substitution.minmaximum_grad 0.65% : 0.000010s : 11: substitution.partial_eliminate 0.91% : 0.000014s : 16: substitution.remove_not_recompute_node 1.86% : 0.000029s : 9: substitution.replace_applicator 0.70% : 0.000011s : 8: substitution.replace_old_param 0.31% : 0.000005s : 1: substitution.set_cell_output_no_recompute 1.41% : 0.000022s : 8: substitution.switch_simplify 1.69% : 0.000026s : 7: substitution.tuple_list_convert_item_index_to_positive 0.73% : 0.000011s : 7: substitution.tuple_list_get_item_const_eliminator 1.04% : 0.000016s : 7: substitution.tuple_list_get_item_depend_reorder 4.38% : 0.000067s : 19: substitution.tuple_list_get_item_eliminator 1.12% : 0.000017s : 7: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.304162 2 98.02% : 0.298145s : 1: type_inference.infer 1.98% : 0.006017s : 1: type_inference.specialize ------[replace.] 0.000513 54 1.74% : 0.000009s : 1: replace.cast_eliminate 60.11% : 0.000308s : 36: replace.inline 22.89% : 0.000117s : 8: replace.switch_simplify 15.27% : 0.000078s : 9: replace.tuple_list_get_item_eliminator ------[match.] 0.001057 54 1.19% : 0.000013s : 1: match.cast_eliminate 94.59% : 0.001000s : 36: match.inline 1.70% : 0.000018s : 8: match.switch_simplify 2.51% : 0.000027s : 9: match.tuple_list_get_item_eliminator ------[predicate.] 0.000824 5282 1.21% : 0.000010s : 69: predicate.accumulaten_eliminater 0.22% : 0.000002s : 4: predicate.ad_related_special_op_eliminate 0.41% : 0.000003s : 21: predicate.addn_check_dump 1.21% : 0.000010s : 69: predicate.addn_zero_filter 1.10% : 0.000009s : 69: predicate.adjust_all_reduce_mul_add 2.33% : 0.000019s : 90: predicate.arithmetic_simplify 1.37% : 0.000011s : 70: predicate.cast_eliminate 0.84% : 0.000007s : 51: predicate.check_bprop_eliminate 0.38% : 0.000003s : 21: predicate.compare_switch_simplify 0.04% : 0.000000s : 4: predicate.const_output_eliminate 0.44% : 0.000004s : 21: predicate.depend_value_elim 1.31% : 0.000011s : 70: predicate.dict_get_item_const_eliminator 1.33% : 0.000011s : 70: predicate.dict_get_item_eliminator 1.21% : 0.000010s : 70: predicate.dict_set_item_eliminator 0.21% : 0.000002s : 8: predicate.dumpgradient_eliminate 0.06% : 0.000001s : 4: predicate.elim_not_effective 0.09% : 0.000001s : 4: predicate.elim_shapecalc_of_broadcastargs 1.26% : 0.000010s : 74: predicate.environ_add_const_eliminate 1.26% : 0.000010s : 74: predicate.environ_get_add_eliminate 1.20% : 0.000010s : 74: predicate.environ_get_depend_swap 1.58% : 0.000013s : 95: predicate.environ_get_eliminate 1.23% : 0.000010s : 74: predicate.environ_get_set_eliminate 1.95% : 0.000016s : 115: predicate.exchange_switch_depend_value 2.86% : 0.000024s : 115: predicate.float_depend_g_call 0.36% : 0.000003s : 21: predicate.float_environ_get_switch 0.46% : 0.000004s : 25: predicate.float_tuple_getitem_switch 0.04% : 0.000000s : 4: predicate.fold_const_symbol 0.48% : 0.000004s : 21: predicate.get_grad_eliminate 0.05% : 0.000000s : 4: predicate.graph_param_transform 0.37% : 0.000003s : 21: predicate.incorporate_call 0.32% : 0.000003s : 21: predicate.incorporate_call_switch 5.60% : 0.000046s : 234: predicate.inline 1.08% : 0.000009s : 43: predicate.inline_without_move 0.19% : 0.000002s : 21: predicate.j_node_and_user_rematch 0.58% : 0.000005s : 21: predicate.less_batch_normalization 1.62% : 0.000013s : 87: predicate.list_to_tuple_eliminator_ 2.49% : 0.000021s : 156: predicate.load_eliminater 0.24% : 0.000002s : 4: predicate.loop_unroll_after_grad 3.26% : 0.000027s : 179: predicate.loop_unroll_before_grad 1.36% : 0.000011s : 78: predicate.make_slice_get_slice_eliminator 0.38% : 0.000003s : 21: predicate.merge_addn 0.81% : 0.000007s : 51: predicate.micro_step_allgather_replace 0.83% : 0.000007s : 51: predicate.mini_step_allgather_replace 1.18% : 0.000010s : 69: predicate.minmaximum_grad 0.27% : 0.000002s : 4: predicate.mutable_eliminate 0.10% : 0.000001s : 4: predicate.opt_reshape 0.09% : 0.000001s : 4: predicate.parallel_virtual_node 3.01% : 0.000025s : 115: predicate.partial_defer_inline 1.53% : 0.000013s : 83: predicate.partial_eliminate 1.16% : 0.000010s : 69: predicate.print_const_string_wrapper 0.42% : 0.000004s : 21: predicate.reduce_all_const_elim 1.51% : 0.000012s : 69: predicate.reduce_eliminate 2.56% : 0.000021s : 156: predicate.redundant_stop_gradient_eliminater 0.23% : 0.000002s : 21: predicate.remove_not_recompute_node 1.61% : 0.000013s : 130: predicate.replace_applicator 0.56% : 0.000005s : 43: predicate.replace_old_param 0.06% : 0.000001s : 4: predicate.reset_defer_inline 1.23% : 0.000010s : 69: predicate.reshape_eliminate 0.86% : 0.000007s : 51: predicate.row_tensor_add_zeros_like 0.11% : 0.000001s : 4: predicate.row_tensor_eliminate 1.08% : 0.000009s : 51: predicate.same_eliminate 0.23% : 0.000002s : 21: predicate.set_cell_output_no_recompute 0.54% : 0.000004s : 21: predicate.shard_identity_eliminate 0.23% : 0.000002s : 8: predicate.special_op_eliminate 0.48% : 0.000004s : 21: predicate.specialize_transform 1.14% : 0.000009s : 51: predicate.split_environ_get_set_with_tuple_value 1.12% : 0.000009s : 43: predicate.stack_unstack_eliminate 0.08% : 0.000001s : 4: predicate.switch_call_monad_eliminater 2.19% : 0.000018s : 115: predicate.switch_defer_inline 3.15% : 0.000026s : 166: predicate.switch_layer_defer_inline 6.47% : 0.000053s : 335: predicate.switch_simplify 1.25% : 0.000010s : 69: predicate.tile_eliminate 1.17% : 0.000010s : 69: predicate.transpose_eliminate 1.51% : 0.000012s : 78: predicate.tuple_list_convert_item_index_to_positive 1.49% : 0.000012s : 78: predicate.tuple_list_get_item_const_eliminator 1.39% : 0.000011s : 78: predicate.tuple_list_get_item_depend_reorder 2.53% : 0.000021s : 108: predicate.tuple_list_get_item_eliminator 1.47% : 0.000012s : 78: predicate.tuple_list_get_set_item_eliminator 2.13% : 0.000018s : 99: predicate.tuple_list_set_item_eliminator 1.58% : 0.000013s : 87: predicate.tuple_to_list_eliminator_ 2.47% : 0.000020s : 156: predicate.updatestate_pure_node_eliminater 2.93% : 0.000024s : 177: predicate.updatestate_useless_node_eliminater 0.10% : 0.000001s : 4: predicate.value_based_eliminate 0.45% : 0.000004s : 21: predicate.virtual_dataset_eliminate 0.45% : 0.000004s : 21: predicate.virtual_output_eliminate 0.07% : 0.000001s : 4: predicate.virtual_view_grad_eliminate 0.14% : 0.000001s : 4: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.004003 61 62.15% : 0.002488s : 21: func_graph_cloner_run.FuncGraphClonerGraph 37.85% : 0.001515s : 40: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.707138 233 0.00% : 0.000004s : 1: ForceFp32Comm 4.11% : 0.029065s : 1: add_attr 4.11% : 0.029048s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000059s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000256s : 1: auto_monad 0.00% : 0.000023s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.08% : 0.000544s : 1: bootstrap 0.00% : 0.000032s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000019s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000027s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.01% : 0.000082s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000007s : 1: get_jit_bprop_graph 0.00% : 0.000024s : 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% : 0.000006s : 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% : 0.000019s : 1: label_micro_interleaved_index 0.07% : 0.000478s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.11% : 0.000768s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000016s : 1: opt.transform.mutable_eliminate 0.92% : 0.006475s : 117: opt.transform.opt_a 0.00% : 0.000030s : 1: opt.transform.opt_after_cconv 0.00% : 0.000025s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000107s : 28: opt.transform.opt_b 0.01% : 0.000048s : 2: opt.transform.opt_trans_graph 0.01% : 0.000041s : 4: opt.transform.symbol_engine_opt 16.12% : 0.113974s : 1: opt_a 0.02% : 0.000109s : 1: opt_after_cconv 0.07% : 0.000515s : 1: opt_after_jit_grad 0.03% : 0.000218s : 1: opt_b 16.52% : 0.116792s : 1: optimize 0.00% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000029s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.02% : 0.000119s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000010s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000020s : 1: remove_dup_value 14.22% : 0.100539s : 2: renormalize.infer 0.30% : 0.002150s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000027s : 1: rewriter_after_opt_a 0.08% : 0.000576s : 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% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000083s : 1: symbol_engine_optimizer 0.01% : 0.000078s : 1: tuple_transform 43.05% : 0.304432s : 1: type_inference TotalTime = 0.27794, [21] [bootstrap]: 0.00093455 [type_inference]: 0.240123 [event_method]: 0.00031225 [auto_monad]: 0.00017309 [graph_reusing]: 1.161e-05 [inline]: 3.81001e-06 [add_attr]: 0.0248548, [1] [add_attr_with_inline]: 0.0247954, [1] [Cycle 1]: 0.00012603, [2] [tag_attr]: 6.245e-05 [meta_addattr_fg_expand]: 1.252e-05 [parallel-infer-symbol]: 3.76001e-06 [pre_auto_parallel]: 7.953e-05 [insert-virtual-dataset]: 2.76e-06 [parallel-infer-symbol-second]: 1.89e-06 [dataset_repeat_opt]: 2.01e-06 [pipeline_split]: 1.49998e-06 [optimize]: 0.0103229, [53] [py_interpret_to_execute]: 1.145e-05 [rewriter_before_opt_a]: 0.00037269 [opt_a]: 0.00724399, [2] [Cycle 1]: 0.00649401, [45] [expand_dump_flag]: 7.76001e-06 [switch_simplify]: 0.0002174 [loop_unroll]: 7.073e-05 [a_1]: 0.00165901 [with_stream_mark]: 3.049e-05 [recompute_prepare]: 1.233e-05 [updatestate_depend_eliminate]: 4.65999e-06 [updatestate_assign_eliminate]: 4.63999e-06 [updatestate_loads_eliminate]: 3.42002e-06 [parameter_eliminate]: 2.61e-06 [a_2]: 8.712e-05 [accelerated_algorithm]: 2.107e-05 [shard]: 2.14e-06 [meta_shard_fg_expand]: 5.00001e-06 [shard_inline]: 7.31001e-06 [merge_send_recv]: 9.72999e-06 [auto_parallel]: 1.068e-05 [parallel]: 2.261e-05 [flash_sp]: 9.05999e-06 [merge_comm]: 3.65e-06 [allreduce_fusion]: 3.53e-06 [matmul_add_comm_reduction]: 1.202e-05 [allreduce_slice_to_reducescatter]: 9.39996e-07 [virtual_shard_identity]: 9.04998e-06 [virtual_dataset]: 7.00002e-06 [get_grad_eliminate_]: 7e-06 [virtual_output]: 7.37997e-06 [merge_forward]: 4.82e-06 [cell_reuse_recompute_pass]: 1.91998e-06 [offload_activation]: 1.02e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.43e-05 [merge_recompute_call_nodes]: 1.55999e-06 [before_grad]: 1.439e-05 [set_forward_comm_id_for_comm_node_pass]: 3.77002e-06 [meta_fg_expand]: 3.91999e-06 [flash_sp_send_recv_attached]: 5.22e-06 [receive_attached]: 2.34999e-06 [after_resolve]: 1.213e-05 [a_after_grad]: 9.69999e-06 [renormalize]: 0.00370526 [add_forward_monad_depend]: 1.255e-05 [auto_monad_grad]: 3.03e-06 [auto_monad_eliminator]: 2.526e-05 [cse]: 3.816e-05 [a_3]: 5.965e-05 [Cycle 2]: 0.00073224, [45] [expand_dump_flag]: 2.21998e-06 [switch_simplify]: 8.43999e-06 [loop_unroll]: 5.87001e-06 [a_1]: 0.00016669 [with_stream_mark]: 1.734e-05 [recompute_prepare]: 6.41e-06 [updatestate_depend_eliminate]: 4.07e-06 [updatestate_assign_eliminate]: 2.99001e-06 [updatestate_loads_eliminate]: 3.4e-06 [parameter_eliminate]: 2.80002e-06 [a_2]: 7.606e-05 [accelerated_algorithm]: 1.157e-05 [shard]: 2.25002e-06 [meta_shard_fg_expand]: 2.08002e-06 [shard_inline]: 5.89e-06 [merge_send_recv]: 8.53001e-06 [auto_parallel]: 9.14998e-06 [parallel]: 1.021e-05 [flash_sp]: 4.80999e-06 [merge_comm]: 3.16001e-06 [allreduce_fusion]: 3.10998e-06 [matmul_add_comm_reduction]: 1.088e-05 [allreduce_slice_to_reducescatter]: 1.10999e-06 [virtual_shard_identity]: 6.39999e-06 [virtual_dataset]: 5.94999e-06 [get_grad_eliminate_]: 6.02001e-06 [virtual_output]: 5.11002e-06 [merge_forward]: 4.51002e-06 [cell_reuse_recompute_pass]: 3.03998e-06 [offload_activation]: 1.051e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.226e-05 [merge_recompute_call_nodes]: 1.38002e-06 [before_grad]: 1.007e-05 [set_forward_comm_id_for_comm_node_pass]: 4.37e-06 [meta_fg_expand]: 2.93e-06 [flash_sp_send_recv_attached]: 1.99e-06 [receive_attached]: 2.44001e-06 [after_resolve]: 1.201e-05 [a_after_grad]: 8.36002e-06 [renormalize]: 7.00238e-08 [add_forward_monad_depend]: 1.82001e-06 [auto_monad_grad]: 1.84e-06 [auto_monad_eliminator]: 9.84999e-06 [cse]: 1.891e-05 [a_3]: 3.218e-05 [py_interpret_to_execute_after_opt_a]: 1.165e-05 [slice_cell_reuse_recomputed_activation]: 1.77001e-06 [rewriter_after_opt_a]: 2.258e-05 [convert_after_rewriter]: 1.43002e-06 [order_py_execute_after_rewriter]: 1.10999e-06 [mutable_eliminate]: 0.00082955 [opt_b]: 0.00021708, [1] [Cycle 1]: 0.00020655, [7] [b_1]: 0.00011806 [b_2]: 8.53001e-06 [updatestate_depend_eliminate]: 9.02e-06 [updatestate_assign_eliminate]: 2.49001e-06 [updatestate_loads_eliminate]: 2.91e-06 [renormalize]: 1.17e-06 [cse]: 2.907e-05 [optimize_parallel_all_gather_comm]: 8.017e-05 [overlap_param_gather]: 2.56e-06 [cconv]: 0.00010673 [loop_unroll]: 0.00062841 [opt_after_cconv]: 0.00011504, [1] [Cycle 1]: 0.00010651, [7] [c_1]: 2.877e-05 [parameter_eliminate]: 5.45001e-06 [updatestate_depend_eliminate]: 7.11999e-06 [updatestate_assign_eliminate]: 2.51e-06 [updatestate_loads_eliminate]: 3.58999e-06 [cse]: 2.579e-05 [renormalize]: 8.09989e-07 [remove_dup_value]: 1.765e-05 [tuple_transform]: 7.699e-05, [1] [Cycle 1]: 7.106e-05, [4] [d_1]: 4.398e-05 [none_parameter_eliminate]: 1.91e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 7.22997e-06 [partial_unused_args_eliminate]: 1.89e-06 [add_recomputation]: 0.00010646 [cse_after_recomputation]: 2.725e-05, [1] [Cycle 1]: 2.22e-05, [1] [cse]: 1.598e-05 [environ_conv]: 6.18002e-06 [swap_dp_allreduce_reducescatter]: 5.24998e-06 [bias_add_comm_swap]: 3.65e-06 [label_micro_interleaved_index]: 6.81001e-06 [label_fine_grained_interleaved_index]: 2.39001e-06 [merge_cast_opt]: 1.42999e-06 [slice_recompute_activation]: 2.34001e-06 [micro_interleaved_order_control]: 2.59001e-06 [assign_add_opt]: 1.34e-06 [ForceFp32Comm]: 8.79983e-07 [remove_cast_before_assign_add]: 1.16002e-06 [full_micro_interleaved_order_control]: 2.48e-06 [reorder_send_recv_between_fp_bp]: 2.81e-06 [comm_op_add_attrs]: 1.20001e-06 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.12999e-06 [interleave_parallel_branches]: 1.07e-06 [overlap_opt_shard_in_pipeline]: 1.67001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.74998e-06 [control_data_broadcast_order]: 1.486e-05 [grouped_pairwise_exchange_alltoall]: 1.39e-06 [offloading_packed_experts]: 4.17e-06 [overlap_recompute_and_grad_model_parallel]: 5.83997e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.14e-06 [overlap_recompute_allgather_and_fa_grad]: 1.30001e-06 [overlap_recompute_comm]: 2.29001e-06 [overlap_grad_ring_attention]: 4.97999e-06 [overlap_grad_flash_sp]: 2.229e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.32001e-06 [split_layernorm_comm]: 2.37001e-06 [handle_group_info]: 1.37999e-06 [symbol_engine_optimizer]: 7.893e-05, [1] [Cycle 1]: 7.395e-05, [6] [build]: 3.86001e-06 [elim_shapecalc]: 1.17e-05 [elim_not_effective]: 1.302e-05 [opt_reshape]: 7.03e-06 [fold_const_symbol]: 9.67001e-06 [renormalize]: 2.00002e-07 [detach_backward]: 9.207e-05 [pipeline_parallel_scheduler]: 1.97001e-06 [auto_monad_reorder]: 2.083e-05 [get_jit_bprop_graph]: 2.36e-06 [rewriter_after_jit_bprop_graph]: 8.08999e-06 [opt_after_jit_grad]: 0.00069117 [validate]: 5.244e-05 Sums bootstrap : 0.000935s : 0.37% type_inference : 0.240123s : 95.30% event_method : 0.000312s : 0.12% auto_monad : 0.000173s : 0.07% graph_reusing : 0.000012s : 0.00% inline : 0.000004s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000062s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000013s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000080s : 0.03% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000001s : 0.00% optimize.py_interpret_to_execute : 0.000011s : 0.00% optimize.rewriter_before_opt_a : 0.000373s : 0.15% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000226s : 0.09% optimize.opt_a.loop_unroll : 0.000077s : 0.03% optimize.opt_a.a_1 : 0.001826s : 0.72% optimize.opt_a.with_stream_mark : 0.000048s : 0.02% optimize.opt_a.recompute_prepare : 0.000019s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000007s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000163s : 0.06% optimize.opt_a.accelerated_algorithm : 0.000033s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000013s : 0.01% optimize.opt_a.merge_send_recv : 0.000018s : 0.01% optimize.opt_a.auto_parallel : 0.000020s : 0.01% optimize.opt_a.parallel : 0.000033s : 0.01% optimize.opt_a.flash_sp : 0.000014s : 0.01% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000007s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000023s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000015s : 0.01% optimize.opt_a.virtual_dataset : 0.000013s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000013s : 0.01% optimize.opt_a.virtual_output : 0.000012s : 0.00% optimize.opt_a.merge_forward : 0.000009s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000021s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000027s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000024s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.00% optimize.opt_a.meta_fg_expand : 0.000007s : 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 : 0.000024s : 0.01% optimize.opt_a.a_after_grad : 0.000018s : 0.01% optimize.opt_a.renormalize : 0.003705s : 1.47% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000035s : 0.01% optimize.opt_a.cse : 0.000057s : 0.02% optimize.opt_a.a_3 : 0.000092s : 0.04% optimize.py_interpret_to_execute_after_opt_a : 0.000012s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000023s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000830s : 0.33% optimize.opt_b.b_1 : 0.000118s : 0.05% optimize.opt_b.b_2 : 0.000009s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000029s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000080s : 0.03% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000107s : 0.04% optimize.loop_unroll : 0.000628s : 0.25% optimize.opt_after_cconv.c_1 : 0.000029s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000026s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000018s : 0.01% optimize.tuple_transform.d_1 : 0.000044s : 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 : 0.000007s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000106s : 0.04% optimize.cse_after_recomputation.cse : 0.000016s : 0.01% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000002s : 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 : 0.000015s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000022s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000012s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000013s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000010s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000092s : 0.04% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000021s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000691s : 0.27% validate : 0.000052s : 0.02% Time group info: ------[substitution.] 0.000649 51 19.23% : 0.000125s : 5: substitution.arithmetic_simplify 2.36% : 0.000015s : 1: substitution.cast_eliminate 0.26% : 0.000002s : 2: substitution.elim_not_effective 0.32% : 0.000002s : 2: substitution.fold_const_symbol 1.00% : 0.000006s : 3: substitution.graph_param_transform 67.33% : 0.000437s : 19: substitution.inline 0.90% : 0.000006s : 4: substitution.j_node_and_user_rematch 2.28% : 0.000015s : 2: substitution.less_batch_normalization 0.96% : 0.000006s : 4: substitution.remove_not_recompute_node 0.94% : 0.000006s : 2: substitution.replace_old_param 2.58% : 0.000017s : 5: substitution.switch_simplify 1.84% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.239917 2 98.26% : 0.235741s : 1: type_inference.infer 1.74% : 0.004175s : 1: type_inference.specialize ------[replace.] 0.000246 27 2.67% : 0.000007s : 1: replace.cast_eliminate 58.77% : 0.000144s : 19: replace.inline 29.35% : 0.000072s : 5: replace.switch_simplify 9.21% : 0.000023s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000467 27 3.06% : 0.000014s : 1: match.cast_eliminate 91.57% : 0.000428s : 19: match.inline 3.01% : 0.000014s : 5: match.switch_simplify 2.36% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000341 1960 1.22% : 0.000004s : 26: predicate.accumulaten_eliminater 0.62% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.28% : 0.000001s : 6: predicate.addn_check_dump 1.16% : 0.000004s : 26: predicate.addn_zero_filter 1.10% : 0.000004s : 26: predicate.adjust_all_reduce_mul_add 2.42% : 0.000008s : 32: predicate.arithmetic_simplify 1.41% : 0.000005s : 27: predicate.cast_eliminate 0.30% : 0.000001s : 6: predicate.check_bprop_eliminate 0.29% : 0.000001s : 6: predicate.compare_switch_simplify 0.08% : 0.000000s : 3: predicate.const_output_eliminate 0.39% : 0.000001s : 6: predicate.depend_value_elim 1.35% : 0.000005s : 27: predicate.dict_get_item_const_eliminator 1.29% : 0.000004s : 27: predicate.dict_get_item_eliminator 1.33% : 0.000005s : 27: predicate.dict_set_item_eliminator 0.53% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.15% : 0.000001s : 3: predicate.elim_not_effective 0.26% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.37% : 0.000005s : 30: predicate.environ_add_const_eliminate 1.29% : 0.000004s : 30: predicate.environ_get_add_eliminate 1.24% : 0.000004s : 30: predicate.environ_get_depend_swap 1.53% : 0.000005s : 36: predicate.environ_get_eliminate 1.23% : 0.000004s : 30: predicate.environ_get_set_eliminate 2.12% : 0.000007s : 48: predicate.exchange_switch_depend_value 3.23% : 0.000011s : 48: predicate.float_depend_g_call 0.26% : 0.000001s : 6: predicate.float_environ_get_switch 0.45% : 0.000002s : 9: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 3: predicate.fold_const_symbol 0.46% : 0.000002s : 6: predicate.get_grad_eliminate 0.16% : 0.000001s : 3: predicate.graph_param_transform 0.27% : 0.000001s : 6: predicate.incorporate_call 0.23% : 0.000001s : 6: predicate.incorporate_call_switch 6.19% : 0.000021s : 92: predicate.inline 0.47% : 0.000002s : 6: predicate.inline_without_move 0.14% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.64% : 0.000002s : 6: predicate.less_batch_normalization 1.59% : 0.000005s : 35: predicate.list_to_tuple_eliminator_ 2.60% : 0.000009s : 61: predicate.load_eliminater 0.61% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.57% : 0.000012s : 84: predicate.loop_unroll_before_grad 1.75% : 0.000006s : 33: predicate.make_slice_get_slice_eliminator 0.27% : 0.000001s : 6: predicate.merge_addn 0.24% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.27% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.17% : 0.000004s : 26: predicate.minmaximum_grad 0.90% : 0.000003s : 3: predicate.mutable_eliminate 0.20% : 0.000001s : 3: predicate.opt_reshape 0.25% : 0.000001s : 3: predicate.parallel_virtual_node 2.87% : 0.000010s : 48: predicate.partial_defer_inline 1.48% : 0.000005s : 32: predicate.partial_eliminate 1.06% : 0.000004s : 26: predicate.print_const_string_wrapper 0.27% : 0.000001s : 6: predicate.reduce_all_const_elim 1.53% : 0.000005s : 26: predicate.reduce_eliminate 2.62% : 0.000009s : 61: predicate.redundant_stop_gradient_eliminater 0.25% : 0.000001s : 6: predicate.remove_not_recompute_node 1.22% : 0.000004s : 35: predicate.replace_applicator 0.20% : 0.000001s : 6: predicate.replace_old_param 0.14% : 0.000000s : 3: predicate.reset_defer_inline 1.17% : 0.000004s : 26: predicate.reshape_eliminate 0.30% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.28% : 0.000001s : 3: predicate.row_tensor_eliminate 0.45% : 0.000002s : 6: predicate.same_eliminate 0.20% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.56% : 0.000002s : 6: predicate.shard_identity_eliminate 0.33% : 0.000001s : 6: predicate.special_op_eliminate 0.32% : 0.000001s : 6: predicate.specialize_transform 0.51% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.47% : 0.000002s : 6: predicate.stack_unstack_eliminate 0.14% : 0.000000s : 3: predicate.switch_call_monad_eliminater 2.30% : 0.000008s : 48: predicate.switch_defer_inline 2.59% : 0.000009s : 54: predicate.switch_layer_defer_inline 7.29% : 0.000025s : 151: predicate.switch_simplify 1.24% : 0.000004s : 26: predicate.tile_eliminate 1.21% : 0.000004s : 26: predicate.transpose_eliminate 1.60% : 0.000005s : 33: predicate.tuple_list_convert_item_index_to_positive 1.63% : 0.000006s : 33: predicate.tuple_list_get_item_const_eliminator 1.55% : 0.000005s : 33: predicate.tuple_list_get_item_depend_reorder 2.95% : 0.000010s : 41: predicate.tuple_list_get_item_eliminator 1.57% : 0.000005s : 33: predicate.tuple_list_get_set_item_eliminator 2.12% : 0.000007s : 39: predicate.tuple_list_set_item_eliminator 1.74% : 0.000006s : 35: predicate.tuple_to_list_eliminator_ 2.46% : 0.000008s : 61: predicate.updatestate_pure_node_eliminater 2.85% : 0.000010s : 67: predicate.updatestate_useless_node_eliminater 0.18% : 0.000001s : 3: predicate.value_based_eliminate 0.58% : 0.000002s : 6: predicate.virtual_dataset_eliminate 0.35% : 0.000001s : 6: predicate.virtual_output_eliminate 0.13% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.32% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002379 33 53.05% : 0.001262s : 12: func_graph_cloner_run.FuncGraphClonerGraph 46.95% : 0.001117s : 21: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.319369 192 0.00% : 0.000003s : 1: ForceFp32Comm 7.78% : 0.024862s : 1: add_attr 7.77% : 0.024801s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 0.03% : 0.000112s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.06% : 0.000187s : 1: auto_monad 0.01% : 0.000025s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.31% : 0.000981s : 1: bootstrap 0.03% : 0.000112s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000018s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000031s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.03% : 0.000098s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.10% : 0.000330s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000016s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000005s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.20% : 0.000641s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.26% : 0.000846s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000018s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000023s : 1: opt.transform.mutable_eliminate 0.78% : 0.002490s : 78: opt.transform.opt_a 0.01% : 0.000028s : 1: opt.transform.opt_after_cconv 0.01% : 0.000029s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000094s : 28: opt.transform.opt_b 0.02% : 0.000049s : 2: opt.transform.opt_trans_graph 0.01% : 0.000038s : 4: opt.transform.symbol_engine_opt 2.27% : 0.007248s : 1: opt_a 0.04% : 0.000119s : 1: opt_after_cconv 0.22% : 0.000707s : 1: opt_after_jit_grad 0.07% : 0.000221s : 1: opt_b 3.23% : 0.010331s : 1: optimize 0.03% : 0.000086s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000025s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 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% : 0.000005s : 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 0.03% : 0.000085s : 1: pre_auto_parallel 0.00% : 0.000016s : 1: py_interpret_to_execute 0.00% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000022s : 1: remove_dup_value 0.46% : 0.001471s : 1: renormalize.infer 0.69% : 0.002218s : 1: renormalize.specialize 0.00% : 0.000008s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000012s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000026s : 1: rewriter_after_opt_a 0.12% : 0.000382s : 1: rewriter_before_opt_a 0.00% : 0.000004s : 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% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.03% : 0.000081s : 1: symbol_engine_optimizer 0.02% : 0.000080s : 1: tuple_transform 75.20% : 0.240165s : 1: type_inference TotalTime = 0.229327, [21] [bootstrap]: 0.0004484 [type_inference]: 0.170831 [event_method]: 6.564e-05 [auto_monad]: 0.00024742 [graph_reusing]: 1.956e-05 [inline]: 2.83998e-06 [add_attr]: 0.00446414, [1] [add_attr_with_inline]: 0.00444942, [1] [Cycle 1]: 0.00017452, [2] [tag_attr]: 9.537e-05 [meta_addattr_fg_expand]: 2.136e-05 [parallel-infer-symbol]: 3.96001e-06 [pre_auto_parallel]: 0.0001186 [insert-virtual-dataset]: 3.08998e-06 [parallel-infer-symbol-second]: 9.09989e-07 [dataset_repeat_opt]: 1.99e-06 [pipeline_split]: 1.59998e-06 [optimize]: 0.0523351, [53] [py_interpret_to_execute]: 1.165e-05 [rewriter_before_opt_a]: 0.00055551 [opt_a]: 0.0494599, [3] [Cycle 1]: 0.0451197, [45] [expand_dump_flag]: 1.396e-05 [switch_simplify]: 0.00031199 [loop_unroll]: 0.00011174 [a_1]: 0.00302817 [with_stream_mark]: 4.549e-05 [recompute_prepare]: 3.821e-05 [updatestate_depend_eliminate]: 1.146e-05 [updatestate_assign_eliminate]: 8.50999e-06 [updatestate_loads_eliminate]: 7.53e-06 [parameter_eliminate]: 5.14998e-06 [a_2]: 0.00028785 [accelerated_algorithm]: 3.816e-05 [shard]: 2.49001e-06 [meta_shard_fg_expand]: 1.281e-05 [shard_inline]: 1.771e-05 [merge_send_recv]: 2.188e-05 [auto_parallel]: 1.934e-05 [parallel]: 2.367e-05 [flash_sp]: 1.547e-05 [merge_comm]: 1.024e-05 [allreduce_fusion]: 9.15001e-06 [matmul_add_comm_reduction]: 4.229e-05 [allreduce_slice_to_reducescatter]: 1.10999e-06 [virtual_shard_identity]: 2.194e-05 [virtual_dataset]: 1.641e-05 [get_grad_eliminate_]: 1.547e-05 [virtual_output]: 1.556e-05 [merge_forward]: 1.169e-05 [cell_reuse_recompute_pass]: 1.86e-06 [offload_activation]: 2.02e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.422e-05 [merge_recompute_call_nodes]: 1.96e-06 [before_grad]: 3.177e-05 [set_forward_comm_id_for_comm_node_pass]: 1.429e-05 [meta_fg_expand]: 0.0349479 [flash_sp_send_recv_attached]: 7.83999e-06 [receive_attached]: 2.42001e-06 [after_resolve]: 9.086e-05 [a_after_grad]: 9.388e-05 [renormalize]: 0.00453197 [add_forward_monad_depend]: 1.396e-05 [auto_monad_grad]: 6.17999e-06 [auto_monad_eliminator]: 5.93e-05 [cse]: 0.00029327 [a_3]: 0.00034692 [Cycle 2]: 0.00352024, [45] [expand_dump_flag]: 2.83998e-06 [switch_simplify]: 4.438e-05 [loop_unroll]: 3.941e-05 [a_1]: 0.00141414 [with_stream_mark]: 2.373e-05 [recompute_prepare]: 1.092e-05 [updatestate_depend_eliminate]: 5.57999e-06 [updatestate_assign_eliminate]: 4e-06 [updatestate_loads_eliminate]: 3.26001e-06 [parameter_eliminate]: 2.36998e-06 [a_2]: 0.00015223 [accelerated_algorithm]: 2.307e-05 [shard]: 2.34001e-06 [meta_shard_fg_expand]: 2.89999e-06 [shard_inline]: 7.22002e-06 [merge_send_recv]: 8.85001e-06 [auto_parallel]: 1.077e-05 [parallel]: 9.47999e-06 [flash_sp]: 4.38999e-06 [merge_comm]: 4.13001e-06 [allreduce_fusion]: 3.75e-06 [matmul_add_comm_reduction]: 1.181e-05 [allreduce_slice_to_reducescatter]: 1.07998e-06 [virtual_shard_identity]: 8.35001e-06 [virtual_dataset]: 6.98998e-06 [get_grad_eliminate_]: 6.43e-06 [virtual_output]: 6.12001e-06 [merge_forward]: 4.68001e-06 [cell_reuse_recompute_pass]: 1.39998e-06 [offload_activation]: 1.089e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.422e-05 [merge_recompute_call_nodes]: 1.50999e-06 [before_grad]: 1.193e-05 [set_forward_comm_id_for_comm_node_pass]: 3.9e-06 [meta_fg_expand]: 0.00030609 [flash_sp_send_recv_attached]: 2.32999e-06 [receive_attached]: 2.73e-06 [after_resolve]: 1.841e-05 [a_after_grad]: 1.049e-05 [renormalize]: 0.0009313 [add_forward_monad_depend]: 6.11e-06 [auto_monad_grad]: 2.68998e-06 [auto_monad_eliminator]: 1.779e-05 [cse]: 3.616e-05 [a_3]: 5.318e-05 [Cycle 3]: 0.00079419, [45] [expand_dump_flag]: 2.03997e-06 [switch_simplify]: 8.88002e-06 [loop_unroll]: 6.56e-06 [a_1]: 0.00017361 [with_stream_mark]: 1.167e-05 [recompute_prepare]: 7.13998e-06 [updatestate_depend_eliminate]: 4.80001e-06 [updatestate_assign_eliminate]: 3.34001e-06 [updatestate_loads_eliminate]: 2.58003e-06 [parameter_eliminate]: 1.66e-06 [a_2]: 8.99e-05 [accelerated_algorithm]: 1.056e-05 [shard]: 1.13001e-06 [meta_shard_fg_expand]: 1.94e-06 [shard_inline]: 7.1e-06 [merge_send_recv]: 6.68e-06 [auto_parallel]: 7.33999e-06 [parallel]: 6.69999e-06 [flash_sp]: 9.20001e-07 [merge_comm]: 3.98999e-06 [allreduce_fusion]: 4.03001e-06 [matmul_add_comm_reduction]: 8.02003e-06 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 8.03001e-06 [virtual_dataset]: 3.71e-05 [get_grad_eliminate_]: 8.3e-06 [virtual_output]: 7.08e-06 [merge_forward]: 4.33001e-06 [cell_reuse_recompute_pass]: 2.06e-06 [offload_activation]: 9.48002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.5e-05 [merge_recompute_call_nodes]: 9.90025e-07 [before_grad]: 1.24e-05 [set_forward_comm_id_for_comm_node_pass]: 4.55999e-06 [meta_fg_expand]: 2.46998e-06 [flash_sp_send_recv_attached]: 1.15001e-06 [receive_attached]: 1.24e-06 [after_resolve]: 1.129e-05 [a_after_grad]: 9.99999e-06 [renormalize]: 6.00121e-08 [add_forward_monad_depend]: 1.87999e-06 [auto_monad_grad]: 1.44e-06 [auto_monad_eliminator]: 9.44e-06 [cse]: 2.149e-05 [a_3]: 4.069e-05 [py_interpret_to_execute_after_opt_a]: 7.55e-06 [slice_cell_reuse_recomputed_activation]: 2.39999e-06 [rewriter_after_opt_a]: 2.45e-05 [convert_after_rewriter]: 1.15001e-06 [order_py_execute_after_rewriter]: 1.56002e-06 [mutable_eliminate]: 0.00076116 [opt_b]: 0.0002128, [1] [Cycle 1]: 0.00020433, [7] [b_1]: 0.00012483 [b_2]: 7.38999e-06 [updatestate_depend_eliminate]: 6.88e-06 [updatestate_assign_eliminate]: 3.05002e-06 [updatestate_loads_eliminate]: 3.37002e-06 [renormalize]: 6.39993e-07 [cse]: 2.462e-05 [optimize_parallel_all_gather_comm]: 1.953e-05 [overlap_param_gather]: 2.22999e-06 [cconv]: 3.146e-05 [loop_unroll]: 0.00053105 [opt_after_cconv]: 0.00010919, [1] [Cycle 1]: 0.00010209, [7] [c_1]: 3.074e-05 [parameter_eliminate]: 4.13999e-06 [updatestate_depend_eliminate]: 6.14001e-06 [updatestate_assign_eliminate]: 2.69001e-06 [updatestate_loads_eliminate]: 2.54001e-06 [cse]: 2.263e-05 [renormalize]: 2.70025e-07 [remove_dup_value]: 1.685e-05 [tuple_transform]: 7.465e-05, [1] [Cycle 1]: 6.986e-05, [4] [d_1]: 4.367e-05 [none_parameter_eliminate]: 1.59e-06 [renormalize]: 1.69995e-07 [switch_simplify]: 6.91001e-06 [partial_unused_args_eliminate]: 2.32999e-06 [add_recomputation]: 5.519e-05 [cse_after_recomputation]: 2.35e-05, [1] [Cycle 1]: 1.909e-05, [1] [cse]: 1.412e-05 [environ_conv]: 5.64998e-06 [swap_dp_allreduce_reducescatter]: 6.46999e-06 [bias_add_comm_swap]: 3.01001e-06 [label_micro_interleaved_index]: 4.90001e-06 [label_fine_grained_interleaved_index]: 2.73003e-06 [merge_cast_opt]: 1.45001e-06 [slice_recompute_activation]: 2.16998e-06 [micro_interleaved_order_control]: 2.41998e-06 [assign_add_opt]: 1.37e-06 [ForceFp32Comm]: 1.19998e-06 [remove_cast_before_assign_add]: 9.79984e-07 [full_micro_interleaved_order_control]: 2.16998e-06 [reorder_send_recv_between_fp_bp]: 2.85998e-06 [comm_op_add_attrs]: 1.02e-06 [add_comm_op_reuse_tag]: 1.22e-06 [interleave_split_concat_branches]: 1.09e-06 [interleave_parallel_branches]: 1.04e-06 [overlap_opt_shard_in_pipeline]: 1.29e-06 [overlap_opt_shard_grad_in_pipeline]: 1.91e-06 [control_data_broadcast_order]: 1.459e-05 [grouped_pairwise_exchange_alltoall]: 1.30001e-06 [offloading_packed_experts]: 4.50999e-06 [overlap_recompute_and_grad_model_parallel]: 5.63002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.28002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.24e-06 [overlap_recompute_comm]: 2.43e-06 [overlap_grad_ring_attention]: 4.40999e-06 [overlap_grad_flash_sp]: 2.358e-05 [begin_end_overlap_inline]: 5.3001e-07 [split_matmul_comm_elemetwise]: 2.05002e-06 [split_layernorm_comm]: 1.97999e-06 [handle_group_info]: 1.66002e-06 [symbol_engine_optimizer]: 7.623e-05, [1] [Cycle 1]: 7.118e-05, [6] [build]: 2.76999e-06 [elim_shapecalc]: 1.011e-05 [elim_not_effective]: 1.394e-05 [opt_reshape]: 7.09001e-06 [fold_const_symbol]: 1.098e-05 [renormalize]: 1.79978e-07 [detach_backward]: 2.31e-06 [pipeline_parallel_scheduler]: 1.36002e-06 [auto_monad_reorder]: 2.134e-05 [get_jit_bprop_graph]: 1.83002e-06 [rewriter_after_jit_bprop_graph]: 4.15999e-06 [opt_after_jit_grad]: 0.00050472 [validate]: 4.963e-05 Sums bootstrap : 0.000448s : 0.20% type_inference : 0.170831s : 76.46% event_method : 0.000066s : 0.03% auto_monad : 0.000247s : 0.11% graph_reusing : 0.000020s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000095s : 0.04% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000021s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000119s : 0.05% 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 : 0.000012s : 0.01% optimize.rewriter_before_opt_a : 0.000556s : 0.25% optimize.opt_a.expand_dump_flag : 0.000019s : 0.01% optimize.opt_a.switch_simplify : 0.000365s : 0.16% optimize.opt_a.loop_unroll : 0.000158s : 0.07% optimize.opt_a.a_1 : 0.004616s : 2.07% optimize.opt_a.with_stream_mark : 0.000081s : 0.04% optimize.opt_a.recompute_prepare : 0.000056s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000022s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000016s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000013s : 0.01% optimize.opt_a.parameter_eliminate : 0.000009s : 0.00% optimize.opt_a.a_2 : 0.000530s : 0.24% optimize.opt_a.accelerated_algorithm : 0.000072s : 0.03% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000018s : 0.01% optimize.opt_a.shard_inline : 0.000032s : 0.01% optimize.opt_a.merge_send_recv : 0.000037s : 0.02% optimize.opt_a.auto_parallel : 0.000037s : 0.02% optimize.opt_a.parallel : 0.000040s : 0.02% optimize.opt_a.flash_sp : 0.000021s : 0.01% optimize.opt_a.merge_comm : 0.000018s : 0.01% optimize.opt_a.allreduce_fusion : 0.000017s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000062s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000038s : 0.02% optimize.opt_a.virtual_dataset : 0.000060s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000030s : 0.01% optimize.opt_a.virtual_output : 0.000029s : 0.01% optimize.opt_a.merge_forward : 0.000021s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000041s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000063s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000056s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000023s : 0.01% optimize.opt_a.meta_fg_expand : 0.035256s : 15.78% optimize.opt_a.flash_sp_send_recv_attached : 0.000011s : 0.01% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000121s : 0.05% optimize.opt_a.a_after_grad : 0.000114s : 0.05% optimize.opt_a.renormalize : 0.005463s : 2.45% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.01% optimize.opt_a.auto_monad_grad : 0.000010s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000087s : 0.04% optimize.opt_a.cse : 0.000351s : 0.16% optimize.opt_a.a_3 : 0.000441s : 0.20% optimize.py_interpret_to_execute_after_opt_a : 0.000008s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000024s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000002s : 0.00% optimize.mutable_eliminate : 0.000761s : 0.34% optimize.opt_b.b_1 : 0.000125s : 0.06% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000025s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000031s : 0.01% optimize.loop_unroll : 0.000531s : 0.24% optimize.opt_after_cconv.c_1 : 0.000031s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000023s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000044s : 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 : 0.000007s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000055s : 0.02% optimize.cse_after_recomputation.cse : 0.000014s : 0.01% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000015s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000024s : 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 : 0.000002s : 0.00% optimize.symbol_engine_optimizer.build : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000011s : 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 : 0.000021s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000505s : 0.23% validate : 0.000050s : 0.02% Time group info: ------[substitution.] 0.001583 199 11.25% : 0.000178s : 6: substitution.arithmetic_simplify 0.96% : 0.000015s : 1: substitution.cast_eliminate 0.13% : 0.000002s : 3: substitution.elim_not_effective 0.70% : 0.000011s : 11: substitution.float_depend_g_call 0.35% : 0.000005s : 3: substitution.float_tuple_getitem_switch 0.11% : 0.000002s : 3: substitution.fold_const_symbol 0.40% : 0.000006s : 4: substitution.graph_param_transform 0.22% : 0.000004s : 2: substitution.incorporate_call 0.19% : 0.000003s : 2: substitution.incorporate_call_switch 66.11% : 0.001046s : 36: substitution.inline 1.74% : 0.000028s : 2: substitution.inline_without_move 0.82% : 0.000013s : 16: substitution.j_node_and_user_rematch 1.71% : 0.000027s : 3: substitution.less_batch_normalization 0.84% : 0.000013s : 7: substitution.minmaximum_grad 0.62% : 0.000010s : 11: substitution.partial_eliminate 0.90% : 0.000014s : 16: substitution.remove_not_recompute_node 1.72% : 0.000027s : 9: substitution.replace_applicator 0.77% : 0.000012s : 8: substitution.replace_old_param 0.35% : 0.000006s : 1: substitution.set_cell_output_no_recompute 1.47% : 0.000023s : 8: substitution.switch_simplify 1.48% : 0.000023s : 7: substitution.tuple_list_convert_item_index_to_positive 0.63% : 0.000010s : 7: substitution.tuple_list_get_item_const_eliminator 1.08% : 0.000017s : 7: substitution.tuple_list_get_item_depend_reorder 4.42% : 0.000070s : 19: substitution.tuple_list_get_item_eliminator 1.02% : 0.000016s : 7: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.170619 2 96.52% : 0.164683s : 1: type_inference.infer 3.48% : 0.005936s : 1: type_inference.specialize ------[replace.] 0.000504 54 1.92% : 0.000010s : 1: replace.cast_eliminate 62.00% : 0.000313s : 36: replace.inline 21.62% : 0.000109s : 8: replace.switch_simplify 14.46% : 0.000073s : 9: replace.tuple_list_get_item_eliminator ------[match.] 0.001087 54 1.31% : 0.000014s : 1: match.cast_eliminate 94.38% : 0.001025s : 36: match.inline 1.81% : 0.000020s : 8: match.switch_simplify 2.50% : 0.000027s : 9: match.tuple_list_get_item_eliminator ------[predicate.] 0.000840 5282 1.18% : 0.000010s : 69: predicate.accumulaten_eliminater 0.24% : 0.000002s : 4: predicate.ad_related_special_op_eliminate 0.37% : 0.000003s : 21: predicate.addn_check_dump 1.30% : 0.000011s : 69: predicate.addn_zero_filter 1.10% : 0.000009s : 69: predicate.adjust_all_reduce_mul_add 2.12% : 0.000018s : 90: predicate.arithmetic_simplify 1.24% : 0.000010s : 70: predicate.cast_eliminate 0.86% : 0.000007s : 51: predicate.check_bprop_eliminate 0.35% : 0.000003s : 21: predicate.compare_switch_simplify 0.04% : 0.000000s : 4: predicate.const_output_eliminate 0.39% : 0.000003s : 21: predicate.depend_value_elim 1.30% : 0.000011s : 70: predicate.dict_get_item_const_eliminator 1.33% : 0.000011s : 70: predicate.dict_get_item_eliminator 1.18% : 0.000010s : 70: predicate.dict_set_item_eliminator 0.23% : 0.000002s : 8: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 4: predicate.elim_not_effective 0.09% : 0.000001s : 4: predicate.elim_shapecalc_of_broadcastargs 1.25% : 0.000010s : 74: predicate.environ_add_const_eliminate 1.21% : 0.000010s : 74: predicate.environ_get_add_eliminate 1.22% : 0.000010s : 74: predicate.environ_get_depend_swap 1.62% : 0.000014s : 95: predicate.environ_get_eliminate 1.22% : 0.000010s : 74: predicate.environ_get_set_eliminate 1.96% : 0.000016s : 115: predicate.exchange_switch_depend_value 2.98% : 0.000025s : 115: predicate.float_depend_g_call 0.37% : 0.000003s : 21: predicate.float_environ_get_switch 0.50% : 0.000004s : 25: predicate.float_tuple_getitem_switch 0.04% : 0.000000s : 4: predicate.fold_const_symbol 0.49% : 0.000004s : 21: predicate.get_grad_eliminate 0.05% : 0.000000s : 4: predicate.graph_param_transform 0.40% : 0.000003s : 21: predicate.incorporate_call 0.31% : 0.000003s : 21: predicate.incorporate_call_switch 5.90% : 0.000050s : 234: predicate.inline 1.08% : 0.000009s : 43: predicate.inline_without_move 0.21% : 0.000002s : 21: predicate.j_node_and_user_rematch 0.65% : 0.000005s : 21: predicate.less_batch_normalization 1.51% : 0.000013s : 87: predicate.list_to_tuple_eliminator_ 2.51% : 0.000021s : 156: predicate.load_eliminater 0.28% : 0.000002s : 4: predicate.loop_unroll_after_grad 3.11% : 0.000026s : 179: predicate.loop_unroll_before_grad 1.44% : 0.000012s : 78: predicate.make_slice_get_slice_eliminator 0.35% : 0.000003s : 21: predicate.merge_addn 0.87% : 0.000007s : 51: predicate.micro_step_allgather_replace 0.88% : 0.000007s : 51: predicate.mini_step_allgather_replace 1.13% : 0.000009s : 69: predicate.minmaximum_grad 0.42% : 0.000004s : 4: predicate.mutable_eliminate 0.09% : 0.000001s : 4: predicate.opt_reshape 0.10% : 0.000001s : 4: predicate.parallel_virtual_node 3.09% : 0.000026s : 115: predicate.partial_defer_inline 1.51% : 0.000013s : 83: predicate.partial_eliminate 1.14% : 0.000010s : 69: predicate.print_const_string_wrapper 0.37% : 0.000003s : 21: predicate.reduce_all_const_elim 1.50% : 0.000013s : 69: predicate.reduce_eliminate 2.58% : 0.000022s : 156: predicate.redundant_stop_gradient_eliminater 0.24% : 0.000002s : 21: predicate.remove_not_recompute_node 1.58% : 0.000013s : 130: predicate.replace_applicator 0.61% : 0.000005s : 43: predicate.replace_old_param 0.07% : 0.000001s : 4: predicate.reset_defer_inline 1.31% : 0.000011s : 69: predicate.reshape_eliminate 0.91% : 0.000008s : 51: predicate.row_tensor_add_zeros_like 0.10% : 0.000001s : 4: predicate.row_tensor_eliminate 1.03% : 0.000009s : 51: predicate.same_eliminate 0.27% : 0.000002s : 21: predicate.set_cell_output_no_recompute 0.50% : 0.000004s : 21: predicate.shard_identity_eliminate 0.25% : 0.000002s : 8: predicate.special_op_eliminate 0.47% : 0.000004s : 21: predicate.specialize_transform 1.06% : 0.000009s : 51: predicate.split_environ_get_set_with_tuple_value 0.99% : 0.000008s : 43: predicate.stack_unstack_eliminate 0.10% : 0.000001s : 4: predicate.switch_call_monad_eliminater 2.20% : 0.000018s : 115: predicate.switch_defer_inline 3.20% : 0.000027s : 166: predicate.switch_layer_defer_inline 6.81% : 0.000057s : 335: predicate.switch_simplify 1.11% : 0.000009s : 69: predicate.tile_eliminate 1.16% : 0.000010s : 69: predicate.transpose_eliminate 1.38% : 0.000012s : 78: predicate.tuple_list_convert_item_index_to_positive 1.46% : 0.000012s : 78: predicate.tuple_list_get_item_const_eliminator 1.39% : 0.000012s : 78: predicate.tuple_list_get_item_depend_reorder 2.68% : 0.000022s : 108: predicate.tuple_list_get_item_eliminator 1.39% : 0.000012s : 78: predicate.tuple_list_get_set_item_eliminator 2.00% : 0.000017s : 99: predicate.tuple_list_set_item_eliminator 1.46% : 0.000012s : 87: predicate.tuple_to_list_eliminator_ 2.50% : 0.000021s : 156: predicate.updatestate_pure_node_eliminater 2.92% : 0.000025s : 177: predicate.updatestate_useless_node_eliminater 0.10% : 0.000001s : 4: predicate.value_based_eliminate 0.41% : 0.000003s : 21: predicate.virtual_dataset_eliminate 0.41% : 0.000003s : 21: predicate.virtual_output_eliminate 0.06% : 0.000001s : 4: predicate.virtual_view_grad_eliminate 0.13% : 0.000001s : 4: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.004101 61 62.55% : 0.002565s : 21: func_graph_cloner_run.FuncGraphClonerGraph 37.45% : 0.001536s : 40: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.298318 233 0.00% : 0.000004s : 1: ForceFp32Comm 1.50% : 0.004472s : 1: add_attr 1.49% : 0.004455s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000060s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.09% : 0.000257s : 1: auto_monad 0.01% : 0.000025s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.16% : 0.000479s : 1: bootstrap 0.01% : 0.000035s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000018s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000026s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.01% : 0.000016s : 1: environ_conv 0.02% : 0.000074s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000024s : 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% : 0.000006s : 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 0.18% : 0.000540s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.26% : 0.000772s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000015s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000020s : 1: opt.transform.mutable_eliminate 2.22% : 0.006614s : 117: opt.transform.opt_a 0.01% : 0.000030s : 1: opt.transform.opt_after_cconv 0.01% : 0.000028s : 1: opt.transform.opt_after_jit_grad 0.04% : 0.000106s : 28: opt.transform.opt_b 0.02% : 0.000048s : 2: opt.transform.opt_trans_graph 0.01% : 0.000038s : 4: opt.transform.symbol_engine_opt 16.58% : 0.049465s : 1: opt_a 0.04% : 0.000112s : 1: opt_after_cconv 0.17% : 0.000517s : 1: opt_after_jit_grad 0.07% : 0.000216s : 1: opt_b 17.55% : 0.052341s : 1: optimize 0.01% : 0.000023s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000027s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.04% : 0.000126s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000011s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000020s : 1: remove_dup_value 1.19% : 0.003554s : 2: renormalize.infer 0.63% : 0.001886s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000028s : 1: rewriter_after_opt_a 0.19% : 0.000573s : 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% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.03% : 0.000079s : 1: symbol_engine_optimizer 0.03% : 0.000077s : 1: tuple_transform 57.28% : 0.170867s : 1: type_inference skip Bool... Tensor.copy (100, 100) Int8 ... Tensor.copy (100, 100) Int16 ... Tensor.copy (100, 100) Int32 ... Tensor.copy (100, 100) Int64 ... Tensor.copy (100, 100) UInt8 ... Tensor.copy (100, 100) UInt16 ... Tensor.copy (100, 100) UInt32 ... Tensor.copy (100, 100) UInt64 ... Tensor.copy (100, 100) Float16 ... Tensor.copy (100, 100) Float32 ... Tensor.copy (100, 100) Float64 ... skip BFloat16... Tensor.copy (100, 100) Complex64 ... Tensor.copy (100, 100) Complex128 ... skip Int4... skip Float8E4M3FN... skip Float8E5M2... skip HiFloat8... . [hook] pytest_runtest_teardown:test_tensor_copy_limits[kbk] tests/st/tensor/test_copy.py::test_tensor_copy_limits[kbk],max_mem:4.0M [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:42:36.784.241 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.247335, [21] [bootstrap]: 0.00070141 [type_inference]: 0.206106 [event_method]: 8.41e-05 [auto_monad]: 0.00018478 [graph_reusing]: 1.327e-05 [inline]: 3.26001e-06 [add_attr]: 0.0291722, [1] [add_attr_with_inline]: 0.0291575, [1] [Cycle 1]: 0.0001276, [2] [tag_attr]: 6.354e-05 [meta_addattr_fg_expand]: 1.328e-05 [parallel-infer-symbol]: 3.53e-06 [pre_auto_parallel]: 9.467e-05 [insert-virtual-dataset]: 3.11001e-06 [parallel-infer-symbol-second]: 8.29983e-07 [dataset_repeat_opt]: 2.04e-06 [pipeline_split]: 1.77999e-06 [optimize]: 0.0100351, [53] [py_interpret_to_execute]: 7.38e-06 [rewriter_before_opt_a]: 0.00041414 [opt_a]: 0.00702793, [2] [Cycle 1]: 0.00609468, [45] [expand_dump_flag]: 6.35002e-06 [switch_simplify]: 0.00025337 [loop_unroll]: 8.353e-05 [a_1]: 0.00187772 [with_stream_mark]: 2.917e-05 [recompute_prepare]: 1.665e-05 [updatestate_depend_eliminate]: 5.95002e-06 [updatestate_assign_eliminate]: 4.92e-06 [updatestate_loads_eliminate]: 4.42998e-06 [parameter_eliminate]: 3.19001e-06 [a_2]: 0.0001098 [accelerated_algorithm]: 9.17001e-06 [shard]: 2.26e-06 [meta_shard_fg_expand]: 4.08999e-06 [shard_inline]: 8.05e-06 [merge_send_recv]: 1.245e-05 [auto_parallel]: 1.098e-05 [parallel]: 3.296e-05 [flash_sp]: 1.133e-05 [merge_comm]: 5.45001e-06 [allreduce_fusion]: 4.82998e-06 [matmul_add_comm_reduction]: 1.307e-05 [allreduce_slice_to_reducescatter]: 8.80013e-07 [virtual_shard_identity]: 1.381e-05 [virtual_dataset]: 7.87e-06 [get_grad_eliminate_]: 8.19002e-06 [virtual_output]: 7.87998e-06 [merge_forward]: 5.77999e-06 [cell_reuse_recompute_pass]: 2.29001e-06 [offload_activation]: 1.339e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.065e-05 [merge_recompute_call_nodes]: 1.43002e-06 [before_grad]: 1.495e-05 [set_forward_comm_id_for_comm_node_pass]: 6.00002e-06 [meta_fg_expand]: 5.10001e-06 [flash_sp_send_recv_attached]: 3.01001e-06 [receive_attached]: 2.06998e-06 [after_resolve]: 1.486e-05 [a_after_grad]: 1.234e-05 [renormalize]: 0.00295083 [add_forward_monad_depend]: 1.07e-05 [auto_monad_grad]: 2.75997e-06 [auto_monad_eliminator]: 2.707e-05 [cse]: 5.227e-05 [a_3]: 7.33e-05 [Cycle 2]: 0.00091507, [45] [expand_dump_flag]: 2.22999e-06 [switch_simplify]: 1.115e-05 [loop_unroll]: 7.52002e-06 [a_1]: 0.00018978 [with_stream_mark]: 2.332e-05 [recompute_prepare]: 8.92e-06 [updatestate_depend_eliminate]: 5.15999e-06 [updatestate_assign_eliminate]: 4.85001e-06 [updatestate_loads_eliminate]: 4.28999e-06 [parameter_eliminate]: 2.59999e-06 [a_2]: 9.535e-05 [accelerated_algorithm]: 8.81002e-06 [shard]: 2.55002e-06 [meta_shard_fg_expand]: 2.79999e-06 [shard_inline]: 8.16002e-06 [merge_send_recv]: 9.72001e-06 [auto_parallel]: 1.091e-05 [parallel]: 8.70001e-06 [flash_sp]: 3.76999e-06 [merge_comm]: 4.79002e-06 [allreduce_fusion]: 4.43999e-06 [matmul_add_comm_reduction]: 1.201e-05 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 1.053e-05 [virtual_dataset]: 7.6e-06 [get_grad_eliminate_]: 4.427e-05 [virtual_output]: 7.85e-06 [merge_forward]: 7.83001e-06 [cell_reuse_recompute_pass]: 2.71e-06 [offload_activation]: 1.274e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.915e-05 [merge_recompute_call_nodes]: 1.50999e-06 [before_grad]: 1.361e-05 [set_forward_comm_id_for_comm_node_pass]: 5.76e-06 [meta_fg_expand]: 4.07e-06 [flash_sp_send_recv_attached]: 2.16e-06 [receive_attached]: 2.12999e-06 [after_resolve]: 1.224e-05 [a_after_grad]: 1.176e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 3.11001e-06 [auto_monad_grad]: 2.06e-06 [auto_monad_eliminator]: 1.206e-05 [cse]: 2.957e-05 [a_3]: 4.838e-05 [py_interpret_to_execute_after_opt_a]: 1.111e-05 [slice_cell_reuse_recomputed_activation]: 2.32001e-06 [rewriter_after_opt_a]: 2.836e-05 [convert_after_rewriter]: 1.16997e-06 [order_py_execute_after_rewriter]: 1.10001e-06 [mutable_eliminate]: 0.00077244 [opt_b]: 0.00027452, [1] [Cycle 1]: 0.00026556, [7] [b_1]: 0.00016197 [b_2]: 9.72001e-06 [updatestate_depend_eliminate]: 9.78002e-06 [updatestate_assign_eliminate]: 3.88001e-06 [updatestate_loads_eliminate]: 3.48e-06 [renormalize]: 8.2e-07 [cse]: 3.801e-05 [optimize_parallel_all_gather_comm]: 2.278e-05 [overlap_param_gather]: 2.12999e-06 [cconv]: 3.85e-05 [loop_unroll]: 0.00054733 [opt_after_cconv]: 0.00013207, [1] [Cycle 1]: 0.00012438, [7] [c_1]: 3.774e-05 [parameter_eliminate]: 4.65999e-06 [updatestate_depend_eliminate]: 8.03999e-06 [updatestate_assign_eliminate]: 3.68999e-06 [updatestate_loads_eliminate]: 4.48001e-06 [cse]: 3.277e-05 [renormalize]: 8.70001e-07 [remove_dup_value]: 5.452e-05 [tuple_transform]: 9.298e-05, [1] [Cycle 1]: 8.704e-05, [4] [d_1]: 5.624e-05 [none_parameter_eliminate]: 1.72001e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 8.44998e-06 [partial_unused_args_eliminate]: 2.16998e-06 [add_recomputation]: 8.125e-05 [cse_after_recomputation]: 3.331e-05, [1] [Cycle 1]: 2.721e-05, [1] [cse]: 2.045e-05 [environ_conv]: 7.08e-06 [swap_dp_allreduce_reducescatter]: 7.80998e-06 [bias_add_comm_swap]: 2.96999e-06 [label_micro_interleaved_index]: 6.28e-06 [label_fine_grained_interleaved_index]: 2.53e-06 [merge_cast_opt]: 1.54e-06 [slice_recompute_activation]: 2.36e-06 [micro_interleaved_order_control]: 2.56998e-06 [assign_add_opt]: 1.29998e-06 [ForceFp32Comm]: 8.39995e-07 [remove_cast_before_assign_add]: 1.12e-06 [full_micro_interleaved_order_control]: 2.48e-06 [reorder_send_recv_between_fp_bp]: 2.53e-06 [comm_op_add_attrs]: 1.44e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.14e-06 [interleave_parallel_branches]: 1.40999e-06 [overlap_opt_shard_in_pipeline]: 1.14e-06 [overlap_opt_shard_grad_in_pipeline]: 1.72999e-06 [control_data_broadcast_order]: 1.764e-05 [grouped_pairwise_exchange_alltoall]: 2.16998e-06 [offloading_packed_experts]: 4.82e-06 [overlap_recompute_and_grad_model_parallel]: 6.02999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.16997e-06 [overlap_recompute_comm]: 2.37001e-06 [overlap_grad_ring_attention]: 4.60001e-06 [overlap_grad_flash_sp]: 2.767e-05 [begin_end_overlap_inline]: 5.60016e-07 [split_matmul_comm_elemetwise]: 2.21e-06 [split_layernorm_comm]: 1.62001e-06 [handle_group_info]: 1.04e-06 [symbol_engine_optimizer]: 9.965e-05, [1] [Cycle 1]: 9.439e-05, [6] [build]: 3.85e-06 [elim_shapecalc]: 1.647e-05 [elim_not_effective]: 1.823e-05 [opt_reshape]: 8.13999e-06 [fold_const_symbol]: 1.373e-05 [renormalize]: 2.59985e-07 [detach_backward]: 2.09e-06 [pipeline_parallel_scheduler]: 1.49e-06 [auto_monad_reorder]: 2.418e-05 [get_jit_bprop_graph]: 2.39001e-06 [rewriter_after_jit_bprop_graph]: 7.71999e-06 [opt_after_jit_grad]: 0.00062472 [validate]: 5.596e-05 Sums bootstrap : 0.000701s : 0.32% type_inference : 0.206106s : 94.97% event_method : 0.000084s : 0.04% auto_monad : 0.000185s : 0.09% graph_reusing : 0.000013s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000064s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000013s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000095s : 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 : 0.000007s : 0.00% optimize.rewriter_before_opt_a : 0.000414s : 0.19% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000265s : 0.12% optimize.opt_a.loop_unroll : 0.000091s : 0.04% optimize.opt_a.a_1 : 0.002067s : 0.95% optimize.opt_a.with_stream_mark : 0.000052s : 0.02% optimize.opt_a.recompute_prepare : 0.000026s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000205s : 0.09% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.01% optimize.opt_a.merge_send_recv : 0.000022s : 0.01% optimize.opt_a.auto_parallel : 0.000022s : 0.01% optimize.opt_a.parallel : 0.000042s : 0.02% optimize.opt_a.flash_sp : 0.000015s : 0.01% optimize.opt_a.merge_comm : 0.000010s : 0.00% optimize.opt_a.allreduce_fusion : 0.000009s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000025s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000024s : 0.01% optimize.opt_a.virtual_dataset : 0.000015s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000052s : 0.02% optimize.opt_a.virtual_output : 0.000016s : 0.01% optimize.opt_a.merge_forward : 0.000014s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000040s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000029s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.01% optimize.opt_a.meta_fg_expand : 0.000009s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000027s : 0.01% optimize.opt_a.a_after_grad : 0.000024s : 0.01% optimize.opt_a.renormalize : 0.002951s : 1.36% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000039s : 0.02% optimize.opt_a.cse : 0.000082s : 0.04% optimize.opt_a.a_3 : 0.000122s : 0.06% optimize.py_interpret_to_execute_after_opt_a : 0.000011s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000028s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000772s : 0.36% optimize.opt_b.b_1 : 0.000162s : 0.07% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000038s : 0.02% optimize.optimize_parallel_all_gather_comm : 0.000023s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000038s : 0.02% optimize.loop_unroll : 0.000547s : 0.25% optimize.opt_after_cconv.c_1 : 0.000038s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000033s : 0.02% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000055s : 0.03% optimize.tuple_transform.d_1 : 0.000056s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000081s : 0.04% optimize.cse_after_recomputation.cse : 0.000020s : 0.01% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000008s : 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 : 0.000003s : 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 : 0.000018s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000028s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000016s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000008s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 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 : 0.000024s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000625s : 0.29% validate : 0.000056s : 0.03% Time group info: ------[substitution.] 0.000638 75 8.65% : 0.000055s : 11: substitution.cast_eliminate 0.37% : 0.000002s : 4: substitution.elim_not_effective 0.46% : 0.000003s : 4: substitution.fold_const_symbol 1.31% : 0.000008s : 5: substitution.graph_param_transform 81.54% : 0.000521s : 24: substitution.inline 1.03% : 0.000007s : 8: substitution.j_node_and_user_rematch 1.22% : 0.000008s : 8: substitution.remove_not_recompute_node 0.88% : 0.000006s : 2: substitution.replace_old_param 2.90% : 0.000019s : 7: substitution.switch_simplify 1.66% : 0.000011s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.205978 2 97.60% : 0.201040s : 1: type_inference.infer 2.40% : 0.004938s : 1: type_inference.specialize ------[replace.] 0.000277 33 61.64% : 0.000171s : 24: replace.inline 31.46% : 0.000087s : 7: replace.switch_simplify 6.89% : 0.000019s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000530 33 95.66% : 0.000507s : 24: match.inline 2.64% : 0.000014s : 7: match.switch_simplify 1.70% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000411 2544 1.11% : 0.000005s : 32: predicate.accumulaten_eliminater 0.65% : 0.000003s : 5: predicate.ad_related_special_op_eliminate 0.34% : 0.000001s : 10: predicate.addn_check_dump 1.22% : 0.000005s : 32: predicate.addn_zero_filter 1.08% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.13% : 0.000009s : 42: predicate.arithmetic_simplify 1.53% : 0.000006s : 32: predicate.cast_eliminate 0.43% : 0.000002s : 10: predicate.check_bprop_eliminate 0.34% : 0.000001s : 10: predicate.compare_switch_simplify 0.11% : 0.000000s : 5: predicate.const_output_eliminate 0.40% : 0.000002s : 10: predicate.depend_value_elim 1.20% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.16% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.12% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.78% : 0.000003s : 10: predicate.dumpgradient_eliminate 0.17% : 0.000001s : 5: predicate.elim_not_effective 0.24% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.26% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.22% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.21% : 0.000005s : 37: predicate.environ_get_depend_swap 1.66% : 0.000007s : 47: predicate.environ_get_eliminate 1.36% : 0.000006s : 37: predicate.environ_get_set_eliminate 2.09% : 0.000009s : 58: predicate.exchange_switch_depend_value 3.23% : 0.000013s : 58: predicate.float_depend_g_call 0.34% : 0.000001s : 10: predicate.float_environ_get_switch 0.50% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.10% : 0.000000s : 5: predicate.fold_const_symbol 0.56% : 0.000002s : 10: predicate.get_grad_eliminate 0.12% : 0.000000s : 5: predicate.graph_param_transform 0.36% : 0.000001s : 10: predicate.incorporate_call 0.30% : 0.000001s : 10: predicate.incorporate_call_switch 6.20% : 0.000025s : 120: predicate.inline 0.51% : 0.000002s : 10: predicate.inline_without_move 0.17% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.60% : 0.000002s : 10: predicate.less_batch_normalization 1.68% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.45% : 0.000010s : 76: predicate.load_eliminater 0.66% : 0.000003s : 5: predicate.loop_unroll_after_grad 3.50% : 0.000014s : 104: predicate.loop_unroll_before_grad 1.62% : 0.000007s : 42: predicate.make_slice_get_slice_eliminator 0.34% : 0.000001s : 10: predicate.merge_addn 0.37% : 0.000001s : 10: predicate.micro_step_allgather_replace 0.35% : 0.000001s : 10: predicate.mini_step_allgather_replace 1.06% : 0.000004s : 32: predicate.minmaximum_grad 0.92% : 0.000004s : 5: predicate.mutable_eliminate 0.23% : 0.000001s : 5: predicate.opt_reshape 0.37% : 0.000002s : 5: predicate.parallel_virtual_node 2.60% : 0.000011s : 58: predicate.partial_defer_inline 1.45% : 0.000006s : 39: predicate.partial_eliminate 1.08% : 0.000004s : 32: predicate.print_const_string_wrapper 0.34% : 0.000001s : 10: predicate.reduce_all_const_elim 1.47% : 0.000006s : 32: predicate.reduce_eliminate 2.54% : 0.000010s : 76: predicate.redundant_stop_gradient_eliminater 0.29% : 0.000001s : 10: predicate.remove_not_recompute_node 1.31% : 0.000005s : 44: predicate.replace_applicator 0.39% : 0.000002s : 10: predicate.replace_old_param 0.30% : 0.000001s : 5: predicate.reset_defer_inline 1.15% : 0.000005s : 32: predicate.reshape_eliminate 0.36% : 0.000001s : 10: predicate.row_tensor_add_zeros_like 0.21% : 0.000001s : 5: predicate.row_tensor_eliminate 0.57% : 0.000002s : 10: predicate.same_eliminate 0.34% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.52% : 0.000002s : 10: predicate.shard_identity_eliminate 0.55% : 0.000002s : 10: predicate.special_op_eliminate 0.43% : 0.000002s : 10: predicate.specialize_transform 0.59% : 0.000002s : 10: predicate.split_environ_get_set_with_tuple_value 0.51% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.25% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.24% : 0.000009s : 58: predicate.switch_defer_inline 2.55% : 0.000010s : 68: predicate.switch_layer_defer_inline 7.32% : 0.000030s : 191: predicate.switch_simplify 1.10% : 0.000005s : 32: predicate.tile_eliminate 1.12% : 0.000005s : 32: predicate.transpose_eliminate 1.65% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.54% : 0.000006s : 42: predicate.tuple_list_get_item_const_eliminator 1.44% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.55% : 0.000010s : 54: predicate.tuple_list_get_item_eliminator 1.51% : 0.000006s : 42: predicate.tuple_list_get_set_item_eliminator 1.97% : 0.000008s : 52: predicate.tuple_list_set_item_eliminator 1.49% : 0.000006s : 44: predicate.tuple_to_list_eliminator_ 2.41% : 0.000010s : 76: predicate.updatestate_pure_node_eliminater 3.01% : 0.000012s : 86: predicate.updatestate_useless_node_eliminater 0.22% : 0.000001s : 5: predicate.value_based_eliminate 0.45% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.42% : 0.000002s : 10: predicate.virtual_output_eliminate 0.18% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.25% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003367 42 57.18% : 0.001925s : 16: func_graph_cloner_run.FuncGraphClonerGraph 42.82% : 0.001442s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.292595 192 0.00% : 0.000004s : 1: ForceFp32Comm 9.97% : 0.029179s : 1: add_attr 9.97% : 0.029162s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.03% : 0.000088s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.07% : 0.000194s : 1: auto_monad 0.01% : 0.000029s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.26% : 0.000746s : 1: bootstrap 0.01% : 0.000043s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000023s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000037s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.03% : 0.000093s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000017s : 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% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.19% : 0.000562s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.27% : 0.000790s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.01% : 0.000020s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000025s : 1: opt.transform.mutable_eliminate 0.99% : 0.002902s : 78: opt.transform.opt_a 0.01% : 0.000036s : 1: opt.transform.opt_after_cconv 0.01% : 0.000035s : 1: opt.transform.opt_after_jit_grad 0.05% : 0.000143s : 28: opt.transform.opt_b 0.02% : 0.000062s : 2: opt.transform.opt_trans_graph 0.02% : 0.000052s : 4: opt.transform.symbol_engine_opt 2.40% : 0.007032s : 1: opt_a 0.05% : 0.000136s : 1: opt_after_cconv 0.22% : 0.000642s : 1: opt_after_jit_grad 0.10% : 0.000278s : 1: opt_b 3.43% : 0.010041s : 1: optimize 0.01% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000031s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 1: overlap_grad_ring_attention 0.00% : 0.000006s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000101s : 1: pre_auto_parallel 0.00% : 0.000011s : 1: py_interpret_to_execute 0.00% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000060s : 1: remove_dup_value 0.54% : 0.001575s : 1: renormalize.infer 0.47% : 0.001361s : 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% : 0.000033s : 1: rewriter_after_opt_a 0.15% : 0.000425s : 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% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.04% : 0.000103s : 1: symbol_engine_optimizer 0.03% : 0.000096s : 1: tuple_transform 70.45% : 0.206137s : 1: type_inference .......... TotalTime = 0.55814, [21] [bootstrap]: 0.0137187 [type_inference]: 0.413223 [event_method]: 7.612e-05 [auto_monad]: 0.00027467 [graph_reusing]: 2.152e-05 [inline]: 3.06999e-06 [add_attr]: 0.00389365, [1] [add_attr_with_inline]: 0.00388414, [1] [Cycle 1]: 0.00012291, [2] [tag_attr]: 6.892e-05 [meta_addattr_fg_expand]: 2.09e-05 [parallel-infer-symbol]: 3.45003e-06 [pre_auto_parallel]: 0.00010204 [insert-virtual-dataset]: 2.93e-06 [parallel-infer-symbol-second]: 8.2e-07 [dataset_repeat_opt]: 2.02999e-06 [pipeline_split]: 1.54e-06 [optimize]: 0.12598, [53] [py_interpret_to_execute]: 5.27001e-06 [rewriter_before_opt_a]: 0.0006042 [opt_a]: 0.123347, [2] [Cycle 1]: 0.122219, [45] [expand_dump_flag]: 9.46e-06 [switch_simplify]: 0.0003147 [loop_unroll]: 0.00012145 [a_1]: 0.00263753 [with_stream_mark]: 3.299e-05 [recompute_prepare]: 2.449e-05 [updatestate_depend_eliminate]: 9.72001e-06 [updatestate_assign_eliminate]: 8.05e-06 [updatestate_loads_eliminate]: 7.28999e-06 [parameter_eliminate]: 2.83998e-06 [a_2]: 0.00023422 [accelerated_algorithm]: 1.566e-05 [shard]: 1.72001e-06 [meta_shard_fg_expand]: 7.11999e-06 [shard_inline]: 1.529e-05 [merge_send_recv]: 1.785e-05 [auto_parallel]: 1.229e-05 [parallel]: 2.034e-05 [flash_sp]: 1.144e-05 [merge_comm]: 9.27999e-06 [allreduce_fusion]: 8.57e-06 [matmul_add_comm_reduction]: 3.004e-05 [allreduce_slice_to_reducescatter]: 7.79983e-07 [virtual_shard_identity]: 1.69e-05 [virtual_dataset]: 1.522e-05 [get_grad_eliminate_]: 1.525e-05 [virtual_output]: 1.468e-05 [merge_forward]: 9.20001e-06 [cell_reuse_recompute_pass]: 1.16002e-06 [offload_activation]: 1.85e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.854e-05 [merge_recompute_call_nodes]: 1.50001e-06 [before_grad]: 4.701e-05 [set_forward_comm_id_for_comm_node_pass]: 1.086e-05 [meta_fg_expand]: 0.00345638 [flash_sp_send_recv_attached]: 5.40999e-06 [receive_attached]: 2.69001e-06 [after_resolve]: 0.0001168 [a_after_grad]: 0.00015607 [renormalize]: 0.114314 [add_forward_monad_depend]: 8.04002e-06 [auto_monad_grad]: 2.83e-06 [auto_monad_eliminator]: 1.38e-05 [cse]: 2.709e-05 [a_3]: 4.846e-05 [Cycle 2]: 0.00111316, [45] [expand_dump_flag]: 1.81e-06 [switch_simplify]: 6.74001e-06 [loop_unroll]: 4.42e-06 [a_1]: 5.384e-05 [with_stream_mark]: 1.418e-05 [recompute_prepare]: 4.00998e-06 [updatestate_depend_eliminate]: 3.35e-06 [updatestate_assign_eliminate]: 2.78e-06 [updatestate_loads_eliminate]: 2.56998e-06 [parameter_eliminate]: 1.99999e-06 [a_2]: 3.85e-05 [accelerated_algorithm]: 4.34997e-06 [shard]: 2.34999e-06 [meta_shard_fg_expand]: 1.97001e-06 [shard_inline]: 3.55003e-06 [merge_send_recv]: 7.5e-06 [auto_parallel]: 8.47e-06 [parallel]: 9.04e-06 [flash_sp]: 3.78001e-06 [merge_comm]: 2.41e-06 [allreduce_fusion]: 2.73998e-06 [matmul_add_comm_reduction]: 7.61999e-06 [allreduce_slice_to_reducescatter]: 6.09987e-07 [virtual_shard_identity]: 4.83001e-06 [virtual_dataset]: 3.85998e-06 [get_grad_eliminate_]: 3.86001e-06 [virtual_output]: 3.51001e-06 [merge_forward]: 3.28e-06 [cell_reuse_recompute_pass]: 2.72001e-06 [offload_activation]: 7.1e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.205e-05 [merge_recompute_call_nodes]: 1.49e-06 [before_grad]: 6.89001e-06 [set_forward_comm_id_for_comm_node_pass]: 2.44999e-06 [meta_fg_expand]: 0.00059396 [flash_sp_send_recv_attached]: 2.26e-06 [receive_attached]: 2.75002e-06 [after_resolve]: 1.179e-05 [a_after_grad]: 5.57001e-06 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 2.52001e-06 [auto_monad_grad]: 1.42e-06 [auto_monad_eliminator]: 7.31001e-06 [cse]: 1.378e-05 [a_3]: 2.328e-05 [py_interpret_to_execute_after_opt_a]: 5.91e-06 [slice_cell_reuse_recomputed_activation]: 2.43e-06 [rewriter_after_opt_a]: 1.723e-05 [convert_after_rewriter]: 1.17999e-06 [order_py_execute_after_rewriter]: 1.14e-06 [mutable_eliminate]: 0.00075292 [opt_b]: 0.00014911, [1] [Cycle 1]: 0.00014102, [7] [b_1]: 7.583e-05 [b_2]: 5.34e-06 [updatestate_depend_eliminate]: 4.53001e-06 [updatestate_assign_eliminate]: 2.01998e-06 [updatestate_loads_eliminate]: 1.69e-06 [renormalize]: 7.49977e-07 [cse]: 1.645e-05 [optimize_parallel_all_gather_comm]: 1.615e-05 [overlap_param_gather]: 1.64998e-06 [cconv]: 2.823e-05 [loop_unroll]: 0.00047233 [opt_after_cconv]: 8.165e-05, [1] [Cycle 1]: 7.52e-05, [7] [c_1]: 1.621e-05 [parameter_eliminate]: 3.2e-06 [updatestate_depend_eliminate]: 4.28999e-06 [updatestate_assign_eliminate]: 1.68002e-06 [updatestate_loads_eliminate]: 1.49998e-06 [cse]: 1.552e-05 [renormalize]: 5.50004e-07 [remove_dup_value]: 1.444e-05 [tuple_transform]: 5.352e-05, [1] [Cycle 1]: 4.872e-05, [4] [d_1]: 2.502e-05 [none_parameter_eliminate]: 1.57001e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 4.49998e-06 [partial_unused_args_eliminate]: 1.71e-06 [add_recomputation]: 3.252e-05 [cse_after_recomputation]: 1.799e-05, [1] [Cycle 1]: 1.332e-05, [1] [cse]: 8.29002e-06 [environ_conv]: 4.35999e-06 [swap_dp_allreduce_reducescatter]: 3.68e-06 [bias_add_comm_swap]: 2.57001e-06 [label_micro_interleaved_index]: 4.55001e-06 [label_fine_grained_interleaved_index]: 2.41e-06 [merge_cast_opt]: 1.30001e-06 [slice_recompute_activation]: 2.21003e-06 [micro_interleaved_order_control]: 2.61e-06 [assign_add_opt]: 1.27999e-06 [ForceFp32Comm]: 8.00006e-07 [remove_cast_before_assign_add]: 1.27999e-06 [full_micro_interleaved_order_control]: 2.09e-06 [reorder_send_recv_between_fp_bp]: 2.93e-06 [comm_op_add_attrs]: 1.58002e-06 [add_comm_op_reuse_tag]: 9.70002e-07 [interleave_split_concat_branches]: 1.36002e-06 [interleave_parallel_branches]: 9.99979e-07 [overlap_opt_shard_in_pipeline]: 1.15001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.89999e-06 [control_data_broadcast_order]: 1.016e-05 [grouped_pairwise_exchange_alltoall]: 1.43002e-06 [offloading_packed_experts]: 3.09999e-06 [overlap_recompute_and_grad_model_parallel]: 3.95998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.36998e-06 [overlap_recompute_comm]: 2.61e-06 [overlap_grad_ring_attention]: 3.26001e-06 [overlap_grad_flash_sp]: 1.465e-05 [begin_end_overlap_inline]: 6.00005e-07 [split_matmul_comm_elemetwise]: 2.12999e-06 [split_layernorm_comm]: 1.62001e-06 [handle_group_info]: 1.00001e-06 [symbol_engine_optimizer]: 6.014e-05, [1] [Cycle 1]: 5.602e-05, [6] [build]: 2.36e-06 [elim_shapecalc]: 7.00998e-06 [elim_not_effective]: 9.27001e-06 [opt_reshape]: 4.25e-06 [fold_const_symbol]: 6.81999e-06 [renormalize]: 1.39989e-07 [detach_backward]: 1.97001e-06 [pipeline_parallel_scheduler]: 0.00011036 [auto_monad_reorder]: 1.462e-05 [get_jit_bprop_graph]: 1.87999e-06 [rewriter_after_jit_bprop_graph]: 4.84e-06 [opt_after_jit_grad]: 0.00047618 [validate]: 3.409e-05 Sums bootstrap : 0.013719s : 2.48% type_inference : 0.413223s : 74.69% event_method : 0.000076s : 0.01% auto_monad : 0.000275s : 0.05% graph_reusing : 0.000022s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000069s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000021s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000102s : 0.02% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000604s : 0.11% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000321s : 0.06% optimize.opt_a.loop_unroll : 0.000126s : 0.02% optimize.opt_a.a_1 : 0.002691s : 0.49% optimize.opt_a.with_stream_mark : 0.000047s : 0.01% optimize.opt_a.recompute_prepare : 0.000028s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000273s : 0.05% optimize.opt_a.accelerated_algorithm : 0.000020s : 0.00% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000009s : 0.00% optimize.opt_a.shard_inline : 0.000019s : 0.00% optimize.opt_a.merge_send_recv : 0.000025s : 0.00% optimize.opt_a.auto_parallel : 0.000021s : 0.00% optimize.opt_a.parallel : 0.000029s : 0.01% optimize.opt_a.flash_sp : 0.000015s : 0.00% optimize.opt_a.merge_comm : 0.000012s : 0.00% optimize.opt_a.allreduce_fusion : 0.000011s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000038s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000022s : 0.00% optimize.opt_a.virtual_dataset : 0.000019s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000019s : 0.00% optimize.opt_a.virtual_output : 0.000018s : 0.00% optimize.opt_a.merge_forward : 0.000012s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000041s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000054s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000013s : 0.00% optimize.opt_a.meta_fg_expand : 0.004050s : 0.73% 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 : 0.000129s : 0.02% optimize.opt_a.a_after_grad : 0.000162s : 0.03% optimize.opt_a.renormalize : 0.114314s : 20.66% optimize.opt_a.add_forward_monad_depend : 0.000011s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000021s : 0.00% optimize.opt_a.cse : 0.000041s : 0.01% optimize.opt_a.a_3 : 0.000072s : 0.01% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000017s : 0.00% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000753s : 0.14% optimize.opt_b.b_1 : 0.000076s : 0.01% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000016s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000016s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000028s : 0.01% optimize.loop_unroll : 0.000472s : 0.09% optimize.opt_after_cconv.c_1 : 0.000016s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000014s : 0.00% optimize.tuple_transform.d_1 : 0.000025s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000033s : 0.01% optimize.cse_after_recomputation.cse : 0.000008s : 0.00% optimize.environ_conv : 0.000004s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 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 : 0.000002s : 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 : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000010s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000015s : 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 : 0.000002s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000007s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000009s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000004s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000110s : 0.02% auto_monad_reorder : 0.000015s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000476s : 0.09% validate : 0.000034s : 0.01% Time group info: ------[substitution.] 0.000875 134 3.80% : 0.000033s : 8: substitution.cast_eliminate 0.26% : 0.000002s : 1: substitution.elim_not_effective 0.70% : 0.000006s : 9: substitution.float_depend_g_call 0.46% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.14% : 0.000001s : 1: substitution.fold_const_symbol 0.61% : 0.000005s : 1: substitution.graph_param_transform 0.31% : 0.000003s : 2: substitution.incorporate_call 0.29% : 0.000003s : 2: substitution.incorporate_call_switch 77.27% : 0.000676s : 33: substitution.inline 3.95% : 0.000035s : 4: substitution.inline_without_move 0.93% : 0.000008s : 11: substitution.j_node_and_user_rematch 0.40% : 0.000004s : 2: substitution.minmaximum_grad 0.74% : 0.000006s : 9: substitution.partial_eliminate 1.04% : 0.000009s : 11: substitution.remove_not_recompute_node 0.25% : 0.000002s : 1: substitution.replace_applicator 0.93% : 0.000008s : 12: substitution.replace_old_param 0.38% : 0.000003s : 1: substitution.set_cell_output_no_recompute 2.77% : 0.000024s : 10: substitution.switch_simplify 0.86% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.35% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.52% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.55% : 0.000022s : 6: substitution.tuple_list_get_item_eliminator 0.49% : 0.000004s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.413007 2 82.29% : 0.339877s : 1: type_inference.infer 17.71% : 0.073130s : 1: type_inference.specialize ------[replace.] 0.000333 45 62.88% : 0.000209s : 33: replace.inline 32.50% : 0.000108s : 10: replace.switch_simplify 4.63% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000630 45 95.21% : 0.000599s : 33: match.inline 2.98% : 0.000019s : 10: match.switch_simplify 1.81% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000501 3189 1.20% : 0.000006s : 43: predicate.accumulaten_eliminater 0.23% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.41% : 0.000002s : 14: predicate.addn_check_dump 1.22% : 0.000006s : 43: predicate.addn_zero_filter 1.16% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 2.02% : 0.000010s : 57: predicate.arithmetic_simplify 1.51% : 0.000008s : 43: predicate.cast_eliminate 0.13% : 0.000001s : 2: predicate.check_bprop_eliminate 0.38% : 0.000002s : 14: predicate.compare_switch_simplify 0.03% : 0.000000s : 1: predicate.const_output_eliminate 0.41% : 0.000002s : 14: predicate.depend_value_elim 1.23% : 0.000006s : 43: predicate.dict_get_item_const_eliminator 1.30% : 0.000007s : 43: predicate.dict_get_item_eliminator 1.25% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.18% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 1: predicate.elim_not_effective 0.09% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.23% : 0.000006s : 44: predicate.environ_add_const_eliminate 1.30% : 0.000007s : 44: predicate.environ_get_add_eliminate 1.14% : 0.000006s : 44: predicate.environ_get_depend_swap 1.65% : 0.000008s : 58: predicate.environ_get_eliminate 1.19% : 0.000006s : 44: predicate.environ_get_set_eliminate 2.17% : 0.000011s : 78: predicate.exchange_switch_depend_value 3.13% : 0.000016s : 78: predicate.float_depend_g_call 0.40% : 0.000002s : 14: predicate.float_environ_get_switch 0.50% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.49% : 0.000002s : 14: predicate.get_grad_eliminate 0.04% : 0.000000s : 1: predicate.graph_param_transform 0.39% : 0.000002s : 14: predicate.incorporate_call 0.35% : 0.000002s : 14: predicate.incorporate_call_switch 5.88% : 0.000029s : 151: predicate.inline 2.75% : 0.000014s : 70: predicate.inline_without_move 0.21% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.57% : 0.000003s : 14: predicate.less_batch_normalization 1.43% : 0.000007s : 47: predicate.list_to_tuple_eliminator_ 2.40% : 0.000012s : 90: predicate.load_eliminater 0.32% : 0.000002s : 1: predicate.loop_unroll_after_grad 4.21% : 0.000021s : 141: predicate.loop_unroll_before_grad 1.28% : 0.000006s : 45: predicate.make_slice_get_slice_eliminator 0.42% : 0.000002s : 14: predicate.merge_addn 0.14% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.09% : 0.000000s : 2: predicate.mini_step_allgather_replace 1.21% : 0.000006s : 43: predicate.minmaximum_grad 0.32% : 0.000002s : 1: predicate.mutable_eliminate 0.08% : 0.000000s : 1: predicate.opt_reshape 0.14% : 0.000001s : 1: predicate.parallel_virtual_node 2.96% : 0.000015s : 78: predicate.partial_defer_inline 1.47% : 0.000007s : 46: predicate.partial_eliminate 1.13% : 0.000006s : 43: predicate.print_const_string_wrapper 0.42% : 0.000002s : 14: predicate.reduce_all_const_elim 1.48% : 0.000007s : 43: predicate.reduce_eliminate 2.37% : 0.000012s : 90: predicate.redundant_stop_gradient_eliminater 0.31% : 0.000002s : 14: predicate.remove_not_recompute_node 1.05% : 0.000005s : 47: predicate.replace_applicator 1.34% : 0.000007s : 70: predicate.replace_old_param 0.04% : 0.000000s : 1: predicate.reset_defer_inline 1.18% : 0.000006s : 43: predicate.reshape_eliminate 0.18% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.07% : 0.000000s : 1: predicate.row_tensor_eliminate 0.24% : 0.000001s : 2: predicate.same_eliminate 0.24% : 0.000001s : 14: predicate.set_cell_output_no_recompute 0.51% : 0.000003s : 14: predicate.shard_identity_eliminate 0.18% : 0.000001s : 2: predicate.special_op_eliminate 0.46% : 0.000002s : 14: predicate.specialize_transform 0.38% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 2.37% : 0.000012s : 70: predicate.stack_unstack_eliminate 0.04% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.46% : 0.000012s : 78: predicate.switch_defer_inline 2.52% : 0.000013s : 80: predicate.switch_layer_defer_inline 8.22% : 0.000041s : 254: predicate.switch_simplify 1.21% : 0.000006s : 43: predicate.tile_eliminate 1.25% : 0.000006s : 43: predicate.transpose_eliminate 1.43% : 0.000007s : 45: predicate.tuple_list_convert_item_index_to_positive 1.42% : 0.000007s : 45: predicate.tuple_list_get_item_const_eliminator 1.24% : 0.000006s : 45: predicate.tuple_list_get_item_depend_reorder 2.34% : 0.000012s : 61: predicate.tuple_list_get_item_eliminator 1.40% : 0.000007s : 45: predicate.tuple_list_get_set_item_eliminator 2.16% : 0.000011s : 59: predicate.tuple_list_set_item_eliminator 1.29% : 0.000006s : 47: predicate.tuple_to_list_eliminator_ 2.31% : 0.000012s : 90: predicate.updatestate_pure_node_eliminater 2.87% : 0.000014s : 104: predicate.updatestate_useless_node_eliminater 0.08% : 0.000000s : 1: predicate.value_based_eliminate 0.55% : 0.000003s : 14: predicate.virtual_dataset_eliminate 0.45% : 0.000002s : 14: predicate.virtual_output_eliminate 0.03% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.13% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.120064 69 44.11% : 0.052962s : 33: func_graph_cloner_run.FuncGraphClonerGraph 55.89% : 0.067101s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.806253 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.48% : 0.003899s : 1: add_attr 0.48% : 0.003888s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000036s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000284s : 1: auto_monad 0.00% : 0.000019s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 1.71% : 0.013755s : 1: bootstrap 0.00% : 0.000032s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000013s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.00% : 0.000021s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000007s : 1: environ_conv 0.01% : 0.000085s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000026s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000005s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.06% : 0.000482s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000763s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.00% : 0.000011s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000013s : 1: opt.transform.mutable_eliminate 0.48% : 0.003907s : 78: opt.transform.opt_a 0.00% : 0.000015s : 1: opt.transform.opt_after_cconv 0.00% : 0.000016s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000051s : 28: opt.transform.opt_b 0.00% : 0.000027s : 2: opt.transform.opt_trans_graph 0.00% : 0.000024s : 4: opt.transform.symbol_engine_opt 15.30% : 0.123351s : 1: opt_a 0.01% : 0.000085s : 1: opt_after_cconv 0.06% : 0.000486s : 1: opt_after_jit_grad 0.02% : 0.000153s : 1: opt_b 15.63% : 0.125986s : 1: optimize 0.00% : 0.000019s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.00% : 0.000018s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000004s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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 0.01% : 0.000115s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.01% : 0.000108s : 1: pre_auto_parallel 0.00% : 0.000009s : 1: py_interpret_to_execute 0.00% : 0.000009s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000018s : 1: remove_dup_value 14.08% : 0.113509s : 1: renormalize.infer 0.10% : 0.000792s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000020s : 1: rewriter_after_opt_a 0.08% : 0.000614s : 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% : 0.000006s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000063s : 1: symbol_engine_optimizer 0.01% : 0.000056s : 1: tuple_transform 51.26% : 0.413254s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:44:15.147.905 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.494098, [21] [bootstrap]: 0.00045316 [type_inference]: 0.47921 [event_method]: 9.44e-05 [auto_monad]: 0.00020125 [graph_reusing]: 1.301e-05 [inline]: 3.25998e-06 [add_attr]: 0.00403454, [1] [add_attr_with_inline]: 0.00402506, [1] [Cycle 1]: 0.00011016, [2] [tag_attr]: 6.125e-05 [meta_addattr_fg_expand]: 1.591e-05 [parallel-infer-symbol]: 3.15998e-06 [pre_auto_parallel]: 8.343e-05 [insert-virtual-dataset]: 2.29999e-06 [parallel-infer-symbol-second]: 7.59988e-07 [dataset_repeat_opt]: 1.96998e-06 [pipeline_split]: 1.51002e-06 [optimize]: 0.00923732, [53] [py_interpret_to_execute]: 5.12999e-06 [rewriter_before_opt_a]: 0.00043986 [opt_a]: 0.00647877, [2] [Cycle 1]: 0.00568235, [45] [expand_dump_flag]: 8.24002e-06 [switch_simplify]: 0.00024174 [loop_unroll]: 8.456e-05 [a_1]: 0.00183818 [with_stream_mark]: 2.029e-05 [recompute_prepare]: 1.096e-05 [updatestate_depend_eliminate]: 6.33e-06 [updatestate_assign_eliminate]: 4.81002e-06 [updatestate_loads_eliminate]: 4.32e-06 [parameter_eliminate]: 2.33002e-06 [a_2]: 0.00011002 [accelerated_algorithm]: 9.62999e-06 [shard]: 1.73997e-06 [meta_shard_fg_expand]: 4.58999e-06 [shard_inline]: 8.52998e-06 [merge_send_recv]: 1.027e-05 [auto_parallel]: 7.58001e-06 [parallel]: 1.936e-05 [flash_sp]: 9.88002e-06 [merge_comm]: 5.42001e-06 [allreduce_fusion]: 4.63999e-06 [matmul_add_comm_reduction]: 1.105e-05 [allreduce_slice_to_reducescatter]: 6.89994e-07 [virtual_shard_identity]: 9.84999e-06 [virtual_dataset]: 8.37e-06 [get_grad_eliminate_]: 8.39002e-06 [virtual_output]: 8.04997e-06 [merge_forward]: 5.41002e-06 [cell_reuse_recompute_pass]: 1.13001e-06 [offload_activation]: 1.116e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.787e-05 [merge_recompute_call_nodes]: 1.57999e-06 [before_grad]: 1.349e-05 [set_forward_comm_id_for_comm_node_pass]: 4.75999e-06 [meta_fg_expand]: 5.66e-06 [flash_sp_send_recv_attached]: 2.23998e-06 [receive_attached]: 2.11e-06 [after_resolve]: 1.198e-05 [a_after_grad]: 1.266e-05 [renormalize]: 0.00272746 [add_forward_monad_depend]: 6.23998e-06 [auto_monad_grad]: 2.69001e-06 [auto_monad_eliminator]: 1.869e-05 [cse]: 4.703e-05 [a_3]: 6.221e-05 [Cycle 2]: 0.00078377, [45] [expand_dump_flag]: 1.45999e-06 [switch_simplify]: 9.25999e-06 [loop_unroll]: 7.51999e-06 [a_1]: 0.00019091 [with_stream_mark]: 1.392e-05 [recompute_prepare]: 7.77e-06 [updatestate_depend_eliminate]: 4.36002e-06 [updatestate_assign_eliminate]: 3.95e-06 [updatestate_loads_eliminate]: 3.58e-06 [parameter_eliminate]: 1.11002e-06 [a_2]: 9.711e-05 [accelerated_algorithm]: 7.92e-06 [shard]: 1.30001e-06 [meta_shard_fg_expand]: 1.97001e-06 [shard_inline]: 7.83999e-06 [merge_send_recv]: 6.66999e-06 [auto_parallel]: 6.78e-06 [parallel]: 5.72001e-06 [flash_sp]: 3.25e-06 [merge_comm]: 4.45e-06 [allreduce_fusion]: 4.42e-06 [matmul_add_comm_reduction]: 7.8e-06 [allreduce_slice_to_reducescatter]: 4.19997e-07 [virtual_shard_identity]: 8.48999e-06 [virtual_dataset]: 7.46999e-06 [get_grad_eliminate_]: 7.41999e-06 [virtual_output]: 6.90998e-06 [merge_forward]: 4.43001e-06 [cell_reuse_recompute_pass]: 1.82001e-06 [offload_activation]: 8.80001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.56e-05 [merge_recompute_call_nodes]: 8.79983e-07 [before_grad]: 1.285e-05 [set_forward_comm_id_for_comm_node_pass]: 4.82998e-06 [meta_fg_expand]: 3.35003e-06 [flash_sp_send_recv_attached]: 1.29998e-06 [receive_attached]: 1.43002e-06 [after_resolve]: 1.098e-05 [a_after_grad]: 1.077e-05 [renormalize]: 5.9983e-08 [add_forward_monad_depend]: 1.37e-06 [auto_monad_grad]: 1.03001e-06 [auto_monad_eliminator]: 9.39998e-06 [cse]: 2.283e-05 [a_3]: 4.685e-05 [py_interpret_to_execute_after_opt_a]: 5.54998e-06 [slice_cell_reuse_recomputed_activation]: 2.11e-06 [rewriter_after_opt_a]: 2.625e-05 [convert_after_rewriter]: 1.35999e-06 [order_py_execute_after_rewriter]: 1.14e-06 [mutable_eliminate]: 0.00071023 [opt_b]: 0.00026018, [1] [Cycle 1]: 0.000253, [7] [b_1]: 0.00016482 [b_2]: 9.09e-06 [updatestate_depend_eliminate]: 6.98e-06 [updatestate_assign_eliminate]: 3.6e-06 [updatestate_loads_eliminate]: 3.53999e-06 [renormalize]: 6.89994e-07 [cse]: 2.987e-05 [optimize_parallel_all_gather_comm]: 1.9e-05 [overlap_param_gather]: 2.11998e-06 [cconv]: 2.444e-05 [loop_unroll]: 0.0004926 [opt_after_cconv]: 0.0001234, [1] [Cycle 1]: 0.00011712, [7] [c_1]: 3.704e-05 [parameter_eliminate]: 2.89001e-06 [updatestate_depend_eliminate]: 7.01001e-06 [updatestate_assign_eliminate]: 3.85e-06 [updatestate_loads_eliminate]: 4.2e-06 [cse]: 2.85e-05 [renormalize]: 7.00005e-07 [remove_dup_value]: 4.145e-05 [tuple_transform]: 8.677e-05, [1] [Cycle 1]: 8.203e-05, [4] [d_1]: 5.32e-05 [none_parameter_eliminate]: 1.67999e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 8.33999e-06 [partial_unused_args_eliminate]: 1.71002e-06 [add_recomputation]: 6.056e-05 [cse_after_recomputation]: 2.792e-05, [1] [Cycle 1]: 2.345e-05, [1] [cse]: 1.823e-05 [environ_conv]: 6.80002e-06 [swap_dp_allreduce_reducescatter]: 6.49999e-06 [bias_add_comm_swap]: 2.64001e-06 [label_micro_interleaved_index]: 4.92e-06 [label_fine_grained_interleaved_index]: 2.81e-06 [merge_cast_opt]: 1.45999e-06 [slice_recompute_activation]: 2.09e-06 [micro_interleaved_order_control]: 2.19001e-06 [assign_add_opt]: 1.27e-06 [ForceFp32Comm]: 8.90024e-07 [remove_cast_before_assign_add]: 1.00999e-06 [full_micro_interleaved_order_control]: 2.54999e-06 [reorder_send_recv_between_fp_bp]: 2.73e-06 [comm_op_add_attrs]: 1.06002e-06 [add_comm_op_reuse_tag]: 1.37e-06 [interleave_split_concat_branches]: 1.18001e-06 [interleave_parallel_branches]: 1.02e-06 [overlap_opt_shard_in_pipeline]: 1.17e-06 [overlap_opt_shard_grad_in_pipeline]: 1.63002e-06 [control_data_broadcast_order]: 1.622e-05 [grouped_pairwise_exchange_alltoall]: 1.57999e-06 [offloading_packed_experts]: 4.52e-06 [overlap_recompute_and_grad_model_parallel]: 6.19001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.44e-06 [overlap_recompute_allgather_and_fa_grad]: 1.35999e-06 [overlap_recompute_comm]: 2.17999e-06 [overlap_grad_ring_attention]: 5.67999e-06 [overlap_grad_flash_sp]: 2.302e-05 [begin_end_overlap_inline]: 4.90021e-07 [split_matmul_comm_elemetwise]: 2.44999e-06 [split_layernorm_comm]: 1.79e-06 [handle_group_info]: 1.19e-06 [symbol_engine_optimizer]: 8.481e-05, [1] [Cycle 1]: 8.048e-05, [6] [build]: 3.9e-06 [elim_shapecalc]: 1.134e-05 [elim_not_effective]: 1.577e-05 [opt_reshape]: 9.15999e-06 [fold_const_symbol]: 1.311e-05 [renormalize]: 2.09984e-07 [detach_backward]: 2.09e-06 [pipeline_parallel_scheduler]: 1.48002e-06 [auto_monad_reorder]: 2.021e-05 [get_jit_bprop_graph]: 2.68e-06 [rewriter_after_jit_bprop_graph]: 3.7e-06 [opt_after_jit_grad]: 0.00049574 [validate]: 4.802e-05 Sums bootstrap : 0.000453s : 0.09% type_inference : 0.479210s : 97.98% event_method : 0.000094s : 0.02% auto_monad : 0.000201s : 0.04% graph_reusing : 0.000013s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000061s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000016s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000083s : 0.02% 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 : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000440s : 0.09% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000251s : 0.05% optimize.opt_a.loop_unroll : 0.000092s : 0.02% optimize.opt_a.a_1 : 0.002029s : 0.41% optimize.opt_a.with_stream_mark : 0.000034s : 0.01% optimize.opt_a.recompute_prepare : 0.000019s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000008s : 0.00% optimize.opt_a.parameter_eliminate : 0.000003s : 0.00% optimize.opt_a.a_2 : 0.000207s : 0.04% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.00% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.00% optimize.opt_a.merge_send_recv : 0.000017s : 0.00% optimize.opt_a.auto_parallel : 0.000014s : 0.00% optimize.opt_a.parallel : 0.000025s : 0.01% optimize.opt_a.flash_sp : 0.000013s : 0.00% optimize.opt_a.merge_comm : 0.000010s : 0.00% optimize.opt_a.allreduce_fusion : 0.000009s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000019s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000018s : 0.00% optimize.opt_a.virtual_dataset : 0.000016s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.00% optimize.opt_a.virtual_output : 0.000015s : 0.00% optimize.opt_a.merge_forward : 0.000010s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000020s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000033s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000026s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000010s : 0.00% optimize.opt_a.meta_fg_expand : 0.000009s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000023s : 0.00% optimize.opt_a.a_after_grad : 0.000023s : 0.00% optimize.opt_a.renormalize : 0.002728s : 0.56% optimize.opt_a.add_forward_monad_depend : 0.000008s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000028s : 0.01% optimize.opt_a.cse : 0.000070s : 0.01% optimize.opt_a.a_3 : 0.000109s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000026s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000710s : 0.15% optimize.opt_b.b_1 : 0.000165s : 0.03% optimize.opt_b.b_2 : 0.000009s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000030s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000019s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000024s : 0.00% optimize.loop_unroll : 0.000493s : 0.10% optimize.opt_after_cconv.c_1 : 0.000037s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000028s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000041s : 0.01% optimize.tuple_transform.d_1 : 0.000053s : 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 : 0.000008s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000061s : 0.01% optimize.cse_after_recomputation.cse : 0.000018s : 0.00% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000016s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 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 : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000023s : 0.00% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000016s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000013s : 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 : 0.000020s : 0.00% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000496s : 0.10% validate : 0.000048s : 0.01% Time group info: ------[substitution.] 0.000592 75 8.23% : 0.000049s : 11: substitution.cast_eliminate 0.46% : 0.000003s : 4: substitution.elim_not_effective 0.43% : 0.000003s : 4: substitution.fold_const_symbol 1.12% : 0.000007s : 5: substitution.graph_param_transform 81.76% : 0.000484s : 24: substitution.inline 0.78% : 0.000005s : 8: substitution.j_node_and_user_rematch 1.30% : 0.000008s : 8: substitution.remove_not_recompute_node 0.60% : 0.000004s : 2: substitution.replace_old_param 3.58% : 0.000021s : 7: substitution.switch_simplify 1.74% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.479073 2 98.94% : 0.473997s : 1: type_inference.infer 1.06% : 0.005077s : 1: type_inference.specialize ------[replace.] 0.000244 33 60.17% : 0.000147s : 24: replace.inline 33.88% : 0.000083s : 7: replace.switch_simplify 5.96% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000497 33 95.06% : 0.000472s : 24: match.inline 3.12% : 0.000015s : 7: match.switch_simplify 1.82% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000409 2544 1.10% : 0.000005s : 32: predicate.accumulaten_eliminater 0.48% : 0.000002s : 5: predicate.ad_related_special_op_eliminate 0.35% : 0.000001s : 10: predicate.addn_check_dump 1.27% : 0.000005s : 32: predicate.addn_zero_filter 1.07% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.20% : 0.000009s : 42: predicate.arithmetic_simplify 1.65% : 0.000007s : 32: predicate.cast_eliminate 0.35% : 0.000001s : 10: predicate.check_bprop_eliminate 0.35% : 0.000001s : 10: predicate.compare_switch_simplify 0.12% : 0.000000s : 5: predicate.const_output_eliminate 0.39% : 0.000002s : 10: predicate.depend_value_elim 1.21% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.24% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.12% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.52% : 0.000002s : 10: predicate.dumpgradient_eliminate 0.12% : 0.000000s : 5: predicate.elim_not_effective 0.23% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.37% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.25% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.20% : 0.000005s : 37: predicate.environ_get_depend_swap 1.73% : 0.000007s : 47: predicate.environ_get_eliminate 1.27% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.05% : 0.000008s : 58: predicate.exchange_switch_depend_value 2.91% : 0.000012s : 58: predicate.float_depend_g_call 0.33% : 0.000001s : 10: predicate.float_environ_get_switch 0.49% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.11% : 0.000000s : 5: predicate.fold_const_symbol 0.42% : 0.000002s : 10: predicate.get_grad_eliminate 0.13% : 0.000001s : 5: predicate.graph_param_transform 0.37% : 0.000001s : 10: predicate.incorporate_call 0.32% : 0.000001s : 10: predicate.incorporate_call_switch 5.87% : 0.000024s : 120: predicate.inline 0.43% : 0.000002s : 10: predicate.inline_without_move 0.19% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.52% : 0.000002s : 10: predicate.less_batch_normalization 1.75% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.52% : 0.000010s : 76: predicate.load_eliminater 0.53% : 0.000002s : 5: predicate.loop_unroll_after_grad 3.73% : 0.000015s : 104: predicate.loop_unroll_before_grad 1.56% : 0.000006s : 42: predicate.make_slice_get_slice_eliminator 0.39% : 0.000002s : 10: predicate.merge_addn 0.38% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.34% : 0.000001s : 10: predicate.mini_step_allgather_replace 1.06% : 0.000004s : 32: predicate.minmaximum_grad 0.66% : 0.000003s : 5: predicate.mutable_eliminate 0.23% : 0.000001s : 5: predicate.opt_reshape 0.22% : 0.000001s : 5: predicate.parallel_virtual_node 2.76% : 0.000011s : 58: predicate.partial_defer_inline 1.47% : 0.000006s : 39: predicate.partial_eliminate 1.12% : 0.000005s : 32: predicate.print_const_string_wrapper 0.35% : 0.000001s : 10: predicate.reduce_all_const_elim 1.55% : 0.000006s : 32: predicate.reduce_eliminate 2.54% : 0.000010s : 76: predicate.redundant_stop_gradient_eliminater 0.27% : 0.000001s : 10: predicate.remove_not_recompute_node 1.13% : 0.000005s : 44: predicate.replace_applicator 0.31% : 0.000001s : 10: predicate.replace_old_param 0.14% : 0.000001s : 5: predicate.reset_defer_inline 1.23% : 0.000005s : 32: predicate.reshape_eliminate 0.39% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.22% : 0.000001s : 5: predicate.row_tensor_eliminate 0.46% : 0.000002s : 10: predicate.same_eliminate 0.24% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.51% : 0.000002s : 10: predicate.shard_identity_eliminate 0.46% : 0.000002s : 10: predicate.special_op_eliminate 0.43% : 0.000002s : 10: predicate.specialize_transform 0.53% : 0.000002s : 10: predicate.split_environ_get_set_with_tuple_value 0.52% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.24% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.28% : 0.000009s : 58: predicate.switch_defer_inline 2.62% : 0.000011s : 68: predicate.switch_layer_defer_inline 7.82% : 0.000032s : 191: predicate.switch_simplify 1.17% : 0.000005s : 32: predicate.tile_eliminate 1.14% : 0.000005s : 32: predicate.transpose_eliminate 1.66% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.64% : 0.000007s : 42: predicate.tuple_list_get_item_const_eliminator 1.59% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.61% : 0.000011s : 54: predicate.tuple_list_get_item_eliminator 1.58% : 0.000006s : 42: predicate.tuple_list_get_set_item_eliminator 2.19% : 0.000009s : 52: predicate.tuple_list_set_item_eliminator 1.58% : 0.000006s : 44: predicate.tuple_to_list_eliminator_ 2.52% : 0.000010s : 76: predicate.updatestate_pure_node_eliminater 3.04% : 0.000012s : 86: predicate.updatestate_useless_node_eliminater 0.26% : 0.000001s : 5: predicate.value_based_eliminate 0.41% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.42% : 0.000002s : 10: predicate.virtual_output_eliminate 0.17% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.33% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002945 42 52.53% : 0.001547s : 16: func_graph_cloner_run.FuncGraphClonerGraph 47.47% : 0.001398s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.513126 192 0.00% : 0.000003s : 1: ForceFp32Comm 0.79% : 0.004040s : 1: add_attr 0.79% : 0.004029s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000065s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000211s : 1: auto_monad 0.00% : 0.000024s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.09% : 0.000482s : 1: bootstrap 0.01% : 0.000028s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000019s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000031s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.02% : 0.000104s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000017s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.10% : 0.000503s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.14% : 0.000721s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 0.55% : 0.002824s : 78: opt.transform.opt_a 0.01% : 0.000036s : 1: opt.transform.opt_after_cconv 0.01% : 0.000030s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000144s : 28: opt.transform.opt_b 0.01% : 0.000059s : 2: opt.transform.opt_trans_graph 0.01% : 0.000047s : 4: opt.transform.symbol_engine_opt 1.26% : 0.006483s : 1: opt_a 0.02% : 0.000127s : 1: opt_after_cconv 0.10% : 0.000507s : 1: opt_after_jit_grad 0.05% : 0.000264s : 1: opt_b 1.80% : 0.009243s : 1: optimize 0.00% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000026s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000009s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.00% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.02% : 0.000089s : 1: pre_auto_parallel 0.00% : 0.000008s : 1: py_interpret_to_execute 0.00% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000046s : 1: remove_dup_value 0.29% : 0.001463s : 1: renormalize.infer 0.24% : 0.001254s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000029s : 1: rewriter_after_opt_a 0.09% : 0.000450s : 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% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000088s : 1: symbol_engine_optimizer 0.02% : 0.000090s : 1: tuple_transform 93.40% : 0.479241s : 1: type_inference .... TotalTime = 0.484521, [21] [bootstrap]: 0.00044705 [type_inference]: 0.408217 [event_method]: 9.812e-05 [auto_monad]: 0.0003361 [graph_reusing]: 2.299e-05 [inline]: 3.26999e-06 [add_attr]: 0.00422645, [1] [add_attr_with_inline]: 0.00421302, [1] [Cycle 1]: 0.00015484, [2] [tag_attr]: 8.526e-05 [meta_addattr_fg_expand]: 2.068e-05 [parallel-infer-symbol]: 3.45e-06 [pre_auto_parallel]: 0.00011718 [insert-virtual-dataset]: 2.58e-06 [parallel-infer-symbol-second]: 1.02998e-06 [dataset_repeat_opt]: 1.99999e-06 [pipeline_split]: 1.53002e-06 [optimize]: 0.0700021, [53] [py_interpret_to_execute]: 9.29998e-06 [rewriter_before_opt_a]: 0.00061662 [opt_a]: 0.0666441, [2] [Cycle 1]: 0.040933, [45] [expand_dump_flag]: 1.584e-05 [switch_simplify]: 0.00033602 [loop_unroll]: 0.00012035 [a_1]: 0.00305265 [with_stream_mark]: 4.081e-05 [recompute_prepare]: 3.323e-05 [updatestate_depend_eliminate]: 1.097e-05 [updatestate_assign_eliminate]: 7.83999e-06 [updatestate_loads_eliminate]: 7.77002e-06 [parameter_eliminate]: 3.78999e-06 [a_2]: 0.00024352 [accelerated_algorithm]: 1.763e-05 [shard]: 2.61e-06 [meta_shard_fg_expand]: 8.59e-06 [shard_inline]: 1.597e-05 [merge_send_recv]: 2.436e-05 [auto_parallel]: 1.873e-05 [parallel]: 2.134e-05 [flash_sp]: 1.427e-05 [merge_comm]: 1.087e-05 [allreduce_fusion]: 9.92999e-06 [matmul_add_comm_reduction]: 3.733e-05 [allreduce_slice_to_reducescatter]: 1.05001e-06 [virtual_shard_identity]: 2.396e-05 [virtual_dataset]: 1.638e-05 [get_grad_eliminate_]: 1.582e-05 [virtual_output]: 1.603e-05 [merge_forward]: 1.13e-05 [cell_reuse_recompute_pass]: 1.48002e-06 [offload_activation]: 2.162e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.642e-05 [merge_recompute_call_nodes]: 1.86e-06 [before_grad]: 3.092e-05 [set_forward_comm_id_for_comm_node_pass]: 1.231e-05 [meta_fg_expand]: 0.0265408 [flash_sp_send_recv_attached]: 6.06998e-06 [receive_attached]: 2.26998e-06 [after_resolve]: 0.00012463 [a_after_grad]: 0.00017216 [renormalize]: 0.00928643 [add_forward_monad_depend]: 1.139e-05 [auto_monad_grad]: 3.36001e-06 [auto_monad_eliminator]: 1.879e-05 [cse]: 3.197e-05 [a_3]: 5.231e-05 [Cycle 2]: 0.0256934, [45] [expand_dump_flag]: 2.59999e-06 [switch_simplify]: 7.38e-06 [loop_unroll]: 4.85001e-06 [a_1]: 6.082e-05 [with_stream_mark]: 2.024e-05 [recompute_prepare]: 5.27001e-06 [updatestate_depend_eliminate]: 3.71001e-06 [updatestate_assign_eliminate]: 2.39001e-06 [updatestate_loads_eliminate]: 2.79001e-06 [parameter_eliminate]: 2.02001e-06 [a_2]: 4.093e-05 [accelerated_algorithm]: 4.72998e-06 [shard]: 2.81e-06 [meta_shard_fg_expand]: 2.65002e-06 [shard_inline]: 4.00998e-06 [merge_send_recv]: 9.21002e-06 [auto_parallel]: 8.48999e-06 [parallel]: 9.64999e-06 [flash_sp]: 4.99998e-06 [merge_comm]: 3.41999e-06 [allreduce_fusion]: 2.52001e-06 [matmul_add_comm_reduction]: 2.055e-05 [allreduce_slice_to_reducescatter]: 8.50006e-07 [virtual_shard_identity]: 4.484e-05 [virtual_dataset]: 7.69002e-06 [get_grad_eliminate_]: 6.49001e-06 [virtual_output]: 5.23002e-06 [merge_forward]: 6.76e-06 [cell_reuse_recompute_pass]: 2.88998e-06 [offload_activation]: 1.08e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.191e-05 [merge_recompute_call_nodes]: 3.04001e-06 [before_grad]: 9.17001e-06 [set_forward_comm_id_for_comm_node_pass]: 7.92e-06 [meta_fg_expand]: 0.00075471 [flash_sp_send_recv_attached]: 3.31001e-06 [receive_attached]: 3.5e-06 [after_resolve]: 1.491e-05 [a_after_grad]: 7.92e-06 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 6.79001e-06 [auto_monad_grad]: 3.16999e-06 [auto_monad_eliminator]: 1.7e-05 [cse]: 3.853e-05 [a_3]: 2.959e-05 [py_interpret_to_execute_after_opt_a]: 1.256e-05 [slice_cell_reuse_recomputed_activation]: 2.88998e-06 [rewriter_after_opt_a]: 2.116e-05 [convert_after_rewriter]: 1.63002e-06 [order_py_execute_after_rewriter]: 1.19e-06 [mutable_eliminate]: 0.00120908 [opt_b]: 0.00017655, [1] [Cycle 1]: 0.00016542, [7] [b_1]: 8.424e-05 [b_2]: 6.41998e-06 [updatestate_depend_eliminate]: 8.17998e-06 [updatestate_assign_eliminate]: 2.51e-06 [updatestate_loads_eliminate]: 2.14999e-06 [renormalize]: 6.00005e-07 [cse]: 2.545e-05 [optimize_parallel_all_gather_comm]: 2.106e-05 [overlap_param_gather]: 1.87999e-06 [cconv]: 3.593e-05 [loop_unroll]: 0.00050318 [opt_after_cconv]: 9.375e-05, [1] [Cycle 1]: 8.632e-05, [7] [c_1]: 1.685e-05 [parameter_eliminate]: 5.47001e-06 [updatestate_depend_eliminate]: 6.36e-06 [updatestate_assign_eliminate]: 1.99e-06 [updatestate_loads_eliminate]: 1.71e-06 [cse]: 1.889e-05 [renormalize]: 5.50004e-07 [remove_dup_value]: 1.56e-05 [tuple_transform]: 6.208e-05, [1] [Cycle 1]: 5.669e-05, [4] [d_1]: 2.893e-05 [none_parameter_eliminate]: 1.86998e-06 [renormalize]: 1.99972e-07 [switch_simplify]: 5.37999e-06 [partial_unused_args_eliminate]: 2.25002e-06 [add_recomputation]: 4.145e-05 [cse_after_recomputation]: 2.495e-05, [1] [Cycle 1]: 1.81e-05, [1] [cse]: 1.162e-05 [environ_conv]: 5.87001e-06 [swap_dp_allreduce_reducescatter]: 4.79e-06 [bias_add_comm_swap]: 3.11999e-06 [label_micro_interleaved_index]: 6.04999e-06 [label_fine_grained_interleaved_index]: 3.12002e-06 [merge_cast_opt]: 1.57999e-06 [slice_recompute_activation]: 1.99e-06 [micro_interleaved_order_control]: 2.39001e-06 [assign_add_opt]: 1.67999e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.04e-06 [full_micro_interleaved_order_control]: 2.46e-06 [reorder_send_recv_between_fp_bp]: 3.03e-06 [comm_op_add_attrs]: 1.22999e-06 [add_comm_op_reuse_tag]: 1.00001e-06 [interleave_split_concat_branches]: 1.41002e-06 [interleave_parallel_branches]: 1.22999e-06 [overlap_opt_shard_in_pipeline]: 1.27999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.31e-06 [control_data_broadcast_order]: 1.282e-05 [grouped_pairwise_exchange_alltoall]: 1.69998e-06 [offloading_packed_experts]: 4.3e-06 [overlap_recompute_and_grad_model_parallel]: 4.23001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.09998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.75001e-06 [overlap_recompute_comm]: 2.40002e-06 [overlap_grad_ring_attention]: 3.81001e-06 [overlap_grad_flash_sp]: 1.879e-05 [begin_end_overlap_inline]: 5.09986e-07 [split_matmul_comm_elemetwise]: 2.69001e-06 [split_layernorm_comm]: 1.97001e-06 [handle_group_info]: 1.08001e-06 [symbol_engine_optimizer]: 0.00013862, [1] [Cycle 1]: 0.00013281, [6] [build]: 3.31999e-06 [elim_shapecalc]: 9.62999e-06 [elim_not_effective]: 1.158e-05 [opt_reshape]: 6.722e-05 [fold_const_symbol]: 7.97003e-06 [renormalize]: 2.19996e-07 [detach_backward]: 2.76e-06 [pipeline_parallel_scheduler]: 1.43002e-06 [auto_monad_reorder]: 1.857e-05 [get_jit_bprop_graph]: 2.66e-06 [rewriter_after_jit_bprop_graph]: 8.51002e-06 [opt_after_jit_grad]: 0.00056243 [validate]: 4.227e-05 Sums bootstrap : 0.000447s : 0.10% type_inference : 0.408217s : 89.79% event_method : 0.000098s : 0.02% auto_monad : 0.000336s : 0.07% graph_reusing : 0.000023s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000085s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000021s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000117s : 0.03% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000009s : 0.00% optimize.rewriter_before_opt_a : 0.000617s : 0.14% optimize.opt_a.expand_dump_flag : 0.000018s : 0.00% optimize.opt_a.switch_simplify : 0.000343s : 0.08% optimize.opt_a.loop_unroll : 0.000125s : 0.03% optimize.opt_a.a_1 : 0.003113s : 0.68% optimize.opt_a.with_stream_mark : 0.000061s : 0.01% optimize.opt_a.recompute_prepare : 0.000039s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000284s : 0.06% optimize.opt_a.accelerated_algorithm : 0.000022s : 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 : 0.000020s : 0.00% optimize.opt_a.merge_send_recv : 0.000034s : 0.01% optimize.opt_a.auto_parallel : 0.000027s : 0.01% optimize.opt_a.parallel : 0.000031s : 0.01% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000014s : 0.00% optimize.opt_a.allreduce_fusion : 0.000012s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000058s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000069s : 0.02% optimize.opt_a.virtual_dataset : 0.000024s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000022s : 0.00% optimize.opt_a.virtual_output : 0.000021s : 0.00% optimize.opt_a.merge_forward : 0.000018s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000032s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000058s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000005s : 0.00% optimize.opt_a.before_grad : 0.000040s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000020s : 0.00% optimize.opt_a.meta_fg_expand : 0.027296s : 6.00% 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 : 0.000140s : 0.03% optimize.opt_a.a_after_grad : 0.000180s : 0.04% optimize.opt_a.renormalize : 0.009287s : 2.04% optimize.opt_a.add_forward_monad_depend : 0.000018s : 0.00% optimize.opt_a.auto_monad_grad : 0.000007s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000036s : 0.01% optimize.opt_a.cse : 0.000070s : 0.02% optimize.opt_a.a_3 : 0.000082s : 0.02% optimize.py_interpret_to_execute_after_opt_a : 0.000013s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.00% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.001209s : 0.27% optimize.opt_b.b_1 : 0.000084s : 0.02% optimize.opt_b.b_2 : 0.000006s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000025s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.01% optimize.loop_unroll : 0.000503s : 0.11% optimize.opt_after_cconv.c_1 : 0.000017s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000019s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000016s : 0.00% optimize.tuple_transform.d_1 : 0.000029s : 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 : 0.000005s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000041s : 0.01% optimize.cse_after_recomputation.cse : 0.000012s : 0.00% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000013s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000019s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000012s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000067s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000008s : 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 : 0.000019s : 0.00% get_jit_bprop_graph : 0.000003s : 0.00% rewriter_after_jit_bprop_graph : 0.000009s : 0.00% opt_after_jit_grad : 0.000562s : 0.12% validate : 0.000042s : 0.01% Time group info: ------[substitution.] 0.001108 134 3.87% : 0.000043s : 8: substitution.cast_eliminate 0.28% : 0.000003s : 1: substitution.elim_not_effective 0.66% : 0.000007s : 9: substitution.float_depend_g_call 0.47% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.12% : 0.000001s : 1: substitution.fold_const_symbol 0.50% : 0.000006s : 1: substitution.graph_param_transform 0.44% : 0.000005s : 2: substitution.incorporate_call 0.17% : 0.000002s : 2: substitution.incorporate_call_switch 78.06% : 0.000865s : 33: substitution.inline 3.99% : 0.000044s : 4: substitution.inline_without_move 0.93% : 0.000010s : 11: substitution.j_node_and_user_rematch 0.40% : 0.000004s : 2: substitution.minmaximum_grad 0.64% : 0.000007s : 9: substitution.partial_eliminate 0.94% : 0.000010s : 11: substitution.remove_not_recompute_node 0.41% : 0.000005s : 1: substitution.replace_applicator 0.94% : 0.000010s : 12: substitution.replace_old_param 0.45% : 0.000005s : 1: substitution.set_cell_output_no_recompute 2.14% : 0.000024s : 10: substitution.switch_simplify 0.87% : 0.000010s : 2: substitution.tuple_list_convert_item_index_to_positive 0.31% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.54% : 0.000006s : 2: substitution.tuple_list_get_item_depend_reorder 2.40% : 0.000027s : 6: substitution.tuple_list_get_item_eliminator 0.47% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.407839 2 85.86% : 0.350183s : 1: type_inference.infer 14.14% : 0.057657s : 1: type_inference.specialize ------[replace.] 0.000421 45 65.97% : 0.000278s : 33: replace.inline 30.15% : 0.000127s : 10: replace.switch_simplify 3.88% : 0.000016s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000875 45 96.47% : 0.000844s : 33: match.inline 2.21% : 0.000019s : 10: match.switch_simplify 1.33% : 0.000012s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000540 3189 1.14% : 0.000006s : 43: predicate.accumulaten_eliminater 0.44% : 0.000002s : 1: predicate.ad_related_special_op_eliminate 0.40% : 0.000002s : 14: predicate.addn_check_dump 1.21% : 0.000007s : 43: predicate.addn_zero_filter 1.03% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 1.91% : 0.000010s : 57: predicate.arithmetic_simplify 1.46% : 0.000008s : 43: predicate.cast_eliminate 0.13% : 0.000001s : 2: predicate.check_bprop_eliminate 0.37% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.35% : 0.000002s : 14: predicate.depend_value_elim 1.18% : 0.000006s : 43: predicate.dict_get_item_const_eliminator 1.22% : 0.000007s : 43: predicate.dict_get_item_eliminator 1.07% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.44% : 0.000002s : 2: predicate.dumpgradient_eliminate 0.06% : 0.000000s : 1: predicate.elim_not_effective 0.11% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.23% : 0.000007s : 44: predicate.environ_add_const_eliminate 1.11% : 0.000006s : 44: predicate.environ_get_add_eliminate 1.12% : 0.000006s : 44: predicate.environ_get_depend_swap 1.48% : 0.000008s : 58: predicate.environ_get_eliminate 1.17% : 0.000006s : 44: predicate.environ_get_set_eliminate 2.05% : 0.000011s : 78: predicate.exchange_switch_depend_value 3.75% : 0.000020s : 78: predicate.float_depend_g_call 0.40% : 0.000002s : 14: predicate.float_environ_get_switch 0.54% : 0.000003s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.50% : 0.000003s : 14: predicate.get_grad_eliminate 0.09% : 0.000000s : 1: predicate.graph_param_transform 0.47% : 0.000003s : 14: predicate.incorporate_call 0.34% : 0.000002s : 14: predicate.incorporate_call_switch 5.90% : 0.000032s : 151: predicate.inline 2.70% : 0.000015s : 70: predicate.inline_without_move 0.19% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.63% : 0.000003s : 14: predicate.less_batch_normalization 1.67% : 0.000009s : 47: predicate.list_to_tuple_eliminator_ 2.26% : 0.000012s : 90: predicate.load_eliminater 0.50% : 0.000003s : 1: predicate.loop_unroll_after_grad 3.93% : 0.000021s : 141: predicate.loop_unroll_before_grad 1.37% : 0.000007s : 45: predicate.make_slice_get_slice_eliminator 0.40% : 0.000002s : 14: predicate.merge_addn 0.12% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.12% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.05% : 0.000006s : 43: predicate.minmaximum_grad 0.65% : 0.000003s : 1: predicate.mutable_eliminate 0.15% : 0.000001s : 1: predicate.opt_reshape 0.11% : 0.000001s : 1: predicate.parallel_virtual_node 3.12% : 0.000017s : 78: predicate.partial_defer_inline 1.30% : 0.000007s : 46: predicate.partial_eliminate 1.10% : 0.000006s : 43: predicate.print_const_string_wrapper 0.39% : 0.000002s : 14: predicate.reduce_all_const_elim 1.44% : 0.000008s : 43: predicate.reduce_eliminate 2.49% : 0.000013s : 90: predicate.redundant_stop_gradient_eliminater 0.37% : 0.000002s : 14: predicate.remove_not_recompute_node 1.04% : 0.000006s : 47: predicate.replace_applicator 1.22% : 0.000007s : 70: predicate.replace_old_param 0.08% : 0.000000s : 1: predicate.reset_defer_inline 1.23% : 0.000007s : 43: predicate.reshape_eliminate 0.17% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.11% : 0.000001s : 1: predicate.row_tensor_eliminate 0.29% : 0.000002s : 2: predicate.same_eliminate 0.27% : 0.000001s : 14: predicate.set_cell_output_no_recompute 0.80% : 0.000004s : 14: predicate.shard_identity_eliminate 0.11% : 0.000001s : 2: predicate.special_op_eliminate 0.47% : 0.000003s : 14: predicate.specialize_transform 0.35% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 2.53% : 0.000014s : 70: predicate.stack_unstack_eliminate 0.08% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.27% : 0.000012s : 78: predicate.switch_defer_inline 2.37% : 0.000013s : 80: predicate.switch_layer_defer_inline 7.95% : 0.000043s : 254: predicate.switch_simplify 1.23% : 0.000007s : 43: predicate.tile_eliminate 1.17% : 0.000006s : 43: predicate.transpose_eliminate 1.38% : 0.000007s : 45: predicate.tuple_list_convert_item_index_to_positive 1.37% : 0.000007s : 45: predicate.tuple_list_get_item_const_eliminator 1.25% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.77% : 0.000015s : 61: predicate.tuple_list_get_item_eliminator 1.29% : 0.000007s : 45: predicate.tuple_list_get_set_item_eliminator 1.99% : 0.000011s : 59: predicate.tuple_list_set_item_eliminator 1.29% : 0.000007s : 47: predicate.tuple_to_list_eliminator_ 2.20% : 0.000012s : 90: predicate.updatestate_pure_node_eliminater 2.67% : 0.000014s : 104: predicate.updatestate_useless_node_eliminater 0.07% : 0.000000s : 1: predicate.value_based_eliminate 0.56% : 0.000003s : 14: predicate.virtual_dataset_eliminate 0.42% : 0.000002s : 14: predicate.virtual_output_eliminate 0.03% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.18% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.005643 69 70.07% : 0.003954s : 33: func_graph_cloner_run.FuncGraphClonerGraph 29.93% : 0.001689s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.572517 192 0.00% : 0.000004s : 1: ForceFp32Comm 0.74% : 0.004233s : 1: add_attr 0.74% : 0.004217s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000045s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.06% : 0.000350s : 1: auto_monad 0.03% : 0.000193s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.08% : 0.000480s : 1: bootstrap 0.01% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000016s : 1: control_data_broadcast_order 0.00% : 0.000005s : 1: convert_after_rewriter 0.00% : 0.000028s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000009s : 1: environ_conv 0.02% : 0.000108s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000008s : 1: get_jit_bprop_graph 0.00% : 0.000028s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.09% : 0.000515s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.21% : 0.001227s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000020s : 1: opt.transform.mutable_eliminate 0.78% : 0.004455s : 78: opt.transform.opt_a 0.00% : 0.000015s : 1: opt.transform.opt_after_cconv 0.00% : 0.000023s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000057s : 28: opt.transform.opt_b 0.01% : 0.000032s : 2: opt.transform.opt_trans_graph 0.01% : 0.000032s : 4: opt.transform.symbol_engine_opt 11.64% : 0.066649s : 1: opt_a 0.02% : 0.000097s : 1: opt_after_cconv 0.10% : 0.000579s : 1: opt_after_jit_grad 0.03% : 0.000180s : 1: opt_b 12.23% : 0.070009s : 1: optimize 0.00% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000005s : 1: order_py_execute_after_rewriter 0.00% : 0.000022s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.02% : 0.000123s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000016s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000020s : 1: remove_dup_value 1.45% : 0.008329s : 1: renormalize.infer 0.16% : 0.000941s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000013s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000025s : 1: rewriter_after_opt_a 0.11% : 0.000629s : 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% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000142s : 1: symbol_engine_optimizer 0.01% : 0.000065s : 1: tuple_transform 71.31% : 0.408270s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:44:54.127.64 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.251255, [21] [bootstrap]: 0.00050031 [type_inference]: 0.183815 [event_method]: 0.00049901 [auto_monad]: 0.00027198 [graph_reusing]: 1.17e-05 [inline]: 3.08998e-06 [add_attr]: 0.0560144, [1] [add_attr_with_inline]: 0.0559987, [1] [Cycle 1]: 0.00013019, [2] [tag_attr]: 6.563e-05 [meta_addattr_fg_expand]: 1.37e-05 [parallel-infer-symbol]: 4.21001e-06 [pre_auto_parallel]: 0.00035891 [insert-virtual-dataset]: 2.48e-06 [parallel-infer-symbol-second]: 9.5999e-07 [dataset_repeat_opt]: 3.02002e-06 [pipeline_split]: 1.69998e-06 [optimize]: 0.0087636, [53] [py_interpret_to_execute]: 1.139e-05 [rewriter_before_opt_a]: 0.00037044 [opt_a]: 0.00568622, [2] [Cycle 1]: 0.00501315, [45] [expand_dump_flag]: 5.70001e-06 [switch_simplify]: 0.0001726 [loop_unroll]: 5.778e-05 [a_1]: 0.00131059 [with_stream_mark]: 2.781e-05 [recompute_prepare]: 1.213e-05 [updatestate_depend_eliminate]: 4.52e-06 [updatestate_assign_eliminate]: 4.19002e-06 [updatestate_loads_eliminate]: 2.98998e-06 [parameter_eliminate]: 2.69001e-06 [a_2]: 0.00011588 [accelerated_algorithm]: 7.68001e-06 [shard]: 1.74e-06 [meta_shard_fg_expand]: 4.85001e-06 [shard_inline]: 6.68e-06 [merge_send_recv]: 9.79e-06 [auto_parallel]: 1.015e-05 [parallel]: 2.21e-05 [flash_sp]: 1.284e-05 [merge_comm]: 3.7e-06 [allreduce_fusion]: 3.00002e-06 [matmul_add_comm_reduction]: 1.159e-05 [allreduce_slice_to_reducescatter]: 9.10019e-07 [virtual_shard_identity]: 9.47001e-06 [virtual_dataset]: 6.00002e-06 [get_grad_eliminate_]: 6.07001e-06 [virtual_output]: 5.98002e-06 [merge_forward]: 3.65e-06 [cell_reuse_recompute_pass]: 2.01e-06 [offload_activation]: 1.042e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.46e-05 [merge_recompute_call_nodes]: 1.39e-06 [before_grad]: 1.12e-05 [set_forward_comm_id_for_comm_node_pass]: 3.66999e-06 [meta_fg_expand]: 3.85e-06 [flash_sp_send_recv_attached]: 2.73998e-06 [receive_attached]: 2.36e-06 [after_resolve]: 1.119e-05 [a_after_grad]: 9.74e-06 [renormalize]: 0.00263291 [add_forward_monad_depend]: 1.204e-05 [auto_monad_grad]: 3.48e-06 [auto_monad_eliminator]: 2.491e-05 [cse]: 3.599e-05 [a_3]: 6.174e-05 [Cycle 2]: 0.00065586, [45] [expand_dump_flag]: 2.41e-06 [switch_simplify]: 9.42001e-06 [loop_unroll]: 5.89999e-06 [a_1]: 0.00011421 [with_stream_mark]: 1.876e-05 [recompute_prepare]: 6.77002e-06 [updatestate_depend_eliminate]: 3.63e-06 [updatestate_assign_eliminate]: 3.16001e-06 [updatestate_loads_eliminate]: 3.43e-06 [parameter_eliminate]: 2.23002e-06 [a_2]: 5.995e-05 [accelerated_algorithm]: 6.07999e-06 [shard]: 2.36998e-06 [meta_shard_fg_expand]: 2.26998e-06 [shard_inline]: 4.92e-06 [merge_send_recv]: 8.75999e-06 [auto_parallel]: 9.27001e-06 [parallel]: 1.035e-05 [flash_sp]: 3.97e-06 [merge_comm]: 3.34001e-06 [allreduce_fusion]: 2.93e-06 [matmul_add_comm_reduction]: 9.97001e-06 [allreduce_slice_to_reducescatter]: 7.99977e-07 [virtual_shard_identity]: 7.31001e-06 [virtual_dataset]: 5.26998e-06 [get_grad_eliminate_]: 5.36002e-06 [virtual_output]: 5.19998e-06 [merge_forward]: 3.5e-06 [cell_reuse_recompute_pass]: 2.89001e-06 [offload_activation]: 9.46998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.751e-05 [merge_recompute_call_nodes]: 1.25001e-06 [before_grad]: 9.24e-06 [set_forward_comm_id_for_comm_node_pass]: 3.61999e-06 [meta_fg_expand]: 3.24001e-06 [flash_sp_send_recv_attached]: 1.62999e-06 [receive_attached]: 2.44001e-06 [after_resolve]: 1.182e-05 [a_after_grad]: 8.45001e-06 [renormalize]: 5.9983e-08 [add_forward_monad_depend]: 1.62999e-06 [auto_monad_grad]: 1.45999e-06 [auto_monad_eliminator]: 8.22e-06 [cse]: 1.762e-05 [a_3]: 3.238e-05 [py_interpret_to_execute_after_opt_a]: 1.161e-05 [slice_cell_reuse_recomputed_activation]: 2.17999e-06 [rewriter_after_opt_a]: 3.801e-05 [convert_after_rewriter]: 3.86999e-06 [order_py_execute_after_rewriter]: 1.27999e-06 [mutable_eliminate]: 0.00082569 [opt_b]: 0.00021019, [1] [Cycle 1]: 0.00020137, [7] [b_1]: 0.00011757 [b_2]: 7.87998e-06 [updatestate_depend_eliminate]: 7.04001e-06 [updatestate_assign_eliminate]: 2.78998e-06 [updatestate_loads_eliminate]: 2.12999e-06 [renormalize]: 6.80011e-07 [cse]: 2.599e-05 [optimize_parallel_all_gather_comm]: 2.202e-05 [overlap_param_gather]: 2.02999e-06 [cconv]: 3.583e-05 [loop_unroll]: 0.00062151 [opt_after_cconv]: 0.00011488, [1] [Cycle 1]: 0.00010588, [7] [c_1]: 2.883e-05 [parameter_eliminate]: 5.21998e-06 [updatestate_depend_eliminate]: 6.73e-06 [updatestate_assign_eliminate]: 2.36e-06 [updatestate_loads_eliminate]: 2.11998e-06 [cse]: 2.469e-05 [renormalize]: 6.69999e-07 [remove_dup_value]: 1.373e-05 [tuple_transform]: 7.683e-05, [1] [Cycle 1]: 7.205e-05, [4] [d_1]: 4.395e-05 [none_parameter_eliminate]: 1.62999e-06 [renormalize]: 2.19996e-07 [switch_simplify]: 6.38e-06 [partial_unused_args_eliminate]: 1.76e-06 [add_recomputation]: 5.089e-05 [cse_after_recomputation]: 2.188e-05, [1] [Cycle 1]: 1.63e-05, [1] [cse]: 1.091e-05 [environ_conv]: 6.02999e-06 [swap_dp_allreduce_reducescatter]: 5.08002e-06 [bias_add_comm_swap]: 3.41001e-06 [label_micro_interleaved_index]: 6.39001e-06 [label_fine_grained_interleaved_index]: 2.56998e-06 [merge_cast_opt]: 1.40001e-06 [slice_recompute_activation]: 2.09999e-06 [micro_interleaved_order_control]: 2.89001e-06 [assign_add_opt]: 1.50001e-06 [ForceFp32Comm]: 8.39995e-07 [remove_cast_before_assign_add]: 1.00001e-06 [full_micro_interleaved_order_control]: 2.31e-06 [reorder_send_recv_between_fp_bp]: 2.47001e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 9.70002e-07 [interleave_split_concat_branches]: 1.15999e-06 [interleave_parallel_branches]: 1.20999e-06 [overlap_opt_shard_in_pipeline]: 1.25999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.00002e-06 [control_data_broadcast_order]: 1.294e-05 [grouped_pairwise_exchange_alltoall]: 1.82999e-06 [offloading_packed_experts]: 3.73999e-06 [overlap_recompute_and_grad_model_parallel]: 4.72998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.09e-06 [overlap_recompute_allgather_and_fa_grad]: 1.32e-06 [overlap_recompute_comm]: 2.07999e-06 [overlap_grad_ring_attention]: 3.66001e-06 [overlap_grad_flash_sp]: 2.228e-05 [begin_end_overlap_inline]: 4.69998e-07 [split_matmul_comm_elemetwise]: 2.96001e-06 [split_layernorm_comm]: 1.94e-06 [handle_group_info]: 1.05999e-06 [symbol_engine_optimizer]: 0.00021363, [1] [Cycle 1]: 0.00020866, [6] [build]: 4.13999e-06 [elim_shapecalc]: 1.038e-05 [elim_not_effective]: 1.294e-05 [opt_reshape]: 0.00013533 [fold_const_symbol]: 1.281e-05 [renormalize]: 2.50002e-07 [detach_backward]: 2.37999e-06 [pipeline_parallel_scheduler]: 1.67001e-06 [auto_monad_reorder]: 1.968e-05 [get_jit_bprop_graph]: 2.27001e-06 [rewriter_after_jit_bprop_graph]: 7.16999e-06 [opt_after_jit_grad]: 0.00067856 [validate]: 5.086e-05 Sums bootstrap : 0.000500s : 0.26% type_inference : 0.183815s : 94.73% event_method : 0.000499s : 0.26% auto_monad : 0.000272s : 0.14% graph_reusing : 0.000012s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000066s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000014s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000359s : 0.18% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000003s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000011s : 0.01% optimize.rewriter_before_opt_a : 0.000370s : 0.19% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000182s : 0.09% optimize.opt_a.loop_unroll : 0.000064s : 0.03% optimize.opt_a.a_1 : 0.001425s : 0.73% optimize.opt_a.with_stream_mark : 0.000047s : 0.02% optimize.opt_a.recompute_prepare : 0.000019s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000176s : 0.09% optimize.opt_a.accelerated_algorithm : 0.000014s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000012s : 0.01% optimize.opt_a.merge_send_recv : 0.000019s : 0.01% optimize.opt_a.auto_parallel : 0.000019s : 0.01% optimize.opt_a.parallel : 0.000032s : 0.02% optimize.opt_a.flash_sp : 0.000017s : 0.01% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000006s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000022s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000017s : 0.01% optimize.opt_a.virtual_dataset : 0.000011s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.01% optimize.opt_a.virtual_output : 0.000011s : 0.01% optimize.opt_a.merge_forward : 0.000007s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000020s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000032s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000020s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.00% optimize.opt_a.meta_fg_expand : 0.000007s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000023s : 0.01% optimize.opt_a.a_after_grad : 0.000018s : 0.01% optimize.opt_a.renormalize : 0.002633s : 1.36% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000033s : 0.02% optimize.opt_a.cse : 0.000054s : 0.03% optimize.opt_a.a_3 : 0.000094s : 0.05% optimize.py_interpret_to_execute_after_opt_a : 0.000012s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000038s : 0.02% optimize.convert_after_rewriter : 0.000004s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000826s : 0.43% optimize.opt_b.b_1 : 0.000118s : 0.06% optimize.opt_b.b_2 : 0.000008s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000026s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.02% optimize.loop_unroll : 0.000622s : 0.32% optimize.opt_after_cconv.c_1 : 0.000029s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000025s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000044s : 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 : 0.000006s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000051s : 0.03% optimize.cse_after_recomputation.cse : 0.000011s : 0.01% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000013s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000022s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000013s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000135s : 0.07% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000013s : 0.01% 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 : 0.000020s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000679s : 0.35% validate : 0.000051s : 0.03% Time group info: ------[substitution.] 0.000419 39 4.00% : 0.000017s : 1: substitution.cast_eliminate 0.57% : 0.000002s : 2: substitution.elim_not_effective 0.43% : 0.000002s : 2: substitution.fold_const_symbol 1.44% : 0.000006s : 3: substitution.graph_param_transform 83.10% : 0.000348s : 15: substitution.inline 1.13% : 0.000005s : 4: substitution.j_node_and_user_rematch 1.39% : 0.000006s : 4: substitution.remove_not_recompute_node 1.37% : 0.000006s : 2: substitution.replace_old_param 3.60% : 0.000015s : 4: substitution.switch_simplify 2.96% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.183688 2 97.74% : 0.179531s : 1: type_inference.infer 2.26% : 0.004157s : 1: type_inference.specialize ------[replace.] 0.000194 22 4.57% : 0.000009s : 1: replace.cast_eliminate 56.78% : 0.000110s : 15: replace.inline 29.42% : 0.000057s : 4: replace.switch_simplify 9.24% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000379 22 3.89% : 0.000015s : 1: match.cast_eliminate 89.69% : 0.000339s : 15: match.inline 3.43% : 0.000013s : 4: match.switch_simplify 2.99% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000303 1716 1.14% : 0.000003s : 22: predicate.accumulaten_eliminater 0.65% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.32% : 0.000001s : 6: predicate.addn_check_dump 1.04% : 0.000003s : 22: predicate.addn_zero_filter 1.05% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.38% : 0.000007s : 28: predicate.arithmetic_simplify 1.43% : 0.000004s : 23: predicate.cast_eliminate 0.33% : 0.000001s : 6: predicate.check_bprop_eliminate 0.34% : 0.000001s : 6: predicate.compare_switch_simplify 0.09% : 0.000000s : 3: predicate.const_output_eliminate 0.39% : 0.000001s : 6: predicate.depend_value_elim 1.12% : 0.000003s : 23: predicate.dict_get_item_const_eliminator 1.21% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.07% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.62% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.12% : 0.000000s : 3: predicate.elim_not_effective 0.32% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.25% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.25% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.26% : 0.000004s : 26: predicate.environ_get_depend_swap 1.57% : 0.000005s : 32: predicate.environ_get_eliminate 1.20% : 0.000004s : 26: predicate.environ_get_set_eliminate 1.90% : 0.000006s : 40: predicate.exchange_switch_depend_value 2.69% : 0.000008s : 40: predicate.float_depend_g_call 0.37% : 0.000001s : 6: predicate.float_environ_get_switch 0.51% : 0.000002s : 9: predicate.float_tuple_getitem_switch 0.11% : 0.000000s : 3: predicate.fold_const_symbol 0.41% : 0.000001s : 6: predicate.get_grad_eliminate 0.16% : 0.000000s : 3: predicate.graph_param_transform 0.28% : 0.000001s : 6: predicate.incorporate_call 0.25% : 0.000001s : 6: predicate.incorporate_call_switch 5.81% : 0.000018s : 80: predicate.inline 0.39% : 0.000001s : 6: predicate.inline_without_move 0.19% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.65% : 0.000002s : 6: predicate.less_batch_normalization 1.71% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.58% : 0.000008s : 53: predicate.load_eliminater 0.92% : 0.000003s : 3: predicate.loop_unroll_after_grad 3.26% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.63% : 0.000005s : 29: predicate.make_slice_get_slice_eliminator 0.34% : 0.000001s : 6: predicate.merge_addn 0.36% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.43% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.05% : 0.000003s : 22: predicate.minmaximum_grad 0.98% : 0.000003s : 3: predicate.mutable_eliminate 0.29% : 0.000001s : 3: predicate.opt_reshape 0.25% : 0.000001s : 3: predicate.parallel_virtual_node 3.01% : 0.000009s : 40: predicate.partial_defer_inline 1.38% : 0.000004s : 28: predicate.partial_eliminate 1.06% : 0.000003s : 22: predicate.print_const_string_wrapper 0.39% : 0.000001s : 6: predicate.reduce_all_const_elim 1.57% : 0.000005s : 22: predicate.reduce_eliminate 2.74% : 0.000008s : 53: predicate.redundant_stop_gradient_eliminater 0.29% : 0.000001s : 6: predicate.remove_not_recompute_node 1.23% : 0.000004s : 31: predicate.replace_applicator 0.35% : 0.000001s : 6: predicate.replace_old_param 0.16% : 0.000000s : 3: predicate.reset_defer_inline 1.44% : 0.000004s : 22: predicate.reshape_eliminate 0.48% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.28% : 0.000001s : 3: predicate.row_tensor_eliminate 0.66% : 0.000002s : 6: predicate.same_eliminate 0.22% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.60% : 0.000002s : 6: predicate.shard_identity_eliminate 0.59% : 0.000002s : 6: predicate.special_op_eliminate 0.38% : 0.000001s : 6: predicate.specialize_transform 0.74% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.52% : 0.000002s : 6: predicate.stack_unstack_eliminate 0.20% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.09% : 0.000006s : 40: predicate.switch_defer_inline 2.37% : 0.000007s : 46: predicate.switch_layer_defer_inline 7.07% : 0.000021s : 126: predicate.switch_simplify 1.20% : 0.000004s : 22: predicate.tile_eliminate 1.08% : 0.000003s : 22: predicate.transpose_eliminate 1.76% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.68% : 0.000005s : 29: predicate.tuple_list_get_item_const_eliminator 1.63% : 0.000005s : 29: predicate.tuple_list_get_item_depend_reorder 2.95% : 0.000009s : 37: predicate.tuple_list_get_item_eliminator 1.54% : 0.000005s : 29: predicate.tuple_list_get_set_item_eliminator 2.27% : 0.000007s : 35: predicate.tuple_list_set_item_eliminator 1.68% : 0.000005s : 31: predicate.tuple_to_list_eliminator_ 2.29% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 2.91% : 0.000009s : 59: predicate.updatestate_useless_node_eliminater 0.26% : 0.000001s : 3: predicate.value_based_eliminate 0.44% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.43% : 0.000001s : 6: predicate.virtual_output_eliminate 0.15% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.24% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002474 27 44.32% : 0.001096s : 10: func_graph_cloner_run.FuncGraphClonerGraph 55.68% : 0.001377s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.320918 192 0.00% : 0.000003s : 1: ForceFp32Comm 17.46% : 0.056021s : 1: add_attr 17.45% : 0.056003s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 0.02% : 0.000055s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.09% : 0.000288s : 1: auto_monad 0.01% : 0.000025s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.17% : 0.000540s : 1: bootstrap 0.01% : 0.000039s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000016s : 1: control_data_broadcast_order 0.00% : 0.000008s : 1: convert_after_rewriter 0.01% : 0.000025s : 1: cse_after_recomputation 0.00% : 0.000006s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000010s : 1: environ_conv 0.16% : 0.000526s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.01% : 0.000017s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000009s : 1: label_micro_interleaved_index 0.20% : 0.000634s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.26% : 0.000841s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000019s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000025s : 1: opt.transform.mutable_eliminate 0.63% : 0.002027s : 78: opt.transform.opt_a 0.01% : 0.000027s : 1: opt.transform.opt_after_cconv 0.01% : 0.000032s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000092s : 28: opt.transform.opt_b 0.01% : 0.000047s : 2: opt.transform.opt_trans_graph 0.05% : 0.000168s : 4: opt.transform.symbol_engine_opt 1.77% : 0.005691s : 1: opt_a 0.04% : 0.000119s : 1: opt_after_cconv 0.22% : 0.000697s : 1: opt_after_jit_grad 0.07% : 0.000214s : 1: opt_b 2.73% : 0.008770s : 1: optimize 0.01% : 0.000026s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000026s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000006s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000008s : 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% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.11% : 0.000368s : 1: pre_auto_parallel 0.00% : 0.000015s : 1: py_interpret_to_execute 0.00% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000017s : 1: remove_dup_value 0.39% : 0.001267s : 1: renormalize.infer 0.42% : 0.001347s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.03% : 0.000111s : 1: rewriter_after_opt_a 0.12% : 0.000380s : 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% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.07% : 0.000217s : 1: symbol_engine_optimizer 0.02% : 0.000079s : 1: tuple_transform 57.29% : 0.183847s : 1: type_inference TotalTime = 0.277639, [21] [bootstrap]: 0.00044695 [type_inference]: 0.234658 [event_method]: 9.384e-05 [auto_monad]: 0.00024177 [graph_reusing]: 1.391e-05 [inline]: 3.29001e-06 [add_attr]: 0.0285521, [1] [add_attr_with_inline]: 0.0285383, [1] [Cycle 1]: 0.00013329, [2] [tag_attr]: 7.132e-05 [meta_addattr_fg_expand]: 1.61e-05 [parallel-infer-symbol]: 3.31999e-06 [pre_auto_parallel]: 9.652e-05 [insert-virtual-dataset]: 2.48998e-06 [parallel-infer-symbol-second]: 2.04999e-06 [dataset_repeat_opt]: 1.89e-06 [pipeline_split]: 1.65001e-06 [optimize]: 0.0127127, [53] [py_interpret_to_execute]: 9.49999e-06 [rewriter_before_opt_a]: 0.00048884 [opt_a]: 0.0101583, [2] [Cycle 1]: 0.0093062, [45] [expand_dump_flag]: 8.02e-06 [switch_simplify]: 0.00025198 [loop_unroll]: 9.063e-05 [a_1]: 0.00215858 [with_stream_mark]: 3.465e-05 [recompute_prepare]: 2.551e-05 [updatestate_depend_eliminate]: 9.37999e-06 [updatestate_assign_eliminate]: 6.62002e-06 [updatestate_loads_eliminate]: 6.48e-06 [parameter_eliminate]: 3.38e-06 [a_2]: 0.00020731 [accelerated_algorithm]: 1.561e-05 [shard]: 1.89e-06 [meta_shard_fg_expand]: 5.42001e-06 [shard_inline]: 1.26e-05 [merge_send_recv]: 1.877e-05 [auto_parallel]: 1.389e-05 [parallel]: 2.032e-05 [flash_sp]: 1.222e-05 [merge_comm]: 8.07e-06 [allreduce_fusion]: 7.93001e-06 [matmul_add_comm_reduction]: 3.226e-05 [allreduce_slice_to_reducescatter]: 1.34e-06 [virtual_shard_identity]: 1.812e-05 [virtual_dataset]: 1.339e-05 [get_grad_eliminate_]: 1.478e-05 [virtual_output]: 1.278e-05 [merge_forward]: 8.23001e-06 [cell_reuse_recompute_pass]: 1.59e-06 [offload_activation]: 1.779e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.672e-05 [merge_recompute_call_nodes]: 1.64e-06 [before_grad]: 2.611e-05 [set_forward_comm_id_for_comm_node_pass]: 8.03001e-06 [meta_fg_expand]: 0.00217336 [flash_sp_send_recv_attached]: 5.37001e-06 [receive_attached]: 2.44999e-06 [after_resolve]: 6.951e-05 [a_after_grad]: 8.939e-05 [renormalize]: 0.00341619 [add_forward_monad_depend]: 8.13001e-06 [auto_monad_grad]: 2.70997e-06 [auto_monad_eliminator]: 1.55e-05 [cse]: 3.09e-05 [a_3]: 4.624e-05 [Cycle 2]: 0.00083546, [45] [expand_dump_flag]: 2.54999e-06 [switch_simplify]: 6.61e-06 [loop_unroll]: 4.97999e-06 [a_1]: 5.462e-05 [with_stream_mark]: 1.522e-05 [recompute_prepare]: 4.48999e-06 [updatestate_depend_eliminate]: 3.50998e-06 [updatestate_assign_eliminate]: 3.21001e-06 [updatestate_loads_eliminate]: 2.71999e-06 [parameter_eliminate]: 2.07999e-06 [a_2]: 3.873e-05 [accelerated_algorithm]: 4.10998e-06 [shard]: 2.29001e-06 [meta_shard_fg_expand]: 2.32001e-06 [shard_inline]: 3.58e-06 [merge_send_recv]: 7.15e-06 [auto_parallel]: 8.75999e-06 [parallel]: 8.76002e-06 [flash_sp]: 3.48999e-06 [merge_comm]: 2.48e-06 [allreduce_fusion]: 2.30002e-06 [matmul_add_comm_reduction]: 7.93001e-06 [allreduce_slice_to_reducescatter]: 8.99978e-07 [virtual_shard_identity]: 5.17e-06 [virtual_dataset]: 3.96001e-06 [get_grad_eliminate_]: 3.52997e-06 [virtual_output]: 3.45e-06 [merge_forward]: 3.46001e-06 [cell_reuse_recompute_pass]: 3.20002e-06 [offload_activation]: 7.97e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.32e-05 [merge_recompute_call_nodes]: 1.44e-06 [before_grad]: 6.59999e-06 [set_forward_comm_id_for_comm_node_pass]: 2.51e-06 [meta_fg_expand]: 0.00030983 [flash_sp_send_recv_attached]: 2.74001e-06 [receive_attached]: 2.41e-06 [after_resolve]: 9.81e-06 [a_after_grad]: 6.04001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.69001e-06 [auto_monad_grad]: 1.72001e-06 [auto_monad_eliminator]: 8.54e-06 [cse]: 1.59e-05 [a_3]: 2.293e-05 [py_interpret_to_execute_after_opt_a]: 6.39999e-06 [slice_cell_reuse_recomputed_activation]: 1.95001e-06 [rewriter_after_opt_a]: 1.908e-05 [convert_after_rewriter]: 1.28002e-06 [order_py_execute_after_rewriter]: 1.17999e-06 [mutable_eliminate]: 0.00073505 [opt_b]: 0.00015134, [1] [Cycle 1]: 0.00014152, [7] [b_1]: 7.371e-05 [b_2]: 5.42999e-06 [updatestate_depend_eliminate]: 6.40997e-06 [updatestate_assign_eliminate]: 1.79e-06 [updatestate_loads_eliminate]: 1.81998e-06 [renormalize]: 8.2e-07 [cse]: 1.831e-05 [optimize_parallel_all_gather_comm]: 1.8e-05 [overlap_param_gather]: 1.74998e-06 [cconv]: 3.14e-05 [loop_unroll]: 0.00047888 [opt_after_cconv]: 8.471e-05, [1] [Cycle 1]: 7.816e-05, [7] [c_1]: 1.582e-05 [parameter_eliminate]: 4.52e-06 [updatestate_depend_eliminate]: 5.47001e-06 [updatestate_assign_eliminate]: 1.59e-06 [updatestate_loads_eliminate]: 1.46002e-06 [cse]: 1.619e-05 [renormalize]: 8.70001e-07 [remove_dup_value]: 1.597e-05 [tuple_transform]: 5.539e-05, [1] [Cycle 1]: 5.005e-05, [4] [d_1]: 2.581e-05 [none_parameter_eliminate]: 1.76e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 4.48999e-06 [partial_unused_args_eliminate]: 1.84998e-06 [add_recomputation]: 3.766e-05 [cse_after_recomputation]: 1.897e-05, [1] [Cycle 1]: 1.419e-05, [1] [cse]: 8.21002e-06 [environ_conv]: 5.31002e-06 [swap_dp_allreduce_reducescatter]: 5.87999e-06 [bias_add_comm_swap]: 2.71e-06 [label_micro_interleaved_index]: 4.72e-06 [label_fine_grained_interleaved_index]: 3.06001e-06 [merge_cast_opt]: 1.51998e-06 [slice_recompute_activation]: 1.99e-06 [micro_interleaved_order_control]: 2.77002e-06 [assign_add_opt]: 1.35001e-06 [ForceFp32Comm]: 1.02998e-06 [remove_cast_before_assign_add]: 1.30001e-06 [full_micro_interleaved_order_control]: 2.04e-06 [reorder_send_recv_between_fp_bp]: 2.74001e-06 [comm_op_add_attrs]: 1.04998e-06 [add_comm_op_reuse_tag]: 1.00999e-06 [interleave_split_concat_branches]: 1.14e-06 [interleave_parallel_branches]: 1.07e-06 [overlap_opt_shard_in_pipeline]: 1.10001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.96e-06 [control_data_broadcast_order]: 1.037e-05 [grouped_pairwise_exchange_alltoall]: 1.51002e-06 [offloading_packed_experts]: 3.35e-06 [overlap_recompute_and_grad_model_parallel]: 4.03001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.19e-06 [overlap_recompute_allgather_and_fa_grad]: 1.54998e-06 [overlap_recompute_comm]: 2.22999e-06 [overlap_grad_ring_attention]: 3.16001e-06 [overlap_grad_flash_sp]: 1.798e-05 [begin_end_overlap_inline]: 5.10016e-07 [split_matmul_comm_elemetwise]: 2.59001e-06 [split_layernorm_comm]: 1.76e-06 [handle_group_info]: 1.03001e-06 [symbol_engine_optimizer]: 6.284e-05, [1] [Cycle 1]: 5.872e-05, [6] [build]: 3.35e-06 [elim_shapecalc]: 6.78e-06 [elim_not_effective]: 9.77001e-06 [opt_reshape]: 4.40999e-06 [fold_const_symbol]: 6.94001e-06 [renormalize]: 1.80007e-07 [detach_backward]: 2.32999e-06 [pipeline_parallel_scheduler]: 1.54e-06 [auto_monad_reorder]: 1.575e-05 [get_jit_bprop_graph]: 1.55001e-06 [rewriter_after_jit_bprop_graph]: 5.28002e-06 [opt_after_jit_grad]: 0.00053122 [validate]: 3.963e-05 Sums bootstrap : 0.000447s : 0.18% type_inference : 0.234658s : 94.62% event_method : 0.000094s : 0.04% auto_monad : 0.000242s : 0.10% graph_reusing : 0.000014s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000071s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000016s : 0.01% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000097s : 0.04% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000009s : 0.00% optimize.rewriter_before_opt_a : 0.000489s : 0.20% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000259s : 0.10% optimize.opt_a.loop_unroll : 0.000096s : 0.04% optimize.opt_a.a_1 : 0.002213s : 0.89% optimize.opt_a.with_stream_mark : 0.000050s : 0.02% optimize.opt_a.recompute_prepare : 0.000030s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000246s : 0.10% optimize.opt_a.accelerated_algorithm : 0.000020s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.00% optimize.opt_a.shard_inline : 0.000016s : 0.01% optimize.opt_a.merge_send_recv : 0.000026s : 0.01% optimize.opt_a.auto_parallel : 0.000023s : 0.01% optimize.opt_a.parallel : 0.000029s : 0.01% optimize.opt_a.flash_sp : 0.000016s : 0.01% optimize.opt_a.merge_comm : 0.000011s : 0.00% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000040s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000023s : 0.01% optimize.opt_a.virtual_dataset : 0.000017s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000018s : 0.01% optimize.opt_a.virtual_output : 0.000016s : 0.01% optimize.opt_a.merge_forward : 0.000012s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000040s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000033s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.00% optimize.opt_a.meta_fg_expand : 0.002483s : 1.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000079s : 0.03% optimize.opt_a.a_after_grad : 0.000095s : 0.04% optimize.opt_a.renormalize : 0.003416s : 1.38% optimize.opt_a.add_forward_monad_depend : 0.000011s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000024s : 0.01% optimize.opt_a.cse : 0.000047s : 0.02% optimize.opt_a.a_3 : 0.000069s : 0.03% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000019s : 0.01% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000735s : 0.30% optimize.opt_b.b_1 : 0.000074s : 0.03% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000018s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000018s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000031s : 0.01% optimize.loop_unroll : 0.000479s : 0.19% optimize.opt_after_cconv.c_1 : 0.000016s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000016s : 0.01% optimize.tuple_transform.d_1 : 0.000026s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000038s : 0.02% optimize.cse_after_recomputation.cse : 0.000008s : 0.00% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 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 : 0.000002s : 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 : 0.000010s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000018s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000007s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000010s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000004s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 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 : 0.000016s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000531s : 0.21% validate : 0.000040s : 0.02% Time group info: ------[substitution.] 0.000693 102 1.65% : 0.000011s : 1: substitution.cast_eliminate 0.35% : 0.000002s : 1: substitution.elim_not_effective 0.91% : 0.000006s : 9: substitution.float_depend_g_call 0.58% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.17% : 0.000001s : 1: substitution.fold_const_symbol 0.79% : 0.000005s : 1: substitution.graph_param_transform 0.44% : 0.000003s : 2: substitution.incorporate_call 0.34% : 0.000002s : 2: substitution.incorporate_call_switch 71.76% : 0.000498s : 24: substitution.inline 3.09% : 0.000021s : 2: substitution.inline_without_move 1.07% : 0.000007s : 9: substitution.j_node_and_user_rematch 0.51% : 0.000004s : 2: substitution.minmaximum_grad 0.93% : 0.000006s : 9: substitution.partial_eliminate 1.20% : 0.000008s : 9: substitution.remove_not_recompute_node 0.32% : 0.000002s : 1: substitution.replace_applicator 0.78% : 0.000005s : 5: substitution.replace_old_param 0.48% : 0.000003s : 1: substitution.set_cell_output_no_recompute 2.78% : 0.000019s : 7: substitution.switch_simplify 1.09% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.47% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.71% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 8.96% : 0.000062s : 6: substitution.tuple_list_get_item_eliminator 0.64% : 0.000004s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.234024 2 97.37% : 0.227880s : 1: type_inference.infer 2.63% : 0.006144s : 1: type_inference.specialize ------[replace.] 0.000277 34 2.28% : 0.000006s : 1: replace.cast_eliminate 59.77% : 0.000166s : 24: replace.inline 31.93% : 0.000089s : 7: replace.switch_simplify 6.02% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000561 34 1.90% : 0.000011s : 1: match.cast_eliminate 86.59% : 0.000486s : 24: match.inline 2.65% : 0.000015s : 7: match.switch_simplify 8.87% : 0.000050s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000412 2554 1.20% : 0.000005s : 35: predicate.accumulaten_eliminater 0.41% : 0.000002s : 1: predicate.ad_related_special_op_eliminate 0.41% : 0.000002s : 12: predicate.addn_check_dump 1.24% : 0.000005s : 35: predicate.addn_zero_filter 1.10% : 0.000005s : 35: predicate.adjust_all_reduce_mul_add 2.18% : 0.000009s : 47: predicate.arithmetic_simplify 1.38% : 0.000006s : 36: predicate.cast_eliminate 0.17% : 0.000001s : 2: predicate.check_bprop_eliminate 0.51% : 0.000002s : 12: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.46% : 0.000002s : 12: predicate.depend_value_elim 1.25% : 0.000005s : 36: predicate.dict_get_item_const_eliminator 1.35% : 0.000006s : 36: predicate.dict_get_item_eliminator 1.16% : 0.000005s : 36: predicate.dict_set_item_eliminator 0.35% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 1: predicate.elim_not_effective 0.09% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.28% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.17% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.19% : 0.000005s : 37: predicate.environ_get_depend_swap 1.73% : 0.000007s : 49: predicate.environ_get_eliminate 1.20% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.10% : 0.000009s : 62: predicate.exchange_switch_depend_value 3.07% : 0.000013s : 62: predicate.float_depend_g_call 0.44% : 0.000002s : 12: predicate.float_environ_get_switch 0.52% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.52% : 0.000002s : 12: predicate.get_grad_eliminate 0.05% : 0.000000s : 1: predicate.graph_param_transform 0.40% : 0.000002s : 12: predicate.incorporate_call 0.36% : 0.000001s : 12: predicate.incorporate_call_switch 6.13% : 0.000025s : 123: predicate.inline 1.68% : 0.000007s : 33: predicate.inline_without_move 0.24% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.71% : 0.000003s : 12: predicate.less_batch_normalization 1.59% : 0.000007s : 40: predicate.list_to_tuple_eliminator_ 2.36% : 0.000010s : 75: predicate.load_eliminater 0.40% : 0.000002s : 1: predicate.loop_unroll_after_grad 3.89% : 0.000016s : 108: predicate.loop_unroll_before_grad 1.35% : 0.000006s : 38: predicate.make_slice_get_slice_eliminator 0.46% : 0.000002s : 12: predicate.merge_addn 0.22% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.14% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.10% : 0.000005s : 35: predicate.minmaximum_grad 0.50% : 0.000002s : 1: predicate.mutable_eliminate 0.09% : 0.000000s : 1: predicate.opt_reshape 0.08% : 0.000000s : 1: predicate.parallel_virtual_node 2.79% : 0.000012s : 62: predicate.partial_defer_inline 1.48% : 0.000006s : 39: predicate.partial_eliminate 1.18% : 0.000005s : 35: predicate.print_const_string_wrapper 0.49% : 0.000002s : 12: predicate.reduce_all_const_elim 1.59% : 0.000007s : 35: predicate.reduce_eliminate 2.46% : 0.000010s : 75: predicate.redundant_stop_gradient_eliminater 0.34% : 0.000001s : 12: predicate.remove_not_recompute_node 1.17% : 0.000005s : 40: predicate.replace_applicator 0.83% : 0.000003s : 33: predicate.replace_old_param 0.06% : 0.000000s : 1: predicate.reset_defer_inline 1.18% : 0.000005s : 35: predicate.reshape_eliminate 0.14% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.10% : 0.000000s : 1: predicate.row_tensor_eliminate 0.27% : 0.000001s : 2: predicate.same_eliminate 0.29% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.63% : 0.000003s : 12: predicate.shard_identity_eliminate 0.20% : 0.000001s : 2: predicate.special_op_eliminate 0.48% : 0.000002s : 12: predicate.specialize_transform 0.38% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 1.62% : 0.000007s : 33: predicate.stack_unstack_eliminate 0.09% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.36% : 0.000010s : 62: predicate.switch_defer_inline 2.54% : 0.000010s : 64: predicate.switch_layer_defer_inline 7.67% : 0.000032s : 197: predicate.switch_simplify 1.18% : 0.000005s : 35: predicate.tile_eliminate 1.43% : 0.000006s : 35: predicate.transpose_eliminate 1.57% : 0.000006s : 38: predicate.tuple_list_convert_item_index_to_positive 1.48% : 0.000006s : 38: predicate.tuple_list_get_item_const_eliminator 1.35% : 0.000006s : 38: predicate.tuple_list_get_item_depend_reorder 2.79% : 0.000011s : 52: predicate.tuple_list_get_item_eliminator 1.51% : 0.000006s : 38: predicate.tuple_list_get_set_item_eliminator 1.96% : 0.000008s : 50: predicate.tuple_list_set_item_eliminator 1.33% : 0.000005s : 40: predicate.tuple_to_list_eliminator_ 2.39% : 0.000010s : 75: predicate.updatestate_pure_node_eliminater 2.94% : 0.000012s : 87: predicate.updatestate_useless_node_eliminater 0.08% : 0.000000s : 1: predicate.value_based_eliminate 0.64% : 0.000003s : 12: predicate.virtual_dataset_eliminate 0.51% : 0.000002s : 12: predicate.virtual_output_eliminate 0.07% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.13% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.049358 46 97.54% : 0.048145s : 19: func_graph_cloner_run.FuncGraphClonerGraph 2.46% : 0.001213s : 27: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.325490 192 0.00% : 0.000004s : 1: ForceFp32Comm 8.77% : 0.028558s : 1: add_attr 8.77% : 0.028543s : 1: add_attr_with_inline 0.00% : 0.000005s : 1: add_comm_op_reuse_tag 0.01% : 0.000042s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.08% : 0.000256s : 1: auto_monad 0.01% : 0.000020s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.15% : 0.000482s : 1: bootstrap 0.01% : 0.000035s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000013s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000022s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.00% : 0.000008s : 1: environ_conv 0.03% : 0.000107s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000018s : 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% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.15% : 0.000489s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.23% : 0.000747s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.00% : 0.000011s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000014s : 1: opt.transform.mutable_eliminate 0.97% : 0.003164s : 78: opt.transform.opt_a 0.00% : 0.000014s : 1: opt.transform.opt_after_cconv 0.01% : 0.000019s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000049s : 28: opt.transform.opt_b 0.01% : 0.000028s : 2: opt.transform.opt_trans_graph 0.01% : 0.000025s : 4: opt.transform.symbol_engine_opt 3.12% : 0.010162s : 1: opt_a 0.03% : 0.000088s : 1: opt_after_cconv 0.17% : 0.000544s : 1: opt_after_jit_grad 0.05% : 0.000155s : 1: opt_b 3.91% : 0.012719s : 1: optimize 0.01% : 0.000021s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000021s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.03% : 0.000102s : 1: pre_auto_parallel 0.00% : 0.000013s : 1: py_interpret_to_execute 0.00% : 0.000010s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000020s : 1: remove_dup_value 0.80% : 0.002599s : 1: renormalize.infer 0.25% : 0.000804s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000023s : 1: rewriter_after_opt_a 0.15% : 0.000499s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000004s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000065s : 1: symbol_engine_optimizer 0.02% : 0.000058s : 1: tuple_transform 72.11% : 0.234697s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:44:58.479.370 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.158074, [21] [bootstrap]: 0.00044545 [type_inference]: 0.144661 [event_method]: 0.00034483 [auto_monad]: 0.00015718 [graph_reusing]: 9.97001e-06 [inline]: 3.23998e-06 [add_attr]: 0.00397082, [1] [add_attr_with_inline]: 0.0039583, [1] [Cycle 1]: 9.968e-05, [2] [tag_attr]: 4.636e-05 [meta_addattr_fg_expand]: 1.07e-05 [parallel-infer-symbol]: 4.03001e-06 [pre_auto_parallel]: 6.342e-05 [insert-virtual-dataset]: 2.50002e-06 [parallel-infer-symbol-second]: 7.10017e-07 [dataset_repeat_opt]: 1.92999e-06 [pipeline_split]: 1.68002e-06 [optimize]: 0.00754647, [53] [py_interpret_to_execute]: 7.05e-06 [rewriter_before_opt_a]: 0.00033667 [opt_a]: 0.0048571, [2] [Cycle 1]: 0.00425108, [45] [expand_dump_flag]: 5.10001e-06 [switch_simplify]: 0.00016411 [loop_unroll]: 5.721e-05 [a_1]: 0.00125672 [with_stream_mark]: 2.099e-05 [recompute_prepare]: 1.07e-05 [updatestate_depend_eliminate]: 4.03001e-06 [updatestate_assign_eliminate]: 3.13e-06 [updatestate_loads_eliminate]: 2.76e-06 [parameter_eliminate]: 2.32999e-06 [a_2]: 7.311e-05 [accelerated_algorithm]: 6.82002e-06 [shard]: 1.62999e-06 [meta_shard_fg_expand]: 2.36998e-06 [shard_inline]: 6.05002e-06 [merge_send_recv]: 9.52999e-06 [auto_parallel]: 7.44002e-06 [parallel]: 2.015e-05 [flash_sp]: 9.54999e-06 [merge_comm]: 3.75e-06 [allreduce_fusion]: 3.14999e-06 [matmul_add_comm_reduction]: 9.98998e-06 [allreduce_slice_to_reducescatter]: 7.89994e-07 [virtual_shard_identity]: 8.05e-06 [virtual_dataset]: 6.24001e-06 [get_grad_eliminate_]: 5.76e-06 [virtual_output]: 5.94999e-06 [merge_forward]: 4.08001e-06 [cell_reuse_recompute_pass]: 1.12e-06 [offload_activation]: 9.70002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.308e-05 [merge_recompute_call_nodes]: 1.47001e-06 [before_grad]: 9.59999e-06 [set_forward_comm_id_for_comm_node_pass]: 3.31999e-06 [meta_fg_expand]: 4.19002e-06 [flash_sp_send_recv_attached]: 2.37001e-06 [receive_attached]: 2.04e-06 [after_resolve]: 9.47999e-06 [a_after_grad]: 8.56002e-06 [renormalize]: 0.00208237 [add_forward_monad_depend]: 8.96002e-06 [auto_monad_grad]: 2.31e-06 [auto_monad_eliminator]: 1.872e-05 [cse]: 3.249e-05 [a_3]: 4.959e-05 [Cycle 2]: 0.00059194, [45] [expand_dump_flag]: 2.28002e-06 [switch_simplify]: 7.53e-06 [loop_unroll]: 5.36002e-06 [a_1]: 0.0001099 [with_stream_mark]: 1.535e-05 [recompute_prepare]: 5.74e-06 [updatestate_depend_eliminate]: 3.16999e-06 [updatestate_assign_eliminate]: 2.85002e-06 [updatestate_loads_eliminate]: 2.76e-06 [parameter_eliminate]: 1.19e-06 [a_2]: 5.943e-05 [accelerated_algorithm]: 5.19e-06 [shard]: 1.57999e-06 [meta_shard_fg_expand]: 1.72001e-06 [shard_inline]: 5.15001e-06 [merge_send_recv]: 6.34001e-06 [auto_parallel]: 7.05e-06 [parallel]: 7.05e-06 [flash_sp]: 3.94002e-06 [merge_comm]: 3.18e-06 [allreduce_fusion]: 2.74999e-06 [matmul_add_comm_reduction]: 7.28e-06 [allreduce_slice_to_reducescatter]: 3.9002e-07 [virtual_shard_identity]: 6.17999e-06 [virtual_dataset]: 4.85999e-06 [get_grad_eliminate_]: 4.94e-06 [virtual_output]: 4.79e-06 [merge_forward]: 3.55e-06 [cell_reuse_recompute_pass]: 2.14999e-06 [offload_activation]: 7.82e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.361e-05 [merge_recompute_call_nodes]: 1.17999e-06 [before_grad]: 8.75999e-06 [set_forward_comm_id_for_comm_node_pass]: 3.89002e-06 [meta_fg_expand]: 2.24001e-06 [flash_sp_send_recv_attached]: 1.44e-06 [receive_attached]: 1.89999e-06 [after_resolve]: 9.05001e-06 [a_after_grad]: 8.38001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.19998e-06 [auto_monad_grad]: 1.18001e-06 [auto_monad_eliminator]: 5.49e-06 [cse]: 1.305e-05 [a_3]: 2.938e-05 [py_interpret_to_execute_after_opt_a]: 8.45001e-06 [slice_cell_reuse_recomputed_activation]: 2.22999e-06 [rewriter_after_opt_a]: 1.87e-05 [convert_after_rewriter]: 1.99e-06 [order_py_execute_after_rewriter]: 1.14e-06 [mutable_eliminate]: 0.000755 [opt_b]: 0.00019753, [1] [Cycle 1]: 0.00018925, [7] [b_1]: 0.00011228 [b_2]: 7.56999e-06 [updatestate_depend_eliminate]: 7.13998e-06 [updatestate_assign_eliminate]: 2.22001e-06 [updatestate_loads_eliminate]: 2.71e-06 [renormalize]: 7.50006e-07 [cse]: 2.081e-05 [optimize_parallel_all_gather_comm]: 2.001e-05 [overlap_param_gather]: 1.84e-06 [cconv]: 3.071e-05 [loop_unroll]: 0.00052531 [opt_after_cconv]: 0.00010457, [1] [Cycle 1]: 9.654e-05, [7] [c_1]: 2.732e-05 [parameter_eliminate]: 4.3e-06 [updatestate_depend_eliminate]: 7.00998e-06 [updatestate_assign_eliminate]: 2.41e-06 [updatestate_loads_eliminate]: 1.92999e-06 [cse]: 1.926e-05 [renormalize]: 5.00004e-07 [remove_dup_value]: 1.448e-05 [tuple_transform]: 7.429e-05, [1] [Cycle 1]: 6.909e-05, [4] [d_1]: 4.114e-05 [none_parameter_eliminate]: 1.59998e-06 [renormalize]: 2.3999e-07 [switch_simplify]: 6.27001e-06 [partial_unused_args_eliminate]: 2.11998e-06 [add_recomputation]: 4.89e-05 [cse_after_recomputation]: 2.193e-05, [1] [Cycle 1]: 1.731e-05, [1] [cse]: 1.149e-05 [environ_conv]: 5.87001e-06 [swap_dp_allreduce_reducescatter]: 4.75999e-06 [bias_add_comm_swap]: 3.23998e-06 [label_micro_interleaved_index]: 5.70001e-06 [label_fine_grained_interleaved_index]: 2.83e-06 [merge_cast_opt]: 1.28002e-06 [slice_recompute_activation]: 2.11e-06 [micro_interleaved_order_control]: 2.26e-06 [assign_add_opt]: 1.45001e-06 [ForceFp32Comm]: 8.49977e-07 [remove_cast_before_assign_add]: 1.08001e-06 [full_micro_interleaved_order_control]: 2.19001e-06 [reorder_send_recv_between_fp_bp]: 2.47001e-06 [comm_op_add_attrs]: 1.08001e-06 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.17e-06 [interleave_parallel_branches]: 1.17999e-06 [overlap_opt_shard_in_pipeline]: 1.64998e-06 [overlap_opt_shard_grad_in_pipeline]: 1.72999e-06 [control_data_broadcast_order]: 1.507e-05 [grouped_pairwise_exchange_alltoall]: 1.57999e-06 [offloading_packed_experts]: 3.86001e-06 [overlap_recompute_and_grad_model_parallel]: 4.23999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.40999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.38002e-06 [overlap_recompute_comm]: 2.83998e-06 [overlap_grad_ring_attention]: 4.18999e-06 [overlap_grad_flash_sp]: 2.108e-05 [begin_end_overlap_inline]: 6.50005e-07 [split_matmul_comm_elemetwise]: 2.54999e-06 [split_layernorm_comm]: 1.54e-06 [handle_group_info]: 1.64e-06 [symbol_engine_optimizer]: 0.00016576, [1] [Cycle 1]: 0.00016008, [6] [build]: 4.3e-06 [elim_shapecalc]: 1.2e-05 [elim_not_effective]: 1.418e-05 [opt_reshape]: 6.78e-06 [fold_const_symbol]: 8.413e-05 [renormalize]: 3.00002e-07 [detach_backward]: 2.57001e-06 [pipeline_parallel_scheduler]: 1.79e-06 [auto_monad_reorder]: 2.066e-05 [get_jit_bprop_graph]: 2.18002e-06 [rewriter_after_jit_bprop_graph]: 7.55e-06 [opt_after_jit_grad]: 0.00058447 [validate]: 4.824e-05 Sums bootstrap : 0.000445s : 0.29% type_inference : 0.144661s : 94.50% event_method : 0.000345s : 0.23% auto_monad : 0.000157s : 0.10% graph_reusing : 0.000010s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000046s : 0.03% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000011s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000063s : 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 : 0.000007s : 0.00% optimize.rewriter_before_opt_a : 0.000337s : 0.22% optimize.opt_a.expand_dump_flag : 0.000007s : 0.00% optimize.opt_a.switch_simplify : 0.000172s : 0.11% optimize.opt_a.loop_unroll : 0.000063s : 0.04% optimize.opt_a.a_1 : 0.001367s : 0.89% optimize.opt_a.with_stream_mark : 0.000036s : 0.02% optimize.opt_a.recompute_prepare : 0.000016s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000133s : 0.09% optimize.opt_a.accelerated_algorithm : 0.000012s : 0.01% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000004s : 0.00% optimize.opt_a.shard_inline : 0.000011s : 0.01% optimize.opt_a.merge_send_recv : 0.000016s : 0.01% optimize.opt_a.auto_parallel : 0.000014s : 0.01% optimize.opt_a.parallel : 0.000027s : 0.02% optimize.opt_a.flash_sp : 0.000013s : 0.01% optimize.opt_a.merge_comm : 0.000007s : 0.00% optimize.opt_a.allreduce_fusion : 0.000006s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000017s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000014s : 0.01% optimize.opt_a.virtual_dataset : 0.000011s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.01% optimize.opt_a.virtual_output : 0.000011s : 0.01% optimize.opt_a.merge_forward : 0.000008s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000018s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000027s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000018s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.00% optimize.opt_a.meta_fg_expand : 0.000006s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000019s : 0.01% optimize.opt_a.a_after_grad : 0.000017s : 0.01% optimize.opt_a.renormalize : 0.002082s : 1.36% optimize.opt_a.add_forward_monad_depend : 0.000010s : 0.01% optimize.opt_a.auto_monad_grad : 0.000003s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000024s : 0.02% optimize.opt_a.cse : 0.000046s : 0.03% optimize.opt_a.a_3 : 0.000079s : 0.05% optimize.py_interpret_to_execute_after_opt_a : 0.000008s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000019s : 0.01% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000755s : 0.49% optimize.opt_b.b_1 : 0.000112s : 0.07% optimize.opt_b.b_2 : 0.000008s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000021s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000031s : 0.02% optimize.loop_unroll : 0.000525s : 0.34% optimize.opt_after_cconv.c_1 : 0.000027s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000019s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000041s : 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 : 0.000006s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000049s : 0.03% optimize.cse_after_recomputation.cse : 0.000011s : 0.01% optimize.environ_conv : 0.000006s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000002s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000015s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000021s : 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 : 0.000002s : 0.00% optimize.symbol_engine_optimizer.build : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000012s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000084s : 0.05% 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 : 0.000021s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000584s : 0.38% validate : 0.000048s : 0.03% Time group info: ------[substitution.] 0.000375 39 3.95% : 0.000015s : 1: substitution.cast_eliminate 0.49% : 0.000002s : 2: substitution.elim_not_effective 0.45% : 0.000002s : 2: substitution.fold_const_symbol 1.56% : 0.000006s : 3: substitution.graph_param_transform 83.70% : 0.000314s : 15: substitution.inline 1.05% : 0.000004s : 4: substitution.j_node_and_user_rematch 1.37% : 0.000005s : 4: substitution.remove_not_recompute_node 1.12% : 0.000004s : 2: substitution.replace_old_param 3.68% : 0.000014s : 4: substitution.switch_simplify 2.63% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.144543 2 86.40% : 0.124882s : 1: type_inference.infer 13.60% : 0.019661s : 1: type_inference.specialize ------[replace.] 0.000177 22 3.76% : 0.000007s : 1: replace.cast_eliminate 55.75% : 0.000099s : 15: replace.inline 31.95% : 0.000057s : 4: replace.switch_simplify 8.54% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000340 22 4.03% : 0.000014s : 1: match.cast_eliminate 90.04% : 0.000306s : 15: match.inline 3.34% : 0.000011s : 4: match.switch_simplify 2.59% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000288 1716 1.13% : 0.000003s : 22: predicate.accumulaten_eliminater 0.86% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.29% : 0.000001s : 6: predicate.addn_check_dump 1.05% : 0.000003s : 22: predicate.addn_zero_filter 0.96% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.03% : 0.000006s : 28: predicate.arithmetic_simplify 1.36% : 0.000004s : 23: predicate.cast_eliminate 0.39% : 0.000001s : 6: predicate.check_bprop_eliminate 0.31% : 0.000001s : 6: predicate.compare_switch_simplify 0.09% : 0.000000s : 3: predicate.const_output_eliminate 0.48% : 0.000001s : 6: predicate.depend_value_elim 1.17% : 0.000003s : 23: predicate.dict_get_item_const_eliminator 1.40% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.11% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.79% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.18% : 0.000001s : 3: predicate.elim_not_effective 0.31% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.27% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.33% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.24% : 0.000004s : 26: predicate.environ_get_depend_swap 1.67% : 0.000005s : 32: predicate.environ_get_eliminate 1.24% : 0.000004s : 26: predicate.environ_get_set_eliminate 1.93% : 0.000006s : 40: predicate.exchange_switch_depend_value 2.85% : 0.000008s : 40: predicate.float_depend_g_call 0.29% : 0.000001s : 6: predicate.float_environ_get_switch 0.57% : 0.000002s : 9: predicate.float_tuple_getitem_switch 0.15% : 0.000000s : 3: predicate.fold_const_symbol 0.41% : 0.000001s : 6: predicate.get_grad_eliminate 0.10% : 0.000000s : 3: predicate.graph_param_transform 0.32% : 0.000001s : 6: predicate.incorporate_call 0.28% : 0.000001s : 6: predicate.incorporate_call_switch 6.04% : 0.000017s : 80: predicate.inline 0.43% : 0.000001s : 6: predicate.inline_without_move 0.18% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.63% : 0.000002s : 6: predicate.less_batch_normalization 1.85% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.50% : 0.000007s : 53: predicate.load_eliminater 1.10% : 0.000003s : 3: predicate.loop_unroll_after_grad 3.42% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.76% : 0.000005s : 29: predicate.make_slice_get_slice_eliminator 0.30% : 0.000001s : 6: predicate.merge_addn 0.40% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.41% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.04% : 0.000003s : 22: predicate.minmaximum_grad 0.88% : 0.000003s : 3: predicate.mutable_eliminate 0.23% : 0.000001s : 3: predicate.opt_reshape 0.27% : 0.000001s : 3: predicate.parallel_virtual_node 2.79% : 0.000008s : 40: predicate.partial_defer_inline 1.51% : 0.000004s : 28: predicate.partial_eliminate 1.05% : 0.000003s : 22: predicate.print_const_string_wrapper 0.37% : 0.000001s : 6: predicate.reduce_all_const_elim 1.66% : 0.000005s : 22: predicate.reduce_eliminate 2.63% : 0.000008s : 53: predicate.redundant_stop_gradient_eliminater 0.28% : 0.000001s : 6: predicate.remove_not_recompute_node 1.19% : 0.000003s : 31: predicate.replace_applicator 0.29% : 0.000001s : 6: predicate.replace_old_param 0.15% : 0.000000s : 3: predicate.reset_defer_inline 1.11% : 0.000003s : 22: predicate.reshape_eliminate 0.38% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.25% : 0.000001s : 3: predicate.row_tensor_eliminate 0.48% : 0.000001s : 6: predicate.same_eliminate 0.22% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.56% : 0.000002s : 6: predicate.shard_identity_eliminate 0.43% : 0.000001s : 6: predicate.special_op_eliminate 0.40% : 0.000001s : 6: predicate.specialize_transform 0.60% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.50% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.24% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.20% : 0.000006s : 40: predicate.switch_defer_inline 2.47% : 0.000007s : 46: predicate.switch_layer_defer_inline 6.89% : 0.000020s : 126: predicate.switch_simplify 1.17% : 0.000003s : 22: predicate.tile_eliminate 1.18% : 0.000003s : 22: predicate.transpose_eliminate 1.67% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.63% : 0.000005s : 29: predicate.tuple_list_get_item_const_eliminator 1.59% : 0.000005s : 29: predicate.tuple_list_get_item_depend_reorder 2.71% : 0.000008s : 37: predicate.tuple_list_get_item_eliminator 1.74% : 0.000005s : 29: predicate.tuple_list_get_set_item_eliminator 2.24% : 0.000006s : 35: predicate.tuple_list_set_item_eliminator 1.59% : 0.000005s : 31: predicate.tuple_to_list_eliminator_ 2.47% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 2.91% : 0.000008s : 59: predicate.updatestate_useless_node_eliminater 0.23% : 0.000001s : 3: predicate.value_based_eliminate 0.41% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.41% : 0.000001s : 6: predicate.virtual_output_eliminate 0.15% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.28% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002010 27 49.85% : 0.001002s : 10: func_graph_cloner_run.FuncGraphClonerGraph 50.15% : 0.001008s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.173738 192 0.00% : 0.000006s : 1: ForceFp32Comm 2.29% : 0.003977s : 1: add_attr 2.28% : 0.003963s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.03% : 0.000054s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.10% : 0.000168s : 1: auto_monad 0.01% : 0.000025s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.27% : 0.000473s : 1: bootstrap 0.02% : 0.000034s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.01% : 0.000019s : 1: control_data_broadcast_order 0.00% : 0.000005s : 1: convert_after_rewriter 0.01% : 0.000025s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000010s : 1: environ_conv 0.21% : 0.000359s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000014s : 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% : 0.000005s : 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 0.01% : 0.000010s : 1: label_micro_interleaved_index 0.31% : 0.000538s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.44% : 0.000767s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000019s : 1: opt.transform.mutable_eliminate 1.09% : 0.001891s : 78: opt.transform.opt_a 0.01% : 0.000026s : 1: opt.transform.opt_after_cconv 0.02% : 0.000028s : 1: opt.transform.opt_after_jit_grad 0.05% : 0.000090s : 28: opt.transform.opt_b 0.03% : 0.000045s : 2: opt.transform.opt_trans_graph 0.06% : 0.000112s : 4: opt.transform.symbol_engine_opt 2.80% : 0.004860s : 1: opt_a 0.06% : 0.000108s : 1: opt_after_cconv 0.34% : 0.000599s : 1: opt_after_jit_grad 0.12% : 0.000201s : 1: opt_b 4.35% : 0.007553s : 1: optimize 0.01% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000025s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000004s : 1: pipeline_split 0.04% : 0.000068s : 1: pre_auto_parallel 0.01% : 0.000010s : 1: py_interpret_to_execute 0.01% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000018s : 1: remove_dup_value 0.65% : 0.001129s : 1: renormalize.infer 0.54% : 0.000941s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000022s : 1: rewriter_after_opt_a 0.20% : 0.000344s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000006s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.10% : 0.000169s : 1: symbol_engine_optimizer 0.04% : 0.000077s : 1: tuple_transform 83.28% : 0.144688s : 1: type_inference . TotalTime = 0.210943, [21] [bootstrap]: 0.00043921 [type_inference]: 0.177707 [event_method]: 8.022e-05 [auto_monad]: 0.00022498 [graph_reusing]: 1.257e-05 [inline]: 2.87002e-06 [add_attr]: 0.00421693, [1] [add_attr_with_inline]: 0.00420387, [1] [Cycle 1]: 0.00013622, [2] [tag_attr]: 7.263e-05 [meta_addattr_fg_expand]: 1.631e-05 [parallel-infer-symbol]: 3.85e-06 [pre_auto_parallel]: 9.668e-05 [insert-virtual-dataset]: 3.13e-06 [parallel-infer-symbol-second]: 1.55001e-06 [dataset_repeat_opt]: 2.16e-06 [pipeline_split]: 1.48002e-06 [optimize]: 0.0270405, [53] [py_interpret_to_execute]: 8.42e-06 [rewriter_before_opt_a]: 0.00049985 [opt_a]: 0.0240326, [2] [Cycle 1]: 0.0230604, [45] [expand_dump_flag]: 7.10998e-06 [switch_simplify]: 0.00026853 [loop_unroll]: 9.395e-05 [a_1]: 0.00236525 [with_stream_mark]: 3.797e-05 [recompute_prepare]: 3.103e-05 [updatestate_depend_eliminate]: 8.79003e-06 [updatestate_assign_eliminate]: 6.73998e-06 [updatestate_loads_eliminate]: 6.68e-06 [parameter_eliminate]: 3.27002e-06 [a_2]: 0.00021105 [accelerated_algorithm]: 1.456e-05 [shard]: 2.09e-06 [meta_shard_fg_expand]: 7.02002e-06 [shard_inline]: 1.39e-05 [merge_send_recv]: 2.056e-05 [auto_parallel]: 1.496e-05 [parallel]: 2.132e-05 [flash_sp]: 1.47e-05 [merge_comm]: 9.30001e-06 [allreduce_fusion]: 7.51999e-06 [matmul_add_comm_reduction]: 3.481e-05 [allreduce_slice_to_reducescatter]: 1.07998e-06 [virtual_shard_identity]: 2.031e-05 [virtual_dataset]: 1.425e-05 [get_grad_eliminate_]: 1.279e-05 [virtual_output]: 1.296e-05 [merge_forward]: 9.64e-06 [cell_reuse_recompute_pass]: 1.74998e-06 [offload_activation]: 1.992e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.994e-05 [merge_recompute_call_nodes]: 2.09e-06 [before_grad]: 2.676e-05 [set_forward_comm_id_for_comm_node_pass]: 1.13e-05 [meta_fg_expand]: 0.00276261 [flash_sp_send_recv_attached]: 7.63999e-06 [receive_attached]: 2.48e-06 [after_resolve]: 8.192e-05 [a_after_grad]: 9.181e-05 [renormalize]: 0.0161845 [add_forward_monad_depend]: 1.344e-05 [auto_monad_grad]: 3.02002e-06 [auto_monad_eliminator]: 2.231e-05 [cse]: 3.548e-05 [a_3]: 5.427e-05 [Cycle 2]: 0.00095266, [45] [expand_dump_flag]: 2.71999e-06 [switch_simplify]: 6.91001e-06 [loop_unroll]: 4.35999e-06 [a_1]: 5.843e-05 [with_stream_mark]: 2.029e-05 [recompute_prepare]: 5.15999e-06 [updatestate_depend_eliminate]: 3.71001e-06 [updatestate_assign_eliminate]: 2.53e-06 [updatestate_loads_eliminate]: 2.98e-06 [parameter_eliminate]: 1.96e-06 [a_2]: 4.237e-05 [accelerated_algorithm]: 4.48999e-06 [shard]: 3.18e-06 [meta_shard_fg_expand]: 3.14001e-06 [shard_inline]: 3.78999e-06 [merge_send_recv]: 7.82002e-06 [auto_parallel]: 9.55001e-06 [parallel]: 9.53002e-06 [flash_sp]: 5.14998e-06 [merge_comm]: 2.43998e-06 [allreduce_fusion]: 2.41998e-06 [matmul_add_comm_reduction]: 9.01002e-06 [allreduce_slice_to_reducescatter]: 7.60017e-07 [virtual_shard_identity]: 6.66e-06 [virtual_dataset]: 4.27e-06 [get_grad_eliminate_]: 3.75e-06 [virtual_output]: 4.28999e-06 [merge_forward]: 4.57e-06 [cell_reuse_recompute_pass]: 3.66001e-06 [offload_activation]: 9.02999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.51e-05 [merge_recompute_call_nodes]: 1.41998e-06 [before_grad]: 8.47e-06 [set_forward_comm_id_for_comm_node_pass]: 3.68e-06 [meta_fg_expand]: 0.00035605 [flash_sp_send_recv_attached]: 3.03e-06 [receive_attached]: 2.66999e-06 [after_resolve]: 1.25e-05 [a_after_grad]: 6.31e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 4.07e-06 [auto_monad_grad]: 2.09e-06 [auto_monad_eliminator]: 1.304e-05 [cse]: 2.168e-05 [a_3]: 2.508e-05 [py_interpret_to_execute_after_opt_a]: 9.98998e-06 [slice_cell_reuse_recomputed_activation]: 2.89999e-06 [rewriter_after_opt_a]: 2.218e-05 [convert_after_rewriter]: 1.57001e-06 [order_py_execute_after_rewriter]: 1.28002e-06 [mutable_eliminate]: 0.00083013 [opt_b]: 0.00017923, [1] [Cycle 1]: 0.00016792, [7] [b_1]: 7.941e-05 [b_2]: 6.65002e-06 [updatestate_depend_eliminate]: 9.10001e-06 [updatestate_assign_eliminate]: 2.60002e-06 [updatestate_loads_eliminate]: 2.47001e-06 [renormalize]: 1.89e-06 [cse]: 2.813e-05 [optimize_parallel_all_gather_comm]: 2.243e-05 [overlap_param_gather]: 2.12999e-06 [cconv]: 3.984e-05 [loop_unroll]: 0.0006582 [opt_after_cconv]: 0.00010852, [1] [Cycle 1]: 0.0001, [7] [c_1]: 1.782e-05 [parameter_eliminate]: 6.59999e-06 [updatestate_depend_eliminate]: 7.62998e-06 [updatestate_assign_eliminate]: 2.04e-06 [updatestate_loads_eliminate]: 1.96e-06 [cse]: 2.564e-05 [renormalize]: 9.00007e-07 [remove_dup_value]: 1.591e-05 [tuple_transform]: 6.423e-05, [1] [Cycle 1]: 5.843e-05, [4] [d_1]: 3.179e-05 [none_parameter_eliminate]: 1.85001e-06 [renormalize]: 1.90019e-07 [switch_simplify]: 4.97e-06 [partial_unused_args_eliminate]: 2.50002e-06 [add_recomputation]: 4.266e-05 [cse_after_recomputation]: 2.37e-05, [1] [Cycle 1]: 1.83e-05, [1] [cse]: 1.046e-05 [environ_conv]: 6.57002e-06 [swap_dp_allreduce_reducescatter]: 5.05999e-06 [bias_add_comm_swap]: 3.33e-06 [label_micro_interleaved_index]: 6.14001e-06 [label_fine_grained_interleaved_index]: 2.73998e-06 [merge_cast_opt]: 1.71e-06 [slice_recompute_activation]: 2.48e-06 [micro_interleaved_order_control]: 2.09e-06 [assign_add_opt]: 1.46002e-06 [ForceFp32Comm]: 8.39995e-07 [remove_cast_before_assign_add]: 1.04e-06 [full_micro_interleaved_order_control]: 2.59001e-06 [reorder_send_recv_between_fp_bp]: 2.71e-06 [comm_op_add_attrs]: 1.04998e-06 [add_comm_op_reuse_tag]: 9.80013e-07 [interleave_split_concat_branches]: 1.25999e-06 [interleave_parallel_branches]: 1.03001e-06 [overlap_opt_shard_in_pipeline]: 1.52999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.55001e-06 [control_data_broadcast_order]: 1.462e-05 [grouped_pairwise_exchange_alltoall]: 1.96e-06 [offloading_packed_experts]: 3.98999e-06 [overlap_recompute_and_grad_model_parallel]: 4.53999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.20999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.35999e-06 [overlap_recompute_comm]: 2.43002e-06 [overlap_grad_ring_attention]: 3.46999e-06 [overlap_grad_flash_sp]: 2.146e-05 [begin_end_overlap_inline]: 4.7998e-07 [split_matmul_comm_elemetwise]: 2.55002e-06 [split_layernorm_comm]: 2.14999e-06 [handle_group_info]: 1.02e-06 [symbol_engine_optimizer]: 8.408e-05, [1] [Cycle 1]: 7.641e-05, [6] [build]: 3.71001e-06 [elim_shapecalc]: 1.299e-05 [elim_not_effective]: 1.067e-05 [opt_reshape]: 5.25999e-06 [fold_const_symbol]: 6.28998e-06 [renormalize]: 1.69995e-07 [detach_backward]: 2.49999e-06 [pipeline_parallel_scheduler]: 2.26998e-06 [auto_monad_reorder]: 1.9e-05 [get_jit_bprop_graph]: 1.99e-06 [rewriter_after_jit_bprop_graph]: 8.30999e-06 [opt_after_jit_grad]: 0.00064718 [validate]: 0.00020584 Sums bootstrap : 0.000439s : 0.21% type_inference : 0.177707s : 86.50% event_method : 0.000080s : 0.04% auto_monad : 0.000225s : 0.11% graph_reusing : 0.000013s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000073s : 0.04% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000016s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000097s : 0.05% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000001s : 0.00% optimize.py_interpret_to_execute : 0.000008s : 0.00% optimize.rewriter_before_opt_a : 0.000500s : 0.24% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000275s : 0.13% optimize.opt_a.loop_unroll : 0.000098s : 0.05% optimize.opt_a.a_1 : 0.002424s : 1.18% optimize.opt_a.with_stream_mark : 0.000058s : 0.03% optimize.opt_a.recompute_prepare : 0.000036s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000253s : 0.12% optimize.opt_a.accelerated_algorithm : 0.000019s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000018s : 0.01% optimize.opt_a.merge_send_recv : 0.000028s : 0.01% optimize.opt_a.auto_parallel : 0.000025s : 0.01% optimize.opt_a.parallel : 0.000031s : 0.02% optimize.opt_a.flash_sp : 0.000020s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000044s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000027s : 0.01% optimize.opt_a.virtual_dataset : 0.000019s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000017s : 0.01% optimize.opt_a.virtual_output : 0.000017s : 0.01% optimize.opt_a.merge_forward : 0.000014s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000029s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000045s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000035s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000015s : 0.01% optimize.opt_a.meta_fg_expand : 0.003119s : 1.52% optimize.opt_a.flash_sp_send_recv_attached : 0.000011s : 0.01% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000094s : 0.05% optimize.opt_a.a_after_grad : 0.000098s : 0.05% optimize.opt_a.renormalize : 0.016185s : 7.88% optimize.opt_a.add_forward_monad_depend : 0.000018s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000035s : 0.02% optimize.opt_a.cse : 0.000057s : 0.03% optimize.opt_a.a_3 : 0.000079s : 0.04% optimize.py_interpret_to_execute_after_opt_a : 0.000010s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000022s : 0.01% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000830s : 0.40% optimize.opt_b.b_1 : 0.000079s : 0.04% optimize.opt_b.b_2 : 0.000007s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000002s : 0.00% optimize.opt_b.cse : 0.000028s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000040s : 0.02% optimize.loop_unroll : 0.000658s : 0.32% optimize.opt_after_cconv.c_1 : 0.000018s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000026s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000016s : 0.01% optimize.tuple_transform.d_1 : 0.000032s : 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 : 0.000005s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000043s : 0.02% optimize.cse_after_recomputation.cse : 0.000010s : 0.01% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 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 : 0.000002s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000015s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000021s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000013s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000006s : 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 : 0.000019s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000008s : 0.00% opt_after_jit_grad : 0.000647s : 0.32% validate : 0.000206s : 0.10% Time group info: ------[substitution.] 0.000833 102 1.39% : 0.000012s : 1: substitution.cast_eliminate 0.16% : 0.000001s : 1: substitution.elim_not_effective 0.98% : 0.000008s : 9: substitution.float_depend_g_call 0.58% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.12% : 0.000001s : 1: substitution.fold_const_symbol 0.55% : 0.000005s : 1: substitution.graph_param_transform 0.44% : 0.000004s : 2: substitution.incorporate_call 0.31% : 0.000003s : 2: substitution.incorporate_call_switch 78.60% : 0.000655s : 24: substitution.inline 3.31% : 0.000028s : 2: substitution.inline_without_move 1.02% : 0.000008s : 9: substitution.j_node_and_user_rematch 0.51% : 0.000004s : 2: substitution.minmaximum_grad 0.90% : 0.000007s : 9: substitution.partial_eliminate 1.02% : 0.000009s : 9: substitution.remove_not_recompute_node 0.45% : 0.000004s : 1: substitution.replace_applicator 0.73% : 0.000006s : 5: substitution.replace_old_param 0.59% : 0.000005s : 1: substitution.set_cell_output_no_recompute 2.22% : 0.000018s : 7: substitution.switch_simplify 1.09% : 0.000009s : 2: substitution.tuple_list_convert_item_index_to_positive 0.40% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.64% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 3.35% : 0.000028s : 6: substitution.tuple_list_get_item_eliminator 0.64% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.177162 2 87.58% : 0.155152s : 1: type_inference.infer 12.42% : 0.022010s : 1: type_inference.specialize ------[replace.] 0.000310 34 3.14% : 0.000010s : 1: replace.cast_eliminate 61.86% : 0.000192s : 24: replace.inline 29.72% : 0.000092s : 7: replace.switch_simplify 5.28% : 0.000016s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000676 34 1.56% : 0.000011s : 1: match.cast_eliminate 94.62% : 0.000640s : 24: match.inline 2.17% : 0.000015s : 7: match.switch_simplify 1.65% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000438 2554 1.20% : 0.000005s : 35: predicate.accumulaten_eliminater 0.58% : 0.000003s : 1: predicate.ad_related_special_op_eliminate 0.38% : 0.000002s : 12: predicate.addn_check_dump 1.10% : 0.000005s : 35: predicate.addn_zero_filter 1.03% : 0.000005s : 35: predicate.adjust_all_reduce_mul_add 2.07% : 0.000009s : 47: predicate.arithmetic_simplify 1.34% : 0.000006s : 36: predicate.cast_eliminate 0.13% : 0.000001s : 2: predicate.check_bprop_eliminate 0.37% : 0.000002s : 12: predicate.compare_switch_simplify 0.03% : 0.000000s : 1: predicate.const_output_eliminate 0.43% : 0.000002s : 12: predicate.depend_value_elim 1.23% : 0.000005s : 36: predicate.dict_get_item_const_eliminator 1.29% : 0.000006s : 36: predicate.dict_get_item_eliminator 1.11% : 0.000005s : 36: predicate.dict_set_item_eliminator 0.59% : 0.000003s : 2: predicate.dumpgradient_eliminate 0.08% : 0.000000s : 1: predicate.elim_not_effective 0.36% : 0.000002s : 1: predicate.elim_shapecalc_of_broadcastargs 1.19% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.18% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.14% : 0.000005s : 37: predicate.environ_get_depend_swap 1.58% : 0.000007s : 49: predicate.environ_get_eliminate 1.20% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.04% : 0.000009s : 62: predicate.exchange_switch_depend_value 3.38% : 0.000015s : 62: predicate.float_depend_g_call 0.41% : 0.000002s : 12: predicate.float_environ_get_switch 0.51% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.03% : 0.000000s : 1: predicate.fold_const_symbol 0.47% : 0.000002s : 12: predicate.get_grad_eliminate 0.09% : 0.000000s : 1: predicate.graph_param_transform 0.43% : 0.000002s : 12: predicate.incorporate_call 0.33% : 0.000001s : 12: predicate.incorporate_call_switch 6.22% : 0.000027s : 123: predicate.inline 1.85% : 0.000008s : 33: predicate.inline_without_move 0.21% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.57% : 0.000002s : 12: predicate.less_batch_normalization 1.42% : 0.000006s : 40: predicate.list_to_tuple_eliminator_ 2.37% : 0.000010s : 75: predicate.load_eliminater 0.77% : 0.000003s : 1: predicate.loop_unroll_after_grad 3.74% : 0.000016s : 108: predicate.loop_unroll_before_grad 1.43% : 0.000006s : 38: predicate.make_slice_get_slice_eliminator 0.50% : 0.000002s : 12: predicate.merge_addn 0.16% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.15% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.00% : 0.000004s : 35: predicate.minmaximum_grad 0.87% : 0.000004s : 1: predicate.mutable_eliminate 0.10% : 0.000000s : 1: predicate.opt_reshape 0.16% : 0.000001s : 1: predicate.parallel_virtual_node 3.07% : 0.000013s : 62: predicate.partial_defer_inline 1.40% : 0.000006s : 39: predicate.partial_eliminate 1.14% : 0.000005s : 35: predicate.print_const_string_wrapper 0.41% : 0.000002s : 12: predicate.reduce_all_const_elim 1.51% : 0.000007s : 35: predicate.reduce_eliminate 2.35% : 0.000010s : 75: predicate.redundant_stop_gradient_eliminater 0.33% : 0.000001s : 12: predicate.remove_not_recompute_node 1.12% : 0.000005s : 40: predicate.replace_applicator 0.83% : 0.000004s : 33: predicate.replace_old_param 0.14% : 0.000001s : 1: predicate.reset_defer_inline 1.15% : 0.000005s : 35: predicate.reshape_eliminate 0.14% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.09% : 0.000000s : 1: predicate.row_tensor_eliminate 0.24% : 0.000001s : 2: predicate.same_eliminate 0.33% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.71% : 0.000003s : 12: predicate.shard_identity_eliminate 0.16% : 0.000001s : 2: predicate.special_op_eliminate 0.57% : 0.000003s : 12: predicate.specialize_transform 0.39% : 0.000002s : 2: predicate.split_environ_get_set_with_tuple_value 1.86% : 0.000008s : 33: predicate.stack_unstack_eliminate 0.09% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.19% : 0.000010s : 62: predicate.switch_defer_inline 2.31% : 0.000010s : 64: predicate.switch_layer_defer_inline 7.90% : 0.000035s : 197: predicate.switch_simplify 1.21% : 0.000005s : 35: predicate.tile_eliminate 1.17% : 0.000005s : 35: predicate.transpose_eliminate 1.39% : 0.000006s : 38: predicate.tuple_list_convert_item_index_to_positive 1.40% : 0.000006s : 38: predicate.tuple_list_get_item_const_eliminator 1.33% : 0.000006s : 38: predicate.tuple_list_get_item_depend_reorder 3.02% : 0.000013s : 52: predicate.tuple_list_get_item_eliminator 1.50% : 0.000007s : 38: predicate.tuple_list_get_set_item_eliminator 1.83% : 0.000008s : 50: predicate.tuple_list_set_item_eliminator 1.35% : 0.000006s : 40: predicate.tuple_to_list_eliminator_ 2.21% : 0.000010s : 75: predicate.updatestate_pure_node_eliminater 2.82% : 0.000012s : 87: predicate.updatestate_useless_node_eliminater 0.21% : 0.000001s : 1: predicate.value_based_eliminate 0.62% : 0.000003s : 12: predicate.virtual_dataset_eliminate 0.50% : 0.000002s : 12: predicate.virtual_output_eliminate 0.05% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.11% : 0.000000s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003509 46 61.74% : 0.002166s : 19: func_graph_cloner_run.FuncGraphClonerGraph 38.26% : 0.001343s : 27: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.261680 192 0.00% : 0.000004s : 1: ForceFp32Comm 1.61% : 0.004224s : 1: add_attr 1.61% : 0.004209s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000048s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.09% : 0.000235s : 1: auto_monad 0.01% : 0.000024s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.18% : 0.000476s : 1: bootstrap 0.02% : 0.000045s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.01% : 0.000018s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000027s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000011s : 1: environ_conv 0.04% : 0.000092s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000017s : 1: graph_reusing 0.00% : 0.000007s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000005s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000010s : 1: label_micro_interleaved_index 0.26% : 0.000673s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.33% : 0.000852s : 1: mutable_eliminate 0.00% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000019s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000023s : 1: opt.transform.mutable_eliminate 1.31% : 0.003435s : 78: opt.transform.opt_a 0.01% : 0.000017s : 1: opt.transform.opt_after_cconv 0.01% : 0.000027s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000055s : 28: opt.transform.opt_b 0.01% : 0.000034s : 2: opt.transform.opt_trans_graph 0.01% : 0.000030s : 4: opt.transform.symbol_engine_opt 9.19% : 0.024037s : 1: opt_a 0.04% : 0.000113s : 1: opt_after_cconv 0.25% : 0.000666s : 1: opt_after_jit_grad 0.07% : 0.000183s : 1: opt_b 10.34% : 0.027047s : 1: optimize 0.01% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000025s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000007s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 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% : 0.000006s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.04% : 0.000103s : 1: pre_auto_parallel 0.00% : 0.000013s : 1: py_interpret_to_execute 0.01% : 0.000013s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000020s : 1: remove_dup_value 5.77% : 0.015111s : 1: renormalize.infer 0.40% : 0.001049s : 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% : 0.000026s : 1: rewriter_after_opt_a 0.20% : 0.000513s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000006s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.03% : 0.000087s : 1: symbol_engine_optimizer 0.03% : 0.000068s : 1: tuple_transform 67.93% : 0.177746s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:45:02.552.104 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.145823, [21] [bootstrap]: 0.00048059 [type_inference]: 0.126622 [event_method]: 0.00011472 [auto_monad]: 0.0002305 [graph_reusing]: 1.319e-05 [inline]: 3.25e-06 [add_attr]: 0.00434836, [1] [add_attr_with_inline]: 0.00433364, [1] [Cycle 1]: 0.00013381, [2] [tag_attr]: 6.698e-05 [meta_addattr_fg_expand]: 1.71e-05 [parallel-infer-symbol]: 4.32998e-06 [pre_auto_parallel]: 9.867e-05 [insert-virtual-dataset]: 3.2e-06 [parallel-infer-symbol-second]: 8.39995e-07 [dataset_repeat_opt]: 2.02001e-06 [pipeline_split]: 1.84998e-06 [optimize]: 0.0128463, [53] [py_interpret_to_execute]: 9.66e-06 [rewriter_before_opt_a]: 0.00049035 [opt_a]: 0.00914553, [2] [Cycle 1]: 0.00824257, [45] [expand_dump_flag]: 1.152e-05 [switch_simplify]: 0.0012602 [loop_unroll]: 0.00010666 [a_1]: 0.0021038 [with_stream_mark]: 3.009e-05 [recompute_prepare]: 1.671e-05 [updatestate_depend_eliminate]: 6.34001e-06 [updatestate_assign_eliminate]: 4.80001e-06 [updatestate_loads_eliminate]: 4.22998e-06 [parameter_eliminate]: 2.73998e-06 [a_2]: 0.00011807 [accelerated_algorithm]: 1.032e-05 [shard]: 2.21e-06 [meta_shard_fg_expand]: 8.84998e-06 [shard_inline]: 8.92e-06 [merge_send_recv]: 1.099e-05 [auto_parallel]: 1.144e-05 [parallel]: 2.252e-05 [flash_sp]: 1.293e-05 [merge_comm]: 5.86998e-06 [allreduce_fusion]: 5.30999e-06 [matmul_add_comm_reduction]: 1.198e-05 [allreduce_slice_to_reducescatter]: 9.20001e-07 [virtual_shard_identity]: 1.192e-05 [virtual_dataset]: 8.52998e-06 [get_grad_eliminate_]: 8.59e-06 [virtual_output]: 9.04998e-06 [merge_forward]: 5.27001e-06 [cell_reuse_recompute_pass]: 2.22999e-06 [offload_activation]: 1.338e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.066e-05 [merge_recompute_call_nodes]: 1.59998e-06 [before_grad]: 1.591e-05 [set_forward_comm_id_for_comm_node_pass]: 6.49001e-06 [meta_fg_expand]: 6.31e-06 [flash_sp_send_recv_attached]: 2.64999e-06 [receive_attached]: 2.95002e-06 [after_resolve]: 1.461e-05 [a_after_grad]: 1.359e-05 [renormalize]: 0.00377408 [add_forward_monad_depend]: 1.232e-05 [auto_monad_grad]: 3.16001e-06 [auto_monad_eliminator]: 2.852e-05 [cse]: 5.307e-05 [a_3]: 7.958e-05 [Cycle 2]: 0.00088407, [45] [expand_dump_flag]: 2.59001e-06 [switch_simplify]: 1.148e-05 [loop_unroll]: 8.32998e-06 [a_1]: 0.0002094 [with_stream_mark]: 2.077e-05 [recompute_prepare]: 8.74e-06 [updatestate_depend_eliminate]: 5.34e-06 [updatestate_assign_eliminate]: 4.27003e-06 [updatestate_loads_eliminate]: 4.72e-06 [parameter_eliminate]: 2.74001e-06 [a_2]: 9.842e-05 [accelerated_algorithm]: 8.18001e-06 [shard]: 2.27999e-06 [meta_shard_fg_expand]: 2.98e-06 [shard_inline]: 8.07e-06 [merge_send_recv]: 1.025e-05 [auto_parallel]: 1.028e-05 [parallel]: 1.001e-05 [flash_sp]: 3.56999e-06 [merge_comm]: 4.58001e-06 [allreduce_fusion]: 4.52e-06 [matmul_add_comm_reduction]: 1.146e-05 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 9.60001e-06 [virtual_dataset]: 8.32998e-06 [get_grad_eliminate_]: 7.22002e-06 [virtual_output]: 7.7e-06 [merge_forward]: 5.49e-06 [cell_reuse_recompute_pass]: 3.26001e-06 [offload_activation]: 1.267e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.691e-05 [merge_recompute_call_nodes]: 1.31998e-06 [before_grad]: 1.375e-05 [set_forward_comm_id_for_comm_node_pass]: 5.70001e-06 [meta_fg_expand]: 3.95e-06 [flash_sp_send_recv_attached]: 1.89e-06 [receive_attached]: 2.83e-06 [after_resolve]: 1.334e-05 [a_after_grad]: 1.178e-05 [renormalize]: 6.99947e-08 [add_forward_monad_depend]: 2.01998e-06 [auto_monad_grad]: 1.60001e-06 [auto_monad_eliminator]: 1.331e-05 [cse]: 3.199e-05 [a_3]: 4.983e-05 [py_interpret_to_execute_after_opt_a]: 8.85001e-06 [slice_cell_reuse_recomputed_activation]: 2.34999e-06 [rewriter_after_opt_a]: 3.303e-05 [convert_after_rewriter]: 1.35001e-06 [order_py_execute_after_rewriter]: 1.35999e-06 [mutable_eliminate]: 0.00085073 [opt_b]: 0.00063004, [1] [Cycle 1]: 0.00061927, [7] [b_1]: 0.0004803 [b_2]: 1.462e-05 [updatestate_depend_eliminate]: 1.253e-05 [updatestate_assign_eliminate]: 4.55999e-06 [updatestate_loads_eliminate]: 3.95998e-06 [renormalize]: 5.3001e-07 [cse]: 4.901e-05 [optimize_parallel_all_gather_comm]: 2.434e-05 [overlap_param_gather]: 1.97999e-06 [cconv]: 3.811e-05 [loop_unroll]: 0.00070638 [opt_after_cconv]: 0.00015495, [1] [Cycle 1]: 0.00014667, [7] [c_1]: 4.263e-05 [parameter_eliminate]: 6.74001e-06 [updatestate_depend_eliminate]: 9.74e-06 [updatestate_assign_eliminate]: 3.61999e-06 [updatestate_loads_eliminate]: 5.40999e-06 [cse]: 4.339e-05 [renormalize]: 6.49976e-07 [remove_dup_value]: 5.362e-05 [tuple_transform]: 9.765e-05, [1] [Cycle 1]: 9.222e-05, [4] [d_1]: 6.226e-05 [none_parameter_eliminate]: 2.00002e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 9.01002e-06 [partial_unused_args_eliminate]: 1.89999e-06 [add_recomputation]: 6.948e-05 [cse_after_recomputation]: 3.384e-05, [1] [Cycle 1]: 2.716e-05, [1] [cse]: 2.17e-05 [environ_conv]: 8.84e-06 [swap_dp_allreduce_reducescatter]: 7.26999e-06 [bias_add_comm_swap]: 3.8e-06 [label_micro_interleaved_index]: 6.99001e-06 [label_fine_grained_interleaved_index]: 2.57001e-06 [merge_cast_opt]: 1.66002e-06 [slice_recompute_activation]: 2.56998e-06 [micro_interleaved_order_control]: 2.32999e-06 [assign_add_opt]: 1.61002e-06 [ForceFp32Comm]: 1.12e-06 [remove_cast_before_assign_add]: 1.31002e-06 [full_micro_interleaved_order_control]: 2.21e-06 [reorder_send_recv_between_fp_bp]: 2.64999e-06 [comm_op_add_attrs]: 1.57001e-06 [add_comm_op_reuse_tag]: 1.05999e-06 [interleave_split_concat_branches]: 1.15001e-06 [interleave_parallel_branches]: 1.08001e-06 [overlap_opt_shard_in_pipeline]: 1.27e-06 [overlap_opt_shard_grad_in_pipeline]: 1.82001e-06 [control_data_broadcast_order]: 1.814e-05 [grouped_pairwise_exchange_alltoall]: 1.87999e-06 [offloading_packed_experts]: 4.91002e-06 [overlap_recompute_and_grad_model_parallel]: 7.57002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.19998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.38002e-06 [overlap_recompute_comm]: 2.46998e-06 [overlap_grad_ring_attention]: 5.07e-06 [overlap_grad_flash_sp]: 2.78e-05 [begin_end_overlap_inline]: 5.49975e-07 [split_matmul_comm_elemetwise]: 2.09999e-06 [split_layernorm_comm]: 1.74e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 9.467e-05, [1] [Cycle 1]: 8.972e-05, [6] [build]: 4.64002e-06 [elim_shapecalc]: 1.433e-05 [elim_not_effective]: 1.687e-05 [opt_reshape]: 9.52001e-06 [fold_const_symbol]: 1.405e-05 [renormalize]: 1.80007e-07 [detach_backward]: 2.93e-06 [pipeline_parallel_scheduler]: 1.71e-06 [auto_monad_reorder]: 2.144e-05 [get_jit_bprop_graph]: 2.07999e-06 [rewriter_after_jit_bprop_graph]: 5.23002e-06 [opt_after_jit_grad]: 0.00074991 [validate]: 5.702e-05 Sums bootstrap : 0.000481s : 0.34% type_inference : 0.126622s : 90.25% event_method : 0.000115s : 0.08% auto_monad : 0.000230s : 0.16% graph_reusing : 0.000013s : 0.01% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000067s : 0.05% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000017s : 0.01% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000099s : 0.07% 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 : 0.000010s : 0.01% optimize.rewriter_before_opt_a : 0.000490s : 0.35% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.001272s : 0.91% optimize.opt_a.loop_unroll : 0.000115s : 0.08% optimize.opt_a.a_1 : 0.002313s : 1.65% optimize.opt_a.with_stream_mark : 0.000051s : 0.04% optimize.opt_a.recompute_prepare : 0.000025s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000216s : 0.15% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000012s : 0.01% optimize.opt_a.shard_inline : 0.000017s : 0.01% optimize.opt_a.merge_send_recv : 0.000021s : 0.02% optimize.opt_a.auto_parallel : 0.000022s : 0.02% optimize.opt_a.parallel : 0.000033s : 0.02% optimize.opt_a.flash_sp : 0.000017s : 0.01% optimize.opt_a.merge_comm : 0.000010s : 0.01% optimize.opt_a.allreduce_fusion : 0.000010s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000023s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000022s : 0.02% optimize.opt_a.virtual_dataset : 0.000017s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.01% optimize.opt_a.virtual_output : 0.000017s : 0.01% optimize.opt_a.merge_forward : 0.000011s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000038s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000030s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.01% optimize.opt_a.meta_fg_expand : 0.000010s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000028s : 0.02% optimize.opt_a.a_after_grad : 0.000025s : 0.02% optimize.opt_a.renormalize : 0.003774s : 2.69% optimize.opt_a.add_forward_monad_depend : 0.000014s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000042s : 0.03% optimize.opt_a.cse : 0.000085s : 0.06% optimize.opt_a.a_3 : 0.000129s : 0.09% optimize.py_interpret_to_execute_after_opt_a : 0.000009s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000033s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000851s : 0.61% optimize.opt_b.b_1 : 0.000480s : 0.34% optimize.opt_b.b_2 : 0.000015s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000013s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000049s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000024s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000038s : 0.03% optimize.loop_unroll : 0.000706s : 0.50% optimize.opt_after_cconv.c_1 : 0.000043s : 0.03% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.cse : 0.000043s : 0.03% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000054s : 0.04% optimize.tuple_transform.d_1 : 0.000062s : 0.04% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000009s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000069s : 0.05% optimize.cse_after_recomputation.cse : 0.000022s : 0.02% optimize.environ_conv : 0.000009s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000007s : 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 : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000003s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000018s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000008s : 0.01% 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000028s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000017s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 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 : 0.000021s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000750s : 0.53% validate : 0.000057s : 0.04% Time group info: ------[substitution.] 0.000824 75 7.36% : 0.000061s : 11: substitution.cast_eliminate 0.44% : 0.000004s : 4: substitution.elim_not_effective 0.23% : 0.000002s : 4: substitution.fold_const_symbol 1.08% : 0.000009s : 5: substitution.graph_param_transform 82.98% : 0.000684s : 24: substitution.inline 0.75% : 0.000006s : 8: substitution.j_node_and_user_rematch 0.95% : 0.000008s : 8: substitution.remove_not_recompute_node 0.76% : 0.000006s : 2: substitution.replace_old_param 3.98% : 0.000033s : 7: substitution.switch_simplify 1.47% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.126448 2 91.64% : 0.115879s : 1: type_inference.infer 8.36% : 0.010569s : 1: type_inference.specialize ------[replace.] 0.000340 33 54.35% : 0.000185s : 24: replace.inline 39.69% : 0.000135s : 7: replace.switch_simplify 5.97% : 0.000020s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000708 33 94.71% : 0.000670s : 24: match.inline 3.82% : 0.000027s : 7: match.switch_simplify 1.47% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000470 2544 1.21% : 0.000006s : 32: predicate.accumulaten_eliminater 0.56% : 0.000003s : 5: predicate.ad_related_special_op_eliminate 0.34% : 0.000002s : 10: predicate.addn_check_dump 1.16% : 0.000005s : 32: predicate.addn_zero_filter 1.03% : 0.000005s : 32: predicate.adjust_all_reduce_mul_add 2.09% : 0.000010s : 42: predicate.arithmetic_simplify 1.75% : 0.000008s : 32: predicate.cast_eliminate 0.43% : 0.000002s : 10: predicate.check_bprop_eliminate 0.34% : 0.000002s : 10: predicate.compare_switch_simplify 0.10% : 0.000000s : 5: predicate.const_output_eliminate 0.36% : 0.000002s : 10: predicate.depend_value_elim 1.13% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.25% : 0.000006s : 32: predicate.dict_get_item_eliminator 1.01% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.64% : 0.000003s : 10: predicate.dumpgradient_eliminate 0.15% : 0.000001s : 5: predicate.elim_not_effective 0.22% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.28% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.15% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.09% : 0.000005s : 37: predicate.environ_get_depend_swap 1.62% : 0.000008s : 47: predicate.environ_get_eliminate 1.13% : 0.000005s : 37: predicate.environ_get_set_eliminate 1.82% : 0.000009s : 58: predicate.exchange_switch_depend_value 2.65% : 0.000012s : 58: predicate.float_depend_g_call 0.31% : 0.000001s : 10: predicate.float_environ_get_switch 0.50% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.09% : 0.000000s : 5: predicate.fold_const_symbol 0.45% : 0.000002s : 10: predicate.get_grad_eliminate 0.15% : 0.000001s : 5: predicate.graph_param_transform 0.32% : 0.000002s : 10: predicate.incorporate_call 0.27% : 0.000001s : 10: predicate.incorporate_call_switch 5.80% : 0.000027s : 120: predicate.inline 0.36% : 0.000002s : 10: predicate.inline_without_move 0.17% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.62% : 0.000003s : 10: predicate.less_batch_normalization 1.57% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.23% : 0.000010s : 76: predicate.load_eliminater 0.76% : 0.000004s : 5: predicate.loop_unroll_after_grad 5.10% : 0.000024s : 104: predicate.loop_unroll_before_grad 1.64% : 0.000008s : 42: predicate.make_slice_get_slice_eliminator 0.32% : 0.000002s : 10: predicate.merge_addn 0.40% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.33% : 0.000002s : 10: predicate.mini_step_allgather_replace 0.95% : 0.000004s : 32: predicate.minmaximum_grad 0.89% : 0.000004s : 5: predicate.mutable_eliminate 0.24% : 0.000001s : 5: predicate.opt_reshape 0.28% : 0.000001s : 5: predicate.parallel_virtual_node 2.63% : 0.000012s : 58: predicate.partial_defer_inline 1.27% : 0.000006s : 39: predicate.partial_eliminate 1.18% : 0.000006s : 32: predicate.print_const_string_wrapper 0.33% : 0.000002s : 10: predicate.reduce_all_const_elim 1.42% : 0.000007s : 32: predicate.reduce_eliminate 2.44% : 0.000011s : 76: predicate.redundant_stop_gradient_eliminater 0.26% : 0.000001s : 10: predicate.remove_not_recompute_node 1.12% : 0.000005s : 44: predicate.replace_applicator 0.37% : 0.000002s : 10: predicate.replace_old_param 0.32% : 0.000001s : 5: predicate.reset_defer_inline 1.09% : 0.000005s : 32: predicate.reshape_eliminate 0.42% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.32% : 0.000002s : 5: predicate.row_tensor_eliminate 0.48% : 0.000002s : 10: predicate.same_eliminate 0.21% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.48% : 0.000002s : 10: predicate.shard_identity_eliminate 0.50% : 0.000002s : 10: predicate.special_op_eliminate 0.38% : 0.000002s : 10: predicate.specialize_transform 0.52% : 0.000002s : 10: predicate.split_environ_get_set_with_tuple_value 0.50% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.20% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.00% : 0.000009s : 58: predicate.switch_defer_inline 2.34% : 0.000011s : 68: predicate.switch_layer_defer_inline 9.78% : 0.000046s : 191: predicate.switch_simplify 1.21% : 0.000006s : 32: predicate.tile_eliminate 1.10% : 0.000005s : 32: predicate.transpose_eliminate 1.64% : 0.000008s : 42: predicate.tuple_list_convert_item_index_to_positive 1.47% : 0.000007s : 42: predicate.tuple_list_get_item_const_eliminator 1.43% : 0.000007s : 42: predicate.tuple_list_get_item_depend_reorder 2.47% : 0.000012s : 54: predicate.tuple_list_get_item_eliminator 1.54% : 0.000007s : 42: predicate.tuple_list_get_set_item_eliminator 2.25% : 0.000011s : 52: predicate.tuple_list_set_item_eliminator 1.47% : 0.000007s : 44: predicate.tuple_to_list_eliminator_ 2.27% : 0.000011s : 76: predicate.updatestate_pure_node_eliminater 2.98% : 0.000014s : 86: predicate.updatestate_useless_node_eliminater 0.19% : 0.000001s : 5: predicate.value_based_eliminate 0.39% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.39% : 0.000002s : 10: predicate.virtual_output_eliminate 0.15% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.23% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003809 42 48.76% : 0.001858s : 16: func_graph_cloner_run.FuncGraphClonerGraph 51.24% : 0.001952s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.171499 192 0.00% : 0.000004s : 1: ForceFp32Comm 2.54% : 0.004355s : 1: add_attr 2.53% : 0.004338s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.04% : 0.000074s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.14% : 0.000243s : 1: auto_monad 0.01% : 0.000026s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.30% : 0.000514s : 1: bootstrap 0.02% : 0.000042s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.02% : 0.000037s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000012s : 1: environ_conv 0.07% : 0.000126s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000018s : 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 0.01% : 0.000010s : 1: label_micro_interleaved_index 0.42% : 0.000720s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.51% : 0.000871s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.02% : 0.000026s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000027s : 1: opt.transform.mutable_eliminate 2.44% : 0.004189s : 78: opt.transform.opt_a 0.02% : 0.000041s : 1: opt.transform.opt_after_cconv 0.02% : 0.000037s : 1: opt.transform.opt_after_jit_grad 0.26% : 0.000454s : 28: opt.transform.opt_b 0.04% : 0.000069s : 2: opt.transform.opt_trans_graph 0.03% : 0.000051s : 4: opt.transform.symbol_engine_opt 5.34% : 0.009150s : 1: opt_a 0.09% : 0.000159s : 1: opt_after_cconv 0.45% : 0.000767s : 1: opt_after_jit_grad 0.37% : 0.000635s : 1: opt_b 7.49% : 0.012853s : 1: optimize 0.02% : 0.000028s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.02% : 0.000031s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000008s : 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% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.01% : 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% : 0.000005s : 1: pipeline_split 0.06% : 0.000104s : 1: pre_auto_parallel 0.01% : 0.000013s : 1: py_interpret_to_execute 0.01% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.03% : 0.000058s : 1: remove_dup_value 1.11% : 0.001900s : 1: renormalize.infer 1.08% : 0.001856s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000010s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000036s : 1: rewriter_after_opt_a 0.29% : 0.000503s : 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% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.06% : 0.000097s : 1: symbol_engine_optimizer 0.06% : 0.000102s : 1: tuple_transform 73.85% : 0.126654s : 1: type_inference . TotalTime = 0.115693, [21] [bootstrap]: 0.00045856 [type_inference]: 0.0910244 [event_method]: 7.783e-05 [auto_monad]: 0.00026915 [graph_reusing]: 2.254e-05 [inline]: 3.20002e-06 [add_attr]: 0.00401253, [1] [add_attr_with_inline]: 0.00400053, [1] [Cycle 1]: 0.00014289, [2] [tag_attr]: 8.056e-05 [meta_addattr_fg_expand]: 2.278e-05 [parallel-infer-symbol]: 3.52997e-06 [pre_auto_parallel]: 0.00010886 [insert-virtual-dataset]: 2.53e-06 [parallel-infer-symbol-second]: 1.99e-06 [dataset_repeat_opt]: 2.32001e-06 [pipeline_split]: 1.55001e-06 [optimize]: 0.0189968, [53] [py_interpret_to_execute]: 6.38e-06 [rewriter_before_opt_a]: 0.00063842 [opt_a]: 0.0163025, [2] [Cycle 1]: 0.0151737, [45] [expand_dump_flag]: 1.279e-05 [switch_simplify]: 0.00033939 [loop_unroll]: 0.00012936 [a_1]: 0.00302339 [with_stream_mark]: 3.724e-05 [recompute_prepare]: 2.944e-05 [updatestate_depend_eliminate]: 1.057e-05 [updatestate_assign_eliminate]: 8.64e-06 [updatestate_loads_eliminate]: 7.58999e-06 [parameter_eliminate]: 3.66999e-06 [a_2]: 0.00024705 [accelerated_algorithm]: 1.63e-05 [shard]: 1.95001e-06 [meta_shard_fg_expand]: 6.38e-06 [shard_inline]: 1.578e-05 [merge_send_recv]: 1.829e-05 [auto_parallel]: 1.391e-05 [parallel]: 1.938e-05 [flash_sp]: 1.269e-05 [merge_comm]: 9.80002e-06 [allreduce_fusion]: 8.81002e-06 [matmul_add_comm_reduction]: 3.232e-05 [allreduce_slice_to_reducescatter]: 8.59989e-07 [virtual_shard_identity]: 1.825e-05 [virtual_dataset]: 1.573e-05 [get_grad_eliminate_]: 1.516e-05 [virtual_output]: 1.548e-05 [merge_forward]: 9.79999e-06 [cell_reuse_recompute_pass]: 1.12e-06 [offload_activation]: 1.905e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.107e-05 [merge_recompute_call_nodes]: 1.36998e-06 [before_grad]: 2.869e-05 [set_forward_comm_id_for_comm_node_pass]: 1.031e-05 [meta_fg_expand]: 0.00317604 [flash_sp_send_recv_attached]: 4.52998e-06 [receive_attached]: 2.71e-06 [after_resolve]: 0.00011278 [a_after_grad]: 0.00015405 [renormalize]: 0.00709998 [add_forward_monad_depend]: 7.92e-06 [auto_monad_grad]: 2.42001e-06 [auto_monad_eliminator]: 1.471e-05 [cse]: 2.786e-05 [a_3]: 4.733e-05 [Cycle 2]: 0.00111227, [45] [expand_dump_flag]: 2.16003e-06 [switch_simplify]: 7.13e-06 [loop_unroll]: 4.26001e-06 [a_1]: 5.27e-05 [with_stream_mark]: 1.352e-05 [recompute_prepare]: 4.27e-06 [updatestate_depend_eliminate]: 2.93998e-06 [updatestate_assign_eliminate]: 2.54999e-06 [updatestate_loads_eliminate]: 2.64999e-06 [parameter_eliminate]: 1.78997e-06 [a_2]: 3.963e-05 [accelerated_algorithm]: 4.79998e-06 [shard]: 2.38002e-06 [meta_shard_fg_expand]: 1.98997e-06 [shard_inline]: 3.59002e-06 [merge_send_recv]: 6.87002e-06 [auto_parallel]: 7.73999e-06 [parallel]: 9.02e-06 [flash_sp]: 3.99002e-06 [merge_comm]: 2.26003e-06 [allreduce_fusion]: 2.24999e-06 [matmul_add_comm_reduction]: 7.48999e-06 [allreduce_slice_to_reducescatter]: 5.8001e-07 [virtual_shard_identity]: 4.72998e-06 [virtual_dataset]: 4e-06 [get_grad_eliminate_]: 3.4e-06 [virtual_output]: 3.7e-06 [merge_forward]: 4.25e-06 [cell_reuse_recompute_pass]: 2.62001e-06 [offload_activation]: 8.06001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.331e-05 [merge_recompute_call_nodes]: 1.89e-06 [before_grad]: 7.61001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.28e-06 [meta_fg_expand]: 0.00057994 [flash_sp_send_recv_attached]: 2.18002e-06 [receive_attached]: 2.78e-06 [after_resolve]: 1.144e-05 [a_after_grad]: 6.31998e-06 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 3.4e-06 [auto_monad_grad]: 1.92999e-06 [auto_monad_eliminator]: 9.14998e-06 [cse]: 1.566e-05 [a_3]: 2.505e-05 [py_interpret_to_execute_after_opt_a]: 7.00998e-06 [slice_cell_reuse_recomputed_activation]: 2.26e-06 [rewriter_after_opt_a]: 1.886e-05 [convert_after_rewriter]: 1.25999e-06 [order_py_execute_after_rewriter]: 1.29e-06 [mutable_eliminate]: 0.00070741 [opt_b]: 0.00014524, [1] [Cycle 1]: 0.00013693, [7] [b_1]: 7.347e-05 [b_2]: 4.85999e-06 [updatestate_depend_eliminate]: 4.81002e-06 [updatestate_assign_eliminate]: 2.22999e-06 [updatestate_loads_eliminate]: 1.76998e-06 [renormalize]: 5.3001e-07 [cse]: 1.523e-05 [optimize_parallel_all_gather_comm]: 1.561e-05 [overlap_param_gather]: 2.07999e-06 [cconv]: 2.841e-05 [loop_unroll]: 0.00043461 [opt_after_cconv]: 7.684e-05, [1] [Cycle 1]: 7.104e-05, [7] [c_1]: 1.476e-05 [parameter_eliminate]: 3.30003e-06 [updatestate_depend_eliminate]: 4.15e-06 [updatestate_assign_eliminate]: 1.64e-06 [updatestate_loads_eliminate]: 1.45001e-06 [cse]: 1.353e-05 [renormalize]: 4.10015e-07 [remove_dup_value]: 1.401e-05 [tuple_transform]: 5.101e-05, [1] [Cycle 1]: 4.626e-05, [4] [d_1]: 2.293e-05 [none_parameter_eliminate]: 1.39998e-06 [renormalize]: 1.39989e-07 [switch_simplify]: 4.23999e-06 [partial_unused_args_eliminate]: 1.59998e-06 [add_recomputation]: 3.121e-05 [cse_after_recomputation]: 1.808e-05, [1] [Cycle 1]: 1.362e-05, [1] [cse]: 8.37e-06 [environ_conv]: 5.15999e-06 [swap_dp_allreduce_reducescatter]: 2.439e-05 [bias_add_comm_swap]: 2.58998e-06 [label_micro_interleaved_index]: 4.59002e-06 [label_fine_grained_interleaved_index]: 2.48e-06 [merge_cast_opt]: 1.50999e-06 [slice_recompute_activation]: 1.96998e-06 [micro_interleaved_order_control]: 2.74001e-06 [assign_add_opt]: 1.39e-06 [ForceFp32Comm]: 6.99976e-07 [remove_cast_before_assign_add]: 9.20001e-07 [full_micro_interleaved_order_control]: 2.05002e-06 [reorder_send_recv_between_fp_bp]: 2.21e-06 [comm_op_add_attrs]: 1.05001e-06 [add_comm_op_reuse_tag]: 1.02998e-06 [interleave_split_concat_branches]: 1.07e-06 [interleave_parallel_branches]: 1.02e-06 [overlap_opt_shard_in_pipeline]: 1.17999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.71e-06 [control_data_broadcast_order]: 9.44e-06 [grouped_pairwise_exchange_alltoall]: 1.34e-06 [offloading_packed_experts]: 3.06999e-06 [overlap_recompute_and_grad_model_parallel]: 3.85e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.47999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.15999e-06 [overlap_recompute_comm]: 2.01e-06 [overlap_grad_ring_attention]: 3.05002e-06 [overlap_grad_flash_sp]: 1.541e-05 [begin_end_overlap_inline]: 4.89992e-07 [split_matmul_comm_elemetwise]: 2.47001e-06 [split_layernorm_comm]: 1.91e-06 [handle_group_info]: 1.18001e-06 [symbol_engine_optimizer]: 6.245e-05, [1] [Cycle 1]: 5.815e-05, [6] [build]: 2.84999e-06 [elim_shapecalc]: 8.41002e-06 [elim_not_effective]: 9.31002e-06 [opt_reshape]: 4.21001e-06 [fold_const_symbol]: 7.06001e-06 [renormalize]: 2.69996e-07 [detach_backward]: 2.25002e-06 [pipeline_parallel_scheduler]: 1.76998e-06 [auto_monad_reorder]: 1.6e-05 [get_jit_bprop_graph]: 1.92001e-06 [rewriter_after_jit_bprop_graph]: 5.37001e-06 [opt_after_jit_grad]: 0.00045748 [validate]: 3.44e-05 Sums bootstrap : 0.000459s : 0.41% type_inference : 0.091024s : 82.34% event_method : 0.000078s : 0.07% auto_monad : 0.000269s : 0.24% graph_reusing : 0.000023s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000081s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000023s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000109s : 0.10% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000638s : 0.58% optimize.opt_a.expand_dump_flag : 0.000015s : 0.01% optimize.opt_a.switch_simplify : 0.000347s : 0.31% optimize.opt_a.loop_unroll : 0.000134s : 0.12% optimize.opt_a.a_1 : 0.003076s : 2.78% optimize.opt_a.with_stream_mark : 0.000051s : 0.05% optimize.opt_a.recompute_prepare : 0.000034s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000014s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000287s : 0.26% optimize.opt_a.accelerated_algorithm : 0.000021s : 0.02% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.01% optimize.opt_a.shard_inline : 0.000019s : 0.02% optimize.opt_a.merge_send_recv : 0.000025s : 0.02% optimize.opt_a.auto_parallel : 0.000022s : 0.02% optimize.opt_a.parallel : 0.000028s : 0.03% optimize.opt_a.flash_sp : 0.000017s : 0.02% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000011s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000040s : 0.04% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000023s : 0.02% optimize.opt_a.virtual_dataset : 0.000020s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000019s : 0.02% optimize.opt_a.virtual_output : 0.000019s : 0.02% optimize.opt_a.merge_forward : 0.000014s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000027s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000044s : 0.04% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000036s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000014s : 0.01% optimize.opt_a.meta_fg_expand : 0.003756s : 3.40% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.01% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000124s : 0.11% optimize.opt_a.a_after_grad : 0.000160s : 0.15% optimize.opt_a.renormalize : 0.007100s : 6.42% optimize.opt_a.add_forward_monad_depend : 0.000011s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000024s : 0.02% optimize.opt_a.cse : 0.000044s : 0.04% optimize.opt_a.a_3 : 0.000072s : 0.07% optimize.py_interpret_to_execute_after_opt_a : 0.000007s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000019s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000707s : 0.64% optimize.opt_b.b_1 : 0.000073s : 0.07% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000015s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000016s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000028s : 0.03% optimize.loop_unroll : 0.000435s : 0.39% optimize.opt_after_cconv.c_1 : 0.000015s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000014s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.01% optimize.tuple_transform.d_1 : 0.000023s : 0.02% optimize.tuple_transform.none_parameter_eliminate : 0.000001s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000031s : 0.03% optimize.cse_after_recomputation.cse : 0.000008s : 0.01% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000024s : 0.02% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000002s : 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 : 0.000002s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000009s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000015s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000008s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000009s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000004s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 0.01% 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 : 0.000016s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000457s : 0.41% validate : 0.000034s : 0.03% Time group info: ------[substitution.] 0.000944 134 3.72% : 0.000035s : 8: substitution.cast_eliminate 0.30% : 0.000003s : 1: substitution.elim_not_effective 0.80% : 0.000008s : 9: substitution.float_depend_g_call 0.52% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.10% : 0.000001s : 1: substitution.fold_const_symbol 0.53% : 0.000005s : 1: substitution.graph_param_transform 0.33% : 0.000003s : 2: substitution.incorporate_call 0.21% : 0.000002s : 2: substitution.incorporate_call_switch 78.14% : 0.000737s : 33: substitution.inline 3.33% : 0.000031s : 4: substitution.inline_without_move 0.93% : 0.000009s : 11: substitution.j_node_and_user_rematch 0.38% : 0.000004s : 2: substitution.minmaximum_grad 0.70% : 0.000007s : 9: substitution.partial_eliminate 0.96% : 0.000009s : 11: substitution.remove_not_recompute_node 0.30% : 0.000003s : 1: substitution.replace_applicator 0.91% : 0.000009s : 12: substitution.replace_old_param 0.39% : 0.000004s : 1: substitution.set_cell_output_no_recompute 2.54% : 0.000024s : 10: substitution.switch_simplify 0.89% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.38% : 0.000004s : 2: substitution.tuple_list_get_item_const_eliminator 0.54% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.64% : 0.000025s : 6: substitution.tuple_list_get_item_eliminator 0.48% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.090793 2 92.76% : 0.084221s : 1: type_inference.infer 7.24% : 0.006572s : 1: type_inference.specialize ------[replace.] 0.000368 45 62.82% : 0.000231s : 33: replace.inline 32.35% : 0.000119s : 10: replace.switch_simplify 4.83% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000750 45 95.65% : 0.000718s : 33: match.inline 2.59% : 0.000019s : 10: match.switch_simplify 1.76% : 0.000013s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000545 3189 1.22% : 0.000007s : 43: predicate.accumulaten_eliminater 0.15% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.39% : 0.000002s : 14: predicate.addn_check_dump 1.16% : 0.000006s : 43: predicate.addn_zero_filter 1.09% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 2.18% : 0.000012s : 57: predicate.arithmetic_simplify 1.52% : 0.000008s : 43: predicate.cast_eliminate 0.13% : 0.000001s : 2: predicate.check_bprop_eliminate 0.41% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.38% : 0.000002s : 14: predicate.depend_value_elim 1.29% : 0.000007s : 43: predicate.dict_get_item_const_eliminator 1.39% : 0.000008s : 43: predicate.dict_get_item_eliminator 1.19% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.22% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.03% : 0.000000s : 1: predicate.elim_not_effective 0.08% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.27% : 0.000007s : 44: predicate.environ_add_const_eliminate 1.25% : 0.000007s : 44: predicate.environ_get_add_eliminate 1.16% : 0.000006s : 44: predicate.environ_get_depend_swap 1.60% : 0.000009s : 58: predicate.environ_get_eliminate 1.17% : 0.000006s : 44: predicate.environ_get_set_eliminate 2.20% : 0.000012s : 78: predicate.exchange_switch_depend_value 3.29% : 0.000018s : 78: predicate.float_depend_g_call 0.35% : 0.000002s : 14: predicate.float_environ_get_switch 0.52% : 0.000003s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.42% : 0.000002s : 14: predicate.get_grad_eliminate 0.06% : 0.000000s : 1: predicate.graph_param_transform 0.36% : 0.000002s : 14: predicate.incorporate_call 0.33% : 0.000002s : 14: predicate.incorporate_call_switch 6.16% : 0.000034s : 151: predicate.inline 2.46% : 0.000013s : 70: predicate.inline_without_move 0.19% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.50% : 0.000003s : 14: predicate.less_batch_normalization 1.53% : 0.000008s : 47: predicate.list_to_tuple_eliminator_ 2.29% : 0.000012s : 90: predicate.load_eliminater 0.22% : 0.000001s : 1: predicate.loop_unroll_after_grad 4.25% : 0.000023s : 141: predicate.loop_unroll_before_grad 1.39% : 0.000008s : 45: predicate.make_slice_get_slice_eliminator 0.44% : 0.000002s : 14: predicate.merge_addn 0.08% : 0.000000s : 2: predicate.micro_step_allgather_replace 0.10% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.18% : 0.000006s : 43: predicate.minmaximum_grad 0.33% : 0.000002s : 1: predicate.mutable_eliminate 0.07% : 0.000000s : 1: predicate.opt_reshape 0.07% : 0.000000s : 1: predicate.parallel_virtual_node 3.20% : 0.000017s : 78: predicate.partial_defer_inline 1.49% : 0.000008s : 46: predicate.partial_eliminate 1.16% : 0.000006s : 43: predicate.print_const_string_wrapper 0.37% : 0.000002s : 14: predicate.reduce_all_const_elim 1.44% : 0.000008s : 43: predicate.reduce_eliminate 2.46% : 0.000013s : 90: predicate.redundant_stop_gradient_eliminater 0.24% : 0.000001s : 14: predicate.remove_not_recompute_node 1.04% : 0.000006s : 47: predicate.replace_applicator 1.18% : 0.000006s : 70: predicate.replace_old_param 0.05% : 0.000000s : 1: predicate.reset_defer_inline 1.23% : 0.000007s : 43: predicate.reshape_eliminate 0.12% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.09% : 0.000000s : 1: predicate.row_tensor_eliminate 0.23% : 0.000001s : 2: predicate.same_eliminate 0.32% : 0.000002s : 14: predicate.set_cell_output_no_recompute 0.56% : 0.000003s : 14: predicate.shard_identity_eliminate 0.14% : 0.000001s : 2: predicate.special_op_eliminate 0.51% : 0.000003s : 14: predicate.specialize_transform 0.26% : 0.000001s : 2: predicate.split_environ_get_set_with_tuple_value 2.21% : 0.000012s : 70: predicate.stack_unstack_eliminate 0.03% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.52% : 0.000014s : 78: predicate.switch_defer_inline 2.68% : 0.000015s : 80: predicate.switch_layer_defer_inline 8.50% : 0.000046s : 254: predicate.switch_simplify 1.15% : 0.000006s : 43: predicate.tile_eliminate 1.27% : 0.000007s : 43: predicate.transpose_eliminate 1.37% : 0.000007s : 45: predicate.tuple_list_convert_item_index_to_positive 1.46% : 0.000008s : 45: predicate.tuple_list_get_item_const_eliminator 1.35% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.35% : 0.000013s : 61: predicate.tuple_list_get_item_eliminator 1.34% : 0.000007s : 45: predicate.tuple_list_get_set_item_eliminator 1.94% : 0.000011s : 59: predicate.tuple_list_set_item_eliminator 1.25% : 0.000007s : 47: predicate.tuple_to_list_eliminator_ 2.34% : 0.000013s : 90: predicate.updatestate_pure_node_eliminater 2.87% : 0.000016s : 104: predicate.updatestate_useless_node_eliminater 0.07% : 0.000000s : 1: predicate.value_based_eliminate 0.44% : 0.000002s : 14: predicate.virtual_dataset_eliminate 0.52% : 0.000003s : 14: predicate.virtual_output_eliminate 0.03% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.11% : 0.000001s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.004831 69 71.77% : 0.003467s : 33: func_graph_cloner_run.FuncGraphClonerGraph 28.23% : 0.001364s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.150138 192 0.00% : 0.000003s : 1: ForceFp32Comm 2.68% : 0.004019s : 1: add_attr 2.67% : 0.004006s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000035s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.19% : 0.000279s : 1: auto_monad 0.01% : 0.000020s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.32% : 0.000488s : 1: bootstrap 0.02% : 0.000032s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000013s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000021s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000008s : 1: environ_conv 0.06% : 0.000087s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000027s : 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% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000005s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.30% : 0.000444s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.48% : 0.000718s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.01% : 0.000010s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000012s : 1: opt.transform.mutable_eliminate 2.88% : 0.004328s : 78: opt.transform.opt_a 0.01% : 0.000013s : 1: opt.transform.opt_after_cconv 0.01% : 0.000016s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000048s : 28: opt.transform.opt_b 0.02% : 0.000025s : 2: opt.transform.opt_trans_graph 0.02% : 0.000026s : 4: opt.transform.symbol_engine_opt 10.86% : 0.016307s : 1: opt_a 0.05% : 0.000080s : 1: opt_after_cconv 0.31% : 0.000468s : 1: opt_after_jit_grad 0.10% : 0.000148s : 1: opt_b 12.66% : 0.019003s : 1: optimize 0.01% : 0.000019s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000019s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000004s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.08% : 0.000115s : 1: pre_auto_parallel 0.01% : 0.000010s : 1: py_interpret_to_execute 0.01% : 0.000010s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000018s : 1: remove_dup_value 4.21% : 0.006316s : 1: renormalize.infer 0.51% : 0.000771s : 1: renormalize.specialize 0.00% : 0.000005s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000022s : 1: rewriter_after_opt_a 0.43% : 0.000649s : 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 0.02% : 0.000028s : 1: swap_dp_allreduce_reducescatter 0.10% : 0.000157s : 1: symbol_engine_optimizer 0.04% : 0.000054s : 1: tuple_transform 60.65% : 0.091054s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:45:18.664.842 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.0840361, [21] [bootstrap]: 0.00046349 [type_inference]: 0.0686225 [event_method]: 9.376e-05 [auto_monad]: 0.00018839 [graph_reusing]: 1.348e-05 [inline]: 2.43e-06 [add_attr]: 0.00371242, [1] [add_attr_with_inline]: 0.0037027, [1] [Cycle 1]: 0.00011118, [2] [tag_attr]: 5.879e-05 [meta_addattr_fg_expand]: 1.577e-05 [parallel-infer-symbol]: 3.58999e-06 [pre_auto_parallel]: 8.073e-05 [insert-virtual-dataset]: 2.84001e-06 [parallel-infer-symbol-second]: 7.00005e-07 [dataset_repeat_opt]: 1.64e-06 [pipeline_split]: 1.99999e-06 [optimize]: 0.00999254, [53] [py_interpret_to_execute]: 5.66e-06 [rewriter_before_opt_a]: 0.00042754 [opt_a]: 0.00705605, [2] [Cycle 1]: 0.00616187, [45] [expand_dump_flag]: 6.72002e-06 [switch_simplify]: 0.00024276 [loop_unroll]: 9.77e-05 [a_1]: 0.00190019 [with_stream_mark]: 2.575e-05 [recompute_prepare]: 1.474e-05 [updatestate_depend_eliminate]: 5.95002e-06 [updatestate_assign_eliminate]: 4.58001e-06 [updatestate_loads_eliminate]: 4.90999e-06 [parameter_eliminate]: 2.53e-06 [a_2]: 0.0001159 [accelerated_algorithm]: 9.64e-06 [shard]: 1.74998e-06 [meta_shard_fg_expand]: 5.17999e-06 [shard_inline]: 9.19e-06 [merge_send_recv]: 9.80002e-06 [auto_parallel]: 1.065e-05 [parallel]: 1.891e-05 [flash_sp]: 1.166e-05 [merge_comm]: 5.60001e-06 [allreduce_fusion]: 4.85001e-06 [matmul_add_comm_reduction]: 1.108e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 1.262e-05 [virtual_dataset]: 9.10001e-06 [get_grad_eliminate_]: 8.72e-06 [virtual_output]: 9.08002e-06 [merge_forward]: 5.39998e-06 [cell_reuse_recompute_pass]: 1.60001e-06 [offload_activation]: 2.535e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.142e-05 [merge_recompute_call_nodes]: 1.36002e-06 [before_grad]: 1.634e-05 [set_forward_comm_id_for_comm_node_pass]: 5.72999e-06 [meta_fg_expand]: 5.84999e-06 [flash_sp_send_recv_attached]: 3.18e-06 [receive_attached]: 2.31e-06 [after_resolve]: 1.305e-05 [a_after_grad]: 1.417e-05 [renormalize]: 0.00303039 [add_forward_monad_depend]: 8.49002e-06 [auto_monad_grad]: 3.02002e-06 [auto_monad_eliminator]: 2.43e-05 [cse]: 4.858e-05 [a_3]: 7.402e-05 [Cycle 2]: 0.00087821, [45] [expand_dump_flag]: 2.66999e-06 [switch_simplify]: 1.147e-05 [loop_unroll]: 9.00999e-06 [a_1]: 0.00020396 [with_stream_mark]: 1.506e-05 [recompute_prepare]: 8.1e-06 [updatestate_depend_eliminate]: 5.37999e-06 [updatestate_assign_eliminate]: 4.32998e-06 [updatestate_loads_eliminate]: 4.80001e-06 [parameter_eliminate]: 2.05002e-06 [a_2]: 0.00010564 [accelerated_algorithm]: 8.21002e-06 [shard]: 1.82999e-06 [meta_shard_fg_expand]: 2.91e-06 [shard_inline]: 8.84003e-06 [merge_send_recv]: 9.44e-06 [auto_parallel]: 1.003e-05 [parallel]: 8.69e-06 [flash_sp]: 3.89002e-06 [merge_comm]: 4.96997e-06 [allreduce_fusion]: 4.68001e-06 [matmul_add_comm_reduction]: 9.07001e-06 [allreduce_slice_to_reducescatter]: 6.19999e-07 [virtual_shard_identity]: 9.94999e-06 [virtual_dataset]: 8.03999e-06 [get_grad_eliminate_]: 8.02e-06 [virtual_output]: 7.91001e-06 [merge_forward]: 5.34998e-06 [cell_reuse_recompute_pass]: 2.63e-06 [offload_activation]: 1.171e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.735e-05 [merge_recompute_call_nodes]: 1.41002e-06 [before_grad]: 1.434e-05 [set_forward_comm_id_for_comm_node_pass]: 5.99999e-06 [meta_fg_expand]: 3.37997e-06 [flash_sp_send_recv_attached]: 1.79e-06 [receive_attached]: 2.36e-06 [after_resolve]: 1.288e-05 [a_after_grad]: 1.239e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 1.85001e-06 [auto_monad_grad]: 1.67999e-06 [auto_monad_eliminator]: 1.099e-05 [cse]: 2.68e-05 [a_3]: 5.268e-05 [py_interpret_to_execute_after_opt_a]: 8.17998e-06 [slice_cell_reuse_recomputed_activation]: 2.01e-06 [rewriter_after_opt_a]: 3.099e-05 [convert_after_rewriter]: 1.25999e-06 [order_py_execute_after_rewriter]: 1.27e-06 [mutable_eliminate]: 0.00075636 [opt_b]: 0.00027566, [1] [Cycle 1]: 0.00026615, [7] [b_1]: 0.00016466 [b_2]: 8.90999e-06 [updatestate_depend_eliminate]: 1.085e-05 [updatestate_assign_eliminate]: 3.81001e-06 [updatestate_loads_eliminate]: 3.34001e-06 [renormalize]: 5.40022e-07 [cse]: 3.703e-05 [optimize_parallel_all_gather_comm]: 2.019e-05 [overlap_param_gather]: 2.02001e-06 [cconv]: 3.623e-05 [loop_unroll]: 0.00049893 [opt_after_cconv]: 0.00012754, [1] [Cycle 1]: 0.00012087, [7] [c_1]: 3.759e-05 [parameter_eliminate]: 4.76002e-06 [updatestate_depend_eliminate]: 7.67002e-06 [updatestate_assign_eliminate]: 3.64002e-06 [updatestate_loads_eliminate]: 3.96001e-06 [cse]: 2.922e-05 [renormalize]: 5.69999e-07 [remove_dup_value]: 5.198e-05 [tuple_transform]: 9.151e-05, [1] [Cycle 1]: 8.61e-05, [4] [d_1]: 5.681e-05 [none_parameter_eliminate]: 1.94999e-06 [renormalize]: 1.29978e-07 [switch_simplify]: 8.53001e-06 [partial_unused_args_eliminate]: 1.64998e-06 [add_recomputation]: 6.486e-05 [cse_after_recomputation]: 3.028e-05, [1] [Cycle 1]: 2.493e-05, [1] [cse]: 1.877e-05 [environ_conv]: 7.9e-06 [swap_dp_allreduce_reducescatter]: 6.62002e-06 [bias_add_comm_swap]: 2.74001e-06 [label_micro_interleaved_index]: 5.03002e-06 [label_fine_grained_interleaved_index]: 2.43998e-06 [merge_cast_opt]: 1.52999e-06 [slice_recompute_activation]: 2.24001e-06 [micro_interleaved_order_control]: 2.66e-06 [assign_add_opt]: 1.69e-06 [ForceFp32Comm]: 9.20001e-07 [remove_cast_before_assign_add]: 1.07e-06 [full_micro_interleaved_order_control]: 2.40002e-06 [reorder_send_recv_between_fp_bp]: 2.94001e-06 [comm_op_add_attrs]: 1.02998e-06 [add_comm_op_reuse_tag]: 1.23002e-06 [interleave_split_concat_branches]: 1.12e-06 [interleave_parallel_branches]: 1.06002e-06 [overlap_opt_shard_in_pipeline]: 1.22999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.97999e-06 [control_data_broadcast_order]: 1.811e-05 [grouped_pairwise_exchange_alltoall]: 1.50001e-06 [offloading_packed_experts]: 4.78001e-06 [overlap_recompute_and_grad_model_parallel]: 5.74999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.18001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.18001e-06 [overlap_recompute_comm]: 2.32001e-06 [overlap_grad_ring_attention]: 5.00001e-06 [overlap_grad_flash_sp]: 2.951e-05 [begin_end_overlap_inline]: 5.09986e-07 [split_matmul_comm_elemetwise]: 2.28998e-06 [split_layernorm_comm]: 1.96e-06 [handle_group_info]: 1.23002e-06 [symbol_engine_optimizer]: 9.864e-05, [1] [Cycle 1]: 9.334e-05, [6] [build]: 4.92e-06 [elim_shapecalc]: 1.462e-05 [elim_not_effective]: 1.769e-05 [opt_reshape]: 9.59e-06 [fold_const_symbol]: 1.489e-05 [renormalize]: 2.50002e-07 [detach_backward]: 2.29001e-06 [pipeline_parallel_scheduler]: 1.42e-06 [auto_monad_reorder]: 2.297e-05 [get_jit_bprop_graph]: 2.41998e-06 [rewriter_after_jit_bprop_graph]: 6.79999e-06 [opt_after_jit_grad]: 0.00058675 [validate]: 5.31e-05 Sums bootstrap : 0.000463s : 0.58% type_inference : 0.068622s : 86.59% event_method : 0.000094s : 0.12% auto_monad : 0.000188s : 0.24% graph_reusing : 0.000013s : 0.02% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000059s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000016s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000081s : 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 : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000428s : 0.54% optimize.opt_a.expand_dump_flag : 0.000009s : 0.01% optimize.opt_a.switch_simplify : 0.000254s : 0.32% optimize.opt_a.loop_unroll : 0.000107s : 0.13% optimize.opt_a.a_1 : 0.002104s : 2.65% optimize.opt_a.with_stream_mark : 0.000041s : 0.05% optimize.opt_a.recompute_prepare : 0.000023s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.01% optimize.opt_a.a_2 : 0.000222s : 0.28% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.02% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000008s : 0.01% optimize.opt_a.shard_inline : 0.000018s : 0.02% optimize.opt_a.merge_send_recv : 0.000019s : 0.02% optimize.opt_a.auto_parallel : 0.000021s : 0.03% optimize.opt_a.parallel : 0.000028s : 0.03% optimize.opt_a.flash_sp : 0.000016s : 0.02% optimize.opt_a.merge_comm : 0.000011s : 0.01% optimize.opt_a.allreduce_fusion : 0.000010s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000023s : 0.03% optimize.opt_a.virtual_dataset : 0.000017s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000017s : 0.02% optimize.opt_a.virtual_output : 0.000017s : 0.02% optimize.opt_a.merge_forward : 0.000011s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.01% optimize.opt_a.offload_activation : 0.000037s : 0.05% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000039s : 0.05% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000031s : 0.04% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.01% optimize.opt_a.meta_fg_expand : 0.000009s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000005s : 0.01% optimize.opt_a.receive_attached : 0.000005s : 0.01% optimize.opt_a.after_resolve : 0.000026s : 0.03% optimize.opt_a.a_after_grad : 0.000027s : 0.03% optimize.opt_a.renormalize : 0.003030s : 3.82% optimize.opt_a.add_forward_monad_depend : 0.000010s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000035s : 0.04% optimize.opt_a.cse : 0.000075s : 0.10% optimize.opt_a.a_3 : 0.000127s : 0.16% optimize.py_interpret_to_execute_after_opt_a : 0.000008s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000031s : 0.04% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000756s : 0.95% optimize.opt_b.b_1 : 0.000165s : 0.21% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000037s : 0.05% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.03% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000036s : 0.05% optimize.loop_unroll : 0.000499s : 0.63% optimize.opt_after_cconv.c_1 : 0.000038s : 0.05% optimize.opt_after_cconv.parameter_eliminate : 0.000005s : 0.01% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000029s : 0.04% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000052s : 0.07% optimize.tuple_transform.d_1 : 0.000057s : 0.07% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000009s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000065s : 0.08% optimize.cse_after_recomputation.cse : 0.000019s : 0.02% optimize.environ_conv : 0.000008s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000007s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.01% optimize.label_fine_grained_interleaved_index : 0.000002s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000018s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 0.01% 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 : 0.000005s : 0.01% optimize.overlap_grad_flash_sp : 0.000030s : 0.04% 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 : 0.000005s : 0.01% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000015s : 0.02% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 0.02% 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 : 0.000023s : 0.03% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.01% opt_after_jit_grad : 0.000587s : 0.74% validate : 0.000053s : 0.07% Time group info: ------[substitution.] 0.000614 75 8.75% : 0.000054s : 11: substitution.cast_eliminate 0.40% : 0.000002s : 4: substitution.elim_not_effective 0.34% : 0.000002s : 4: substitution.fold_const_symbol 1.14% : 0.000007s : 5: substitution.graph_param_transform 81.29% : 0.000499s : 24: substitution.inline 0.96% : 0.000006s : 8: substitution.j_node_and_user_rematch 1.48% : 0.000009s : 8: substitution.remove_not_recompute_node 0.88% : 0.000005s : 2: substitution.replace_old_param 2.95% : 0.000018s : 7: substitution.switch_simplify 1.81% : 0.000011s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.068502 2 92.50% : 0.063366s : 1: type_inference.infer 7.50% : 0.005136s : 1: type_inference.specialize ------[replace.] 0.000270 33 57.81% : 0.000156s : 24: replace.inline 35.49% : 0.000096s : 7: replace.switch_simplify 6.70% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000508 33 95.43% : 0.000485s : 24: match.inline 2.69% : 0.000014s : 7: match.switch_simplify 1.88% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000429 2544 1.31% : 0.000006s : 32: predicate.accumulaten_eliminater 0.57% : 0.000002s : 5: predicate.ad_related_special_op_eliminate 0.36% : 0.000002s : 10: predicate.addn_check_dump 1.15% : 0.000005s : 32: predicate.addn_zero_filter 1.02% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.06% : 0.000009s : 42: predicate.arithmetic_simplify 1.53% : 0.000007s : 32: predicate.cast_eliminate 0.41% : 0.000002s : 10: predicate.check_bprop_eliminate 0.38% : 0.000002s : 10: predicate.compare_switch_simplify 0.10% : 0.000000s : 5: predicate.const_output_eliminate 0.39% : 0.000002s : 10: predicate.depend_value_elim 1.23% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.34% : 0.000006s : 32: predicate.dict_get_item_eliminator 1.13% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.63% : 0.000003s : 10: predicate.dumpgradient_eliminate 0.13% : 0.000001s : 5: predicate.elim_not_effective 0.27% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.40% : 0.000006s : 37: predicate.environ_add_const_eliminate 1.20% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.26% : 0.000005s : 37: predicate.environ_get_depend_swap 1.73% : 0.000007s : 47: predicate.environ_get_eliminate 1.20% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.03% : 0.000009s : 58: predicate.exchange_switch_depend_value 2.95% : 0.000013s : 58: predicate.float_depend_g_call 0.36% : 0.000002s : 10: predicate.float_environ_get_switch 0.51% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.11% : 0.000000s : 5: predicate.fold_const_symbol 0.44% : 0.000002s : 10: predicate.get_grad_eliminate 0.11% : 0.000000s : 5: predicate.graph_param_transform 0.34% : 0.000001s : 10: predicate.incorporate_call 0.31% : 0.000001s : 10: predicate.incorporate_call_switch 6.29% : 0.000027s : 120: predicate.inline 0.51% : 0.000002s : 10: predicate.inline_without_move 0.20% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.65% : 0.000003s : 10: predicate.less_batch_normalization 1.58% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.38% : 0.000010s : 76: predicate.load_eliminater 0.67% : 0.000003s : 5: predicate.loop_unroll_after_grad 3.59% : 0.000015s : 104: predicate.loop_unroll_before_grad 1.82% : 0.000008s : 42: predicate.make_slice_get_slice_eliminator 0.37% : 0.000002s : 10: predicate.merge_addn 0.36% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.37% : 0.000002s : 10: predicate.mini_step_allgather_replace 1.14% : 0.000005s : 32: predicate.minmaximum_grad 0.89% : 0.000004s : 5: predicate.mutable_eliminate 0.22% : 0.000001s : 5: predicate.opt_reshape 0.25% : 0.000001s : 5: predicate.parallel_virtual_node 2.60% : 0.000011s : 58: predicate.partial_defer_inline 1.49% : 0.000006s : 39: predicate.partial_eliminate 1.10% : 0.000005s : 32: predicate.print_const_string_wrapper 0.40% : 0.000002s : 10: predicate.reduce_all_const_elim 1.54% : 0.000007s : 32: predicate.reduce_eliminate 2.50% : 0.000011s : 76: predicate.redundant_stop_gradient_eliminater 0.22% : 0.000001s : 10: predicate.remove_not_recompute_node 1.14% : 0.000005s : 44: predicate.replace_applicator 0.24% : 0.000001s : 10: predicate.replace_old_param 0.19% : 0.000001s : 5: predicate.reset_defer_inline 1.18% : 0.000005s : 32: predicate.reshape_eliminate 0.42% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.21% : 0.000001s : 5: predicate.row_tensor_eliminate 0.49% : 0.000002s : 10: predicate.same_eliminate 0.30% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.58% : 0.000002s : 10: predicate.shard_identity_eliminate 0.49% : 0.000002s : 10: predicate.special_op_eliminate 0.50% : 0.000002s : 10: predicate.specialize_transform 0.67% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.54% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.19% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.20% : 0.000009s : 58: predicate.switch_defer_inline 2.60% : 0.000011s : 68: predicate.switch_layer_defer_inline 7.38% : 0.000032s : 191: predicate.switch_simplify 1.11% : 0.000005s : 32: predicate.tile_eliminate 1.20% : 0.000005s : 32: predicate.transpose_eliminate 1.53% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.54% : 0.000007s : 42: predicate.tuple_list_get_item_const_eliminator 1.44% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.51% : 0.000011s : 54: predicate.tuple_list_get_item_eliminator 1.53% : 0.000007s : 42: predicate.tuple_list_get_set_item_eliminator 2.04% : 0.000009s : 52: predicate.tuple_list_set_item_eliminator 1.59% : 0.000007s : 44: predicate.tuple_to_list_eliminator_ 2.40% : 0.000010s : 76: predicate.updatestate_pure_node_eliminater 3.04% : 0.000013s : 86: predicate.updatestate_useless_node_eliminater 0.21% : 0.000001s : 5: predicate.value_based_eliminate 0.47% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.47% : 0.000002s : 10: predicate.virtual_output_eliminate 0.17% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.35% : 0.000002s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002981 42 49.96% : 0.001489s : 16: func_graph_cloner_run.FuncGraphClonerGraph 50.04% : 0.001492s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.103915 192 0.00% : 0.000003s : 1: ForceFp32Comm 3.58% : 0.003718s : 1: add_attr 3.57% : 0.003706s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.07% : 0.000070s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.19% : 0.000198s : 1: auto_monad 0.03% : 0.000027s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000005s : 1: bias_add_comm_swap 0.47% : 0.000489s : 1: bootstrap 0.04% : 0.000040s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000033s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000006s : 1: detach_backward 0.01% : 0.000012s : 1: environ_conv 0.10% : 0.000103s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000006s : 1: get_jit_bprop_graph 0.02% : 0.000018s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000005s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000008s : 1: label_micro_interleaved_index 0.49% : 0.000510s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.74% : 0.000769s : 1: mutable_eliminate 0.01% : 0.000008s : 1: offloading_packed_experts 0.02% : 0.000020s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000023s : 1: opt.transform.mutable_eliminate 2.85% : 0.002966s : 78: opt.transform.opt_a 0.03% : 0.000036s : 1: opt.transform.opt_after_cconv 0.03% : 0.000035s : 1: opt.transform.opt_after_jit_grad 0.14% : 0.000144s : 28: opt.transform.opt_b 0.06% : 0.000063s : 2: opt.transform.opt_trans_graph 0.05% : 0.000053s : 4: opt.transform.symbol_engine_opt 6.79% : 0.007061s : 1: opt_a 0.13% : 0.000131s : 1: opt_after_cconv 0.58% : 0.000601s : 1: opt_after_jit_grad 0.27% : 0.000279s : 1: opt_b 9.62% : 0.009998s : 1: optimize 0.02% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.03% : 0.000033s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000008s : 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 0.01% : 0.000006s : 1: overlap_param_gather 0.01% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.01% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.01% : 0.000006s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.08% : 0.000086s : 1: pre_auto_parallel 0.01% : 0.000009s : 1: py_interpret_to_execute 0.01% : 0.000012s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.05% : 0.000056s : 1: remove_dup_value 1.49% : 0.001551s : 1: renormalize.infer 1.41% : 0.001467s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.03% : 0.000034s : 1: rewriter_after_opt_a 0.42% : 0.000435s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.01% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.10% : 0.000102s : 1: symbol_engine_optimizer 0.09% : 0.000094s : 1: tuple_transform 66.06% : 0.068648s : 1: type_inference .[WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:45:19.247.029 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.0681046, [21] [bootstrap]: 0.00051231 [type_inference]: 0.0556922 [event_method]: 0.00027368 [auto_monad]: 0.00014433 [graph_reusing]: 1.013e-05 [inline]: 2.51e-06 [add_attr]: 0.00398594, [1] [add_attr_with_inline]: 0.00397551, [1] [Cycle 1]: 8.393e-05, [2] [tag_attr]: 3.96e-05 [meta_addattr_fg_expand]: 1.091e-05 [parallel-infer-symbol]: 3.33e-06 [pre_auto_parallel]: 5.903e-05 [insert-virtual-dataset]: 2.25002e-06 [parallel-infer-symbol-second]: 6.69999e-07 [dataset_repeat_opt]: 1.90001e-06 [pipeline_split]: 1.87999e-06 [optimize]: 0.00664226, [53] [py_interpret_to_execute]: 4.77998e-06 [rewriter_before_opt_a]: 0.00030672 [opt_a]: 0.00429644, [2] [Cycle 1]: 0.00372209, [45] [expand_dump_flag]: 4.70001e-06 [switch_simplify]: 0.00015425 [loop_unroll]: 5.799e-05 [a_1]: 0.0012268 [with_stream_mark]: 1.62e-05 [recompute_prepare]: 7.98001e-06 [updatestate_depend_eliminate]: 3.88001e-06 [updatestate_assign_eliminate]: 3.43e-06 [updatestate_loads_eliminate]: 2.68e-06 [parameter_eliminate]: 2.07001e-06 [a_2]: 6.948e-05 [accelerated_algorithm]: 5.83002e-06 [shard]: 1.70001e-06 [meta_shard_fg_expand]: 2.74001e-06 [shard_inline]: 5.54e-06 [merge_send_recv]: 8.14997e-06 [auto_parallel]: 6.12001e-06 [parallel]: 4.475e-05 [flash_sp]: 8.09002e-06 [merge_comm]: 4.05e-06 [allreduce_fusion]: 3.17002e-06 [matmul_add_comm_reduction]: 9.20999e-06 [allreduce_slice_to_reducescatter]: 6.10016e-07 [virtual_shard_identity]: 7.68999e-06 [virtual_dataset]: 6.17999e-06 [get_grad_eliminate_]: 5.71e-06 [virtual_output]: 5.69999e-06 [merge_forward]: 3.61999e-06 [cell_reuse_recompute_pass]: 1.30001e-06 [offload_activation]: 8.86002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.212e-05 [merge_recompute_call_nodes]: 1.44003e-06 [before_grad]: 9.56e-06 [set_forward_comm_id_for_comm_node_pass]: 3.23e-06 [meta_fg_expand]: 3.56999e-06 [flash_sp_send_recv_attached]: 2.74999e-06 [receive_attached]: 2.41e-06 [after_resolve]: 8.57998e-06 [a_after_grad]: 8.68001e-06 [renormalize]: 0.00163723 [add_forward_monad_depend]: 6.20002e-06 [auto_monad_grad]: 2.57001e-06 [auto_monad_eliminator]: 1.409e-05 [cse]: 2.755e-05 [a_3]: 4.372e-05 [Cycle 2]: 0.0005619, [45] [expand_dump_flag]: 1.60001e-06 [switch_simplify]: 7.46999e-06 [loop_unroll]: 5.39e-06 [a_1]: 0.00010262 [with_stream_mark]: 1.176e-05 [recompute_prepare]: 5.46002e-06 [updatestate_depend_eliminate]: 2.74001e-06 [updatestate_assign_eliminate]: 2.34001e-06 [updatestate_loads_eliminate]: 2.56998e-06 [parameter_eliminate]: 9.39996e-07 [a_2]: 5.885e-05 [accelerated_algorithm]: 5.32001e-06 [shard]: 1.35999e-06 [meta_shard_fg_expand]: 1.20999e-06 [shard_inline]: 5.04e-06 [merge_send_recv]: 5.15999e-06 [auto_parallel]: 5.47999e-06 [parallel]: 6.06e-06 [flash_sp]: 3.26999e-06 [merge_comm]: 3.56999e-06 [allreduce_fusion]: 2.95002e-06 [matmul_add_comm_reduction]: 5.96e-06 [allreduce_slice_to_reducescatter]: 3.4002e-07 [virtual_shard_identity]: 6.06998e-06 [virtual_dataset]: 5.14e-06 [get_grad_eliminate_]: 5.39e-06 [virtual_output]: 4.92999e-06 [merge_forward]: 2.98998e-06 [cell_reuse_recompute_pass]: 1.81e-06 [offload_activation]: 6.63e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.207e-05 [merge_recompute_call_nodes]: 9.60019e-07 [before_grad]: 8.82e-06 [set_forward_comm_id_for_comm_node_pass]: 3.13e-06 [meta_fg_expand]: 1.97001e-06 [flash_sp_send_recv_attached]: 1.07998e-06 [receive_attached]: 1.22e-06 [after_resolve]: 7.97e-06 [a_after_grad]: 8.26002e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.17e-06 [auto_monad_grad]: 1.04e-06 [auto_monad_eliminator]: 5.65001e-06 [cse]: 1.195e-05 [a_3]: 2.934e-05 [py_interpret_to_execute_after_opt_a]: 4.99998e-06 [slice_cell_reuse_recomputed_activation]: 1.78002e-06 [rewriter_after_opt_a]: 1.551e-05 [convert_after_rewriter]: 1.77001e-06 [order_py_execute_after_rewriter]: 1.14e-06 [mutable_eliminate]: 0.00063329 [opt_b]: 0.00017578, [1] [Cycle 1]: 0.00016853, [7] [b_1]: 0.00010293 [b_2]: 6.28002e-06 [updatestate_depend_eliminate]: 5.71e-06 [updatestate_assign_eliminate]: 2.17999e-06 [updatestate_loads_eliminate]: 2.06e-06 [renormalize]: 5.69999e-07 [cse]: 1.544e-05 [optimize_parallel_all_gather_comm]: 1.614e-05 [overlap_param_gather]: 2.17999e-06 [cconv]: 2.403e-05 [loop_unroll]: 0.00045169 [opt_after_cconv]: 9.118e-05, [1] [Cycle 1]: 8.511e-05, [7] [c_1]: 2.525e-05 [parameter_eliminate]: 2.44999e-06 [updatestate_depend_eliminate]: 5.09e-06 [updatestate_assign_eliminate]: 2.28998e-06 [updatestate_loads_eliminate]: 2.04999e-06 [cse]: 1.543e-05 [renormalize]: 4.09986e-07 [remove_dup_value]: 1.275e-05 [tuple_transform]: 6.768e-05, [1] [Cycle 1]: 6.349e-05, [4] [d_1]: 3.743e-05 [none_parameter_eliminate]: 1.61002e-06 [renormalize]: 1.50001e-07 [switch_simplify]: 6.02999e-06 [partial_unused_args_eliminate]: 1.70001e-06 [add_recomputation]: 8.518e-05 [cse_after_recomputation]: 2.181e-05, [1] [Cycle 1]: 1.675e-05, [1] [cse]: 1.126e-05 [environ_conv]: 5.17e-06 [swap_dp_allreduce_reducescatter]: 4.89e-06 [bias_add_comm_swap]: 2.29001e-06 [label_micro_interleaved_index]: 4.35e-06 [label_fine_grained_interleaved_index]: 2.71e-06 [merge_cast_opt]: 1.37e-06 [slice_recompute_activation]: 1.87999e-06 [micro_interleaved_order_control]: 2.13998e-06 [assign_add_opt]: 1.26997e-06 [ForceFp32Comm]: 8.89995e-07 [remove_cast_before_assign_add]: 1.02e-06 [full_micro_interleaved_order_control]: 2.07001e-06 [reorder_send_recv_between_fp_bp]: 2.72001e-06 [comm_op_add_attrs]: 1.15999e-06 [add_comm_op_reuse_tag]: 9.50007e-07 [interleave_split_concat_branches]: 1.12e-06 [interleave_parallel_branches]: 1.07e-06 [overlap_opt_shard_in_pipeline]: 1.15001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.76e-06 [control_data_broadcast_order]: 1.138e-05 [grouped_pairwise_exchange_alltoall]: 1.44e-06 [offloading_packed_experts]: 3.76001e-06 [overlap_recompute_and_grad_model_parallel]: 4.79e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.10999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37e-06 [overlap_recompute_comm]: 2.76e-06 [overlap_grad_ring_attention]: 3.83999e-06 [overlap_grad_flash_sp]: 1.911e-05 [begin_end_overlap_inline]: 5.19998e-07 [split_matmul_comm_elemetwise]: 2.14e-06 [split_layernorm_comm]: 1.77001e-06 [handle_group_info]: 1.34e-06 [symbol_engine_optimizer]: 0.00010956, [1] [Cycle 1]: 0.00010532, [6] [build]: 3.06001e-06 [elim_shapecalc]: 9.07999e-06 [elim_not_effective]: 4.797e-05 [opt_reshape]: 7.01999e-06 [fold_const_symbol]: 9.39e-06 [renormalize]: 3.00002e-07 [detach_backward]: 2.19001e-06 [pipeline_parallel_scheduler]: 1.61002e-06 [auto_monad_reorder]: 1.724e-05 [get_jit_bprop_graph]: 1.90001e-06 [rewriter_after_jit_bprop_graph]: 4.46002e-06 [opt_after_jit_grad]: 0.00049362 [validate]: 7.497e-05 Sums bootstrap : 0.000512s : 0.81% type_inference : 0.055692s : 88.15% event_method : 0.000274s : 0.43% auto_monad : 0.000144s : 0.23% graph_reusing : 0.000010s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000040s : 0.06% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000011s : 0.02% parallel-infer-symbol : 0.000003s : 0.01% pre_auto_parallel : 0.000059s : 0.09% 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 : 0.000005s : 0.01% optimize.rewriter_before_opt_a : 0.000307s : 0.49% optimize.opt_a.expand_dump_flag : 0.000006s : 0.01% optimize.opt_a.switch_simplify : 0.000162s : 0.26% optimize.opt_a.loop_unroll : 0.000063s : 0.10% optimize.opt_a.a_1 : 0.001329s : 2.10% optimize.opt_a.with_stream_mark : 0.000028s : 0.04% optimize.opt_a.recompute_prepare : 0.000013s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000007s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000006s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000005s : 0.01% optimize.opt_a.parameter_eliminate : 0.000003s : 0.00% optimize.opt_a.a_2 : 0.000128s : 0.20% optimize.opt_a.accelerated_algorithm : 0.000011s : 0.02% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000004s : 0.01% optimize.opt_a.shard_inline : 0.000011s : 0.02% optimize.opt_a.merge_send_recv : 0.000013s : 0.02% optimize.opt_a.auto_parallel : 0.000012s : 0.02% optimize.opt_a.parallel : 0.000051s : 0.08% optimize.opt_a.flash_sp : 0.000011s : 0.02% optimize.opt_a.merge_comm : 0.000008s : 0.01% optimize.opt_a.allreduce_fusion : 0.000006s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000015s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000014s : 0.02% optimize.opt_a.virtual_dataset : 0.000011s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.02% optimize.opt_a.virtual_output : 0.000011s : 0.02% optimize.opt_a.merge_forward : 0.000007s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000015s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000024s : 0.04% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000018s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000006s : 0.01% optimize.opt_a.meta_fg_expand : 0.000006s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.01% optimize.opt_a.receive_attached : 0.000004s : 0.01% optimize.opt_a.after_resolve : 0.000017s : 0.03% optimize.opt_a.a_after_grad : 0.000017s : 0.03% optimize.opt_a.renormalize : 0.001637s : 2.59% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000020s : 0.03% optimize.opt_a.cse : 0.000039s : 0.06% optimize.opt_a.a_3 : 0.000073s : 0.12% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000016s : 0.02% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000633s : 1.00% optimize.opt_b.b_1 : 0.000103s : 0.16% optimize.opt_b.b_2 : 0.000006s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000015s : 0.02% optimize.optimize_parallel_all_gather_comm : 0.000016s : 0.03% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000024s : 0.04% optimize.loop_unroll : 0.000452s : 0.71% optimize.opt_after_cconv.c_1 : 0.000025s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000015s : 0.02% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.02% optimize.tuple_transform.d_1 : 0.000037s : 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 : 0.000006s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000085s : 0.13% optimize.cse_after_recomputation.cse : 0.000011s : 0.02% optimize.environ_conv : 0.000005s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000002s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.01% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000011s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000004s : 0.01% optimize.overlap_grad_flash_sp : 0.000019s : 0.03% 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000048s : 0.08% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 0.01% 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 : 0.000017s : 0.03% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.01% opt_after_jit_grad : 0.000494s : 0.78% validate : 0.000075s : 0.12% Time group info: ------[substitution.] 0.000360 39 2.90% : 0.000010s : 1: substitution.cast_eliminate 0.54% : 0.000002s : 2: substitution.elim_not_effective 0.38% : 0.000001s : 2: substitution.fold_const_symbol 1.44% : 0.000005s : 3: substitution.graph_param_transform 85.56% : 0.000308s : 15: substitution.inline 1.01% : 0.000004s : 4: substitution.j_node_and_user_rematch 1.23% : 0.000004s : 4: substitution.remove_not_recompute_node 0.83% : 0.000003s : 2: substitution.replace_old_param 3.37% : 0.000012s : 4: substitution.switch_simplify 2.75% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.055605 2 94.41% : 0.052495s : 1: type_inference.infer 5.59% : 0.003110s : 1: type_inference.specialize ------[replace.] 0.000163 22 3.54% : 0.000006s : 1: replace.cast_eliminate 57.55% : 0.000094s : 15: replace.inline 30.14% : 0.000049s : 4: replace.switch_simplify 8.76% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000328 22 2.89% : 0.000009s : 1: match.cast_eliminate 91.33% : 0.000300s : 15: match.inline 3.06% : 0.000010s : 4: match.switch_simplify 2.71% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000278 1716 1.16% : 0.000003s : 22: predicate.accumulaten_eliminater 0.49% : 0.000001s : 3: predicate.ad_related_special_op_eliminate 0.31% : 0.000001s : 6: predicate.addn_check_dump 1.23% : 0.000003s : 22: predicate.addn_zero_filter 1.06% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.11% : 0.000006s : 28: predicate.arithmetic_simplify 1.29% : 0.000004s : 23: predicate.cast_eliminate 0.41% : 0.000001s : 6: predicate.check_bprop_eliminate 0.31% : 0.000001s : 6: predicate.compare_switch_simplify 0.12% : 0.000000s : 3: predicate.const_output_eliminate 0.37% : 0.000001s : 6: predicate.depend_value_elim 1.29% : 0.000004s : 23: predicate.dict_get_item_const_eliminator 1.31% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.22% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.62% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.13% : 0.000000s : 3: predicate.elim_not_effective 0.24% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.43% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.27% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.32% : 0.000004s : 26: predicate.environ_get_depend_swap 1.77% : 0.000005s : 32: predicate.environ_get_eliminate 1.33% : 0.000004s : 26: predicate.environ_get_set_eliminate 2.00% : 0.000006s : 40: predicate.exchange_switch_depend_value 3.02% : 0.000008s : 40: predicate.float_depend_g_call 0.32% : 0.000001s : 6: predicate.float_environ_get_switch 0.44% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.13% : 0.000000s : 3: predicate.fold_const_symbol 0.58% : 0.000002s : 6: predicate.get_grad_eliminate 0.19% : 0.000001s : 3: predicate.graph_param_transform 0.33% : 0.000001s : 6: predicate.incorporate_call 0.28% : 0.000001s : 6: predicate.incorporate_call_switch 5.82% : 0.000016s : 80: predicate.inline 0.50% : 0.000001s : 6: predicate.inline_without_move 0.18% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.51% : 0.000001s : 6: predicate.less_batch_normalization 1.73% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.52% : 0.000007s : 53: predicate.load_eliminater 0.57% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.61% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.61% : 0.000004s : 29: predicate.make_slice_get_slice_eliminator 0.34% : 0.000001s : 6: predicate.merge_addn 0.35% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.35% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.07% : 0.000003s : 22: predicate.minmaximum_grad 0.63% : 0.000002s : 3: predicate.mutable_eliminate 0.28% : 0.000001s : 3: predicate.opt_reshape 0.24% : 0.000001s : 3: predicate.parallel_virtual_node 2.79% : 0.000008s : 40: predicate.partial_defer_inline 1.55% : 0.000004s : 28: predicate.partial_eliminate 1.09% : 0.000003s : 22: predicate.print_const_string_wrapper 0.37% : 0.000001s : 6: predicate.reduce_all_const_elim 1.55% : 0.000004s : 22: predicate.reduce_eliminate 2.70% : 0.000008s : 53: predicate.redundant_stop_gradient_eliminater 0.23% : 0.000001s : 6: predicate.remove_not_recompute_node 1.13% : 0.000003s : 31: predicate.replace_applicator 0.33% : 0.000001s : 6: predicate.replace_old_param 0.17% : 0.000000s : 3: predicate.reset_defer_inline 1.13% : 0.000003s : 22: predicate.reshape_eliminate 0.37% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.24% : 0.000001s : 3: predicate.row_tensor_eliminate 0.56% : 0.000002s : 6: predicate.same_eliminate 0.23% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.52% : 0.000001s : 6: predicate.shard_identity_eliminate 0.49% : 0.000001s : 6: predicate.special_op_eliminate 0.40% : 0.000001s : 6: predicate.specialize_transform 0.55% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.46% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.19% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.26% : 0.000006s : 40: predicate.switch_defer_inline 2.49% : 0.000007s : 46: predicate.switch_layer_defer_inline 7.10% : 0.000020s : 126: predicate.switch_simplify 1.20% : 0.000003s : 22: predicate.tile_eliminate 1.19% : 0.000003s : 22: predicate.transpose_eliminate 1.75% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.68% : 0.000005s : 29: predicate.tuple_list_get_item_const_eliminator 1.56% : 0.000004s : 29: predicate.tuple_list_get_item_depend_reorder 2.81% : 0.000008s : 37: predicate.tuple_list_get_item_eliminator 1.47% : 0.000004s : 29: predicate.tuple_list_get_set_item_eliminator 2.07% : 0.000006s : 35: predicate.tuple_list_set_item_eliminator 1.78% : 0.000005s : 31: predicate.tuple_to_list_eliminator_ 2.63% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 3.11% : 0.000009s : 59: predicate.updatestate_useless_node_eliminater 0.22% : 0.000001s : 3: predicate.value_based_eliminate 0.40% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.41% : 0.000001s : 6: predicate.virtual_output_eliminate 0.16% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.28% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.001830 27 48.42% : 0.000886s : 10: func_graph_cloner_run.FuncGraphClonerGraph 51.58% : 0.000944s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.082291 192 0.00% : 0.000004s : 1: ForceFp32Comm 4.85% : 0.003991s : 1: add_attr 4.84% : 0.003980s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 0.11% : 0.000090s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.19% : 0.000155s : 1: auto_monad 0.03% : 0.000021s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.01% : 0.000005s : 1: bias_add_comm_swap 0.66% : 0.000540s : 1: bootstrap 0.03% : 0.000028s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000014s : 1: control_data_broadcast_order 0.01% : 0.000005s : 1: convert_after_rewriter 0.03% : 0.000025s : 1: cse_after_recomputation 0.01% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000005s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.35% : 0.000286s : 1: event_method 0.01% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000014s : 1: graph_reusing 0.01% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000006s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000007s : 1: label_micro_interleaved_index 0.56% : 0.000461s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.78% : 0.000643s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.02% : 0.000013s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000013s : 1: opt.transform.mutable_eliminate 2.22% : 0.001831s : 78: opt.transform.opt_a 0.03% : 0.000024s : 1: opt.transform.opt_after_cconv 0.03% : 0.000022s : 1: opt.transform.opt_after_jit_grad 0.10% : 0.000082s : 28: opt.transform.opt_b 0.05% : 0.000041s : 2: opt.transform.opt_trans_graph 0.08% : 0.000070s : 4: opt.transform.symbol_engine_opt 5.23% : 0.004300s : 1: opt_a 0.11% : 0.000094s : 1: opt_after_cconv 0.61% : 0.000504s : 1: opt_after_jit_grad 0.22% : 0.000179s : 1: opt_b 8.08% : 0.006647s : 1: optimize 0.02% : 0.000020s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.03% : 0.000022s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000007s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.01% : 0.000005s : 1: overlap_param_gather 0.01% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.01% : 0.000006s : 1: partial_unused_args_eliminate 0.01% : 0.000005s : 1: pipeline_parallel_scheduler 0.01% : 0.000005s : 1: pipeline_split 0.08% : 0.000064s : 1: pre_auto_parallel 0.01% : 0.000008s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000016s : 1: remove_dup_value 1.07% : 0.000884s : 1: renormalize.infer 0.90% : 0.000744s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000019s : 1: rewriter_after_opt_a 0.38% : 0.000313s : 1: rewriter_before_opt_a 0.01% : 0.000004s : 1: slice_cell_reuse_recomputed_activation 0.01% : 0.000004s : 1: slice_recompute_activation 0.01% : 0.000005s : 1: split_layernorm_comm 0.01% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.14% : 0.000112s : 1: symbol_engine_optimizer 0.09% : 0.000071s : 1: tuple_transform 67.70% : 0.055714s : 1: type_inference .[WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:45:31.146.243 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.0632106, [21] [bootstrap]: 0.00039698 [type_inference]: 0.0517978 [event_method]: 0.00027086 [auto_monad]: 0.00013428 [graph_reusing]: 9.74e-06 [inline]: 2.58e-06 [add_attr]: 0.00331837, [1] [add_attr_with_inline]: 0.00331026, [1] [Cycle 1]: 7.55e-05, [2] [tag_attr]: 3.539e-05 [meta_addattr_fg_expand]: 1.194e-05 [parallel-infer-symbol]: 3.01999e-06 [pre_auto_parallel]: 5.62e-05 [insert-virtual-dataset]: 2.49999e-06 [parallel-infer-symbol-second]: 6.39993e-07 [dataset_repeat_opt]: 1.93002e-06 [pipeline_split]: 1.54e-06 [optimize]: 0.00645321, [53] [py_interpret_to_execute]: 5.06997e-06 [rewriter_before_opt_a]: 0.00029202 [opt_a]: 0.00430085, [2] [Cycle 1]: 0.00373472, [45] [expand_dump_flag]: 4.77e-06 [switch_simplify]: 0.00015328 [loop_unroll]: 5.803e-05 [a_1]: 0.00118752 [with_stream_mark]: 1.533e-05 [recompute_prepare]: 7.90998e-06 [updatestate_depend_eliminate]: 3.47997e-06 [updatestate_assign_eliminate]: 2.96999e-06 [updatestate_loads_eliminate]: 2.73e-06 [parameter_eliminate]: 1.85001e-06 [a_2]: 7.12e-05 [accelerated_algorithm]: 6.56999e-06 [shard]: 1.94e-06 [meta_shard_fg_expand]: 2.27999e-06 [shard_inline]: 6.02001e-06 [merge_send_recv]: 8.55001e-06 [auto_parallel]: 5.57001e-06 [parallel]: 1.875e-05 [flash_sp]: 8.57998e-06 [merge_comm]: 3.63e-06 [allreduce_fusion]: 3.24001e-06 [matmul_add_comm_reduction]: 9.58997e-06 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 6.68998e-06 [virtual_dataset]: 5.91e-06 [get_grad_eliminate_]: 6.02001e-06 [virtual_output]: 5.35999e-06 [merge_forward]: 3.76999e-06 [cell_reuse_recompute_pass]: 1.10001e-06 [offload_activation]: 9.47999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.166e-05 [merge_recompute_call_nodes]: 1.62001e-06 [before_grad]: 9.31e-06 [set_forward_comm_id_for_comm_node_pass]: 3.28998e-06 [meta_fg_expand]: 3.5e-06 [flash_sp_send_recv_attached]: 2.54001e-06 [receive_attached]: 2.27999e-06 [after_resolve]: 8.29002e-06 [a_after_grad]: 8.82999e-06 [renormalize]: 0.00171847 [add_forward_monad_depend]: 6.09001e-06 [auto_monad_grad]: 2.54001e-06 [auto_monad_eliminator]: 1.431e-05 [cse]: 2.705e-05 [a_3]: 4.308e-05 [Cycle 2]: 0.00055471, [45] [expand_dump_flag]: 1.00999e-06 [switch_simplify]: 6.58e-06 [loop_unroll]: 5.19e-06 [a_1]: 0.00010185 [with_stream_mark]: 1.195e-05 [recompute_prepare]: 5.42001e-06 [updatestate_depend_eliminate]: 2.57001e-06 [updatestate_assign_eliminate]: 2.32001e-06 [updatestate_loads_eliminate]: 2.39001e-06 [parameter_eliminate]: 1.57999e-06 [a_2]: 5.837e-05 [accelerated_algorithm]: 5.48997e-06 [shard]: 1.04998e-06 [meta_shard_fg_expand]: 1.55001e-06 [shard_inline]: 4.89e-06 [merge_send_recv]: 4.75001e-06 [auto_parallel]: 5.22e-06 [parallel]: 5.14e-06 [flash_sp]: 3.51001e-06 [merge_comm]: 2.75997e-06 [allreduce_fusion]: 2.81e-06 [matmul_add_comm_reduction]: 6.26e-06 [allreduce_slice_to_reducescatter]: 5.60016e-07 [virtual_shard_identity]: 5.79999e-06 [virtual_dataset]: 5.01002e-06 [get_grad_eliminate_]: 4.90999e-06 [virtual_output]: 4.84e-06 [merge_forward]: 2.59001e-06 [cell_reuse_recompute_pass]: 1.40999e-06 [offload_activation]: 6.24001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.232e-05 [merge_recompute_call_nodes]: 6.99976e-07 [before_grad]: 8.1e-06 [set_forward_comm_id_for_comm_node_pass]: 3.21001e-06 [meta_fg_expand]: 1.97999e-06 [flash_sp_send_recv_attached]: 1.09e-06 [receive_attached]: 1.50999e-06 [after_resolve]: 7.88001e-06 [a_after_grad]: 7.73001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.71e-06 [auto_monad_grad]: 1.02e-06 [auto_monad_eliminator]: 5.32001e-06 [cse]: 1.174e-05 [a_3]: 2.956e-05 [py_interpret_to_execute_after_opt_a]: 4.72e-06 [slice_cell_reuse_recomputed_activation]: 1.86e-06 [rewriter_after_opt_a]: 1.553e-05 [convert_after_rewriter]: 1.69e-06 [order_py_execute_after_rewriter]: 1.12e-06 [mutable_eliminate]: 0.00054199 [opt_b]: 0.00018018, [1] [Cycle 1]: 0.000174, [7] [b_1]: 0.00010421 [b_2]: 6.96999e-06 [updatestate_depend_eliminate]: 5.55001e-06 [updatestate_assign_eliminate]: 2.41e-06 [updatestate_loads_eliminate]: 2.39999e-06 [renormalize]: 6.69999e-07 [cse]: 1.687e-05 [optimize_parallel_all_gather_comm]: 1.436e-05 [overlap_param_gather]: 2.19001e-06 [cconv]: 2.351e-05 [loop_unroll]: 0.00043886 [opt_after_cconv]: 9.284e-05, [1] [Cycle 1]: 8.669e-05, [7] [c_1]: 2.573e-05 [parameter_eliminate]: 2.61e-06 [updatestate_depend_eliminate]: 4.89e-06 [updatestate_assign_eliminate]: 2.29001e-06 [updatestate_loads_eliminate]: 2.17999e-06 [cse]: 1.637e-05 [renormalize]: 4.49974e-07 [remove_dup_value]: 1.374e-05 [tuple_transform]: 6.916e-05, [1] [Cycle 1]: 6.41e-05, [4] [d_1]: 3.753e-05 [none_parameter_eliminate]: 1.45999e-06 [renormalize]: 3.00002e-07 [switch_simplify]: 6.60002e-06 [partial_unused_args_eliminate]: 1.66998e-06 [add_recomputation]: 4.442e-05 [cse_after_recomputation]: 2.163e-05, [1] [Cycle 1]: 1.669e-05, [1] [cse]: 1.085e-05 [environ_conv]: 5.49e-06 [swap_dp_allreduce_reducescatter]: 4.75001e-06 [bias_add_comm_swap]: 2.69999e-06 [label_micro_interleaved_index]: 4.22e-06 [label_fine_grained_interleaved_index]: 2.76999e-06 [merge_cast_opt]: 1.40001e-06 [slice_recompute_activation]: 2.36e-06 [micro_interleaved_order_control]: 2.56e-06 [assign_add_opt]: 1.43002e-06 [ForceFp32Comm]: 7.80012e-07 [remove_cast_before_assign_add]: 9.89996e-07 [full_micro_interleaved_order_control]: 2.39001e-06 [reorder_send_recv_between_fp_bp]: 2.99999e-06 [comm_op_add_attrs]: 1.05999e-06 [add_comm_op_reuse_tag]: 9.60019e-07 [interleave_split_concat_branches]: 1.31002e-06 [interleave_parallel_branches]: 1.10001e-06 [overlap_opt_shard_in_pipeline]: 1.13001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.87999e-06 [control_data_broadcast_order]: 1.204e-05 [grouped_pairwise_exchange_alltoall]: 1.57001e-06 [offloading_packed_experts]: 3.73001e-06 [overlap_recompute_and_grad_model_parallel]: 4.73001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.18001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.66e-06 [overlap_recompute_comm]: 2.03997e-06 [overlap_grad_ring_attention]: 3.61001e-06 [overlap_grad_flash_sp]: 1.739e-05 [begin_end_overlap_inline]: 4.70027e-07 [split_matmul_comm_elemetwise]: 2.31998e-06 [split_layernorm_comm]: 1.77999e-06 [handle_group_info]: 1.04003e-06 [symbol_engine_optimizer]: 6.61e-05, [1] [Cycle 1]: 6.22e-05, [6] [build]: 2.74999e-06 [elim_shapecalc]: 8.62e-06 [elim_not_effective]: 1.082e-05 [opt_reshape]: 5.57001e-06 [fold_const_symbol]: 8.79e-06 [renormalize]: 2.80008e-07 [detach_backward]: 2.09e-06 [pipeline_parallel_scheduler]: 2.02001e-06 [auto_monad_reorder]: 1.769e-05 [get_jit_bprop_graph]: 1.29e-06 [rewriter_after_jit_bprop_graph]: 5.56e-06 [opt_after_jit_grad]: 0.00049803 [validate]: 3.794e-05 Sums bootstrap : 0.000397s : 0.67% type_inference : 0.051798s : 87.90% event_method : 0.000271s : 0.46% auto_monad : 0.000134s : 0.23% graph_reusing : 0.000010s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000035s : 0.06% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000012s : 0.02% parallel-infer-symbol : 0.000003s : 0.01% pre_auto_parallel : 0.000056s : 0.10% 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 : 0.000005s : 0.01% optimize.rewriter_before_opt_a : 0.000292s : 0.50% optimize.opt_a.expand_dump_flag : 0.000006s : 0.01% optimize.opt_a.switch_simplify : 0.000160s : 0.27% optimize.opt_a.loop_unroll : 0.000063s : 0.11% optimize.opt_a.a_1 : 0.001289s : 2.19% optimize.opt_a.with_stream_mark : 0.000027s : 0.05% optimize.opt_a.recompute_prepare : 0.000013s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000005s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000005s : 0.01% optimize.opt_a.parameter_eliminate : 0.000003s : 0.01% optimize.opt_a.a_2 : 0.000130s : 0.22% optimize.opt_a.accelerated_algorithm : 0.000012s : 0.02% optimize.opt_a.shard : 0.000003s : 0.01% optimize.opt_a.meta_shard_fg_expand : 0.000004s : 0.01% optimize.opt_a.shard_inline : 0.000011s : 0.02% optimize.opt_a.merge_send_recv : 0.000013s : 0.02% optimize.opt_a.auto_parallel : 0.000011s : 0.02% optimize.opt_a.parallel : 0.000024s : 0.04% optimize.opt_a.flash_sp : 0.000012s : 0.02% optimize.opt_a.merge_comm : 0.000006s : 0.01% optimize.opt_a.allreduce_fusion : 0.000006s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000016s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000012s : 0.02% optimize.opt_a.virtual_dataset : 0.000011s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.02% optimize.opt_a.virtual_output : 0.000010s : 0.02% optimize.opt_a.merge_forward : 0.000006s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000016s : 0.03% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000024s : 0.04% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000017s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000006s : 0.01% optimize.opt_a.meta_fg_expand : 0.000005s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.01% optimize.opt_a.receive_attached : 0.000004s : 0.01% optimize.opt_a.after_resolve : 0.000016s : 0.03% optimize.opt_a.a_after_grad : 0.000017s : 0.03% optimize.opt_a.renormalize : 0.001719s : 2.92% optimize.opt_a.add_forward_monad_depend : 0.000008s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000020s : 0.03% optimize.opt_a.cse : 0.000039s : 0.07% optimize.opt_a.a_3 : 0.000073s : 0.12% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000016s : 0.03% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000542s : 0.92% optimize.opt_b.b_1 : 0.000104s : 0.18% optimize.opt_b.b_2 : 0.000007s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000017s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000014s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000024s : 0.04% optimize.loop_unroll : 0.000439s : 0.74% optimize.opt_after_cconv.c_1 : 0.000026s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000016s : 0.03% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000014s : 0.02% optimize.tuple_transform.d_1 : 0.000038s : 0.06% optimize.tuple_transform.none_parameter_eliminate : 0.000001s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000007s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000044s : 0.08% optimize.cse_after_recomputation.cse : 0.000011s : 0.02% optimize.environ_conv : 0.000005s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.01% 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 : 0.01% 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 : 0.000012s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000004s : 0.01% optimize.overlap_grad_flash_sp : 0.000017s : 0.03% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000006s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 0.01% 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 : 0.000018s : 0.03% get_jit_bprop_graph : 0.000001s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.01% opt_after_jit_grad : 0.000498s : 0.85% validate : 0.000038s : 0.06% Time group info: ------[substitution.] 0.000324 39 3.51% : 0.000011s : 1: substitution.cast_eliminate 0.54% : 0.000002s : 2: substitution.elim_not_effective 0.54% : 0.000002s : 2: substitution.fold_const_symbol 1.69% : 0.000005s : 3: substitution.graph_param_transform 83.49% : 0.000271s : 15: substitution.inline 1.02% : 0.000003s : 4: substitution.j_node_and_user_rematch 1.45% : 0.000005s : 4: substitution.remove_not_recompute_node 0.88% : 0.000003s : 2: substitution.replace_old_param 4.02% : 0.000013s : 4: substitution.switch_simplify 2.86% : 0.000009s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.051716 2 94.10% : 0.048667s : 1: type_inference.infer 5.90% : 0.003049s : 1: type_inference.specialize ------[replace.] 0.000156 22 3.41% : 0.000005s : 1: replace.cast_eliminate 57.58% : 0.000090s : 15: replace.inline 30.49% : 0.000047s : 4: replace.switch_simplify 8.53% : 0.000013s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000292 22 3.51% : 0.000010s : 1: match.cast_eliminate 89.93% : 0.000263s : 15: match.inline 3.74% : 0.000011s : 4: match.switch_simplify 2.82% : 0.000008s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000280 1716 1.15% : 0.000003s : 22: predicate.accumulaten_eliminater 0.51% : 0.000001s : 3: predicate.ad_related_special_op_eliminate 0.31% : 0.000001s : 6: predicate.addn_check_dump 1.18% : 0.000003s : 22: predicate.addn_zero_filter 1.05% : 0.000003s : 22: predicate.adjust_all_reduce_mul_add 2.24% : 0.000006s : 28: predicate.arithmetic_simplify 1.24% : 0.000003s : 23: predicate.cast_eliminate 0.34% : 0.000001s : 6: predicate.check_bprop_eliminate 0.32% : 0.000001s : 6: predicate.compare_switch_simplify 0.10% : 0.000000s : 3: predicate.const_output_eliminate 0.35% : 0.000001s : 6: predicate.depend_value_elim 1.22% : 0.000003s : 23: predicate.dict_get_item_const_eliminator 1.29% : 0.000004s : 23: predicate.dict_get_item_eliminator 1.16% : 0.000003s : 23: predicate.dict_set_item_eliminator 0.67% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.10% : 0.000000s : 3: predicate.elim_not_effective 0.24% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.33% : 0.000004s : 26: predicate.environ_add_const_eliminate 1.37% : 0.000004s : 26: predicate.environ_get_add_eliminate 1.32% : 0.000004s : 26: predicate.environ_get_depend_swap 1.89% : 0.000005s : 32: predicate.environ_get_eliminate 1.38% : 0.000004s : 26: predicate.environ_get_set_eliminate 2.00% : 0.000006s : 40: predicate.exchange_switch_depend_value 3.06% : 0.000009s : 40: predicate.float_depend_g_call 0.36% : 0.000001s : 6: predicate.float_environ_get_switch 0.50% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.11% : 0.000000s : 3: predicate.fold_const_symbol 0.48% : 0.000001s : 6: predicate.get_grad_eliminate 0.14% : 0.000000s : 3: predicate.graph_param_transform 0.31% : 0.000001s : 6: predicate.incorporate_call 0.28% : 0.000001s : 6: predicate.incorporate_call_switch 6.36% : 0.000018s : 80: predicate.inline 0.44% : 0.000001s : 6: predicate.inline_without_move 0.19% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.69% : 0.000002s : 6: predicate.less_batch_normalization 1.74% : 0.000005s : 31: predicate.list_to_tuple_eliminator_ 2.52% : 0.000007s : 53: predicate.load_eliminater 0.68% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.51% : 0.000010s : 69: predicate.loop_unroll_before_grad 1.71% : 0.000005s : 29: predicate.make_slice_get_slice_eliminator 0.34% : 0.000001s : 6: predicate.merge_addn 0.31% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.35% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.07% : 0.000003s : 22: predicate.minmaximum_grad 0.62% : 0.000002s : 3: predicate.mutable_eliminate 0.20% : 0.000001s : 3: predicate.opt_reshape 0.24% : 0.000001s : 3: predicate.parallel_virtual_node 2.65% : 0.000007s : 40: predicate.partial_defer_inline 1.51% : 0.000004s : 28: predicate.partial_eliminate 1.12% : 0.000003s : 22: predicate.print_const_string_wrapper 0.38% : 0.000001s : 6: predicate.reduce_all_const_elim 1.57% : 0.000004s : 22: predicate.reduce_eliminate 2.58% : 0.000007s : 53: predicate.redundant_stop_gradient_eliminater 0.26% : 0.000001s : 6: predicate.remove_not_recompute_node 1.18% : 0.000003s : 31: predicate.replace_applicator 0.25% : 0.000001s : 6: predicate.replace_old_param 0.14% : 0.000000s : 3: predicate.reset_defer_inline 1.18% : 0.000003s : 22: predicate.reshape_eliminate 0.38% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.23% : 0.000001s : 3: predicate.row_tensor_eliminate 0.39% : 0.000001s : 6: predicate.same_eliminate 0.23% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.47% : 0.000001s : 6: predicate.shard_identity_eliminate 0.42% : 0.000001s : 6: predicate.special_op_eliminate 0.41% : 0.000001s : 6: predicate.specialize_transform 0.51% : 0.000001s : 6: predicate.split_environ_get_set_with_tuple_value 0.54% : 0.000002s : 6: predicate.stack_unstack_eliminate 0.19% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.20% : 0.000006s : 40: predicate.switch_defer_inline 2.65% : 0.000007s : 46: predicate.switch_layer_defer_inline 7.12% : 0.000020s : 126: predicate.switch_simplify 1.20% : 0.000003s : 22: predicate.tile_eliminate 1.19% : 0.000003s : 22: predicate.transpose_eliminate 1.73% : 0.000005s : 29: predicate.tuple_list_convert_item_index_to_positive 1.67% : 0.000005s : 29: predicate.tuple_list_get_item_const_eliminator 1.63% : 0.000005s : 29: predicate.tuple_list_get_item_depend_reorder 2.61% : 0.000007s : 37: predicate.tuple_list_get_item_eliminator 1.53% : 0.000004s : 29: predicate.tuple_list_get_set_item_eliminator 2.17% : 0.000006s : 35: predicate.tuple_list_set_item_eliminator 1.58% : 0.000004s : 31: predicate.tuple_to_list_eliminator_ 2.66% : 0.000007s : 53: predicate.updatestate_pure_node_eliminater 3.11% : 0.000009s : 59: predicate.updatestate_useless_node_eliminater 0.24% : 0.000001s : 3: predicate.value_based_eliminate 0.37% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.37% : 0.000001s : 6: predicate.virtual_output_eliminate 0.16% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.26% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.001665 27 46.20% : 0.000769s : 10: func_graph_cloner_run.FuncGraphClonerGraph 53.80% : 0.000896s : 17: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.076555 192 0.00% : 0.000003s : 1: ForceFp32Comm 4.34% : 0.003323s : 1: add_attr 4.33% : 0.003314s : 1: add_attr_with_inline 0.01% : 0.000004s : 1: add_comm_op_reuse_tag 0.06% : 0.000049s : 1: add_recomputation 0.01% : 0.000004s : 1: assign_add_opt 0.19% : 0.000144s : 1: auto_monad 0.03% : 0.000022s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000005s : 1: bias_add_comm_swap 0.55% : 0.000417s : 1: bootstrap 0.04% : 0.000027s : 1: cconv 0.01% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000016s : 1: control_data_broadcast_order 0.01% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000025s : 1: cse_after_recomputation 0.01% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000005s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.37% : 0.000284s : 1: event_method 0.01% : 0.000006s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000014s : 1: graph_reusing 0.01% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.01% : 0.000004s : 1: handle_group_info 0.01% : 0.000006s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.01% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000007s : 1: label_micro_interleaved_index 0.59% : 0.000448s : 1: loop_unroll 0.01% : 0.000004s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.72% : 0.000551s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.02% : 0.000013s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000013s : 1: opt.transform.mutable_eliminate 2.33% : 0.001787s : 78: opt.transform.opt_a 0.03% : 0.000024s : 1: opt.transform.opt_after_cconv 0.03% : 0.000022s : 1: opt.transform.opt_after_jit_grad 0.11% : 0.000083s : 28: opt.transform.opt_b 0.06% : 0.000042s : 2: opt.transform.opt_trans_graph 0.04% : 0.000031s : 4: opt.transform.symbol_engine_opt 5.62% : 0.004304s : 1: opt_a 0.13% : 0.000097s : 1: opt_after_cconv 0.66% : 0.000508s : 1: opt_after_jit_grad 0.24% : 0.000184s : 1: opt_b 8.44% : 0.006458s : 1: optimize 0.02% : 0.000018s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.03% : 0.000020s : 1: overlap_grad_flash_sp 0.01% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000006s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.01% : 0.000006s : 1: overlap_param_gather 0.01% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000006s : 1: parallel-infer-symbol 0.00% : 0.000003s : 1: parallel-infer-symbol-second 0.01% : 0.000005s : 1: partial_unused_args_eliminate 0.01% : 0.000008s : 1: pipeline_parallel_scheduler 0.01% : 0.000004s : 1: pipeline_split 0.08% : 0.000062s : 1: pre_auto_parallel 0.01% : 0.000008s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000017s : 1: remove_dup_value 1.21% : 0.000928s : 1: renormalize.infer 1.02% : 0.000781s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000019s : 1: rewriter_after_opt_a 0.39% : 0.000299s : 1: rewriter_before_opt_a 0.01% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.01% : 0.000005s : 1: slice_recompute_activation 0.01% : 0.000004s : 1: split_layernorm_comm 0.01% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.09% : 0.000069s : 1: symbol_engine_optimizer 0.09% : 0.000072s : 1: tuple_transform 67.69% : 0.051819s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:45:33.323.263 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.08353, [21] [bootstrap]: 0.00036799 [type_inference]: 0.069405 [event_method]: 8.615e-05 [auto_monad]: 0.00018474 [graph_reusing]: 1.241e-05 [inline]: 2.93e-06 [add_attr]: 0.00344264, [1] [add_attr_with_inline]: 0.00343363, [1] [Cycle 1]: 0.00010165, [2] [tag_attr]: 5.32e-05 [meta_addattr_fg_expand]: 1.51e-05 [parallel-infer-symbol]: 3.35e-06 [pre_auto_parallel]: 7.547e-05 [insert-virtual-dataset]: 2.22999e-06 [parallel-infer-symbol-second]: 1.24e-06 [dataset_repeat_opt]: 1.96998e-06 [pipeline_split]: 1.62999e-06 [optimize]: 0.00918605, [53] [py_interpret_to_execute]: 5.07e-06 [rewriter_before_opt_a]: 0.00042055 [opt_a]: 0.00644006, [2] [Cycle 1]: 0.00561262, [45] [expand_dump_flag]: 6.83998e-06 [switch_simplify]: 0.00023528 [loop_unroll]: 8.461e-05 [a_1]: 0.00177896 [with_stream_mark]: 2.251e-05 [recompute_prepare]: 1.35e-05 [updatestate_depend_eliminate]: 5.84e-06 [updatestate_assign_eliminate]: 4.79e-06 [updatestate_loads_eliminate]: 4.40999e-06 [parameter_eliminate]: 2.36e-06 [a_2]: 0.00011315 [accelerated_algorithm]: 9.89001e-06 [shard]: 1.84e-06 [meta_shard_fg_expand]: 3.97998e-06 [shard_inline]: 8.62998e-06 [merge_send_recv]: 9.62001e-06 [auto_parallel]: 9.86e-06 [parallel]: 1.854e-05 [flash_sp]: 1.095e-05 [merge_comm]: 5.16998e-06 [allreduce_fusion]: 4.83001e-06 [matmul_add_comm_reduction]: 1.092e-05 [allreduce_slice_to_reducescatter]: 6.89994e-07 [virtual_shard_identity]: 9.89999e-06 [virtual_dataset]: 7.78999e-06 [get_grad_eliminate_]: 7.8e-06 [virtual_output]: 7.58001e-06 [merge_forward]: 5.00999e-06 [cell_reuse_recompute_pass]: 1.32999e-06 [offload_activation]: 1.009e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.791e-05 [merge_recompute_call_nodes]: 1.45999e-06 [before_grad]: 1.377e-05 [set_forward_comm_id_for_comm_node_pass]: 5.14e-06 [meta_fg_expand]: 4.90999e-06 [flash_sp_send_recv_attached]: 3.01001e-06 [receive_attached]: 2.12001e-06 [after_resolve]: 1.135e-05 [a_after_grad]: 1.204e-05 [renormalize]: 0.00270256 [add_forward_monad_depend]: 7.68999e-06 [auto_monad_grad]: 2.86e-06 [auto_monad_eliminator]: 2.095e-05 [cse]: 4.543e-05 [a_3]: 6.701e-05 [Cycle 2]: 0.00081348, [45] [expand_dump_flag]: 1.83002e-06 [switch_simplify]: 9.96998e-06 [loop_unroll]: 7.80998e-06 [a_1]: 0.00020096 [with_stream_mark]: 1.659e-05 [recompute_prepare]: 8.38999e-06 [updatestate_depend_eliminate]: 4.68999e-06 [updatestate_assign_eliminate]: 3.94002e-06 [updatestate_loads_eliminate]: 4.1e-06 [parameter_eliminate]: 1.77999e-06 [a_2]: 9.691e-05 [accelerated_algorithm]: 7.66999e-06 [shard]: 1.91998e-06 [meta_shard_fg_expand]: 2.36998e-06 [shard_inline]: 7.48e-06 [merge_send_recv]: 8.70001e-06 [auto_parallel]: 8.88002e-06 [parallel]: 7.17002e-06 [flash_sp]: 4.08999e-06 [merge_comm]: 4.45e-06 [allreduce_fusion]: 4.53999e-06 [matmul_add_comm_reduction]: 9.29998e-06 [allreduce_slice_to_reducescatter]: 1.09003e-06 [virtual_shard_identity]: 8.69003e-06 [virtual_dataset]: 7.49002e-06 [get_grad_eliminate_]: 7.00002e-06 [virtual_output]: 7.03e-06 [merge_forward]: 4.72e-06 [cell_reuse_recompute_pass]: 2.47001e-06 [offload_activation]: 9.41998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.456e-05 [merge_recompute_call_nodes]: 1.42999e-06 [before_grad]: 1.283e-05 [set_forward_comm_id_for_comm_node_pass]: 5.38002e-06 [meta_fg_expand]: 3.26001e-06 [flash_sp_send_recv_attached]: 1.37e-06 [receive_attached]: 1.79998e-06 [after_resolve]: 1.173e-05 [a_after_grad]: 1.103e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.41998e-06 [auto_monad_grad]: 1.79998e-06 [auto_monad_eliminator]: 9.56e-06 [cse]: 2.51e-05 [a_3]: 4.7e-05 [py_interpret_to_execute_after_opt_a]: 5.82001e-06 [slice_cell_reuse_recomputed_activation]: 1.84e-06 [rewriter_after_opt_a]: 2.79e-05 [convert_after_rewriter]: 1.29e-06 [order_py_execute_after_rewriter]: 1.09003e-06 [mutable_eliminate]: 0.00066477 [opt_b]: 0.00025586, [1] [Cycle 1]: 0.00024818, [7] [b_1]: 0.00016009 [b_2]: 9.34998e-06 [updatestate_depend_eliminate]: 6.83998e-06 [updatestate_assign_eliminate]: 3.43999e-06 [updatestate_loads_eliminate]: 3.38e-06 [renormalize]: 6.30011e-07 [cse]: 3.028e-05 [optimize_parallel_all_gather_comm]: 1.855e-05 [overlap_param_gather]: 2.12001e-06 [cconv]: 2.938e-05 [loop_unroll]: 0.00052777 [opt_after_cconv]: 0.00012603, [1] [Cycle 1]: 0.00012004, [7] [c_1]: 3.784e-05 [parameter_eliminate]: 4.10998e-06 [updatestate_depend_eliminate]: 7.04001e-06 [updatestate_assign_eliminate]: 3.43e-06 [updatestate_loads_eliminate]: 4.08001e-06 [cse]: 3.034e-05 [renormalize]: 5.89993e-07 [remove_dup_value]: 4.705e-05 [tuple_transform]: 8.857e-05, [1] [Cycle 1]: 8.296e-05, [4] [d_1]: 5.472e-05 [none_parameter_eliminate]: 1.69e-06 [renormalize]: 1.60013e-07 [switch_simplify]: 8.53001e-06 [partial_unused_args_eliminate]: 1.79e-06 [add_recomputation]: 5.897e-05 [cse_after_recomputation]: 2.961e-05, [1] [Cycle 1]: 2.454e-05, [1] [cse]: 1.917e-05 [environ_conv]: 6.59001e-06 [swap_dp_allreduce_reducescatter]: 7.1e-06 [bias_add_comm_swap]: 2.69999e-06 [label_micro_interleaved_index]: 4.85001e-06 [label_fine_grained_interleaved_index]: 2.45002e-06 [merge_cast_opt]: 2.06e-06 [slice_recompute_activation]: 2.17999e-06 [micro_interleaved_order_control]: 2.45002e-06 [assign_add_opt]: 1.27e-06 [ForceFp32Comm]: 6.90023e-07 [remove_cast_before_assign_add]: 8.99978e-07 [full_micro_interleaved_order_control]: 2.16998e-06 [reorder_send_recv_between_fp_bp]: 2.74999e-06 [comm_op_add_attrs]: 1.02e-06 [add_comm_op_reuse_tag]: 9.80013e-07 [interleave_split_concat_branches]: 1.15001e-06 [interleave_parallel_branches]: 1.19e-06 [overlap_opt_shard_in_pipeline]: 1.30001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.82001e-06 [control_data_broadcast_order]: 1.551e-05 [grouped_pairwise_exchange_alltoall]: 1.35999e-06 [offloading_packed_experts]: 5.39e-06 [overlap_recompute_and_grad_model_parallel]: 5.64998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.13001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.48002e-06 [overlap_recompute_comm]: 2.45002e-06 [overlap_grad_ring_attention]: 4.78001e-06 [overlap_grad_flash_sp]: 2.427e-05 [begin_end_overlap_inline]: 8.50006e-07 [split_matmul_comm_elemetwise]: 2.37001e-06 [split_layernorm_comm]: 1.54e-06 [handle_group_info]: 1.43002e-06 [symbol_engine_optimizer]: 8.69e-05, [1] [Cycle 1]: 8.21e-05, [6] [build]: 3.98001e-06 [elim_shapecalc]: 1.313e-05 [elim_not_effective]: 1.599e-05 [opt_reshape]: 8.54998e-06 [fold_const_symbol]: 1.394e-05 [renormalize]: 2.09984e-07 [detach_backward]: 2.02999e-06 [pipeline_parallel_scheduler]: 1.37999e-06 [auto_monad_reorder]: 2.227e-05 [get_jit_bprop_graph]: 1.49e-06 [rewriter_after_jit_bprop_graph]: 3.65e-06 [opt_after_jit_grad]: 0.00049915 [validate]: 4.333e-05 Sums bootstrap : 0.000368s : 0.47% type_inference : 0.069405s : 87.73% event_method : 0.000086s : 0.11% auto_monad : 0.000185s : 0.23% graph_reusing : 0.000012s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000053s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000015s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000075s : 0.10% 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 : 0.000005s : 0.01% optimize.rewriter_before_opt_a : 0.000421s : 0.53% optimize.opt_a.expand_dump_flag : 0.000009s : 0.01% optimize.opt_a.switch_simplify : 0.000245s : 0.31% optimize.opt_a.loop_unroll : 0.000092s : 0.12% optimize.opt_a.a_1 : 0.001980s : 2.50% optimize.opt_a.with_stream_mark : 0.000039s : 0.05% optimize.opt_a.recompute_prepare : 0.000022s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000009s : 0.01% optimize.opt_a.parameter_eliminate : 0.000004s : 0.01% optimize.opt_a.a_2 : 0.000210s : 0.27% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.02% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000006s : 0.01% optimize.opt_a.shard_inline : 0.000016s : 0.02% optimize.opt_a.merge_send_recv : 0.000018s : 0.02% optimize.opt_a.auto_parallel : 0.000019s : 0.02% optimize.opt_a.parallel : 0.000026s : 0.03% optimize.opt_a.flash_sp : 0.000015s : 0.02% optimize.opt_a.merge_comm : 0.000010s : 0.01% optimize.opt_a.allreduce_fusion : 0.000009s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000019s : 0.02% optimize.opt_a.virtual_dataset : 0.000015s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000015s : 0.02% optimize.opt_a.virtual_output : 0.000015s : 0.02% optimize.opt_a.merge_forward : 0.000010s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000020s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000032s : 0.04% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000027s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.01% optimize.opt_a.meta_fg_expand : 0.000008s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.01% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000023s : 0.03% optimize.opt_a.a_after_grad : 0.000023s : 0.03% optimize.opt_a.renormalize : 0.002703s : 3.42% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.01% optimize.opt_a.auto_monad_grad : 0.000005s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000031s : 0.04% optimize.opt_a.cse : 0.000071s : 0.09% optimize.opt_a.a_3 : 0.000114s : 0.14% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000028s : 0.04% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000665s : 0.84% optimize.opt_b.b_1 : 0.000160s : 0.20% optimize.opt_b.b_2 : 0.000009s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000030s : 0.04% optimize.optimize_parallel_all_gather_comm : 0.000019s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000029s : 0.04% optimize.loop_unroll : 0.000528s : 0.67% optimize.opt_after_cconv.c_1 : 0.000038s : 0.05% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.01% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000007s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.01% optimize.opt_after_cconv.cse : 0.000030s : 0.04% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000047s : 0.06% optimize.tuple_transform.d_1 : 0.000055s : 0.07% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000009s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000059s : 0.07% optimize.cse_after_recomputation.cse : 0.000019s : 0.02% optimize.environ_conv : 0.000007s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000007s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.01% optimize.label_fine_grained_interleaved_index : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000016s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 0.01% 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 : 0.000005s : 0.01% optimize.overlap_grad_flash_sp : 0.000024s : 0.03% 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 : 0.000004s : 0.01% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000013s : 0.02% optimize.symbol_engine_optimizer.elim_not_effective : 0.000016s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000014s : 0.02% 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 : 0.000022s : 0.03% get_jit_bprop_graph : 0.000001s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000499s : 0.63% validate : 0.000043s : 0.05% Time group info: ------[substitution.] 0.000576 75 8.69% : 0.000050s : 11: substitution.cast_eliminate 0.51% : 0.000003s : 4: substitution.elim_not_effective 0.36% : 0.000002s : 4: substitution.fold_const_symbol 1.18% : 0.000007s : 5: substitution.graph_param_transform 81.65% : 0.000470s : 24: substitution.inline 0.97% : 0.000006s : 8: substitution.j_node_and_user_rematch 1.24% : 0.000007s : 8: substitution.remove_not_recompute_node 0.78% : 0.000004s : 2: substitution.replace_old_param 2.97% : 0.000017s : 7: substitution.switch_simplify 1.66% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.069252 2 93.03% : 0.064427s : 1: type_inference.infer 6.97% : 0.004825s : 1: type_inference.specialize ------[replace.] 0.000243 33 60.98% : 0.000148s : 24: replace.inline 33.06% : 0.000080s : 7: replace.switch_simplify 5.96% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000479 33 95.54% : 0.000458s : 24: match.inline 2.76% : 0.000013s : 7: match.switch_simplify 1.71% : 0.000008s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000404 2544 1.23% : 0.000005s : 32: predicate.accumulaten_eliminater 0.53% : 0.000002s : 5: predicate.ad_related_special_op_eliminate 0.36% : 0.000001s : 10: predicate.addn_check_dump 1.21% : 0.000005s : 32: predicate.addn_zero_filter 1.05% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.17% : 0.000009s : 42: predicate.arithmetic_simplify 1.46% : 0.000006s : 32: predicate.cast_eliminate 0.37% : 0.000002s : 10: predicate.check_bprop_eliminate 0.36% : 0.000001s : 10: predicate.compare_switch_simplify 0.11% : 0.000000s : 5: predicate.const_output_eliminate 0.39% : 0.000002s : 10: predicate.depend_value_elim 1.20% : 0.000005s : 32: predicate.dict_get_item_const_eliminator 1.29% : 0.000005s : 32: predicate.dict_get_item_eliminator 1.12% : 0.000005s : 32: predicate.dict_set_item_eliminator 0.56% : 0.000002s : 10: predicate.dumpgradient_eliminate 0.13% : 0.000001s : 5: predicate.elim_not_effective 0.26% : 0.000001s : 5: predicate.elim_shapecalc_of_broadcastargs 1.33% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.19% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.26% : 0.000005s : 37: predicate.environ_get_depend_swap 1.73% : 0.000007s : 47: predicate.environ_get_eliminate 1.28% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.02% : 0.000008s : 58: predicate.exchange_switch_depend_value 2.91% : 0.000012s : 58: predicate.float_depend_g_call 0.37% : 0.000001s : 10: predicate.float_environ_get_switch 0.52% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.10% : 0.000000s : 5: predicate.fold_const_symbol 0.41% : 0.000002s : 10: predicate.get_grad_eliminate 0.11% : 0.000000s : 5: predicate.graph_param_transform 0.37% : 0.000001s : 10: predicate.incorporate_call 0.32% : 0.000001s : 10: predicate.incorporate_call_switch 6.12% : 0.000025s : 120: predicate.inline 0.45% : 0.000002s : 10: predicate.inline_without_move 0.20% : 0.000001s : 10: predicate.j_node_and_user_rematch 0.60% : 0.000002s : 10: predicate.less_batch_normalization 1.72% : 0.000007s : 44: predicate.list_to_tuple_eliminator_ 2.52% : 0.000010s : 76: predicate.load_eliminater 0.61% : 0.000002s : 5: predicate.loop_unroll_after_grad 3.65% : 0.000015s : 104: predicate.loop_unroll_before_grad 1.59% : 0.000006s : 42: predicate.make_slice_get_slice_eliminator 0.39% : 0.000002s : 10: predicate.merge_addn 0.37% : 0.000002s : 10: predicate.micro_step_allgather_replace 0.44% : 0.000002s : 10: predicate.mini_step_allgather_replace 1.09% : 0.000004s : 32: predicate.minmaximum_grad 0.55% : 0.000002s : 5: predicate.mutable_eliminate 0.25% : 0.000001s : 5: predicate.opt_reshape 0.23% : 0.000001s : 5: predicate.parallel_virtual_node 2.81% : 0.000011s : 58: predicate.partial_defer_inline 1.53% : 0.000006s : 39: predicate.partial_eliminate 1.09% : 0.000004s : 32: predicate.print_const_string_wrapper 0.39% : 0.000002s : 10: predicate.reduce_all_const_elim 1.56% : 0.000006s : 32: predicate.reduce_eliminate 2.59% : 0.000010s : 76: predicate.redundant_stop_gradient_eliminater 0.26% : 0.000001s : 10: predicate.remove_not_recompute_node 1.14% : 0.000005s : 44: predicate.replace_applicator 0.23% : 0.000001s : 10: predicate.replace_old_param 0.15% : 0.000001s : 5: predicate.reset_defer_inline 1.21% : 0.000005s : 32: predicate.reshape_eliminate 0.40% : 0.000002s : 10: predicate.row_tensor_add_zeros_like 0.23% : 0.000001s : 5: predicate.row_tensor_eliminate 0.52% : 0.000002s : 10: predicate.same_eliminate 0.32% : 0.000001s : 10: predicate.set_cell_output_no_recompute 0.52% : 0.000002s : 10: predicate.shard_identity_eliminate 0.44% : 0.000002s : 10: predicate.special_op_eliminate 0.46% : 0.000002s : 10: predicate.specialize_transform 0.64% : 0.000003s : 10: predicate.split_environ_get_set_with_tuple_value 0.53% : 0.000002s : 10: predicate.stack_unstack_eliminate 0.22% : 0.000001s : 5: predicate.switch_call_monad_eliminater 2.23% : 0.000009s : 58: predicate.switch_defer_inline 2.57% : 0.000010s : 68: predicate.switch_layer_defer_inline 7.42% : 0.000030s : 191: predicate.switch_simplify 1.12% : 0.000005s : 32: predicate.tile_eliminate 1.22% : 0.000005s : 32: predicate.transpose_eliminate 1.65% : 0.000007s : 42: predicate.tuple_list_convert_item_index_to_positive 1.55% : 0.000006s : 42: predicate.tuple_list_get_item_const_eliminator 1.53% : 0.000006s : 42: predicate.tuple_list_get_item_depend_reorder 2.59% : 0.000010s : 54: predicate.tuple_list_get_item_eliminator 1.54% : 0.000006s : 42: predicate.tuple_list_get_set_item_eliminator 2.05% : 0.000008s : 52: predicate.tuple_list_set_item_eliminator 1.62% : 0.000007s : 44: predicate.tuple_to_list_eliminator_ 2.68% : 0.000011s : 76: predicate.updatestate_pure_node_eliminater 3.02% : 0.000012s : 86: predicate.updatestate_useless_node_eliminater 0.23% : 0.000001s : 5: predicate.value_based_eliminate 0.41% : 0.000002s : 10: predicate.virtual_dataset_eliminate 0.39% : 0.000002s : 10: predicate.virtual_output_eliminate 0.16% : 0.000001s : 5: predicate.virtual_view_grad_eliminate 0.25% : 0.000001s : 5: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003827 42 64.11% : 0.002453s : 16: func_graph_cloner_run.FuncGraphClonerGraph 35.89% : 0.001374s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.101827 192 0.00% : 0.000003s : 1: ForceFp32Comm 3.39% : 0.003447s : 1: add_attr 3.38% : 0.003438s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.06% : 0.000063s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.19% : 0.000193s : 1: auto_monad 0.03% : 0.000026s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000005s : 1: bias_add_comm_swap 0.38% : 0.000391s : 1: bootstrap 0.03% : 0.000033s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000019s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000032s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.01% : 0.000010s : 1: environ_conv 0.09% : 0.000095s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000016s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000006s : 1: inline 0.01% : 0.000005s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000008s : 1: label_micro_interleaved_index 0.53% : 0.000538s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.66% : 0.000676s : 1: mutable_eliminate 0.01% : 0.000008s : 1: offloading_packed_experts 0.02% : 0.000018s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000018s : 1: opt.transform.mutable_eliminate 2.72% : 0.002773s : 78: opt.transform.opt_a 0.04% : 0.000036s : 1: opt.transform.opt_after_cconv 0.03% : 0.000031s : 1: opt.transform.opt_after_jit_grad 0.14% : 0.000142s : 28: opt.transform.opt_b 0.06% : 0.000061s : 2: opt.transform.opt_trans_graph 0.05% : 0.000048s : 4: opt.transform.symbol_engine_opt 6.33% : 0.006444s : 1: opt_a 0.13% : 0.000130s : 1: opt_after_cconv 0.50% : 0.000509s : 1: opt_after_jit_grad 0.25% : 0.000259s : 1: opt_b 9.03% : 0.009192s : 1: optimize 0.02% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.03% : 0.000027s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000008s : 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 0.01% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000009s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000008s : 1: parallel-infer-symbol 0.01% : 0.000005s : 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 0.08% : 0.000081s : 1: pre_auto_parallel 0.01% : 0.000009s : 1: py_interpret_to_execute 0.01% : 0.000009s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.05% : 0.000052s : 1: remove_dup_value 1.44% : 0.001466s : 1: renormalize.infer 1.20% : 0.001226s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.03% : 0.000032s : 1: rewriter_after_opt_a 0.42% : 0.000429s : 1: rewriter_before_opt_a 0.00% : 0.000004s : 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% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.09% : 0.000090s : 1: symbol_engine_optimizer 0.09% : 0.000091s : 1: tuple_transform 68.18% : 0.069428s : 1: type_inference TotalTime = 0.110898, [21] [bootstrap]: 0.00043031 [type_inference]: 0.0875614 [event_method]: 7.653e-05 [auto_monad]: 0.00025927 [graph_reusing]: 2.034e-05 [inline]: 2.21e-06 [add_attr]: 0.00361595, [1] [add_attr_with_inline]: 0.00360658, [1] [Cycle 1]: 0.00012695, [2] [tag_attr]: 7.119e-05 [meta_addattr_fg_expand]: 2.153e-05 [parallel-infer-symbol]: 3.3e-06 [pre_auto_parallel]: 9.982e-05 [insert-virtual-dataset]: 2.56e-06 [parallel-infer-symbol-second]: 1.27999e-06 [dataset_repeat_opt]: 2.16e-06 [pipeline_split]: 1.72001e-06 [optimize]: 0.0180952, [53] [py_interpret_to_execute]: 6.33e-06 [rewriter_before_opt_a]: 0.00056158 [opt_a]: 0.0156146, [2] [Cycle 1]: 0.014561, [45] [expand_dump_flag]: 1.176e-05 [switch_simplify]: 0.00031204 [loop_unroll]: 0.00012605 [a_1]: 0.00272814 [with_stream_mark]: 3.089e-05 [recompute_prepare]: 2.609e-05 [updatestate_depend_eliminate]: 9.73002e-06 [updatestate_assign_eliminate]: 8.40001e-06 [updatestate_loads_eliminate]: 8.28001e-06 [parameter_eliminate]: 3.03e-06 [a_2]: 0.00024856 [accelerated_algorithm]: 1.716e-05 [shard]: 1.74e-06 [meta_shard_fg_expand]: 7.87e-06 [shard_inline]: 1.621e-05 [merge_send_recv]: 1.894e-05 [auto_parallel]: 1.284e-05 [parallel]: 1.902e-05 [flash_sp]: 1.137e-05 [merge_comm]: 9.56e-06 [allreduce_fusion]: 9.37001e-06 [matmul_add_comm_reduction]: 3.024e-05 [allreduce_slice_to_reducescatter]: 6.19999e-07 [virtual_shard_identity]: 1.953e-05 [virtual_dataset]: 1.651e-05 [get_grad_eliminate_]: 1.504e-05 [virtual_output]: 1.639e-05 [merge_forward]: 1.006e-05 [cell_reuse_recompute_pass]: 1.17999e-06 [offload_activation]: 1.877e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.301e-05 [merge_recompute_call_nodes]: 1.51998e-06 [before_grad]: 2.781e-05 [set_forward_comm_id_for_comm_node_pass]: 9.84001e-06 [meta_fg_expand]: 0.00315404 [flash_sp_send_recv_attached]: 4.62e-06 [receive_attached]: 2.26003e-06 [after_resolve]: 0.00010614 [a_after_grad]: 0.00015121 [renormalize]: 0.00684981 [add_forward_monad_depend]: 5.97999e-06 [auto_monad_grad]: 2.69001e-06 [auto_monad_eliminator]: 1.381e-05 [cse]: 2.67e-05 [a_3]: 3.916e-05 [Cycle 2]: 0.00103888, [45] [expand_dump_flag]: 2.09e-06 [switch_simplify]: 5.82999e-06 [loop_unroll]: 4.99e-06 [a_1]: 5.134e-05 [with_stream_mark]: 1.196e-05 [recompute_prepare]: 4.33001e-06 [updatestate_depend_eliminate]: 2.99999e-06 [updatestate_assign_eliminate]: 2.12999e-06 [updatestate_loads_eliminate]: 2.21e-06 [parameter_eliminate]: 1.23002e-06 [a_2]: 3.796e-05 [accelerated_algorithm]: 3.63999e-06 [shard]: 1.77999e-06 [meta_shard_fg_expand]: 1.72999e-06 [shard_inline]: 3.44001e-06 [merge_send_recv]: 5.49e-06 [auto_parallel]: 6.60002e-06 [parallel]: 6.96001e-06 [flash_sp]: 3.71999e-06 [merge_comm]: 2.48998e-06 [allreduce_fusion]: 2.27001e-06 [matmul_add_comm_reduction]: 5.77001e-06 [allreduce_slice_to_reducescatter]: 4.00003e-07 [virtual_shard_identity]: 4.3e-06 [virtual_dataset]: 3.7e-06 [get_grad_eliminate_]: 3.68e-06 [virtual_output]: 3.45e-06 [merge_forward]: 3.37002e-06 [cell_reuse_recompute_pass]: 1.74998e-06 [offload_activation]: 6.97002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.044e-05 [merge_recompute_call_nodes]: 1.10999e-06 [before_grad]: 6.44001e-06 [set_forward_comm_id_for_comm_node_pass]: 2.56998e-06 [meta_fg_expand]: 0.00054103 [flash_sp_send_recv_attached]: 1.89e-06 [receive_attached]: 2.06e-06 [after_resolve]: 1.078e-05 [a_after_grad]: 6.04999e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.79001e-06 [auto_monad_grad]: 1.22e-06 [auto_monad_eliminator]: 7.75e-06 [cse]: 1.419e-05 [a_3]: 2.399e-05 [py_interpret_to_execute_after_opt_a]: 5.51e-06 [slice_cell_reuse_recomputed_activation]: 2.12001e-06 [rewriter_after_opt_a]: 1.75e-05 [convert_after_rewriter]: 1.25999e-06 [order_py_execute_after_rewriter]: 1.10999e-06 [mutable_eliminate]: 0.00060625 [opt_b]: 0.00014042, [1] [Cycle 1]: 0.00013339, [7] [b_1]: 7.252e-05 [b_2]: 4.82e-06 [updatestate_depend_eliminate]: 4.25e-06 [updatestate_assign_eliminate]: 1.68002e-06 [updatestate_loads_eliminate]: 1.64e-06 [renormalize]: 7.89994e-07 [cse]: 1.489e-05 [optimize_parallel_all_gather_comm]: 1.444e-05 [overlap_param_gather]: 2.21e-06 [cconv]: 2.208e-05 [loop_unroll]: 0.00045548 [opt_after_cconv]: 7.825e-05, [1] [Cycle 1]: 7.256e-05, [7] [c_1]: 1.626e-05 [parameter_eliminate]: 2.31e-06 [updatestate_depend_eliminate]: 4.15e-06 [updatestate_assign_eliminate]: 1.76003e-06 [updatestate_loads_eliminate]: 1.52001e-06 [cse]: 1.447e-05 [renormalize]: 3.50003e-07 [remove_dup_value]: 1.331e-05 [tuple_transform]: 5.122e-05, [1] [Cycle 1]: 4.667e-05, [4] [d_1]: 2.363e-05 [none_parameter_eliminate]: 1.46002e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 4.18999e-06 [partial_unused_args_eliminate]: 1.67001e-06 [add_recomputation]: 3.071e-05 [cse_after_recomputation]: 1.788e-05, [1] [Cycle 1]: 1.335e-05, [1] [cse]: 8.15e-06 [environ_conv]: 4.94e-06 [swap_dp_allreduce_reducescatter]: 3.68e-06 [bias_add_comm_swap]: 2.27999e-06 [label_micro_interleaved_index]: 4.05998e-06 [label_fine_grained_interleaved_index]: 2.51998e-06 [merge_cast_opt]: 1.32e-06 [slice_recompute_activation]: 1.98002e-06 [micro_interleaved_order_control]: 2.17999e-06 [assign_add_opt]: 1.28002e-06 [ForceFp32Comm]: 8.70001e-07 [remove_cast_before_assign_add]: 1.27999e-06 [full_micro_interleaved_order_control]: 2.13998e-06 [reorder_send_recv_between_fp_bp]: 2.88e-06 [comm_op_add_attrs]: 9.49978e-07 [add_comm_op_reuse_tag]: 9.80013e-07 [interleave_split_concat_branches]: 1.12e-06 [interleave_parallel_branches]: 1.08001e-06 [overlap_opt_shard_in_pipeline]: 1.10001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.91e-06 [control_data_broadcast_order]: 8.49002e-06 [grouped_pairwise_exchange_alltoall]: 1.44e-06 [offloading_packed_experts]: 2.89001e-06 [overlap_recompute_and_grad_model_parallel]: 4.33999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.10999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.29e-06 [overlap_recompute_comm]: 2.34001e-06 [overlap_grad_ring_attention]: 3.05002e-06 [overlap_grad_flash_sp]: 1.57e-05 [begin_end_overlap_inline]: 4.89992e-07 [split_matmul_comm_elemetwise]: 2.14e-06 [split_layernorm_comm]: 1.60001e-06 [handle_group_info]: 9.60019e-07 [symbol_engine_optimizer]: 6.009e-05, [1] [Cycle 1]: 5.577e-05, [6] [build]: 2.53e-06 [elim_shapecalc]: 6.73e-06 [elim_not_effective]: 9.39998e-06 [opt_reshape]: 4.52e-06 [fold_const_symbol]: 6.43998e-06 [renormalize]: 1.8999e-07 [detach_backward]: 2.10002e-06 [pipeline_parallel_scheduler]: 1.56998e-06 [auto_monad_reorder]: 1.541e-05 [get_jit_bprop_graph]: 1.86e-06 [rewriter_after_jit_bprop_graph]: 4.13001e-06 [opt_after_jit_grad]: 0.00049384 [validate]: 3.135e-05 Sums bootstrap : 0.000430s : 0.41% type_inference : 0.087561s : 82.48% event_method : 0.000077s : 0.07% auto_monad : 0.000259s : 0.24% graph_reusing : 0.000020s : 0.02% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000071s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000022s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000100s : 0.09% 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 : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000562s : 0.53% optimize.opt_a.expand_dump_flag : 0.000014s : 0.01% optimize.opt_a.switch_simplify : 0.000318s : 0.30% optimize.opt_a.loop_unroll : 0.000131s : 0.12% optimize.opt_a.a_1 : 0.002779s : 2.62% optimize.opt_a.with_stream_mark : 0.000043s : 0.04% optimize.opt_a.recompute_prepare : 0.000030s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.01% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000287s : 0.27% optimize.opt_a.accelerated_algorithm : 0.000021s : 0.02% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.01% optimize.opt_a.shard_inline : 0.000020s : 0.02% optimize.opt_a.merge_send_recv : 0.000024s : 0.02% optimize.opt_a.auto_parallel : 0.000019s : 0.02% optimize.opt_a.parallel : 0.000026s : 0.02% optimize.opt_a.flash_sp : 0.000015s : 0.01% optimize.opt_a.merge_comm : 0.000012s : 0.01% optimize.opt_a.allreduce_fusion : 0.000012s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000036s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000024s : 0.02% optimize.opt_a.virtual_dataset : 0.000020s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000019s : 0.02% optimize.opt_a.virtual_output : 0.000020s : 0.02% optimize.opt_a.merge_forward : 0.000013s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000043s : 0.04% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000034s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.01% optimize.opt_a.meta_fg_expand : 0.003695s : 3.48% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.01% optimize.opt_a.receive_attached : 0.000004s : 0.00% optimize.opt_a.after_resolve : 0.000117s : 0.11% optimize.opt_a.a_after_grad : 0.000157s : 0.15% optimize.opt_a.renormalize : 0.006850s : 6.45% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000022s : 0.02% optimize.opt_a.cse : 0.000041s : 0.04% optimize.opt_a.a_3 : 0.000063s : 0.06% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000018s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000606s : 0.57% optimize.opt_b.b_1 : 0.000073s : 0.07% optimize.opt_b.b_2 : 0.000005s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000015s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000014s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000022s : 0.02% optimize.loop_unroll : 0.000455s : 0.43% optimize.opt_after_cconv.c_1 : 0.000016s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000014s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000024s : 0.02% optimize.tuple_transform.none_parameter_eliminate : 0.000001s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000031s : 0.03% optimize.cse_after_recomputation.cse : 0.000008s : 0.01% optimize.environ_conv : 0.000005s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000002s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000008s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000016s : 0.01% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000007s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000009s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000006s : 0.01% 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 : 0.000015s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000494s : 0.47% validate : 0.000031s : 0.03% Time group info: ------[substitution.] 0.000867 134 3.45% : 0.000030s : 8: substitution.cast_eliminate 0.25% : 0.000002s : 1: substitution.elim_not_effective 0.83% : 0.000007s : 9: substitution.float_depend_g_call 0.43% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.15% : 0.000001s : 1: substitution.fold_const_symbol 0.56% : 0.000005s : 1: substitution.graph_param_transform 0.37% : 0.000003s : 2: substitution.incorporate_call 0.24% : 0.000002s : 2: substitution.incorporate_call_switch 78.05% : 0.000677s : 33: substitution.inline 3.30% : 0.000029s : 4: substitution.inline_without_move 0.95% : 0.000008s : 11: substitution.j_node_and_user_rematch 0.43% : 0.000004s : 2: substitution.minmaximum_grad 0.77% : 0.000007s : 9: substitution.partial_eliminate 1.07% : 0.000009s : 11: substitution.remove_not_recompute_node 0.26% : 0.000002s : 1: substitution.replace_applicator 0.85% : 0.000007s : 12: substitution.replace_old_param 0.32% : 0.000003s : 1: substitution.set_cell_output_no_recompute 2.65% : 0.000023s : 10: substitution.switch_simplify 0.88% : 0.000008s : 2: substitution.tuple_list_convert_item_index_to_positive 0.36% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.59% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 2.71% : 0.000023s : 6: substitution.tuple_list_get_item_eliminator 0.52% : 0.000005s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.087370 2 92.61% : 0.080911s : 1: type_inference.infer 7.39% : 0.006459s : 1: type_inference.specialize ------[replace.] 0.000335 45 63.45% : 0.000212s : 33: replace.inline 32.02% : 0.000107s : 10: replace.switch_simplify 4.53% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000688 45 95.76% : 0.000659s : 33: match.inline 2.66% : 0.000018s : 10: match.switch_simplify 1.58% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000520 3189 1.17% : 0.000006s : 43: predicate.accumulaten_eliminater 0.20% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.41% : 0.000002s : 14: predicate.addn_check_dump 1.24% : 0.000006s : 43: predicate.addn_zero_filter 1.10% : 0.000006s : 43: predicate.adjust_all_reduce_mul_add 2.20% : 0.000011s : 57: predicate.arithmetic_simplify 1.48% : 0.000008s : 43: predicate.cast_eliminate 0.09% : 0.000000s : 2: predicate.check_bprop_eliminate 0.41% : 0.000002s : 14: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.42% : 0.000002s : 14: predicate.depend_value_elim 1.20% : 0.000006s : 43: predicate.dict_get_item_const_eliminator 1.39% : 0.000007s : 43: predicate.dict_get_item_eliminator 1.13% : 0.000006s : 43: predicate.dict_set_item_eliminator 0.18% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.03% : 0.000000s : 1: predicate.elim_not_effective 0.07% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.25% : 0.000006s : 44: predicate.environ_add_const_eliminate 1.14% : 0.000006s : 44: predicate.environ_get_add_eliminate 1.27% : 0.000007s : 44: predicate.environ_get_depend_swap 1.52% : 0.000008s : 58: predicate.environ_get_eliminate 1.18% : 0.000006s : 44: predicate.environ_get_set_eliminate 2.24% : 0.000012s : 78: predicate.exchange_switch_depend_value 3.16% : 0.000016s : 78: predicate.float_depend_g_call 0.38% : 0.000002s : 14: predicate.float_environ_get_switch 0.44% : 0.000002s : 15: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.45% : 0.000002s : 14: predicate.get_grad_eliminate 0.06% : 0.000000s : 1: predicate.graph_param_transform 0.39% : 0.000002s : 14: predicate.incorporate_call 0.37% : 0.000002s : 14: predicate.incorporate_call_switch 6.22% : 0.000032s : 151: predicate.inline 2.58% : 0.000013s : 70: predicate.inline_without_move 0.21% : 0.000001s : 14: predicate.j_node_and_user_rematch 0.59% : 0.000003s : 14: predicate.less_batch_normalization 1.49% : 0.000008s : 47: predicate.list_to_tuple_eliminator_ 2.44% : 0.000013s : 90: predicate.load_eliminater 0.24% : 0.000001s : 1: predicate.loop_unroll_after_grad 4.11% : 0.000021s : 141: predicate.loop_unroll_before_grad 1.42% : 0.000007s : 45: predicate.make_slice_get_slice_eliminator 0.42% : 0.000002s : 14: predicate.merge_addn 0.11% : 0.000001s : 2: predicate.micro_step_allgather_replace 0.10% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.30% : 0.000007s : 43: predicate.minmaximum_grad 0.26% : 0.000001s : 1: predicate.mutable_eliminate 0.05% : 0.000000s : 1: predicate.opt_reshape 0.08% : 0.000000s : 1: predicate.parallel_virtual_node 3.06% : 0.000016s : 78: predicate.partial_defer_inline 1.46% : 0.000008s : 46: predicate.partial_eliminate 1.17% : 0.000006s : 43: predicate.print_const_string_wrapper 0.46% : 0.000002s : 14: predicate.reduce_all_const_elim 1.52% : 0.000008s : 43: predicate.reduce_eliminate 2.54% : 0.000013s : 90: predicate.redundant_stop_gradient_eliminater 0.28% : 0.000001s : 14: predicate.remove_not_recompute_node 0.99% : 0.000005s : 47: predicate.replace_applicator 1.17% : 0.000006s : 70: predicate.replace_old_param 0.05% : 0.000000s : 1: predicate.reset_defer_inline 1.23% : 0.000006s : 43: predicate.reshape_eliminate 0.13% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.06% : 0.000000s : 1: predicate.row_tensor_eliminate 0.18% : 0.000001s : 2: predicate.same_eliminate 0.27% : 0.000001s : 14: predicate.set_cell_output_no_recompute 0.62% : 0.000003s : 14: predicate.shard_identity_eliminate 0.12% : 0.000001s : 2: predicate.special_op_eliminate 0.44% : 0.000002s : 14: predicate.specialize_transform 0.26% : 0.000001s : 2: predicate.split_environ_get_set_with_tuple_value 2.29% : 0.000012s : 70: predicate.stack_unstack_eliminate 0.07% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.46% : 0.000013s : 78: predicate.switch_defer_inline 2.60% : 0.000014s : 80: predicate.switch_layer_defer_inline 7.99% : 0.000042s : 254: predicate.switch_simplify 1.21% : 0.000006s : 43: predicate.tile_eliminate 1.23% : 0.000006s : 43: predicate.transpose_eliminate 1.40% : 0.000007s : 45: predicate.tuple_list_convert_item_index_to_positive 1.38% : 0.000007s : 45: predicate.tuple_list_get_item_const_eliminator 1.27% : 0.000007s : 45: predicate.tuple_list_get_item_depend_reorder 2.47% : 0.000013s : 61: predicate.tuple_list_get_item_eliminator 1.50% : 0.000008s : 45: predicate.tuple_list_get_set_item_eliminator 1.95% : 0.000010s : 59: predicate.tuple_list_set_item_eliminator 1.31% : 0.000007s : 47: predicate.tuple_to_list_eliminator_ 2.42% : 0.000013s : 90: predicate.updatestate_pure_node_eliminater 3.06% : 0.000016s : 104: predicate.updatestate_useless_node_eliminater 0.08% : 0.000000s : 1: predicate.value_based_eliminate 0.47% : 0.000002s : 14: predicate.virtual_dataset_eliminate 0.51% : 0.000003s : 14: predicate.virtual_output_eliminate 0.04% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.09% : 0.000000s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.004677 69 70.58% : 0.003301s : 33: func_graph_cloner_run.FuncGraphClonerGraph 29.42% : 0.001376s : 36: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.143376 192 0.00% : 0.000004s : 1: ForceFp32Comm 2.53% : 0.003621s : 1: add_attr 2.52% : 0.003611s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.02% : 0.000035s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.19% : 0.000270s : 1: auto_monad 0.01% : 0.000019s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.32% : 0.000455s : 1: bootstrap 0.02% : 0.000025s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000012s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.01% : 0.000021s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000008s : 1: environ_conv 0.06% : 0.000085s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000025s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000005s : 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% : 0.000005s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.32% : 0.000464s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.43% : 0.000616s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.01% : 0.000010s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000011s : 1: opt.transform.mutable_eliminate 2.78% : 0.003981s : 78: opt.transform.opt_a 0.01% : 0.000015s : 1: opt.transform.opt_after_cconv 0.01% : 0.000016s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000046s : 28: opt.transform.opt_b 0.02% : 0.000025s : 2: opt.transform.opt_trans_graph 0.02% : 0.000024s : 4: opt.transform.symbol_engine_opt 10.89% : 0.015618s : 1: opt_a 0.06% : 0.000081s : 1: opt_after_cconv 0.35% : 0.000504s : 1: opt_after_jit_grad 0.10% : 0.000144s : 1: opt_b 12.63% : 0.018102s : 1: optimize 0.01% : 0.000018s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000019s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000005s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.07% : 0.000106s : 1: pre_auto_parallel 0.01% : 0.000010s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000017s : 1: remove_dup_value 4.27% : 0.006127s : 1: renormalize.infer 0.50% : 0.000711s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000021s : 1: rewriter_after_opt_a 0.40% : 0.000570s : 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% : 0.000006s : 1: swap_dp_allreduce_reducescatter 0.04% : 0.000063s : 1: symbol_engine_optimizer 0.04% : 0.000054s : 1: tuple_transform 61.09% : 0.087588s : 1: type_inference [WARNING] ME(168422:281472943677232,MainProcess):2026-01-29-17:45:35.718.406 [mindspore/common/dtype.py:295] The interface 'mindspore.dtype_to_nptype' is deprecated from version 2.7 and will be removed in a future version. TotalTime = 0.0788852, [21] [bootstrap]: 0.00044685 [type_inference]: 0.0656028 [event_method]: 7.842e-05 [auto_monad]: 0.00017885 [graph_reusing]: 1.298e-05 [inline]: 2.20002e-06 [add_attr]: 0.00360567, [1] [add_attr_with_inline]: 0.00359683, [1] [Cycle 1]: 0.00010051, [2] [tag_attr]: 5.307e-05 [meta_addattr_fg_expand]: 1.477e-05 [parallel-infer-symbol]: 3.13e-06 [pre_auto_parallel]: 7.429e-05 [insert-virtual-dataset]: 2.63e-06 [parallel-infer-symbol-second]: 1.40001e-06 [dataset_repeat_opt]: 1.99e-06 [pipeline_split]: 1.64e-06 [optimize]: 0.00814513, [53] [py_interpret_to_execute]: 5.49e-06 [rewriter_before_opt_a]: 0.00038257 [opt_a]: 0.00580377, [2] [Cycle 1]: 0.00521926, [45] [expand_dump_flag]: 5.69e-06 [switch_simplify]: 0.0002098 [loop_unroll]: 9.4e-05 [a_1]: 0.00191526 [with_stream_mark]: 1.26e-05 [recompute_prepare]: 8.65999e-06 [updatestate_depend_eliminate]: 3.21999e-06 [updatestate_assign_eliminate]: 2.34999e-06 [updatestate_loads_eliminate]: 2.27999e-06 [parameter_eliminate]: 1.25999e-06 [a_2]: 7.082e-05 [accelerated_algorithm]: 6.39001e-06 [shard]: 1.49998e-06 [meta_shard_fg_expand]: 3.01001e-06 [shard_inline]: 6.01e-06 [merge_send_recv]: 7.46999e-06 [auto_parallel]: 6.06e-06 [parallel]: 1.794e-05 [flash_sp]: 6.89999e-06 [merge_comm]: 3.31999e-06 [allreduce_fusion]: 3.18e-06 [matmul_add_comm_reduction]: 9.50001e-06 [allreduce_slice_to_reducescatter]: 6.79982e-07 [virtual_shard_identity]: 7.18998e-06 [virtual_dataset]: 6.23e-06 [get_grad_eliminate_]: 5.74999e-06 [virtual_output]: 5.66998e-06 [merge_forward]: 2.96999e-06 [cell_reuse_recompute_pass]: 1.04998e-06 [offload_activation]: 8.92e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.235e-05 [merge_recompute_call_nodes]: 1.64e-06 [before_grad]: 8.78001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.34001e-06 [meta_fg_expand]: 4.17e-06 [flash_sp_send_recv_attached]: 2.31e-06 [receive_attached]: 2.21e-06 [after_resolve]: 8.84e-06 [a_after_grad]: 8.74e-06 [renormalize]: 0.00235649 [add_forward_monad_depend]: 5.02e-06 [auto_monad_grad]: 2.45002e-06 [auto_monad_eliminator]: 1.639e-05 [cse]: 5.18e-05 [a_3]: 4.605e-05 [Cycle 2]: 0.00057252, [45] [expand_dump_flag]: 1.34e-06 [switch_simplify]: 7.08e-06 [loop_unroll]: 5.57999e-06 [a_1]: 0.00010222 [with_stream_mark]: 1.239e-05 [recompute_prepare]: 5.88998e-06 [updatestate_depend_eliminate]: 3.18e-06 [updatestate_assign_eliminate]: 2.15002e-06 [updatestate_loads_eliminate]: 2.61999e-06 [parameter_eliminate]: 1.00001e-06 [a_2]: 6.25e-05 [accelerated_algorithm]: 5.25999e-06 [shard]: 1.14003e-06 [meta_shard_fg_expand]: 1.60001e-06 [shard_inline]: 5.22e-06 [merge_send_recv]: 4.97e-06 [auto_parallel]: 5.05999e-06 [parallel]: 5.97999e-06 [flash_sp]: 3.55998e-06 [merge_comm]: 2.74999e-06 [allreduce_fusion]: 3.39001e-06 [matmul_add_comm_reduction]: 6.83e-06 [allreduce_slice_to_reducescatter]: 3.39991e-07 [virtual_shard_identity]: 5.72999e-06 [virtual_dataset]: 5.14998e-06 [get_grad_eliminate_]: 5.04e-06 [virtual_output]: 5.02999e-06 [merge_forward]: 2.56e-06 [cell_reuse_recompute_pass]: 1.62001e-06 [offload_activation]: 5.89e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.281e-05 [merge_recompute_call_nodes]: 7.89994e-07 [before_grad]: 8.25e-06 [set_forward_comm_id_for_comm_node_pass]: 3.40998e-06 [meta_fg_expand]: 2.11998e-06 [flash_sp_send_recv_attached]: 9.20001e-07 [receive_attached]: 1.51998e-06 [after_resolve]: 8.43999e-06 [a_after_grad]: 7.63001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.09998e-06 [auto_monad_grad]: 1.34e-06 [auto_monad_eliminator]: 6.58998e-06 [cse]: 1.39e-05 [a_3]: 3.177e-05 [py_interpret_to_execute_after_opt_a]: 5.09e-06 [slice_cell_reuse_recomputed_activation]: 2.20002e-06 [rewriter_after_opt_a]: 1.775e-05 [convert_after_rewriter]: 1.24e-06 [order_py_execute_after_rewriter]: 1.49e-06 [mutable_eliminate]: 0.0005659 [opt_b]: 0.0001836, [1] [Cycle 1]: 0.00017705, [7] [b_1]: 0.00010704 [b_2]: 6.80002e-06 [updatestate_depend_eliminate]: 5.19e-06 [updatestate_assign_eliminate]: 2.34999e-06 [updatestate_loads_eliminate]: 2.18002e-06 [renormalize]: 3.60014e-07 [cse]: 1.929e-05 [optimize_parallel_all_gather_comm]: 1.638e-05 [overlap_param_gather]: 1.87999e-06 [cconv]: 2.196e-05 [loop_unroll]: 0.00047692 [opt_after_cconv]: 0.00010143, [1] [Cycle 1]: 9.539e-05, [7] [c_1]: 3e-05 [parameter_eliminate]: 2.34999e-06 [updatestate_depend_eliminate]: 4.93001e-06 [updatestate_assign_eliminate]: 2.31e-06 [updatestate_loads_eliminate]: 2.69001e-06 [cse]: 1.954e-05 [renormalize]: 4.39992e-07 [remove_dup_value]: 1.517e-05 [tuple_transform]: 7.097e-05, [1] [Cycle 1]: 6.606e-05, [4] [d_1]: 3.992e-05 [none_parameter_eliminate]: 1.76e-06 [renormalize]: 1.70025e-07 [switch_simplify]: 6.03998e-06 [partial_unused_args_eliminate]: 1.69e-06 [add_recomputation]: 4.383e-05 [cse_after_recomputation]: 2.272e-05, [1] [Cycle 1]: 1.829e-05, [1] [cse]: 1.268e-05 [environ_conv]: 5.39e-06 [swap_dp_allreduce_reducescatter]: 5.24e-06 [bias_add_comm_swap]: 2.83e-06 [label_micro_interleaved_index]: 4.85001e-06 [label_fine_grained_interleaved_index]: 2.73e-06 [merge_cast_opt]: 1.24998e-06 [slice_recompute_activation]: 1.91e-06 [micro_interleaved_order_control]: 2.35002e-06 [assign_add_opt]: 1.42e-06 [ForceFp32Comm]: 8.59989e-07 [remove_cast_before_assign_add]: 1.07e-06 [full_micro_interleaved_order_control]: 2.22001e-06 [reorder_send_recv_between_fp_bp]: 2.58998e-06 [comm_op_add_attrs]: 1.29998e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.19e-06 [interleave_parallel_branches]: 1.12999e-06 [overlap_opt_shard_in_pipeline]: 1.17999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.17999e-06 [control_data_broadcast_order]: 1.149e-05 [grouped_pairwise_exchange_alltoall]: 1.44e-06 [offloading_packed_experts]: 3.98999e-06 [overlap_recompute_and_grad_model_parallel]: 4.59998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.09e-06 [overlap_recompute_allgather_and_fa_grad]: 1.30999e-06 [overlap_recompute_comm]: 2.52001e-06 [overlap_grad_ring_attention]: 3.67002e-06 [overlap_grad_flash_sp]: 1.885e-05 [begin_end_overlap_inline]: 4.90021e-07 [split_matmul_comm_elemetwise]: 2.39001e-06 [split_layernorm_comm]: 1.77999e-06 [handle_group_info]: 9.60019e-07 [symbol_engine_optimizer]: 7.286e-05, [1] [Cycle 1]: 6.868e-05, [6] [build]: 2.88e-06 [elim_shapecalc]: 9.74999e-06 [elim_not_effective]: 1.257e-05 [opt_reshape]: 6.70002e-06 [fold_const_symbol]: 9.43002e-06 [renormalize]: 1.80007e-07 [detach_backward]: 1.92999e-06 [pipeline_parallel_scheduler]: 1.55001e-06 [auto_monad_reorder]: 1.732e-05 [get_jit_bprop_graph]: 1.87999e-06 [rewriter_after_jit_bprop_graph]: 3.91999e-06 [opt_after_jit_grad]: 0.00049226 [validate]: 3.778e-05 Sums bootstrap : 0.000447s : 0.60% type_inference : 0.065603s : 88.24% event_method : 0.000078s : 0.11% auto_monad : 0.000179s : 0.24% graph_reusing : 0.000013s : 0.02% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000053s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000015s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000074s : 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 : 0.000005s : 0.01% optimize.rewriter_before_opt_a : 0.000383s : 0.51% optimize.opt_a.expand_dump_flag : 0.000007s : 0.01% optimize.opt_a.switch_simplify : 0.000217s : 0.29% optimize.opt_a.loop_unroll : 0.000100s : 0.13% optimize.opt_a.a_1 : 0.002017s : 2.71% optimize.opt_a.with_stream_mark : 0.000025s : 0.03% optimize.opt_a.recompute_prepare : 0.000015s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000005s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000005s : 0.01% optimize.opt_a.parameter_eliminate : 0.000002s : 0.00% optimize.opt_a.a_2 : 0.000133s : 0.18% optimize.opt_a.accelerated_algorithm : 0.000012s : 0.02% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.01% optimize.opt_a.shard_inline : 0.000011s : 0.02% optimize.opt_a.merge_send_recv : 0.000012s : 0.02% optimize.opt_a.auto_parallel : 0.000011s : 0.01% optimize.opt_a.parallel : 0.000024s : 0.03% optimize.opt_a.flash_sp : 0.000010s : 0.01% optimize.opt_a.merge_comm : 0.000006s : 0.01% optimize.opt_a.allreduce_fusion : 0.000007s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000016s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000013s : 0.02% optimize.opt_a.virtual_dataset : 0.000011s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.01% optimize.opt_a.virtual_output : 0.000011s : 0.01% optimize.opt_a.merge_forward : 0.000006s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000015s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000025s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000017s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.01% optimize.opt_a.meta_fg_expand : 0.000006s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000003s : 0.00% optimize.opt_a.receive_attached : 0.000004s : 0.01% optimize.opt_a.after_resolve : 0.000017s : 0.02% optimize.opt_a.a_after_grad : 0.000016s : 0.02% optimize.opt_a.renormalize : 0.002357s : 3.17% optimize.opt_a.add_forward_monad_depend : 0.000006s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000023s : 0.03% optimize.opt_a.cse : 0.000066s : 0.09% optimize.opt_a.a_3 : 0.000078s : 0.10% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000018s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000566s : 0.76% optimize.opt_b.b_1 : 0.000107s : 0.14% optimize.opt_b.b_2 : 0.000007s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000005s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000019s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000016s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000022s : 0.03% optimize.loop_unroll : 0.000477s : 0.64% optimize.opt_after_cconv.c_1 : 0.000030s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000020s : 0.03% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000015s : 0.02% optimize.tuple_transform.d_1 : 0.000040s : 0.05% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000006s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000044s : 0.06% optimize.cse_after_recomputation.cse : 0.000013s : 0.02% optimize.environ_conv : 0.000005s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.01% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000011s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000004s : 0.00% optimize.overlap_grad_flash_sp : 0.000019s : 0.03% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000013s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 0.01% 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 : 0.000017s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.01% opt_after_jit_grad : 0.000492s : 0.66% validate : 0.000038s : 0.05% Time group info: ------[substitution.] 0.000554 52 2.90% : 0.000016s : 2: substitution.cast_eliminate 0.45% : 0.000002s : 2: substitution.elim_not_effective 0.27% : 0.000002s : 2: substitution.fold_const_symbol 1.00% : 0.000006s : 3: substitution.graph_param_transform 89.34% : 0.000495s : 24: substitution.inline 0.46% : 0.000003s : 4: substitution.j_node_and_user_rematch 0.78% : 0.000004s : 4: substitution.remove_not_recompute_node 0.50% : 0.000003s : 2: substitution.replace_old_param 2.95% : 0.000016s : 7: substitution.switch_simplify 1.35% : 0.000007s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.065508 2 93.37% : 0.061165s : 1: type_inference.infer 6.63% : 0.004344s : 1: type_inference.specialize ------[replace.] 0.000245 35 4.06% : 0.000010s : 2: replace.cast_eliminate 60.91% : 0.000149s : 24: replace.inline 29.17% : 0.000072s : 7: replace.switch_simplify 5.85% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000517 35 2.83% : 0.000015s : 2: match.cast_eliminate 93.50% : 0.000484s : 24: match.inline 2.48% : 0.000013s : 7: match.switch_simplify 1.18% : 0.000006s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000374 2336 1.18% : 0.000004s : 32: predicate.accumulaten_eliminater 0.46% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.22% : 0.000001s : 6: predicate.addn_check_dump 1.27% : 0.000005s : 32: predicate.addn_zero_filter 1.12% : 0.000004s : 32: predicate.adjust_all_reduce_mul_add 2.06% : 0.000008s : 38: predicate.arithmetic_simplify 1.44% : 0.000005s : 34: predicate.cast_eliminate 0.25% : 0.000001s : 6: predicate.check_bprop_eliminate 0.22% : 0.000001s : 6: predicate.compare_switch_simplify 0.07% : 0.000000s : 3: predicate.const_output_eliminate 0.29% : 0.000001s : 6: predicate.depend_value_elim 1.36% : 0.000005s : 34: predicate.dict_get_item_const_eliminator 1.45% : 0.000005s : 34: predicate.dict_get_item_eliminator 1.26% : 0.000005s : 34: predicate.dict_set_item_eliminator 0.38% : 0.000001s : 6: predicate.dumpgradient_eliminate 0.09% : 0.000000s : 3: predicate.elim_not_effective 0.18% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.39% : 0.000005s : 37: predicate.environ_add_const_eliminate 1.42% : 0.000005s : 37: predicate.environ_get_add_eliminate 1.33% : 0.000005s : 37: predicate.environ_get_depend_swap 1.77% : 0.000007s : 43: predicate.environ_get_eliminate 1.45% : 0.000005s : 37: predicate.environ_get_set_eliminate 2.31% : 0.000009s : 60: predicate.exchange_switch_depend_value 3.19% : 0.000012s : 60: predicate.float_depend_g_call 0.22% : 0.000001s : 6: predicate.float_environ_get_switch 0.32% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.09% : 0.000000s : 3: predicate.fold_const_symbol 0.27% : 0.000001s : 6: predicate.get_grad_eliminate 0.10% : 0.000000s : 3: predicate.graph_param_transform 0.23% : 0.000001s : 6: predicate.incorporate_call 0.21% : 0.000001s : 6: predicate.incorporate_call_switch 6.21% : 0.000023s : 110: predicate.inline 0.29% : 0.000001s : 6: predicate.inline_without_move 0.13% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.44% : 0.000002s : 6: predicate.less_batch_normalization 1.77% : 0.000007s : 42: predicate.list_to_tuple_eliminator_ 2.77% : 0.000010s : 74: predicate.load_eliminater 0.52% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.87% : 0.000014s : 99: predicate.loop_unroll_before_grad 1.67% : 0.000006s : 40: predicate.make_slice_get_slice_eliminator 0.28% : 0.000001s : 6: predicate.merge_addn 0.23% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.24% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.22% : 0.000005s : 32: predicate.minmaximum_grad 0.52% : 0.000002s : 3: predicate.mutable_eliminate 0.15% : 0.000001s : 3: predicate.opt_reshape 0.17% : 0.000001s : 3: predicate.parallel_virtual_node 3.02% : 0.000011s : 60: predicate.partial_defer_inline 1.64% : 0.000006s : 39: predicate.partial_eliminate 1.45% : 0.000005s : 32: predicate.print_const_string_wrapper 0.25% : 0.000001s : 6: predicate.reduce_all_const_elim 1.64% : 0.000006s : 32: predicate.reduce_eliminate 2.75% : 0.000010s : 74: predicate.redundant_stop_gradient_eliminater 0.19% : 0.000001s : 6: predicate.remove_not_recompute_node 1.14% : 0.000004s : 42: predicate.replace_applicator 0.18% : 0.000001s : 6: predicate.replace_old_param 0.12% : 0.000000s : 3: predicate.reset_defer_inline 1.32% : 0.000005s : 32: predicate.reshape_eliminate 0.26% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.16% : 0.000001s : 3: predicate.row_tensor_eliminate 0.35% : 0.000001s : 6: predicate.same_eliminate 0.17% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.33% : 0.000001s : 6: predicate.shard_identity_eliminate 0.36% : 0.000001s : 6: predicate.special_op_eliminate 0.30% : 0.000001s : 6: predicate.specialize_transform 0.37% : 0.000001s : 6: predicate.split_environ_get_set_with_tuple_value 0.32% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.14% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.54% : 0.000009s : 60: predicate.switch_defer_inline 2.74% : 0.000010s : 66: predicate.switch_layer_defer_inline 7.63% : 0.000029s : 182: predicate.switch_simplify 1.31% : 0.000005s : 32: predicate.tile_eliminate 1.36% : 0.000005s : 32: predicate.transpose_eliminate 1.71% : 0.000006s : 40: predicate.tuple_list_convert_item_index_to_positive 1.76% : 0.000007s : 40: predicate.tuple_list_get_item_const_eliminator 1.64% : 0.000006s : 40: predicate.tuple_list_get_item_depend_reorder 2.48% : 0.000009s : 48: predicate.tuple_list_get_item_eliminator 1.55% : 0.000006s : 40: predicate.tuple_list_get_set_item_eliminator 2.09% : 0.000008s : 46: predicate.tuple_list_set_item_eliminator 1.62% : 0.000006s : 42: predicate.tuple_to_list_eliminator_ 2.72% : 0.000010s : 74: predicate.updatestate_pure_node_eliminater 3.22% : 0.000012s : 80: predicate.updatestate_useless_node_eliminater 0.17% : 0.000001s : 3: predicate.value_based_eliminate 0.28% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.30% : 0.000001s : 6: predicate.virtual_output_eliminate 0.11% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.22% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002710 42 51.21% : 0.001388s : 16: func_graph_cloner_run.FuncGraphClonerGraph 2.22% : 0.000060s : 1: func_graph_cloner_run.FuncGraphClonerNode 46.56% : 0.001262s : 25: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.095721 192 0.00% : 0.000003s : 1: ForceFp32Comm 3.77% : 0.003611s : 1: add_attr 3.76% : 0.003601s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.05% : 0.000048s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.20% : 0.000187s : 1: auto_monad 0.02% : 0.000021s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.01% : 0.000005s : 1: bias_add_comm_swap 0.50% : 0.000476s : 1: bootstrap 0.03% : 0.000026s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000015s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000026s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000005s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.09% : 0.000087s : 1: event_method 0.01% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000017s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000005s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000008s : 1: label_micro_interleaved_index 0.51% : 0.000487s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.60% : 0.000576s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.02% : 0.000015s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000014s : 1: opt.transform.mutable_eliminate 2.74% : 0.002620s : 78: opt.transform.opt_a 0.03% : 0.000029s : 1: opt.transform.opt_after_cconv 0.02% : 0.000023s : 1: opt.transform.opt_after_jit_grad 0.09% : 0.000087s : 28: opt.transform.opt_b 0.05% : 0.000043s : 2: opt.transform.opt_trans_graph 0.04% : 0.000035s : 4: opt.transform.symbol_engine_opt 6.07% : 0.005807s : 1: opt_a 0.11% : 0.000106s : 1: opt_after_cconv 0.52% : 0.000502s : 1: opt_after_jit_grad 0.20% : 0.000187s : 1: opt_b 8.52% : 0.008151s : 1: optimize 0.02% : 0.000021s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.02% : 0.000022s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000006s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.01% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.01% : 0.000005s : 1: partial_unused_args_eliminate 0.01% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.08% : 0.000080s : 1: pre_auto_parallel 0.01% : 0.000009s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000019s : 1: remove_dup_value 1.30% : 0.001244s : 1: renormalize.infer 1.15% : 0.001103s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000021s : 1: rewriter_after_opt_a 0.41% : 0.000390s : 1: rewriter_before_opt_a 0.01% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000004s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.01% : 0.000006s : 1: split_matmul_comm_elemetwise 0.01% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.08% : 0.000076s : 1: symbol_engine_optimizer 0.08% : 0.000074s : 1: tuple_transform 68.56% : 0.065627s : 1: type_inference TotalTime = 0.104553, [21] [bootstrap]: 0.00040812 [type_inference]: 0.0856286 [event_method]: 7.379e-05 [auto_monad]: 0.00024971 [graph_reusing]: 2.057e-05 [inline]: 3.14999e-06 [add_attr]: 0.00371237, [1] [add_attr_with_inline]: 0.0037018, [1] [Cycle 1]: 0.00013124, [2] [tag_attr]: 7.418e-05 [meta_addattr_fg_expand]: 2.177e-05 [parallel-infer-symbol]: 3.26999e-06 [pre_auto_parallel]: 0.00010147 [insert-virtual-dataset]: 2.66e-06 [parallel-infer-symbol-second]: 7.29982e-07 [dataset_repeat_opt]: 1.90001e-06 [pipeline_split]: 1.84e-06 [optimize]: 0.0136326, [53] [py_interpret_to_execute]: 6.19001e-06 [rewriter_before_opt_a]: 0.00059962 [opt_a]: 0.0112396, [2] [Cycle 1]: 0.0102156, [45] [expand_dump_flag]: 9.40001e-06 [switch_simplify]: 0.00034346 [loop_unroll]: 0.00012433 [a_1]: 0.00274839 [with_stream_mark]: 2.589e-05 [recompute_prepare]: 2.01e-05 [updatestate_depend_eliminate]: 8.1e-06 [updatestate_assign_eliminate]: 6.66e-06 [updatestate_loads_eliminate]: 6.38e-06 [parameter_eliminate]: 3.14001e-06 [a_2]: 0.00020174 [accelerated_algorithm]: 1.511e-05 [shard]: 1.67001e-06 [meta_shard_fg_expand]: 7.26999e-06 [shard_inline]: 3.609e-05 [merge_send_recv]: 1.599e-05 [auto_parallel]: 1.079e-05 [parallel]: 1.799e-05 [flash_sp]: 9.93002e-06 [merge_comm]: 8.47998e-06 [allreduce_fusion]: 7.53e-06 [matmul_add_comm_reduction]: 2.982e-05 [allreduce_slice_to_reducescatter]: 6.39993e-07 [virtual_shard_identity]: 1.522e-05 [virtual_dataset]: 1.285e-05 [get_grad_eliminate_]: 2.826e-05 [virtual_output]: 1.349e-05 [merge_forward]: 8.77e-06 [cell_reuse_recompute_pass]: 1.29e-06 [offload_activation]: 1.705e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.533e-05 [merge_recompute_call_nodes]: 2.12999e-06 [before_grad]: 2.38e-05 [set_forward_comm_id_for_comm_node_pass]: 8.48999e-06 [meta_fg_expand]: 0.00169392 [flash_sp_send_recv_attached]: 4.13999e-06 [receive_attached]: 2.40002e-06 [after_resolve]: 5.904e-05 [a_after_grad]: 7.931e-05 [renormalize]: 0.0041792 [add_forward_monad_depend]: 6.00002e-06 [auto_monad_grad]: 2.64999e-06 [auto_monad_eliminator]: 1.296e-05 [cse]: 2.556e-05 [a_3]: 3.56e-05 [Cycle 2]: 0.00101011, [45] [expand_dump_flag]: 1.40999e-06 [switch_simplify]: 5.27001e-06 [loop_unroll]: 4.43999e-06 [a_1]: 7.134e-05 [with_stream_mark]: 1.138e-05 [recompute_prepare]: 4.35e-06 [updatestate_depend_eliminate]: 2.37999e-06 [updatestate_assign_eliminate]: 1.92999e-06 [updatestate_loads_eliminate]: 2.05002e-06 [parameter_eliminate]: 1.37999e-06 [a_2]: 3.891e-05 [accelerated_algorithm]: 3.89002e-06 [shard]: 1.49998e-06 [meta_shard_fg_expand]: 1.47001e-06 [shard_inline]: 3.32002e-06 [merge_send_recv]: 5.56e-06 [auto_parallel]: 5.41998e-06 [parallel]: 6.15002e-06 [flash_sp]: 3.88001e-06 [merge_comm]: 2.24001e-06 [allreduce_fusion]: 2.33002e-06 [matmul_add_comm_reduction]: 5.84e-06 [allreduce_slice_to_reducescatter]: 4.69998e-07 [virtual_shard_identity]: 4.42e-06 [virtual_dataset]: 3.77002e-06 [get_grad_eliminate_]: 3.45e-06 [virtual_output]: 3.60998e-06 [merge_forward]: 2.56e-06 [cell_reuse_recompute_pass]: 1.75001e-06 [offload_activation]: 5.22999e-06 [cell_reuse_handle_not_recompute_node_pass]: 9.43002e-06 [merge_recompute_call_nodes]: 1.14003e-06 [before_grad]: 6.49001e-06 [set_forward_comm_id_for_comm_node_pass]: 2.51998e-06 [meta_fg_expand]: 0.00049962 [flash_sp_send_recv_attached]: 1.65001e-06 [receive_attached]: 2.11998e-06 [after_resolve]: 1.17e-05 [a_after_grad]: 6.01e-06 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 2.59001e-06 [auto_monad_grad]: 1.40001e-06 [auto_monad_eliminator]: 7.48e-06 [cse]: 1.445e-05 [a_3]: 2.308e-05 [py_interpret_to_execute_after_opt_a]: 4.53999e-06 [slice_cell_reuse_recomputed_activation]: 2.51998e-06 [rewriter_after_opt_a]: 1.706e-05 [convert_after_rewriter]: 1.26002e-06 [order_py_execute_after_rewriter]: 1.19e-06 [mutable_eliminate]: 0.00056094 [opt_b]: 0.00013906, [1] [Cycle 1]: 0.00013207, [7] [b_1]: 7.041e-05 [b_2]: 6.00002e-06 [updatestate_depend_eliminate]: 4.42e-06 [updatestate_assign_eliminate]: 1.84998e-06 [updatestate_loads_eliminate]: 1.59e-06 [renormalize]: 4.69998e-07 [cse]: 1.467e-05 [optimize_parallel_all_gather_comm]: 1.384e-05 [overlap_param_gather]: 1.79e-06 [cconv]: 2.295e-05 [loop_unroll]: 0.00044491 [opt_after_cconv]: 7.681e-05, [1] [Cycle 1]: 7.115e-05, [7] [c_1]: 1.475e-05 [parameter_eliminate]: 2.82002e-06 [updatestate_depend_eliminate]: 4.2e-06 [updatestate_assign_eliminate]: 1.63002e-06 [updatestate_loads_eliminate]: 1.54e-06 [cse]: 1.424e-05 [renormalize]: 2.69996e-07 [remove_dup_value]: 1.289e-05 [tuple_transform]: 5.005e-05, [1] [Cycle 1]: 4.511e-05, [4] [d_1]: 2.231e-05 [none_parameter_eliminate]: 1.52999e-06 [renormalize]: 1.69995e-07 [switch_simplify]: 3.78999e-06 [partial_unused_args_eliminate]: 1.60999e-06 [add_recomputation]: 3.072e-05 [cse_after_recomputation]: 1.762e-05, [1] [Cycle 1]: 1.302e-05, [1] [cse]: 7.87e-06 [environ_conv]: 4.48001e-06 [swap_dp_allreduce_reducescatter]: 3.97e-06 [bias_add_comm_swap]: 2.41e-06 [label_micro_interleaved_index]: 4.54002e-06 [label_fine_grained_interleaved_index]: 2.96001e-06 [merge_cast_opt]: 1.31002e-06 [slice_recompute_activation]: 2.04999e-06 [micro_interleaved_order_control]: 2.12001e-06 [assign_add_opt]: 1.23002e-06 [ForceFp32Comm]: 1.15001e-06 [remove_cast_before_assign_add]: 1.39e-06 [full_micro_interleaved_order_control]: 2.04e-06 [reorder_send_recv_between_fp_bp]: 2.67001e-06 [comm_op_add_attrs]: 1.27e-06 [add_comm_op_reuse_tag]: 9.30013e-07 [interleave_split_concat_branches]: 1.07e-06 [interleave_parallel_branches]: 1.35999e-06 [overlap_opt_shard_in_pipeline]: 1.27e-06 [overlap_opt_shard_grad_in_pipeline]: 1.56998e-06 [control_data_broadcast_order]: 9.34e-06 [grouped_pairwise_exchange_alltoall]: 1.45999e-06 [offloading_packed_experts]: 3.16001e-06 [overlap_recompute_and_grad_model_parallel]: 3.56001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37999e-06 [overlap_recompute_comm]: 2.22001e-06 [overlap_grad_ring_attention]: 3.04001e-06 [overlap_grad_flash_sp]: 1.507e-05 [begin_end_overlap_inline]: 5.10016e-07 [split_matmul_comm_elemetwise]: 1.99e-06 [split_layernorm_comm]: 1.72001e-06 [handle_group_info]: 1.03001e-06 [symbol_engine_optimizer]: 5.911e-05, [1] [Cycle 1]: 5.498e-05, [6] [build]: 2.69999e-06 [elim_shapecalc]: 6.76999e-06 [elim_not_effective]: 8.80001e-06 [opt_reshape]: 4.1e-06 [fold_const_symbol]: 6.43e-06 [renormalize]: 1.40019e-07 [detach_backward]: 1.92001e-06 [pipeline_parallel_scheduler]: 1.64e-06 [auto_monad_reorder]: 1.409e-05 [get_jit_bprop_graph]: 1.30999e-06 [rewriter_after_jit_bprop_graph]: 3.7e-06 [opt_after_jit_grad]: 0.00048254 [validate]: 3.132e-05 Sums bootstrap : 0.000408s : 0.41% type_inference : 0.085629s : 85.75% event_method : 0.000074s : 0.07% auto_monad : 0.000250s : 0.25% graph_reusing : 0.000021s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000074s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000022s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000101s : 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 : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000600s : 0.60% optimize.opt_a.expand_dump_flag : 0.000011s : 0.01% optimize.opt_a.switch_simplify : 0.000349s : 0.35% optimize.opt_a.loop_unroll : 0.000129s : 0.13% optimize.opt_a.a_1 : 0.002820s : 2.82% optimize.opt_a.with_stream_mark : 0.000037s : 0.04% optimize.opt_a.recompute_prepare : 0.000024s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000008s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000241s : 0.24% optimize.opt_a.accelerated_algorithm : 0.000019s : 0.02% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000009s : 0.01% optimize.opt_a.shard_inline : 0.000039s : 0.04% optimize.opt_a.merge_send_recv : 0.000022s : 0.02% optimize.opt_a.auto_parallel : 0.000016s : 0.02% optimize.opt_a.parallel : 0.000024s : 0.02% optimize.opt_a.flash_sp : 0.000014s : 0.01% optimize.opt_a.merge_comm : 0.000011s : 0.01% optimize.opt_a.allreduce_fusion : 0.000010s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000036s : 0.04% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000020s : 0.02% optimize.opt_a.virtual_dataset : 0.000017s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000032s : 0.03% optimize.opt_a.virtual_output : 0.000017s : 0.02% optimize.opt_a.merge_forward : 0.000011s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000022s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000035s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000030s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.01% optimize.opt_a.meta_fg_expand : 0.002194s : 2.20% optimize.opt_a.flash_sp_send_recv_attached : 0.000006s : 0.01% optimize.opt_a.receive_attached : 0.000005s : 0.00% optimize.opt_a.after_resolve : 0.000071s : 0.07% optimize.opt_a.a_after_grad : 0.000085s : 0.09% optimize.opt_a.renormalize : 0.004179s : 4.19% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000020s : 0.02% optimize.opt_a.cse : 0.000040s : 0.04% optimize.opt_a.a_3 : 0.000059s : 0.06% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000017s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000561s : 0.56% optimize.opt_b.b_1 : 0.000070s : 0.07% optimize.opt_b.b_2 : 0.000006s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000015s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000014s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000023s : 0.02% optimize.loop_unroll : 0.000445s : 0.45% optimize.opt_after_cconv.c_1 : 0.000015s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000014s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000022s : 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 : 0.000004s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000031s : 0.03% optimize.cse_after_recomputation.cse : 0.000008s : 0.01% optimize.environ_conv : 0.000004s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000002s : 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 : 0.000002s : 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 : 0.000009s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000015s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000007s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000009s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000004s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000006s : 0.01% 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 : 0.000014s : 0.01% get_jit_bprop_graph : 0.000001s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000483s : 0.48% validate : 0.000031s : 0.03% Time group info: ------[substitution.] 0.000858 116 1.67% : 0.000014s : 2: substitution.cast_eliminate 0.23% : 0.000002s : 1: substitution.elim_not_effective 0.76% : 0.000007s : 9: substitution.float_depend_g_call 0.38% : 0.000003s : 2: substitution.float_tuple_getitem_switch 0.13% : 0.000001s : 1: substitution.fold_const_symbol 0.51% : 0.000004s : 1: substitution.graph_param_transform 0.35% : 0.000003s : 2: substitution.incorporate_call 0.27% : 0.000002s : 2: substitution.incorporate_call_switch 81.13% : 0.000696s : 33: substitution.inline 2.08% : 0.000018s : 2: substitution.inline_without_move 0.73% : 0.000006s : 9: substitution.j_node_and_user_rematch 0.48% : 0.000004s : 2: substitution.minmaximum_grad 0.77% : 0.000007s : 9: substitution.partial_eliminate 0.81% : 0.000007s : 9: substitution.remove_not_recompute_node 0.27% : 0.000002s : 1: substitution.replace_applicator 0.53% : 0.000005s : 6: substitution.replace_old_param 0.31% : 0.000003s : 1: substitution.set_cell_output_no_recompute 2.80% : 0.000024s : 10: substitution.switch_simplify 1.15% : 0.000010s : 2: substitution.tuple_list_convert_item_index_to_positive 0.50% : 0.000004s : 2: substitution.tuple_list_get_item_const_eliminator 0.71% : 0.000006s : 2: substitution.tuple_list_get_item_depend_reorder 2.75% : 0.000024s : 6: substitution.tuple_list_get_item_eliminator 0.68% : 0.000006s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.085435 2 92.52% : 0.079044s : 1: type_inference.infer 7.48% : 0.006391s : 1: type_inference.specialize ------[replace.] 0.000348 47 2.93% : 0.000010s : 2: replace.cast_eliminate 60.50% : 0.000211s : 33: replace.inline 32.43% : 0.000113s : 10: replace.switch_simplify 4.14% : 0.000014s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000723 47 1.83% : 0.000013s : 2: match.cast_eliminate 93.99% : 0.000680s : 33: match.inline 2.56% : 0.000018s : 10: match.switch_simplify 1.62% : 0.000012s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000495 3131 1.20% : 0.000006s : 44: predicate.accumulaten_eliminater 0.18% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.33% : 0.000002s : 12: predicate.addn_check_dump 1.28% : 0.000006s : 44: predicate.addn_zero_filter 1.15% : 0.000006s : 44: predicate.adjust_all_reduce_mul_add 2.11% : 0.000010s : 56: predicate.arithmetic_simplify 1.40% : 0.000007s : 46: predicate.cast_eliminate 0.11% : 0.000001s : 2: predicate.check_bprop_eliminate 0.35% : 0.000002s : 12: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.36% : 0.000002s : 12: predicate.depend_value_elim 1.43% : 0.000007s : 46: predicate.dict_get_item_const_eliminator 1.47% : 0.000007s : 46: predicate.dict_get_item_eliminator 1.27% : 0.000006s : 46: predicate.dict_set_item_eliminator 0.18% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.04% : 0.000000s : 1: predicate.elim_not_effective 0.09% : 0.000000s : 1: predicate.elim_shapecalc_of_broadcastargs 1.31% : 0.000007s : 47: predicate.environ_add_const_eliminate 1.31% : 0.000006s : 47: predicate.environ_get_add_eliminate 1.23% : 0.000006s : 47: predicate.environ_get_depend_swap 1.62% : 0.000008s : 59: predicate.environ_get_eliminate 1.29% : 0.000006s : 47: predicate.environ_get_set_eliminate 2.28% : 0.000011s : 81: predicate.exchange_switch_depend_value 3.23% : 0.000016s : 81: predicate.float_depend_g_call 0.35% : 0.000002s : 12: predicate.float_environ_get_switch 0.41% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.42% : 0.000002s : 12: predicate.get_grad_eliminate 0.03% : 0.000000s : 1: predicate.graph_param_transform 0.35% : 0.000002s : 12: predicate.incorporate_call 0.30% : 0.000001s : 12: predicate.incorporate_call_switch 6.15% : 0.000030s : 151: predicate.inline 1.39% : 0.000007s : 34: predicate.inline_without_move 0.18% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.48% : 0.000002s : 12: predicate.less_batch_normalization 1.65% : 0.000008s : 50: predicate.list_to_tuple_eliminator_ 2.49% : 0.000012s : 94: predicate.load_eliminater 0.25% : 0.000001s : 1: predicate.loop_unroll_after_grad 4.56% : 0.000023s : 138: predicate.loop_unroll_before_grad 1.49% : 0.000007s : 48: predicate.make_slice_get_slice_eliminator 0.35% : 0.000002s : 12: predicate.merge_addn 0.09% : 0.000000s : 2: predicate.micro_step_allgather_replace 0.09% : 0.000000s : 2: predicate.mini_step_allgather_replace 1.23% : 0.000006s : 44: predicate.minmaximum_grad 0.22% : 0.000001s : 1: predicate.mutable_eliminate 0.08% : 0.000000s : 1: predicate.opt_reshape 0.09% : 0.000000s : 1: predicate.parallel_virtual_node 3.35% : 0.000017s : 81: predicate.partial_defer_inline 1.53% : 0.000008s : 49: predicate.partial_eliminate 1.23% : 0.000006s : 44: predicate.print_const_string_wrapper 0.36% : 0.000002s : 12: predicate.reduce_all_const_elim 1.63% : 0.000008s : 44: predicate.reduce_eliminate 2.61% : 0.000013s : 94: predicate.redundant_stop_gradient_eliminater 0.22% : 0.000001s : 12: predicate.remove_not_recompute_node 1.08% : 0.000005s : 50: predicate.replace_applicator 0.62% : 0.000003s : 34: predicate.replace_old_param 0.04% : 0.000000s : 1: predicate.reset_defer_inline 1.22% : 0.000006s : 44: predicate.reshape_eliminate 0.17% : 0.000001s : 2: predicate.row_tensor_add_zeros_like 0.13% : 0.000001s : 1: predicate.row_tensor_eliminate 0.17% : 0.000001s : 2: predicate.same_eliminate 0.22% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.43% : 0.000002s : 12: predicate.shard_identity_eliminate 0.20% : 0.000001s : 2: predicate.special_op_eliminate 0.39% : 0.000002s : 12: predicate.specialize_transform 0.22% : 0.000001s : 2: predicate.split_environ_get_set_with_tuple_value 1.18% : 0.000006s : 34: predicate.stack_unstack_eliminate 0.04% : 0.000000s : 1: predicate.switch_call_monad_eliminater 2.68% : 0.000013s : 81: predicate.switch_defer_inline 2.78% : 0.000014s : 83: predicate.switch_layer_defer_inline 8.56% : 0.000042s : 252: predicate.switch_simplify 1.24% : 0.000006s : 44: predicate.tile_eliminate 1.31% : 0.000006s : 44: predicate.transpose_eliminate 1.58% : 0.000008s : 48: predicate.tuple_list_convert_item_index_to_positive 1.59% : 0.000008s : 48: predicate.tuple_list_get_item_const_eliminator 1.44% : 0.000007s : 48: predicate.tuple_list_get_item_depend_reorder 2.49% : 0.000012s : 62: predicate.tuple_list_get_item_eliminator 1.45% : 0.000007s : 48: predicate.tuple_list_get_set_item_eliminator 2.01% : 0.000010s : 60: predicate.tuple_list_set_item_eliminator 1.42% : 0.000007s : 50: predicate.tuple_to_list_eliminator_ 2.47% : 0.000012s : 94: predicate.updatestate_pure_node_eliminater 3.00% : 0.000015s : 106: predicate.updatestate_useless_node_eliminater 0.08% : 0.000000s : 1: predicate.value_based_eliminate 0.39% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.42% : 0.000002s : 12: predicate.virtual_output_eliminate 0.04% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.09% : 0.000000s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003856 61 63.52% : 0.002449s : 25: func_graph_cloner_run.FuncGraphClonerGraph 1.83% : 0.000071s : 1: func_graph_cloner_run.FuncGraphClonerNode 34.65% : 0.001336s : 35: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.129986 192 0.00% : 0.000004s : 1: ForceFp32Comm 2.86% : 0.003718s : 1: add_attr 2.85% : 0.003705s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.03% : 0.000034s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.20% : 0.000259s : 1: auto_monad 0.01% : 0.000018s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.34% : 0.000439s : 1: bootstrap 0.02% : 0.000027s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000012s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.02% : 0.000020s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.01% : 0.000007s : 1: environ_conv 0.06% : 0.000082s : 1: event_method 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000004s : 1: get_jit_bprop_graph 0.02% : 0.000025s : 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% : 0.000006s : 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 0.02% : 0.000027s : 1: label_micro_interleaved_index 0.35% : 0.000454s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.44% : 0.000570s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.01% : 0.000010s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000011s : 1: opt.transform.mutable_eliminate 2.99% : 0.003892s : 78: opt.transform.opt_a 0.01% : 0.000014s : 1: opt.transform.opt_after_cconv 0.01% : 0.000016s : 1: opt.transform.opt_after_jit_grad 0.04% : 0.000048s : 28: opt.transform.opt_b 0.02% : 0.000024s : 2: opt.transform.opt_trans_graph 0.02% : 0.000023s : 4: opt.transform.symbol_engine_opt 8.65% : 0.011243s : 1: opt_a 0.06% : 0.000080s : 1: opt_after_cconv 0.38% : 0.000493s : 1: opt_after_jit_grad 0.11% : 0.000142s : 1: opt_b 10.49% : 0.013638s : 1: optimize 0.01% : 0.000017s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.01% : 0.000018s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000006s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.01% : 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% : 0.000005s : 1: pipeline_split 0.08% : 0.000107s : 1: pre_auto_parallel 0.01% : 0.000009s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000017s : 1: remove_dup_value 2.69% : 0.003492s : 1: renormalize.infer 0.52% : 0.000678s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000020s : 1: rewriter_after_opt_a 0.47% : 0.000608s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.05% : 0.000062s : 1: symbol_engine_optimizer 0.04% : 0.000053s : 1: tuple_transform 65.90% : 0.085656s : 1: type_inference TotalTime = 0.0754567, [21] [bootstrap]: 0.00042771 [type_inference]: 0.062781 [event_method]: 7.183e-05 [auto_monad]: 0.00016769 [graph_reusing]: 1.153e-05 [inline]: 2.81e-06 [add_attr]: 0.00365132, [1] [add_attr_with_inline]: 0.00364169, [1] [Cycle 1]: 9.754e-05, [2] [tag_attr]: 5.047e-05 [meta_addattr_fg_expand]: 1.386e-05 [parallel-infer-symbol]: 3.06001e-06 [pre_auto_parallel]: 6.634e-05 [insert-virtual-dataset]: 2.66999e-06 [parallel-infer-symbol-second]: 6.80011e-07 [dataset_repeat_opt]: 2.02999e-06 [pipeline_split]: 1.48002e-06 [optimize]: 0.00751165, [53] [py_interpret_to_execute]: 4.58999e-06 [rewriter_before_opt_a]: 0.0003763 [opt_a]: 0.00515822, [2] [Cycle 1]: 0.00456153, [45] [expand_dump_flag]: 5.82999e-06 [switch_simplify]: 0.00020017 [loop_unroll]: 7.111e-05 [a_1]: 0.00153387 [with_stream_mark]: 1.659e-05 [recompute_prepare]: 8.61997e-06 [updatestate_depend_eliminate]: 4.12e-06 [updatestate_assign_eliminate]: 3.28998e-06 [updatestate_loads_eliminate]: 2.86999e-06 [parameter_eliminate]: 2.07999e-06 [a_2]: 7.168e-05 [accelerated_algorithm]: 6.17001e-06 [shard]: 1.75001e-06 [meta_shard_fg_expand]: 3.16001e-06 [shard_inline]: 6.10002e-06 [merge_send_recv]: 8.27e-06 [auto_parallel]: 6.25002e-06 [parallel]: 1.844e-05 [flash_sp]: 8.63001e-06 [merge_comm]: 3.48999e-06 [allreduce_fusion]: 3.10002e-06 [matmul_add_comm_reduction]: 8.94998e-06 [allreduce_slice_to_reducescatter]: 6.09987e-07 [virtual_shard_identity]: 7.12002e-06 [virtual_dataset]: 5.62999e-06 [get_grad_eliminate_]: 5.37999e-06 [virtual_output]: 5.42999e-06 [merge_forward]: 3.64002e-06 [cell_reuse_recompute_pass]: 1.26997e-06 [offload_activation]: 9.11002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.14e-05 [merge_recompute_call_nodes]: 1.37e-06 [before_grad]: 9.33002e-06 [set_forward_comm_id_for_comm_node_pass]: 3.81999e-06 [meta_fg_expand]: 3.9e-06 [flash_sp_send_recv_attached]: 2.33998e-06 [receive_attached]: 2.23998e-06 [after_resolve]: 9.02999e-06 [a_after_grad]: 8.08001e-06 [renormalize]: 0.00212619 [add_forward_monad_depend]: 5.74e-06 [auto_monad_grad]: 2.69999e-06 [auto_monad_eliminator]: 1.759e-05 [cse]: 3.188e-05 [a_3]: 4.45e-05 [Cycle 2]: 0.0005855, [45] [expand_dump_flag]: 1.25001e-06 [switch_simplify]: 6.56e-06 [loop_unroll]: 5.29e-06 [a_1]: 0.0001023 [with_stream_mark]: 1.18e-05 [recompute_prepare]: 5.39e-06 [updatestate_depend_eliminate]: 3.26999e-06 [updatestate_assign_eliminate]: 2.20002e-06 [updatestate_loads_eliminate]: 3.09999e-06 [parameter_eliminate]: 1.37e-06 [a_2]: 6.196e-05 [accelerated_algorithm]: 5.32999e-06 [shard]: 1.13001e-06 [meta_shard_fg_expand]: 1.47999e-06 [shard_inline]: 5.32999e-06 [merge_send_recv]: 5.05001e-06 [auto_parallel]: 5.46e-06 [parallel]: 5.53002e-06 [flash_sp]: 3.03e-06 [merge_comm]: 3.17002e-06 [allreduce_fusion]: 3.04001e-06 [matmul_add_comm_reduction]: 6.63998e-06 [allreduce_slice_to_reducescatter]: 3.50003e-07 [virtual_shard_identity]: 5.74e-06 [virtual_dataset]: 5.19e-06 [get_grad_eliminate_]: 4.82e-06 [virtual_output]: 4.80001e-06 [merge_forward]: 3.28e-06 [cell_reuse_recompute_pass]: 1.64998e-06 [offload_activation]: 7.41001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.287e-05 [merge_recompute_call_nodes]: 1.02e-06 [before_grad]: 8.63001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.81001e-06 [meta_fg_expand]: 2.04e-06 [flash_sp_send_recv_attached]: 1.69e-06 [receive_attached]: 1.14e-06 [after_resolve]: 8.67998e-06 [a_after_grad]: 7.44002e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.07998e-06 [auto_monad_grad]: 9.89996e-07 [auto_monad_eliminator]: 6.63998e-06 [cse]: 1.399e-05 [a_3]: 4.288e-05 [py_interpret_to_execute_after_opt_a]: 4.87e-06 [slice_cell_reuse_recomputed_activation]: 2.09999e-06 [rewriter_after_opt_a]: 1.929e-05 [convert_after_rewriter]: 1.65001e-06 [order_py_execute_after_rewriter]: 1.37999e-06 [mutable_eliminate]: 0.00063018 [opt_b]: 0.00018749, [1] [Cycle 1]: 0.00018087, [7] [b_1]: 0.00010906 [b_2]: 7.50998e-06 [updatestate_depend_eliminate]: 5.79e-06 [updatestate_assign_eliminate]: 2.49999e-06 [updatestate_loads_eliminate]: 2.20002e-06 [renormalize]: 6.80011e-07 [cse]: 1.89e-05 [optimize_parallel_all_gather_comm]: 1.56e-05 [overlap_param_gather]: 1.89e-06 [cconv]: 2.711e-05 [loop_unroll]: 0.00045118 [opt_after_cconv]: 9.232e-05, [1] [Cycle 1]: 8.65e-05, [7] [c_1]: 2.517e-05 [parameter_eliminate]: 2.81e-06 [updatestate_depend_eliminate]: 4.62998e-06 [updatestate_assign_eliminate]: 2.31e-06 [updatestate_loads_eliminate]: 2.58e-06 [cse]: 1.772e-05 [renormalize]: 4.30009e-07 [remove_dup_value]: 1.561e-05 [tuple_transform]: 6.795e-05, [1] [Cycle 1]: 6.301e-05, [4] [d_1]: 3.734e-05 [none_parameter_eliminate]: 1.99999e-06 [renormalize]: 2.00002e-07 [switch_simplify]: 6.10002e-06 [partial_unused_args_eliminate]: 1.85001e-06 [add_recomputation]: 4.456e-05 [cse_after_recomputation]: 2.04e-05, [1] [Cycle 1]: 1.604e-05, [1] [cse]: 1.108e-05 [environ_conv]: 6.51999e-06 [swap_dp_allreduce_reducescatter]: 4.69998e-06 [bias_add_comm_swap]: 2.69001e-06 [label_micro_interleaved_index]: 4.60999e-06 [label_fine_grained_interleaved_index]: 2.72001e-06 [merge_cast_opt]: 1.27e-06 [slice_recompute_activation]: 2.11e-06 [micro_interleaved_order_control]: 2.22999e-06 [assign_add_opt]: 1.32e-06 [ForceFp32Comm]: 7.10017e-07 [remove_cast_before_assign_add]: 9.90025e-07 [full_micro_interleaved_order_control]: 2.10002e-06 [reorder_send_recv_between_fp_bp]: 2.53e-06 [comm_op_add_attrs]: 1.12999e-06 [add_comm_op_reuse_tag]: 9.50007e-07 [interleave_split_concat_branches]: 1.07998e-06 [interleave_parallel_branches]: 1.04e-06 [overlap_opt_shard_in_pipeline]: 1.72999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.66e-06 [control_data_broadcast_order]: 1.225e-05 [grouped_pairwise_exchange_alltoall]: 1.64e-06 [offloading_packed_experts]: 3.73001e-06 [overlap_recompute_and_grad_model_parallel]: 4.89e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.15999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.39e-06 [overlap_recompute_comm]: 2.25002e-06 [overlap_grad_ring_attention]: 4.2e-06 [overlap_grad_flash_sp]: 1.964e-05 [begin_end_overlap_inline]: 4.60015e-07 [split_matmul_comm_elemetwise]: 2.53e-06 [split_layernorm_comm]: 2.16e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 6.874e-05, [1] [Cycle 1]: 6.422e-05, [6] [build]: 2.79999e-06 [elim_shapecalc]: 8.67e-06 [elim_not_effective]: 1.069e-05 [opt_reshape]: 6.41e-06 [fold_const_symbol]: 9.04e-06 [renormalize]: 2.30008e-07 [detach_backward]: 1.96e-06 [pipeline_parallel_scheduler]: 1.49998e-06 [auto_monad_reorder]: 1.663e-05 [get_jit_bprop_graph]: 2.11e-06 [rewriter_after_jit_bprop_graph]: 3.98001e-06 [opt_after_jit_grad]: 0.00048286 [validate]: 7.897e-05 Sums bootstrap : 0.000428s : 0.60% type_inference : 0.062781s : 88.56% event_method : 0.000072s : 0.10% auto_monad : 0.000168s : 0.24% graph_reusing : 0.000012s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000050s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000014s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000066s : 0.09% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000001s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.01% optimize.rewriter_before_opt_a : 0.000376s : 0.53% optimize.opt_a.expand_dump_flag : 0.000007s : 0.01% optimize.opt_a.switch_simplify : 0.000207s : 0.29% optimize.opt_a.loop_unroll : 0.000076s : 0.11% optimize.opt_a.a_1 : 0.001636s : 2.31% optimize.opt_a.with_stream_mark : 0.000028s : 0.04% optimize.opt_a.recompute_prepare : 0.000014s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000007s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000005s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.01% optimize.opt_a.parameter_eliminate : 0.000003s : 0.00% optimize.opt_a.a_2 : 0.000134s : 0.19% optimize.opt_a.accelerated_algorithm : 0.000012s : 0.02% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.01% optimize.opt_a.shard_inline : 0.000011s : 0.02% optimize.opt_a.merge_send_recv : 0.000013s : 0.02% optimize.opt_a.auto_parallel : 0.000012s : 0.02% optimize.opt_a.parallel : 0.000024s : 0.03% optimize.opt_a.flash_sp : 0.000012s : 0.02% optimize.opt_a.merge_comm : 0.000007s : 0.01% optimize.opt_a.allreduce_fusion : 0.000006s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000016s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000013s : 0.02% optimize.opt_a.virtual_dataset : 0.000011s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000010s : 0.01% optimize.opt_a.virtual_output : 0.000010s : 0.01% optimize.opt_a.merge_forward : 0.000007s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000017s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000024s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000018s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.01% optimize.opt_a.meta_fg_expand : 0.000006s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.01% optimize.opt_a.receive_attached : 0.000003s : 0.00% optimize.opt_a.after_resolve : 0.000018s : 0.02% optimize.opt_a.a_after_grad : 0.000016s : 0.02% optimize.opt_a.renormalize : 0.002126s : 3.00% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000024s : 0.03% optimize.opt_a.cse : 0.000046s : 0.06% optimize.opt_a.a_3 : 0.000087s : 0.12% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000019s : 0.03% optimize.convert_after_rewriter : 0.000002s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000630s : 0.89% optimize.opt_b.b_1 : 0.000109s : 0.15% optimize.opt_b.b_2 : 0.000008s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000019s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000016s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000027s : 0.04% optimize.loop_unroll : 0.000451s : 0.64% optimize.opt_after_cconv.c_1 : 0.000025s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000018s : 0.02% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.02% optimize.tuple_transform.d_1 : 0.000037s : 0.05% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000006s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000045s : 0.06% optimize.cse_after_recomputation.cse : 0.000011s : 0.02% optimize.environ_conv : 0.000007s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.01% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000002s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000012s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000004s : 0.01% optimize.overlap_grad_flash_sp : 0.000020s : 0.03% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000009s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000011s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000006s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 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 : 0.000017s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.01% opt_after_jit_grad : 0.000483s : 0.68% validate : 0.000079s : 0.11% Time group info: ------[substitution.] 0.000434 46 2.58% : 0.000011s : 1: substitution.cast_eliminate 0.37% : 0.000002s : 2: substitution.elim_not_effective 0.41% : 0.000002s : 2: substitution.fold_const_symbol 1.22% : 0.000005s : 3: substitution.graph_param_transform 87.04% : 0.000378s : 20: substitution.inline 0.82% : 0.000004s : 4: substitution.j_node_and_user_rematch 1.00% : 0.000004s : 4: substitution.remove_not_recompute_node 0.84% : 0.000004s : 2: substitution.replace_old_param 3.49% : 0.000015s : 6: substitution.switch_simplify 2.23% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.062684 2 93.63% : 0.058691s : 1: type_inference.infer 6.37% : 0.003993s : 1: type_inference.specialize ------[replace.] 0.000204 29 3.18% : 0.000006s : 1: replace.cast_eliminate 57.55% : 0.000117s : 20: replace.inline 32.80% : 0.000067s : 6: replace.switch_simplify 6.47% : 0.000013s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000399 29 2.56% : 0.000010s : 1: match.cast_eliminate 92.34% : 0.000368s : 20: match.inline 3.02% : 0.000012s : 6: match.switch_simplify 2.08% : 0.000008s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000323 2021 1.25% : 0.000004s : 27: predicate.accumulaten_eliminater 0.49% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.28% : 0.000001s : 6: predicate.addn_check_dump 1.23% : 0.000004s : 27: predicate.addn_zero_filter 1.10% : 0.000004s : 27: predicate.adjust_all_reduce_mul_add 2.27% : 0.000007s : 33: predicate.arithmetic_simplify 1.33% : 0.000004s : 28: predicate.cast_eliminate 0.29% : 0.000001s : 6: predicate.check_bprop_eliminate 0.27% : 0.000001s : 6: predicate.compare_switch_simplify 0.09% : 0.000000s : 3: predicate.const_output_eliminate 0.30% : 0.000001s : 6: predicate.depend_value_elim 1.27% : 0.000004s : 28: predicate.dict_get_item_const_eliminator 1.33% : 0.000004s : 28: predicate.dict_get_item_eliminator 1.20% : 0.000004s : 28: predicate.dict_set_item_eliminator 0.46% : 0.000001s : 6: predicate.dumpgradient_eliminate 0.10% : 0.000000s : 3: predicate.elim_not_effective 0.20% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.36% : 0.000004s : 31: predicate.environ_add_const_eliminate 1.28% : 0.000004s : 31: predicate.environ_get_add_eliminate 1.33% : 0.000004s : 31: predicate.environ_get_depend_swap 1.63% : 0.000005s : 37: predicate.environ_get_eliminate 1.28% : 0.000004s : 31: predicate.environ_get_set_eliminate 2.28% : 0.000007s : 50: predicate.exchange_switch_depend_value 3.11% : 0.000010s : 50: predicate.float_depend_g_call 0.24% : 0.000001s : 6: predicate.float_environ_get_switch 0.37% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 3: predicate.fold_const_symbol 0.31% : 0.000001s : 6: predicate.get_grad_eliminate 0.26% : 0.000001s : 3: predicate.graph_param_transform 0.27% : 0.000001s : 6: predicate.incorporate_call 0.24% : 0.000001s : 6: predicate.incorporate_call_switch 5.99% : 0.000019s : 95: predicate.inline 0.31% : 0.000001s : 6: predicate.inline_without_move 0.15% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.48% : 0.000002s : 6: predicate.less_batch_normalization 1.72% : 0.000006s : 36: predicate.list_to_tuple_eliminator_ 2.66% : 0.000009s : 63: predicate.load_eliminater 0.50% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.91% : 0.000013s : 87: predicate.loop_unroll_before_grad 1.51% : 0.000005s : 34: predicate.make_slice_get_slice_eliminator 0.30% : 0.000001s : 6: predicate.merge_addn 0.27% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.29% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.17% : 0.000004s : 27: predicate.minmaximum_grad 0.63% : 0.000002s : 3: predicate.mutable_eliminate 0.20% : 0.000001s : 3: predicate.opt_reshape 0.19% : 0.000001s : 3: predicate.parallel_virtual_node 2.94% : 0.000009s : 50: predicate.partial_defer_inline 1.61% : 0.000005s : 33: predicate.partial_eliminate 1.23% : 0.000004s : 27: predicate.print_const_string_wrapper 0.28% : 0.000001s : 6: predicate.reduce_all_const_elim 1.54% : 0.000005s : 27: predicate.reduce_eliminate 2.76% : 0.000009s : 63: predicate.redundant_stop_gradient_eliminater 0.19% : 0.000001s : 6: predicate.remove_not_recompute_node 1.16% : 0.000004s : 36: predicate.replace_applicator 0.27% : 0.000001s : 6: predicate.replace_old_param 0.13% : 0.000000s : 3: predicate.reset_defer_inline 1.30% : 0.000004s : 27: predicate.reshape_eliminate 0.32% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.23% : 0.000001s : 3: predicate.row_tensor_eliminate 0.39% : 0.000001s : 6: predicate.same_eliminate 0.20% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.46% : 0.000001s : 6: predicate.shard_identity_eliminate 0.39% : 0.000001s : 6: predicate.special_op_eliminate 0.49% : 0.000002s : 6: predicate.specialize_transform 0.40% : 0.000001s : 6: predicate.split_environ_get_set_with_tuple_value 0.40% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.16% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.48% : 0.000008s : 50: predicate.switch_defer_inline 2.70% : 0.000009s : 56: predicate.switch_layer_defer_inline 7.76% : 0.000025s : 158: predicate.switch_simplify 1.22% : 0.000004s : 27: predicate.tile_eliminate 1.28% : 0.000004s : 27: predicate.transpose_eliminate 1.68% : 0.000005s : 34: predicate.tuple_list_convert_item_index_to_positive 1.59% : 0.000005s : 34: predicate.tuple_list_get_item_const_eliminator 1.63% : 0.000005s : 34: predicate.tuple_list_get_item_depend_reorder 2.63% : 0.000008s : 42: predicate.tuple_list_get_item_eliminator 1.62% : 0.000005s : 34: predicate.tuple_list_get_set_item_eliminator 2.12% : 0.000007s : 40: predicate.tuple_list_set_item_eliminator 1.70% : 0.000005s : 36: predicate.tuple_to_list_eliminator_ 2.54% : 0.000008s : 63: predicate.updatestate_pure_node_eliminater 3.13% : 0.000010s : 69: predicate.updatestate_useless_node_eliminater 0.17% : 0.000001s : 3: predicate.value_based_eliminate 0.34% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.35% : 0.000001s : 6: predicate.virtual_output_eliminate 0.13% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.28% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002216 36 50.01% : 0.001108s : 14: func_graph_cloner_run.FuncGraphClonerGraph 49.99% : 0.001108s : 22: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.091009 192 0.00% : 0.000003s : 1: ForceFp32Comm 4.02% : 0.003656s : 1: add_attr 4.01% : 0.003646s : 1: add_attr_with_inline 0.00% : 0.000003s : 1: add_comm_op_reuse_tag 0.05% : 0.000048s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.19% : 0.000176s : 1: auto_monad 0.02% : 0.000020s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000006s : 1: bias_add_comm_swap 0.50% : 0.000458s : 1: bootstrap 0.03% : 0.000031s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000015s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000024s : 1: cse_after_recomputation 0.01% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000005s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.09% : 0.000080s : 1: event_method 0.01% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000015s : 1: graph_reusing 0.01% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000006s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000007s : 1: label_micro_interleaved_index 0.51% : 0.000460s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.70% : 0.000640s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000013s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000015s : 1: opt.transform.mutable_eliminate 2.42% : 0.002203s : 78: opt.transform.opt_a 0.03% : 0.000024s : 1: opt.transform.opt_after_cconv 0.02% : 0.000022s : 1: opt.transform.opt_after_jit_grad 0.10% : 0.000088s : 28: opt.transform.opt_b 0.05% : 0.000041s : 2: opt.transform.opt_trans_graph 0.03% : 0.000032s : 4: opt.transform.symbol_engine_opt 5.67% : 0.005161s : 1: opt_a 0.11% : 0.000096s : 1: opt_after_cconv 0.54% : 0.000493s : 1: opt_after_jit_grad 0.21% : 0.000191s : 1: opt_b 8.26% : 0.007517s : 1: optimize 0.02% : 0.000019s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.02% : 0.000023s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000007s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.01% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000005s : 1: overlap_recompute_comm 0.01% : 0.000006s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.01% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.08% : 0.000071s : 1: pre_auto_parallel 0.01% : 0.000008s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000020s : 1: remove_dup_value 1.40% : 0.001276s : 1: renormalize.infer 0.92% : 0.000840s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000023s : 1: rewriter_after_opt_a 0.42% : 0.000384s : 1: rewriter_before_opt_a 0.01% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.01% : 0.000005s : 1: slice_recompute_activation 0.01% : 0.000005s : 1: split_layernorm_comm 0.01% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.08% : 0.000071s : 1: symbol_engine_optimizer 0.08% : 0.000071s : 1: tuple_transform 69.01% : 0.062804s : 1: type_inference TotalTime = 0.0998226, [21] [bootstrap]: 0.00042232 [type_inference]: 0.0808758 [event_method]: 7.76e-05 [auto_monad]: 0.00023921 [graph_reusing]: 1.847e-05 [inline]: 2.01e-06 [add_attr]: 0.00390681, [1] [add_attr_with_inline]: 0.00389811, [1] [Cycle 1]: 0.00012031, [2] [tag_attr]: 6.916e-05 [meta_addattr_fg_expand]: 1.931e-05 [parallel-infer-symbol]: 4.12e-06 [pre_auto_parallel]: 9.697e-05 [insert-virtual-dataset]: 3.33e-06 [parallel-infer-symbol-second]: 8.39995e-07 [dataset_repeat_opt]: 2.27001e-06 [pipeline_split]: 2.54999e-06 [optimize]: 0.0134481, [53] [py_interpret_to_execute]: 5.72001e-06 [rewriter_before_opt_a]: 0.00062935 [opt_a]: 0.0107991, [2] [Cycle 1]: 0.0100121, [45] [expand_dump_flag]: 8.16002e-06 [switch_simplify]: 0.0003103 [loop_unroll]: 0.00010673 [a_1]: 0.00243043 [with_stream_mark]: 2.357e-05 [recompute_prepare]: 1.933e-05 [updatestate_depend_eliminate]: 8.32e-06 [updatestate_assign_eliminate]: 6.68e-06 [updatestate_loads_eliminate]: 7.25e-06 [parameter_eliminate]: 2.58e-06 [a_2]: 0.0002257 [accelerated_algorithm]: 1.376e-05 [shard]: 2.02001e-06 [meta_shard_fg_expand]: 5.81e-06 [shard_inline]: 1.356e-05 [merge_send_recv]: 1.743e-05 [auto_parallel]: 1.171e-05 [parallel]: 1.849e-05 [flash_sp]: 1.124e-05 [merge_comm]: 8.32e-06 [allreduce_fusion]: 7.62002e-06 [matmul_add_comm_reduction]: 2.828e-05 [allreduce_slice_to_reducescatter]: 7.39994e-07 [virtual_shard_identity]: 1.553e-05 [virtual_dataset]: 1.269e-05 [get_grad_eliminate_]: 1.253e-05 [virtual_output]: 1.266e-05 [merge_forward]: 7.35003e-06 [cell_reuse_recompute_pass]: 1.23002e-06 [offload_activation]: 1.725e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.36e-05 [merge_recompute_call_nodes]: 1.39003e-06 [before_grad]: 2.351e-05 [set_forward_comm_id_for_comm_node_pass]: 8.12e-06 [meta_fg_expand]: 0.00178477 [flash_sp_send_recv_attached]: 4.48999e-06 [receive_attached]: 2.12999e-06 [after_resolve]: 5.798e-05 [a_after_grad]: 7.839e-05 [renormalize]: 0.00424964 [add_forward_monad_depend]: 6.71e-06 [auto_monad_grad]: 2.57001e-06 [auto_monad_eliminator]: 1.471e-05 [cse]: 2.564e-05 [a_3]: 4.079e-05 [Cycle 2]: 0.00077053, [45] [expand_dump_flag]: 2.32999e-06 [switch_simplify]: 6.01998e-06 [loop_unroll]: 4.42998e-06 [a_1]: 5.223e-05 [with_stream_mark]: 1.395e-05 [recompute_prepare]: 4.28001e-06 [updatestate_depend_eliminate]: 3.03e-06 [updatestate_assign_eliminate]: 2.52001e-06 [updatestate_loads_eliminate]: 2.74001e-06 [parameter_eliminate]: 1.45999e-06 [a_2]: 3.881e-05 [accelerated_algorithm]: 4.38001e-06 [shard]: 2.18002e-06 [meta_shard_fg_expand]: 1.68002e-06 [shard_inline]: 3.45e-06 [merge_send_recv]: 7.52002e-06 [auto_parallel]: 7.26001e-06 [parallel]: 7.93999e-06 [flash_sp]: 3.70998e-06 [merge_comm]: 2.22001e-06 [allreduce_fusion]: 2.58e-06 [matmul_add_comm_reduction]: 7.6e-06 [allreduce_slice_to_reducescatter]: 6.60017e-07 [virtual_shard_identity]: 4.42e-06 [virtual_dataset]: 3.81001e-06 [get_grad_eliminate_]: 3.40998e-06 [virtual_output]: 3.85e-06 [merge_forward]: 2.92002e-06 [cell_reuse_recompute_pass]: 2.04999e-06 [offload_activation]: 7.71001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.061e-05 [merge_recompute_call_nodes]: 1.17e-06 [before_grad]: 6.37001e-06 [set_forward_comm_id_for_comm_node_pass]: 2.67001e-06 [meta_fg_expand]: 0.00027488 [flash_sp_send_recv_attached]: 2.29001e-06 [receive_attached]: 3.04001e-06 [after_resolve]: 7.6e-06 [a_after_grad]: 5.52001e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 2.15002e-06 [auto_monad_grad]: 1.19e-06 [auto_monad_eliminator]: 6.32001e-06 [cse]: 1.42e-05 [a_3]: 2.134e-05 [py_interpret_to_execute_after_opt_a]: 5.71e-06 [slice_cell_reuse_recomputed_activation]: 2.27001e-06 [rewriter_after_opt_a]: 1.81e-05 [convert_after_rewriter]: 1.24e-06 [order_py_execute_after_rewriter]: 1.15999e-06 [mutable_eliminate]: 0.00065139 [opt_b]: 0.00013994, [1] [Cycle 1]: 0.00013135, [7] [b_1]: 7.005e-05 [b_2]: 5.34e-06 [updatestate_depend_eliminate]: 4.66997e-06 [updatestate_assign_eliminate]: 1.69e-06 [updatestate_loads_eliminate]: 1.52001e-06 [renormalize]: 8.49977e-07 [cse]: 1.412e-05 [optimize_parallel_all_gather_comm]: 1.765e-05 [overlap_param_gather]: 2.22001e-06 [cconv]: 5.933e-05 [loop_unroll]: 0.0004324 [opt_after_cconv]: 7.747e-05, [1] [Cycle 1]: 7.148e-05, [7] [c_1]: 1.522e-05 [parameter_eliminate]: 2.63e-06 [updatestate_depend_eliminate]: 4.35e-06 [updatestate_assign_eliminate]: 1.71e-06 [updatestate_loads_eliminate]: 1.42e-06 [cse]: 1.419e-05 [renormalize]: 4.19997e-07 [remove_dup_value]: 1.348e-05 [tuple_transform]: 8.925e-05, [1] [Cycle 1]: 8.385e-05, [4] [d_1]: 5.674e-05 [none_parameter_eliminate]: 2.06998e-06 [renormalize]: 1.8999e-07 [switch_simplify]: 4.65001e-06 [partial_unused_args_eliminate]: 1.62999e-06 [add_recomputation]: 3.382e-05 [cse_after_recomputation]: 2.07e-05, [1] [Cycle 1]: 1.579e-05, [1] [cse]: 1.03e-05 [environ_conv]: 5.09e-06 [swap_dp_allreduce_reducescatter]: 4.17e-06 [bias_add_comm_swap]: 2.62001e-06 [label_micro_interleaved_index]: 4.45e-06 [label_fine_grained_interleaved_index]: 2.52001e-06 [merge_cast_opt]: 1.37999e-06 [slice_recompute_activation]: 2.29999e-06 [micro_interleaved_order_control]: 2.36e-06 [assign_add_opt]: 1.68997e-06 [ForceFp32Comm]: 1.07e-06 [remove_cast_before_assign_add]: 1.02e-06 [full_micro_interleaved_order_control]: 2.22999e-06 [reorder_send_recv_between_fp_bp]: 2.63998e-06 [comm_op_add_attrs]: 9.60019e-07 [add_comm_op_reuse_tag]: 1.45999e-06 [interleave_split_concat_branches]: 1.20001e-06 [interleave_parallel_branches]: 1.00999e-06 [overlap_opt_shard_in_pipeline]: 1.25001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.70001e-06 [control_data_broadcast_order]: 9.74999e-06 [grouped_pairwise_exchange_alltoall]: 1.45999e-06 [offloading_packed_experts]: 3.04001e-06 [overlap_recompute_and_grad_model_parallel]: 3.9e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.11002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.33002e-06 [overlap_recompute_comm]: 2.17999e-06 [overlap_grad_ring_attention]: 3.09999e-06 [overlap_grad_flash_sp]: 1.577e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.68e-06 [split_layernorm_comm]: 1.57001e-06 [handle_group_info]: 9.09989e-07 [symbol_engine_optimizer]: 0.00012175, [1] [Cycle 1]: 0.00011716, [6] [build]: 2.43e-06 [elim_shapecalc]: 6.521e-05 [elim_not_effective]: 9.27999e-06 [opt_reshape]: 4.68999e-06 [fold_const_symbol]: 6.69999e-06 [renormalize]: 2.80008e-07 [detach_backward]: 2.20002e-06 [pipeline_parallel_scheduler]: 1.90001e-06 [auto_monad_reorder]: 1.486e-05 [get_jit_bprop_graph]: 1.82001e-06 [rewriter_after_jit_bprop_graph]: 4.17e-06 [opt_after_jit_grad]: 0.00049489 [validate]: 3.453e-05 Sums bootstrap : 0.000422s : 0.44% type_inference : 0.080876s : 85.21% event_method : 0.000078s : 0.08% auto_monad : 0.000239s : 0.25% graph_reusing : 0.000018s : 0.02% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000069s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000019s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000097s : 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 : 0.000003s : 0.00% optimize.py_interpret_to_execute : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000629s : 0.66% optimize.opt_a.expand_dump_flag : 0.000010s : 0.01% optimize.opt_a.switch_simplify : 0.000316s : 0.33% optimize.opt_a.loop_unroll : 0.000111s : 0.12% optimize.opt_a.a_1 : 0.002483s : 2.62% optimize.opt_a.with_stream_mark : 0.000038s : 0.04% optimize.opt_a.recompute_prepare : 0.000024s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000009s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.01% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000265s : 0.28% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.02% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.01% optimize.opt_a.shard_inline : 0.000017s : 0.02% optimize.opt_a.merge_send_recv : 0.000025s : 0.03% optimize.opt_a.auto_parallel : 0.000019s : 0.02% optimize.opt_a.parallel : 0.000026s : 0.03% optimize.opt_a.flash_sp : 0.000015s : 0.02% optimize.opt_a.merge_comm : 0.000011s : 0.01% optimize.opt_a.allreduce_fusion : 0.000010s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000036s : 0.04% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000020s : 0.02% optimize.opt_a.virtual_dataset : 0.000017s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.02% optimize.opt_a.virtual_output : 0.000017s : 0.02% optimize.opt_a.merge_forward : 0.000010s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000025s : 0.03% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000034s : 0.04% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000030s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.01% optimize.opt_a.meta_fg_expand : 0.002060s : 2.17% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.01% optimize.opt_a.receive_attached : 0.000005s : 0.01% optimize.opt_a.after_resolve : 0.000066s : 0.07% optimize.opt_a.a_after_grad : 0.000084s : 0.09% optimize.opt_a.renormalize : 0.004250s : 4.48% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000021s : 0.02% optimize.opt_a.cse : 0.000040s : 0.04% optimize.opt_a.a_3 : 0.000062s : 0.07% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000018s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000651s : 0.69% optimize.opt_b.b_1 : 0.000070s : 0.07% optimize.opt_b.b_2 : 0.000005s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000014s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000018s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000059s : 0.06% optimize.loop_unroll : 0.000432s : 0.46% optimize.opt_after_cconv.c_1 : 0.000015s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000001s : 0.00% optimize.opt_after_cconv.cse : 0.000014s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000013s : 0.01% optimize.tuple_transform.d_1 : 0.000057s : 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 : 0.000005s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000034s : 0.04% optimize.cse_after_recomputation.cse : 0.000010s : 0.01% optimize.environ_conv : 0.000005s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000004s : 0.00% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000010s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000003s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000004s : 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 : 0.000003s : 0.00% optimize.overlap_grad_flash_sp : 0.000016s : 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 : 0.000002s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000065s : 0.07% optimize.symbol_engine_optimizer.elim_not_effective : 0.000009s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000005s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000007s : 0.01% 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 : 0.000015s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000495s : 0.52% validate : 0.000035s : 0.04% Time group info: ------[substitution.] 0.000759 110 1.25% : 0.000009s : 1: substitution.cast_eliminate 0.17% : 0.000001s : 1: substitution.elim_not_effective 0.80% : 0.000006s : 9: substitution.float_depend_g_call 0.48% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.14% : 0.000001s : 1: substitution.fold_const_symbol 4.77% : 0.000036s : 1: substitution.graph_param_transform 0.37% : 0.000003s : 2: substitution.incorporate_call 0.29% : 0.000002s : 2: substitution.incorporate_call_switch 76.14% : 0.000578s : 29: substitution.inline 2.29% : 0.000017s : 2: substitution.inline_without_move 0.82% : 0.000006s : 9: substitution.j_node_and_user_rematch 0.53% : 0.000004s : 2: substitution.minmaximum_grad 0.81% : 0.000006s : 9: substitution.partial_eliminate 0.95% : 0.000007s : 9: substitution.remove_not_recompute_node 0.34% : 0.000003s : 1: substitution.replace_applicator 0.62% : 0.000005s : 6: substitution.replace_old_param 0.31% : 0.000002s : 1: substitution.set_cell_output_no_recompute 3.13% : 0.000024s : 9: substitution.switch_simplify 0.97% : 0.000007s : 2: substitution.tuple_list_convert_item_index_to_positive 0.41% : 0.000003s : 2: substitution.tuple_list_get_item_const_eliminator 0.62% : 0.000005s : 2: substitution.tuple_list_get_item_depend_reorder 3.26% : 0.000025s : 6: substitution.tuple_list_get_item_eliminator 0.55% : 0.000004s : 2: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.080701 2 92.62% : 0.074744s : 1: type_inference.infer 7.38% : 0.005957s : 1: type_inference.specialize ------[replace.] 0.000313 41 1.70% : 0.000005s : 1: replace.cast_eliminate 59.08% : 0.000185s : 29: replace.inline 34.42% : 0.000108s : 9: replace.switch_simplify 4.79% : 0.000015s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000602 41 1.46% : 0.000009s : 1: match.cast_eliminate 93.41% : 0.000563s : 29: match.inline 3.02% : 0.000018s : 9: match.switch_simplify 2.11% : 0.000013s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000450 2862 1.21% : 0.000005s : 40: predicate.accumulaten_eliminater 0.22% : 0.000001s : 1: predicate.ad_related_special_op_eliminate 0.36% : 0.000002s : 12: predicate.addn_check_dump 1.30% : 0.000006s : 40: predicate.addn_zero_filter 1.26% : 0.000006s : 40: predicate.adjust_all_reduce_mul_add 2.20% : 0.000010s : 52: predicate.arithmetic_simplify 1.35% : 0.000006s : 41: predicate.cast_eliminate 0.12% : 0.000001s : 2: predicate.check_bprop_eliminate 0.44% : 0.000002s : 12: predicate.compare_switch_simplify 0.02% : 0.000000s : 1: predicate.const_output_eliminate 0.39% : 0.000002s : 12: predicate.depend_value_elim 1.28% : 0.000006s : 41: predicate.dict_get_item_const_eliminator 1.41% : 0.000006s : 41: predicate.dict_get_item_eliminator 1.26% : 0.000006s : 41: predicate.dict_set_item_eliminator 0.24% : 0.000001s : 2: predicate.dumpgradient_eliminate 0.04% : 0.000000s : 1: predicate.elim_not_effective 0.11% : 0.000001s : 1: predicate.elim_shapecalc_of_broadcastargs 1.30% : 0.000006s : 42: predicate.environ_add_const_eliminate 1.21% : 0.000005s : 42: predicate.environ_get_add_eliminate 1.23% : 0.000006s : 42: predicate.environ_get_depend_swap 1.75% : 0.000008s : 54: predicate.environ_get_eliminate 1.24% : 0.000006s : 42: predicate.environ_get_set_eliminate 2.26% : 0.000010s : 72: predicate.exchange_switch_depend_value 3.20% : 0.000014s : 72: predicate.float_depend_g_call 0.38% : 0.000002s : 12: predicate.float_environ_get_switch 0.46% : 0.000002s : 13: predicate.float_tuple_getitem_switch 0.02% : 0.000000s : 1: predicate.fold_const_symbol 0.43% : 0.000002s : 12: predicate.get_grad_eliminate 0.08% : 0.000000s : 1: predicate.graph_param_transform 0.36% : 0.000002s : 12: predicate.incorporate_call 0.34% : 0.000002s : 12: predicate.incorporate_call_switch 5.93% : 0.000027s : 138: predicate.inline 1.46% : 0.000007s : 34: predicate.inline_without_move 0.20% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.52% : 0.000002s : 12: predicate.less_batch_normalization 1.46% : 0.000007s : 45: predicate.list_to_tuple_eliminator_ 2.52% : 0.000011s : 85: predicate.load_eliminater 0.30% : 0.000001s : 1: predicate.loop_unroll_after_grad 4.29% : 0.000019s : 126: predicate.loop_unroll_before_grad 1.47% : 0.000007s : 43: predicate.make_slice_get_slice_eliminator 0.38% : 0.000002s : 12: predicate.merge_addn 0.11% : 0.000000s : 2: predicate.micro_step_allgather_replace 0.12% : 0.000001s : 2: predicate.mini_step_allgather_replace 1.14% : 0.000005s : 40: predicate.minmaximum_grad 0.31% : 0.000001s : 1: predicate.mutable_eliminate 0.08% : 0.000000s : 1: predicate.opt_reshape 0.08% : 0.000000s : 1: predicate.parallel_virtual_node 3.26% : 0.000015s : 72: predicate.partial_defer_inline 1.49% : 0.000007s : 44: predicate.partial_eliminate 1.24% : 0.000006s : 40: predicate.print_const_string_wrapper 0.38% : 0.000002s : 12: predicate.reduce_all_const_elim 1.63% : 0.000007s : 40: predicate.reduce_eliminate 2.52% : 0.000011s : 85: predicate.redundant_stop_gradient_eliminater 0.26% : 0.000001s : 12: predicate.remove_not_recompute_node 1.16% : 0.000005s : 45: predicate.replace_applicator 0.68% : 0.000003s : 34: predicate.replace_old_param 0.06% : 0.000000s : 1: predicate.reset_defer_inline 1.25% : 0.000006s : 40: predicate.reshape_eliminate 0.10% : 0.000000s : 2: predicate.row_tensor_add_zeros_like 0.12% : 0.000001s : 1: predicate.row_tensor_eliminate 0.22% : 0.000001s : 2: predicate.same_eliminate 0.24% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.49% : 0.000002s : 12: predicate.shard_identity_eliminate 0.14% : 0.000001s : 2: predicate.special_op_eliminate 0.43% : 0.000002s : 12: predicate.specialize_transform 0.27% : 0.000001s : 2: predicate.split_environ_get_set_with_tuple_value 1.33% : 0.000006s : 34: predicate.stack_unstack_eliminate 0.13% : 0.000001s : 1: predicate.switch_call_monad_eliminater 2.63% : 0.000012s : 72: predicate.switch_defer_inline 2.72% : 0.000012s : 74: predicate.switch_layer_defer_inline 8.60% : 0.000039s : 229: predicate.switch_simplify 1.27% : 0.000006s : 40: predicate.tile_eliminate 1.37% : 0.000006s : 40: predicate.transpose_eliminate 1.61% : 0.000007s : 43: predicate.tuple_list_convert_item_index_to_positive 1.46% : 0.000007s : 43: predicate.tuple_list_get_item_const_eliminator 1.36% : 0.000006s : 43: predicate.tuple_list_get_item_depend_reorder 2.49% : 0.000011s : 57: predicate.tuple_list_get_item_eliminator 1.44% : 0.000006s : 43: predicate.tuple_list_get_set_item_eliminator 1.97% : 0.000009s : 55: predicate.tuple_list_set_item_eliminator 1.32% : 0.000006s : 45: predicate.tuple_to_list_eliminator_ 2.42% : 0.000011s : 85: predicate.updatestate_pure_node_eliminater 2.92% : 0.000013s : 97: predicate.updatestate_useless_node_eliminater 0.13% : 0.000001s : 1: predicate.value_based_eliminate 0.43% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.44% : 0.000002s : 12: predicate.virtual_output_eliminate 0.04% : 0.000000s : 1: predicate.virtual_view_grad_eliminate 0.10% : 0.000000s : 1: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003453 55 63.94% : 0.002208s : 23: func_graph_cloner_run.FuncGraphClonerGraph 36.06% : 0.001245s : 32: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.124962 192 0.00% : 0.000004s : 1: ForceFp32Comm 3.13% : 0.003913s : 1: add_attr 3.12% : 0.003902s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.03% : 0.000038s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.20% : 0.000248s : 1: auto_monad 0.01% : 0.000019s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.36% : 0.000449s : 1: bootstrap 0.05% : 0.000064s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000013s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.02% : 0.000024s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.01% : 0.000008s : 1: environ_conv 0.07% : 0.000086s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000023s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000005s : 1: inline 0.01% : 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 0.01% : 0.000007s : 1: label_micro_interleaved_index 0.35% : 0.000442s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.53% : 0.000661s : 1: mutable_eliminate 0.00% : 0.000006s : 1: offloading_packed_experts 0.01% : 0.000010s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000012s : 1: opt.transform.mutable_eliminate 2.79% : 0.003485s : 78: opt.transform.opt_a 0.01% : 0.000014s : 1: opt.transform.opt_after_cconv 0.01% : 0.000016s : 1: opt.transform.opt_after_jit_grad 0.04% : 0.000047s : 28: opt.transform.opt_b 0.05% : 0.000059s : 2: opt.transform.opt_trans_graph 0.02% : 0.000024s : 4: opt.transform.symbol_engine_opt 8.65% : 0.010804s : 1: opt_a 0.06% : 0.000081s : 1: opt_after_cconv 0.40% : 0.000505s : 1: opt_after_jit_grad 0.11% : 0.000143s : 1: opt_b 10.77% : 0.013454s : 1: optimize 0.02% : 0.000021s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.02% : 0.000019s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000006s : 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% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000007s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000005s : 1: overlap_recompute_comm 0.01% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000004s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.08% : 0.000103s : 1: pre_auto_parallel 0.01% : 0.000009s : 1: py_interpret_to_execute 0.01% : 0.000009s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000017s : 1: remove_dup_value 2.83% : 0.003534s : 1: renormalize.infer 0.56% : 0.000705s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000021s : 1: rewriter_after_opt_a 0.52% : 0.000647s : 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% : 0.000007s : 1: swap_dp_allreduce_reducescatter 0.10% : 0.000125s : 1: symbol_engine_optimizer 0.07% : 0.000092s : 1: tuple_transform 64.74% : 0.080898s : 1: type_inference TotalTime = 0.0737487, [21] [bootstrap]: 0.00044348 [type_inference]: 0.060308 [event_method]: 0.00030811 [auto_monad]: 0.00016409 [graph_reusing]: 1.139e-05 [inline]: 2.61e-06 [add_attr]: 0.00386188, [1] [add_attr_with_inline]: 0.00385206, [1] [Cycle 1]: 0.00010172, [2] [tag_attr]: 5.308e-05 [meta_addattr_fg_expand]: 1.253e-05 [parallel-infer-symbol]: 3.04001e-06 [pre_auto_parallel]: 6.847e-05 [insert-virtual-dataset]: 2.36e-06 [parallel-infer-symbol-second]: 1.42e-06 [dataset_repeat_opt]: 2.11e-06 [pipeline_split]: 1.69e-06 [optimize]: 0.0077735, [53] [py_interpret_to_execute]: 6.39999e-06 [rewriter_before_opt_a]: 0.00036786 [opt_a]: 0.00529733, [2] [Cycle 1]: 0.00461566, [45] [expand_dump_flag]: 5.91e-06 [switch_simplify]: 0.00019414 [loop_unroll]: 6.905e-05 [a_1]: 0.00164146 [with_stream_mark]: 2.052e-05 [recompute_prepare]: 1.162e-05 [updatestate_depend_eliminate]: 4.22998e-06 [updatestate_assign_eliminate]: 3.51999e-06 [updatestate_loads_eliminate]: 3.01001e-06 [parameter_eliminate]: 2.46e-06 [a_2]: 0.00011097 [accelerated_algorithm]: 2.141e-05 [shard]: 2.02001e-06 [meta_shard_fg_expand]: 3.26001e-06 [shard_inline]: 6.79999e-06 [merge_send_recv]: 7.68001e-06 [auto_parallel]: 8.38999e-06 [parallel]: 1.82e-05 [flash_sp]: 8.36002e-06 [merge_comm]: 3.71999e-06 [allreduce_fusion]: 3.48e-06 [matmul_add_comm_reduction]: 1.063e-05 [allreduce_slice_to_reducescatter]: 8.70001e-07 [virtual_shard_identity]: 8.37e-06 [virtual_dataset]: 6.64999e-06 [get_grad_eliminate_]: 5.74e-06 [virtual_output]: 6.02999e-06 [merge_forward]: 4.91002e-06 [cell_reuse_recompute_pass]: 1.58002e-06 [offload_activation]: 1.013e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.233e-05 [merge_recompute_call_nodes]: 1.42e-06 [before_grad]: 9.31e-06 [set_forward_comm_id_for_comm_node_pass]: 3.31001e-06 [meta_fg_expand]: 4.07e-06 [flash_sp_send_recv_attached]: 5.19e-06 [receive_attached]: 2.20002e-06 [after_resolve]: 1.083e-05 [a_after_grad]: 9.22999e-06 [renormalize]: 0.00195503 [add_forward_monad_depend]: 7.83001e-06 [auto_monad_grad]: 2.40002e-06 [auto_monad_eliminator]: 1.977e-05 [cse]: 3.432e-05 [a_3]: 4.825e-05 [Cycle 2]: 0.00066686, [45] [expand_dump_flag]: 1.90001e-06 [switch_simplify]: 7.63001e-06 [loop_unroll]: 5.74e-06 [a_1]: 0.00013101 [with_stream_mark]: 1.312e-05 [recompute_prepare]: 6.33e-06 [updatestate_depend_eliminate]: 4.04002e-06 [updatestate_assign_eliminate]: 2.59999e-06 [updatestate_loads_eliminate]: 3.11001e-06 [parameter_eliminate]: 1.49e-06 [a_2]: 7.802e-05 [accelerated_algorithm]: 1.037e-05 [shard]: 1.39e-06 [meta_shard_fg_expand]: 1.69e-06 [shard_inline]: 5.91e-06 [merge_send_recv]: 6.81001e-06 [auto_parallel]: 7.45e-06 [parallel]: 6.65998e-06 [flash_sp]: 3.71999e-06 [merge_comm]: 3.41999e-06 [allreduce_fusion]: 2.93e-06 [matmul_add_comm_reduction]: 8.65999e-06 [allreduce_slice_to_reducescatter]: 7.2e-07 [virtual_shard_identity]: 7.06001e-06 [virtual_dataset]: 5.19e-06 [get_grad_eliminate_]: 5.07e-06 [virtual_output]: 4.74998e-06 [merge_forward]: 4.52998e-06 [cell_reuse_recompute_pass]: 1.93002e-06 [offload_activation]: 8.17998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.125e-05 [merge_recompute_call_nodes]: 1.01002e-06 [before_grad]: 8.67998e-06 [set_forward_comm_id_for_comm_node_pass]: 4.43999e-06 [meta_fg_expand]: 2.11e-06 [flash_sp_send_recv_attached]: 1.57999e-06 [receive_attached]: 2.14e-06 [after_resolve]: 8.84998e-06 [a_after_grad]: 7.85998e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.87001e-06 [auto_monad_grad]: 1.43002e-06 [auto_monad_eliminator]: 9.47999e-06 [cse]: 1.859e-05 [a_3]: 3.292e-05 [py_interpret_to_execute_after_opt_a]: 4.67e-06 [slice_cell_reuse_recomputed_activation]: 2.29001e-06 [rewriter_after_opt_a]: 2.054e-05 [convert_after_rewriter]: 1.21997e-06 [order_py_execute_after_rewriter]: 1.14e-06 [mutable_eliminate]: 0.00064085 [opt_b]: 0.00019189, [1] [Cycle 1]: 0.00018407, [7] [b_1]: 0.00010859 [b_2]: 7.55e-06 [updatestate_depend_eliminate]: 6.71999e-06 [updatestate_assign_eliminate]: 2.58e-06 [updatestate_loads_eliminate]: 2.22999e-06 [renormalize]: 5.69999e-07 [cse]: 2.115e-05 [optimize_parallel_all_gather_comm]: 1.654e-05 [overlap_param_gather]: 1.86998e-06 [cconv]: 2.829e-05 [loop_unroll]: 0.00046914 [opt_after_cconv]: 0.00010246, [1] [Cycle 1]: 9.546e-05, [7] [c_1]: 2.794e-05 [parameter_eliminate]: 3.66999e-06 [updatestate_depend_eliminate]: 6.01e-06 [updatestate_assign_eliminate]: 2.53e-06 [updatestate_loads_eliminate]: 2.29999e-06 [cse]: 2.06e-05 [renormalize]: 4.19997e-07 [remove_dup_value]: 1.683e-05 [tuple_transform]: 0.0001023, [1] [Cycle 1]: 9.722e-05, [4] [d_1]: 6.435e-05 [none_parameter_eliminate]: 2.01998e-06 [renormalize]: 2.80008e-07 [switch_simplify]: 8.18999e-06 [partial_unused_args_eliminate]: 1.77999e-06 [add_recomputation]: 4.832e-05 [cse_after_recomputation]: 2.708e-05, [1] [Cycle 1]: 2.149e-05, [1] [cse]: 1.476e-05 [environ_conv]: 5.15999e-06 [swap_dp_allreduce_reducescatter]: 5.05999e-06 [bias_add_comm_swap]: 2.74001e-06 [label_micro_interleaved_index]: 5.35001e-06 [label_fine_grained_interleaved_index]: 2.51998e-06 [merge_cast_opt]: 1.32e-06 [slice_recompute_activation]: 2.21e-06 [micro_interleaved_order_control]: 2.33002e-06 [assign_add_opt]: 1.38002e-06 [ForceFp32Comm]: 1.30999e-06 [remove_cast_before_assign_add]: 1.03001e-06 [full_micro_interleaved_order_control]: 2.32999e-06 [reorder_send_recv_between_fp_bp]: 2.60002e-06 [comm_op_add_attrs]: 1.00001e-06 [add_comm_op_reuse_tag]: 1.02e-06 [interleave_split_concat_branches]: 1.17e-06 [interleave_parallel_branches]: 9.89996e-07 [overlap_opt_shard_in_pipeline]: 1.24998e-06 [overlap_opt_shard_grad_in_pipeline]: 2.37999e-06 [control_data_broadcast_order]: 1.373e-05 [grouped_pairwise_exchange_alltoall]: 1.54998e-06 [offloading_packed_experts]: 3.76999e-06 [overlap_recompute_and_grad_model_parallel]: 5.01002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.29e-06 [overlap_recompute_allgather_and_fa_grad]: 1.35999e-06 [overlap_recompute_comm]: 2.54999e-06 [overlap_grad_ring_attention]: 4.53999e-06 [overlap_grad_flash_sp]: 1.933e-05 [begin_end_overlap_inline]: 5.50004e-07 [split_matmul_comm_elemetwise]: 2.09999e-06 [split_layernorm_comm]: 1.76e-06 [handle_group_info]: 1.15001e-06 [symbol_engine_optimizer]: 7.721e-05, [1] [Cycle 1]: 7.176e-05, [6] [build]: 2.98998e-06 [elim_shapecalc]: 1.011e-05 [elim_not_effective]: 1.364e-05 [opt_reshape]: 6.94001e-06 [fold_const_symbol]: 8.94e-06 [renormalize]: 2.00002e-07 [detach_backward]: 2.31e-06 [pipeline_parallel_scheduler]: 1.83002e-06 [auto_monad_reorder]: 1.722e-05 [get_jit_bprop_graph]: 1.65001e-06 [rewriter_after_jit_bprop_graph]: 3.63e-06 [opt_after_jit_grad]: 0.00054441 [validate]: 4.184e-05 Sums bootstrap : 0.000443s : 0.64% type_inference : 0.060308s : 87.57% event_method : 0.000308s : 0.45% auto_monad : 0.000164s : 0.24% graph_reusing : 0.000011s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000053s : 0.08% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000013s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000068s : 0.10% 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 : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000368s : 0.53% optimize.opt_a.expand_dump_flag : 0.000008s : 0.01% optimize.opt_a.switch_simplify : 0.000202s : 0.29% optimize.opt_a.loop_unroll : 0.000075s : 0.11% optimize.opt_a.a_1 : 0.001772s : 2.57% optimize.opt_a.with_stream_mark : 0.000034s : 0.05% optimize.opt_a.recompute_prepare : 0.000018s : 0.03% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000006s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000006s : 0.01% optimize.opt_a.parameter_eliminate : 0.000004s : 0.01% optimize.opt_a.a_2 : 0.000189s : 0.27% optimize.opt_a.accelerated_algorithm : 0.000032s : 0.05% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.01% optimize.opt_a.shard_inline : 0.000013s : 0.02% optimize.opt_a.merge_send_recv : 0.000014s : 0.02% optimize.opt_a.auto_parallel : 0.000016s : 0.02% optimize.opt_a.parallel : 0.000025s : 0.04% optimize.opt_a.flash_sp : 0.000012s : 0.02% optimize.opt_a.merge_comm : 0.000007s : 0.01% optimize.opt_a.allreduce_fusion : 0.000006s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000019s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000015s : 0.02% optimize.opt_a.virtual_dataset : 0.000012s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.02% optimize.opt_a.virtual_output : 0.000011s : 0.02% optimize.opt_a.merge_forward : 0.000009s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.01% optimize.opt_a.offload_activation : 0.000018s : 0.03% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000024s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000018s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000008s : 0.01% optimize.opt_a.meta_fg_expand : 0.000006s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000007s : 0.01% optimize.opt_a.receive_attached : 0.000004s : 0.01% optimize.opt_a.after_resolve : 0.000020s : 0.03% optimize.opt_a.a_after_grad : 0.000017s : 0.02% optimize.opt_a.renormalize : 0.001955s : 2.84% optimize.opt_a.add_forward_monad_depend : 0.000010s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000029s : 0.04% optimize.opt_a.cse : 0.000053s : 0.08% optimize.opt_a.a_3 : 0.000081s : 0.12% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000021s : 0.03% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000641s : 0.93% optimize.opt_b.b_1 : 0.000109s : 0.16% optimize.opt_b.b_2 : 0.000008s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000021s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000017s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000028s : 0.04% optimize.loop_unroll : 0.000469s : 0.68% optimize.opt_after_cconv.c_1 : 0.000028s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.01% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000021s : 0.03% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.02% optimize.tuple_transform.d_1 : 0.000064s : 0.09% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000008s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000048s : 0.07% optimize.cse_after_recomputation.cse : 0.000015s : 0.02% optimize.environ_conv : 0.000005s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.01% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000014s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000005s : 0.01% optimize.overlap_grad_flash_sp : 0.000019s : 0.03% 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000010s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000009s : 0.01% 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 : 0.000017s : 0.03% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.01% opt_after_jit_grad : 0.000544s : 0.79% validate : 0.000042s : 0.06% Time group info: ------[substitution.] 0.000581 51 14.08% : 0.000082s : 5: substitution.arithmetic_simplify 2.18% : 0.000013s : 1: substitution.cast_eliminate 0.35% : 0.000002s : 2: substitution.elim_not_effective 0.24% : 0.000001s : 2: substitution.fold_const_symbol 1.03% : 0.000006s : 3: substitution.graph_param_transform 73.57% : 0.000428s : 19: substitution.inline 0.60% : 0.000003s : 4: substitution.j_node_and_user_rematch 2.16% : 0.000013s : 2: substitution.less_batch_normalization 0.86% : 0.000005s : 4: substitution.remove_not_recompute_node 0.60% : 0.000003s : 2: substitution.replace_old_param 2.54% : 0.000015s : 5: substitution.switch_simplify 1.80% : 0.000010s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.060208 2 93.77% : 0.056460s : 1: type_inference.infer 6.23% : 0.003748s : 1: type_inference.specialize ------[replace.] 0.000217 27 3.34% : 0.000007s : 1: replace.cast_eliminate 60.21% : 0.000131s : 19: replace.inline 28.64% : 0.000062s : 5: replace.switch_simplify 7.80% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000448 27 2.57% : 0.000012s : 1: match.cast_eliminate 92.87% : 0.000416s : 19: match.inline 2.54% : 0.000011s : 5: match.switch_simplify 2.02% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000344 1960 1.06% : 0.000004s : 26: predicate.accumulaten_eliminater 0.40% : 0.000001s : 3: predicate.ad_related_special_op_eliminate 0.23% : 0.000001s : 6: predicate.addn_check_dump 1.16% : 0.000004s : 26: predicate.addn_zero_filter 1.00% : 0.000003s : 26: predicate.adjust_all_reduce_mul_add 2.26% : 0.000008s : 32: predicate.arithmetic_simplify 1.23% : 0.000004s : 27: predicate.cast_eliminate 0.30% : 0.000001s : 6: predicate.check_bprop_eliminate 0.24% : 0.000001s : 6: predicate.compare_switch_simplify 0.08% : 0.000000s : 3: predicate.const_output_eliminate 0.27% : 0.000001s : 6: predicate.depend_value_elim 1.17% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 1.24% : 0.000004s : 27: predicate.dict_get_item_eliminator 1.08% : 0.000004s : 27: predicate.dict_set_item_eliminator 0.62% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.14% : 0.000000s : 3: predicate.elim_not_effective 0.25% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.23% : 0.000004s : 30: predicate.environ_add_const_eliminate 1.14% : 0.000004s : 30: predicate.environ_get_add_eliminate 1.27% : 0.000004s : 30: predicate.environ_get_depend_swap 1.55% : 0.000005s : 36: predicate.environ_get_eliminate 1.20% : 0.000004s : 30: predicate.environ_get_set_eliminate 1.98% : 0.000007s : 48: predicate.exchange_switch_depend_value 3.28% : 0.000011s : 48: predicate.float_depend_g_call 0.23% : 0.000001s : 6: predicate.float_environ_get_switch 0.42% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 3: predicate.fold_const_symbol 0.34% : 0.000001s : 6: predicate.get_grad_eliminate 0.11% : 0.000000s : 3: predicate.graph_param_transform 0.29% : 0.000001s : 6: predicate.incorporate_call 0.23% : 0.000001s : 6: predicate.incorporate_call_switch 6.24% : 0.000021s : 92: predicate.inline 0.46% : 0.000002s : 6: predicate.inline_without_move 0.14% : 0.000000s : 6: predicate.j_node_and_user_rematch 0.53% : 0.000002s : 6: predicate.less_batch_normalization 1.59% : 0.000005s : 35: predicate.list_to_tuple_eliminator_ 2.39% : 0.000008s : 61: predicate.load_eliminater 0.79% : 0.000003s : 3: predicate.loop_unroll_after_grad 3.45% : 0.000012s : 84: predicate.loop_unroll_before_grad 1.43% : 0.000005s : 33: predicate.make_slice_get_slice_eliminator 0.24% : 0.000001s : 6: predicate.merge_addn 0.28% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.30% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.10% : 0.000004s : 26: predicate.minmaximum_grad 0.68% : 0.000002s : 3: predicate.mutable_eliminate 0.17% : 0.000001s : 3: predicate.opt_reshape 0.19% : 0.000001s : 3: predicate.parallel_virtual_node 2.93% : 0.000010s : 48: predicate.partial_defer_inline 1.42% : 0.000005s : 32: predicate.partial_eliminate 1.22% : 0.000004s : 26: predicate.print_const_string_wrapper 0.26% : 0.000001s : 6: predicate.reduce_all_const_elim 1.52% : 0.000005s : 26: predicate.reduce_eliminate 2.44% : 0.000008s : 61: predicate.redundant_stop_gradient_eliminater 0.29% : 0.000001s : 6: predicate.remove_not_recompute_node 1.11% : 0.000004s : 35: predicate.replace_applicator 0.25% : 0.000001s : 6: predicate.replace_old_param 0.19% : 0.000001s : 3: predicate.reset_defer_inline 1.08% : 0.000004s : 26: predicate.reshape_eliminate 0.30% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.22% : 0.000001s : 3: predicate.row_tensor_eliminate 0.43% : 0.000001s : 6: predicate.same_eliminate 0.29% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.54% : 0.000002s : 6: predicate.shard_identity_eliminate 0.36% : 0.000001s : 6: predicate.special_op_eliminate 0.39% : 0.000001s : 6: predicate.specialize_transform 0.49% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.53% : 0.000002s : 6: predicate.stack_unstack_eliminate 0.20% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.24% : 0.000008s : 48: predicate.switch_defer_inline 2.40% : 0.000008s : 54: predicate.switch_layer_defer_inline 7.14% : 0.000025s : 151: predicate.switch_simplify 1.05% : 0.000004s : 26: predicate.tile_eliminate 1.19% : 0.000004s : 26: predicate.transpose_eliminate 1.63% : 0.000006s : 33: predicate.tuple_list_convert_item_index_to_positive 1.43% : 0.000005s : 33: predicate.tuple_list_get_item_const_eliminator 1.44% : 0.000005s : 33: predicate.tuple_list_get_item_depend_reorder 2.62% : 0.000009s : 41: predicate.tuple_list_get_item_eliminator 1.42% : 0.000005s : 33: predicate.tuple_list_get_set_item_eliminator 1.95% : 0.000007s : 39: predicate.tuple_list_set_item_eliminator 1.42% : 0.000005s : 35: predicate.tuple_to_list_eliminator_ 2.39% : 0.000008s : 61: predicate.updatestate_pure_node_eliminater 8.10% : 0.000028s : 67: predicate.updatestate_useless_node_eliminater 0.18% : 0.000001s : 3: predicate.value_based_eliminate 0.32% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.27% : 0.000001s : 6: predicate.virtual_output_eliminate 0.12% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.19% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002114 33 48.14% : 0.001018s : 12: func_graph_cloner_run.FuncGraphClonerGraph 51.86% : 0.001096s : 21: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.089860 192 0.00% : 0.000004s : 1: ForceFp32Comm 4.30% : 0.003867s : 1: add_attr 4.29% : 0.003856s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.06% : 0.000053s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.20% : 0.000175s : 1: auto_monad 0.02% : 0.000022s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000006s : 1: bias_add_comm_swap 0.52% : 0.000469s : 1: bootstrap 0.04% : 0.000032s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000017s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000030s : 1: cse_after_recomputation 0.01% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000005s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.36% : 0.000323s : 1: event_method 0.01% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000016s : 1: graph_reusing 0.01% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000006s : 1: inline 0.01% : 0.000005s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000009s : 1: label_micro_interleaved_index 0.53% : 0.000479s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.73% : 0.000652s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.02% : 0.000015s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000016s : 1: opt.transform.mutable_eliminate 2.69% : 0.002415s : 78: opt.transform.opt_a 0.03% : 0.000026s : 1: opt.transform.opt_after_cconv 0.03% : 0.000025s : 1: opt.transform.opt_after_jit_grad 0.10% : 0.000088s : 28: opt.transform.opt_b 0.05% : 0.000044s : 2: opt.transform.opt_trans_graph 0.04% : 0.000036s : 4: opt.transform.symbol_engine_opt 5.90% : 0.005301s : 1: opt_a 0.12% : 0.000106s : 1: opt_after_cconv 0.62% : 0.000556s : 1: opt_after_jit_grad 0.22% : 0.000195s : 1: opt_b 8.66% : 0.007779s : 1: optimize 0.02% : 0.000020s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.03% : 0.000023s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000007s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.01% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000005s : 1: overlap_recompute_comm 0.01% : 0.000006s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.01% : 0.000005s : 1: partial_unused_args_eliminate 0.01% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.08% : 0.000073s : 1: pre_auto_parallel 0.01% : 0.000010s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.04% : 0.000034s : 1: remove_dup_value 1.20% : 0.001081s : 1: renormalize.infer 0.96% : 0.000863s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.03% : 0.000024s : 1: rewriter_after_opt_a 0.42% : 0.000378s : 1: rewriter_before_opt_a 0.01% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.01% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.01% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.09% : 0.000080s : 1: symbol_engine_optimizer 0.12% : 0.000106s : 1: tuple_transform 67.15% : 0.060339s : 1: type_inference TotalTime = 0.103101, [21] [bootstrap]: 0.0003896 [type_inference]: 0.0795038 [event_method]: 7.14e-05 [auto_monad]: 0.00025631 [graph_reusing]: 2.078e-05 [inline]: 2.58003e-06 [add_attr]: 0.00357668, [1] [add_attr_with_inline]: 0.00356721, [1] [Cycle 1]: 0.00011723, [2] [tag_attr]: 6.35e-05 [meta_addattr_fg_expand]: 2.026e-05 [parallel-infer-symbol]: 3.54002e-06 [pre_auto_parallel]: 8.946e-05 [insert-virtual-dataset]: 2.37001e-06 [parallel-infer-symbol-second]: 7.00005e-07 [dataset_repeat_opt]: 1.62001e-06 [pipeline_split]: 1.45999e-06 [optimize]: 0.0184894, [53] [py_interpret_to_execute]: 4.70999e-06 [rewriter_before_opt_a]: 0.00050767 [opt_a]: 0.0159044, [3] [Cycle 1]: 0.0116379, [45] [expand_dump_flag]: 1.118e-05 [switch_simplify]: 0.00028069 [loop_unroll]: 0.0001109 [a_1]: 0.00277015 [with_stream_mark]: 3.101e-05 [recompute_prepare]: 3.016e-05 [updatestate_depend_eliminate]: 9.89001e-06 [updatestate_assign_eliminate]: 8.32e-06 [updatestate_loads_eliminate]: 7.71999e-06 [parameter_eliminate]: 3.38e-06 [a_2]: 0.00029962 [accelerated_algorithm]: 3.488e-05 [shard]: 2.78e-06 [meta_shard_fg_expand]: 7.55e-06 [shard_inline]: 1.744e-05 [merge_send_recv]: 1.975e-05 [auto_parallel]: 1.58e-05 [parallel]: 4.243e-05 [flash_sp]: 1.224e-05 [merge_comm]: 1.185e-05 [allreduce_fusion]: 8.65999e-06 [matmul_add_comm_reduction]: 3.316e-05 [allreduce_slice_to_reducescatter]: 7.89994e-07 [virtual_shard_identity]: 2.134e-05 [virtual_dataset]: 1.7e-05 [get_grad_eliminate_]: 1.697e-05 [virtual_output]: 1.66e-05 [merge_forward]: 1.049e-05 [cell_reuse_recompute_pass]: 1.82999e-06 [offload_activation]: 1.774e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.431e-05 [merge_recompute_call_nodes]: 1.39998e-06 [before_grad]: 3.159e-05 [set_forward_comm_id_for_comm_node_pass]: 1.076e-05 [meta_fg_expand]: 0.00174836 [flash_sp_send_recv_attached]: 4.83001e-06 [receive_attached]: 2.14999e-06 [after_resolve]: 6.112e-05 [a_after_grad]: 8.938e-05 [renormalize]: 0.0046485 [add_forward_monad_depend]: 1.418e-05 [auto_monad_grad]: 6.59001e-06 [auto_monad_eliminator]: 6.298e-05 [cse]: 0.00033712 [a_3]: 0.00034318 [Cycle 2]: 0.00353287, [45] [expand_dump_flag]: 3.46001e-06 [switch_simplify]: 4.679e-05 [loop_unroll]: 4.194e-05 [a_1]: 0.00144021 [with_stream_mark]: 2.131e-05 [recompute_prepare]: 1.111e-05 [updatestate_depend_eliminate]: 4.82998e-06 [updatestate_assign_eliminate]: 4e-06 [updatestate_loads_eliminate]: 3.66999e-06 [parameter_eliminate]: 2.73998e-06 [a_2]: 0.00010241 [accelerated_algorithm]: 2.204e-05 [shard]: 2.26998e-06 [meta_shard_fg_expand]: 2.61e-06 [shard_inline]: 7.36999e-06 [merge_send_recv]: 8.89e-06 [auto_parallel]: 9.10001e-06 [parallel]: 9.35001e-06 [flash_sp]: 4.88001e-06 [merge_comm]: 4.19002e-06 [allreduce_fusion]: 3.78001e-06 [matmul_add_comm_reduction]: 1.167e-05 [allreduce_slice_to_reducescatter]: 9.09989e-07 [virtual_shard_identity]: 9.71998e-06 [virtual_dataset]: 7.05002e-06 [get_grad_eliminate_]: 6.53998e-06 [virtual_output]: 6.07001e-06 [merge_forward]: 5.25999e-06 [cell_reuse_recompute_pass]: 1.67001e-06 [offload_activation]: 1.08e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.395e-05 [merge_recompute_call_nodes]: 2.04e-06 [before_grad]: 1.12e-05 [set_forward_comm_id_for_comm_node_pass]: 3.96001e-06 [meta_fg_expand]: 0.00033594 [flash_sp_send_recv_attached]: 2.91999e-06 [receive_attached]: 2.48998e-06 [after_resolve]: 2.126e-05 [a_after_grad]: 1.112e-05 [renormalize]: 0.0009399 [add_forward_monad_depend]: 5.79e-06 [auto_monad_grad]: 2.07001e-06 [auto_monad_eliminator]: 1.622e-05 [cse]: 3.341e-05 [a_3]: 5.106e-05 [Cycle 3]: 0.00071259, [45] [expand_dump_flag]: 2.10002e-06 [switch_simplify]: 7.47998e-06 [loop_unroll]: 6.43998e-06 [a_1]: 0.00015919 [with_stream_mark]: 1.243e-05 [recompute_prepare]: 7.1e-06 [updatestate_depend_eliminate]: 4.67998e-06 [updatestate_assign_eliminate]: 2.98998e-06 [updatestate_loads_eliminate]: 2.81999e-06 [parameter_eliminate]: 1.20001e-06 [a_2]: 8.631e-05 [accelerated_algorithm]: 1.1e-05 [shard]: 1.32e-06 [meta_shard_fg_expand]: 1.87001e-06 [shard_inline]: 6.45002e-06 [merge_send_recv]: 6.91001e-06 [auto_parallel]: 7.77e-06 [parallel]: 7.79002e-06 [flash_sp]: 1.02e-06 [merge_comm]: 3.62002e-06 [allreduce_fusion]: 3.48e-06 [matmul_add_comm_reduction]: 7.92e-06 [allreduce_slice_to_reducescatter]: 3.09985e-07 [virtual_shard_identity]: 7.65e-06 [virtual_dataset]: 6.19999e-06 [get_grad_eliminate_]: 5.92999e-06 [virtual_output]: 6.32001e-06 [merge_forward]: 3.85e-06 [cell_reuse_recompute_pass]: 2.30002e-06 [offload_activation]: 8.42e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.248e-05 [merge_recompute_call_nodes]: 8.40024e-07 [before_grad]: 1.044e-05 [set_forward_comm_id_for_comm_node_pass]: 4.12998e-06 [meta_fg_expand]: 2.45002e-06 [flash_sp_send_recv_attached]: 1.17e-06 [receive_attached]: 1.87001e-06 [after_resolve]: 9.95002e-06 [a_after_grad]: 9.13002e-06 [renormalize]: 5.00004e-08 [add_forward_monad_depend]: 1.76998e-06 [auto_monad_grad]: 1.27e-06 [auto_monad_eliminator]: 9.20001e-06 [cse]: 1.878e-05 [a_3]: 3.881e-05 [py_interpret_to_execute_after_opt_a]: 5.07e-06 [slice_cell_reuse_recomputed_activation]: 1.97999e-06 [rewriter_after_opt_a]: 2.177e-05 [convert_after_rewriter]: 1.30999e-06 [order_py_execute_after_rewriter]: 1.12999e-06 [mutable_eliminate]: 0.0006547 [opt_b]: 0.00021315, [1] [Cycle 1]: 0.0002054, [7] [b_1]: 0.00012944 [b_2]: 7.28e-06 [updatestate_depend_eliminate]: 5.66998e-06 [updatestate_assign_eliminate]: 2.83e-06 [updatestate_loads_eliminate]: 2.78e-06 [renormalize]: 6.69999e-07 [cse]: 2.277e-05 [optimize_parallel_all_gather_comm]: 1.688e-05 [overlap_param_gather]: 1.94e-06 [cconv]: 2.444e-05 [loop_unroll]: 0.00042745 [opt_after_cconv]: 0.00010652, [1] [Cycle 1]: 0.00010055, [7] [c_1]: 3.195e-05 [parameter_eliminate]: 2.42001e-06 [updatestate_depend_eliminate]: 5.77999e-06 [updatestate_assign_eliminate]: 2.83e-06 [updatestate_loads_eliminate]: 2.54001e-06 [cse]: 2.12e-05 [renormalize]: 4.49974e-07 [remove_dup_value]: 1.54e-05 [tuple_transform]: 7.797e-05, [1] [Cycle 1]: 7.353e-05, [4] [d_1]: 4.622e-05 [none_parameter_eliminate]: 1.71e-06 [renormalize]: 2.10013e-07 [switch_simplify]: 7.36001e-06 [partial_unused_args_eliminate]: 1.99e-06 [add_recomputation]: 4.774e-05 [cse_after_recomputation]: 2.551e-05, [1] [Cycle 1]: 2.105e-05, [1] [cse]: 1.536e-05 [environ_conv]: 5.57001e-06 [swap_dp_allreduce_reducescatter]: 5.39e-06 [bias_add_comm_swap]: 2.65997e-06 [label_micro_interleaved_index]: 3.98999e-06 [label_fine_grained_interleaved_index]: 2.78e-06 [merge_cast_opt]: 1.52999e-06 [slice_recompute_activation]: 1.91998e-06 [micro_interleaved_order_control]: 2.34001e-06 [assign_add_opt]: 1.21997e-06 [ForceFp32Comm]: 9.60019e-07 [remove_cast_before_assign_add]: 1.15001e-06 [full_micro_interleaved_order_control]: 2.37001e-06 [reorder_send_recv_between_fp_bp]: 2.76999e-06 [comm_op_add_attrs]: 9.70002e-07 [add_comm_op_reuse_tag]: 1.03001e-06 [interleave_split_concat_branches]: 1.17e-06 [interleave_parallel_branches]: 1.06002e-06 [overlap_opt_shard_in_pipeline]: 1.25999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.88002e-06 [control_data_broadcast_order]: 1.515e-05 [grouped_pairwise_exchange_alltoall]: 1.49e-06 [offloading_packed_experts]: 4.34997e-06 [overlap_recompute_and_grad_model_parallel]: 5.29003e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.13001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.33002e-06 [overlap_recompute_comm]: 2.78e-06 [overlap_grad_ring_attention]: 4.60001e-06 [overlap_grad_flash_sp]: 2.105e-05 [begin_end_overlap_inline]: 4.80009e-07 [split_matmul_comm_elemetwise]: 1.90001e-06 [split_layernorm_comm]: 1.81998e-06 [handle_group_info]: 9.20001e-07 [symbol_engine_optimizer]: 8.042e-05, [1] [Cycle 1]: 7.541e-05, [6] [build]: 3.23998e-06 [elim_shapecalc]: 1.104e-05 [elim_not_effective]: 1.534e-05 [opt_reshape]: 7.49002e-06 [fold_const_symbol]: 1.091e-05 [renormalize]: 1.60013e-07 [detach_backward]: 2.44001e-06 [pipeline_parallel_scheduler]: 1.45999e-06 [auto_monad_reorder]: 1.939e-05 [get_jit_bprop_graph]: 1.93002e-06 [rewriter_after_jit_bprop_graph]: 3.70998e-06 [opt_after_jit_grad]: 0.00045641 [validate]: 3.808e-05 Sums bootstrap : 0.000390s : 0.40% type_inference : 0.079504s : 80.96% event_method : 0.000071s : 0.07% auto_monad : 0.000256s : 0.26% graph_reusing : 0.000021s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000063s : 0.06% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000020s : 0.02% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000089s : 0.09% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000001s : 0.00% optimize.py_interpret_to_execute : 0.000005s : 0.00% optimize.rewriter_before_opt_a : 0.000508s : 0.52% optimize.opt_a.expand_dump_flag : 0.000017s : 0.02% optimize.opt_a.switch_simplify : 0.000335s : 0.34% optimize.opt_a.loop_unroll : 0.000159s : 0.16% optimize.opt_a.a_1 : 0.004370s : 4.45% optimize.opt_a.with_stream_mark : 0.000065s : 0.07% optimize.opt_a.recompute_prepare : 0.000048s : 0.05% optimize.opt_a.updatestate_depend_eliminate : 0.000019s : 0.02% optimize.opt_a.updatestate_assign_eliminate : 0.000015s : 0.02% optimize.opt_a.updatestate_loads_eliminate : 0.000014s : 0.01% optimize.opt_a.parameter_eliminate : 0.000007s : 0.01% optimize.opt_a.a_2 : 0.000488s : 0.50% optimize.opt_a.accelerated_algorithm : 0.000068s : 0.07% optimize.opt_a.shard : 0.000006s : 0.01% optimize.opt_a.meta_shard_fg_expand : 0.000012s : 0.01% optimize.opt_a.shard_inline : 0.000031s : 0.03% optimize.opt_a.merge_send_recv : 0.000036s : 0.04% optimize.opt_a.auto_parallel : 0.000033s : 0.03% optimize.opt_a.parallel : 0.000060s : 0.06% optimize.opt_a.flash_sp : 0.000018s : 0.02% optimize.opt_a.merge_comm : 0.000020s : 0.02% optimize.opt_a.allreduce_fusion : 0.000016s : 0.02% optimize.opt_a.matmul_add_comm_reduction : 0.000053s : 0.05% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000039s : 0.04% optimize.opt_a.virtual_dataset : 0.000030s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000029s : 0.03% optimize.opt_a.virtual_output : 0.000029s : 0.03% optimize.opt_a.merge_forward : 0.000020s : 0.02% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.01% optimize.opt_a.offload_activation : 0.000037s : 0.04% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000061s : 0.06% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000053s : 0.05% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000019s : 0.02% optimize.opt_a.meta_fg_expand : 0.002087s : 2.12% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.01% optimize.opt_a.receive_attached : 0.000007s : 0.01% optimize.opt_a.after_resolve : 0.000092s : 0.09% optimize.opt_a.a_after_grad : 0.000110s : 0.11% optimize.opt_a.renormalize : 0.005588s : 5.69% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.02% optimize.opt_a.auto_monad_grad : 0.000010s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000088s : 0.09% optimize.opt_a.cse : 0.000389s : 0.40% optimize.opt_a.a_3 : 0.000433s : 0.44% optimize.py_interpret_to_execute_after_opt_a : 0.000005s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000022s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000655s : 0.67% optimize.opt_b.b_1 : 0.000129s : 0.13% optimize.opt_b.b_2 : 0.000007s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000023s : 0.02% optimize.optimize_parallel_all_gather_comm : 0.000017s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000024s : 0.02% optimize.loop_unroll : 0.000427s : 0.44% optimize.opt_after_cconv.c_1 : 0.000032s : 0.03% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.cse : 0.000021s : 0.02% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000015s : 0.02% optimize.tuple_transform.d_1 : 0.000046s : 0.05% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000007s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000048s : 0.05% optimize.cse_after_recomputation.cse : 0.000015s : 0.02% optimize.environ_conv : 0.000006s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 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 : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000015s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000021s : 0.02% optimize.begin_end_overlap_inline : 0.000000s : 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 : 0.000003s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000015s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000007s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000011s : 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 : 0.000019s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000456s : 0.46% validate : 0.000038s : 0.04% Time group info: ------[substitution.] 0.001290 199 9.37% : 0.000121s : 6: substitution.arithmetic_simplify 0.69% : 0.000009s : 1: substitution.cast_eliminate 0.21% : 0.000003s : 3: substitution.elim_not_effective 0.64% : 0.000008s : 11: substitution.float_depend_g_call 0.37% : 0.000005s : 3: substitution.float_tuple_getitem_switch 0.13% : 0.000002s : 3: substitution.fold_const_symbol 0.48% : 0.000006s : 4: substitution.graph_param_transform 0.32% : 0.000004s : 2: substitution.incorporate_call 0.16% : 0.000002s : 2: substitution.incorporate_call_switch 65.93% : 0.000851s : 36: substitution.inline 1.54% : 0.000020s : 2: substitution.inline_without_move 0.81% : 0.000011s : 16: substitution.j_node_and_user_rematch 1.92% : 0.000025s : 3: substitution.less_batch_normalization 0.95% : 0.000012s : 7: substitution.minmaximum_grad 0.73% : 0.000009s : 11: substitution.partial_eliminate 1.05% : 0.000014s : 16: substitution.remove_not_recompute_node 2.13% : 0.000028s : 9: substitution.replace_applicator 0.85% : 0.000011s : 8: substitution.replace_old_param 0.32% : 0.000004s : 1: substitution.set_cell_output_no_recompute 1.41% : 0.000018s : 8: substitution.switch_simplify 1.81% : 0.000023s : 7: substitution.tuple_list_convert_item_index_to_positive 0.81% : 0.000010s : 7: substitution.tuple_list_get_item_const_eliminator 1.16% : 0.000015s : 7: substitution.tuple_list_get_item_depend_reorder 4.93% : 0.000064s : 19: substitution.tuple_list_get_item_eliminator 1.27% : 0.000016s : 7: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.079330 2 92.35% : 0.073259s : 1: type_inference.infer 7.65% : 0.006071s : 1: type_inference.specialize ------[replace.] 0.000413 54 1.41% : 0.000006s : 1: replace.cast_eliminate 59.71% : 0.000247s : 36: replace.inline 20.62% : 0.000085s : 8: replace.switch_simplify 18.26% : 0.000075s : 9: replace.tuple_list_get_item_eliminator ------[match.] 0.000879 54 0.96% : 0.000008s : 1: match.cast_eliminate 94.41% : 0.000830s : 36: match.inline 1.65% : 0.000015s : 8: match.switch_simplify 2.98% : 0.000026s : 9: match.tuple_list_get_item_eliminator ------[predicate.] 0.000850 5282 1.06% : 0.000009s : 69: predicate.accumulaten_eliminater 0.20% : 0.000002s : 4: predicate.ad_related_special_op_eliminate 0.36% : 0.000003s : 21: predicate.addn_check_dump 1.24% : 0.000011s : 69: predicate.addn_zero_filter 1.07% : 0.000009s : 69: predicate.adjust_all_reduce_mul_add 2.04% : 0.000017s : 90: predicate.arithmetic_simplify 1.17% : 0.000010s : 70: predicate.cast_eliminate 0.86% : 0.000007s : 51: predicate.check_bprop_eliminate 0.42% : 0.000004s : 21: predicate.compare_switch_simplify 0.04% : 0.000000s : 4: predicate.const_output_eliminate 0.38% : 0.000003s : 21: predicate.depend_value_elim 1.22% : 0.000010s : 70: predicate.dict_get_item_const_eliminator 1.25% : 0.000011s : 70: predicate.dict_get_item_eliminator 1.15% : 0.000010s : 70: predicate.dict_set_item_eliminator 0.22% : 0.000002s : 8: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 4: predicate.elim_not_effective 0.10% : 0.000001s : 4: predicate.elim_shapecalc_of_broadcastargs 1.19% : 0.000010s : 74: predicate.environ_add_const_eliminate 1.14% : 0.000010s : 74: predicate.environ_get_add_eliminate 1.22% : 0.000010s : 74: predicate.environ_get_depend_swap 1.64% : 0.000014s : 95: predicate.environ_get_eliminate 1.19% : 0.000010s : 74: predicate.environ_get_set_eliminate 1.94% : 0.000016s : 115: predicate.exchange_switch_depend_value 2.77% : 0.000024s : 115: predicate.float_depend_g_call 0.36% : 0.000003s : 21: predicate.float_environ_get_switch 0.47% : 0.000004s : 25: predicate.float_tuple_getitem_switch 0.04% : 0.000000s : 4: predicate.fold_const_symbol 0.46% : 0.000004s : 21: predicate.get_grad_eliminate 0.06% : 0.000001s : 4: predicate.graph_param_transform 0.39% : 0.000003s : 21: predicate.incorporate_call 0.33% : 0.000003s : 21: predicate.incorporate_call_switch 5.62% : 0.000048s : 234: predicate.inline 1.04% : 0.000009s : 43: predicate.inline_without_move 0.18% : 0.000002s : 21: predicate.j_node_and_user_rematch 0.59% : 0.000005s : 21: predicate.less_batch_normalization 1.52% : 0.000013s : 87: predicate.list_to_tuple_eliminator_ 2.53% : 0.000021s : 156: predicate.load_eliminater 0.20% : 0.000002s : 4: predicate.loop_unroll_after_grad 3.03% : 0.000026s : 179: predicate.loop_unroll_before_grad 1.30% : 0.000011s : 78: predicate.make_slice_get_slice_eliminator 0.39% : 0.000003s : 21: predicate.merge_addn 0.81% : 0.000007s : 51: predicate.micro_step_allgather_replace 0.81% : 0.000007s : 51: predicate.mini_step_allgather_replace 1.16% : 0.000010s : 69: predicate.minmaximum_grad 0.30% : 0.000003s : 4: predicate.mutable_eliminate 0.09% : 0.000001s : 4: predicate.opt_reshape 0.09% : 0.000001s : 4: predicate.parallel_virtual_node 2.67% : 0.000023s : 115: predicate.partial_defer_inline 1.52% : 0.000013s : 83: predicate.partial_eliminate 1.17% : 0.000010s : 69: predicate.print_const_string_wrapper 0.38% : 0.000003s : 21: predicate.reduce_all_const_elim 1.39% : 0.000012s : 69: predicate.reduce_eliminate 2.45% : 0.000021s : 156: predicate.redundant_stop_gradient_eliminater 0.33% : 0.000003s : 21: predicate.remove_not_recompute_node 1.50% : 0.000013s : 130: predicate.replace_applicator 0.47% : 0.000004s : 43: predicate.replace_old_param 0.08% : 0.000001s : 4: predicate.reset_defer_inline 1.19% : 0.000010s : 69: predicate.reshape_eliminate 0.87% : 0.000007s : 51: predicate.row_tensor_add_zeros_like 0.09% : 0.000001s : 4: predicate.row_tensor_eliminate 1.06% : 0.000009s : 51: predicate.same_eliminate 0.28% : 0.000002s : 21: predicate.set_cell_output_no_recompute 0.54% : 0.000005s : 21: predicate.shard_identity_eliminate 0.18% : 0.000002s : 8: predicate.special_op_eliminate 0.41% : 0.000004s : 21: predicate.specialize_transform 1.02% : 0.000009s : 51: predicate.split_environ_get_set_with_tuple_value 0.94% : 0.000008s : 43: predicate.stack_unstack_eliminate 0.08% : 0.000001s : 4: predicate.switch_call_monad_eliminater 2.10% : 0.000018s : 115: predicate.switch_defer_inline 6.92% : 0.000059s : 166: predicate.switch_layer_defer_inline 6.28% : 0.000053s : 335: predicate.switch_simplify 1.12% : 0.000010s : 69: predicate.tile_eliminate 1.13% : 0.000010s : 69: predicate.transpose_eliminate 1.43% : 0.000012s : 78: predicate.tuple_list_convert_item_index_to_positive 1.44% : 0.000012s : 78: predicate.tuple_list_get_item_const_eliminator 1.37% : 0.000012s : 78: predicate.tuple_list_get_item_depend_reorder 2.64% : 0.000022s : 108: predicate.tuple_list_get_item_eliminator 1.51% : 0.000013s : 78: predicate.tuple_list_get_set_item_eliminator 1.97% : 0.000017s : 99: predicate.tuple_list_set_item_eliminator 1.41% : 0.000012s : 87: predicate.tuple_to_list_eliminator_ 2.39% : 0.000020s : 156: predicate.updatestate_pure_node_eliminater 2.86% : 0.000024s : 177: predicate.updatestate_useless_node_eliminater 0.10% : 0.000001s : 4: predicate.value_based_eliminate 0.45% : 0.000004s : 21: predicate.virtual_dataset_eliminate 0.40% : 0.000003s : 21: predicate.virtual_output_eliminate 0.06% : 0.000001s : 4: predicate.virtual_view_grad_eliminate 0.10% : 0.000001s : 4: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003852 61 58.83% : 0.002266s : 21: func_graph_cloner_run.FuncGraphClonerGraph 41.17% : 0.001586s : 40: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.137111 233 0.01% : 0.000015s : 1: ForceFp32Comm 2.61% : 0.003581s : 1: add_attr 2.60% : 0.003571s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.04% : 0.000052s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.19% : 0.000267s : 1: auto_monad 0.02% : 0.000023s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.30% : 0.000416s : 1: bootstrap 0.02% : 0.000028s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000019s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.02% : 0.000028s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.06% : 0.000080s : 1: event_method 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000025s : 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% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.32% : 0.000436s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.48% : 0.000664s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.01% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000016s : 1: opt.transform.mutable_eliminate 4.54% : 0.006219s : 117: opt.transform.opt_a 0.02% : 0.000031s : 1: opt.transform.opt_after_cconv 0.02% : 0.000025s : 1: opt.transform.opt_after_jit_grad 0.08% : 0.000108s : 28: opt.transform.opt_b 0.04% : 0.000052s : 2: opt.transform.opt_trans_graph 0.03% : 0.000041s : 4: opt.transform.symbol_engine_opt 11.60% : 0.015908s : 1: opt_a 0.08% : 0.000111s : 1: opt_after_cconv 0.34% : 0.000466s : 1: opt_after_jit_grad 0.16% : 0.000217s : 1: opt_b 13.49% : 0.018496s : 1: optimize 0.01% : 0.000020s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.02% : 0.000025s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000008s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 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 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000006s : 1: overlap_recompute_comm 0.01% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000004s : 1: pipeline_parallel_scheduler 0.00% : 0.000004s : 1: pipeline_split 0.07% : 0.000094s : 1: pre_auto_parallel 0.01% : 0.000008s : 1: py_interpret_to_execute 0.01% : 0.000008s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000005s : 1: remove_cast_before_assign_add 0.01% : 0.000019s : 1: remove_dup_value 2.56% : 0.003505s : 2: renormalize.infer 1.51% : 0.002064s : 2: renormalize.specialize 0.00% : 0.000005s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000007s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000025s : 1: rewriter_after_opt_a 0.38% : 0.000516s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000004s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000008s : 1: swap_dp_allreduce_reducescatter 0.06% : 0.000083s : 1: symbol_engine_optimizer 0.06% : 0.000081s : 1: tuple_transform 58.00% : 0.079527s : 1: type_inference TotalTime = 0.0761258, [21] [bootstrap]: 0.00039713 [type_inference]: 0.0620624 [event_method]: 0.00046673 [auto_monad]: 0.00019073 [graph_reusing]: 1.295e-05 [inline]: 3.61001e-06 [add_attr]: 0.00381136, [1] [add_attr_with_inline]: 0.00379918, [1] [Cycle 1]: 0.00010081, [2] [tag_attr]: 5.48e-05 [meta_addattr_fg_expand]: 1.261e-05 [parallel-infer-symbol]: 3.41999e-06 [pre_auto_parallel]: 7.134e-05 [insert-virtual-dataset]: 2.32001e-06 [parallel-infer-symbol-second]: 1.86e-06 [dataset_repeat_opt]: 1.64e-06 [pipeline_split]: 1.78002e-06 [optimize]: 0.00829309, [53] [py_interpret_to_execute]: 6.06e-06 [rewriter_before_opt_a]: 0.00038042 [opt_a]: 0.00581444, [2] [Cycle 1]: 0.00508319, [45] [expand_dump_flag]: 5.88002e-06 [switch_simplify]: 0.0002021 [loop_unroll]: 7.067e-05 [a_1]: 0.00167192 [with_stream_mark]: 2.107e-05 [recompute_prepare]: 9.91e-06 [updatestate_depend_eliminate]: 4.22e-06 [updatestate_assign_eliminate]: 3.25e-06 [updatestate_loads_eliminate]: 3.40998e-06 [parameter_eliminate]: 2.68998e-06 [a_2]: 8.84e-05 [accelerated_algorithm]: 0.00031858 [shard]: 2.53e-06 [meta_shard_fg_expand]: 4.20999e-06 [shard_inline]: 8.19002e-06 [merge_send_recv]: 8.84e-06 [auto_parallel]: 9.40001e-06 [parallel]: 2.155e-05 [flash_sp]: 1.016e-05 [merge_comm]: 3.36999e-06 [allreduce_fusion]: 3.2e-06 [matmul_add_comm_reduction]: 1.145e-05 [allreduce_slice_to_reducescatter]: 7.39994e-07 [virtual_shard_identity]: 7.85998e-06 [virtual_dataset]: 6.07001e-06 [get_grad_eliminate_]: 6.12001e-06 [virtual_output]: 5.85002e-06 [merge_forward]: 4.26001e-06 [cell_reuse_recompute_pass]: 1.75001e-06 [offload_activation]: 9.26998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.306e-05 [merge_recompute_call_nodes]: 1.77999e-06 [before_grad]: 9.99001e-06 [set_forward_comm_id_for_comm_node_pass]: 3.41999e-06 [meta_fg_expand]: 4.12e-06 [flash_sp_send_recv_attached]: 4.89998e-06 [receive_attached]: 2.17001e-06 [after_resolve]: 1.058e-05 [a_after_grad]: 9.20999e-06 [renormalize]: 0.00209919 [add_forward_monad_depend]: 8.50001e-06 [auto_monad_grad]: 2.61999e-06 [auto_monad_eliminator]: 2.104e-05 [cse]: 3.15e-05 [a_3]: 5.306e-05 [Cycle 2]: 0.00071543, [45] [expand_dump_flag]: 2.59001e-06 [switch_simplify]: 8.47e-06 [loop_unroll]: 5.88002e-06 [a_1]: 0.00014567 [with_stream_mark]: 1.393e-05 [recompute_prepare]: 5.92001e-06 [updatestate_depend_eliminate]: 4.07e-06 [updatestate_assign_eliminate]: 3.18e-06 [updatestate_loads_eliminate]: 3.75998e-06 [parameter_eliminate]: 1.50001e-06 [a_2]: 7.857e-05 [accelerated_algorithm]: 1.048e-05 [shard]: 2.09e-06 [meta_shard_fg_expand]: 2.27999e-06 [shard_inline]: 6.64001e-06 [merge_send_recv]: 8.05999e-06 [auto_parallel]: 8.85999e-06 [parallel]: 6.88e-06 [flash_sp]: 3.28e-06 [merge_comm]: 3.06001e-06 [allreduce_fusion]: 3.08998e-06 [matmul_add_comm_reduction]: 9.41e-06 [allreduce_slice_to_reducescatter]: 7.39994e-07 [virtual_shard_identity]: 6.24999e-06 [virtual_dataset]: 5.24e-06 [get_grad_eliminate_]: 5.15001e-06 [virtual_output]: 5.09e-06 [merge_forward]: 3.85e-06 [cell_reuse_recompute_pass]: 1.99e-06 [offload_activation]: 8.64998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.054e-05 [merge_recompute_call_nodes]: 1.28002e-06 [before_grad]: 8.15e-06 [set_forward_comm_id_for_comm_node_pass]: 3.81001e-06 [meta_fg_expand]: 2.24001e-06 [flash_sp_send_recv_attached]: 1.39998e-06 [receive_attached]: 1.89e-06 [after_resolve]: 8.84998e-06 [a_after_grad]: 7.58999e-06 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 1.39998e-06 [auto_monad_grad]: 1.37999e-06 [auto_monad_eliminator]: 3.843e-05 [cse]: 1.731e-05 [a_3]: 3.362e-05 [py_interpret_to_execute_after_opt_a]: 6.04001e-06 [slice_cell_reuse_recomputed_activation]: 2.51998e-06 [rewriter_after_opt_a]: 2.027e-05 [convert_after_rewriter]: 1.22e-06 [order_py_execute_after_rewriter]: 1.17e-06 [mutable_eliminate]: 0.00068603 [opt_b]: 0.00019159, [1] [Cycle 1]: 0.00018398, [7] [b_1]: 0.00010973 [b_2]: 8.02e-06 [updatestate_depend_eliminate]: 6.28002e-06 [updatestate_assign_eliminate]: 2.46998e-06 [updatestate_loads_eliminate]: 2.12999e-06 [renormalize]: 8.09989e-07 [cse]: 2.031e-05 [optimize_parallel_all_gather_comm]: 1.805e-05 [overlap_param_gather]: 1.86998e-06 [cconv]: 2.928e-05 [loop_unroll]: 0.00045936 [opt_after_cconv]: 0.00010149, [1] [Cycle 1]: 9.438e-05, [7] [c_1]: 2.757e-05 [parameter_eliminate]: 4.31002e-06 [updatestate_depend_eliminate]: 5.33002e-06 [updatestate_assign_eliminate]: 2.36e-06 [updatestate_loads_eliminate]: 2.12999e-06 [cse]: 1.965e-05 [renormalize]: 4.89992e-07 [remove_dup_value]: 1.678e-05 [tuple_transform]: 7.044e-05, [1] [Cycle 1]: 6.592e-05, [4] [d_1]: 3.883e-05 [none_parameter_eliminate]: 1.65001e-06 [renormalize]: 1.60013e-07 [switch_simplify]: 6.21998e-06 [partial_unused_args_eliminate]: 2.41e-06 [add_recomputation]: 4.612e-05 [cse_after_recomputation]: 2.236e-05, [1] [Cycle 1]: 1.73e-05, [1] [cse]: 1.218e-05 [environ_conv]: 5.42999e-06 [swap_dp_allreduce_reducescatter]: 5.27999e-06 [bias_add_comm_swap]: 2.53e-06 [label_micro_interleaved_index]: 4.43001e-06 [label_fine_grained_interleaved_index]: 2.59001e-06 [merge_cast_opt]: 1.41998e-06 [slice_recompute_activation]: 2.17999e-06 [micro_interleaved_order_control]: 2.63998e-06 [assign_add_opt]: 1.25999e-06 [ForceFp32Comm]: 1.10001e-06 [remove_cast_before_assign_add]: 1.00999e-06 [full_micro_interleaved_order_control]: 2.24999e-06 [reorder_send_recv_between_fp_bp]: 2.63e-06 [comm_op_add_attrs]: 1.44003e-06 [add_comm_op_reuse_tag]: 1.28002e-06 [interleave_split_concat_branches]: 1.10999e-06 [interleave_parallel_branches]: 9.89996e-07 [overlap_opt_shard_in_pipeline]: 1.90001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.03997e-06 [control_data_broadcast_order]: 1.312e-05 [grouped_pairwise_exchange_alltoall]: 1.71e-06 [offloading_packed_experts]: 4.22e-06 [overlap_recompute_and_grad_model_parallel]: 5.19e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.20999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.36002e-06 [overlap_recompute_comm]: 2.36e-06 [overlap_grad_ring_attention]: 3.82002e-06 [overlap_grad_flash_sp]: 2.009e-05 [begin_end_overlap_inline]: 8.70001e-07 [split_matmul_comm_elemetwise]: 2.08002e-06 [split_layernorm_comm]: 1.64e-06 [handle_group_info]: 1.05001e-06 [symbol_engine_optimizer]: 8.08e-05, [1] [Cycle 1]: 7.536e-05, [6] [build]: 3.51001e-06 [elim_shapecalc]: 1.172e-05 [elim_not_effective]: 1.251e-05 [opt_reshape]: 5.71e-06 [fold_const_symbol]: 9.74e-06 [renormalize]: 1.60013e-07 [detach_backward]: 2.34001e-06 [pipeline_parallel_scheduler]: 1.40001e-06 [auto_monad_reorder]: 1.794e-05 [get_jit_bprop_graph]: 2.12999e-06 [rewriter_after_jit_bprop_graph]: 5.47999e-06 [opt_after_jit_grad]: 0.00053977 [validate]: 4.664e-05 Sums bootstrap : 0.000397s : 0.56% type_inference : 0.062062s : 87.03% event_method : 0.000467s : 0.65% auto_monad : 0.000191s : 0.27% graph_reusing : 0.000013s : 0.02% inline : 0.000004s : 0.01% add_attr.add_attr_with_inline.tag_attr : 0.000055s : 0.08% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000013s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000071s : 0.10% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000006s : 0.01% optimize.rewriter_before_opt_a : 0.000380s : 0.53% optimize.opt_a.expand_dump_flag : 0.000008s : 0.01% optimize.opt_a.switch_simplify : 0.000211s : 0.30% optimize.opt_a.loop_unroll : 0.000077s : 0.11% optimize.opt_a.a_1 : 0.001818s : 2.55% optimize.opt_a.with_stream_mark : 0.000035s : 0.05% optimize.opt_a.recompute_prepare : 0.000016s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000008s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000006s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000007s : 0.01% optimize.opt_a.parameter_eliminate : 0.000004s : 0.01% optimize.opt_a.a_2 : 0.000167s : 0.23% optimize.opt_a.accelerated_algorithm : 0.000329s : 0.46% optimize.opt_a.shard : 0.000005s : 0.01% optimize.opt_a.meta_shard_fg_expand : 0.000006s : 0.01% optimize.opt_a.shard_inline : 0.000015s : 0.02% optimize.opt_a.merge_send_recv : 0.000017s : 0.02% optimize.opt_a.auto_parallel : 0.000018s : 0.03% optimize.opt_a.parallel : 0.000028s : 0.04% optimize.opt_a.flash_sp : 0.000013s : 0.02% optimize.opt_a.merge_comm : 0.000006s : 0.01% optimize.opt_a.allreduce_fusion : 0.000006s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000021s : 0.03% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000014s : 0.02% optimize.opt_a.virtual_dataset : 0.000011s : 0.02% optimize.opt_a.get_grad_eliminate_ : 0.000011s : 0.02% optimize.opt_a.virtual_output : 0.000011s : 0.02% optimize.opt_a.merge_forward : 0.000008s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.01% optimize.opt_a.offload_activation : 0.000018s : 0.03% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000024s : 0.03% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000018s : 0.03% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000007s : 0.01% optimize.opt_a.meta_fg_expand : 0.000006s : 0.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000006s : 0.01% optimize.opt_a.receive_attached : 0.000004s : 0.01% optimize.opt_a.after_resolve : 0.000019s : 0.03% optimize.opt_a.a_after_grad : 0.000017s : 0.02% optimize.opt_a.renormalize : 0.002099s : 2.94% optimize.opt_a.add_forward_monad_depend : 0.000010s : 0.01% optimize.opt_a.auto_monad_grad : 0.000004s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000059s : 0.08% optimize.opt_a.cse : 0.000049s : 0.07% optimize.opt_a.a_3 : 0.000087s : 0.12% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000020s : 0.03% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000686s : 0.96% optimize.opt_b.b_1 : 0.000110s : 0.15% optimize.opt_b.b_2 : 0.000008s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000020s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000018s : 0.03% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000029s : 0.04% optimize.loop_unroll : 0.000459s : 0.64% optimize.opt_after_cconv.c_1 : 0.000028s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.01% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000005s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.cse : 0.000020s : 0.03% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.02% optimize.tuple_transform.d_1 : 0.000039s : 0.05% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000006s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000046s : 0.06% optimize.cse_after_recomputation.cse : 0.000012s : 0.02% optimize.environ_conv : 0.000005s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000005s : 0.01% optimize.bias_add_comm_swap : 0.000003s : 0.00% optimize.label_micro_interleaved_index : 0.000004s : 0.01% 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 : 0.000013s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000004s : 0.01% optimize.overlap_recompute_and_grad_model_parallel : 0.000005s : 0.01% 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 : 0.000004s : 0.01% optimize.overlap_grad_flash_sp : 0.000020s : 0.03% 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000012s : 0.02% optimize.symbol_engine_optimizer.elim_not_effective : 0.000013s : 0.02% optimize.symbol_engine_optimizer.opt_reshape : 0.000006s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000010s : 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 : 0.000018s : 0.03% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.01% opt_after_jit_grad : 0.000540s : 0.76% validate : 0.000047s : 0.07% Time group info: ------[substitution.] 0.000914 51 9.74% : 0.000089s : 5: substitution.arithmetic_simplify 1.35% : 0.000012s : 1: substitution.cast_eliminate 0.20% : 0.000002s : 2: substitution.elim_not_effective 0.12% : 0.000001s : 2: substitution.fold_const_symbol 0.62% : 0.000006s : 3: substitution.graph_param_transform 50.18% : 0.000459s : 19: substitution.inline 0.40% : 0.000004s : 4: substitution.j_node_and_user_rematch 33.52% : 0.000306s : 2: substitution.less_batch_normalization 0.58% : 0.000005s : 4: substitution.remove_not_recompute_node 0.46% : 0.000004s : 2: substitution.replace_old_param 1.64% : 0.000015s : 5: substitution.switch_simplify 1.18% : 0.000011s : 2: substitution.tuple_list_get_item_eliminator ------[type_inference.] 0.061941 2 92.90% : 0.057541s : 1: type_inference.infer 7.10% : 0.004400s : 1: type_inference.specialize ------[replace.] 0.000230 27 3.47% : 0.000008s : 1: replace.cast_eliminate 59.99% : 0.000138s : 19: replace.inline 28.88% : 0.000066s : 5: replace.switch_simplify 7.66% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000479 27 2.31% : 0.000011s : 1: match.cast_eliminate 93.37% : 0.000447s : 19: match.inline 2.40% : 0.000011s : 5: match.switch_simplify 1.92% : 0.000009s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000330 1960 1.13% : 0.000004s : 26: predicate.accumulaten_eliminater 0.57% : 0.000002s : 3: predicate.ad_related_special_op_eliminate 0.26% : 0.000001s : 6: predicate.addn_check_dump 1.23% : 0.000004s : 26: predicate.addn_zero_filter 1.05% : 0.000003s : 26: predicate.adjust_all_reduce_mul_add 2.21% : 0.000007s : 32: predicate.arithmetic_simplify 1.37% : 0.000005s : 27: predicate.cast_eliminate 0.32% : 0.000001s : 6: predicate.check_bprop_eliminate 0.27% : 0.000001s : 6: predicate.compare_switch_simplify 0.08% : 0.000000s : 3: predicate.const_output_eliminate 0.34% : 0.000001s : 6: predicate.depend_value_elim 1.34% : 0.000004s : 27: predicate.dict_get_item_const_eliminator 1.38% : 0.000005s : 27: predicate.dict_get_item_eliminator 1.17% : 0.000004s : 27: predicate.dict_set_item_eliminator 0.61% : 0.000002s : 6: predicate.dumpgradient_eliminate 0.12% : 0.000000s : 3: predicate.elim_not_effective 0.22% : 0.000001s : 3: predicate.elim_shapecalc_of_broadcastargs 1.34% : 0.000004s : 30: predicate.environ_add_const_eliminate 1.28% : 0.000004s : 30: predicate.environ_get_add_eliminate 1.34% : 0.000004s : 30: predicate.environ_get_depend_swap 1.61% : 0.000005s : 36: predicate.environ_get_eliminate 1.31% : 0.000004s : 30: predicate.environ_get_set_eliminate 2.08% : 0.000007s : 48: predicate.exchange_switch_depend_value 3.21% : 0.000011s : 48: predicate.float_depend_g_call 0.28% : 0.000001s : 6: predicate.float_environ_get_switch 0.35% : 0.000001s : 9: predicate.float_tuple_getitem_switch 0.08% : 0.000000s : 3: predicate.fold_const_symbol 0.37% : 0.000001s : 6: predicate.get_grad_eliminate 0.11% : 0.000000s : 3: predicate.graph_param_transform 0.33% : 0.000001s : 6: predicate.incorporate_call 0.23% : 0.000001s : 6: predicate.incorporate_call_switch 7.03% : 0.000023s : 92: predicate.inline 0.34% : 0.000001s : 6: predicate.inline_without_move 0.17% : 0.000001s : 6: predicate.j_node_and_user_rematch 0.64% : 0.000002s : 6: predicate.less_batch_normalization 1.64% : 0.000005s : 35: predicate.list_to_tuple_eliminator_ 2.58% : 0.000009s : 61: predicate.load_eliminater 0.55% : 0.000002s : 3: predicate.loop_unroll_after_grad 3.71% : 0.000012s : 84: predicate.loop_unroll_before_grad 1.57% : 0.000005s : 33: predicate.make_slice_get_slice_eliminator 0.27% : 0.000001s : 6: predicate.merge_addn 0.29% : 0.000001s : 6: predicate.micro_step_allgather_replace 0.29% : 0.000001s : 6: predicate.mini_step_allgather_replace 1.06% : 0.000004s : 26: predicate.minmaximum_grad 0.71% : 0.000002s : 3: predicate.mutable_eliminate 0.16% : 0.000001s : 3: predicate.opt_reshape 0.19% : 0.000001s : 3: predicate.parallel_virtual_node 2.94% : 0.000010s : 48: predicate.partial_defer_inline 1.47% : 0.000005s : 32: predicate.partial_eliminate 1.18% : 0.000004s : 26: predicate.print_const_string_wrapper 0.30% : 0.000001s : 6: predicate.reduce_all_const_elim 1.69% : 0.000006s : 26: predicate.reduce_eliminate 2.55% : 0.000008s : 61: predicate.redundant_stop_gradient_eliminater 0.25% : 0.000001s : 6: predicate.remove_not_recompute_node 1.27% : 0.000004s : 35: predicate.replace_applicator 0.26% : 0.000001s : 6: predicate.replace_old_param 0.13% : 0.000000s : 3: predicate.reset_defer_inline 1.25% : 0.000004s : 26: predicate.reshape_eliminate 0.32% : 0.000001s : 6: predicate.row_tensor_add_zeros_like 0.26% : 0.000001s : 3: predicate.row_tensor_eliminate 0.48% : 0.000002s : 6: predicate.same_eliminate 0.27% : 0.000001s : 6: predicate.set_cell_output_no_recompute 0.39% : 0.000001s : 6: predicate.shard_identity_eliminate 0.46% : 0.000002s : 6: predicate.special_op_eliminate 0.35% : 0.000001s : 6: predicate.specialize_transform 0.54% : 0.000002s : 6: predicate.split_environ_get_set_with_tuple_value 0.41% : 0.000001s : 6: predicate.stack_unstack_eliminate 0.21% : 0.000001s : 3: predicate.switch_call_monad_eliminater 2.34% : 0.000008s : 48: predicate.switch_defer_inline 2.54% : 0.000008s : 54: predicate.switch_layer_defer_inline 7.48% : 0.000025s : 151: predicate.switch_simplify 1.16% : 0.000004s : 26: predicate.tile_eliminate 1.13% : 0.000004s : 26: predicate.transpose_eliminate 1.67% : 0.000006s : 33: predicate.tuple_list_convert_item_index_to_positive 1.56% : 0.000005s : 33: predicate.tuple_list_get_item_const_eliminator 1.52% : 0.000005s : 33: predicate.tuple_list_get_item_depend_reorder 2.65% : 0.000009s : 41: predicate.tuple_list_get_item_eliminator 1.69% : 0.000006s : 33: predicate.tuple_list_get_set_item_eliminator 2.21% : 0.000007s : 39: predicate.tuple_list_set_item_eliminator 1.53% : 0.000005s : 35: predicate.tuple_to_list_eliminator_ 2.44% : 0.000008s : 61: predicate.updatestate_pure_node_eliminater 2.94% : 0.000010s : 67: predicate.updatestate_useless_node_eliminater 0.21% : 0.000001s : 3: predicate.value_based_eliminate 0.35% : 0.000001s : 6: predicate.virtual_dataset_eliminate 0.32% : 0.000001s : 6: predicate.virtual_output_eliminate 0.11% : 0.000000s : 3: predicate.virtual_view_grad_eliminate 0.37% : 0.000001s : 3: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002181 33 44.24% : 0.000965s : 12: func_graph_cloner_run.FuncGraphClonerGraph 55.76% : 0.001216s : 21: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.093172 192 0.00% : 0.000004s : 1: ForceFp32Comm 4.10% : 0.003817s : 1: add_attr 4.08% : 0.003804s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.05% : 0.000051s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.22% : 0.000204s : 1: auto_monad 0.02% : 0.000023s : 1: auto_monad_reorder 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000005s : 1: bias_add_comm_swap 0.45% : 0.000418s : 1: bootstrap 0.04% : 0.000033s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.02% : 0.000017s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.03% : 0.000025s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.01% : 0.000007s : 1: detach_backward 0.01% : 0.000010s : 1: environ_conv 0.52% : 0.000484s : 1: event_method 0.01% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000018s : 1: graph_reusing 0.01% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.01% : 0.000007s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.01% : 0.000006s : 1: label_fine_grained_interleaved_index 0.01% : 0.000007s : 1: label_micro_interleaved_index 0.50% : 0.000469s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 0.75% : 0.000696s : 1: mutable_eliminate 0.01% : 0.000007s : 1: offloading_packed_experts 0.02% : 0.000014s : 1: opt.transform.loop_unroll_optimizer 0.02% : 0.000015s : 1: opt.transform.mutable_eliminate 2.95% : 0.002746s : 78: opt.transform.opt_a 0.03% : 0.000026s : 1: opt.transform.opt_after_cconv 0.03% : 0.000025s : 1: opt.transform.opt_after_jit_grad 0.10% : 0.000090s : 28: opt.transform.opt_b 0.05% : 0.000043s : 2: opt.transform.opt_trans_graph 0.04% : 0.000035s : 4: opt.transform.symbol_engine_opt 6.24% : 0.005818s : 1: opt_a 0.11% : 0.000105s : 1: opt_after_cconv 0.60% : 0.000556s : 1: opt_after_jit_grad 0.21% : 0.000195s : 1: opt_b 8.91% : 0.008300s : 1: optimize 0.02% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.03% : 0.000024s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000007s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.01% : 0.000005s : 1: overlap_opt_shard_in_pipeline 0.01% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000008s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000005s : 1: overlap_recompute_comm 0.01% : 0.000007s : 1: parallel-infer-symbol 0.01% : 0.000005s : 1: parallel-infer-symbol-second 0.01% : 0.000005s : 1: partial_unused_args_eliminate 0.01% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.08% : 0.000077s : 1: pre_auto_parallel 0.01% : 0.000010s : 1: py_interpret_to_execute 0.01% : 0.000009s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.02% : 0.000020s : 1: remove_dup_value 1.21% : 0.001131s : 1: renormalize.infer 1.03% : 0.000958s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.03% : 0.000023s : 1: rewriter_after_opt_a 0.42% : 0.000389s : 1: rewriter_before_opt_a 0.01% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.01% : 0.000005s : 1: slice_recompute_activation 0.01% : 0.000005s : 1: split_layernorm_comm 0.01% : 0.000005s : 1: split_matmul_comm_elemetwise 0.01% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.09% : 0.000084s : 1: symbol_engine_optimizer 0.08% : 0.000073s : 1: tuple_transform 66.64% : 0.062089s : 1: type_inference TotalTime = 0.105922, [21] [bootstrap]: 0.00043463 [type_inference]: 0.0814001 [event_method]: 7.161e-05 [auto_monad]: 0.00024711 [graph_reusing]: 1.995e-05 [inline]: 2.67001e-06 [add_attr]: 0.00365957, [1] [add_attr_with_inline]: 0.00364836, [1] [Cycle 1]: 0.00013134, [2] [tag_attr]: 7.125e-05 [meta_addattr_fg_expand]: 1.992e-05 [parallel-infer-symbol]: 3.25e-06 [pre_auto_parallel]: 9.581e-05 [insert-virtual-dataset]: 2.31e-06 [parallel-infer-symbol-second]: 1.71e-06 [dataset_repeat_opt]: 1.89e-06 [pipeline_split]: 1.56002e-06 [optimize]: 0.0192158, [53] [py_interpret_to_execute]: 7.07997e-06 [rewriter_before_opt_a]: 0.0005504 [opt_a]: 0.0163693, [3] [Cycle 1]: 0.0120371, [45] [expand_dump_flag]: 1.109e-05 [switch_simplify]: 0.00029491 [loop_unroll]: 0.00010995 [a_1]: 0.00290114 [with_stream_mark]: 3.081e-05 [recompute_prepare]: 2.705e-05 [updatestate_depend_eliminate]: 9.14e-06 [updatestate_assign_eliminate]: 7.75e-06 [updatestate_loads_eliminate]: 7.75998e-06 [parameter_eliminate]: 3.41999e-06 [a_2]: 0.00025463 [accelerated_algorithm]: 2.881e-05 [shard]: 1.69998e-06 [meta_shard_fg_expand]: 6.84999e-06 [shard_inline]: 1.592e-05 [merge_send_recv]: 1.739e-05 [auto_parallel]: 1.329e-05 [parallel]: 1.958e-05 [flash_sp]: 1.2e-05 [merge_comm]: 9.69999e-06 [allreduce_fusion]: 8.65999e-06 [matmul_add_comm_reduction]: 3.103e-05 [allreduce_slice_to_reducescatter]: 6.69999e-07 [virtual_shard_identity]: 1.882e-05 [virtual_dataset]: 1.464e-05 [get_grad_eliminate_]: 1.494e-05 [virtual_output]: 1.494e-05 [merge_forward]: 9.73002e-06 [cell_reuse_recompute_pass]: 1.01002e-06 [offload_activation]: 1.823e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.907e-05 [merge_recompute_call_nodes]: 1.50001e-06 [before_grad]: 2.835e-05 [set_forward_comm_id_for_comm_node_pass]: 1.02e-05 [meta_fg_expand]: 0.00168489 [flash_sp_send_recv_attached]: 4.82998e-06 [receive_attached]: 1.03e-05 [after_resolve]: 6.681e-05 [a_after_grad]: 8.472e-05 [renormalize]: 0.00506775 [add_forward_monad_depend]: 1.444e-05 [auto_monad_grad]: 6.36e-06 [auto_monad_eliminator]: 6.641e-05 [cse]: 0.00034544 [a_3]: 0.00034102 [Cycle 2]: 0.00359474, [45] [expand_dump_flag]: 3.06001e-06 [switch_simplify]: 4.562e-05 [loop_unroll]: 4.055e-05 [a_1]: 0.00140586 [with_stream_mark]: 2.517e-05 [recompute_prepare]: 1.225e-05 [updatestate_depend_eliminate]: 4.99e-06 [updatestate_assign_eliminate]: 4.14002e-06 [updatestate_loads_eliminate]: 3.51999e-06 [parameter_eliminate]: 2.21e-06 [a_2]: 0.00010091 [accelerated_algorithm]: 3.625e-05 [shard]: 2.37001e-06 [meta_shard_fg_expand]: 2.97002e-06 [shard_inline]: 7.99002e-06 [merge_send_recv]: 9.31998e-06 [auto_parallel]: 1.04e-05 [parallel]: 9.72999e-06 [flash_sp]: 4.52e-06 [merge_comm]: 3.83999e-06 [allreduce_fusion]: 4.00998e-06 [matmul_add_comm_reduction]: 1.116e-05 [allreduce_slice_to_reducescatter]: 9.30013e-07 [virtual_shard_identity]: 1.068e-05 [virtual_dataset]: 7.63999e-06 [get_grad_eliminate_]: 7.43e-06 [virtual_output]: 7.61999e-06 [merge_forward]: 5.69e-06 [cell_reuse_recompute_pass]: 1.52999e-06 [offload_activation]: 1.152e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.527e-05 [merge_recompute_call_nodes]: 1.46002e-06 [before_grad]: 1.196e-05 [set_forward_comm_id_for_comm_node_pass]: 5.12999e-06 [meta_fg_expand]: 0.00033625 [flash_sp_send_recv_attached]: 2.01e-06 [receive_attached]: 3.18e-06 [after_resolve]: 2.056e-05 [a_after_grad]: 1.114e-05 [renormalize]: 0.00099313 [add_forward_monad_depend]: 6.12001e-06 [auto_monad_grad]: 2.49999e-06 [auto_monad_eliminator]: 1.696e-05 [cse]: 3.52e-05 [a_3]: 5.151e-05 [Cycle 3]: 0.00071435, [45] [expand_dump_flag]: 1.82999e-06 [switch_simplify]: 8.18001e-06 [loop_unroll]: 6.56e-06 [a_1]: 0.00016062 [with_stream_mark]: 1.074e-05 [recompute_prepare]: 6.63998e-06 [updatestate_depend_eliminate]: 4.2e-06 [updatestate_assign_eliminate]: 3.45e-06 [updatestate_loads_eliminate]: 3.13e-06 [parameter_eliminate]: 1.74e-06 [a_2]: 8.765e-05 [accelerated_algorithm]: 9.87001e-06 [shard]: 1.27999e-06 [meta_shard_fg_expand]: 1.81e-06 [shard_inline]: 6.78003e-06 [merge_send_recv]: 6.58998e-06 [auto_parallel]: 7.23999e-06 [parallel]: 6.28e-06 [flash_sp]: 1.13001e-06 [merge_comm]: 3.66001e-06 [allreduce_fusion]: 3.7e-06 [matmul_add_comm_reduction]: 7.74002e-06 [allreduce_slice_to_reducescatter]: 7.2e-07 [virtual_shard_identity]: 7.59002e-06 [virtual_dataset]: 5.82999e-06 [get_grad_eliminate_]: 6.02999e-06 [virtual_output]: 6.36e-06 [merge_forward]: 3.81001e-06 [cell_reuse_recompute_pass]: 1.72999e-06 [offload_activation]: 8.96002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.306e-05 [merge_recompute_call_nodes]: 1.06002e-06 [before_grad]: 1.102e-05 [set_forward_comm_id_for_comm_node_pass]: 4.52e-06 [meta_fg_expand]: 2.51e-06 [flash_sp_send_recv_attached]: 1.04998e-06 [receive_attached]: 1.39e-06 [after_resolve]: 9.54e-06 [a_after_grad]: 9.32999e-06 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.53002e-06 [auto_monad_grad]: 1.03001e-06 [auto_monad_eliminator]: 8.2e-06 [cse]: 1.954e-05 [a_3]: 3.864e-05 [py_interpret_to_execute_after_opt_a]: 5.66998e-06 [slice_cell_reuse_recomputed_activation]: 2.48e-06 [rewriter_after_opt_a]: 2.292e-05 [convert_after_rewriter]: 1.29e-06 [order_py_execute_after_rewriter]: 1.15001e-06 [mutable_eliminate]: 0.00073833 [opt_b]: 0.00022696, [1] [Cycle 1]: 0.00021826, [7] [b_1]: 0.00013286 [b_2]: 9.64e-06 [updatestate_depend_eliminate]: 7.14001e-06 [updatestate_assign_eliminate]: 2.83e-06 [updatestate_loads_eliminate]: 2.96999e-06 [renormalize]: 5.79981e-07 [cse]: 2.837e-05 [optimize_parallel_all_gather_comm]: 1.882e-05 [overlap_param_gather]: 1.79998e-06 [cconv]: 2.901e-05 [loop_unroll]: 0.00049658 [opt_after_cconv]: 0.0001122, [1] [Cycle 1]: 0.00010588, [7] [c_1]: 3.204e-05 [parameter_eliminate]: 3.38999e-06 [updatestate_depend_eliminate]: 6.01e-06 [updatestate_assign_eliminate]: 2.83e-06 [updatestate_loads_eliminate]: 3.75998e-06 [cse]: 2.42e-05 [renormalize]: 3.39991e-07 [remove_dup_value]: 1.584e-05 [tuple_transform]: 8.017e-05, [1] [Cycle 1]: 7.372e-05, [4] [d_1]: 4.633e-05 [none_parameter_eliminate]: 1.66998e-06 [renormalize]: 2.09984e-07 [switch_simplify]: 7.6e-06 [partial_unused_args_eliminate]: 1.66e-06 [add_recomputation]: 5.413e-05 [cse_after_recomputation]: 2.597e-05, [1] [Cycle 1]: 2.062e-05, [1] [cse]: 1.533e-05 [environ_conv]: 6.36e-06 [swap_dp_allreduce_reducescatter]: 5.98002e-06 [bias_add_comm_swap]: 2.58e-06 [label_micro_interleaved_index]: 4.60001e-06 [label_fine_grained_interleaved_index]: 2.56998e-06 [merge_cast_opt]: 1.20001e-06 [slice_recompute_activation]: 1.96e-06 [micro_interleaved_order_control]: 2.35002e-06 [assign_add_opt]: 1.37999e-06 [ForceFp32Comm]: 1.19003e-06 [remove_cast_before_assign_add]: 1.15001e-06 [full_micro_interleaved_order_control]: 2.02999e-06 [reorder_send_recv_between_fp_bp]: 2.58003e-06 [comm_op_add_attrs]: 1.00001e-06 [add_comm_op_reuse_tag]: 9.29984e-07 [interleave_split_concat_branches]: 1.24e-06 [interleave_parallel_branches]: 1.00999e-06 [overlap_opt_shard_in_pipeline]: 1.60001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.03997e-06 [control_data_broadcast_order]: 1.722e-05 [grouped_pairwise_exchange_alltoall]: 1.45001e-06 [offloading_packed_experts]: 4.75001e-06 [overlap_recompute_and_grad_model_parallel]: 5.58002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.20001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.35999e-06 [overlap_recompute_comm]: 2.56e-06 [overlap_grad_ring_attention]: 4.57998e-06 [overlap_grad_flash_sp]: 2.447e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.17999e-06 [split_layernorm_comm]: 1.60001e-06 [handle_group_info]: 9.50007e-07 [symbol_engine_optimizer]: 8.219e-05, [1] [Cycle 1]: 7.74e-05, [6] [build]: 4.1e-06 [elim_shapecalc]: 1.13e-05 [elim_not_effective]: 1.438e-05 [opt_reshape]: 8.18999e-06 [fold_const_symbol]: 1.117e-05 [renormalize]: 2.00002e-07 [detach_backward]: 2.19001e-06 [pipeline_parallel_scheduler]: 1.84998e-06 [auto_monad_reorder]: 2.044e-05 [get_jit_bprop_graph]: 1.42e-06 [rewriter_after_jit_bprop_graph]: 4.43999e-06 [opt_after_jit_grad]: 0.00049828 [validate]: 4.038e-05 Sums bootstrap : 0.000435s : 0.43% type_inference : 0.081400s : 80.68% event_method : 0.000072s : 0.07% auto_monad : 0.000247s : 0.24% graph_reusing : 0.000020s : 0.02% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000071s : 0.07% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000020s : 0.02% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000096s : 0.09% insert-virtual-dataset : 0.000002s : 0.00% parallel-infer-symbol-second : 0.000002s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000007s : 0.01% optimize.rewriter_before_opt_a : 0.000550s : 0.55% optimize.opt_a.expand_dump_flag : 0.000016s : 0.02% optimize.opt_a.switch_simplify : 0.000349s : 0.35% optimize.opt_a.loop_unroll : 0.000157s : 0.16% optimize.opt_a.a_1 : 0.004468s : 4.43% optimize.opt_a.with_stream_mark : 0.000067s : 0.07% optimize.opt_a.recompute_prepare : 0.000046s : 0.05% optimize.opt_a.updatestate_depend_eliminate : 0.000018s : 0.02% optimize.opt_a.updatestate_assign_eliminate : 0.000015s : 0.02% optimize.opt_a.updatestate_loads_eliminate : 0.000014s : 0.01% optimize.opt_a.parameter_eliminate : 0.000007s : 0.01% optimize.opt_a.a_2 : 0.000443s : 0.44% optimize.opt_a.accelerated_algorithm : 0.000075s : 0.07% optimize.opt_a.shard : 0.000005s : 0.01% optimize.opt_a.meta_shard_fg_expand : 0.000012s : 0.01% optimize.opt_a.shard_inline : 0.000031s : 0.03% optimize.opt_a.merge_send_recv : 0.000033s : 0.03% optimize.opt_a.auto_parallel : 0.000031s : 0.03% optimize.opt_a.parallel : 0.000036s : 0.04% optimize.opt_a.flash_sp : 0.000018s : 0.02% optimize.opt_a.merge_comm : 0.000017s : 0.02% optimize.opt_a.allreduce_fusion : 0.000016s : 0.02% optimize.opt_a.matmul_add_comm_reduction : 0.000050s : 0.05% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000037s : 0.04% optimize.opt_a.virtual_dataset : 0.000028s : 0.03% optimize.opt_a.get_grad_eliminate_ : 0.000028s : 0.03% optimize.opt_a.virtual_output : 0.000029s : 0.03% optimize.opt_a.merge_forward : 0.000019s : 0.02% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000039s : 0.04% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000057s : 0.06% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000051s : 0.05% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000020s : 0.02% optimize.opt_a.meta_fg_expand : 0.002024s : 2.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.01% optimize.opt_a.receive_attached : 0.000015s : 0.01% optimize.opt_a.after_resolve : 0.000097s : 0.10% optimize.opt_a.a_after_grad : 0.000105s : 0.10% optimize.opt_a.renormalize : 0.006061s : 6.01% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.02% optimize.opt_a.auto_monad_grad : 0.000010s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000092s : 0.09% optimize.opt_a.cse : 0.000400s : 0.40% optimize.opt_a.a_3 : 0.000431s : 0.43% optimize.py_interpret_to_execute_after_opt_a : 0.000006s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000023s : 0.02% optimize.convert_after_rewriter : 0.000001s : 0.00% optimize.order_py_execute_after_rewriter : 0.000001s : 0.00% optimize.mutable_eliminate : 0.000738s : 0.73% optimize.opt_b.b_1 : 0.000133s : 0.13% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000007s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000003s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000028s : 0.03% optimize.optimize_parallel_all_gather_comm : 0.000019s : 0.02% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000029s : 0.03% optimize.loop_unroll : 0.000497s : 0.49% optimize.opt_after_cconv.c_1 : 0.000032s : 0.03% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000006s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000024s : 0.02% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.02% optimize.tuple_transform.d_1 : 0.000046s : 0.05% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000008s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000054s : 0.05% optimize.cse_after_recomputation.cse : 0.000015s : 0.02% optimize.environ_conv : 0.000006s : 0.01% optimize.swap_dp_allreduce_reducescatter : 0.000006s : 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 : 0.000002s : 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 : 0.000017s : 0.02% optimize.grouped_pairwise_exchange_alltoall : 0.000001s : 0.00% optimize.offloading_packed_experts : 0.000005s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 0.01% 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 : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000024s : 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 : 0.000004s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000011s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000014s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000008s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000011s : 0.01% 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 : 0.000020s : 0.02% get_jit_bprop_graph : 0.000001s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000498s : 0.49% validate : 0.000040s : 0.04% Time group info: ------[substitution.] 0.001341 199 8.76% : 0.000117s : 6: substitution.arithmetic_simplify 0.74% : 0.000010s : 1: substitution.cast_eliminate 0.19% : 0.000003s : 3: substitution.elim_not_effective 0.63% : 0.000009s : 11: substitution.float_depend_g_call 0.31% : 0.000004s : 3: substitution.float_tuple_getitem_switch 0.13% : 0.000002s : 3: substitution.fold_const_symbol 0.53% : 0.000007s : 4: substitution.graph_param_transform 0.24% : 0.000003s : 2: substitution.incorporate_call 0.15% : 0.000002s : 2: substitution.incorporate_call_switch 67.39% : 0.000903s : 36: substitution.inline 1.40% : 0.000019s : 2: substitution.inline_without_move 0.81% : 0.000011s : 16: substitution.j_node_and_user_rematch 1.72% : 0.000023s : 3: substitution.less_batch_normalization 0.87% : 0.000012s : 7: substitution.minmaximum_grad 0.70% : 0.000009s : 11: substitution.partial_eliminate 1.01% : 0.000014s : 16: substitution.remove_not_recompute_node 2.11% : 0.000028s : 9: substitution.replace_applicator 0.78% : 0.000010s : 8: substitution.replace_old_param 0.24% : 0.000003s : 1: substitution.set_cell_output_no_recompute 1.56% : 0.000021s : 8: substitution.switch_simplify 1.88% : 0.000025s : 7: substitution.tuple_list_convert_item_index_to_positive 0.76% : 0.000010s : 7: substitution.tuple_list_get_item_const_eliminator 1.12% : 0.000015s : 7: substitution.tuple_list_get_item_depend_reorder 4.83% : 0.000065s : 19: substitution.tuple_list_get_item_eliminator 1.13% : 0.000015s : 7: substitution.tuple_list_get_set_item_eliminator ------[type_inference.] 0.081221 2 92.62% : 0.075225s : 1: type_inference.infer 7.38% : 0.005996s : 1: type_inference.specialize ------[replace.] 0.000440 54 1.41% : 0.000006s : 1: replace.cast_eliminate 59.71% : 0.000263s : 36: replace.inline 21.87% : 0.000096s : 8: replace.switch_simplify 17.01% : 0.000075s : 9: replace.tuple_list_get_item_eliminator ------[match.] 0.000933 54 0.98% : 0.000009s : 1: match.cast_eliminate 94.40% : 0.000881s : 36: match.inline 1.75% : 0.000016s : 8: match.switch_simplify 2.88% : 0.000027s : 9: match.tuple_list_get_item_eliminator ------[predicate.] 0.000825 5282 1.15% : 0.000010s : 69: predicate.accumulaten_eliminater 0.25% : 0.000002s : 4: predicate.ad_related_special_op_eliminate 0.35% : 0.000003s : 21: predicate.addn_check_dump 1.19% : 0.000010s : 69: predicate.addn_zero_filter 1.10% : 0.000009s : 69: predicate.adjust_all_reduce_mul_add 2.08% : 0.000017s : 90: predicate.arithmetic_simplify 1.36% : 0.000011s : 70: predicate.cast_eliminate 0.90% : 0.000007s : 51: predicate.check_bprop_eliminate 0.36% : 0.000003s : 21: predicate.compare_switch_simplify 0.04% : 0.000000s : 4: predicate.const_output_eliminate 0.42% : 0.000003s : 21: predicate.depend_value_elim 1.23% : 0.000010s : 70: predicate.dict_get_item_const_eliminator 1.37% : 0.000011s : 70: predicate.dict_get_item_eliminator 1.27% : 0.000010s : 70: predicate.dict_set_item_eliminator 0.22% : 0.000002s : 8: predicate.dumpgradient_eliminate 0.05% : 0.000000s : 4: predicate.elim_not_effective 0.11% : 0.000001s : 4: predicate.elim_shapecalc_of_broadcastargs 1.24% : 0.000010s : 74: predicate.environ_add_const_eliminate 1.28% : 0.000011s : 74: predicate.environ_get_add_eliminate 1.18% : 0.000010s : 74: predicate.environ_get_depend_swap 1.61% : 0.000013s : 95: predicate.environ_get_eliminate 1.20% : 0.000010s : 74: predicate.environ_get_set_eliminate 2.01% : 0.000017s : 115: predicate.exchange_switch_depend_value 2.81% : 0.000023s : 115: predicate.float_depend_g_call 0.36% : 0.000003s : 21: predicate.float_environ_get_switch 0.47% : 0.000004s : 25: predicate.float_tuple_getitem_switch 0.04% : 0.000000s : 4: predicate.fold_const_symbol 0.45% : 0.000004s : 21: predicate.get_grad_eliminate 0.06% : 0.000000s : 4: predicate.graph_param_transform 0.38% : 0.000003s : 21: predicate.incorporate_call 0.32% : 0.000003s : 21: predicate.incorporate_call_switch 5.77% : 0.000048s : 234: predicate.inline 1.08% : 0.000009s : 43: predicate.inline_without_move 0.19% : 0.000002s : 21: predicate.j_node_and_user_rematch 0.54% : 0.000004s : 21: predicate.less_batch_normalization 1.79% : 0.000015s : 87: predicate.list_to_tuple_eliminator_ 2.48% : 0.000020s : 156: predicate.load_eliminater 0.31% : 0.000003s : 4: predicate.loop_unroll_after_grad 3.14% : 0.000026s : 179: predicate.loop_unroll_before_grad 1.42% : 0.000012s : 78: predicate.make_slice_get_slice_eliminator 0.38% : 0.000003s : 21: predicate.merge_addn 0.84% : 0.000007s : 51: predicate.micro_step_allgather_replace 0.82% : 0.000007s : 51: predicate.mini_step_allgather_replace 1.19% : 0.000010s : 69: predicate.minmaximum_grad 0.27% : 0.000002s : 4: predicate.mutable_eliminate 0.10% : 0.000001s : 4: predicate.opt_reshape 0.10% : 0.000001s : 4: predicate.parallel_virtual_node 2.78% : 0.000023s : 115: predicate.partial_defer_inline 1.55% : 0.000013s : 83: predicate.partial_eliminate 1.15% : 0.000009s : 69: predicate.print_const_string_wrapper 0.40% : 0.000003s : 21: predicate.reduce_all_const_elim 1.59% : 0.000013s : 69: predicate.reduce_eliminate 2.57% : 0.000021s : 156: predicate.redundant_stop_gradient_eliminater 0.23% : 0.000002s : 21: predicate.remove_not_recompute_node 1.58% : 0.000013s : 130: predicate.replace_applicator 0.53% : 0.000004s : 43: predicate.replace_old_param 0.07% : 0.000001s : 4: predicate.reset_defer_inline 1.21% : 0.000010s : 69: predicate.reshape_eliminate 0.87% : 0.000007s : 51: predicate.row_tensor_add_zeros_like 0.13% : 0.000001s : 4: predicate.row_tensor_eliminate 1.16% : 0.000010s : 51: predicate.same_eliminate 0.27% : 0.000002s : 21: predicate.set_cell_output_no_recompute 0.50% : 0.000004s : 21: predicate.shard_identity_eliminate 0.21% : 0.000002s : 8: predicate.special_op_eliminate 0.47% : 0.000004s : 21: predicate.specialize_transform 1.01% : 0.000008s : 51: predicate.split_environ_get_set_with_tuple_value 0.97% : 0.000008s : 43: predicate.stack_unstack_eliminate 0.08% : 0.000001s : 4: predicate.switch_call_monad_eliminater 2.24% : 0.000018s : 115: predicate.switch_defer_inline 3.07% : 0.000025s : 166: predicate.switch_layer_defer_inline 6.86% : 0.000057s : 335: predicate.switch_simplify 1.15% : 0.000009s : 69: predicate.tile_eliminate 1.32% : 0.000011s : 69: predicate.transpose_eliminate 1.53% : 0.000013s : 78: predicate.tuple_list_convert_item_index_to_positive 1.54% : 0.000013s : 78: predicate.tuple_list_get_item_const_eliminator 1.39% : 0.000011s : 78: predicate.tuple_list_get_item_depend_reorder 2.53% : 0.000021s : 108: predicate.tuple_list_get_item_eliminator 1.50% : 0.000012s : 78: predicate.tuple_list_get_set_item_eliminator 2.09% : 0.000017s : 99: predicate.tuple_list_set_item_eliminator 1.46% : 0.000012s : 87: predicate.tuple_to_list_eliminator_ 2.47% : 0.000020s : 156: predicate.updatestate_pure_node_eliminater 3.14% : 0.000026s : 177: predicate.updatestate_useless_node_eliminater 0.10% : 0.000001s : 4: predicate.value_based_eliminate 0.39% : 0.000003s : 21: predicate.virtual_dataset_eliminate 0.40% : 0.000003s : 21: predicate.virtual_output_eliminate 0.07% : 0.000001s : 4: predicate.virtual_view_grad_eliminate 0.14% : 0.000001s : 4: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003717 61 57.23% : 0.002127s : 21: func_graph_cloner_run.FuncGraphClonerGraph 42.77% : 0.001590s : 40: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.141273 233 0.00% : 0.000004s : 1: ForceFp32Comm 2.59% : 0.003665s : 1: add_attr 2.59% : 0.003653s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.04% : 0.000059s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.18% : 0.000257s : 1: auto_monad 0.02% : 0.000024s : 1: auto_monad_reorder 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.34% : 0.000486s : 1: bootstrap 0.02% : 0.000033s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.02% : 0.000022s : 1: control_data_broadcast_order 0.00% : 0.000004s : 1: convert_after_rewriter 0.02% : 0.000029s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000005s : 1: detach_backward 0.01% : 0.000009s : 1: environ_conv 0.06% : 0.000079s : 1: event_method 0.00% : 0.000007s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.02% : 0.000024s : 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% : 0.000006s : 1: insert-virtual-dataset 0.01% : 0.000021s : 1: interleave_parallel_branches 0.00% : 0.000005s : 1: interleave_split_concat_branches 0.00% : 0.000005s : 1: label_fine_grained_interleaved_index 0.01% : 0.000007s : 1: label_micro_interleaved_index 0.36% : 0.000507s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.53% : 0.000751s : 1: mutable_eliminate 0.01% : 0.000008s : 1: offloading_packed_experts 0.01% : 0.000016s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000018s : 1: opt.transform.mutable_eliminate 4.44% : 0.006273s : 117: opt.transform.opt_a 0.02% : 0.000031s : 1: opt.transform.opt_after_cconv 0.02% : 0.000028s : 1: opt.transform.opt_after_jit_grad 0.08% : 0.000112s : 28: opt.transform.opt_b 0.04% : 0.000052s : 2: opt.transform.opt_trans_graph 0.03% : 0.000041s : 4: opt.transform.symbol_engine_opt 11.59% : 0.016374s : 1: opt_a 0.08% : 0.000116s : 1: opt_after_cconv 0.36% : 0.000509s : 1: opt_after_jit_grad 0.16% : 0.000231s : 1: opt_b 13.61% : 0.019222s : 1: optimize 0.02% : 0.000022s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000004s : 1: order_py_execute_after_rewriter 0.02% : 0.000029s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.01% : 0.000008s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000005s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000005s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.01% : 0.000009s : 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% : 0.000005s : 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% : 0.000101s : 1: pre_auto_parallel 0.01% : 0.000011s : 1: py_interpret_to_execute 0.01% : 0.000010s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000020s : 1: remove_dup_value 2.73% : 0.003857s : 2: renormalize.infer 1.54% : 0.002182s : 2: renormalize.specialize 0.00% : 0.000005s : 1: reorder_send_recv_between_fp_bp 0.01% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000026s : 1: rewriter_after_opt_a 0.40% : 0.000559s : 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% : 0.000009s : 1: swap_dp_allreduce_reducescatter 0.06% : 0.000085s : 1: symbol_engine_optimizer 0.06% : 0.000083s : 1: tuple_transform 57.64% : 0.081424s : 1: type_inference skip Bool... Tensor.copy (100, 100) Int8 ... Tensor.copy (100, 100) Int16 ... Tensor.copy (100, 100) Int32 ... Tensor.copy (100, 100) Int64 ... Tensor.copy (100, 100) UInt8 ... Tensor.copy (100, 100) UInt16 ... Tensor.copy (100, 100) UInt32 ... Tensor.copy (100, 100) UInt64 ... Tensor.copy (100, 100) Float16 ... Tensor.copy (100, 100) Float32 ... Tensor.copy (100, 100) Float64 ... skip BFloat16... Tensor.copy (100, 100) Complex64 ... Tensor.copy (100, 100) Complex128 ... skip Int4... skip Float8E4M3FN... skip Float8E5M2... skip HiFloat8... . [hook] pytest_runtest_teardown:test_tensor_copy_limits[ge] tests/st/tensor/test_copy.py::test_tensor_copy_limits[ge],max_mem:6.0M =============================== 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_copy.py::test_tensor_copy_limits[pynative] /usr/local/Ascend/cann-8.5.0/python/site-packages/asc_op_compile_base/asc_op_compiler/ascendc_compile_gen_code.py:161: DeprecationWarning: invalid escape sequence \w match = re.search(f'{option}=(\w+)', ' '.join(compile_options)) -- Docs: https://docs.pytest.org/en/stable/warnings.html ================== 3 passed, 26 warnings in 528.84s (0:08:48) ==================