==================================================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_008/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 1 item test_sigmoid_.py ************************************************************************************************** *************************************TEST_OP_GENERALIZE Start************************************* ************************************************************************************************** *****************Start Generalize test PYNATIVE_MODE DiscontiguousInput case***************** [INFO]: TEST_OP_GENERALIZE generate discontiguous input with shape (2, 3), is_contiguous False, origin input with shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test discontiguous input [INFO]: TEST_OP_GENERALIZE Start to test contiguous input [INFO]: TEST_OP_GENERALIZE Start to compare discontiguous out with contiguous out Compare ['Success'] for 0'th output. *****************End Generalize test PYNATIVE_MODE DiscontiguousInput case***************** *****************Start Generalize test PYNATIVE_MODE ViewTensor case***************** [INFO]: TEST_OP_GENERALIZE generate view input with shape (2, 3), origin input with shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test no view input [INFO]: TEST_OP_GENERALIZE Start to test view input [INFO]: TEST_OP_GENERALIZE Start to compare view out with no view out Compare ['Success'] for 0'th output. *****************End Generalize test PYNATIVE_MODE ViewTensor case***************** *****************Start Generalize test PYNATIVE_MODE EmptyTensor case***************** [INFO]: TEST_OP_GENERALIZE EmptyTensor random seed 4060807799. [INFO]: TEST_OP_GENERALIZE EmptyTensor test random empty shape, loop 0 [INFO]: TEST_OP_GENERALIZE EmptyTensor generate random shape with shape (0, 0), origin shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test empty tensor [INFO]: TEST_OP_GENERALIZE Start to compare empty out with expect empty out [INFO]: TEST_OP_GENERALIZE EmptyTensor expect output shape (0,), actual output shape (0, 0) Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE EmptyTensor test random empty shape, loop 1 [INFO]: TEST_OP_GENERALIZE EmptyTensor generate random shape with shape (2, 0), origin shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test empty tensor [INFO]: TEST_OP_GENERALIZE Start to compare empty out with expect empty out [INFO]: TEST_OP_GENERALIZE EmptyTensor expect output shape (0,), actual output shape (2, 0) Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE EmptyTensor test random empty shape, loop 2 [INFO]: TEST_OP_GENERALIZE EmptyTensor generate random shape with shape (0, 3), origin shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test empty tensor [INFO]: TEST_OP_GENERALIZE Start to compare empty out with expect empty out [INFO]: TEST_OP_GENERALIZE EmptyTensor expect output shape (0,), actual output shape (0, 3) Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE EmptyTensor test random empty shape, loop 3 [INFO]: TEST_OP_GENERALIZE EmptyTensor generate random shape with shape (0, 0), origin shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test empty tensor [INFO]: TEST_OP_GENERALIZE Start to compare empty out with expect empty out [INFO]: TEST_OP_GENERALIZE EmptyTensor expect output shape (0,), actual output shape (0, 0) Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE EmptyTensor test random empty shape, loop 4 [INFO]: TEST_OP_GENERALIZE EmptyTensor generate random shape with shape (0, 0), origin shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test empty tensor [INFO]: TEST_OP_GENERALIZE Start to compare empty out with expect empty out [INFO]: TEST_OP_GENERALIZE EmptyTensor expect output shape (0,), actual output shape (0, 0) Compare ['Success'] for 0'th output. *****************End Generalize test PYNATIVE_MODE EmptyTensor case***************** *****************Start Generalize test PYNATIVE_MODE ScalarTensor case***************** [INFO]: TEST_OP_GENERALIZE Start to test scalar tensor [INFO]: TEST_OP_GENERALIZE Start to test 1d tensor [INFO]: TEST_OP_GENERALIZE Start to compare scalar tensor out with 1d tensor out Compare ['Success'] for 0'th output. *****************End Generalize test PYNATIVE_MODE ScalarTensor case***************** *****************Start Generalize test PYNATIVE_MODE GradByRequirement case***************** [INFO]: TEST_OP_GENERALIZE Start to test all grads [INFO]: TEST_OP_GENERALIZE Start to test grad by requirement on position 0 [INFO]: TEST_OP_GENERALIZE Start to compare grad by requirement with all grads Compare ['Success'] for 0'th output. *****************End Generalize test PYNATIVE_MODE GradByRequirement case***************** *****************Start Generalize test PYNATIVE_MODE Deterministic case***************** [INFO]: TEST_OP_GENERALIZE test deterministic case loop 0 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 1 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 2 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 3 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 4 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out TotalTime = 39.4859, [24] [bootstrap]: 1.02782 [type_inference]: 2.63161 [event_method]: 2.426e-05 [auto_monad]: 0.00086836 [graph_reusing]: 9.86e-06 [inline]: 3.63e-06 [add_attr]: 1.69636, [1] [add_attr_with_inline]: 1.69634, [1] [Cycle 1]: 0.00132443, [2] [tag_attr]: 2.858e-05 [meta_addattr_fg_expand]: 6.84999e-06 [parallel-infer-symbol]: 4.42e-06 [pre_auto_parallel]: 0.0001099 [insert-virtual-dataset]: 3.16001e-06 [parallel-infer-symbol-second]: 7.49977e-07 [dataset_repeat_opt]: 2.01e-06 [pipeline_split]: 1.76998e-06 [optimize]: 0.0135771, [53] [py_interpret_to_execute]: 4.304e-05 [rewriter_before_opt_a]: 0.00046681 [opt_a]: 0.00772286, [2] [Cycle 1]: 0.00658772, [45] [expand_dump_flag]: 3.86001e-06 [switch_simplify]: 4.842e-05 [loop_unroll]: 3.253e-05 [a_1]: 0.00071791 [with_stream_mark]: 1.983e-05 [recompute_prepare]: 1.099e-05 [updatestate_depend_eliminate]: 6.74999e-06 [updatestate_assign_eliminate]: 0.00016084 [updatestate_loads_eliminate]: 5.72999e-06 [parameter_eliminate]: 2.04e-06 [a_2]: 0.00012815 [accelerated_algorithm]: 1.093e-05 [shard]: 2.24999e-06 [meta_shard_fg_expand]: 2.51e-06 [shard_inline]: 8.91002e-06 [merge_send_recv]: 1.231e-05 [auto_parallel]: 9.49e-06 [parallel]: 0.00104668 [flash_sp]: 1.278e-05 [merge_comm]: 6.62002e-06 [allreduce_fusion]: 5.55001e-06 [matmul_add_comm_reduction]: 1.236e-05 [allreduce_slice_to_reducescatter]: 8.30012e-07 [virtual_shard_identity]: 1.388e-05 [virtual_dataset]: 9.09e-06 [get_grad_eliminate_]: 2.843e-05 [virtual_output]: 8.75001e-06 [merge_forward]: 7.18998e-06 [cell_reuse_recompute_pass]: 1.42999e-06 [offload_activation]: 1.524e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.93e-05 [merge_recompute_call_nodes]: 1.63002e-06 [before_grad]: 1.711e-05 [set_forward_comm_id_for_comm_node_pass]: 5.64998e-06 [meta_fg_expand]: 4.38999e-06 [flash_sp_send_recv_attached]: 3.04999e-06 [receive_attached]: 3.864e-05 [after_resolve]: 1.316e-05 [a_after_grad]: 1.4e-05 [renormalize]: 0.00264635 [add_forward_monad_depend]: 7.73999e-06 [auto_monad_grad]: 2.60997e-06 [auto_monad_eliminator]: 0.00065004 [cse]: 0.00011316 [a_3]: 7.586e-05 [Cycle 2]: 0.00112106, [45] [expand_dump_flag]: 2.27999e-06 [switch_simplify]: 1.2e-05 [loop_unroll]: 0.00013495 [a_1]: 0.00021452 [with_stream_mark]: 2.25e-05 [recompute_prepare]: 9.17001e-06 [updatestate_depend_eliminate]: 6.30002e-06 [updatestate_assign_eliminate]: 5.05999e-06 [updatestate_loads_eliminate]: 4.54002e-06 [parameter_eliminate]: 1.52001e-06 [a_2]: 0.00011199 [accelerated_algorithm]: 8.89e-06 [shard]: 3.01999e-06 [meta_shard_fg_expand]: 3.52997e-06 [shard_inline]: 8.75999e-06 [merge_send_recv]: 9.72999e-06 [auto_parallel]: 8.59998e-06 [parallel]: 8.85001e-06 [flash_sp]: 4.72e-06 [merge_comm]: 5.46e-06 [allreduce_fusion]: 5.26002e-06 [matmul_add_comm_reduction]: 9.15999e-06 [allreduce_slice_to_reducescatter]: 4.69998e-07 [virtual_shard_identity]: 1.037e-05 [virtual_dataset]: 8.54e-06 [get_grad_eliminate_]: 7.88001e-06 [virtual_output]: 7.98999e-06 [merge_forward]: 5.40001e-06 [cell_reuse_recompute_pass]: 3.73999e-06 [offload_activation]: 1.289e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.928e-05 [merge_recompute_call_nodes]: 9.70002e-07 [before_grad]: 1.639e-05 [set_forward_comm_id_for_comm_node_pass]: 6.09999e-06 [meta_fg_expand]: 3.83001e-06 [flash_sp_send_recv_attached]: 1.32999e-06 [receive_attached]: 2.43998e-06 [after_resolve]: 1.351e-05 [a_after_grad]: 1.338e-05 [renormalize]: 1.30007e-07 [add_forward_monad_depend]: 1.72001e-06 [auto_monad_grad]: 2.41e-06 [auto_monad_eliminator]: 1.419e-05 [cse]: 2.445e-05 [a_3]: 5.134e-05 [py_interpret_to_execute_after_opt_a]: 1.72e-05 [slice_cell_reuse_recomputed_activation]: 2.22001e-06 [rewriter_after_opt_a]: 0.00080139 [convert_after_rewriter]: 1.221e-05 [order_py_execute_after_rewriter]: 7.2e-06 [mutable_eliminate]: 0.00112738 [opt_b]: 0.00038611, [1] [Cycle 1]: 0.00037802, [7] [b_1]: 0.00018916 [b_2]: 1.096e-05 [updatestate_depend_eliminate]: 9.97001e-06 [updatestate_assign_eliminate]: 4.52e-06 [updatestate_loads_eliminate]: 4.43999e-06 [renormalize]: 5.00004e-07 [cse]: 0.00011745 [optimize_parallel_all_gather_comm]: 2.316e-05 [overlap_param_gather]: 2.272e-05 [cconv]: 2.876e-05 [loop_unroll]: 0.0004841 [opt_after_cconv]: 0.00014532, [1] [Cycle 1]: 0.00013936, [7] [c_1]: 5.75e-05 [parameter_eliminate]: 2.22999e-06 [updatestate_depend_eliminate]: 8.27e-06 [updatestate_assign_eliminate]: 4.84003e-06 [updatestate_loads_eliminate]: 4.00998e-06 [cse]: 2.606e-05 [renormalize]: 2.10013e-07 [remove_dup_value]: 1.631e-05 [tuple_transform]: 0.00013711, [1] [Cycle 1]: 0.00013166, [4] [d_1]: 0.00010165 [none_parameter_eliminate]: 1.67999e-06 [renormalize]: 2.00002e-07 [switch_simplify]: 9.49e-06 [partial_unused_args_eliminate]: 1.86e-06 [add_recomputation]: 0.00018181 [cse_after_recomputation]: 3.353e-05, [1] [Cycle 1]: 2.806e-05, [1] [cse]: 2.127e-05 [environ_conv]: 0.00015792 [swap_dp_allreduce_reducescatter]: 9.24e-06 [bias_add_comm_swap]: 4.51002e-06 [label_micro_interleaved_index]: 4.38999e-06 [label_fine_grained_interleaved_index]: 2.73998e-06 [merge_cast_opt]: 1.59998e-06 [slice_recompute_activation]: 2.71e-06 [micro_interleaved_order_control]: 2.51998e-06 [assign_add_opt]: 1.36998e-06 [ForceFp32Comm]: 8.50006e-07 [remove_cast_before_assign_add]: 1.25999e-06 [full_micro_interleaved_order_control]: 2.29999e-06 [reorder_send_recv_between_fp_bp]: 2.77002e-06 [comm_op_add_attrs]: 1.39998e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.14e-06 [interleave_parallel_branches]: 1.07998e-06 [overlap_opt_shard_in_pipeline]: 8.801e-05 [overlap_opt_shard_grad_in_pipeline]: 1.80001e-06 [control_data_broadcast_order]: 1.787e-05 [grouped_pairwise_exchange_alltoall]: 1.66e-06 [offloading_packed_experts]: 5.30001e-06 [overlap_recompute_and_grad_model_parallel]: 6.05002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.20001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.75001e-06 [overlap_recompute_comm]: 2.48e-06 [overlap_grad_ring_attention]: 5.49e-06 [overlap_grad_flash_sp]: 0.00017294 [begin_end_overlap_inline]: 6.39993e-07 [split_matmul_comm_elemetwise]: 2.34999e-06 [split_layernorm_comm]: 2.26998e-06 [handle_group_info]: 1.32999e-06 [symbol_engine_optimizer]: 0.00010051, [1] [Cycle 1]: 9.557e-05, [6] [build]: 5.47001e-06 [elim_shapecalc]: 1.456e-05 [elim_not_effective]: 1.955e-05 [opt_reshape]: 9.49999e-06 [fold_const_symbol]: 1.494e-05 [renormalize]: 2.19996e-07 [detach_backward]: 2.34001e-06 [pipeline_parallel_scheduler]: 1.55001e-06 [auto_monad_reorder]: 3.39e-05 [get_jit_bprop_graph]: 1.79998e-06 [rewriter_after_jit_bprop_graph]: 0.00049626 [opt_after_jit_grad]: 0.00062885 [validate]: 0.00068276 [backend_pass]: 1.12999e-06 [task_emit]: 34.1133 [execute]: 1.065e-05 Sums bootstrap : 1.027822s : 2.72% type_inference : 2.631614s : 6.96% event_method : 0.000024s : 0.00% auto_monad : 0.000868s : 0.00% graph_reusing : 0.000010s : 0.00% inline : 0.000004s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000029s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000007s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000110s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000043s : 0.00% optimize.rewriter_before_opt_a : 0.000467s : 0.00% optimize.opt_a.expand_dump_flag : 0.000006s : 0.00% optimize.opt_a.switch_simplify : 0.000060s : 0.00% optimize.opt_a.loop_unroll : 0.000167s : 0.00% optimize.opt_a.a_1 : 0.000932s : 0.00% optimize.opt_a.with_stream_mark : 0.000042s : 0.00% optimize.opt_a.recompute_prepare : 0.000020s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000013s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000166s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000240s : 0.00% optimize.opt_a.accelerated_algorithm : 0.000020s : 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 : 0.000018s : 0.00% optimize.opt_a.merge_send_recv : 0.000022s : 0.00% optimize.opt_a.auto_parallel : 0.000018s : 0.00% optimize.opt_a.parallel : 0.001056s : 0.00% 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 : 0.000022s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000024s : 0.00% optimize.opt_a.virtual_dataset : 0.000018s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000036s : 0.00% optimize.opt_a.virtual_output : 0.000017s : 0.00% optimize.opt_a.merge_forward : 0.000013s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000028s : 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 : 0.000034s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.00% optimize.opt_a.meta_fg_expand : 0.000008s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.00% optimize.opt_a.receive_attached : 0.000041s : 0.00% optimize.opt_a.after_resolve : 0.000027s : 0.00% optimize.opt_a.a_after_grad : 0.000027s : 0.00% optimize.opt_a.renormalize : 0.002646s : 0.01% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.00% optimize.opt_a.auto_monad_grad : 0.000005s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000664s : 0.00% optimize.opt_a.cse : 0.000138s : 0.00% optimize.opt_a.a_3 : 0.000127s : 0.00% optimize.py_interpret_to_execute_after_opt_a : 0.000017s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000801s : 0.00% optimize.convert_after_rewriter : 0.000012s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.001127s : 0.00% optimize.opt_b.b_1 : 0.000189s : 0.00% optimize.opt_b.b_2 : 0.000011s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.00% 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 : 0.000117s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000023s : 0.00% optimize.overlap_param_gather : 0.000023s : 0.00% optimize.cconv : 0.000029s : 0.00% optimize.loop_unroll : 0.000484s : 0.00% optimize.opt_after_cconv.c_1 : 0.000058s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000026s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.00% optimize.tuple_transform.d_1 : 0.000102s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000182s : 0.00% optimize.cse_after_recomputation.cse : 0.000021s : 0.00% optimize.environ_conv : 0.000158s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 0.00% optimize.bias_add_comm_swap : 0.000005s : 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 : 0.000003s : 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 : 0.000002s : 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 : 0.000088s : 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 : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000173s : 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 : 0.000020s : 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 : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000034s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000496s : 0.00% opt_after_jit_grad : 0.000629s : 0.00% validate : 0.000683s : 0.00% backend_pass : 0.000001s : 0.00% task_emit : 34.113301s : 90.28% execute : 0.000011s : 0.00% Time group info: ------[substitution.] 0.000341 60 2.69% : 0.000009s : 2: substitution.depend_value_elim 0.94% : 0.000003s : 5: substitution.elim_not_effective 0.72% : 0.000002s : 5: substitution.fold_const_symbol 15.52% : 0.000053s : 6: substitution.graph_param_transform 59.69% : 0.000203s : 4: substitution.inline 2.68% : 0.000009s : 10: substitution.j_node_and_user_rematch 2.67% : 0.000009s : 10: substitution.remove_not_recompute_node 1.81% : 0.000006s : 2: substitution.replace_old_param 4.05% : 0.000014s : 2: substitution.tuple_list_get_item_eliminator 4.72% : 0.000016s : 6: substitution.updatestate_pure_node_eliminater 4.50% : 0.000015s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 2.631536 2 99.97% : 2.630718s : 1: type_inference.infer 0.03% : 0.000818s : 1: type_inference.specialize ------[replace.] 0.000059 6 70.39% : 0.000042s : 4: replace.inline 29.61% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000214 6 94.10% : 0.000201s : 4: match.inline 5.90% : 0.000013s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000290 1652 0.97% : 0.000003s : 16: predicate.accumulaten_eliminater 0.85% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.61% : 0.000002s : 12: predicate.addn_check_dump 1.12% : 0.000003s : 16: predicate.addn_zero_filter 0.82% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.30% : 0.000007s : 28: predicate.arithmetic_simplify 0.85% : 0.000002s : 16: predicate.cast_eliminate 0.80% : 0.000002s : 12: predicate.check_bprop_eliminate 0.62% : 0.000002s : 12: predicate.compare_switch_simplify 0.18% : 0.000001s : 6: predicate.const_output_eliminate 0.71% : 0.000002s : 12: predicate.depend_value_elim 0.84% : 0.000002s : 16: predicate.dict_get_item_const_eliminator 0.90% : 0.000003s : 16: predicate.dict_get_item_eliminator 0.87% : 0.000003s : 16: predicate.dict_set_item_eliminator 1.06% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.29% : 0.000001s : 6: predicate.elim_not_effective 0.44% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.15% : 0.000003s : 22: predicate.environ_add_const_eliminate 1.02% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.08% : 0.000003s : 22: predicate.environ_get_depend_swap 1.90% : 0.000006s : 34: predicate.environ_get_eliminate 1.13% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.12% : 0.000003s : 22: predicate.exchange_switch_depend_value 1.92% : 0.000006s : 22: predicate.float_depend_g_call 0.63% : 0.000002s : 12: predicate.float_environ_get_switch 1.01% : 0.000003s : 18: predicate.float_tuple_getitem_switch 0.19% : 0.000001s : 6: predicate.fold_const_symbol 0.77% : 0.000002s : 12: predicate.get_grad_eliminate 0.29% : 0.000001s : 6: predicate.graph_param_transform 0.60% : 0.000002s : 12: predicate.incorporate_call 0.51% : 0.000001s : 12: predicate.incorporate_call_switch 5.15% : 0.000015s : 74: predicate.inline 0.76% : 0.000002s : 12: predicate.inline_without_move 0.30% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.98% : 0.000003s : 12: predicate.less_batch_normalization 1.69% : 0.000005s : 30: predicate.list_to_tuple_eliminator_ 2.27% : 0.000007s : 46: predicate.load_eliminater 0.79% : 0.000002s : 6: predicate.loop_unroll_after_grad 2.35% : 0.000007s : 38: predicate.loop_unroll_before_grad 1.82% : 0.000005s : 28: predicate.make_slice_get_slice_eliminator 0.70% : 0.000002s : 12: predicate.merge_addn 0.81% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.61% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.75% : 0.000002s : 16: predicate.minmaximum_grad 0.87% : 0.000003s : 6: predicate.mutable_eliminate 0.38% : 0.000001s : 6: predicate.opt_reshape 0.69% : 0.000002s : 6: predicate.parallel_virtual_node 1.45% : 0.000004s : 22: predicate.partial_defer_inline 1.33% : 0.000004s : 24: predicate.partial_eliminate 0.92% : 0.000003s : 16: predicate.print_const_string_wrapper 0.72% : 0.000002s : 12: predicate.reduce_all_const_elim 1.36% : 0.000004s : 16: predicate.reduce_eliminate 2.54% : 0.000007s : 46: predicate.redundant_stop_gradient_eliminater 0.55% : 0.000002s : 12: predicate.remove_not_recompute_node 1.13% : 0.000003s : 30: predicate.replace_applicator 0.47% : 0.000001s : 12: predicate.replace_old_param 0.23% : 0.000001s : 6: predicate.reset_defer_inline 0.96% : 0.000003s : 16: predicate.reshape_eliminate 0.73% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.52% : 0.000002s : 6: predicate.row_tensor_eliminate 0.84% : 0.000002s : 12: predicate.same_eliminate 0.39% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.89% : 0.000003s : 12: predicate.shard_identity_eliminate 0.78% : 0.000002s : 12: predicate.special_op_eliminate 0.82% : 0.000002s : 12: predicate.specialize_transform 0.95% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 1.33% : 0.000004s : 12: predicate.stack_unstack_eliminate 0.36% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.46% : 0.000004s : 22: predicate.switch_defer_inline 1.81% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.54% : 0.000013s : 78: predicate.switch_simplify 1.18% : 0.000003s : 16: predicate.tile_eliminate 0.90% : 0.000003s : 16: predicate.transpose_eliminate 1.66% : 0.000005s : 28: predicate.tuple_list_convert_item_index_to_positive 1.72% : 0.000005s : 28: predicate.tuple_list_get_item_const_eliminator 1.47% : 0.000004s : 28: predicate.tuple_list_get_item_depend_reorder 3.32% : 0.000010s : 42: predicate.tuple_list_get_item_eliminator 1.49% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.32% : 0.000007s : 40: predicate.tuple_list_set_item_eliminator 1.64% : 0.000005s : 30: predicate.tuple_to_list_eliminator_ 2.37% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.36% : 0.000010s : 58: predicate.updatestate_useless_node_eliminater 0.57% : 0.000002s : 6: predicate.value_based_eliminate 0.76% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.69% : 0.000002s : 12: predicate.virtual_output_eliminate 0.51% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.48% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000816 11 59.99% : 0.000489s : 5: func_graph_cloner_run.FuncGraphClonerGraph 40.01% : 0.000326s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 41.199520 196 0.00% : 0.000004s : 1: ForceFp32Comm 4.12% : 1.696364s : 1: add_attr 4.12% : 1.696344s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000188s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.00% : 0.000885s : 1: auto_monad 0.00% : 0.000038s : 1: auto_monad_reorder 0.00% : 0.000007s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 2.49% : 1.027891s : 1: bootstrap 0.00% : 0.000032s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000016s : 1: convert_after_rewriter 0.00% : 0.000037s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000162s : 1: environ_conv 0.00% : 0.000031s : 1: event_method 0.00% : 0.000018s : 1: execute 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% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 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% : 0.000007s : 1: label_micro_interleaved_index 0.00% : 0.000569s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.00% : 0.001139s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 0.00% : 0.001702s : 78: opt.transform.opt_a 0.00% : 0.000055s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000173s : 28: opt.transform.opt_b 0.00% : 0.000109s : 2: opt.transform.opt_trans_graph 0.00% : 0.000054s : 4: opt.transform.symbol_engine_opt 0.02% : 0.007727s : 1: opt_a 0.00% : 0.000149s : 1: opt_after_cconv 0.00% : 0.000639s : 1: opt_after_jit_grad 0.00% : 0.000391s : 1: opt_b 0.03% : 0.013582s : 1: optimize 0.00% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000011s : 1: order_py_execute_after_rewriter 0.00% : 0.000179s : 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% : 0.000091s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000026s : 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% : 0.000009s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.00% : 0.000115s : 1: pre_auto_parallel 0.00% : 0.000048s : 1: py_interpret_to_execute 0.00% : 0.000021s : 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 0.01% : 0.002207s : 1: renormalize.infer 0.00% : 0.000430s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000503s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000810s : 1: rewriter_after_opt_a 0.00% : 0.000471s : 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% : 0.000007s : 1: split_matmul_comm_elemetwise 0.00% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000103s : 1: symbol_engine_optimizer 82.80% : 34.113331s : 1: task_emit 0.00% : 0.000140s : 1: tuple_transform 6.39% : 2.631638s : 1: type_inference 0.00% : 0.000723s : 1: validate TotalTime = 2.86701, [24] [bootstrap]: 1.01967 [type_inference]: 0.00642847 [event_method]: 1.912e-05 [auto_monad]: 0.00016018 [graph_reusing]: 6.83e-06 [inline]: 2.78e-06 [add_attr]: 1.04977, [1] [add_attr_with_inline]: 1.04976, [1] [Cycle 1]: 8.351e-05, [2] [tag_attr]: 2.8e-05 [meta_addattr_fg_expand]: 6.01998e-06 [parallel-infer-symbol]: 4.30999e-06 [pre_auto_parallel]: 4.447e-05 [insert-virtual-dataset]: 2.69999e-06 [parallel-infer-symbol-second]: 7.50006e-07 [dataset_repeat_opt]: 2.26998e-06 [pipeline_split]: 1.67999e-06 [optimize]: 0.00637849, [53] [py_interpret_to_execute]: 3.411e-05 [rewriter_before_opt_a]: 9.772e-05 [opt_a]: 0.00352125, [2] [Cycle 1]: 0.00259147, [45] [expand_dump_flag]: 3.08e-06 [switch_simplify]: 4.387e-05 [loop_unroll]: 2.991e-05 [a_1]: 0.00070044 [with_stream_mark]: 2.129e-05 [recompute_prepare]: 1.191e-05 [updatestate_depend_eliminate]: 6.53e-06 [updatestate_assign_eliminate]: 6.48998e-06 [updatestate_loads_eliminate]: 5.00999e-06 [parameter_eliminate]: 2.31e-06 [a_2]: 0.00012263 [accelerated_algorithm]: 9.96998e-06 [shard]: 2.39001e-06 [meta_shard_fg_expand]: 2.21998e-06 [shard_inline]: 8.38001e-06 [merge_send_recv]: 1.1e-05 [auto_parallel]: 8.80999e-06 [parallel]: 3.517e-05 [flash_sp]: 1.008e-05 [merge_comm]: 5.59e-06 [allreduce_fusion]: 5.01997e-06 [matmul_add_comm_reduction]: 1.291e-05 [allreduce_slice_to_reducescatter]: 8.2e-07 [virtual_shard_identity]: 1.025e-05 [virtual_dataset]: 9.41e-06 [get_grad_eliminate_]: 8.38001e-06 [virtual_output]: 8.60001e-06 [merge_forward]: 5.75001e-06 [cell_reuse_recompute_pass]: 1.25999e-06 [offload_activation]: 1.44e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.717e-05 [merge_recompute_call_nodes]: 1.57001e-06 [before_grad]: 1.471e-05 [set_forward_comm_id_for_comm_node_pass]: 5.45001e-06 [meta_fg_expand]: 3.78999e-06 [flash_sp_send_recv_attached]: 2.58998e-06 [receive_attached]: 2.44001e-06 [after_resolve]: 1.179e-05 [a_after_grad]: 1.279e-05 [renormalize]: 0.00100405 [add_forward_monad_depend]: 5.70001e-06 [auto_monad_grad]: 2.63e-06 [auto_monad_eliminator]: 2.437e-05 [cse]: 4.212e-05 [a_3]: 7.46e-05 [Cycle 2]: 0.00091841, [45] [expand_dump_flag]: 1.34e-06 [switch_simplify]: 1.124e-05 [loop_unroll]: 8.87e-06 [a_1]: 0.00022904 [with_stream_mark]: 1.454e-05 [recompute_prepare]: 9.84999e-06 [updatestate_depend_eliminate]: 5.30001e-06 [updatestate_assign_eliminate]: 4.92999e-06 [updatestate_loads_eliminate]: 5.02e-06 [parameter_eliminate]: 1.02e-06 [a_2]: 0.00011966 [accelerated_algorithm]: 9.27999e-06 [shard]: 1.07e-06 [meta_shard_fg_expand]: 2.01998e-06 [shard_inline]: 9.32001e-06 [merge_send_recv]: 7.61001e-06 [auto_parallel]: 8.43001e-06 [parallel]: 5.77999e-06 [flash_sp]: 3.91999e-06 [merge_comm]: 5.48002e-06 [allreduce_fusion]: 5.05001e-06 [matmul_add_comm_reduction]: 8.26002e-06 [allreduce_slice_to_reducescatter]: 3.9002e-07 [virtual_shard_identity]: 1.085e-05 [virtual_dataset]: 9.32001e-06 [get_grad_eliminate_]: 8.84e-06 [virtual_output]: 8.47e-06 [merge_forward]: 4.97e-06 [cell_reuse_recompute_pass]: 1.69998e-06 [offload_activation]: 1.13e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.809e-05 [merge_recompute_call_nodes]: 9.10019e-07 [before_grad]: 1.536e-05 [set_forward_comm_id_for_comm_node_pass]: 5.87001e-06 [meta_fg_expand]: 3.36001e-06 [flash_sp_send_recv_attached]: 8.59989e-07 [receive_attached]: 1.20999e-06 [after_resolve]: 1.196e-05 [a_after_grad]: 1.279e-05 [renormalize]: 9.00181e-08 [add_forward_monad_depend]: 1.38002e-06 [auto_monad_grad]: 1.09998e-06 [auto_monad_eliminator]: 1.248e-05 [cse]: 2.332e-05 [a_3]: 5.504e-05 [py_interpret_to_execute_after_opt_a]: 1.352e-05 [slice_cell_reuse_recomputed_activation]: 2.04999e-06 [rewriter_after_opt_a]: 0.00016794 [convert_after_rewriter]: 1.057e-05 [order_py_execute_after_rewriter]: 7.2e-06 [mutable_eliminate]: 0.00077756 [opt_b]: 0.00030467, [1] [Cycle 1]: 0.00029696, [7] [b_1]: 0.00019916 [b_2]: 1.193e-05 [updatestate_depend_eliminate]: 9.34998e-06 [updatestate_assign_eliminate]: 4.84e-06 [updatestate_loads_eliminate]: 4.68999e-06 [renormalize]: 6.09987e-07 [cse]: 2.759e-05 [optimize_parallel_all_gather_comm]: 2.291e-05 [overlap_param_gather]: 5.34e-06 [cconv]: 2.942e-05 [loop_unroll]: 0.00048925 [opt_after_cconv]: 0.00014588, [1] [Cycle 1]: 0.00013934, [7] [c_1]: 5.732e-05 [parameter_eliminate]: 3.09001e-06 [updatestate_depend_eliminate]: 8.67e-06 [updatestate_assign_eliminate]: 4.62e-06 [updatestate_loads_eliminate]: 4.25e-06 [cse]: 2.455e-05 [renormalize]: 3.00002e-07 [remove_dup_value]: 1.6e-05 [tuple_transform]: 9.429e-05, [1] [Cycle 1]: 8.945e-05, [4] [d_1]: 5.88e-05 [none_parameter_eliminate]: 1.89999e-06 [renormalize]: 2.29978e-07 [switch_simplify]: 9.58002e-06 [partial_unused_args_eliminate]: 1.97001e-06 [add_recomputation]: 7.121e-05 [cse_after_recomputation]: 2.941e-05, [1] [Cycle 1]: 2.442e-05, [1] [cse]: 1.874e-05 [environ_conv]: 7.35e-06 [swap_dp_allreduce_reducescatter]: 8.02998e-06 [bias_add_comm_swap]: 2.96999e-06 [label_micro_interleaved_index]: 4.72e-06 [label_fine_grained_interleaved_index]: 3.15998e-06 [merge_cast_opt]: 1.74998e-06 [slice_recompute_activation]: 2.26e-06 [micro_interleaved_order_control]: 2.66e-06 [assign_add_opt]: 1.50001e-06 [ForceFp32Comm]: 8.80013e-07 [remove_cast_before_assign_add]: 1.35999e-06 [full_micro_interleaved_order_control]: 2.76e-06 [reorder_send_recv_between_fp_bp]: 2.84001e-06 [comm_op_add_attrs]: 1.27999e-06 [add_comm_op_reuse_tag]: 1.47999e-06 [interleave_split_concat_branches]: 1.30999e-06 [interleave_parallel_branches]: 1.14e-06 [overlap_opt_shard_in_pipeline]: 3.203e-05 [overlap_opt_shard_grad_in_pipeline]: 2.34001e-06 [control_data_broadcast_order]: 1.984e-05 [grouped_pairwise_exchange_alltoall]: 1.79998e-06 [offloading_packed_experts]: 5.35999e-06 [overlap_recompute_and_grad_model_parallel]: 6.63998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.34e-06 [overlap_recompute_allgather_and_fa_grad]: 1.50999e-06 [overlap_recompute_comm]: 2.82002e-06 [overlap_grad_ring_attention]: 5.69e-06 [overlap_grad_flash_sp]: 2.834e-05 [begin_end_overlap_inline]: 5.89993e-07 [split_matmul_comm_elemetwise]: 2.64001e-06 [split_layernorm_comm]: 2.09e-06 [handle_group_info]: 1.14003e-06 [symbol_engine_optimizer]: 9.91e-05, [1] [Cycle 1]: 9.425e-05, [6] [build]: 5.09003e-06 [elim_shapecalc]: 1.4e-05 [elim_not_effective]: 1.866e-05 [opt_reshape]: 1.008e-05 [fold_const_symbol]: 1.543e-05 [renormalize]: 1.90019e-07 [detach_backward]: 1.99999e-06 [pipeline_parallel_scheduler]: 1.64e-06 [auto_monad_reorder]: 3.152e-05 [get_jit_bprop_graph]: 2.26e-06 [rewriter_after_jit_bprop_graph]: 4.60001e-06 [opt_after_jit_grad]: 0.00086653 [validate]: 5.657e-05 [backend_pass]: 8.80013e-07 [task_emit]: 0.783296 [execute]: 8.81997e-06 Sums bootstrap : 1.019666s : 56.15% type_inference : 0.006428s : 0.35% event_method : 0.000019s : 0.00% auto_monad : 0.000160s : 0.01% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000028s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000006s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000044s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000034s : 0.00% optimize.rewriter_before_opt_a : 0.000098s : 0.01% optimize.opt_a.expand_dump_flag : 0.000004s : 0.00% optimize.opt_a.switch_simplify : 0.000055s : 0.00% optimize.opt_a.loop_unroll : 0.000039s : 0.00% optimize.opt_a.a_1 : 0.000929s : 0.05% optimize.opt_a.with_stream_mark : 0.000036s : 0.00% optimize.opt_a.recompute_prepare : 0.000022s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 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 : 0.000003s : 0.00% optimize.opt_a.a_2 : 0.000242s : 0.01% optimize.opt_a.accelerated_algorithm : 0.000019s : 0.00% optimize.opt_a.shard : 0.000003s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000004s : 0.00% optimize.opt_a.shard_inline : 0.000018s : 0.00% optimize.opt_a.merge_send_recv : 0.000019s : 0.00% optimize.opt_a.auto_parallel : 0.000017s : 0.00% optimize.opt_a.parallel : 0.000041s : 0.00% optimize.opt_a.flash_sp : 0.000014s : 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 : 0.000021s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000021s : 0.00% optimize.opt_a.virtual_dataset : 0.000019s : 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 : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000026s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000035s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 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 : 0.000007s : 0.00% optimize.opt_a.flash_sp_send_recv_attached : 0.000003s : 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 : 0.000026s : 0.00% optimize.opt_a.renormalize : 0.001004s : 0.06% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000037s : 0.00% optimize.opt_a.cse : 0.000065s : 0.00% optimize.opt_a.a_3 : 0.000130s : 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 : 0.000168s : 0.01% optimize.convert_after_rewriter : 0.000011s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000778s : 0.04% optimize.opt_b.b_1 : 0.000199s : 0.01% optimize.opt_b.b_2 : 0.000012s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000028s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000023s : 0.00% optimize.overlap_param_gather : 0.000005s : 0.00% optimize.cconv : 0.000029s : 0.00% optimize.loop_unroll : 0.000489s : 0.03% optimize.opt_after_cconv.c_1 : 0.000057s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000025s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000016s : 0.00% optimize.tuple_transform.d_1 : 0.000059s : 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 : 0.000010s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000071s : 0.00% optimize.cse_after_recomputation.cse : 0.000019s : 0.00% 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 : 0.000005s : 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 : 0.000002s : 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 : 0.000032s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000020s : 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 : 0.000007s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000028s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000019s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000032s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000867s : 0.05% validate : 0.000057s : 0.00% backend_pass : 0.000001s : 0.00% task_emit : 0.783296s : 43.13% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.000296 60 2.58% : 0.000008s : 2: substitution.depend_value_elim 1.02% : 0.000003s : 5: substitution.elim_not_effective 0.84% : 0.000002s : 5: substitution.fold_const_symbol 2.66% : 0.000008s : 6: substitution.graph_param_transform 71.61% : 0.000212s : 4: substitution.inline 1.88% : 0.000006s : 10: substitution.j_node_and_user_rematch 2.92% : 0.000009s : 10: substitution.remove_not_recompute_node 1.35% : 0.000004s : 2: substitution.replace_old_param 4.54% : 0.000013s : 2: substitution.tuple_list_get_item_eliminator 5.91% : 0.000017s : 6: substitution.updatestate_pure_node_eliminater 4.69% : 0.000014s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.006363 2 89.37% : 0.005687s : 1: type_inference.infer 10.63% : 0.000677s : 1: type_inference.specialize ------[replace.] 0.000062 6 72.80% : 0.000045s : 4: replace.inline 27.20% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000222 6 94.45% : 0.000209s : 4: match.inline 5.55% : 0.000012s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000280 1652 0.90% : 0.000003s : 16: predicate.accumulaten_eliminater 0.81% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.68% : 0.000002s : 12: predicate.addn_check_dump 0.90% : 0.000003s : 16: predicate.addn_zero_filter 0.83% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.23% : 0.000006s : 28: predicate.arithmetic_simplify 0.86% : 0.000002s : 16: predicate.cast_eliminate 0.68% : 0.000002s : 12: predicate.check_bprop_eliminate 0.61% : 0.000002s : 12: predicate.compare_switch_simplify 0.19% : 0.000001s : 6: predicate.const_output_eliminate 0.72% : 0.000002s : 12: predicate.depend_value_elim 0.93% : 0.000003s : 16: predicate.dict_get_item_const_eliminator 0.98% : 0.000003s : 16: predicate.dict_get_item_eliminator 1.04% : 0.000003s : 16: predicate.dict_set_item_eliminator 1.12% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.25% : 0.000001s : 6: predicate.elim_not_effective 0.50% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.27% : 0.000004s : 22: predicate.environ_add_const_eliminate 1.32% : 0.000004s : 22: predicate.environ_get_add_eliminate 1.18% : 0.000003s : 22: predicate.environ_get_depend_swap 1.99% : 0.000006s : 34: predicate.environ_get_eliminate 1.17% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.11% : 0.000003s : 22: predicate.exchange_switch_depend_value 1.66% : 0.000005s : 22: predicate.float_depend_g_call 0.64% : 0.000002s : 12: predicate.float_environ_get_switch 0.93% : 0.000003s : 18: predicate.float_tuple_getitem_switch 0.18% : 0.000001s : 6: predicate.fold_const_symbol 0.80% : 0.000002s : 12: predicate.get_grad_eliminate 0.27% : 0.000001s : 6: predicate.graph_param_transform 0.64% : 0.000002s : 12: predicate.incorporate_call 0.54% : 0.000002s : 12: predicate.incorporate_call_switch 5.74% : 0.000016s : 74: predicate.inline 0.81% : 0.000002s : 12: predicate.inline_without_move 0.30% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.96% : 0.000003s : 12: predicate.less_batch_normalization 1.69% : 0.000005s : 30: predicate.list_to_tuple_eliminator_ 2.43% : 0.000007s : 46: predicate.load_eliminater 0.97% : 0.000003s : 6: predicate.loop_unroll_after_grad 1.89% : 0.000005s : 38: predicate.loop_unroll_before_grad 1.86% : 0.000005s : 28: predicate.make_slice_get_slice_eliminator 0.72% : 0.000002s : 12: predicate.merge_addn 0.67% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.67% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.78% : 0.000002s : 16: predicate.minmaximum_grad 1.12% : 0.000003s : 6: predicate.mutable_eliminate 0.46% : 0.000001s : 6: predicate.opt_reshape 0.56% : 0.000002s : 6: predicate.parallel_virtual_node 1.34% : 0.000004s : 22: predicate.partial_defer_inline 1.40% : 0.000004s : 24: predicate.partial_eliminate 0.88% : 0.000002s : 16: predicate.print_const_string_wrapper 0.68% : 0.000002s : 12: predicate.reduce_all_const_elim 1.21% : 0.000003s : 16: predicate.reduce_eliminate 2.53% : 0.000007s : 46: predicate.redundant_stop_gradient_eliminater 0.45% : 0.000001s : 12: predicate.remove_not_recompute_node 1.10% : 0.000003s : 30: predicate.replace_applicator 0.56% : 0.000002s : 12: predicate.replace_old_param 0.24% : 0.000001s : 6: predicate.reset_defer_inline 0.90% : 0.000003s : 16: predicate.reshape_eliminate 0.83% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.55% : 0.000002s : 6: predicate.row_tensor_eliminate 0.91% : 0.000003s : 12: predicate.same_eliminate 0.43% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.78% : 0.000002s : 12: predicate.shard_identity_eliminate 0.90% : 0.000003s : 12: predicate.special_op_eliminate 0.82% : 0.000002s : 12: predicate.specialize_transform 0.97% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 0.87% : 0.000002s : 12: predicate.stack_unstack_eliminate 0.41% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.21% : 0.000003s : 22: predicate.switch_defer_inline 1.84% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.41% : 0.000012s : 78: predicate.switch_simplify 0.90% : 0.000003s : 16: predicate.tile_eliminate 0.93% : 0.000003s : 16: predicate.transpose_eliminate 1.64% : 0.000005s : 28: predicate.tuple_list_convert_item_index_to_positive 1.59% : 0.000004s : 28: predicate.tuple_list_get_item_const_eliminator 1.66% : 0.000005s : 28: predicate.tuple_list_get_item_depend_reorder 3.18% : 0.000009s : 42: predicate.tuple_list_get_item_eliminator 1.64% : 0.000005s : 28: predicate.tuple_list_get_set_item_eliminator 2.32% : 0.000006s : 40: predicate.tuple_list_set_item_eliminator 1.65% : 0.000005s : 30: predicate.tuple_to_list_eliminator_ 2.46% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.31% : 0.000009s : 58: predicate.updatestate_useless_node_eliminater 0.43% : 0.000001s : 6: predicate.value_based_eliminate 0.78% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.84% : 0.000002s : 12: predicate.virtual_output_eliminate 0.35% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.52% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000552 11 52.61% : 0.000290s : 5: func_graph_cloner_run.FuncGraphClonerGraph 47.39% : 0.000262s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 3.926038 196 0.00% : 0.000004s : 1: ForceFp32Comm 26.74% : 1.049782s : 1: add_attr 26.74% : 1.049763s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000076s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.00% : 0.000168s : 1: auto_monad 0.00% : 0.000035s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 25.97% : 1.019709s : 1: bootstrap 0.00% : 0.000033s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000023s : 1: control_data_broadcast_order 0.00% : 0.000014s : 1: convert_after_rewriter 0.00% : 0.000032s : 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 0.00% : 0.000025s : 1: event_method 0.00% : 0.000016s : 1: execute 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000006s : 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% : 0.000008s : 1: label_micro_interleaved_index 0.01% : 0.000498s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.02% : 0.000786s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000020s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000021s : 1: opt.transform.mutable_eliminate 0.04% : 0.001549s : 78: opt.transform.opt_a 0.00% : 0.000055s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000180s : 28: opt.transform.opt_b 0.00% : 0.000066s : 2: opt.transform.opt_trans_graph 0.00% : 0.000054s : 4: opt.transform.symbol_engine_opt 0.09% : 0.003525s : 1: opt_a 0.00% : 0.000150s : 1: opt_after_cconv 0.02% : 0.000878s : 1: opt_after_jit_grad 0.01% : 0.000308s : 1: opt_b 0.16% : 0.006383s : 1: optimize 0.00% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000011s : 1: order_py_execute_after_rewriter 0.00% : 0.000032s : 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% : 0.000036s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000009s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000010s : 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 0.00% : 0.000049s : 1: pre_auto_parallel 0.00% : 0.000074s : 1: py_interpret_to_execute 0.00% : 0.000017s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000005s : 1: remove_cast_before_assign_add 0.00% : 0.000020s : 1: remove_dup_value 0.01% : 0.000578s : 1: renormalize.infer 0.01% : 0.000418s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000174s : 1: rewriter_after_opt_a 0.00% : 0.000103s : 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% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000102s : 1: symbol_engine_optimizer 19.95% : 0.783318s : 1: task_emit 0.00% : 0.000097s : 1: tuple_transform 0.16% : 0.006447s : 1: type_inference 0.00% : 0.000088s : 1: validate TotalTime = 0.599573, [24] [bootstrap]: 0.00072131 [type_inference]: 0.0073308 [event_method]: 1.941e-05 [auto_monad]: 0.00015758 [graph_reusing]: 6.51e-06 [inline]: 2.02999e-06 [add_attr]: 0.576327, [1] [add_attr_with_inline]: 0.576309, [1] [Cycle 1]: 8.85e-05, [2] [tag_attr]: 2.861e-05 [meta_addattr_fg_expand]: 5.92999e-06 [parallel-infer-symbol]: 3.93001e-06 [pre_auto_parallel]: 4.59e-05 [insert-virtual-dataset]: 2.88e-06 [parallel-infer-symbol-second]: 6.59988e-07 [dataset_repeat_opt]: 2.01e-06 [pipeline_split]: 1.64e-06 [optimize]: 0.00609013, [53] [py_interpret_to_execute]: 3.567e-05 [rewriter_before_opt_a]: 9.305e-05 [opt_a]: 0.00335445, [2] [Cycle 1]: 0.00248623, [45] [expand_dump_flag]: 2.67001e-06 [switch_simplify]: 4.471e-05 [loop_unroll]: 2.879e-05 [a_1]: 0.00065578 [with_stream_mark]: 2.082e-05 [recompute_prepare]: 1.073e-05 [updatestate_depend_eliminate]: 6.06e-06 [updatestate_assign_eliminate]: 5.97001e-06 [updatestate_loads_eliminate]: 5.10001e-06 [parameter_eliminate]: 2.07999e-06 [a_2]: 0.00011892 [accelerated_algorithm]: 9.39998e-06 [shard]: 2.07001e-06 [meta_shard_fg_expand]: 2.32999e-06 [shard_inline]: 8.42998e-06 [merge_send_recv]: 1.057e-05 [auto_parallel]: 7.74002e-06 [parallel]: 2.015e-05 [flash_sp]: 8.92e-06 [merge_comm]: 5.19e-06 [allreduce_fusion]: 5.22e-06 [matmul_add_comm_reduction]: 1.274e-05 [allreduce_slice_to_reducescatter]: 1.25001e-06 [virtual_shard_identity]: 1.126e-05 [virtual_dataset]: 9.45001e-06 [get_grad_eliminate_]: 7.92998e-06 [virtual_output]: 8.62e-06 [merge_forward]: 5.42001e-06 [cell_reuse_recompute_pass]: 1.52001e-06 [offload_activation]: 1.487e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.762e-05 [merge_recompute_call_nodes]: 1.90001e-06 [before_grad]: 1.513e-05 [set_forward_comm_id_for_comm_node_pass]: 5.14998e-06 [meta_fg_expand]: 3.76999e-06 [flash_sp_send_recv_attached]: 2.71999e-06 [receive_attached]: 2.21e-06 [after_resolve]: 1.174e-05 [a_after_grad]: 1.308e-05 [renormalize]: 0.00097925 [add_forward_monad_depend]: 5.40001e-06 [auto_monad_grad]: 2.66e-06 [auto_monad_eliminator]: 2.226e-05 [cse]: 4.18e-05 [a_3]: 6.853e-05 [Cycle 2]: 0.00083913, [45] [expand_dump_flag]: 9.39996e-07 [switch_simplify]: 1.2e-05 [loop_unroll]: 8.59998e-06 [a_1]: 0.00019794 [with_stream_mark]: 1.466e-05 [recompute_prepare]: 8.58001e-06 [updatestate_depend_eliminate]: 5.72999e-06 [updatestate_assign_eliminate]: 4.91997e-06 [updatestate_loads_eliminate]: 4.45e-06 [parameter_eliminate]: 1.23002e-06 [a_2]: 0.00010864 [accelerated_algorithm]: 8.67998e-06 [shard]: 1.28002e-06 [meta_shard_fg_expand]: 1.96e-06 [shard_inline]: 8.16002e-06 [merge_send_recv]: 6.91001e-06 [auto_parallel]: 8.05e-06 [parallel]: 5.27001e-06 [flash_sp]: 4.07e-06 [merge_comm]: 4.72998e-06 [allreduce_fusion]: 4.72e-06 [matmul_add_comm_reduction]: 8.14002e-06 [allreduce_slice_to_reducescatter]: 3.69997e-07 [virtual_shard_identity]: 9.92999e-06 [virtual_dataset]: 7.74002e-06 [get_grad_eliminate_]: 7.88999e-06 [virtual_output]: 7.4e-06 [merge_forward]: 4.19002e-06 [cell_reuse_recompute_pass]: 1.32e-06 [offload_activation]: 1.021e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.632e-05 [merge_recompute_call_nodes]: 7.59988e-07 [before_grad]: 1.423e-05 [set_forward_comm_id_for_comm_node_pass]: 5.55001e-06 [meta_fg_expand]: 3.08e-06 [flash_sp_send_recv_attached]: 1.09e-06 [receive_attached]: 1.10999e-06 [after_resolve]: 1.067e-05 [a_after_grad]: 1.256e-05 [renormalize]: 1.10012e-07 [add_forward_monad_depend]: 1.49e-06 [auto_monad_grad]: 1.10999e-06 [auto_monad_eliminator]: 1.174e-05 [cse]: 2.033e-05 [a_3]: 5.219e-05 [py_interpret_to_execute_after_opt_a]: 1.209e-05 [slice_cell_reuse_recomputed_activation]: 1.89e-06 [rewriter_after_opt_a]: 0.00017625 [convert_after_rewriter]: 1.073e-05 [order_py_execute_after_rewriter]: 7.23e-06 [mutable_eliminate]: 0.00076364 [opt_b]: 0.00028707, [1] [Cycle 1]: 0.00028009, [7] [b_1]: 0.00018524 [b_2]: 1.151e-05 [updatestate_depend_eliminate]: 7.9e-06 [updatestate_assign_eliminate]: 4.92e-06 [updatestate_loads_eliminate]: 4.43999e-06 [renormalize]: 7.00005e-07 [cse]: 2.927e-05 [optimize_parallel_all_gather_comm]: 2.238e-05 [overlap_param_gather]: 2.26e-06 [cconv]: 3.122e-05 [loop_unroll]: 0.00049638 [opt_after_cconv]: 0.00013872, [1] [Cycle 1]: 0.00013277, [7] [c_1]: 5.302e-05 [parameter_eliminate]: 3.3e-06 [updatestate_depend_eliminate]: 7.76001e-06 [updatestate_assign_eliminate]: 4.45e-06 [updatestate_loads_eliminate]: 3.93999e-06 [cse]: 2.48e-05 [renormalize]: 3.39991e-07 [remove_dup_value]: 1.753e-05 [tuple_transform]: 8.988e-05, [1] [Cycle 1]: 8.519e-05, [4] [d_1]: 5.608e-05 [none_parameter_eliminate]: 1.64e-06 [renormalize]: 1.59984e-07 [switch_simplify]: 9.16998e-06 [partial_unused_args_eliminate]: 1.81e-06 [add_recomputation]: 7.117e-05 [cse_after_recomputation]: 2.752e-05, [1] [Cycle 1]: 2.266e-05, [1] [cse]: 1.731e-05 [environ_conv]: 7.24001e-06 [swap_dp_allreduce_reducescatter]: 8.03001e-06 [bias_add_comm_swap]: 3.53e-06 [label_micro_interleaved_index]: 4.60999e-06 [label_fine_grained_interleaved_index]: 2.86999e-06 [merge_cast_opt]: 1.99e-06 [slice_recompute_activation]: 2.64999e-06 [micro_interleaved_order_control]: 2.46e-06 [assign_add_opt]: 1.64e-06 [ForceFp32Comm]: 9.99979e-07 [remove_cast_before_assign_add]: 1.09e-06 [full_micro_interleaved_order_control]: 2.26e-06 [reorder_send_recv_between_fp_bp]: 3.01001e-06 [comm_op_add_attrs]: 1.09e-06 [add_comm_op_reuse_tag]: 1.40001e-06 [interleave_split_concat_branches]: 1.22e-06 [interleave_parallel_branches]: 1.09e-06 [overlap_opt_shard_in_pipeline]: 1.20999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.19001e-06 [control_data_broadcast_order]: 1.82e-05 [grouped_pairwise_exchange_alltoall]: 1.76998e-06 [offloading_packed_experts]: 5.43002e-06 [overlap_recompute_and_grad_model_parallel]: 5.89999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.22999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40001e-06 [overlap_recompute_comm]: 2.76999e-06 [overlap_grad_ring_attention]: 5.84e-06 [overlap_grad_flash_sp]: 2.768e-05 [begin_end_overlap_inline]: 6.10016e-07 [split_matmul_comm_elemetwise]: 2.46998e-06 [split_layernorm_comm]: 1.82001e-06 [handle_group_info]: 1.05999e-06 [symbol_engine_optimizer]: 9.426e-05, [1] [Cycle 1]: 8.948e-05, [6] [build]: 4.92e-06 [elim_shapecalc]: 1.412e-05 [elim_not_effective]: 1.813e-05 [opt_reshape]: 9.48002e-06 [fold_const_symbol]: 1.452e-05 [renormalize]: 2.00002e-07 [detach_backward]: 2.20002e-06 [pipeline_parallel_scheduler]: 1.60999e-06 [auto_monad_reorder]: 3.012e-05 [get_jit_bprop_graph]: 1.96e-06 [rewriter_after_jit_bprop_graph]: 5.19e-06 [opt_after_jit_grad]: 0.00057153 [validate]: 8.037e-05 [backend_pass]: 1.04e-06 [task_emit]: 0.00791786 [execute]: 7.82e-06 Sums bootstrap : 0.000721s : 3.25% type_inference : 0.007331s : 33.04% event_method : 0.000019s : 0.09% auto_monad : 0.000158s : 0.71% graph_reusing : 0.000007s : 0.03% inline : 0.000002s : 0.01% add_attr.add_attr_with_inline.tag_attr : 0.000029s : 0.13% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000006s : 0.03% parallel-infer-symbol : 0.000004s : 0.02% pre_auto_parallel : 0.000046s : 0.21% insert-virtual-dataset : 0.000003s : 0.01% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.01% pipeline_split : 0.000002s : 0.01% optimize.py_interpret_to_execute : 0.000036s : 0.16% optimize.rewriter_before_opt_a : 0.000093s : 0.42% optimize.opt_a.expand_dump_flag : 0.000004s : 0.02% optimize.opt_a.switch_simplify : 0.000057s : 0.26% optimize.opt_a.loop_unroll : 0.000037s : 0.17% optimize.opt_a.a_1 : 0.000854s : 3.85% optimize.opt_a.with_stream_mark : 0.000035s : 0.16% optimize.opt_a.recompute_prepare : 0.000019s : 0.09% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 0.05% optimize.opt_a.updatestate_assign_eliminate : 0.000011s : 0.05% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.04% optimize.opt_a.parameter_eliminate : 0.000003s : 0.01% optimize.opt_a.a_2 : 0.000228s : 1.03% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.08% optimize.opt_a.shard : 0.000003s : 0.02% optimize.opt_a.meta_shard_fg_expand : 0.000004s : 0.02% optimize.opt_a.shard_inline : 0.000017s : 0.07% optimize.opt_a.merge_send_recv : 0.000017s : 0.08% optimize.opt_a.auto_parallel : 0.000016s : 0.07% optimize.opt_a.parallel : 0.000025s : 0.11% optimize.opt_a.flash_sp : 0.000013s : 0.06% optimize.opt_a.merge_comm : 0.000010s : 0.04% optimize.opt_a.allreduce_fusion : 0.000010s : 0.04% optimize.opt_a.matmul_add_comm_reduction : 0.000021s : 0.09% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.01% optimize.opt_a.virtual_shard_identity : 0.000021s : 0.10% optimize.opt_a.virtual_dataset : 0.000017s : 0.08% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.07% optimize.opt_a.virtual_output : 0.000016s : 0.07% optimize.opt_a.merge_forward : 0.000010s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.01% optimize.opt_a.offload_activation : 0.000025s : 0.11% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000034s : 0.15% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.01% optimize.opt_a.before_grad : 0.000029s : 0.13% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.05% optimize.opt_a.meta_fg_expand : 0.000007s : 0.03% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.02% optimize.opt_a.receive_attached : 0.000003s : 0.01% optimize.opt_a.after_resolve : 0.000022s : 0.10% optimize.opt_a.a_after_grad : 0.000026s : 0.12% optimize.opt_a.renormalize : 0.000979s : 4.41% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.03% optimize.opt_a.auto_monad_grad : 0.000004s : 0.02% optimize.opt_a.auto_monad_eliminator : 0.000034s : 0.15% optimize.opt_a.cse : 0.000062s : 0.28% optimize.opt_a.a_3 : 0.000121s : 0.54% optimize.py_interpret_to_execute_after_opt_a : 0.000012s : 0.05% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.01% optimize.rewriter_after_opt_a : 0.000176s : 0.79% optimize.convert_after_rewriter : 0.000011s : 0.05% optimize.order_py_execute_after_rewriter : 0.000007s : 0.03% optimize.mutable_eliminate : 0.000764s : 3.44% optimize.opt_b.b_1 : 0.000185s : 0.83% optimize.opt_b.b_2 : 0.000012s : 0.05% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.04% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.02% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.02% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000029s : 0.13% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.10% optimize.overlap_param_gather : 0.000002s : 0.01% optimize.cconv : 0.000031s : 0.14% optimize.loop_unroll : 0.000496s : 2.24% optimize.opt_after_cconv.c_1 : 0.000053s : 0.24% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.01% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.03% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000004s : 0.02% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.02% optimize.opt_after_cconv.cse : 0.000025s : 0.11% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.08% optimize.tuple_transform.d_1 : 0.000056s : 0.25% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000009s : 0.04% optimize.partial_unused_args_eliminate : 0.000002s : 0.01% optimize.add_recomputation : 0.000071s : 0.32% optimize.cse_after_recomputation.cse : 0.000017s : 0.08% optimize.environ_conv : 0.000007s : 0.03% optimize.swap_dp_allreduce_reducescatter : 0.000008s : 0.04% optimize.bias_add_comm_swap : 0.000004s : 0.02% optimize.label_micro_interleaved_index : 0.000005s : 0.02% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.01% optimize.merge_cast_opt : 0.000002s : 0.01% optimize.slice_recompute_activation : 0.000003s : 0.01% optimize.micro_interleaved_order_control : 0.000002s : 0.01% optimize.assign_add_opt : 0.000002s : 0.01% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.01% 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 : 0.01% optimize.interleave_split_concat_branches : 0.000001s : 0.01% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.01% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.01% optimize.control_data_broadcast_order : 0.000018s : 0.08% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.01% optimize.offloading_packed_experts : 0.000005s : 0.02% optimize.overlap_recompute_and_grad_model_parallel : 0.000006s : 0.03% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.01% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.01% optimize.overlap_recompute_comm : 0.000003s : 0.01% optimize.overlap_grad_ring_attention : 0.000006s : 0.03% optimize.overlap_grad_flash_sp : 0.000028s : 0.12% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.01% optimize.split_layernorm_comm : 0.000002s : 0.01% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000005s : 0.02% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.06% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.08% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.04% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 0.07% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.01% pipeline_parallel_scheduler : 0.000002s : 0.01% auto_monad_reorder : 0.000030s : 0.14% get_jit_bprop_graph : 0.000002s : 0.01% rewriter_after_jit_bprop_graph : 0.000005s : 0.02% opt_after_jit_grad : 0.000572s : 2.58% validate : 0.000080s : 0.36% backend_pass : 0.000001s : 0.00% task_emit : 0.007918s : 35.69% execute : 0.000008s : 0.04% Time group info: ------[substitution.] 0.000266 60 2.59% : 0.000007s : 2: substitution.depend_value_elim 1.19% : 0.000003s : 5: substitution.elim_not_effective 0.86% : 0.000002s : 5: substitution.fold_const_symbol 2.85% : 0.000008s : 6: substitution.graph_param_transform 70.56% : 0.000188s : 4: substitution.inline 2.06% : 0.000005s : 10: substitution.j_node_and_user_rematch 2.95% : 0.000008s : 10: substitution.remove_not_recompute_node 1.39% : 0.000004s : 2: substitution.replace_old_param 4.44% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator 5.99% : 0.000016s : 6: substitution.updatestate_pure_node_eliminater 5.12% : 0.000014s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.007273 2 90.37% : 0.006573s : 1: type_inference.infer 9.63% : 0.000700s : 1: type_inference.specialize ------[replace.] 0.000060 6 72.91% : 0.000043s : 4: replace.inline 27.09% : 0.000016s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000196 6 94.55% : 0.000185s : 4: match.inline 5.45% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000267 1652 0.90% : 0.000002s : 16: predicate.accumulaten_eliminater 0.80% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.72% : 0.000002s : 12: predicate.addn_check_dump 0.95% : 0.000003s : 16: predicate.addn_zero_filter 0.77% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.09% : 0.000006s : 28: predicate.arithmetic_simplify 0.90% : 0.000002s : 16: predicate.cast_eliminate 0.83% : 0.000002s : 12: predicate.check_bprop_eliminate 0.69% : 0.000002s : 12: predicate.compare_switch_simplify 0.18% : 0.000000s : 6: predicate.const_output_eliminate 0.71% : 0.000002s : 12: predicate.depend_value_elim 0.92% : 0.000002s : 16: predicate.dict_get_item_const_eliminator 0.95% : 0.000003s : 16: predicate.dict_get_item_eliminator 0.90% : 0.000002s : 16: predicate.dict_set_item_eliminator 1.07% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.22% : 0.000001s : 6: predicate.elim_not_effective 0.51% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.28% : 0.000003s : 22: predicate.environ_add_const_eliminate 1.21% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.16% : 0.000003s : 22: predicate.environ_get_depend_swap 1.86% : 0.000005s : 34: predicate.environ_get_eliminate 1.12% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.15% : 0.000003s : 22: predicate.exchange_switch_depend_value 1.81% : 0.000005s : 22: predicate.float_depend_g_call 0.69% : 0.000002s : 12: predicate.float_environ_get_switch 1.10% : 0.000003s : 18: predicate.float_tuple_getitem_switch 0.18% : 0.000000s : 6: predicate.fold_const_symbol 0.69% : 0.000002s : 12: predicate.get_grad_eliminate 0.27% : 0.000001s : 6: predicate.graph_param_transform 0.66% : 0.000002s : 12: predicate.incorporate_call 0.54% : 0.000001s : 12: predicate.incorporate_call_switch 5.66% : 0.000015s : 74: predicate.inline 0.88% : 0.000002s : 12: predicate.inline_without_move 0.33% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.95% : 0.000003s : 12: predicate.less_batch_normalization 1.65% : 0.000004s : 30: predicate.list_to_tuple_eliminator_ 2.59% : 0.000007s : 46: predicate.load_eliminater 0.85% : 0.000002s : 6: predicate.loop_unroll_after_grad 1.88% : 0.000005s : 38: predicate.loop_unroll_before_grad 1.77% : 0.000005s : 28: predicate.make_slice_get_slice_eliminator 0.77% : 0.000002s : 12: predicate.merge_addn 0.71% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.66% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.81% : 0.000002s : 16: predicate.minmaximum_grad 1.01% : 0.000003s : 6: predicate.mutable_eliminate 0.39% : 0.000001s : 6: predicate.opt_reshape 0.47% : 0.000001s : 6: predicate.parallel_virtual_node 1.56% : 0.000004s : 22: predicate.partial_defer_inline 1.34% : 0.000004s : 24: predicate.partial_eliminate 0.94% : 0.000003s : 16: predicate.print_const_string_wrapper 0.68% : 0.000002s : 12: predicate.reduce_all_const_elim 1.13% : 0.000003s : 16: predicate.reduce_eliminate 2.38% : 0.000006s : 46: predicate.redundant_stop_gradient_eliminater 0.44% : 0.000001s : 12: predicate.remove_not_recompute_node 1.14% : 0.000003s : 30: predicate.replace_applicator 0.52% : 0.000001s : 12: predicate.replace_old_param 0.26% : 0.000001s : 6: predicate.reset_defer_inline 0.99% : 0.000003s : 16: predicate.reshape_eliminate 0.85% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.57% : 0.000002s : 6: predicate.row_tensor_eliminate 0.81% : 0.000002s : 12: predicate.same_eliminate 0.47% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.94% : 0.000003s : 12: predicate.shard_identity_eliminate 0.87% : 0.000002s : 12: predicate.special_op_eliminate 0.80% : 0.000002s : 12: predicate.specialize_transform 0.96% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 0.90% : 0.000002s : 12: predicate.stack_unstack_eliminate 0.44% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.31% : 0.000003s : 22: predicate.switch_defer_inline 2.01% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.61% : 0.000012s : 78: predicate.switch_simplify 0.89% : 0.000002s : 16: predicate.tile_eliminate 1.02% : 0.000003s : 16: predicate.transpose_eliminate 1.64% : 0.000004s : 28: predicate.tuple_list_convert_item_index_to_positive 1.62% : 0.000004s : 28: predicate.tuple_list_get_item_const_eliminator 1.47% : 0.000004s : 28: predicate.tuple_list_get_item_depend_reorder 2.87% : 0.000008s : 42: predicate.tuple_list_get_item_eliminator 1.59% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.19% : 0.000006s : 40: predicate.tuple_list_set_item_eliminator 1.60% : 0.000004s : 30: predicate.tuple_to_list_eliminator_ 2.60% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.30% : 0.000009s : 58: predicate.updatestate_useless_node_eliminater 0.54% : 0.000001s : 6: predicate.value_based_eliminate 0.81% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.77% : 0.000002s : 12: predicate.virtual_output_eliminate 0.41% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.54% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000528 11 47.39% : 0.000250s : 5: func_graph_cloner_run.FuncGraphClonerGraph 52.61% : 0.000278s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.184719 196 0.00% : 0.000004s : 1: ForceFp32Comm 48.65% : 0.576334s : 1: add_attr 48.65% : 0.576314s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000075s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000165s : 1: auto_monad 0.00% : 0.000035s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.06% : 0.000763s : 1: bootstrap 0.00% : 0.000035s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000015s : 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 0.00% : 0.000025s : 1: event_method 0.00% : 0.000014s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000005s : 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% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000007s : 1: label_micro_interleaved_index 0.04% : 0.000505s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.07% : 0.000773s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 0.12% : 0.001450s : 78: opt.transform.opt_a 0.00% : 0.000052s : 1: opt.transform.opt_after_cconv 0.00% : 0.000033s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000169s : 28: opt.transform.opt_b 0.01% : 0.000063s : 2: opt.transform.opt_trans_graph 0.00% : 0.000053s : 4: opt.transform.symbol_engine_opt 0.28% : 0.003358s : 1: opt_a 0.01% : 0.000142s : 1: opt_after_cconv 0.05% : 0.000581s : 1: opt_after_jit_grad 0.02% : 0.000291s : 1: opt_b 0.51% : 0.006095s : 1: optimize 0.00% : 0.000026s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 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% : 0.000009s : 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% : 0.000006s : 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 0.00% : 0.000050s : 1: pre_auto_parallel 0.00% : 0.000039s : 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% : 0.000021s : 1: remove_dup_value 0.05% : 0.000567s : 1: renormalize.infer 0.03% : 0.000404s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.02% : 0.000182s : 1: rewriter_after_opt_a 0.01% : 0.000097s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000097s : 1: symbol_engine_optimizer 0.67% : 0.007939s : 1: task_emit 0.01% : 0.000093s : 1: tuple_transform 0.62% : 0.007346s : 1: type_inference 0.01% : 0.000114s : 1: validate TotalTime = 0.66165, [24] [bootstrap]: 0.00059855 [type_inference]: 0.29367 [event_method]: 5.167e-05 [auto_monad]: 0.00033493 [graph_reusing]: 9.00999e-06 [inline]: 2.83003e-06 [add_attr]: 0.00398099, [1] [add_attr_with_inline]: 0.00395105, [1] [Cycle 1]: 8.617e-05, [2] [tag_attr]: 3.948e-05 [meta_addattr_fg_expand]: 8.75999e-06 [parallel-infer-symbol]: 3.93999e-06 [pre_auto_parallel]: 5.363e-05 [insert-virtual-dataset]: 2.88e-06 [parallel-infer-symbol-second]: 7.89994e-07 [dataset_repeat_opt]: 2.21998e-06 [pipeline_split]: 1.85001e-06 [optimize]: 0.351339, [53] [py_interpret_to_execute]: 3.901e-05 [rewriter_before_opt_a]: 0.00014273 [opt_a]: 0.348358, [3] [Cycle 1]: 0.211511, [45] [expand_dump_flag]: 4.3e-06 [switch_simplify]: 7.118e-05 [loop_unroll]: 5.57e-05 [a_1]: 0.0013386 [with_stream_mark]: 3.041e-05 [recompute_prepare]: 2.632e-05 [updatestate_depend_eliminate]: 1.216e-05 [updatestate_assign_eliminate]: 1.295e-05 [updatestate_loads_eliminate]: 9.81e-06 [parameter_eliminate]: 3.75998e-06 [a_2]: 0.0003091 [accelerated_algorithm]: 2e-05 [shard]: 2.54001e-06 [meta_shard_fg_expand]: 5.10999e-06 [shard_inline]: 1.808e-05 [merge_send_recv]: 2.19e-05 [auto_parallel]: 1.562e-05 [parallel]: 2.734e-05 [flash_sp]: 1.973e-05 [merge_comm]: 1.793e-05 [allreduce_fusion]: 1.196e-05 [matmul_add_comm_reduction]: 3.812e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 2.938e-05 [virtual_dataset]: 1.921e-05 [get_grad_eliminate_]: 1.754e-05 [virtual_output]: 1.811e-05 [merge_forward]: 1.202e-05 [cell_reuse_recompute_pass]: 2.31e-06 [offload_activation]: 2.378e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.093e-05 [merge_recompute_call_nodes]: 1.69e-06 [before_grad]: 3.397e-05 [set_forward_comm_id_for_comm_node_pass]: 1.235e-05 [meta_fg_expand]: 0.199387 [flash_sp_send_recv_attached]: 5.46998e-06 [receive_attached]: 2.50002e-06 [after_resolve]: 0.00012194 [a_after_grad]: 0.00017782 [renormalize]: 0.0074564 [add_forward_monad_depend]: 1.569e-05 [auto_monad_grad]: 8.64e-06 [auto_monad_eliminator]: 7.668e-05 [cse]: 0.00024191 [a_3]: 0.0008675 [Cycle 2]: 0.135926, [45] [expand_dump_flag]: 3.63e-06 [switch_simplify]: 6.028e-05 [loop_unroll]: 5.103e-05 [a_1]: 0.00126895 [with_stream_mark]: 2.384e-05 [recompute_prepare]: 1.302e-05 [updatestate_depend_eliminate]: 3.3e-05 [updatestate_assign_eliminate]: 6.31e-06 [updatestate_loads_eliminate]: 5.04e-06 [parameter_eliminate]: 1.96e-06 [a_2]: 0.00011794 [accelerated_algorithm]: 9.52001e-06 [shard]: 9.82999e-06 [meta_shard_fg_expand]: 1.458e-05 [shard_inline]: 4.944e-05 [merge_send_recv]: 2.356e-05 [auto_parallel]: 2.093e-05 [parallel]: 1.162e-05 [flash_sp]: 6.56999e-06 [merge_comm]: 5.87999e-06 [allreduce_fusion]: 5.20001e-06 [matmul_add_comm_reduction]: 2.547e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 1.441e-05 [virtual_dataset]: 1.099e-05 [get_grad_eliminate_]: 8.65999e-06 [virtual_output]: 8.63001e-06 [merge_forward]: 8.01001e-06 [cell_reuse_recompute_pass]: 3.23998e-06 [offload_activation]: 1.563e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.354e-05 [merge_recompute_call_nodes]: 1.72999e-06 [before_grad]: 1.643e-05 [set_forward_comm_id_for_comm_node_pass]: 6.23e-06 [meta_fg_expand]: 8.613e-05 [flash_sp_send_recv_attached]: 2.04999e-06 [receive_attached]: 2.59999e-06 [after_resolve]: 1.76e-05 [a_after_grad]: 1.391e-05 [renormalize]: 0.0012457 [add_forward_monad_depend]: 9.20999e-06 [auto_monad_grad]: 3.41001e-06 [auto_monad_eliminator]: 2.756e-05 [cse]: 4.715e-05 [a_3]: 6.877e-05 [Cycle 3]: 0.00089923, [45] [expand_dump_flag]: 2.63e-06 [switch_simplify]: 1.086e-05 [loop_unroll]: 9.24e-06 [a_1]: 0.00021253 [with_stream_mark]: 1.794e-05 [recompute_prepare]: 9.27001e-06 [updatestate_depend_eliminate]: 5.99e-06 [updatestate_assign_eliminate]: 5.51e-06 [updatestate_loads_eliminate]: 6.02001e-06 [parameter_eliminate]: 2.49001e-06 [a_2]: 0.00011652 [accelerated_algorithm]: 9.86003e-06 [shard]: 1.85001e-06 [meta_shard_fg_expand]: 2.39999e-06 [shard_inline]: 8.82e-06 [merge_send_recv]: 7.56999e-06 [auto_parallel]: 9.72001e-06 [parallel]: 6.63e-06 [flash_sp]: 1.09e-06 [merge_comm]: 4.98001e-06 [allreduce_fusion]: 4.85001e-06 [matmul_add_comm_reduction]: 9.25999e-06 [allreduce_slice_to_reducescatter]: 6.59988e-07 [virtual_shard_identity]: 9.62001e-06 [virtual_dataset]: 8.25e-06 [get_grad_eliminate_]: 8.15e-06 [virtual_output]: 7.95e-06 [merge_forward]: 5.48002e-06 [cell_reuse_recompute_pass]: 2.21e-06 [offload_activation]: 1.127e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.748e-05 [merge_recompute_call_nodes]: 8.60018e-07 [before_grad]: 1.426e-05 [set_forward_comm_id_for_comm_node_pass]: 5.83002e-06 [meta_fg_expand]: 3.24001e-06 [flash_sp_send_recv_attached]: 1.43002e-06 [receive_attached]: 1.50999e-06 [after_resolve]: 1.301e-05 [a_after_grad]: 1.419e-05 [renormalize]: 7.99773e-08 [add_forward_monad_depend]: 1.43002e-06 [auto_monad_grad]: 1.76e-06 [auto_monad_eliminator]: 1.368e-05 [cse]: 2.492e-05 [a_3]: 5.446e-05 [py_interpret_to_execute_after_opt_a]: 2.013e-05 [slice_cell_reuse_recomputed_activation]: 2.59001e-06 [rewriter_after_opt_a]: 0.00013758 [convert_after_rewriter]: 1.027e-05 [order_py_execute_after_rewriter]: 6.76e-06 [mutable_eliminate]: 0.00082903 [opt_b]: 0.00030249, [1] [Cycle 1]: 0.0002943, [7] [b_1]: 0.00019445 [b_2]: 1.059e-05 [updatestate_depend_eliminate]: 9.24e-06 [updatestate_assign_eliminate]: 4.77e-06 [updatestate_loads_eliminate]: 4.43999e-06 [renormalize]: 6.19999e-07 [cse]: 3.294e-05 [optimize_parallel_all_gather_comm]: 4.773e-05 [overlap_param_gather]: 2.26e-06 [cconv]: 2.96e-05 [loop_unroll]: 0.00049649 [opt_after_cconv]: 0.00016709, [1] [Cycle 1]: 0.00016117, [7] [c_1]: 7.2e-05 [parameter_eliminate]: 3.58999e-06 [updatestate_depend_eliminate]: 8.70999e-06 [updatestate_assign_eliminate]: 5.93002e-06 [updatestate_loads_eliminate]: 4.38001e-06 [cse]: 2.858e-05 [renormalize]: 6.19999e-07 [remove_dup_value]: 1.75e-05 [tuple_transform]: 9.239e-05, [1] [Cycle 1]: 8.713e-05, [4] [d_1]: 5.744e-05 [none_parameter_eliminate]: 1.97999e-06 [renormalize]: 1.79978e-07 [switch_simplify]: 9.05999e-06 [partial_unused_args_eliminate]: 2.16e-06 [add_recomputation]: 9.954e-05 [cse_after_recomputation]: 3.204e-05, [1] [Cycle 1]: 2.59e-05, [1] [cse]: 2.021e-05 [environ_conv]: 6.50002e-06 [swap_dp_allreduce_reducescatter]: 7.6e-06 [bias_add_comm_swap]: 3.31001e-06 [label_micro_interleaved_index]: 4.60001e-06 [label_fine_grained_interleaved_index]: 3.11001e-06 [merge_cast_opt]: 1.60999e-06 [slice_recompute_activation]: 2.32999e-06 [micro_interleaved_order_control]: 2.59999e-06 [assign_add_opt]: 1.35001e-06 [ForceFp32Comm]: 1.13001e-06 [remove_cast_before_assign_add]: 8.89995e-07 [full_micro_interleaved_order_control]: 3.04001e-06 [reorder_send_recv_between_fp_bp]: 3.38999e-06 [comm_op_add_attrs]: 1.09998e-06 [add_comm_op_reuse_tag]: 1.02998e-06 [interleave_split_concat_branches]: 1.44998e-06 [interleave_parallel_branches]: 1.12999e-06 [overlap_opt_shard_in_pipeline]: 1.57999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.75001e-06 [control_data_broadcast_order]: 1.804e-05 [grouped_pairwise_exchange_alltoall]: 1.72001e-06 [offloading_packed_experts]: 5.56e-06 [overlap_recompute_and_grad_model_parallel]: 6.31e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.36998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.50999e-06 [overlap_recompute_comm]: 2.73003e-06 [overlap_grad_ring_attention]: 5.42999e-06 [overlap_grad_flash_sp]: 6.728e-05 [begin_end_overlap_inline]: 5.70028e-07 [split_matmul_comm_elemetwise]: 2.32999e-06 [split_layernorm_comm]: 1.91998e-06 [handle_group_info]: 1.07998e-06 [symbol_engine_optimizer]: 9.613e-05, [1] [Cycle 1]: 9.117e-05, [6] [build]: 5.61998e-06 [elim_shapecalc]: 1.438e-05 [elim_not_effective]: 1.713e-05 [opt_reshape]: 9.09998e-06 [fold_const_symbol]: 1.416e-05 [renormalize]: 1.80007e-07 [detach_backward]: 2.16998e-06 [pipeline_parallel_scheduler]: 2.01e-06 [auto_monad_reorder]: 7.699e-05 [get_jit_bprop_graph]: 6.58e-06 [rewriter_after_jit_bprop_graph]: 5.45001e-06 [opt_after_jit_grad]: 0.00051637 [validate]: 5.408e-05 [backend_pass]: 8.60018e-07 [task_emit]: 0.0106805 [execute]: 8.57998e-06 Sums bootstrap : 0.000599s : 0.11% type_inference : 0.293670s : 56.10% event_method : 0.000052s : 0.01% auto_monad : 0.000335s : 0.06% graph_reusing : 0.000009s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000039s : 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 : 0.000054s : 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 : 0.000039s : 0.01% optimize.rewriter_before_opt_a : 0.000143s : 0.03% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000142s : 0.03% optimize.opt_a.loop_unroll : 0.000116s : 0.02% optimize.opt_a.a_1 : 0.002820s : 0.54% optimize.opt_a.with_stream_mark : 0.000072s : 0.01% optimize.opt_a.recompute_prepare : 0.000049s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000051s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000025s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000021s : 0.00% optimize.opt_a.parameter_eliminate : 0.000008s : 0.00% optimize.opt_a.a_2 : 0.000544s : 0.10% optimize.opt_a.accelerated_algorithm : 0.000039s : 0.01% optimize.opt_a.shard : 0.000014s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000022s : 0.00% optimize.opt_a.shard_inline : 0.000076s : 0.01% optimize.opt_a.merge_send_recv : 0.000053s : 0.01% optimize.opt_a.auto_parallel : 0.000046s : 0.01% optimize.opt_a.parallel : 0.000046s : 0.01% optimize.opt_a.flash_sp : 0.000027s : 0.01% optimize.opt_a.merge_comm : 0.000029s : 0.01% optimize.opt_a.allreduce_fusion : 0.000022s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000073s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000053s : 0.01% optimize.opt_a.virtual_dataset : 0.000038s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000034s : 0.01% optimize.opt_a.virtual_output : 0.000035s : 0.01% optimize.opt_a.merge_forward : 0.000026s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000008s : 0.00% optimize.opt_a.offload_activation : 0.000051s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000092s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000065s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000024s : 0.00% optimize.opt_a.meta_fg_expand : 0.199476s : 38.11% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000153s : 0.03% optimize.opt_a.a_after_grad : 0.000206s : 0.04% optimize.opt_a.renormalize : 0.008702s : 1.66% optimize.opt_a.add_forward_monad_depend : 0.000026s : 0.01% optimize.opt_a.auto_monad_grad : 0.000014s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000118s : 0.02% optimize.opt_a.cse : 0.000314s : 0.06% optimize.opt_a.a_3 : 0.000991s : 0.19% optimize.py_interpret_to_execute_after_opt_a : 0.000020s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000138s : 0.03% optimize.convert_after_rewriter : 0.000010s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000829s : 0.16% optimize.opt_b.b_1 : 0.000194s : 0.04% optimize.opt_b.b_2 : 0.000011s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% 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 : 0.000033s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000048s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000030s : 0.01% optimize.loop_unroll : 0.000496s : 0.09% optimize.opt_after_cconv.c_1 : 0.000072s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000029s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000017s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000100s : 0.02% optimize.cse_after_recomputation.cse : 0.000020s : 0.00% 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 : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000002s : 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 : 0.000006s : 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 : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000067s : 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 : 0.000006s : 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 : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000077s : 0.01% get_jit_bprop_graph : 0.000007s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000516s : 0.10% validate : 0.000054s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.010680s : 2.04% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.001659 224 1.00% : 0.000017s : 4: substitution.depend_value_elim 0.16% : 0.000003s : 5: substitution.elim_not_effective 0.18% : 0.000003s : 3: substitution.float_depend_g_call 0.30% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.13% : 0.000002s : 5: substitution.fold_const_symbol 51.56% : 0.000856s : 4: substitution.getattr_setattr_resolve 0.49% : 0.000008s : 6: substitution.graph_param_transform 0.20% : 0.000003s : 2: substitution.incorporate_call 0.15% : 0.000003s : 2: substitution.incorporate_call_switch 25.54% : 0.000424s : 13: substitution.inline 2.25% : 0.000037s : 5: substitution.inline_without_move 0.91% : 0.000015s : 23: substitution.j_node_and_user_rematch 0.72% : 0.000012s : 6: substitution.minmaximum_grad 0.21% : 0.000004s : 3: substitution.partial_eliminate 1.06% : 0.000018s : 23: substitution.remove_not_recompute_node 3.35% : 0.000056s : 21: substitution.replace_applicator 1.00% : 0.000017s : 16: substitution.replace_old_param 0.23% : 0.000004s : 1: substitution.set_cell_output_no_recompute 1.32% : 0.000022s : 6: substitution.tuple_list_convert_item_index_to_positive 0.59% : 0.000010s : 6: substitution.tuple_list_get_item_const_eliminator 0.80% : 0.000013s : 6: substitution.tuple_list_get_item_depend_reorder 3.00% : 0.000050s : 15: substitution.tuple_list_get_item_eliminator 0.79% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 1.68% : 0.000028s : 18: substitution.updatestate_pure_node_eliminater 2.36% : 0.000039s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.293534 2 99.50% : 0.292075s : 1: type_inference.infer 0.50% : 0.001458s : 1: type_inference.specialize ------[replace.] 0.000306 25 18.60% : 0.000057s : 3: replace.getattr_setattr_resolve 34.62% : 0.000106s : 13: replace.inline 13.98% : 0.000043s : 1: replace.replace_applicator 25.55% : 0.000078s : 7: replace.tuple_list_get_item_eliminator 7.25% : 0.000022s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001269 25 63.74% : 0.000809s : 3: match.getattr_setattr_resolve 32.83% : 0.000417s : 13: match.inline 0.83% : 0.000011s : 1: match.replace_applicator 1.78% : 0.000023s : 7: match.tuple_list_get_item_eliminator 0.81% : 0.000010s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000857 5424 0.89% : 0.000008s : 55: predicate.accumulaten_eliminater 0.27% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.48% : 0.000004s : 29: predicate.addn_check_dump 0.92% : 0.000008s : 55: predicate.addn_zero_filter 0.80% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 5.10% : 0.000044s : 84: predicate.arithmetic_simplify 0.92% : 0.000008s : 55: predicate.cast_eliminate 1.93% : 0.000017s : 130: predicate.check_bprop_eliminate 0.48% : 0.000004s : 29: predicate.compare_switch_simplify 0.06% : 0.000001s : 6: predicate.const_output_eliminate 0.51% : 0.000004s : 29: predicate.depend_value_elim 0.95% : 0.000008s : 55: predicate.dict_get_item_const_eliminator 0.97% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.91% : 0.000008s : 55: predicate.dict_set_item_eliminator 0.43% : 0.000004s : 12: predicate.dumpgradient_eliminate 0.07% : 0.000001s : 6: predicate.elim_not_effective 0.14% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 0.96% : 0.000008s : 61: predicate.environ_add_const_eliminate 1.00% : 0.000009s : 61: predicate.environ_get_add_eliminate 1.01% : 0.000009s : 61: predicate.environ_get_depend_swap 1.54% : 0.000013s : 90: predicate.environ_get_eliminate 0.92% : 0.000008s : 61: predicate.environ_get_set_eliminate 1.26% : 0.000011s : 76: predicate.exchange_switch_depend_value 1.73% : 0.000015s : 76: predicate.float_depend_g_call 0.46% : 0.000004s : 29: predicate.float_environ_get_switch 0.58% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.05% : 0.000000s : 6: predicate.fold_const_symbol 0.56% : 0.000005s : 29: predicate.get_grad_eliminate 0.81% : 0.000007s : 20: predicate.getattr_setattr_resolve 0.09% : 0.000001s : 6: predicate.graph_param_transform 0.46% : 0.000004s : 29: predicate.incorporate_call 0.41% : 0.000004s : 29: predicate.incorporate_call_switch 4.86% : 0.000042s : 201: predicate.inline 2.05% : 0.000018s : 97: predicate.inline_without_move 0.23% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.68% : 0.000006s : 29: predicate.less_batch_normalization 1.29% : 0.000011s : 74: predicate.list_to_tuple_eliminator_ 2.09% : 0.000018s : 129: predicate.load_eliminater 0.32% : 0.000003s : 6: predicate.loop_unroll_after_grad 2.00% : 0.000017s : 129: predicate.loop_unroll_before_grad 1.19% : 0.000010s : 67: predicate.make_slice_get_slice_eliminator 0.51% : 0.000004s : 29: predicate.merge_addn 1.98% : 0.000017s : 126: predicate.micro_step_allgather_replace 1.90% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.88% : 0.000008s : 55: predicate.minmaximum_grad 0.42% : 0.000004s : 6: predicate.mutable_eliminate 0.12% : 0.000001s : 6: predicate.opt_reshape 0.13% : 0.000001s : 6: predicate.parallel_virtual_node 1.50% : 0.000013s : 76: predicate.partial_defer_inline 1.20% : 0.000010s : 68: predicate.partial_eliminate 0.87% : 0.000007s : 55: predicate.print_const_string_wrapper 3.46% : 0.000030s : 29: predicate.reduce_all_const_elim 1.12% : 0.000010s : 55: predicate.reduce_eliminate 2.10% : 0.000018s : 129: predicate.redundant_stop_gradient_eliminater 0.35% : 0.000003s : 29: predicate.remove_not_recompute_node 2.09% : 0.000018s : 190: predicate.replace_applicator 0.94% : 0.000008s : 97: predicate.replace_old_param 0.09% : 0.000001s : 6: predicate.reset_defer_inline 0.96% : 0.000008s : 55: predicate.reshape_eliminate 1.91% : 0.000016s : 126: predicate.row_tensor_add_zeros_like 0.13% : 0.000001s : 6: predicate.row_tensor_eliminate 2.17% : 0.000019s : 130: predicate.same_eliminate 0.34% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.76% : 0.000006s : 29: predicate.shard_identity_eliminate 0.24% : 0.000002s : 12: predicate.special_op_eliminate 0.58% : 0.000005s : 29: predicate.specialize_transform 2.20% : 0.000019s : 126: predicate.split_environ_get_set_with_tuple_value 1.95% : 0.000017s : 97: predicate.stack_unstack_eliminate 0.13% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.34% : 0.000011s : 76: predicate.switch_defer_inline 3.24% : 0.000028s : 206: predicate.switch_layer_defer_inline 4.18% : 0.000036s : 240: predicate.switch_simplify 0.91% : 0.000008s : 55: predicate.tile_eliminate 0.89% : 0.000008s : 55: predicate.transpose_eliminate 1.19% : 0.000010s : 67: predicate.tuple_list_convert_item_index_to_positive 1.26% : 0.000011s : 67: predicate.tuple_list_get_item_const_eliminator 1.08% : 0.000009s : 67: predicate.tuple_list_get_item_depend_reorder 2.16% : 0.000018s : 103: predicate.tuple_list_get_item_eliminator 1.19% : 0.000010s : 67: predicate.tuple_list_get_set_item_eliminator 1.65% : 0.000014s : 96: predicate.tuple_list_set_item_eliminator 1.29% : 0.000011s : 74: predicate.tuple_to_list_eliminator_ 2.03% : 0.000017s : 129: predicate.updatestate_pure_node_eliminater 2.68% : 0.000023s : 159: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 6: predicate.value_based_eliminate 0.55% : 0.000005s : 29: predicate.virtual_dataset_eliminate 0.54% : 0.000005s : 29: predicate.virtual_output_eliminate 0.11% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.17% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002796 48 65.43% : 0.001830s : 24: func_graph_cloner_run.FuncGraphClonerGraph 34.57% : 0.000967s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 1.032214 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.39% : 0.003986s : 1: add_attr 0.38% : 0.003955s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000105s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000345s : 1: auto_monad 0.01% : 0.000081s : 1: auto_monad_reorder 0.00% : 0.000005s : 1: backend_pass 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.06% : 0.000633s : 1: bootstrap 0.00% : 0.000033s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000014s : 1: convert_after_rewriter 0.00% : 0.000035s : 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% : 0.000060s : 1: event_method 0.00% : 0.000015s : 1: execute 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000010s : 1: get_jit_bprop_graph 0.00% : 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% : 0.000006s : 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 0.05% : 0.000505s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.08% : 0.000838s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000019s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000023s : 1: opt.transform.mutable_eliminate 0.51% : 0.005284s : 125: opt.transform.opt_a 0.01% : 0.000071s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000175s : 28: opt.transform.opt_b 0.09% : 0.000967s : 2: opt.transform.opt_resolve 0.01% : 0.000064s : 2: opt.transform.opt_trans_graph 0.00% : 0.000051s : 4: opt.transform.symbol_engine_opt 33.75% : 0.348361s : 1: opt_a 0.02% : 0.000171s : 1: opt_after_cconv 0.05% : 0.000525s : 1: opt_after_jit_grad 0.03% : 0.000306s : 1: opt_b 34.04% : 0.351345s : 1: optimize 0.01% : 0.000052s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 1: order_py_execute_after_rewriter 0.01% : 0.000072s : 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% : 0.000009s : 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 0.01% : 0.000058s : 1: pre_auto_parallel 0.00% : 0.000043s : 1: py_interpret_to_execute 0.00% : 0.000024s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000021s : 1: remove_dup_value 0.59% : 0.006100s : 2: renormalize.infer 0.25% : 0.002578s : 2: 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% : 0.000143s : 1: rewriter_after_opt_a 0.01% : 0.000147s : 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% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000099s : 1: symbol_engine_optimizer 1.04% : 0.010700s : 1: task_emit 0.01% : 0.000095s : 1: tuple_transform 28.45% : 0.293695s : 1: type_inference 0.01% : 0.000083s : 1: validate TotalTime = 1.13874, [24] [bootstrap]: 0.000447 [type_inference]: 0.287259 [event_method]: 5.34e-05 [auto_monad]: 0.00024116 [graph_reusing]: 2.842e-05 [inline]: 6.09999e-06 [add_attr]: 0.00370836, [1] [add_attr_with_inline]: 0.00369968, [1] [Cycle 1]: 7.624e-05, [2] [tag_attr]: 3.64e-05 [meta_addattr_fg_expand]: 9.00999e-06 [parallel-infer-symbol]: 3.43999e-06 [pre_auto_parallel]: 4.962e-05 [insert-virtual-dataset]: 2.69999e-06 [parallel-infer-symbol-second]: 9.20001e-07 [dataset_repeat_opt]: 2.15002e-06 [pipeline_split]: 1.64e-06 [optimize]: 0.837207, [53] [py_interpret_to_execute]: 3.804e-05 [rewriter_before_opt_a]: 0.0001316 [opt_a]: 0.556119, [3] [Cycle 1]: 0.551708, [45] [expand_dump_flag]: 3.76999e-06 [switch_simplify]: 7.084e-05 [loop_unroll]: 5.911e-05 [a_1]: 0.00135351 [with_stream_mark]: 2.88e-05 [recompute_prepare]: 2.449e-05 [updatestate_depend_eliminate]: 1.188e-05 [updatestate_assign_eliminate]: 1.024e-05 [updatestate_loads_eliminate]: 9.42999e-06 [parameter_eliminate]: 2.99999e-06 [a_2]: 0.00027839 [accelerated_algorithm]: 1.797e-05 [shard]: 1.99e-06 [meta_shard_fg_expand]: 4.57e-06 [shard_inline]: 1.76e-05 [merge_send_recv]: 3.002e-05 [auto_parallel]: 1.375e-05 [parallel]: 2.728e-05 [flash_sp]: 1.155e-05 [merge_comm]: 1.153e-05 [allreduce_fusion]: 1.071e-05 [matmul_add_comm_reduction]: 3.119e-05 [allreduce_slice_to_reducescatter]: 9.09989e-07 [virtual_shard_identity]: 2.01e-05 [virtual_dataset]: 1.793e-05 [get_grad_eliminate_]: 1.733e-05 [virtual_output]: 1.721e-05 [merge_forward]: 1.159e-05 [cell_reuse_recompute_pass]: 1.35999e-06 [offload_activation]: 2.205e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.565e-05 [merge_recompute_call_nodes]: 1.46002e-06 [before_grad]: 0.203618 [set_forward_comm_id_for_comm_node_pass]: 3.202e-05 [meta_fg_expand]: 0.337424 [flash_sp_send_recv_attached]: 5.62999e-06 [receive_attached]: 2.79001e-06 [after_resolve]: 0.00013875 [a_after_grad]: 0.00018744 [renormalize]: 0.00644859 [add_forward_monad_depend]: 2.002e-05 [auto_monad_grad]: 9.72001e-06 [auto_monad_eliminator]: 7.967e-05 [cse]: 0.00023929 [a_3]: 0.00089334 [Cycle 2]: 0.00350049, [45] [expand_dump_flag]: 3.51001e-06 [switch_simplify]: 5.669e-05 [loop_unroll]: 5.094e-05 [a_1]: 0.00131301 [with_stream_mark]: 2.872e-05 [recompute_prepare]: 1.45e-05 [updatestate_depend_eliminate]: 3.697e-05 [updatestate_assign_eliminate]: 6.02001e-06 [updatestate_loads_eliminate]: 6.71e-06 [parameter_eliminate]: 3.08e-06 [a_2]: 0.00012542 [accelerated_algorithm]: 9.62001e-06 [shard]: 2.68e-06 [meta_shard_fg_expand]: 3.49001e-06 [shard_inline]: 8.54998e-06 [merge_send_recv]: 1.12e-05 [auto_parallel]: 1.332e-05 [parallel]: 1.07e-05 [flash_sp]: 4.87e-06 [merge_comm]: 5.80002e-06 [allreduce_fusion]: 5.05999e-06 [matmul_add_comm_reduction]: 1.31e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 1.042e-05 [virtual_dataset]: 8.18999e-06 [get_grad_eliminate_]: 7.77e-06 [virtual_output]: 8.05e-06 [merge_forward]: 5.99e-06 [cell_reuse_recompute_pass]: 1.81e-06 [offload_activation]: 1.369e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.77e-05 [merge_recompute_call_nodes]: 1.82999e-06 [before_grad]: 1.429e-05 [set_forward_comm_id_for_comm_node_pass]: 5.58997e-06 [meta_fg_expand]: 6.559e-05 [flash_sp_send_recv_attached]: 2.10002e-06 [receive_attached]: 3.05002e-06 [after_resolve]: 1.426e-05 [a_after_grad]: 1.381e-05 [renormalize]: 0.00115021 [add_forward_monad_depend]: 6.01e-06 [auto_monad_grad]: 2.65002e-06 [auto_monad_eliminator]: 2.452e-05 [cse]: 3.674e-05 [a_3]: 6.66e-05 [Cycle 3]: 0.00089057, [45] [expand_dump_flag]: 1.89999e-06 [switch_simplify]: 1.019e-05 [loop_unroll]: 8.18001e-06 [a_1]: 0.00022733 [with_stream_mark]: 1.503e-05 [recompute_prepare]: 9.24e-06 [updatestate_depend_eliminate]: 5.69999e-06 [updatestate_assign_eliminate]: 5.24998e-06 [updatestate_loads_eliminate]: 4.57998e-06 [parameter_eliminate]: 9.70002e-07 [a_2]: 0.00011553 [accelerated_algorithm]: 9.42001e-06 [shard]: 1.60999e-06 [meta_shard_fg_expand]: 2.26998e-06 [shard_inline]: 8.36002e-06 [merge_send_recv]: 8.73001e-06 [auto_parallel]: 8.92999e-06 [parallel]: 6.41998e-06 [flash_sp]: 1.34998e-06 [merge_comm]: 5.27999e-06 [allreduce_fusion]: 4.91002e-06 [matmul_add_comm_reduction]: 9.19e-06 [allreduce_slice_to_reducescatter]: 7.30011e-07 [virtual_shard_identity]: 9.37001e-06 [virtual_dataset]: 8.17998e-06 [get_grad_eliminate_]: 7.88999e-06 [virtual_output]: 7.99002e-06 [merge_forward]: 4.67998e-06 [cell_reuse_recompute_pass]: 2.16998e-06 [offload_activation]: 1.165e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.657e-05 [merge_recompute_call_nodes]: 9.80013e-07 [before_grad]: 1.416e-05 [set_forward_comm_id_for_comm_node_pass]: 6.04999e-06 [meta_fg_expand]: 3.42002e-06 [flash_sp_send_recv_attached]: 1.04e-06 [receive_attached]: 1.92999e-06 [after_resolve]: 1.157e-05 [a_after_grad]: 1.239e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 1.72001e-06 [auto_monad_grad]: 1.22e-06 [auto_monad_eliminator]: 1.341e-05 [cse]: 2.625e-05 [a_3]: 5.109e-05 [py_interpret_to_execute_after_opt_a]: 1.859e-05 [slice_cell_reuse_recomputed_activation]: 2.09e-06 [rewriter_after_opt_a]: 6.301e-05 [convert_after_rewriter]: 8.71997e-06 [order_py_execute_after_rewriter]: 6.17999e-06 [mutable_eliminate]: 0.00082454 [opt_b]: 0.00029831, [1] [Cycle 1]: 0.00029057, [7] [b_1]: 0.00019069 [b_2]: 1.17e-05 [updatestate_depend_eliminate]: 9.46e-06 [updatestate_assign_eliminate]: 4.76002e-06 [updatestate_loads_eliminate]: 4.23999e-06 [renormalize]: 5.20027e-07 [cse]: 3.105e-05 [optimize_parallel_all_gather_comm]: 2.293e-05 [overlap_param_gather]: 2.66999e-06 [cconv]: 2.948e-05 [loop_unroll]: 0.278673 [opt_after_cconv]: 0.00025045, [1] [Cycle 1]: 0.00020584, [7] [c_1]: 7.292e-05 [parameter_eliminate]: 7.06001e-06 [updatestate_depend_eliminate]: 1.572e-05 [updatestate_assign_eliminate]: 6.45002e-06 [updatestate_loads_eliminate]: 4.90999e-06 [cse]: 5.366e-05 [renormalize]: 7.09988e-07 [remove_dup_value]: 1.843e-05 [tuple_transform]: 9.982e-05, [1] [Cycle 1]: 9.415e-05, [4] [d_1]: 6.196e-05 [none_parameter_eliminate]: 1.82001e-06 [renormalize]: 1.8999e-07 [switch_simplify]: 1.084e-05 [partial_unused_args_eliminate]: 2.64999e-06 [add_recomputation]: 8.113e-05 [cse_after_recomputation]: 3.147e-05, [1] [Cycle 1]: 2.642e-05, [1] [cse]: 2.088e-05 [environ_conv]: 9.47999e-06 [swap_dp_allreduce_reducescatter]: 7.73999e-06 [bias_add_comm_swap]: 3.85e-06 [label_micro_interleaved_index]: 7.33e-06 [label_fine_grained_interleaved_index]: 2.74001e-06 [merge_cast_opt]: 1.41002e-06 [slice_recompute_activation]: 2.36998e-06 [micro_interleaved_order_control]: 2.24001e-06 [assign_add_opt]: 1.50001e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 8.09989e-07 [full_micro_interleaved_order_control]: 2.49999e-06 [reorder_send_recv_between_fp_bp]: 2.81e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 1.07998e-06 [interleave_split_concat_branches]: 1.20999e-06 [interleave_parallel_branches]: 1.12999e-06 [overlap_opt_shard_in_pipeline]: 1.22e-06 [overlap_opt_shard_grad_in_pipeline]: 2.02001e-06 [control_data_broadcast_order]: 2.089e-05 [grouped_pairwise_exchange_alltoall]: 1.77999e-06 [offloading_packed_experts]: 5.67001e-06 [overlap_recompute_and_grad_model_parallel]: 6.31e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.20001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40001e-06 [overlap_recompute_comm]: 2.51998e-06 [overlap_grad_ring_attention]: 6.24001e-06 [overlap_grad_flash_sp]: 2.963e-05 [begin_end_overlap_inline]: 5.00004e-07 [split_matmul_comm_elemetwise]: 2.11998e-06 [split_layernorm_comm]: 1.78002e-06 [handle_group_info]: 9.50007e-07 [symbol_engine_optimizer]: 9.663e-05, [1] [Cycle 1]: 9.152e-05, [6] [build]: 5.54e-06 [elim_shapecalc]: 1.243e-05 [elim_not_effective]: 1.814e-05 [opt_reshape]: 1.028e-05 [fold_const_symbol]: 1.464e-05 [renormalize]: 3.50003e-07 [detach_backward]: 2.61999e-06 [pipeline_parallel_scheduler]: 1.79e-06 [auto_monad_reorder]: 9.003e-05 [get_jit_bprop_graph]: 1.75001e-06 [rewriter_after_jit_bprop_graph]: 7.07997e-06 [opt_after_jit_grad]: 0.00056906 [validate]: 6.279e-05 [backend_pass]: 8.60018e-07 [task_emit]: 0.00871342 [execute]: 8.88002e-06 Sums bootstrap : 0.000447s : 0.04% type_inference : 0.287259s : 25.34% event_method : 0.000053s : 0.00% auto_monad : 0.000241s : 0.02% graph_reusing : 0.000028s : 0.00% inline : 0.000006s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000036s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000009s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000050s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000038s : 0.00% optimize.rewriter_before_opt_a : 0.000132s : 0.01% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000138s : 0.01% optimize.opt_a.loop_unroll : 0.000118s : 0.01% optimize.opt_a.a_1 : 0.002894s : 0.26% optimize.opt_a.with_stream_mark : 0.000073s : 0.01% optimize.opt_a.recompute_prepare : 0.000048s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000055s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000022s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000021s : 0.00% optimize.opt_a.parameter_eliminate : 0.000007s : 0.00% optimize.opt_a.a_2 : 0.000519s : 0.05% optimize.opt_a.accelerated_algorithm : 0.000037s : 0.00% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000035s : 0.00% optimize.opt_a.merge_send_recv : 0.000050s : 0.00% optimize.opt_a.auto_parallel : 0.000036s : 0.00% optimize.opt_a.parallel : 0.000044s : 0.00% optimize.opt_a.flash_sp : 0.000018s : 0.00% optimize.opt_a.merge_comm : 0.000023s : 0.00% optimize.opt_a.allreduce_fusion : 0.000021s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000053s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000040s : 0.00% optimize.opt_a.virtual_dataset : 0.000034s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000033s : 0.00% optimize.opt_a.virtual_output : 0.000033s : 0.00% optimize.opt_a.merge_forward : 0.000022s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000047s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000070s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.203647s : 17.97% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000044s : 0.00% optimize.opt_a.meta_fg_expand : 0.337493s : 29.78% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000008s : 0.00% optimize.opt_a.after_resolve : 0.000165s : 0.01% optimize.opt_a.a_after_grad : 0.000214s : 0.02% optimize.opt_a.renormalize : 0.007599s : 0.67% optimize.opt_a.add_forward_monad_depend : 0.000028s : 0.00% optimize.opt_a.auto_monad_grad : 0.000014s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000118s : 0.01% optimize.opt_a.cse : 0.000302s : 0.03% optimize.opt_a.a_3 : 0.001011s : 0.09% optimize.py_interpret_to_execute_after_opt_a : 0.000019s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000063s : 0.01% optimize.convert_after_rewriter : 0.000009s : 0.00% optimize.order_py_execute_after_rewriter : 0.000006s : 0.00% optimize.mutable_eliminate : 0.000825s : 0.07% optimize.opt_b.b_1 : 0.000191s : 0.02% optimize.opt_b.b_2 : 0.000012s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% 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 : 0.000031s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000023s : 0.00% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000029s : 0.00% optimize.loop_unroll : 0.278673s : 24.59% optimize.opt_after_cconv.c_1 : 0.000073s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000016s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.cse : 0.000054s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000018s : 0.00% optimize.tuple_transform.d_1 : 0.000062s : 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 : 0.000011s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000081s : 0.01% optimize.cse_after_recomputation.cse : 0.000021s : 0.00% optimize.environ_conv : 0.000009s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000008s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000007s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 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 : 0.000021s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 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 : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000030s : 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 : 0.000006s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000012s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 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 : 0.000090s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000569s : 0.05% validate : 0.000063s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.008713s : 0.77% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.012512 224 0.11% : 0.000014s : 4: substitution.depend_value_elim 0.02% : 0.000003s : 5: substitution.elim_not_effective 0.02% : 0.000003s : 3: substitution.float_depend_g_call 0.03% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.02% : 0.000002s : 5: substitution.fold_const_symbol 93.14% : 0.011654s : 4: substitution.getattr_setattr_resolve 0.07% : 0.000008s : 6: substitution.graph_param_transform 0.03% : 0.000003s : 2: substitution.incorporate_call 0.02% : 0.000002s : 2: substitution.incorporate_call_switch 3.67% : 0.000459s : 13: substitution.inline 0.32% : 0.000040s : 5: substitution.inline_without_move 0.15% : 0.000019s : 23: substitution.j_node_and_user_rematch 0.10% : 0.000012s : 6: substitution.minmaximum_grad 0.04% : 0.000004s : 3: substitution.partial_eliminate 0.13% : 0.000017s : 23: substitution.remove_not_recompute_node 0.48% : 0.000061s : 21: substitution.replace_applicator 0.13% : 0.000016s : 16: substitution.replace_old_param 0.02% : 0.000003s : 1: substitution.set_cell_output_no_recompute 0.18% : 0.000022s : 6: substitution.tuple_list_convert_item_index_to_positive 0.07% : 0.000009s : 6: substitution.tuple_list_get_item_const_eliminator 0.11% : 0.000014s : 6: substitution.tuple_list_get_item_depend_reorder 0.40% : 0.000050s : 15: substitution.tuple_list_get_item_eliminator 0.10% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 0.27% : 0.000034s : 18: substitution.updatestate_pure_node_eliminater 0.36% : 0.000045s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.287168 2 99.43% : 0.285525s : 1: type_inference.infer 0.57% : 0.001643s : 1: type_inference.specialize ------[replace.] 0.000341 25 21.81% : 0.000074s : 3: replace.getattr_setattr_resolve 32.63% : 0.000111s : 13: replace.inline 15.35% : 0.000052s : 1: replace.replace_applicator 23.02% : 0.000078s : 7: replace.tuple_list_get_item_eliminator 7.19% : 0.000025s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.012085 25 95.89% : 0.011588s : 3: match.getattr_setattr_resolve 3.73% : 0.000451s : 13: match.inline 0.10% : 0.000012s : 1: match.replace_applicator 0.19% : 0.000023s : 7: match.tuple_list_get_item_eliminator 0.09% : 0.000011s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000807 5424 0.93% : 0.000008s : 55: predicate.accumulaten_eliminater 0.30% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.48% : 0.000004s : 29: predicate.addn_check_dump 1.00% : 0.000008s : 55: predicate.addn_zero_filter 0.88% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 1.93% : 0.000016s : 84: predicate.arithmetic_simplify 0.98% : 0.000008s : 55: predicate.cast_eliminate 2.06% : 0.000017s : 130: predicate.check_bprop_eliminate 0.47% : 0.000004s : 29: predicate.compare_switch_simplify 0.06% : 0.000000s : 6: predicate.const_output_eliminate 0.51% : 0.000004s : 29: predicate.depend_value_elim 0.97% : 0.000008s : 55: predicate.dict_get_item_const_eliminator 1.05% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.96% : 0.000008s : 55: predicate.dict_set_item_eliminator 0.32% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.08% : 0.000001s : 6: predicate.elim_not_effective 0.15% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.07% : 0.000009s : 61: predicate.environ_add_const_eliminate 1.06% : 0.000009s : 61: predicate.environ_get_add_eliminate 1.05% : 0.000009s : 61: predicate.environ_get_depend_swap 1.50% : 0.000012s : 90: predicate.environ_get_eliminate 1.05% : 0.000009s : 61: predicate.environ_get_set_eliminate 1.31% : 0.000011s : 76: predicate.exchange_switch_depend_value 1.89% : 0.000015s : 76: predicate.float_depend_g_call 0.47% : 0.000004s : 29: predicate.float_environ_get_switch 0.60% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.07% : 0.000001s : 6: predicate.fold_const_symbol 0.56% : 0.000004s : 29: predicate.get_grad_eliminate 1.10% : 0.000009s : 20: predicate.getattr_setattr_resolve 0.06% : 0.000000s : 6: predicate.graph_param_transform 0.53% : 0.000004s : 29: predicate.incorporate_call 0.44% : 0.000004s : 29: predicate.incorporate_call_switch 4.93% : 0.000040s : 201: predicate.inline 2.29% : 0.000018s : 97: predicate.inline_without_move 0.40% : 0.000003s : 29: predicate.j_node_and_user_rematch 0.68% : 0.000005s : 29: predicate.less_batch_normalization 1.38% : 0.000011s : 74: predicate.list_to_tuple_eliminator_ 2.18% : 0.000018s : 129: predicate.load_eliminater 0.97% : 0.000008s : 6: predicate.loop_unroll_after_grad 2.19% : 0.000018s : 129: predicate.loop_unroll_before_grad 1.22% : 0.000010s : 67: predicate.make_slice_get_slice_eliminator 0.52% : 0.000004s : 29: predicate.merge_addn 2.02% : 0.000016s : 126: predicate.micro_step_allgather_replace 2.05% : 0.000017s : 126: predicate.mini_step_allgather_replace 0.93% : 0.000008s : 55: predicate.minmaximum_grad 0.47% : 0.000004s : 6: predicate.mutable_eliminate 0.13% : 0.000001s : 6: predicate.opt_reshape 0.14% : 0.000001s : 6: predicate.parallel_virtual_node 1.73% : 0.000014s : 76: predicate.partial_defer_inline 1.28% : 0.000010s : 68: predicate.partial_eliminate 0.90% : 0.000007s : 55: predicate.print_const_string_wrapper 0.53% : 0.000004s : 29: predicate.reduce_all_const_elim 1.24% : 0.000010s : 55: predicate.reduce_eliminate 2.12% : 0.000017s : 129: predicate.redundant_stop_gradient_eliminater 0.34% : 0.000003s : 29: predicate.remove_not_recompute_node 2.16% : 0.000017s : 190: predicate.replace_applicator 1.03% : 0.000008s : 97: predicate.replace_old_param 0.09% : 0.000001s : 6: predicate.reset_defer_inline 1.02% : 0.000008s : 55: predicate.reshape_eliminate 2.14% : 0.000017s : 126: predicate.row_tensor_add_zeros_like 0.19% : 0.000002s : 6: predicate.row_tensor_eliminate 2.30% : 0.000019s : 130: predicate.same_eliminate 0.33% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.58% : 0.000005s : 29: predicate.shard_identity_eliminate 0.26% : 0.000002s : 12: predicate.special_op_eliminate 0.61% : 0.000005s : 29: predicate.specialize_transform 2.25% : 0.000018s : 126: predicate.split_environ_get_set_with_tuple_value 2.10% : 0.000017s : 97: predicate.stack_unstack_eliminate 0.15% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.40% : 0.000011s : 76: predicate.switch_defer_inline 3.40% : 0.000027s : 206: predicate.switch_layer_defer_inline 4.42% : 0.000036s : 240: predicate.switch_simplify 1.01% : 0.000008s : 55: predicate.tile_eliminate 0.95% : 0.000008s : 55: predicate.transpose_eliminate 1.30% : 0.000010s : 67: predicate.tuple_list_convert_item_index_to_positive 1.27% : 0.000010s : 67: predicate.tuple_list_get_item_const_eliminator 1.18% : 0.000010s : 67: predicate.tuple_list_get_item_depend_reorder 2.32% : 0.000019s : 103: predicate.tuple_list_get_item_eliminator 1.19% : 0.000010s : 67: predicate.tuple_list_get_set_item_eliminator 1.76% : 0.000014s : 96: predicate.tuple_list_set_item_eliminator 1.37% : 0.000011s : 74: predicate.tuple_to_list_eliminator_ 2.18% : 0.000018s : 129: predicate.updatestate_pure_node_eliminater 3.04% : 0.000024s : 159: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 6: predicate.value_based_eliminate 0.54% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.53% : 0.000004s : 29: predicate.virtual_output_eliminate 0.13% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.18% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.274407 48 99.62% : 0.273369s : 24: func_graph_cloner_run.FuncGraphClonerGraph 0.38% : 0.001038s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.208274 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.17% : 0.003714s : 1: add_attr 0.17% : 0.003703s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000085s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000250s : 1: auto_monad 0.00% : 0.000095s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.02% : 0.000480s : 1: bootstrap 0.00% : 0.000033s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000024s : 1: control_data_broadcast_order 0.00% : 0.000012s : 1: convert_after_rewriter 0.00% : 0.000034s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.00% : 0.000013s : 1: environ_conv 0.00% : 0.000062s : 1: event_method 0.00% : 0.000015s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000036s : 1: graph_reusing 0.00% : 0.000006s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000010s : 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% : 0.000010s : 1: label_micro_interleaved_index 12.62% : 0.278699s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.04% : 0.000835s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000053s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000024s : 1: opt.transform.mutable_eliminate 9.46% : 0.208872s : 125: opt.transform.opt_a 0.00% : 0.000071s : 1: opt.transform.opt_after_cconv 0.00% : 0.000033s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000174s : 28: opt.transform.opt_b 0.53% : 0.011799s : 2: opt.transform.opt_resolve 0.00% : 0.000070s : 2: opt.transform.opt_trans_graph 0.00% : 0.000052s : 4: opt.transform.symbol_engine_opt 25.18% : 0.556122s : 1: opt_a 0.01% : 0.000256s : 1: opt_after_cconv 0.03% : 0.000579s : 1: opt_after_jit_grad 0.01% : 0.000302s : 1: opt_b 37.91% : 0.837212s : 1: optimize 0.00% : 0.000026s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000009s : 1: order_py_execute_after_rewriter 0.00% : 0.000033s : 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% : 0.000006s : 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% : 0.000007s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000006s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.00% : 0.000054s : 1: pre_auto_parallel 0.00% : 0.000042s : 1: py_interpret_to_execute 0.00% : 0.000022s : 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 0.21% : 0.004700s : 2: renormalize.infer 0.13% : 0.002873s : 2: 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% : 0.000067s : 1: rewriter_after_opt_a 0.01% : 0.000136s : 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% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000099s : 1: symbol_engine_optimizer 0.40% : 0.008731s : 1: task_emit 0.00% : 0.000103s : 1: tuple_transform 13.01% : 0.287302s : 1: type_inference 0.00% : 0.000095s : 1: validate TotalTime = 0.218179, [24] [bootstrap]: 0.00043967 [type_inference]: 0.006763 [event_method]: 2.233e-05 [auto_monad]: 0.00016842 [graph_reusing]: 7.21001e-06 [inline]: 2.91999e-06 [add_attr]: 0.00378767, [1] [add_attr_with_inline]: 0.0037747, [1] [Cycle 1]: 7.349e-05, [2] [tag_attr]: 2.482e-05 [meta_addattr_fg_expand]: 5.44e-06 [parallel-infer-symbol]: 4.01001e-06 [pre_auto_parallel]: 4.255e-05 [insert-virtual-dataset]: 3.01001e-06 [parallel-infer-symbol-second]: 7.2e-07 [dataset_repeat_opt]: 2.21998e-06 [pipeline_split]: 1.84e-06 [optimize]: 0.197022, [53] [py_interpret_to_execute]: 3.124e-05 [rewriter_before_opt_a]: 8.561e-05 [opt_a]: 0.194115, [2] [Cycle 1]: 0.193275, [45] [expand_dump_flag]: 3.96001e-06 [switch_simplify]: 4.382e-05 [loop_unroll]: 2.979e-05 [a_1]: 0.191325 [with_stream_mark]: 3.203e-05 [recompute_prepare]: 1.597e-05 [updatestate_depend_eliminate]: 7.54002e-06 [updatestate_assign_eliminate]: 6.59001e-06 [updatestate_loads_eliminate]: 4.97e-06 [parameter_eliminate]: 2.61999e-06 [a_2]: 0.00012378 [accelerated_algorithm]: 9.59999e-06 [shard]: 2.90002e-06 [meta_shard_fg_expand]: 3.25998e-06 [shard_inline]: 8.38001e-06 [merge_send_recv]: 1.096e-05 [auto_parallel]: 1.303e-05 [parallel]: 2.249e-05 [flash_sp]: 1.234e-05 [merge_comm]: 5.39e-06 [allreduce_fusion]: 4.96997e-06 [matmul_add_comm_reduction]: 1.323e-05 [allreduce_slice_to_reducescatter]: 1.49e-06 [virtual_shard_identity]: 1.039e-05 [virtual_dataset]: 8.57e-06 [get_grad_eliminate_]: 8.25e-06 [virtual_output]: 8.57e-06 [merge_forward]: 6.24001e-06 [cell_reuse_recompute_pass]: 1.28002e-06 [offload_activation]: 1.525e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.771e-05 [merge_recompute_call_nodes]: 1.70001e-06 [before_grad]: 1.601e-05 [set_forward_comm_id_for_comm_node_pass]: 5.37001e-06 [meta_fg_expand]: 4.19002e-06 [flash_sp_send_recv_attached]: 2.68e-06 [receive_attached]: 2.82002e-06 [after_resolve]: 1.242e-05 [a_after_grad]: 1.351e-05 [renormalize]: 0.00101386 [add_forward_monad_depend]: 5.55001e-06 [auto_monad_grad]: 2.43e-06 [auto_monad_eliminator]: 2.289e-05 [cse]: 4.141e-05 [a_3]: 6.548e-05 [Cycle 2]: 0.000826, [45] [expand_dump_flag]: 2.37999e-06 [switch_simplify]: 9.59e-06 [loop_unroll]: 8.73001e-06 [a_1]: 0.00019892 [with_stream_mark]: 1.37e-05 [recompute_prepare]: 8.23999e-06 [updatestate_depend_eliminate]: 5.09998e-06 [updatestate_assign_eliminate]: 4.74e-06 [updatestate_loads_eliminate]: 4.29997e-06 [parameter_eliminate]: 1.07e-06 [a_2]: 0.00011005 [accelerated_algorithm]: 8.55999e-06 [shard]: 1.64e-06 [meta_shard_fg_expand]: 2.19001e-06 [shard_inline]: 8.27e-06 [merge_send_recv]: 6.86001e-06 [auto_parallel]: 8.12003e-06 [parallel]: 5.60001e-06 [flash_sp]: 3.35003e-06 [merge_comm]: 4.74e-06 [allreduce_fusion]: 4.90999e-06 [matmul_add_comm_reduction]: 7.7e-06 [allreduce_slice_to_reducescatter]: 3.19997e-07 [virtual_shard_identity]: 9.76e-06 [virtual_dataset]: 8.12998e-06 [get_grad_eliminate_]: 7.97e-06 [virtual_output]: 7.5e-06 [merge_forward]: 4.18999e-06 [cell_reuse_recompute_pass]: 1.39998e-06 [offload_activation]: 9.14998e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.625e-05 [merge_recompute_call_nodes]: 7.09988e-07 [before_grad]: 1.395e-05 [set_forward_comm_id_for_comm_node_pass]: 5.10999e-06 [meta_fg_expand]: 3.65998e-06 [flash_sp_send_recv_attached]: 9.00007e-07 [receive_attached]: 1.40001e-06 [after_resolve]: 1.09e-05 [a_after_grad]: 1.213e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.21002e-06 [auto_monad_grad]: 1.02e-06 [auto_monad_eliminator]: 1.212e-05 [cse]: 2.048e-05 [a_3]: 5.045e-05 [py_interpret_to_execute_after_opt_a]: 1.534e-05 [slice_cell_reuse_recomputed_activation]: 2.22001e-06 [rewriter_after_opt_a]: 0.00021351 [convert_after_rewriter]: 1.076e-05 [order_py_execute_after_rewriter]: 7.11001e-06 [mutable_eliminate]: 0.00080212 [opt_b]: 0.00033124, [1] [Cycle 1]: 0.00032382, [7] [b_1]: 0.00022714 [b_2]: 1.091e-05 [updatestate_depend_eliminate]: 9.99001e-06 [updatestate_assign_eliminate]: 4.3e-06 [updatestate_loads_eliminate]: 4.23999e-06 [renormalize]: 5.10016e-07 [cse]: 2.744e-05 [optimize_parallel_all_gather_comm]: 2.198e-05 [overlap_param_gather]: 2.16e-06 [cconv]: 2.892e-05 [loop_unroll]: 0.00049722 [opt_after_cconv]: 0.00013931, [1] [Cycle 1]: 0.00013299, [7] [c_1]: 5.353e-05 [parameter_eliminate]: 3.2e-06 [updatestate_depend_eliminate]: 8.83001e-06 [updatestate_assign_eliminate]: 4.58999e-06 [updatestate_loads_eliminate]: 3.93001e-06 [cse]: 2.413e-05 [renormalize]: 3.69997e-07 [remove_dup_value]: 1.68e-05 [tuple_transform]: 0.00013762, [1] [Cycle 1]: 0.00013254, [4] [d_1]: 5.525e-05 [none_parameter_eliminate]: 2.14e-06 [renormalize]: 1.79978e-07 [switch_simplify]: 9.97001e-06 [partial_unused_args_eliminate]: 2.02001e-06 [add_recomputation]: 8.435e-05 [cse_after_recomputation]: 2.837e-05, [1] [Cycle 1]: 2.335e-05, [1] [cse]: 1.756e-05 [environ_conv]: 7.83999e-06 [swap_dp_allreduce_reducescatter]: 8.60999e-06 [bias_add_comm_swap]: 3.3e-06 [label_micro_interleaved_index]: 4.93001e-06 [label_fine_grained_interleaved_index]: 3.46001e-06 [merge_cast_opt]: 1.41998e-06 [slice_recompute_activation]: 2.67001e-06 [micro_interleaved_order_control]: 2.74999e-06 [assign_add_opt]: 1.34e-06 [ForceFp32Comm]: 1.02e-06 [remove_cast_before_assign_add]: 1.41998e-06 [full_micro_interleaved_order_control]: 2.42001e-06 [reorder_send_recv_between_fp_bp]: 3.06999e-06 [comm_op_add_attrs]: 1.05001e-06 [add_comm_op_reuse_tag]: 1.05001e-06 [interleave_split_concat_branches]: 1.18001e-06 [interleave_parallel_branches]: 1.09003e-06 [overlap_opt_shard_in_pipeline]: 1.90001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.72001e-06 [control_data_broadcast_order]: 1.804e-05 [grouped_pairwise_exchange_alltoall]: 1.62999e-06 [offloading_packed_experts]: 5.12999e-06 [overlap_recompute_and_grad_model_parallel]: 5.44998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.12999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37e-06 [overlap_recompute_comm]: 2.56e-06 [overlap_grad_ring_attention]: 5.33002e-06 [overlap_grad_flash_sp]: 2.76e-05 [begin_end_overlap_inline]: 5.59987e-07 [split_matmul_comm_elemetwise]: 2.48002e-06 [split_layernorm_comm]: 2.51e-06 [handle_group_info]: 1.02e-06 [symbol_engine_optimizer]: 9.288e-05, [1] [Cycle 1]: 8.816e-05, [6] [build]: 5.33002e-06 [elim_shapecalc]: 1.234e-05 [elim_not_effective]: 1.772e-05 [opt_reshape]: 8.90999e-06 [fold_const_symbol]: 1.483e-05 [renormalize]: 1.8999e-07 [detach_backward]: 2.48e-06 [pipeline_parallel_scheduler]: 1.55999e-06 [auto_monad_reorder]: 2.901e-05 [get_jit_bprop_graph]: 2.44999e-06 [rewriter_after_jit_bprop_graph]: 4.53999e-06 [opt_after_jit_grad]: 0.00049405 [validate]: 5.274e-05 [backend_pass]: 9.39996e-07 [task_emit]: 0.00908849 [execute]: 8.69e-06 Sums bootstrap : 0.000440s : 0.21% type_inference : 0.006763s : 3.17% event_method : 0.000022s : 0.01% auto_monad : 0.000168s : 0.08% graph_reusing : 0.000007s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000025s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000005s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000043s : 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 : 0.000031s : 0.01% optimize.rewriter_before_opt_a : 0.000086s : 0.04% optimize.opt_a.expand_dump_flag : 0.000006s : 0.00% optimize.opt_a.switch_simplify : 0.000053s : 0.03% optimize.opt_a.loop_unroll : 0.000039s : 0.02% optimize.opt_a.a_1 : 0.191524s : 89.80% optimize.opt_a.with_stream_mark : 0.000046s : 0.02% optimize.opt_a.recompute_prepare : 0.000024s : 0.01% 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 : 0.000009s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000234s : 0.11% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.00% optimize.opt_a.shard_inline : 0.000017s : 0.01% optimize.opt_a.merge_send_recv : 0.000018s : 0.01% optimize.opt_a.auto_parallel : 0.000021s : 0.01% optimize.opt_a.parallel : 0.000028s : 0.01% optimize.opt_a.flash_sp : 0.000016s : 0.01% optimize.opt_a.merge_comm : 0.000010s : 0.00% optimize.opt_a.allreduce_fusion : 0.000010s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000021s : 0.01% 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 : 0.000010s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.00% optimize.opt_a.offload_activation : 0.000024s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000034s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.00% optimize.opt_a.before_grad : 0.000030s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000010s : 0.00% optimize.opt_a.meta_fg_expand : 0.000008s : 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 : 0.01% optimize.opt_a.a_after_grad : 0.000026s : 0.01% optimize.opt_a.renormalize : 0.001014s : 0.48% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.00% optimize.opt_a.auto_monad_grad : 0.000003s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000035s : 0.02% optimize.opt_a.cse : 0.000062s : 0.03% optimize.opt_a.a_3 : 0.000116s : 0.05% optimize.py_interpret_to_execute_after_opt_a : 0.000015s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000214s : 0.10% optimize.convert_after_rewriter : 0.000011s : 0.01% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000802s : 0.38% optimize.opt_b.b_1 : 0.000227s : 0.11% optimize.opt_b.b_2 : 0.000011s : 0.01% 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 : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000027s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000029s : 0.01% optimize.loop_unroll : 0.000497s : 0.23% optimize.opt_after_cconv.c_1 : 0.000054s : 0.03% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000024s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000055s : 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 : 0.000010s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000084s : 0.04% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% optimize.environ_conv : 0.000008s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 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 : 0.000003s : 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 : 0.000002s : 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 : 0.000018s : 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 : 0.000005s : 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 : 0.000005s : 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 : 0.000003s : 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 : 0.000012s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000009s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 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 : 0.000029s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000494s : 0.23% validate : 0.000053s : 0.02% backend_pass : 0.000001s : 0.00% task_emit : 0.009088s : 4.26% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.000395 60 1.94% : 0.000008s : 2: substitution.depend_value_elim 0.80% : 0.000003s : 5: substitution.elim_not_effective 0.59% : 0.000002s : 5: substitution.fold_const_symbol 1.93% : 0.000008s : 6: substitution.graph_param_transform 78.26% : 0.000309s : 4: substitution.inline 1.53% : 0.000006s : 10: substitution.j_node_and_user_rematch 2.00% : 0.000008s : 10: substitution.remove_not_recompute_node 0.88% : 0.000003s : 2: substitution.replace_old_param 4.43% : 0.000017s : 2: substitution.tuple_list_get_item_eliminator 4.50% : 0.000018s : 6: substitution.updatestate_pure_node_eliminater 3.16% : 0.000012s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.006695 2 88.59% : 0.005931s : 1: type_inference.infer 11.41% : 0.000764s : 1: type_inference.specialize ------[replace.] 0.000082 6 74.23% : 0.000061s : 4: replace.inline 25.77% : 0.000021s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000320 6 95.26% : 0.000305s : 4: match.inline 4.74% : 0.000015s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000280 1652 1.27% : 0.000004s : 16: predicate.accumulaten_eliminater 0.78% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.62% : 0.000002s : 12: predicate.addn_check_dump 0.95% : 0.000003s : 16: predicate.addn_zero_filter 0.87% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.55% : 0.000007s : 28: predicate.arithmetic_simplify 1.04% : 0.000003s : 16: predicate.cast_eliminate 0.58% : 0.000002s : 12: predicate.check_bprop_eliminate 0.60% : 0.000002s : 12: predicate.compare_switch_simplify 0.16% : 0.000000s : 6: predicate.const_output_eliminate 0.75% : 0.000002s : 12: predicate.depend_value_elim 0.90% : 0.000003s : 16: predicate.dict_get_item_const_eliminator 1.04% : 0.000003s : 16: predicate.dict_get_item_eliminator 0.87% : 0.000002s : 16: predicate.dict_set_item_eliminator 0.87% : 0.000002s : 12: predicate.dumpgradient_eliminate 0.25% : 0.000001s : 6: predicate.elim_not_effective 0.46% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.43% : 0.000004s : 22: predicate.environ_add_const_eliminate 1.09% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.08% : 0.000003s : 22: predicate.environ_get_depend_swap 1.86% : 0.000005s : 34: predicate.environ_get_eliminate 1.18% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.14% : 0.000003s : 22: predicate.exchange_switch_depend_value 1.98% : 0.000006s : 22: predicate.float_depend_g_call 0.56% : 0.000002s : 12: predicate.float_environ_get_switch 0.87% : 0.000002s : 18: predicate.float_tuple_getitem_switch 0.23% : 0.000001s : 6: predicate.fold_const_symbol 0.67% : 0.000002s : 12: predicate.get_grad_eliminate 0.29% : 0.000001s : 6: predicate.graph_param_transform 0.61% : 0.000002s : 12: predicate.incorporate_call 0.52% : 0.000001s : 12: predicate.incorporate_call_switch 6.62% : 0.000019s : 74: predicate.inline 0.87% : 0.000002s : 12: predicate.inline_without_move 0.30% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.90% : 0.000003s : 12: predicate.less_batch_normalization 1.75% : 0.000005s : 30: predicate.list_to_tuple_eliminator_ 2.25% : 0.000006s : 46: predicate.load_eliminater 1.02% : 0.000003s : 6: predicate.loop_unroll_after_grad 1.90% : 0.000005s : 38: predicate.loop_unroll_before_grad 1.61% : 0.000004s : 28: predicate.make_slice_get_slice_eliminator 0.61% : 0.000002s : 12: predicate.merge_addn 0.59% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.65% : 0.000002s : 12: predicate.mini_step_allgather_replace 3.38% : 0.000009s : 16: predicate.minmaximum_grad 0.94% : 0.000003s : 6: predicate.mutable_eliminate 0.36% : 0.000001s : 6: predicate.opt_reshape 0.42% : 0.000001s : 6: predicate.parallel_virtual_node 1.68% : 0.000005s : 22: predicate.partial_defer_inline 1.34% : 0.000004s : 24: predicate.partial_eliminate 0.86% : 0.000002s : 16: predicate.print_const_string_wrapper 0.61% : 0.000002s : 12: predicate.reduce_all_const_elim 1.05% : 0.000003s : 16: predicate.reduce_eliminate 2.35% : 0.000007s : 46: predicate.redundant_stop_gradient_eliminater 0.37% : 0.000001s : 12: predicate.remove_not_recompute_node 1.20% : 0.000003s : 30: predicate.replace_applicator 0.39% : 0.000001s : 12: predicate.replace_old_param 0.25% : 0.000001s : 6: predicate.reset_defer_inline 0.87% : 0.000002s : 16: predicate.reshape_eliminate 0.61% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.41% : 0.000001s : 6: predicate.row_tensor_eliminate 0.75% : 0.000002s : 12: predicate.same_eliminate 0.38% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.88% : 0.000002s : 12: predicate.shard_identity_eliminate 0.77% : 0.000002s : 12: predicate.special_op_eliminate 0.73% : 0.000002s : 12: predicate.specialize_transform 0.99% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 0.74% : 0.000002s : 12: predicate.stack_unstack_eliminate 0.41% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.15% : 0.000003s : 22: predicate.switch_defer_inline 1.77% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.54% : 0.000013s : 78: predicate.switch_simplify 0.98% : 0.000003s : 16: predicate.tile_eliminate 0.91% : 0.000003s : 16: predicate.transpose_eliminate 1.52% : 0.000004s : 28: predicate.tuple_list_convert_item_index_to_positive 1.54% : 0.000004s : 28: predicate.tuple_list_get_item_const_eliminator 1.47% : 0.000004s : 28: predicate.tuple_list_get_item_depend_reorder 2.79% : 0.000008s : 42: predicate.tuple_list_get_item_eliminator 1.35% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.13% : 0.000006s : 40: predicate.tuple_list_set_item_eliminator 1.62% : 0.000005s : 30: predicate.tuple_to_list_eliminator_ 2.40% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.39% : 0.000009s : 58: predicate.updatestate_useless_node_eliminater 0.46% : 0.000001s : 6: predicate.value_based_eliminate 0.67% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.65% : 0.000002s : 12: predicate.virtual_output_eliminate 0.30% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.44% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000618 11 50.90% : 0.000314s : 5: func_graph_cloner_run.FuncGraphClonerGraph 49.10% : 0.000303s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.612417 196 0.00% : 0.000004s : 1: ForceFp32Comm 0.62% : 0.003794s : 1: add_attr 0.62% : 0.003779s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000089s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000176s : 1: auto_monad 0.01% : 0.000033s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.08% : 0.000471s : 1: bootstrap 0.01% : 0.000032s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000015s : 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% : 0.000011s : 1: environ_conv 0.00% : 0.000029s : 1: event_method 0.00% : 0.000014s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000007s : 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 0.08% : 0.000506s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.13% : 0.000812s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 31.37% : 0.192117s : 78: opt.transform.opt_a 0.01% : 0.000052s : 1: opt.transform.opt_after_cconv 0.00% : 0.000030s : 1: opt.transform.opt_after_jit_grad 0.03% : 0.000169s : 28: opt.transform.opt_b 0.01% : 0.000063s : 2: opt.transform.opt_trans_graph 0.01% : 0.000050s : 4: opt.transform.symbol_engine_opt 31.70% : 0.194119s : 1: opt_a 0.02% : 0.000143s : 1: opt_after_cconv 0.08% : 0.000503s : 1: opt_after_jit_grad 0.05% : 0.000335s : 1: opt_b 32.17% : 0.197028s : 1: optimize 0.00% : 0.000026s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 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% : 0.000004s : 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% : 0.000008s : 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% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.01% : 0.000047s : 1: pre_auto_parallel 0.01% : 0.000036s : 1: py_interpret_to_execute 0.00% : 0.000019s : 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 0.10% : 0.000587s : 1: renormalize.infer 0.07% : 0.000418s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.04% : 0.000219s : 1: rewriter_after_opt_a 0.01% : 0.000091s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.02% : 0.000096s : 1: symbol_engine_optimizer 1.49% : 0.009106s : 1: task_emit 0.02% : 0.000141s : 1: tuple_transform 1.11% : 0.006786s : 1: type_inference 0.01% : 0.000080s : 1: validate TotalTime = 1.38592, [24] [bootstrap]: 0.00043435 [type_inference]: 0.0118485 [event_method]: 5.494e-05 [auto_monad]: 0.00024198 [graph_reusing]: 8.35001e-06 [inline]: 2.68e-06 [add_attr]: 0.654047, [1] [add_attr_with_inline]: 0.654034, [1] [Cycle 1]: 0.00010487, [2] [tag_attr]: 4.445e-05 [meta_addattr_fg_expand]: 9.17999e-06 [parallel-infer-symbol]: 4.35999e-06 [pre_auto_parallel]: 6.267e-05 [insert-virtual-dataset]: 3.21001e-06 [parallel-infer-symbol-second]: 7.7e-07 [dataset_repeat_opt]: 2.14e-06 [pipeline_split]: 1.80001e-06 [optimize]: 0.708909, [53] [py_interpret_to_execute]: 4.684e-05 [rewriter_before_opt_a]: 0.00015588 [opt_a]: 0.706324, [3] [Cycle 1]: 0.0163097, [45] [expand_dump_flag]: 3.79002e-06 [switch_simplify]: 7.634e-05 [loop_unroll]: 5.891e-05 [a_1]: 0.00137564 [with_stream_mark]: 7.098e-05 [recompute_prepare]: 2.71e-05 [updatestate_depend_eliminate]: 1.195e-05 [updatestate_assign_eliminate]: 1.092e-05 [updatestate_loads_eliminate]: 9.55001e-06 [parameter_eliminate]: 3.54002e-06 [a_2]: 0.00027774 [accelerated_algorithm]: 1.861e-05 [shard]: 2.17999e-06 [meta_shard_fg_expand]: 4.37e-06 [shard_inline]: 1.742e-05 [merge_send_recv]: 2.118e-05 [auto_parallel]: 1.501e-05 [parallel]: 2.243e-05 [flash_sp]: 1.198e-05 [merge_comm]: 1.176e-05 [allreduce_fusion]: 1.107e-05 [matmul_add_comm_reduction]: 3.661e-05 [allreduce_slice_to_reducescatter]: 1.44e-06 [virtual_shard_identity]: 2.063e-05 [virtual_dataset]: 1.824e-05 [get_grad_eliminate_]: 1.778e-05 [virtual_output]: 1.728e-05 [merge_forward]: 1.181e-05 [cell_reuse_recompute_pass]: 1.40999e-06 [offload_activation]: 2.59e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.56e-05 [merge_recompute_call_nodes]: 1.65001e-06 [before_grad]: 3.371e-05 [set_forward_comm_id_for_comm_node_pass]: 1.258e-05 [meta_fg_expand]: 0.00632776 [flash_sp_send_recv_attached]: 4.95001e-06 [receive_attached]: 2.49001e-06 [after_resolve]: 0.00011371 [a_after_grad]: 0.00017697 [renormalize]: 0.00589202 [add_forward_monad_depend]: 1.549e-05 [auto_monad_grad]: 9.04998e-06 [auto_monad_eliminator]: 7.655e-05 [cse]: 0.00022696 [a_3]: 0.00083529 [Cycle 2]: 0.689092, [45] [expand_dump_flag]: 3.78001e-06 [switch_simplify]: 5.858e-05 [loop_unroll]: 5.515e-05 [a_1]: 0.00127421 [with_stream_mark]: 2.424e-05 [recompute_prepare]: 1.42e-05 [updatestate_depend_eliminate]: 3.427e-05 [updatestate_assign_eliminate]: 5.67999e-06 [updatestate_loads_eliminate]: 4.97e-06 [parameter_eliminate]: 2.56e-06 [a_2]: 0.00011981 [accelerated_algorithm]: 9.62001e-06 [shard]: 2.25002e-06 [meta_shard_fg_expand]: 3.21999e-06 [shard_inline]: 8.39002e-06 [merge_send_recv]: 1.056e-05 [auto_parallel]: 1.219e-05 [parallel]: 9.22999e-06 [flash_sp]: 4.28999e-06 [merge_comm]: 5.67999e-06 [allreduce_fusion]: 5.18002e-06 [matmul_add_comm_reduction]: 1.197e-05 [allreduce_slice_to_reducescatter]: 1.12e-06 [virtual_shard_identity]: 1.12e-05 [virtual_dataset]: 8.49002e-06 [get_grad_eliminate_]: 7.66001e-06 [virtual_output]: 7.72002e-06 [merge_forward]: 5.62999e-06 [cell_reuse_recompute_pass]: 1.32e-06 [offload_activation]: 0.685455 [cell_reuse_handle_not_recompute_node_pass]: 7.783e-05 [merge_recompute_call_nodes]: 4.60999e-06 [before_grad]: 2.273e-05 [set_forward_comm_id_for_comm_node_pass]: 2.263e-05 [meta_fg_expand]: 8.88e-05 [flash_sp_send_recv_attached]: 2.41e-06 [receive_attached]: 3.01001e-06 [after_resolve]: 1.896e-05 [a_after_grad]: 1.528e-05 [renormalize]: 0.00118725 [add_forward_monad_depend]: 8.02e-06 [auto_monad_grad]: 2.69001e-06 [auto_monad_eliminator]: 2.543e-05 [cse]: 4.478e-05 [a_3]: 6.436e-05 [Cycle 3]: 0.0009019, [45] [expand_dump_flag]: 2.61e-06 [switch_simplify]: 1.029e-05 [loop_unroll]: 8.58001e-06 [a_1]: 0.00023481 [with_stream_mark]: 1.752e-05 [recompute_prepare]: 9.43002e-06 [updatestate_depend_eliminate]: 6.44001e-06 [updatestate_assign_eliminate]: 5.94e-06 [updatestate_loads_eliminate]: 5.37999e-06 [parameter_eliminate]: 2.37001e-06 [a_2]: 0.00011485 [accelerated_algorithm]: 9.24e-06 [shard]: 2.27999e-06 [meta_shard_fg_expand]: 2.43e-06 [shard_inline]: 8.45999e-06 [merge_send_recv]: 7.57002e-06 [auto_parallel]: 1.234e-05 [parallel]: 9.87001e-06 [flash_sp]: 1.60999e-06 [merge_comm]: 5.24e-06 [allreduce_fusion]: 5.20001e-06 [matmul_add_comm_reduction]: 1.158e-05 [allreduce_slice_to_reducescatter]: 7.50006e-07 [virtual_shard_identity]: 1.021e-05 [virtual_dataset]: 8.70999e-06 [get_grad_eliminate_]: 7.59002e-06 [virtual_output]: 8.29998e-06 [merge_forward]: 5.52999e-06 [cell_reuse_recompute_pass]: 1.59e-06 [offload_activation]: 1.302e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.61e-05 [merge_recompute_call_nodes]: 6.89994e-07 [before_grad]: 1.391e-05 [set_forward_comm_id_for_comm_node_pass]: 5.59998e-06 [meta_fg_expand]: 3.13e-06 [flash_sp_send_recv_attached]: 9.70002e-07 [receive_attached]: 1.35999e-06 [after_resolve]: 1.068e-05 [a_after_grad]: 1.272e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 1.37e-06 [auto_monad_grad]: 1.08001e-06 [auto_monad_eliminator]: 1.255e-05 [cse]: 2.142e-05 [a_3]: 5.16e-05 [py_interpret_to_execute_after_opt_a]: 1.656e-05 [slice_cell_reuse_recomputed_activation]: 2.39999e-06 [rewriter_after_opt_a]: 6.092e-05 [convert_after_rewriter]: 8.59e-06 [order_py_execute_after_rewriter]: 6.96999e-06 [mutable_eliminate]: 0.00072276 [opt_b]: 0.00028021, [1] [Cycle 1]: 0.00027265, [7] [b_1]: 0.0001838 [b_2]: 1.008e-05 [updatestate_depend_eliminate]: 8.13999e-06 [updatestate_assign_eliminate]: 4.35999e-06 [updatestate_loads_eliminate]: 4.50999e-06 [renormalize]: 5.00004e-07 [cse]: 2.597e-05 [optimize_parallel_all_gather_comm]: 2.03e-05 [overlap_param_gather]: 1.87001e-06 [cconv]: 2.746e-05 [loop_unroll]: 0.00042666 [opt_after_cconv]: 0.00013476, [1] [Cycle 1]: 0.00012906, [7] [c_1]: 5.093e-05 [parameter_eliminate]: 2.54001e-06 [updatestate_depend_eliminate]: 7.60998e-06 [updatestate_assign_eliminate]: 4.45999e-06 [updatestate_loads_eliminate]: 4.22e-06 [cse]: 2.415e-05 [renormalize]: 4.80009e-07 [remove_dup_value]: 1.806e-05 [tuple_transform]: 9.276e-05, [1] [Cycle 1]: 8.791e-05, [4] [d_1]: 5.792e-05 [none_parameter_eliminate]: 1.62001e-06 [renormalize]: 2.00002e-07 [switch_simplify]: 9.39998e-06 [partial_unused_args_eliminate]: 1.84e-06 [add_recomputation]: 7.562e-05 [cse_after_recomputation]: 3.053e-05, [1] [Cycle 1]: 2.563e-05, [1] [cse]: 1.985e-05 [environ_conv]: 6.21e-06 [swap_dp_allreduce_reducescatter]: 8.11002e-06 [bias_add_comm_swap]: 2.58e-06 [label_micro_interleaved_index]: 4.53999e-06 [label_fine_grained_interleaved_index]: 2.78e-06 [merge_cast_opt]: 1.71e-06 [slice_recompute_activation]: 2.14999e-06 [micro_interleaved_order_control]: 2.54001e-06 [assign_add_opt]: 1.29998e-06 [ForceFp32Comm]: 8.2e-07 [remove_cast_before_assign_add]: 1.23002e-06 [full_micro_interleaved_order_control]: 3.11001e-06 [reorder_send_recv_between_fp_bp]: 2.32999e-06 [comm_op_add_attrs]: 1.30999e-06 [add_comm_op_reuse_tag]: 1.20001e-06 [interleave_split_concat_branches]: 1.14e-06 [interleave_parallel_branches]: 1.10999e-06 [overlap_opt_shard_in_pipeline]: 1.39998e-06 [overlap_opt_shard_grad_in_pipeline]: 1.86e-06 [control_data_broadcast_order]: 1.732e-05 [grouped_pairwise_exchange_alltoall]: 1.93997e-06 [offloading_packed_experts]: 5.56002e-06 [overlap_recompute_and_grad_model_parallel]: 6.33002e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.15999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.37e-06 [overlap_recompute_comm]: 2.36e-06 [overlap_grad_ring_attention]: 5.22e-06 [overlap_grad_flash_sp]: 2.869e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.45002e-06 [split_layernorm_comm]: 2.02999e-06 [handle_group_info]: 1.14e-06 [symbol_engine_optimizer]: 9.116e-05, [1] [Cycle 1]: 8.64e-05, [6] [build]: 4.43999e-06 [elim_shapecalc]: 1.258e-05 [elim_not_effective]: 1.626e-05 [opt_reshape]: 9.11002e-06 [fold_const_symbol]: 1.402e-05 [renormalize]: 1.8999e-07 [detach_backward]: 2.06e-06 [pipeline_parallel_scheduler]: 1.72001e-06 [auto_monad_reorder]: 5.903e-05 [get_jit_bprop_graph]: 1.99e-06 [rewriter_after_jit_bprop_graph]: 4.35e-06 [opt_after_jit_grad]: 0.00046636 [validate]: 4.836e-05 [backend_pass]: 9.80013e-07 [task_emit]: 0.00947627 [execute]: 8.07998e-06 Sums bootstrap : 0.000434s : 0.06% type_inference : 0.011848s : 1.62% event_method : 0.000055s : 0.01% auto_monad : 0.000242s : 0.03% graph_reusing : 0.000008s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000044s : 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 : 0.000063s : 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 : 0.000047s : 0.01% optimize.rewriter_before_opt_a : 0.000156s : 0.02% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000145s : 0.02% optimize.opt_a.loop_unroll : 0.000123s : 0.02% optimize.opt_a.a_1 : 0.002885s : 0.39% optimize.opt_a.with_stream_mark : 0.000113s : 0.02% optimize.opt_a.recompute_prepare : 0.000051s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000053s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000023s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000020s : 0.00% optimize.opt_a.parameter_eliminate : 0.000008s : 0.00% optimize.opt_a.a_2 : 0.000512s : 0.07% optimize.opt_a.accelerated_algorithm : 0.000037s : 0.01% optimize.opt_a.shard : 0.000007s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000034s : 0.00% optimize.opt_a.merge_send_recv : 0.000039s : 0.01% optimize.opt_a.auto_parallel : 0.000040s : 0.01% optimize.opt_a.parallel : 0.000042s : 0.01% optimize.opt_a.flash_sp : 0.000018s : 0.00% optimize.opt_a.merge_comm : 0.000023s : 0.00% optimize.opt_a.allreduce_fusion : 0.000021s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000060s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000042s : 0.01% optimize.opt_a.virtual_dataset : 0.000035s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000033s : 0.00% optimize.opt_a.virtual_output : 0.000033s : 0.00% optimize.opt_a.merge_forward : 0.000023s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.685494s : 93.85% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000130s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000007s : 0.00% optimize.opt_a.before_grad : 0.000070s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000041s : 0.01% optimize.opt_a.meta_fg_expand : 0.006420s : 0.88% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000143s : 0.02% optimize.opt_a.a_after_grad : 0.000205s : 0.03% optimize.opt_a.renormalize : 0.007079s : 0.97% optimize.opt_a.add_forward_monad_depend : 0.000025s : 0.00% optimize.opt_a.auto_monad_grad : 0.000013s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000115s : 0.02% optimize.opt_a.cse : 0.000293s : 0.04% optimize.opt_a.a_3 : 0.000951s : 0.13% optimize.py_interpret_to_execute_after_opt_a : 0.000017s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000061s : 0.01% optimize.convert_after_rewriter : 0.000009s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000723s : 0.10% optimize.opt_b.b_1 : 0.000184s : 0.03% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000026s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000020s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000027s : 0.00% optimize.loop_unroll : 0.000427s : 0.06% optimize.opt_after_cconv.c_1 : 0.000051s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 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 : 0.000024s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.00% optimize.tuple_transform.d_1 : 0.000058s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000076s : 0.01% optimize.cse_after_recomputation.cse : 0.000020s : 0.00% optimize.environ_conv : 0.000006s : 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 : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000003s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000002s : 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 : 0.000017s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 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 : 0.000029s : 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 : 0.000013s : 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 : 0.000014s : 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 : 0.000059s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000466s : 0.06% validate : 0.000048s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.009476s : 1.30% execute : 0.000008s : 0.00% Time group info: ------[substitution.] 0.001575 224 1.04% : 0.000016s : 4: substitution.depend_value_elim 0.17% : 0.000003s : 5: substitution.elim_not_effective 0.18% : 0.000003s : 3: substitution.float_depend_g_call 0.27% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.14% : 0.000002s : 5: substitution.fold_const_symbol 46.23% : 0.000728s : 4: substitution.getattr_setattr_resolve 0.49% : 0.000008s : 6: substitution.graph_param_transform 0.21% : 0.000003s : 2: substitution.incorporate_call 0.13% : 0.000002s : 2: substitution.incorporate_call_switch 29.82% : 0.000470s : 13: substitution.inline 2.25% : 0.000035s : 5: substitution.inline_without_move 1.02% : 0.000016s : 23: substitution.j_node_and_user_rematch 0.68% : 0.000011s : 6: substitution.minmaximum_grad 0.25% : 0.000004s : 3: substitution.partial_eliminate 1.44% : 0.000023s : 23: substitution.remove_not_recompute_node 3.54% : 0.000056s : 21: substitution.replace_applicator 1.00% : 0.000016s : 16: substitution.replace_old_param 0.25% : 0.000004s : 1: substitution.set_cell_output_no_recompute 1.37% : 0.000022s : 6: substitution.tuple_list_convert_item_index_to_positive 0.59% : 0.000009s : 6: substitution.tuple_list_get_item_const_eliminator 0.85% : 0.000013s : 6: substitution.tuple_list_get_item_depend_reorder 3.08% : 0.000048s : 15: substitution.tuple_list_get_item_eliminator 0.86% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 1.76% : 0.000028s : 18: substitution.updatestate_pure_node_eliminater 2.39% : 0.000038s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.011767 2 89.04% : 0.010477s : 1: type_inference.infer 10.96% : 0.001290s : 1: type_inference.specialize ------[replace.] 0.000308 25 16.20% : 0.000050s : 3: replace.getattr_setattr_resolve 37.44% : 0.000115s : 13: replace.inline 13.20% : 0.000041s : 1: replace.replace_applicator 25.59% : 0.000079s : 7: replace.tuple_list_get_item_eliminator 7.57% : 0.000023s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001190 25 57.40% : 0.000683s : 3: match.getattr_setattr_resolve 38.84% : 0.000462s : 13: match.inline 0.99% : 0.000012s : 1: match.replace_applicator 1.89% : 0.000023s : 7: match.tuple_list_get_item_eliminator 0.87% : 0.000010s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000781 5424 0.98% : 0.000008s : 55: predicate.accumulaten_eliminater 0.24% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.54% : 0.000004s : 29: predicate.addn_check_dump 0.95% : 0.000007s : 55: predicate.addn_zero_filter 0.93% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 1.89% : 0.000015s : 84: predicate.arithmetic_simplify 1.00% : 0.000008s : 55: predicate.cast_eliminate 2.13% : 0.000017s : 130: predicate.check_bprop_eliminate 0.49% : 0.000004s : 29: predicate.compare_switch_simplify 0.06% : 0.000000s : 6: predicate.const_output_eliminate 0.53% : 0.000004s : 29: predicate.depend_value_elim 1.04% : 0.000008s : 55: predicate.dict_get_item_const_eliminator 1.03% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.99% : 0.000008s : 55: predicate.dict_set_item_eliminator 0.35% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.08% : 0.000001s : 6: predicate.elim_not_effective 0.15% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.04% : 0.000008s : 61: predicate.environ_add_const_eliminate 1.03% : 0.000008s : 61: predicate.environ_get_add_eliminate 1.07% : 0.000008s : 61: predicate.environ_get_depend_swap 1.54% : 0.000012s : 90: predicate.environ_get_eliminate 1.01% : 0.000008s : 61: predicate.environ_get_set_eliminate 1.33% : 0.000010s : 76: predicate.exchange_switch_depend_value 1.90% : 0.000015s : 76: predicate.float_depend_g_call 0.54% : 0.000004s : 29: predicate.float_environ_get_switch 0.63% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.06% : 0.000000s : 6: predicate.fold_const_symbol 0.60% : 0.000005s : 29: predicate.get_grad_eliminate 0.70% : 0.000005s : 20: predicate.getattr_setattr_resolve 0.10% : 0.000001s : 6: predicate.graph_param_transform 0.51% : 0.000004s : 29: predicate.incorporate_call 0.45% : 0.000004s : 29: predicate.incorporate_call_switch 4.75% : 0.000037s : 201: predicate.inline 2.27% : 0.000018s : 97: predicate.inline_without_move 0.27% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.70% : 0.000005s : 29: predicate.less_batch_normalization 1.32% : 0.000010s : 74: predicate.list_to_tuple_eliminator_ 2.23% : 0.000017s : 129: predicate.load_eliminater 0.30% : 0.000002s : 6: predicate.loop_unroll_after_grad 2.43% : 0.000019s : 129: predicate.loop_unroll_before_grad 1.23% : 0.000010s : 67: predicate.make_slice_get_slice_eliminator 0.50% : 0.000004s : 29: predicate.merge_addn 2.04% : 0.000016s : 126: predicate.micro_step_allgather_replace 2.03% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.96% : 0.000007s : 55: predicate.minmaximum_grad 0.29% : 0.000002s : 6: predicate.mutable_eliminate 0.12% : 0.000001s : 6: predicate.opt_reshape 0.15% : 0.000001s : 6: predicate.parallel_virtual_node 1.60% : 0.000012s : 76: predicate.partial_defer_inline 1.29% : 0.000010s : 68: predicate.partial_eliminate 0.99% : 0.000008s : 55: predicate.print_const_string_wrapper 0.52% : 0.000004s : 29: predicate.reduce_all_const_elim 1.21% : 0.000009s : 55: predicate.reduce_eliminate 2.24% : 0.000017s : 129: predicate.redundant_stop_gradient_eliminater 0.45% : 0.000003s : 29: predicate.remove_not_recompute_node 2.15% : 0.000017s : 190: predicate.replace_applicator 1.05% : 0.000008s : 97: predicate.replace_old_param 0.09% : 0.000001s : 6: predicate.reset_defer_inline 1.10% : 0.000009s : 55: predicate.reshape_eliminate 2.16% : 0.000017s : 126: predicate.row_tensor_add_zeros_like 0.14% : 0.000001s : 6: predicate.row_tensor_eliminate 2.35% : 0.000018s : 130: predicate.same_eliminate 0.39% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.69% : 0.000005s : 29: predicate.shard_identity_eliminate 0.28% : 0.000002s : 12: predicate.special_op_eliminate 0.62% : 0.000005s : 29: predicate.specialize_transform 2.29% : 0.000018s : 126: predicate.split_environ_get_set_with_tuple_value 2.11% : 0.000017s : 97: predicate.stack_unstack_eliminate 0.12% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.45% : 0.000011s : 76: predicate.switch_defer_inline 3.48% : 0.000027s : 206: predicate.switch_layer_defer_inline 4.54% : 0.000035s : 240: predicate.switch_simplify 0.98% : 0.000008s : 55: predicate.tile_eliminate 0.95% : 0.000007s : 55: predicate.transpose_eliminate 1.34% : 0.000010s : 67: predicate.tuple_list_convert_item_index_to_positive 1.36% : 0.000011s : 67: predicate.tuple_list_get_item_const_eliminator 1.18% : 0.000009s : 67: predicate.tuple_list_get_item_depend_reorder 2.32% : 0.000018s : 103: predicate.tuple_list_get_item_eliminator 1.28% : 0.000010s : 67: predicate.tuple_list_get_set_item_eliminator 1.83% : 0.000014s : 96: predicate.tuple_list_set_item_eliminator 1.29% : 0.000010s : 74: predicate.tuple_to_list_eliminator_ 2.25% : 0.000018s : 129: predicate.updatestate_pure_node_eliminater 2.88% : 0.000022s : 159: predicate.updatestate_useless_node_eliminater 0.14% : 0.000001s : 6: predicate.value_based_eliminate 0.62% : 0.000005s : 29: predicate.virtual_dataset_eliminate 0.54% : 0.000004s : 29: predicate.virtual_output_eliminate 0.12% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.14% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002454 48 63.47% : 0.001558s : 24: func_graph_cloner_run.FuncGraphClonerGraph 36.53% : 0.000897s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.762315 247 0.00% : 0.000004s : 1: ForceFp32Comm 23.68% : 0.654055s : 1: add_attr 23.68% : 0.654038s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000080s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000251s : 1: auto_monad 0.00% : 0.000064s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.02% : 0.000459s : 1: bootstrap 0.00% : 0.000031s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000020s : 1: control_data_broadcast_order 0.00% : 0.000012s : 1: convert_after_rewriter 0.00% : 0.000034s : 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 0.00% : 0.000062s : 1: event_method 0.00% : 0.000015s : 1: execute 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000013s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000007s : 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 0.02% : 0.000435s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.03% : 0.000731s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 0.19% : 0.005262s : 125: opt.transform.opt_a 0.00% : 0.000050s : 1: opt.transform.opt_after_cconv 0.00% : 0.000031s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000168s : 28: opt.transform.opt_b 0.03% : 0.000828s : 2: opt.transform.opt_resolve 0.00% : 0.000065s : 2: opt.transform.opt_trans_graph 0.00% : 0.000049s : 4: opt.transform.symbol_engine_opt 25.57% : 0.706328s : 1: opt_a 0.01% : 0.000150s : 1: opt_after_cconv 0.02% : 0.000476s : 1: opt_after_jit_grad 0.01% : 0.000284s : 1: opt_b 25.66% : 0.708915s : 1: optimize 0.00% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 1: order_py_execute_after_rewriter 0.00% : 0.000032s : 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% : 0.000009s : 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 0.00% : 0.000068s : 1: pre_auto_parallel 0.00% : 0.000051s : 1: py_interpret_to_execute 0.00% : 0.000020s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000022s : 1: remove_dup_value 0.16% : 0.004534s : 2: renormalize.infer 0.09% : 0.002521s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000065s : 1: rewriter_after_opt_a 0.01% : 0.000161s : 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% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000094s : 1: symbol_engine_optimizer 0.34% : 0.009492s : 1: task_emit 0.00% : 0.000096s : 1: tuple_transform 0.43% : 0.011870s : 1: type_inference 0.00% : 0.000077s : 1: validate TotalTime = 0.441617, [24] [bootstrap]: 0.00050115 [type_inference]: 0.184711 [event_method]: 5.331e-05 [auto_monad]: 0.00024069 [graph_reusing]: 8.95999e-06 [inline]: 2.37001e-06 [add_attr]: 0.00372452, [1] [add_attr_with_inline]: 0.00371366, [1] [Cycle 1]: 9.262e-05, [2] [tag_attr]: 4.206e-05 [meta_addattr_fg_expand]: 9.62001e-06 [parallel-infer-symbol]: 4.50001e-06 [pre_auto_parallel]: 5.401e-05 [insert-virtual-dataset]: 2.58998e-06 [parallel-infer-symbol-second]: 7.30011e-07 [dataset_repeat_opt]: 2.21e-06 [pipeline_split]: 1.99999e-06 [optimize]: 0.243251, [53] [py_interpret_to_execute]: 4.034e-05 [rewriter_before_opt_a]: 0.0001403 [opt_a]: 0.240521, [3] [Cycle 1]: 0.236362, [45] [expand_dump_flag]: 3.88001e-06 [switch_simplify]: 7.102e-05 [loop_unroll]: 5.596e-05 [a_1]: 0.00133219 [with_stream_mark]: 3.193e-05 [recompute_prepare]: 2.534e-05 [updatestate_depend_eliminate]: 1.262e-05 [updatestate_assign_eliminate]: 1.077e-05 [updatestate_loads_eliminate]: 1.023e-05 [parameter_eliminate]: 2.76e-06 [a_2]: 0.00028748 [accelerated_algorithm]: 1.888e-05 [shard]: 2.87002e-06 [meta_shard_fg_expand]: 4.50999e-06 [shard_inline]: 1.778e-05 [merge_send_recv]: 2.032e-05 [auto_parallel]: 1.428e-05 [parallel]: 2.78e-05 [flash_sp]: 1.279e-05 [merge_comm]: 1.193e-05 [allreduce_fusion]: 1.116e-05 [matmul_add_comm_reduction]: 3.345e-05 [allreduce_slice_to_reducescatter]: 9.50007e-07 [virtual_shard_identity]: 2.149e-05 [virtual_dataset]: 1.87e-05 [get_grad_eliminate_]: 1.769e-05 [virtual_output]: 1.824e-05 [merge_forward]: 1.228e-05 [cell_reuse_recompute_pass]: 1.30999e-06 [offload_activation]: 2.274e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.71e-05 [merge_recompute_call_nodes]: 1.45001e-06 [before_grad]: 3.312e-05 [set_forward_comm_id_for_comm_node_pass]: 1.197e-05 [meta_fg_expand]: 0.00580681 [flash_sp_send_recv_attached]: 5.17999e-06 [receive_attached]: 2.32001e-06 [after_resolve]: 0.00011562 [a_after_grad]: 0.220639 [renormalize]: 0.00600971 [add_forward_monad_depend]: 1.808e-05 [auto_monad_grad]: 8.29998e-06 [auto_monad_eliminator]: 8.022e-05 [cse]: 0.00023739 [a_3]: 0.00084892 [Cycle 2]: 0.00328217, [45] [expand_dump_flag]: 3.63999e-06 [switch_simplify]: 6.657e-05 [loop_unroll]: 5.175e-05 [a_1]: 0.00128971 [with_stream_mark]: 2.877e-05 [recompute_prepare]: 1.348e-05 [updatestate_depend_eliminate]: 3.53e-05 [updatestate_assign_eliminate]: 6.39001e-06 [updatestate_loads_eliminate]: 4.82e-06 [parameter_eliminate]: 2.34001e-06 [a_2]: 0.00011892 [accelerated_algorithm]: 9.79999e-06 [shard]: 2.19999e-06 [meta_shard_fg_expand]: 2.93998e-06 [shard_inline]: 8.79e-06 [merge_send_recv]: 1.107e-05 [auto_parallel]: 1.29e-05 [parallel]: 9.71998e-06 [flash_sp]: 4.58001e-06 [merge_comm]: 5.75001e-06 [allreduce_fusion]: 5.17e-06 [matmul_add_comm_reduction]: 1.315e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 1.158e-05 [virtual_dataset]: 8.13001e-06 [get_grad_eliminate_]: 8.17998e-06 [virtual_output]: 8.23001e-06 [merge_forward]: 5.34e-06 [cell_reuse_recompute_pass]: 1.32999e-06 [offload_activation]: 1.385e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.812e-05 [merge_recompute_call_nodes]: 1.67999e-06 [before_grad]: 1.461e-05 [set_forward_comm_id_for_comm_node_pass]: 5.66998e-06 [meta_fg_expand]: 6.633e-05 [flash_sp_send_recv_attached]: 2.09e-06 [receive_attached]: 3.14999e-06 [after_resolve]: 1.366e-05 [a_after_grad]: 1.285e-05 [renormalize]: 0.00099707 [add_forward_monad_depend]: 5.96e-06 [auto_monad_grad]: 1.93997e-06 [auto_monad_eliminator]: 2.118e-05 [cse]: 3.066e-05 [a_3]: 6.294e-05 [Cycle 3]: 0.00085911, [45] [expand_dump_flag]: 1.84e-06 [switch_simplify]: 1.039e-05 [loop_unroll]: 8.25e-06 [a_1]: 0.00019692 [with_stream_mark]: 1.335e-05 [recompute_prepare]: 8.75999e-06 [updatestate_depend_eliminate]: 5.35999e-06 [updatestate_assign_eliminate]: 5.14e-06 [updatestate_loads_eliminate]: 4.38999e-06 [parameter_eliminate]: 1.19003e-06 [a_2]: 0.00011532 [accelerated_algorithm]: 9.75002e-06 [shard]: 2.16998e-06 [meta_shard_fg_expand]: 2.68e-06 [shard_inline]: 8.27e-06 [merge_send_recv]: 7.66999e-06 [auto_parallel]: 8.77999e-06 [parallel]: 5.94999e-06 [flash_sp]: 1.40999e-06 [merge_comm]: 5.35001e-06 [allreduce_fusion]: 4.92e-06 [matmul_add_comm_reduction]: 8.60999e-06 [allreduce_slice_to_reducescatter]: 4.90021e-07 [virtual_shard_identity]: 1.01e-05 [virtual_dataset]: 8.23001e-06 [get_grad_eliminate_]: 8e-06 [virtual_output]: 7.83001e-06 [merge_forward]: 4.12e-06 [cell_reuse_recompute_pass]: 1.60999e-06 [offload_activation]: 9.87999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.69e-05 [merge_recompute_call_nodes]: 8.69972e-07 [before_grad]: 1.451e-05 [set_forward_comm_id_for_comm_node_pass]: 5.54e-06 [meta_fg_expand]: 3.28e-06 [flash_sp_send_recv_attached]: 1.05999e-06 [receive_attached]: 1.17e-06 [after_resolve]: 1.088e-05 [a_after_grad]: 1.239e-05 [renormalize]: 9.00181e-08 [add_forward_monad_depend]: 1.35001e-06 [auto_monad_grad]: 1.35999e-06 [auto_monad_eliminator]: 1.249e-05 [cse]: 2.357e-05 [a_3]: 5.23e-05 [py_interpret_to_execute_after_opt_a]: 1.55e-05 [slice_cell_reuse_recomputed_activation]: 2.19001e-06 [rewriter_after_opt_a]: 6.22e-05 [convert_after_rewriter]: 8.85001e-06 [order_py_execute_after_rewriter]: 6.31998e-06 [mutable_eliminate]: 0.00077323 [opt_b]: 0.0002875, [1] [Cycle 1]: 0.00028018, [7] [b_1]: 0.00018834 [b_2]: 1.021e-05 [updatestate_depend_eliminate]: 9.10999e-06 [updatestate_assign_eliminate]: 4.47e-06 [updatestate_loads_eliminate]: 4.23999e-06 [renormalize]: 7.99977e-07 [cse]: 2.634e-05 [optimize_parallel_all_gather_comm]: 2.159e-05 [overlap_param_gather]: 7.26999e-06 [cconv]: 2.875e-05 [loop_unroll]: 0.00045376 [opt_after_cconv]: 0.00013863, [1] [Cycle 1]: 0.00013277, [7] [c_1]: 5.267e-05 [parameter_eliminate]: 2.93998e-06 [updatestate_depend_eliminate]: 7.86001e-06 [updatestate_assign_eliminate]: 4.35e-06 [updatestate_loads_eliminate]: 4.43001e-06 [cse]: 2.541e-05 [renormalize]: 5.79981e-07 [remove_dup_value]: 1.736e-05 [tuple_transform]: 8.854e-05, [1] [Cycle 1]: 8.365e-05, [4] [d_1]: 5.559e-05 [none_parameter_eliminate]: 1.67999e-06 [renormalize]: 1.90019e-07 [switch_simplify]: 9.02e-06 [partial_unused_args_eliminate]: 1.77001e-06 [add_recomputation]: 8.573e-05 [cse_after_recomputation]: 3.194e-05, [1] [Cycle 1]: 2.708e-05, [1] [cse]: 2.094e-05 [environ_conv]: 6.74999e-06 [swap_dp_allreduce_reducescatter]: 8.13999e-06 [bias_add_comm_swap]: 2.58e-06 [label_micro_interleaved_index]: 4.68999e-06 [label_fine_grained_interleaved_index]: 2.73e-06 [merge_cast_opt]: 1.52999e-06 [slice_recompute_activation]: 2.51e-06 [micro_interleaved_order_control]: 2.16e-06 [assign_add_opt]: 1.56998e-06 [ForceFp32Comm]: 8.39995e-07 [remove_cast_before_assign_add]: 1.23002e-06 [full_micro_interleaved_order_control]: 2.02001e-06 [reorder_send_recv_between_fp_bp]: 2.79999e-06 [comm_op_add_attrs]: 1.07e-06 [add_comm_op_reuse_tag]: 1.01002e-06 [interleave_split_concat_branches]: 1.14e-06 [interleave_parallel_branches]: 1.10001e-06 [overlap_opt_shard_in_pipeline]: 4.55001e-06 [overlap_opt_shard_grad_in_pipeline]: 1.92001e-06 [control_data_broadcast_order]: 1.783e-05 [grouped_pairwise_exchange_alltoall]: 1.60001e-06 [offloading_packed_experts]: 5.14e-06 [overlap_recompute_and_grad_model_parallel]: 5.69999e-06 [overlap_grad_matmul_and_grad_allreduce]: 6.605e-05 [overlap_recompute_allgather_and_fa_grad]: 1.63002e-06 [overlap_recompute_comm]: 2.74001e-06 [overlap_grad_ring_attention]: 7.31001e-06 [overlap_grad_flash_sp]: 2.963e-05 [begin_end_overlap_inline]: 7.39994e-07 [split_matmul_comm_elemetwise]: 2.32001e-06 [split_layernorm_comm]: 1.73002e-06 [handle_group_info]: 9.50007e-07 [symbol_engine_optimizer]: 9.672e-05, [1] [Cycle 1]: 9.17e-05, [6] [build]: 4.68001e-06 [elim_shapecalc]: 1.441e-05 [elim_not_effective]: 1.826e-05 [opt_reshape]: 9.82999e-06 [fold_const_symbol]: 1.416e-05 [renormalize]: 2.59985e-07 [detach_backward]: 2.22999e-06 [pipeline_parallel_scheduler]: 1.45999e-06 [auto_monad_reorder]: 6.366e-05 [get_jit_bprop_graph]: 1.91e-06 [rewriter_after_jit_bprop_graph]: 4.67e-06 [opt_after_jit_grad]: 0.00050698 [validate]: 5.167e-05 [backend_pass]: 8.99978e-07 [task_emit]: 0.0081737 [execute]: 8.22998e-06 Sums bootstrap : 0.000501s : 0.11% type_inference : 0.184711s : 42.32% event_method : 0.000053s : 0.01% auto_monad : 0.000241s : 0.06% graph_reusing : 0.000009s : 0.00% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000042s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000010s : 0.00% parallel-infer-symbol : 0.000005s : 0.00% pre_auto_parallel : 0.000054s : 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 : 0.000040s : 0.01% optimize.rewriter_before_opt_a : 0.000140s : 0.03% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000148s : 0.03% optimize.opt_a.loop_unroll : 0.000116s : 0.03% optimize.opt_a.a_1 : 0.002819s : 0.65% optimize.opt_a.with_stream_mark : 0.000074s : 0.02% optimize.opt_a.recompute_prepare : 0.000048s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000053s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000022s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000019s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000522s : 0.12% optimize.opt_a.accelerated_algorithm : 0.000038s : 0.01% optimize.opt_a.shard : 0.000007s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000035s : 0.01% optimize.opt_a.merge_send_recv : 0.000039s : 0.01% optimize.opt_a.auto_parallel : 0.000036s : 0.01% optimize.opt_a.parallel : 0.000043s : 0.01% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000023s : 0.01% optimize.opt_a.allreduce_fusion : 0.000021s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000055s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000043s : 0.01% optimize.opt_a.virtual_dataset : 0.000035s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000034s : 0.01% optimize.opt_a.virtual_output : 0.000034s : 0.01% optimize.opt_a.merge_forward : 0.000022s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000046s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000072s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000062s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000023s : 0.01% optimize.opt_a.meta_fg_expand : 0.005876s : 1.35% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000140s : 0.03% optimize.opt_a.a_after_grad : 0.220664s : 50.56% optimize.opt_a.renormalize : 0.007007s : 1.61% optimize.opt_a.add_forward_monad_depend : 0.000025s : 0.01% optimize.opt_a.auto_monad_grad : 0.000012s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000114s : 0.03% optimize.opt_a.cse : 0.000292s : 0.07% optimize.opt_a.a_3 : 0.000964s : 0.22% optimize.py_interpret_to_execute_after_opt_a : 0.000016s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000062s : 0.01% optimize.convert_after_rewriter : 0.000009s : 0.00% optimize.order_py_execute_after_rewriter : 0.000006s : 0.00% optimize.mutable_eliminate : 0.000773s : 0.18% optimize.opt_b.b_1 : 0.000188s : 0.04% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 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 : 0.000026s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.00% optimize.overlap_param_gather : 0.000007s : 0.00% optimize.cconv : 0.000029s : 0.01% optimize.loop_unroll : 0.000454s : 0.10% optimize.opt_after_cconv.c_1 : 0.000053s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 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 : 0.000025s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000017s : 0.00% optimize.tuple_transform.d_1 : 0.000056s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000086s : 0.02% optimize.cse_after_recomputation.cse : 0.000021s : 0.00% 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 : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000003s : 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 : 0.000001s : 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 : 0.000005s : 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 : 0.000066s : 0.02% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000007s : 0.00% optimize.overlap_grad_flash_sp : 0.000030s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 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 : 0.000064s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000507s : 0.12% validate : 0.000052s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.008174s : 1.87% execute : 0.000008s : 0.00% Time group info: ------[substitution.] 0.001611 224 0.84% : 0.000013s : 4: substitution.depend_value_elim 0.18% : 0.000003s : 5: substitution.elim_not_effective 0.17% : 0.000003s : 3: substitution.float_depend_g_call 0.28% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.14% : 0.000002s : 5: substitution.fold_const_symbol 45.72% : 0.000737s : 4: substitution.getattr_setattr_resolve 0.48% : 0.000008s : 6: substitution.graph_param_transform 0.18% : 0.000003s : 2: substitution.incorporate_call 0.18% : 0.000003s : 2: substitution.incorporate_call_switch 29.43% : 0.000474s : 13: substitution.inline 4.40% : 0.000071s : 5: substitution.inline_without_move 0.80% : 0.000013s : 23: substitution.j_node_and_user_rematch 0.67% : 0.000011s : 6: substitution.minmaximum_grad 0.23% : 0.000004s : 3: substitution.partial_eliminate 1.03% : 0.000017s : 23: substitution.remove_not_recompute_node 3.47% : 0.000056s : 21: substitution.replace_applicator 0.80% : 0.000013s : 16: substitution.replace_old_param 0.19% : 0.000003s : 1: substitution.set_cell_output_no_recompute 1.32% : 0.000021s : 6: substitution.tuple_list_convert_item_index_to_positive 0.62% : 0.000010s : 6: substitution.tuple_list_get_item_const_eliminator 0.83% : 0.000013s : 6: substitution.tuple_list_get_item_depend_reorder 3.16% : 0.000051s : 15: substitution.tuple_list_get_item_eliminator 0.83% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 1.58% : 0.000025s : 18: substitution.updatestate_pure_node_eliminater 2.48% : 0.000040s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.184617 2 99.23% : 0.183187s : 1: type_inference.infer 0.77% : 0.001429s : 1: type_inference.specialize ------[replace.] 0.000317 25 16.16% : 0.000051s : 3: replace.getattr_setattr_resolve 35.67% : 0.000113s : 13: replace.inline 15.20% : 0.000048s : 1: replace.replace_applicator 24.72% : 0.000078s : 7: replace.tuple_list_get_item_eliminator 8.24% : 0.000026s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001206 25 57.49% : 0.000693s : 3: match.getattr_setattr_resolve 38.70% : 0.000467s : 13: match.inline 0.93% : 0.000011s : 1: match.replace_applicator 1.86% : 0.000022s : 7: match.tuple_list_get_item_eliminator 1.03% : 0.000012s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000800 5424 0.90% : 0.000007s : 55: predicate.accumulaten_eliminater 0.35% : 0.000003s : 6: predicate.ad_related_special_op_eliminate 0.46% : 0.000004s : 29: predicate.addn_check_dump 0.87% : 0.000007s : 55: predicate.addn_zero_filter 0.87% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 1.75% : 0.000014s : 84: predicate.arithmetic_simplify 0.95% : 0.000008s : 55: predicate.cast_eliminate 2.04% : 0.000016s : 130: predicate.check_bprop_eliminate 0.50% : 0.000004s : 29: predicate.compare_switch_simplify 0.07% : 0.000001s : 6: predicate.const_output_eliminate 0.52% : 0.000004s : 29: predicate.depend_value_elim 0.95% : 0.000008s : 55: predicate.dict_get_item_const_eliminator 1.00% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.96% : 0.000008s : 55: predicate.dict_set_item_eliminator 0.35% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.08% : 0.000001s : 6: predicate.elim_not_effective 0.14% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.03% : 0.000008s : 61: predicate.environ_add_const_eliminate 0.99% : 0.000008s : 61: predicate.environ_get_add_eliminate 0.99% : 0.000008s : 61: predicate.environ_get_depend_swap 1.69% : 0.000014s : 90: predicate.environ_get_eliminate 0.98% : 0.000008s : 61: predicate.environ_get_set_eliminate 1.30% : 0.000010s : 76: predicate.exchange_switch_depend_value 1.95% : 0.000016s : 76: predicate.float_depend_g_call 0.49% : 0.000004s : 29: predicate.float_environ_get_switch 0.60% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.06% : 0.000000s : 6: predicate.fold_const_symbol 0.56% : 0.000004s : 29: predicate.get_grad_eliminate 0.71% : 0.000006s : 20: predicate.getattr_setattr_resolve 0.10% : 0.000001s : 6: predicate.graph_param_transform 0.54% : 0.000004s : 29: predicate.incorporate_call 0.47% : 0.000004s : 29: predicate.incorporate_call_switch 4.67% : 0.000037s : 201: predicate.inline 5.22% : 0.000042s : 97: predicate.inline_without_move 0.25% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.67% : 0.000005s : 29: predicate.less_batch_normalization 1.29% : 0.000010s : 74: predicate.list_to_tuple_eliminator_ 2.08% : 0.000017s : 129: predicate.load_eliminater 0.33% : 0.000003s : 6: predicate.loop_unroll_after_grad 2.10% : 0.000017s : 129: predicate.loop_unroll_before_grad 1.18% : 0.000009s : 67: predicate.make_slice_get_slice_eliminator 0.51% : 0.000004s : 29: predicate.merge_addn 1.98% : 0.000016s : 126: predicate.micro_step_allgather_replace 2.00% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.87% : 0.000007s : 55: predicate.minmaximum_grad 0.38% : 0.000003s : 6: predicate.mutable_eliminate 0.18% : 0.000001s : 6: predicate.opt_reshape 0.14% : 0.000001s : 6: predicate.parallel_virtual_node 1.64% : 0.000013s : 76: predicate.partial_defer_inline 1.28% : 0.000010s : 68: predicate.partial_eliminate 0.89% : 0.000007s : 55: predicate.print_const_string_wrapper 0.53% : 0.000004s : 29: predicate.reduce_all_const_elim 1.17% : 0.000009s : 55: predicate.reduce_eliminate 2.10% : 0.000017s : 129: predicate.redundant_stop_gradient_eliminater 0.36% : 0.000003s : 29: predicate.remove_not_recompute_node 2.27% : 0.000018s : 190: predicate.replace_applicator 0.98% : 0.000008s : 97: predicate.replace_old_param 0.11% : 0.000001s : 6: predicate.reset_defer_inline 0.94% : 0.000007s : 55: predicate.reshape_eliminate 2.05% : 0.000016s : 126: predicate.row_tensor_add_zeros_like 0.13% : 0.000001s : 6: predicate.row_tensor_eliminate 2.24% : 0.000018s : 130: predicate.same_eliminate 0.32% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.64% : 0.000005s : 29: predicate.shard_identity_eliminate 0.28% : 0.000002s : 12: predicate.special_op_eliminate 0.60% : 0.000005s : 29: predicate.specialize_transform 2.26% : 0.000018s : 126: predicate.split_environ_get_set_with_tuple_value 2.45% : 0.000020s : 97: predicate.stack_unstack_eliminate 0.15% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.41% : 0.000011s : 76: predicate.switch_defer_inline 3.36% : 0.000027s : 206: predicate.switch_layer_defer_inline 4.38% : 0.000035s : 240: predicate.switch_simplify 0.89% : 0.000007s : 55: predicate.tile_eliminate 0.92% : 0.000007s : 55: predicate.transpose_eliminate 1.22% : 0.000010s : 67: predicate.tuple_list_convert_item_index_to_positive 1.26% : 0.000010s : 67: predicate.tuple_list_get_item_const_eliminator 1.15% : 0.000009s : 67: predicate.tuple_list_get_item_depend_reorder 2.31% : 0.000018s : 103: predicate.tuple_list_get_item_eliminator 1.15% : 0.000009s : 67: predicate.tuple_list_get_set_item_eliminator 1.75% : 0.000014s : 96: predicate.tuple_list_set_item_eliminator 1.24% : 0.000010s : 74: predicate.tuple_to_list_eliminator_ 2.18% : 0.000017s : 129: predicate.updatestate_pure_node_eliminater 2.86% : 0.000023s : 159: predicate.updatestate_useless_node_eliminater 0.19% : 0.000002s : 6: predicate.value_based_eliminate 0.52% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.52% : 0.000004s : 29: predicate.virtual_output_eliminate 0.12% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.14% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.170545 48 99.46% : 0.169620s : 24: func_graph_cloner_run.FuncGraphClonerGraph 0.54% : 0.000926s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.922343 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.40% : 0.003731s : 1: add_attr 0.40% : 0.003718s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000090s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.03% : 0.000250s : 1: auto_monad 0.01% : 0.000069s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.06% : 0.000535s : 1: bootstrap 0.00% : 0.000032s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000021s : 1: control_data_broadcast_order 0.00% : 0.000012s : 1: convert_after_rewriter 0.00% : 0.000035s : 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% : 0.000061s : 1: event_method 0.00% : 0.000014s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000014s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000006s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000004s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.05% : 0.000462s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.08% : 0.000782s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000018s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 24.46% : 0.225619s : 125: opt.transform.opt_a 0.01% : 0.000051s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000170s : 28: opt.transform.opt_b 0.09% : 0.000838s : 2: opt.transform.opt_resolve 0.01% : 0.000063s : 2: opt.transform.opt_trans_graph 0.01% : 0.000053s : 4: opt.transform.symbol_engine_opt 26.08% : 0.240524s : 1: opt_a 0.02% : 0.000142s : 1: opt_after_cconv 0.06% : 0.000516s : 1: opt_after_jit_grad 0.03% : 0.000291s : 1: opt_b 26.37% : 0.243257s : 1: optimize 0.00% : 0.000025s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000009s : 1: order_py_execute_after_rewriter 0.00% : 0.000033s : 1: overlap_grad_flash_sp 0.01% : 0.000072s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000010s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000008s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000011s : 1: overlap_param_gather 0.00% : 0.000005s : 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 0.01% : 0.000058s : 1: pre_auto_parallel 0.00% : 0.000044s : 1: py_interpret_to_execute 0.00% : 0.000019s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000021s : 1: remove_dup_value 0.45% : 0.004169s : 2: renormalize.infer 0.30% : 0.002813s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000066s : 1: rewriter_after_opt_a 0.02% : 0.000145s : 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% : 0.000100s : 1: symbol_engine_optimizer 0.89% : 0.008191s : 1: task_emit 0.01% : 0.000091s : 1: tuple_transform 20.03% : 0.184735s : 1: type_inference 0.01% : 0.000080s : 1: validate TotalTime = 1.17026, [24] [bootstrap]: 0.00042936 [type_inference]: 0.467577 [event_method]: 5.759e-05 [auto_monad]: 0.00023729 [graph_reusing]: 8.27e-06 [inline]: 2.87002e-06 [add_attr]: 0.00348212, [1] [add_attr_with_inline]: 0.0034726, [1] [Cycle 1]: 8.283e-05, [2] [tag_attr]: 3.993e-05 [meta_addattr_fg_expand]: 9.30001e-06 [parallel-infer-symbol]: 3.89002e-06 [pre_auto_parallel]: 5.263e-05 [insert-virtual-dataset]: 2.77002e-06 [parallel-infer-symbol-second]: 8.39995e-07 [dataset_repeat_opt]: 2.34999e-06 [pipeline_split]: 1.64e-06 [optimize]: 0.688634, [53] [py_interpret_to_execute]: 4.177e-05 [rewriter_before_opt_a]: 0.00013853 [opt_a]: 0.36963, [3] [Cycle 1]: 0.365404, [45] [expand_dump_flag]: 3.75e-06 [switch_simplify]: 7.03e-05 [loop_unroll]: 5.653e-05 [a_1]: 0.0013005 [with_stream_mark]: 4.738e-05 [recompute_prepare]: 3.024e-05 [updatestate_depend_eliminate]: 1.305e-05 [updatestate_assign_eliminate]: 1.131e-05 [updatestate_loads_eliminate]: 1.02e-05 [parameter_eliminate]: 2.93e-06 [a_2]: 0.00030318 [accelerated_algorithm]: 2.152e-05 [shard]: 2.11e-06 [meta_shard_fg_expand]: 5.00001e-06 [shard_inline]: 1.887e-05 [merge_send_recv]: 1.998e-05 [auto_parallel]: 1.446e-05 [parallel]: 1.907e-05 [flash_sp]: 1.251e-05 [merge_comm]: 1.188e-05 [allreduce_fusion]: 1.105e-05 [matmul_add_comm_reduction]: 3.246e-05 [allreduce_slice_to_reducescatter]: 8.00006e-07 [virtual_shard_identity]: 2.147e-05 [virtual_dataset]: 1.848e-05 [get_grad_eliminate_]: 1.801e-05 [virtual_output]: 1.816e-05 [merge_forward]: 1.24e-05 [cell_reuse_recompute_pass]: 1.23002e-06 [offload_activation]: 2.284e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.729e-05 [merge_recompute_call_nodes]: 1.68002e-06 [before_grad]: 3.248e-05 [set_forward_comm_id_for_comm_node_pass]: 1.276e-05 [meta_fg_expand]: 0.353218 [flash_sp_send_recv_attached]: 5.59e-06 [receive_attached]: 2.53e-06 [after_resolve]: 0.00012127 [a_after_grad]: 0.00018235 [renormalize]: 0.00814566 [add_forward_monad_depend]: 1.484e-05 [auto_monad_grad]: 8.95999e-06 [auto_monad_eliminator]: 8.392e-05 [cse]: 0.00023325 [a_3]: 0.00083125 [Cycle 2]: 0.00337411, [45] [expand_dump_flag]: 3.70998e-06 [switch_simplify]: 7.149e-05 [loop_unroll]: 5.234e-05 [a_1]: 0.00124731 [with_stream_mark]: 2.31e-05 [recompute_prepare]: 1.275e-05 [updatestate_depend_eliminate]: 3.382e-05 [updatestate_assign_eliminate]: 7.45e-06 [updatestate_loads_eliminate]: 5.09998e-06 [parameter_eliminate]: 2.14e-06 [a_2]: 0.00017585 [accelerated_algorithm]: 1.426e-05 [shard]: 2.08002e-06 [meta_shard_fg_expand]: 3.3e-06 [shard_inline]: 9.00999e-06 [merge_send_recv]: 1.258e-05 [auto_parallel]: 1.287e-05 [parallel]: 9.66003e-06 [flash_sp]: 4.59002e-06 [merge_comm]: 5.21002e-06 [allreduce_fusion]: 5.80002e-06 [matmul_add_comm_reduction]: 1.277e-05 [allreduce_slice_to_reducescatter]: 6.09987e-07 [virtual_shard_identity]: 1.154e-05 [virtual_dataset]: 8.27998e-06 [get_grad_eliminate_]: 7.91001e-06 [virtual_output]: 8.02e-06 [merge_forward]: 5.87999e-06 [cell_reuse_recompute_pass]: 1.36002e-06 [offload_activation]: 1.309e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.788e-05 [merge_recompute_call_nodes]: 1.50001e-06 [before_grad]: 1.434e-05 [set_forward_comm_id_for_comm_node_pass]: 5.83997e-06 [meta_fg_expand]: 5.572e-05 [flash_sp_send_recv_attached]: 1.20999e-06 [receive_attached]: 2.41e-06 [after_resolve]: 1.385e-05 [a_after_grad]: 1.313e-05 [renormalize]: 0.00109125 [add_forward_monad_depend]: 4.85999e-06 [auto_monad_grad]: 2.00002e-06 [auto_monad_eliminator]: 1.946e-05 [cse]: 2.72e-05 [a_3]: 6.23e-05 [Cycle 3]: 0.00083386, [45] [expand_dump_flag]: 1.15999e-06 [switch_simplify]: 1.029e-05 [loop_unroll]: 8.17998e-06 [a_1]: 0.00019658 [with_stream_mark]: 1.061e-05 [recompute_prepare]: 8.67e-06 [updatestate_depend_eliminate]: 5.29e-06 [updatestate_assign_eliminate]: 4.25e-06 [updatestate_loads_eliminate]: 4.72e-06 [parameter_eliminate]: 9.29984e-07 [a_2]: 0.00011171 [accelerated_algorithm]: 8.21002e-06 [shard]: 1.19e-06 [meta_shard_fg_expand]: 1.92999e-06 [shard_inline]: 8.57e-06 [merge_send_recv]: 6.68e-06 [auto_parallel]: 8.10999e-06 [parallel]: 4.60001e-06 [flash_sp]: 1.17999e-06 [merge_comm]: 5.07e-06 [allreduce_fusion]: 4.74e-06 [matmul_add_comm_reduction]: 7.58999e-06 [allreduce_slice_to_reducescatter]: 3.89991e-07 [virtual_shard_identity]: 9.22999e-06 [virtual_dataset]: 7.73999e-06 [get_grad_eliminate_]: 7.41999e-06 [virtual_output]: 8.24998e-06 [merge_forward]: 4.58999e-06 [cell_reuse_recompute_pass]: 1.30999e-06 [offload_activation]: 9.92999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.599e-05 [merge_recompute_call_nodes]: 7.2e-07 [before_grad]: 1.354e-05 [set_forward_comm_id_for_comm_node_pass]: 5.47999e-06 [meta_fg_expand]: 3.54002e-06 [flash_sp_send_recv_attached]: 1.27e-06 [receive_attached]: 9.20001e-07 [after_resolve]: 2.411e-05 [a_after_grad]: 1.287e-05 [renormalize]: 9.99717e-08 [add_forward_monad_depend]: 1.22999e-06 [auto_monad_grad]: 1.04e-06 [auto_monad_eliminator]: 1.308e-05 [cse]: 2.116e-05 [a_3]: 5.165e-05 [py_interpret_to_execute_after_opt_a]: 1.623e-05 [slice_cell_reuse_recomputed_activation]: 2.32999e-06 [rewriter_after_opt_a]: 0.00014483 [convert_after_rewriter]: 1.214e-05 [order_py_execute_after_rewriter]: 8.57e-06 [mutable_eliminate]: 0.00084149 [opt_b]: 0.0003423, [1] [Cycle 1]: 0.00033218, [7] [b_1]: 0.00020777 [b_2]: 1.043e-05 [updatestate_depend_eliminate]: 1.036e-05 [updatestate_assign_eliminate]: 5.52999e-06 [updatestate_loads_eliminate]: 5.42001e-06 [renormalize]: 7.89994e-07 [cse]: 5.246e-05 [optimize_parallel_all_gather_comm]: 2.936e-05 [overlap_param_gather]: 2.04999e-06 [cconv]: 3.473e-05 [loop_unroll]: 0.00045981 [opt_after_cconv]: 0.00014013, [1] [Cycle 1]: 0.00013417, [7] [c_1]: 5.211e-05 [parameter_eliminate]: 3.73999e-06 [updatestate_depend_eliminate]: 7.65e-06 [updatestate_assign_eliminate]: 4.35999e-06 [updatestate_loads_eliminate]: 4.1e-06 [cse]: 2.665e-05 [renormalize]: 5.89993e-07 [remove_dup_value]: 1.797e-05 [tuple_transform]: 9.311e-05, [1] [Cycle 1]: 8.8e-05, [4] [d_1]: 5.792e-05 [none_parameter_eliminate]: 1.66998e-06 [renormalize]: 1.8999e-07 [switch_simplify]: 9.51e-06 [partial_unused_args_eliminate]: 2.26e-06 [add_recomputation]: 0.00011288 [cse_after_recomputation]: 3.384e-05, [1] [Cycle 1]: 2.844e-05, [1] [cse]: 2.216e-05 [environ_conv]: 7.83001e-06 [swap_dp_allreduce_reducescatter]: 8.14002e-06 [bias_add_comm_swap]: 3.46001e-06 [label_micro_interleaved_index]: 4.62e-06 [label_fine_grained_interleaved_index]: 2.74999e-06 [merge_cast_opt]: 1.44e-06 [slice_recompute_activation]: 2.37001e-06 [micro_interleaved_order_control]: 2.56e-06 [assign_add_opt]: 1.39998e-06 [ForceFp32Comm]: 1.04e-06 [remove_cast_before_assign_add]: 8.10018e-07 [full_micro_interleaved_order_control]: 2.78e-06 [reorder_send_recv_between_fp_bp]: 3.22002e-06 [comm_op_add_attrs]: 1.17999e-06 [add_comm_op_reuse_tag]: 1.40999e-06 [interleave_split_concat_branches]: 1.19e-06 [interleave_parallel_branches]: 1.10001e-06 [overlap_opt_shard_in_pipeline]: 1.45999e-06 [overlap_opt_shard_grad_in_pipeline]: 2.19999e-06 [control_data_broadcast_order]: 1.885e-05 [grouped_pairwise_exchange_alltoall]: 1.67001e-06 [offloading_packed_experts]: 6.38998e-06 [overlap_recompute_and_grad_model_parallel]: 6.61e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.14e-06 [overlap_recompute_allgather_and_fa_grad]: 1.45999e-06 [overlap_recompute_comm]: 2.34001e-06 [overlap_grad_ring_attention]: 5.64998e-06 [overlap_grad_flash_sp]: 3.308e-05 [begin_end_overlap_inline]: 5.69999e-07 [split_matmul_comm_elemetwise]: 2.37999e-06 [split_layernorm_comm]: 1.84e-06 [handle_group_info]: 1.04e-06 [symbol_engine_optimizer]: 9.499e-05, [1] [Cycle 1]: 9.041e-05, [6] [build]: 5.17999e-06 [elim_shapecalc]: 1.329e-05 [elim_not_effective]: 1.857e-05 [opt_reshape]: 9.41e-06 [fold_const_symbol]: 1.449e-05 [renormalize]: 2.10013e-07 [detach_backward]: 2.51e-06 [pipeline_parallel_scheduler]: 1.64e-06 [auto_monad_reorder]: 8.496e-05 [get_jit_bprop_graph]: 2.17999e-06 [rewriter_after_jit_bprop_graph]: 5.17e-06 [opt_after_jit_grad]: 0.00049153 [validate]: 5.497e-05 [backend_pass]: 1.80001e-06 [task_emit]: 0.00858638 [execute]: 9.06002e-06 Sums bootstrap : 0.000429s : 0.05% type_inference : 0.467577s : 55.07% event_method : 0.000058s : 0.01% auto_monad : 0.000237s : 0.03% graph_reusing : 0.000008s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000040s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000009s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000053s : 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 : 0.000042s : 0.00% optimize.rewriter_before_opt_a : 0.000139s : 0.02% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000152s : 0.02% optimize.opt_a.loop_unroll : 0.000117s : 0.01% optimize.opt_a.a_1 : 0.002744s : 0.32% optimize.opt_a.with_stream_mark : 0.000081s : 0.01% optimize.opt_a.recompute_prepare : 0.000052s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000052s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000023s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000020s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000591s : 0.07% optimize.opt_a.accelerated_algorithm : 0.000044s : 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 : 0.000036s : 0.00% optimize.opt_a.merge_send_recv : 0.000039s : 0.00% optimize.opt_a.auto_parallel : 0.000035s : 0.00% optimize.opt_a.parallel : 0.000033s : 0.00% optimize.opt_a.flash_sp : 0.000018s : 0.00% optimize.opt_a.merge_comm : 0.000022s : 0.00% optimize.opt_a.allreduce_fusion : 0.000022s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000053s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000042s : 0.00% optimize.opt_a.virtual_dataset : 0.000034s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000033s : 0.00% optimize.opt_a.virtual_output : 0.000034s : 0.00% optimize.opt_a.merge_forward : 0.000023s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000046s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000071s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000060s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000024s : 0.00% optimize.opt_a.meta_fg_expand : 0.353277s : 41.61% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000159s : 0.02% optimize.opt_a.a_after_grad : 0.000208s : 0.02% optimize.opt_a.renormalize : 0.009237s : 1.09% optimize.opt_a.add_forward_monad_depend : 0.000021s : 0.00% optimize.opt_a.auto_monad_grad : 0.000012s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000116s : 0.01% optimize.opt_a.cse : 0.000282s : 0.03% optimize.opt_a.a_3 : 0.000945s : 0.11% optimize.py_interpret_to_execute_after_opt_a : 0.000016s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000145s : 0.02% optimize.convert_after_rewriter : 0.000012s : 0.00% optimize.order_py_execute_after_rewriter : 0.000009s : 0.00% optimize.mutable_eliminate : 0.000841s : 0.10% optimize.opt_b.b_1 : 0.000208s : 0.02% 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 : 0.000006s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000052s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000029s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000035s : 0.00% optimize.loop_unroll : 0.000460s : 0.05% optimize.opt_after_cconv.c_1 : 0.000052s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 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 : 0.000027s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000018s : 0.00% optimize.tuple_transform.d_1 : 0.000058s : 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 : 0.000010s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000113s : 0.01% optimize.cse_after_recomputation.cse : 0.000022s : 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 : 0.000005s : 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 : 0.000001s : 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 : 0.000019s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000033s : 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 : 0.000019s : 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 : 0.000002s : 0.00% auto_monad_reorder : 0.000085s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000492s : 0.06% validate : 0.000055s : 0.01% backend_pass : 0.000002s : 0.00% task_emit : 0.008586s : 1.01% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.141537 224 0.01% : 0.000013s : 4: substitution.depend_value_elim 0.00% : 0.000003s : 5: substitution.elim_not_effective 0.00% : 0.000003s : 3: substitution.float_depend_g_call 0.00% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.00% : 0.000002s : 5: substitution.fold_const_symbol 99.44% : 0.140748s : 4: substitution.getattr_setattr_resolve 0.01% : 0.000008s : 6: substitution.graph_param_transform 0.00% : 0.000003s : 2: substitution.incorporate_call 0.00% : 0.000002s : 2: substitution.incorporate_call_switch 0.29% : 0.000412s : 13: substitution.inline 0.03% : 0.000038s : 5: substitution.inline_without_move 0.01% : 0.000012s : 23: substitution.j_node_and_user_rematch 0.01% : 0.000011s : 6: substitution.minmaximum_grad 0.00% : 0.000003s : 3: substitution.partial_eliminate 0.01% : 0.000017s : 23: substitution.remove_not_recompute_node 0.04% : 0.000054s : 21: substitution.replace_applicator 0.02% : 0.000028s : 16: substitution.replace_old_param 0.00% : 0.000003s : 1: substitution.set_cell_output_no_recompute 0.01% : 0.000021s : 6: substitution.tuple_list_convert_item_index_to_positive 0.01% : 0.000010s : 6: substitution.tuple_list_get_item_const_eliminator 0.01% : 0.000014s : 6: substitution.tuple_list_get_item_depend_reorder 0.03% : 0.000048s : 15: substitution.tuple_list_get_item_eliminator 0.01% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 0.02% : 0.000028s : 18: substitution.updatestate_pure_node_eliminater 0.03% : 0.000039s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.467488 2 99.65% : 0.465848s : 1: type_inference.infer 0.35% : 0.001640s : 1: type_inference.specialize ------[replace.] 0.000340 25 28.29% : 0.000096s : 3: replace.getattr_setattr_resolve 31.70% : 0.000108s : 13: replace.inline 10.80% : 0.000037s : 1: replace.replace_applicator 22.25% : 0.000076s : 7: replace.tuple_list_get_item_eliminator 6.97% : 0.000024s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.141145 25 99.68% : 0.140700s : 3: match.getattr_setattr_resolve 0.29% : 0.000404s : 13: match.inline 0.01% : 0.000009s : 1: match.replace_applicator 0.02% : 0.000021s : 7: match.tuple_list_get_item_eliminator 0.01% : 0.000010s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000848 5424 0.96% : 0.000008s : 55: predicate.accumulaten_eliminater 0.28% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.48% : 0.000004s : 29: predicate.addn_check_dump 0.97% : 0.000008s : 55: predicate.addn_zero_filter 0.82% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 2.16% : 0.000018s : 84: predicate.arithmetic_simplify 0.97% : 0.000008s : 55: predicate.cast_eliminate 1.99% : 0.000017s : 130: predicate.check_bprop_eliminate 0.47% : 0.000004s : 29: predicate.compare_switch_simplify 0.06% : 0.000001s : 6: predicate.const_output_eliminate 0.49% : 0.000004s : 29: predicate.depend_value_elim 0.92% : 0.000008s : 55: predicate.dict_get_item_const_eliminator 0.93% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.85% : 0.000007s : 55: predicate.dict_set_item_eliminator 0.30% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.08% : 0.000001s : 6: predicate.elim_not_effective 0.14% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 0.95% : 0.000008s : 61: predicate.environ_add_const_eliminate 0.91% : 0.000008s : 61: predicate.environ_get_add_eliminate 0.93% : 0.000008s : 61: predicate.environ_get_depend_swap 1.44% : 0.000012s : 90: predicate.environ_get_eliminate 0.94% : 0.000008s : 61: predicate.environ_get_set_eliminate 1.20% : 0.000010s : 76: predicate.exchange_switch_depend_value 1.73% : 0.000015s : 76: predicate.float_depend_g_call 0.48% : 0.000004s : 29: predicate.float_environ_get_switch 0.59% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.06% : 0.000000s : 6: predicate.fold_const_symbol 0.52% : 0.000004s : 29: predicate.get_grad_eliminate 1.02% : 0.000009s : 20: predicate.getattr_setattr_resolve 0.09% : 0.000001s : 6: predicate.graph_param_transform 0.48% : 0.000004s : 29: predicate.incorporate_call 0.44% : 0.000004s : 29: predicate.incorporate_call_switch 10.00% : 0.000085s : 201: predicate.inline 2.06% : 0.000017s : 97: predicate.inline_without_move 0.25% : 0.000002s : 29: predicate.j_node_and_user_rematch 1.03% : 0.000009s : 29: predicate.less_batch_normalization 1.40% : 0.000012s : 74: predicate.list_to_tuple_eliminator_ 1.97% : 0.000017s : 129: predicate.load_eliminater 0.27% : 0.000002s : 6: predicate.loop_unroll_after_grad 2.02% : 0.000017s : 129: predicate.loop_unroll_before_grad 1.09% : 0.000009s : 67: predicate.make_slice_get_slice_eliminator 0.56% : 0.000005s : 29: predicate.merge_addn 1.87% : 0.000016s : 126: predicate.micro_step_allgather_replace 2.04% : 0.000017s : 126: predicate.mini_step_allgather_replace 0.85% : 0.000007s : 55: predicate.minmaximum_grad 0.41% : 0.000003s : 6: predicate.mutable_eliminate 0.12% : 0.000001s : 6: predicate.opt_reshape 0.16% : 0.000001s : 6: predicate.parallel_virtual_node 1.49% : 0.000013s : 76: predicate.partial_defer_inline 1.20% : 0.000010s : 68: predicate.partial_eliminate 0.94% : 0.000008s : 55: predicate.print_const_string_wrapper 0.49% : 0.000004s : 29: predicate.reduce_all_const_elim 1.36% : 0.000012s : 55: predicate.reduce_eliminate 2.04% : 0.000017s : 129: predicate.redundant_stop_gradient_eliminater 0.30% : 0.000003s : 29: predicate.remove_not_recompute_node 1.96% : 0.000017s : 190: predicate.replace_applicator 0.96% : 0.000008s : 97: predicate.replace_old_param 0.15% : 0.000001s : 6: predicate.reset_defer_inline 1.03% : 0.000009s : 55: predicate.reshape_eliminate 2.13% : 0.000018s : 126: predicate.row_tensor_add_zeros_like 0.12% : 0.000001s : 6: predicate.row_tensor_eliminate 2.25% : 0.000019s : 130: predicate.same_eliminate 0.33% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.69% : 0.000006s : 29: predicate.shard_identity_eliminate 0.26% : 0.000002s : 12: predicate.special_op_eliminate 0.61% : 0.000005s : 29: predicate.specialize_transform 2.21% : 0.000019s : 126: predicate.split_environ_get_set_with_tuple_value 1.90% : 0.000016s : 97: predicate.stack_unstack_eliminate 0.14% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.32% : 0.000011s : 76: predicate.switch_defer_inline 3.21% : 0.000027s : 206: predicate.switch_layer_defer_inline 4.08% : 0.000035s : 240: predicate.switch_simplify 0.87% : 0.000007s : 55: predicate.tile_eliminate 0.91% : 0.000008s : 55: predicate.transpose_eliminate 1.20% : 0.000010s : 67: predicate.tuple_list_convert_item_index_to_positive 1.26% : 0.000011s : 67: predicate.tuple_list_get_item_const_eliminator 1.09% : 0.000009s : 67: predicate.tuple_list_get_item_depend_reorder 2.18% : 0.000019s : 103: predicate.tuple_list_get_item_eliminator 1.20% : 0.000010s : 67: predicate.tuple_list_get_set_item_eliminator 1.84% : 0.000016s : 96: predicate.tuple_list_set_item_eliminator 1.27% : 0.000011s : 74: predicate.tuple_to_list_eliminator_ 2.04% : 0.000017s : 129: predicate.updatestate_pure_node_eliminater 2.73% : 0.000023s : 159: predicate.updatestate_useless_node_eliminater 0.13% : 0.000001s : 6: predicate.value_based_eliminate 0.51% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.52% : 0.000004s : 29: predicate.virtual_output_eliminate 0.14% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.15% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003215 48 58.60% : 0.001884s : 24: func_graph_cloner_run.FuncGraphClonerGraph 41.40% : 0.001331s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 2.017999 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.17% : 0.003487s : 1: add_attr 0.17% : 0.003476s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000118s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.01% : 0.000246s : 1: auto_monad 0.00% : 0.000090s : 1: auto_monad_reorder 0.00% : 0.000010s : 1: backend_pass 0.00% : 0.000005s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.02% : 0.000464s : 1: bootstrap 0.00% : 0.000038s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000022s : 1: control_data_broadcast_order 0.00% : 0.000016s : 1: convert_after_rewriter 0.00% : 0.000037s : 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 0.00% : 0.000065s : 1: event_method 0.00% : 0.000015s : 1: execute 0.00% : 0.000006s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000013s : 1: graph_reusing 0.00% : 0.000006s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000006s : 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% : 0.000468s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.04% : 0.000852s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000027s : 1: opt.transform.mutable_eliminate 0.26% : 0.005170s : 125: opt.transform.opt_a 0.00% : 0.000051s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.01% : 0.000184s : 28: opt.transform.opt_b 6.98% : 0.140922s : 2: opt.transform.opt_resolve 0.00% : 0.000065s : 2: opt.transform.opt_trans_graph 0.00% : 0.000052s : 4: opt.transform.symbol_engine_opt 18.32% : 0.369636s : 1: opt_a 0.01% : 0.000145s : 1: opt_after_cconv 0.02% : 0.000501s : 1: opt_after_jit_grad 0.02% : 0.000346s : 1: opt_b 34.12% : 0.688640s : 1: optimize 0.00% : 0.000033s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000012s : 1: order_py_execute_after_rewriter 0.00% : 0.000036s : 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% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.00% : 0.000058s : 1: pre_auto_parallel 0.00% : 0.000046s : 1: py_interpret_to_execute 0.00% : 0.000020s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000003s : 1: remove_cast_before_assign_add 0.00% : 0.000021s : 1: remove_dup_value 0.33% : 0.006618s : 2: renormalize.infer 0.13% : 0.002600s : 2: 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% : 0.000171s : 1: rewriter_after_opt_a 0.01% : 0.000143s : 1: rewriter_before_opt_a 15.66% : 0.316026s : 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% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000098s : 1: symbol_engine_optimizer 0.43% : 0.008603s : 1: task_emit 0.00% : 0.000096s : 1: tuple_transform 23.17% : 0.467604s : 1: type_inference 0.02% : 0.000359s : 1: validate TotalTime = 0.562139, [24] [bootstrap]: 0.00045233 [type_inference]: 0.216941 [event_method]: 5.676e-05 [auto_monad]: 0.00023677 [graph_reusing]: 9.01002e-06 [inline]: 2.21e-06 [add_attr]: 0.00376675, [1] [add_attr_with_inline]: 0.00375735, [1] [Cycle 1]: 8.104e-05, [2] [tag_attr]: 3.944e-05 [meta_addattr_fg_expand]: 9.32001e-06 [parallel-infer-symbol]: 3.95998e-06 [pre_auto_parallel]: 5.329e-05 [insert-virtual-dataset]: 2.78998e-06 [parallel-infer-symbol-second]: 7.09988e-07 [dataset_repeat_opt]: 2.21e-06 [pipeline_split]: 1.77001e-06 [optimize]: 0.114836, [53] [py_interpret_to_execute]: 4.317e-05 [rewriter_before_opt_a]: 0.00013689 [opt_a]: 0.112232, [3] [Cycle 1]: 0.108331, [45] [expand_dump_flag]: 3.56999e-06 [switch_simplify]: 8.095e-05 [loop_unroll]: 5.666e-05 [a_1]: 0.00152732 [with_stream_mark]: 3.367e-05 [recompute_prepare]: 2.787e-05 [updatestate_depend_eliminate]: 1.276e-05 [updatestate_assign_eliminate]: 1.069e-05 [updatestate_loads_eliminate]: 1.023e-05 [parameter_eliminate]: 3.11999e-06 [a_2]: 0.00028497 [accelerated_algorithm]: 2.041e-05 [shard]: 2.07999e-06 [meta_shard_fg_expand]: 5.14e-06 [shard_inline]: 1.827e-05 [merge_send_recv]: 2.164e-05 [auto_parallel]: 1.577e-05 [parallel]: 2.125e-05 [flash_sp]: 1.311e-05 [merge_comm]: 1.159e-05 [allreduce_fusion]: 1.084e-05 [matmul_add_comm_reduction]: 3.374e-05 [allreduce_slice_to_reducescatter]: 1.27e-06 [virtual_shard_identity]: 2.141e-05 [virtual_dataset]: 1.841e-05 [get_grad_eliminate_]: 1.771e-05 [virtual_output]: 1.765e-05 [merge_forward]: 1.288e-05 [cell_reuse_recompute_pass]: 1.55999e-06 [offload_activation]: 2.329e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.567e-05 [merge_recompute_call_nodes]: 1.57999e-06 [before_grad]: 3.457e-05 [set_forward_comm_id_for_comm_node_pass]: 1.296e-05 [meta_fg_expand]: 0.0987212 [flash_sp_send_recv_attached]: 7.30998e-06 [receive_attached]: 2.34001e-06 [after_resolve]: 0.00014658 [a_after_grad]: 0.00018919 [renormalize]: 0.00530722 [add_forward_monad_depend]: 1.471e-05 [auto_monad_grad]: 7.57002e-06 [auto_monad_eliminator]: 7.185e-05 [cse]: 0.00022721 [a_3]: 0.00084535 [Cycle 2]: 0.0029871, [45] [expand_dump_flag]: 3.58999e-06 [switch_simplify]: 0.00011537 [loop_unroll]: 5.327e-05 [a_1]: 0.00124327 [with_stream_mark]: 2.428e-05 [recompute_prepare]: 1.388e-05 [updatestate_depend_eliminate]: 3.296e-05 [updatestate_assign_eliminate]: 5.97999e-06 [updatestate_loads_eliminate]: 5.12999e-06 [parameter_eliminate]: 1.96e-06 [a_2]: 0.00012179 [accelerated_algorithm]: 9.87001e-06 [shard]: 2.27999e-06 [meta_shard_fg_expand]: 3.65998e-06 [shard_inline]: 9.04e-06 [merge_send_recv]: 1.034e-05 [auto_parallel]: 1.106e-05 [parallel]: 8.85001e-06 [flash_sp]: 4.43999e-06 [merge_comm]: 5.12e-06 [allreduce_fusion]: 5.17e-06 [matmul_add_comm_reduction]: 1.281e-05 [allreduce_slice_to_reducescatter]: 9.89996e-07 [virtual_shard_identity]: 1.026e-05 [virtual_dataset]: 8.50999e-06 [get_grad_eliminate_]: 8.42998e-06 [virtual_output]: 8.01001e-06 [merge_forward]: 5.64998e-06 [cell_reuse_recompute_pass]: 1.37e-06 [offload_activation]: 1.302e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.721e-05 [merge_recompute_call_nodes]: 1.49e-06 [before_grad]: 1.54e-05 [set_forward_comm_id_for_comm_node_pass]: 5.67001e-06 [meta_fg_expand]: 5.033e-05 [flash_sp_send_recv_attached]: 1.69e-06 [receive_attached]: 2.31998e-06 [after_resolve]: 1.287e-05 [a_after_grad]: 1.292e-05 [renormalize]: 0.00074295 [add_forward_monad_depend]: 5.32999e-06 [auto_monad_grad]: 1.71e-06 [auto_monad_eliminator]: 1.823e-05 [cse]: 2.871e-05 [a_3]: 6.114e-05 [Cycle 3]: 0.00089543, [45] [expand_dump_flag]: 1.49e-06 [switch_simplify]: 1.012e-05 [loop_unroll]: 8.25999e-06 [a_1]: 0.00019605 [with_stream_mark]: 1.1e-05 [recompute_prepare]: 8.57998e-06 [updatestate_depend_eliminate]: 5.42001e-06 [updatestate_assign_eliminate]: 4.27e-06 [updatestate_loads_eliminate]: 4.52e-06 [parameter_eliminate]: 9.19972e-07 [a_2]: 0.00011266 [accelerated_algorithm]: 8.89e-06 [shard]: 1.12999e-06 [meta_shard_fg_expand]: 2.19999e-06 [shard_inline]: 8.40001e-06 [merge_send_recv]: 6.77002e-06 [auto_parallel]: 7.3e-06 [parallel]: 4.61002e-06 [flash_sp]: 1.06002e-06 [merge_comm]: 5.07e-06 [allreduce_fusion]: 4.74e-06 [matmul_add_comm_reduction]: 8e-06 [allreduce_slice_to_reducescatter]: 4.00003e-07 [virtual_shard_identity]: 9.92999e-06 [virtual_dataset]: 7.89997e-06 [get_grad_eliminate_]: 7.75e-06 [virtual_output]: 8.07e-06 [merge_forward]: 4.39998e-06 [cell_reuse_recompute_pass]: 1.30001e-06 [offload_activation]: 9.94999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.552e-05 [merge_recompute_call_nodes]: 6.90023e-07 [before_grad]: 1.406e-05 [set_forward_comm_id_for_comm_node_pass]: 5.47999e-06 [meta_fg_expand]: 3.12002e-06 [flash_sp_send_recv_attached]: 9.5999e-07 [receive_attached]: 1.02e-06 [after_resolve]: 1.034e-05 [a_after_grad]: 8.258e-05 [renormalize]: 8.00064e-08 [add_forward_monad_depend]: 1.56002e-06 [auto_monad_grad]: 1.53002e-06 [auto_monad_eliminator]: 1.325e-05 [cse]: 2.377e-05 [a_3]: 5.323e-05 [py_interpret_to_execute_after_opt_a]: 3.936e-05 [slice_cell_reuse_recomputed_activation]: 2.12999e-06 [rewriter_after_opt_a]: 6.334e-05 [convert_after_rewriter]: 9.05001e-06 [order_py_execute_after_rewriter]: 6.74999e-06 [mutable_eliminate]: 0.00072743 [opt_b]: 0.00028396, [1] [Cycle 1]: 0.00027706, [7] [b_1]: 0.00018724 [b_2]: 1.017e-05 [updatestate_depend_eliminate]: 8.47998e-06 [updatestate_assign_eliminate]: 4.31002e-06 [updatestate_loads_eliminate]: 4.26001e-06 [renormalize]: 3.50003e-07 [cse]: 2.641e-05 [optimize_parallel_all_gather_comm]: 2.106e-05 [overlap_param_gather]: 2.00002e-06 [cconv]: 2.489e-05 [loop_unroll]: 0.00044089 [opt_after_cconv]: 0.00013739, [1] [Cycle 1]: 0.0001314, [7] [c_1]: 5.187e-05 [parameter_eliminate]: 2.46e-06 [updatestate_depend_eliminate]: 8.38999e-06 [updatestate_assign_eliminate]: 4.46002e-06 [updatestate_loads_eliminate]: 4.37e-06 [cse]: 2.479e-05 [renormalize]: 2.29978e-07 [remove_dup_value]: 1.704e-05 [tuple_transform]: 8.911e-05, [1] [Cycle 1]: 8.452e-05, [4] [d_1]: 5.561e-05 [none_parameter_eliminate]: 1.92001e-06 [renormalize]: 1.79978e-07 [switch_simplify]: 9.57001e-06 [partial_unused_args_eliminate]: 2.01e-06 [add_recomputation]: 8.085e-05 [cse_after_recomputation]: 3.103e-05, [1] [Cycle 1]: 2.615e-05, [1] [cse]: 2.012e-05 [environ_conv]: 6.68998e-06 [swap_dp_allreduce_reducescatter]: 8.40001e-06 [bias_add_comm_swap]: 2.85002e-06 [label_micro_interleaved_index]: 4.74e-06 [label_fine_grained_interleaved_index]: 3.04999e-06 [merge_cast_opt]: 1.42e-06 [slice_recompute_activation]: 2.28998e-06 [micro_interleaved_order_control]: 2.70002e-06 [assign_add_opt]: 1.27e-06 [ForceFp32Comm]: 1.17999e-06 [remove_cast_before_assign_add]: 1.32999e-06 [full_micro_interleaved_order_control]: 2.66999e-06 [reorder_send_recv_between_fp_bp]: 2.84001e-06 [comm_op_add_attrs]: 1.25001e-06 [add_comm_op_reuse_tag]: 1.17e-06 [interleave_split_concat_branches]: 1.19e-06 [interleave_parallel_branches]: 1.11002e-06 [overlap_opt_shard_in_pipeline]: 1.24e-06 [overlap_opt_shard_grad_in_pipeline]: 1.75001e-06 [control_data_broadcast_order]: 1.821e-05 [grouped_pairwise_exchange_alltoall]: 1.67001e-06 [offloading_packed_experts]: 5.29e-06 [overlap_recompute_and_grad_model_parallel]: 6.22001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.37e-06 [overlap_recompute_allgather_and_fa_grad]: 1.44e-06 [overlap_recompute_comm]: 2.64001e-06 [overlap_grad_ring_attention]: 5.83002e-06 [overlap_grad_flash_sp]: 2.749e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.33002e-06 [split_layernorm_comm]: 1.74e-06 [handle_group_info]: 1.25999e-06 [symbol_engine_optimizer]: 9.22e-05, [1] [Cycle 1]: 8.756e-05, [6] [build]: 4.12e-06 [elim_shapecalc]: 1.211e-05 [elim_not_effective]: 1.816e-05 [opt_reshape]: 9.70002e-06 [fold_const_symbol]: 1.468e-05 [renormalize]: 2.60014e-07 [detach_backward]: 2.54001e-06 [pipeline_parallel_scheduler]: 1.40001e-06 [auto_monad_reorder]: 6.066e-05 [get_jit_bprop_graph]: 2.27999e-06 [rewriter_after_jit_bprop_graph]: 4.17e-06 [opt_after_jit_grad]: 0.00048271 [validate]: 5.2e-05 [backend_pass]: 1.10001e-06 [task_emit]: 0.224925 [execute]: 1.024e-05 Sums bootstrap : 0.000452s : 0.08% type_inference : 0.216941s : 38.95% event_method : 0.000057s : 0.01% auto_monad : 0.000237s : 0.04% graph_reusing : 0.000009s : 0.00% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000039s : 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 : 0.000053s : 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 : 0.000043s : 0.01% optimize.rewriter_before_opt_a : 0.000137s : 0.02% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000206s : 0.04% optimize.opt_a.loop_unroll : 0.000118s : 0.02% optimize.opt_a.a_1 : 0.002967s : 0.53% optimize.opt_a.with_stream_mark : 0.000069s : 0.01% optimize.opt_a.recompute_prepare : 0.000050s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000051s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000021s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000020s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000519s : 0.09% optimize.opt_a.accelerated_algorithm : 0.000039s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000011s : 0.00% optimize.opt_a.shard_inline : 0.000036s : 0.01% optimize.opt_a.merge_send_recv : 0.000039s : 0.01% optimize.opt_a.auto_parallel : 0.000034s : 0.01% optimize.opt_a.parallel : 0.000035s : 0.01% optimize.opt_a.flash_sp : 0.000019s : 0.00% optimize.opt_a.merge_comm : 0.000022s : 0.00% optimize.opt_a.allreduce_fusion : 0.000021s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000055s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000042s : 0.01% optimize.opt_a.virtual_dataset : 0.000035s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000034s : 0.01% optimize.opt_a.virtual_output : 0.000034s : 0.01% optimize.opt_a.merge_forward : 0.000023s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000046s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000068s : 0.01% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000064s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000024s : 0.00% optimize.opt_a.meta_fg_expand : 0.098775s : 17.73% 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 : 0.000170s : 0.03% optimize.opt_a.a_after_grad : 0.000285s : 0.05% optimize.opt_a.renormalize : 0.006050s : 1.09% optimize.opt_a.add_forward_monad_depend : 0.000022s : 0.00% optimize.opt_a.auto_monad_grad : 0.000011s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000103s : 0.02% optimize.opt_a.cse : 0.000280s : 0.05% optimize.opt_a.a_3 : 0.000960s : 0.17% optimize.py_interpret_to_execute_after_opt_a : 0.000039s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000063s : 0.01% optimize.convert_after_rewriter : 0.000009s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000727s : 0.13% optimize.opt_b.b_1 : 0.000187s : 0.03% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 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 : 0.000000s : 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 : 0.000025s : 0.00% optimize.loop_unroll : 0.000441s : 0.08% optimize.opt_after_cconv.c_1 : 0.000052s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 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 : 0.000025s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.00% optimize.tuple_transform.d_1 : 0.000056s : 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 : 0.000010s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000081s : 0.01% optimize.cse_after_recomputation.cse : 0.000020s : 0.00% 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 : 0.000005s : 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 : 0.000001s : 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 : 0.000002s : 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 : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000027s : 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 : 0.000012s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 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 : 0.000001s : 0.00% auto_monad_reorder : 0.000061s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000483s : 0.09% validate : 0.000052s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.224925s : 40.38% execute : 0.000010s : 0.00% Time group info: ------[substitution.] 0.001688 224 0.72% : 0.000012s : 4: substitution.depend_value_elim 0.17% : 0.000003s : 5: substitution.elim_not_effective 0.17% : 0.000003s : 3: substitution.float_depend_g_call 0.25% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.14% : 0.000002s : 5: substitution.fold_const_symbol 42.81% : 0.000723s : 4: substitution.getattr_setattr_resolve 0.41% : 0.000007s : 6: substitution.graph_param_transform 0.21% : 0.000004s : 2: substitution.incorporate_call 0.15% : 0.000002s : 2: substitution.incorporate_call_switch 35.56% : 0.000600s : 13: substitution.inline 2.61% : 0.000044s : 5: substitution.inline_without_move 0.76% : 0.000013s : 23: substitution.j_node_and_user_rematch 0.66% : 0.000011s : 6: substitution.minmaximum_grad 0.21% : 0.000004s : 3: substitution.partial_eliminate 0.92% : 0.000016s : 23: substitution.remove_not_recompute_node 3.15% : 0.000053s : 21: substitution.replace_applicator 0.83% : 0.000014s : 16: substitution.replace_old_param 0.22% : 0.000004s : 1: substitution.set_cell_output_no_recompute 1.28% : 0.000022s : 6: substitution.tuple_list_convert_item_index_to_positive 0.56% : 0.000009s : 6: substitution.tuple_list_get_item_const_eliminator 0.78% : 0.000013s : 6: substitution.tuple_list_get_item_depend_reorder 2.99% : 0.000051s : 15: substitution.tuple_list_get_item_eliminator 0.76% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 1.47% : 0.000025s : 18: substitution.updatestate_pure_node_eliminater 2.20% : 0.000037s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.216856 2 99.36% : 0.215476s : 1: type_inference.infer 0.64% : 0.001380s : 1: type_inference.specialize ------[replace.] 0.000297 25 17.00% : 0.000050s : 3: replace.getattr_setattr_resolve 37.97% : 0.000113s : 13: replace.inline 11.27% : 0.000033s : 1: replace.replace_applicator 26.36% : 0.000078s : 7: replace.tuple_list_get_item_eliminator 7.40% : 0.000022s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001313 25 51.66% : 0.000678s : 3: match.getattr_setattr_resolve 45.13% : 0.000593s : 13: match.inline 0.76% : 0.000010s : 1: match.replace_applicator 1.75% : 0.000023s : 7: match.tuple_list_get_item_eliminator 0.70% : 0.000009s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000833 5424 0.89% : 0.000007s : 55: predicate.accumulaten_eliminater 0.27% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.46% : 0.000004s : 29: predicate.addn_check_dump 0.87% : 0.000007s : 55: predicate.addn_zero_filter 0.83% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 1.83% : 0.000015s : 84: predicate.arithmetic_simplify 0.93% : 0.000008s : 55: predicate.cast_eliminate 2.01% : 0.000017s : 130: predicate.check_bprop_eliminate 0.47% : 0.000004s : 29: predicate.compare_switch_simplify 0.06% : 0.000001s : 6: predicate.const_output_eliminate 0.48% : 0.000004s : 29: predicate.depend_value_elim 0.95% : 0.000008s : 55: predicate.dict_get_item_const_eliminator 0.98% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.88% : 0.000007s : 55: predicate.dict_set_item_eliminator 0.31% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.07% : 0.000001s : 6: predicate.elim_not_effective 0.12% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 0.98% : 0.000008s : 61: predicate.environ_add_const_eliminate 0.96% : 0.000008s : 61: predicate.environ_get_add_eliminate 0.97% : 0.000008s : 61: predicate.environ_get_depend_swap 1.50% : 0.000013s : 90: predicate.environ_get_eliminate 0.94% : 0.000008s : 61: predicate.environ_get_set_eliminate 1.24% : 0.000010s : 76: predicate.exchange_switch_depend_value 1.80% : 0.000015s : 76: predicate.float_depend_g_call 0.48% : 0.000004s : 29: predicate.float_environ_get_switch 0.55% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.06% : 0.000001s : 6: predicate.fold_const_symbol 0.54% : 0.000004s : 29: predicate.get_grad_eliminate 0.69% : 0.000006s : 20: predicate.getattr_setattr_resolve 0.07% : 0.000001s : 6: predicate.graph_param_transform 0.47% : 0.000004s : 29: predicate.incorporate_call 0.43% : 0.000004s : 29: predicate.incorporate_call_switch 4.49% : 0.000037s : 201: predicate.inline 2.20% : 0.000018s : 97: predicate.inline_without_move 0.25% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.59% : 0.000005s : 29: predicate.less_batch_normalization 1.21% : 0.000010s : 74: predicate.list_to_tuple_eliminator_ 2.04% : 0.000017s : 129: predicate.load_eliminater 0.29% : 0.000002s : 6: predicate.loop_unroll_after_grad 2.04% : 0.000017s : 129: predicate.loop_unroll_before_grad 1.17% : 0.000010s : 67: predicate.make_slice_get_slice_eliminator 0.48% : 0.000004s : 29: predicate.merge_addn 1.93% : 0.000016s : 126: predicate.micro_step_allgather_replace 1.88% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.86% : 0.000007s : 55: predicate.minmaximum_grad 0.32% : 0.000003s : 6: predicate.mutable_eliminate 0.13% : 0.000001s : 6: predicate.opt_reshape 0.16% : 0.000001s : 6: predicate.parallel_virtual_node 1.57% : 0.000013s : 76: predicate.partial_defer_inline 1.25% : 0.000010s : 68: predicate.partial_eliminate 0.88% : 0.000007s : 55: predicate.print_const_string_wrapper 0.48% : 0.000004s : 29: predicate.reduce_all_const_elim 1.17% : 0.000010s : 55: predicate.reduce_eliminate 2.06% : 0.000017s : 129: predicate.redundant_stop_gradient_eliminater 0.31% : 0.000003s : 29: predicate.remove_not_recompute_node 2.05% : 0.000017s : 190: predicate.replace_applicator 1.04% : 0.000009s : 97: predicate.replace_old_param 0.09% : 0.000001s : 6: predicate.reset_defer_inline 0.90% : 0.000007s : 55: predicate.reshape_eliminate 2.01% : 0.000017s : 126: predicate.row_tensor_add_zeros_like 0.13% : 0.000001s : 6: predicate.row_tensor_eliminate 2.17% : 0.000018s : 130: predicate.same_eliminate 0.34% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.54% : 0.000005s : 29: predicate.shard_identity_eliminate 0.25% : 0.000002s : 12: predicate.special_op_eliminate 0.58% : 0.000005s : 29: predicate.specialize_transform 2.28% : 0.000019s : 126: predicate.split_environ_get_set_with_tuple_value 1.97% : 0.000016s : 97: predicate.stack_unstack_eliminate 0.14% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.35% : 0.000011s : 76: predicate.switch_defer_inline 3.31% : 0.000028s : 206: predicate.switch_layer_defer_inline 10.90% : 0.000091s : 240: predicate.switch_simplify 0.91% : 0.000008s : 55: predicate.tile_eliminate 0.89% : 0.000007s : 55: predicate.transpose_eliminate 1.24% : 0.000010s : 67: predicate.tuple_list_convert_item_index_to_positive 1.33% : 0.000011s : 67: predicate.tuple_list_get_item_const_eliminator 1.12% : 0.000009s : 67: predicate.tuple_list_get_item_depend_reorder 2.33% : 0.000019s : 103: predicate.tuple_list_get_item_eliminator 1.17% : 0.000010s : 67: predicate.tuple_list_get_set_item_eliminator 1.72% : 0.000014s : 96: predicate.tuple_list_set_item_eliminator 1.18% : 0.000010s : 74: predicate.tuple_to_list_eliminator_ 2.08% : 0.000017s : 129: predicate.updatestate_pure_node_eliminater 2.69% : 0.000022s : 159: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 6: predicate.value_based_eliminate 0.54% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.51% : 0.000004s : 29: predicate.virtual_output_eliminate 0.12% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.14% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002543 48 66.99% : 0.001703s : 24: func_graph_cloner_run.FuncGraphClonerGraph 33.01% : 0.000839s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.693379 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.54% : 0.003772s : 1: add_attr 0.54% : 0.003761s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000085s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000245s : 1: auto_monad 0.01% : 0.000065s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.07% : 0.000480s : 1: bootstrap 0.00% : 0.000028s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000022s : 1: control_data_broadcast_order 0.00% : 0.000012s : 1: convert_after_rewriter 0.00% : 0.000034s : 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% : 0.000065s : 1: event_method 0.00% : 0.000017s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 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% : 0.000005s : 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% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.06% : 0.000449s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.11% : 0.000735s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 0.79% : 0.005473s : 125: opt.transform.opt_a 0.01% : 0.000050s : 1: opt.transform.opt_after_cconv 0.00% : 0.000032s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000170s : 28: opt.transform.opt_b 0.12% : 0.000825s : 2: opt.transform.opt_resolve 0.01% : 0.000063s : 2: opt.transform.opt_trans_graph 0.01% : 0.000051s : 4: opt.transform.symbol_engine_opt 16.19% : 0.112236s : 1: opt_a 0.02% : 0.000141s : 1: opt_after_cconv 0.07% : 0.000492s : 1: opt_after_jit_grad 0.04% : 0.000287s : 1: opt_b 16.56% : 0.114842s : 1: optimize 0.00% : 0.000025s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 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% : 0.000009s : 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% : 0.000009s : 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% : 0.000058s : 1: pre_auto_parallel 0.01% : 0.000047s : 1: py_interpret_to_execute 0.01% : 0.000045s : 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 0.55% : 0.003804s : 2: renormalize.infer 0.32% : 0.002227s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000067s : 1: rewriter_after_opt_a 0.02% : 0.000142s : 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% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000095s : 1: symbol_engine_optimizer 32.44% : 0.224946s : 1: task_emit 0.01% : 0.000092s : 1: tuple_transform 31.29% : 0.216961s : 1: type_inference 0.01% : 0.000081s : 1: validate TotalTime = 0.425384, [24] [bootstrap]: 0.00054544 [type_inference]: 0.114776 [event_method]: 7.443e-05 [auto_monad]: 0.00028379 [graph_reusing]: 9.36e-06 [inline]: 2.99001e-06 [add_attr]: 0.00417171, [1] [add_attr_with_inline]: 0.00416006, [1] [Cycle 1]: 0.00010137, [2] [tag_attr]: 4.948e-05 [meta_addattr_fg_expand]: 9.61e-06 [parallel-infer-symbol]: 3.58e-06 [pre_auto_parallel]: 7.305e-05 [insert-virtual-dataset]: 2.88e-06 [parallel-infer-symbol-second]: 8.50006e-07 [dataset_repeat_opt]: 1.88002e-06 [pipeline_split]: 1.61002e-06 [optimize]: 0.232597, [53] [py_interpret_to_execute]: 4.688e-05 [rewriter_before_opt_a]: 0.00016276 [opt_a]: 0.229673, [3] [Cycle 1]: 0.0904234, [45] [expand_dump_flag]: 3.63999e-06 [switch_simplify]: 7.174e-05 [loop_unroll]: 6.914e-05 [a_1]: 0.00190899 [with_stream_mark]: 3.936e-05 [recompute_prepare]: 2.882e-05 [updatestate_depend_eliminate]: 2.351e-05 [updatestate_assign_eliminate]: 1.251e-05 [updatestate_loads_eliminate]: 1.043e-05 [parameter_eliminate]: 3.32002e-06 [a_2]: 0.0707465 [accelerated_algorithm]: 3.117e-05 [shard]: 5.00999e-06 [meta_shard_fg_expand]: 1.236e-05 [shard_inline]: 2.066e-05 [merge_send_recv]: 4.195e-05 [auto_parallel]: 2.45e-05 [parallel]: 2.433e-05 [flash_sp]: 1.806e-05 [merge_comm]: 1.214e-05 [allreduce_fusion]: 1.103e-05 [matmul_add_comm_reduction]: 4.082e-05 [allreduce_slice_to_reducescatter]: 8.50006e-07 [virtual_shard_identity]: 2.328e-05 [virtual_dataset]: 1.994e-05 [get_grad_eliminate_]: 1.82e-05 [virtual_output]: 1.875e-05 [merge_forward]: 1.326e-05 [cell_reuse_recompute_pass]: 2.96001e-06 [offload_activation]: 2.574e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.367e-05 [merge_recompute_call_nodes]: 1.57001e-06 [before_grad]: 3.614e-05 [set_forward_comm_id_for_comm_node_pass]: 1.33e-05 [meta_fg_expand]: 0.00833531 [flash_sp_send_recv_attached]: 6.61e-06 [receive_attached]: 2.76e-06 [after_resolve]: 0.00012806 [a_after_grad]: 0.00018276 [renormalize]: 0.00662245 [add_forward_monad_depend]: 1.69e-05 [auto_monad_grad]: 9.81998e-06 [auto_monad_eliminator]: 9.139e-05 [cse]: 0.00028075 [a_3]: 0.00093869 [Cycle 2]: 0.138286, [45] [expand_dump_flag]: 3.46999e-06 [switch_simplify]: 5.754e-05 [loop_unroll]: 6.738e-05 [a_1]: 0.135768 [with_stream_mark]: 4.516e-05 [recompute_prepare]: 2.072e-05 [updatestate_depend_eliminate]: 5.111e-05 [updatestate_assign_eliminate]: 7.26001e-06 [updatestate_loads_eliminate]: 5.20001e-06 [parameter_eliminate]: 2.71e-06 [a_2]: 0.00012987 [accelerated_algorithm]: 1.094e-05 [shard]: 3.66999e-06 [meta_shard_fg_expand]: 4.44002e-06 [shard_inline]: 9.49e-06 [merge_send_recv]: 1.338e-05 [auto_parallel]: 1.176e-05 [parallel]: 9.66e-06 [flash_sp]: 4.50999e-06 [merge_comm]: 6.26e-06 [allreduce_fusion]: 5.50001e-06 [matmul_add_comm_reduction]: 1.459e-05 [allreduce_slice_to_reducescatter]: 7.60017e-07 [virtual_shard_identity]: 1.261e-05 [virtual_dataset]: 8.81002e-06 [get_grad_eliminate_]: 9.07001e-06 [virtual_output]: 9.25001e-06 [merge_forward]: 6.59999e-06 [cell_reuse_recompute_pass]: 3.24001e-06 [offload_activation]: 1.438e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.921e-05 [merge_recompute_call_nodes]: 1.44e-06 [before_grad]: 1.473e-05 [set_forward_comm_id_for_comm_node_pass]: 6.16e-06 [meta_fg_expand]: 8.821e-05 [flash_sp_send_recv_attached]: 1.91e-06 [receive_attached]: 2.26e-06 [after_resolve]: 1.991e-05 [a_after_grad]: 1.41e-05 [renormalize]: 0.00128887 [add_forward_monad_depend]: 7.71001e-06 [auto_monad_grad]: 2.73998e-06 [auto_monad_eliminator]: 2.913e-05 [cse]: 4.455e-05 [a_3]: 9.052e-05 [Cycle 3]: 0.00094171, [45] [expand_dump_flag]: 2.78e-06 [switch_simplify]: 1.246e-05 [loop_unroll]: 9.43002e-06 [a_1]: 0.00021415 [with_stream_mark]: 1.838e-05 [recompute_prepare]: 1.214e-05 [updatestate_depend_eliminate]: 7.08e-06 [updatestate_assign_eliminate]: 5.13002e-06 [updatestate_loads_eliminate]: 5.29e-06 [parameter_eliminate]: 1.89e-06 [a_2]: 0.00011734 [accelerated_algorithm]: 1.072e-05 [shard]: 1.76998e-06 [meta_shard_fg_expand]: 2.46998e-06 [shard_inline]: 9.31002e-06 [merge_send_recv]: 1.031e-05 [auto_parallel]: 1.108e-05 [parallel]: 8.30999e-06 [flash_sp]: 1.15999e-06 [merge_comm]: 5.13002e-06 [allreduce_fusion]: 5.12e-06 [matmul_add_comm_reduction]: 9.97999e-06 [allreduce_slice_to_reducescatter]: 4.10015e-07 [virtual_shard_identity]: 1.156e-05 [virtual_dataset]: 8.67e-06 [get_grad_eliminate_]: 8.31002e-06 [virtual_output]: 8.35999e-06 [merge_forward]: 5.69e-06 [cell_reuse_recompute_pass]: 2.10002e-06 [offload_activation]: 1.292e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.916e-05 [merge_recompute_call_nodes]: 1.05001e-06 [before_grad]: 1.51e-05 [set_forward_comm_id_for_comm_node_pass]: 6.98e-06 [meta_fg_expand]: 3.68999e-06 [flash_sp_send_recv_attached]: 1.56998e-06 [receive_attached]: 1.95001e-06 [after_resolve]: 1.22e-05 [a_after_grad]: 1.261e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 2.01e-06 [auto_monad_grad]: 1.80001e-06 [auto_monad_eliminator]: 1.665e-05 [cse]: 3.308e-05 [a_3]: 5.37e-05 [py_interpret_to_execute_after_opt_a]: 2.034e-05 [slice_cell_reuse_recomputed_activation]: 2.12999e-06 [rewriter_after_opt_a]: 6.917e-05 [convert_after_rewriter]: 1.005e-05 [order_py_execute_after_rewriter]: 7.48e-06 [mutable_eliminate]: 0.00078822 [opt_b]: 0.0003207, [1] [Cycle 1]: 0.00031219, [7] [b_1]: 0.0002036 [b_2]: 1.118e-05 [updatestate_depend_eliminate]: 1.238e-05 [updatestate_assign_eliminate]: 4.95001e-06 [updatestate_loads_eliminate]: 4.74e-06 [renormalize]: 6.89994e-07 [cse]: 3.474e-05 [optimize_parallel_all_gather_comm]: 2.368e-05 [overlap_param_gather]: 2.49999e-06 [cconv]: 3.392e-05 [loop_unroll]: 0.00051316 [opt_after_cconv]: 0.00016423, [1] [Cycle 1]: 0.00015649, [7] [c_1]: 6.274e-05 [parameter_eliminate]: 4.37e-06 [updatestate_depend_eliminate]: 1.111e-05 [updatestate_assign_eliminate]: 4.38999e-06 [updatestate_loads_eliminate]: 4.42e-06 [cse]: 3.095e-05 [renormalize]: 4.39992e-07 [remove_dup_value]: 1.888e-05 [tuple_transform]: 9.671e-05, [1] [Cycle 1]: 9.122e-05, [4] [d_1]: 6.012e-05 [none_parameter_eliminate]: 1.86998e-06 [renormalize]: 2.30008e-07 [switch_simplify]: 1.033e-05 [partial_unused_args_eliminate]: 2.14999e-06 [add_recomputation]: 9.94e-05 [cse_after_recomputation]: 3.621e-05, [1] [Cycle 1]: 3.1e-05, [1] [cse]: 2.349e-05 [environ_conv]: 7.28999e-06 [swap_dp_allreduce_reducescatter]: 9.36e-06 [bias_add_comm_swap]: 3.73999e-06 [label_micro_interleaved_index]: 5.03002e-06 [label_fine_grained_interleaved_index]: 3.06001e-06 [merge_cast_opt]: 1.49e-06 [slice_recompute_activation]: 2.19999e-06 [micro_interleaved_order_control]: 2.89001e-06 [assign_add_opt]: 1.77001e-06 [ForceFp32Comm]: 8.00006e-07 [remove_cast_before_assign_add]: 1.10001e-06 [full_micro_interleaved_order_control]: 2.19001e-06 [reorder_send_recv_between_fp_bp]: 2.69999e-06 [comm_op_add_attrs]: 1.28002e-06 [add_comm_op_reuse_tag]: 1.03001e-06 [interleave_split_concat_branches]: 1.36002e-06 [interleave_parallel_branches]: 1.07998e-06 [overlap_opt_shard_in_pipeline]: 1.47999e-06 [overlap_opt_shard_grad_in_pipeline]: 1.65001e-06 [control_data_broadcast_order]: 2.181e-05 [grouped_pairwise_exchange_alltoall]: 2.02999e-06 [offloading_packed_experts]: 6.29999e-06 [overlap_recompute_and_grad_model_parallel]: 6.19001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.37e-06 [overlap_recompute_allgather_and_fa_grad]: 1.72999e-06 [overlap_recompute_comm]: 1.96998e-06 [overlap_grad_ring_attention]: 5.39998e-06 [overlap_grad_flash_sp]: 2.903e-05 [begin_end_overlap_inline]: 5.3001e-07 [split_matmul_comm_elemetwise]: 2.21e-06 [split_layernorm_comm]: 1.94e-06 [handle_group_info]: 9.99979e-07 [symbol_engine_optimizer]: 0.00011347, [1] [Cycle 1]: 0.00010846, [6] [build]: 6.22001e-06 [elim_shapecalc]: 1.839e-05 [elim_not_effective]: 2.213e-05 [opt_reshape]: 1.045e-05 [fold_const_symbol]: 1.609e-05 [renormalize]: 2.29978e-07 [detach_backward]: 3.25e-06 [pipeline_parallel_scheduler]: 1.59998e-06 [auto_monad_reorder]: 8.369e-05 [get_jit_bprop_graph]: 2.21998e-06 [rewriter_after_jit_bprop_graph]: 6.66e-06 [opt_after_jit_grad]: 0.00058337 [validate]: 5.856e-05 [backend_pass]: 9.5999e-07 [task_emit]: 0.0718035 [execute]: 9.03002e-06 Sums bootstrap : 0.000545s : 0.13% type_inference : 0.114776s : 27.36% event_method : 0.000074s : 0.02% auto_monad : 0.000284s : 0.07% graph_reusing : 0.000009s : 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 : 0.000073s : 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 : 0.000047s : 0.01% optimize.rewriter_before_opt_a : 0.000163s : 0.04% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000142s : 0.03% optimize.opt_a.loop_unroll : 0.000146s : 0.03% optimize.opt_a.a_1 : 0.137891s : 32.87% optimize.opt_a.with_stream_mark : 0.000103s : 0.02% optimize.opt_a.recompute_prepare : 0.000062s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000082s : 0.02% optimize.opt_a.updatestate_assign_eliminate : 0.000025s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000021s : 0.00% optimize.opt_a.parameter_eliminate : 0.000008s : 0.00% optimize.opt_a.a_2 : 0.070994s : 16.92% optimize.opt_a.accelerated_algorithm : 0.000053s : 0.01% optimize.opt_a.shard : 0.000010s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000019s : 0.00% optimize.opt_a.shard_inline : 0.000039s : 0.01% optimize.opt_a.merge_send_recv : 0.000066s : 0.02% optimize.opt_a.auto_parallel : 0.000047s : 0.01% optimize.opt_a.parallel : 0.000042s : 0.01% optimize.opt_a.flash_sp : 0.000024s : 0.01% optimize.opt_a.merge_comm : 0.000024s : 0.01% optimize.opt_a.allreduce_fusion : 0.000022s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000065s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000047s : 0.01% optimize.opt_a.virtual_dataset : 0.000037s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000036s : 0.01% optimize.opt_a.virtual_output : 0.000036s : 0.01% optimize.opt_a.merge_forward : 0.000026s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000008s : 0.00% optimize.opt_a.offload_activation : 0.000053s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000082s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000066s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000026s : 0.01% optimize.opt_a.meta_fg_expand : 0.008427s : 2.01% optimize.opt_a.flash_sp_send_recv_attached : 0.000010s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000160s : 0.04% optimize.opt_a.a_after_grad : 0.000209s : 0.05% optimize.opt_a.renormalize : 0.007911s : 1.89% optimize.opt_a.add_forward_monad_depend : 0.000027s : 0.01% optimize.opt_a.auto_monad_grad : 0.000014s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000137s : 0.03% optimize.opt_a.cse : 0.000358s : 0.09% optimize.opt_a.a_3 : 0.001083s : 0.26% optimize.py_interpret_to_execute_after_opt_a : 0.000020s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000069s : 0.02% optimize.convert_after_rewriter : 0.000010s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000788s : 0.19% optimize.opt_b.b_1 : 0.000204s : 0.05% 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 : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000035s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000024s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000034s : 0.01% optimize.loop_unroll : 0.000513s : 0.12% optimize.opt_after_cconv.c_1 : 0.000063s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000011s : 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 : 0.000031s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000019s : 0.00% optimize.tuple_transform.d_1 : 0.000060s : 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 : 0.000010s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000099s : 0.02% optimize.cse_after_recomputation.cse : 0.000023s : 0.01% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 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 : 0.000022s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 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 : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000005s : 0.00% optimize.overlap_grad_flash_sp : 0.000029s : 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 : 0.000006s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000018s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000022s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000016s : 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 : 0.000084s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000583s : 0.14% validate : 0.000059s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.071804s : 17.11% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.002127 224 0.81% : 0.000017s : 4: substitution.depend_value_elim 0.18% : 0.000004s : 5: substitution.elim_not_effective 0.16% : 0.000003s : 3: substitution.float_depend_g_call 0.23% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.12% : 0.000002s : 5: substitution.fold_const_symbol 37.77% : 0.000803s : 4: substitution.getattr_setattr_resolve 0.37% : 0.000008s : 6: substitution.graph_param_transform 0.17% : 0.000004s : 2: substitution.incorporate_call 0.12% : 0.000002s : 2: substitution.incorporate_call_switch 40.92% : 0.000871s : 13: substitution.inline 1.94% : 0.000041s : 5: substitution.inline_without_move 0.68% : 0.000014s : 23: substitution.j_node_and_user_rematch 1.18% : 0.000025s : 6: substitution.minmaximum_grad 0.19% : 0.000004s : 3: substitution.partial_eliminate 0.83% : 0.000018s : 23: substitution.remove_not_recompute_node 2.92% : 0.000062s : 21: substitution.replace_applicator 0.76% : 0.000016s : 16: substitution.replace_old_param 0.18% : 0.000004s : 1: substitution.set_cell_output_no_recompute 1.25% : 0.000027s : 6: substitution.tuple_list_convert_item_index_to_positive 0.54% : 0.000011s : 6: substitution.tuple_list_get_item_const_eliminator 0.80% : 0.000017s : 6: substitution.tuple_list_get_item_depend_reorder 3.46% : 0.000074s : 15: substitution.tuple_list_get_item_eliminator 0.74% : 0.000016s : 6: substitution.tuple_list_get_set_item_eliminator 1.48% : 0.000032s : 18: substitution.updatestate_pure_node_eliminater 2.24% : 0.000048s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.114658 2 98.45% : 0.112878s : 1: type_inference.infer 1.55% : 0.001780s : 1: type_inference.specialize ------[replace.] 0.121419 25 0.05% : 0.000064s : 3: replace.getattr_setattr_resolve 99.79% : 0.121158s : 13: replace.inline 0.05% : 0.000058s : 1: replace.replace_applicator 0.09% : 0.000110s : 7: replace.tuple_list_get_item_eliminator 0.02% : 0.000028s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001674 25 44.96% : 0.000753s : 3: match.getattr_setattr_resolve 51.45% : 0.000861s : 13: match.inline 0.79% : 0.000013s : 1: match.replace_applicator 1.92% : 0.000032s : 7: match.tuple_list_get_item_eliminator 0.88% : 0.000015s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000892 5424 0.93% : 0.000008s : 55: predicate.accumulaten_eliminater 0.30% : 0.000003s : 6: predicate.ad_related_special_op_eliminate 0.43% : 0.000004s : 29: predicate.addn_check_dump 0.94% : 0.000008s : 55: predicate.addn_zero_filter 0.82% : 0.000007s : 55: predicate.adjust_all_reduce_mul_add 2.10% : 0.000019s : 84: predicate.arithmetic_simplify 0.92% : 0.000008s : 55: predicate.cast_eliminate 1.91% : 0.000017s : 130: predicate.check_bprop_eliminate 0.45% : 0.000004s : 29: predicate.compare_switch_simplify 0.05% : 0.000000s : 6: predicate.const_output_eliminate 0.61% : 0.000005s : 29: predicate.depend_value_elim 1.83% : 0.000016s : 55: predicate.dict_get_item_const_eliminator 0.95% : 0.000008s : 55: predicate.dict_get_item_eliminator 0.96% : 0.000009s : 55: predicate.dict_set_item_eliminator 0.36% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.09% : 0.000001s : 6: predicate.elim_not_effective 0.18% : 0.000002s : 6: predicate.elim_shapecalc_of_broadcastargs 1.03% : 0.000009s : 61: predicate.environ_add_const_eliminate 0.93% : 0.000008s : 61: predicate.environ_get_add_eliminate 0.89% : 0.000008s : 61: predicate.environ_get_depend_swap 3.31% : 0.000029s : 90: predicate.environ_get_eliminate 0.99% : 0.000009s : 61: predicate.environ_get_set_eliminate 2.07% : 0.000018s : 76: predicate.exchange_switch_depend_value 2.44% : 0.000022s : 76: predicate.float_depend_g_call 0.48% : 0.000004s : 29: predicate.float_environ_get_switch 0.61% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.06% : 0.000001s : 6: predicate.fold_const_symbol 0.55% : 0.000005s : 29: predicate.get_grad_eliminate 0.84% : 0.000007s : 20: predicate.getattr_setattr_resolve 0.06% : 0.000001s : 6: predicate.graph_param_transform 0.46% : 0.000004s : 29: predicate.incorporate_call 0.40% : 0.000004s : 29: predicate.incorporate_call_switch 5.03% : 0.000045s : 201: predicate.inline 1.96% : 0.000017s : 97: predicate.inline_without_move 0.23% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.82% : 0.000007s : 29: predicate.less_batch_normalization 1.40% : 0.000012s : 74: predicate.list_to_tuple_eliminator_ 1.96% : 0.000018s : 129: predicate.load_eliminater 0.41% : 0.000004s : 6: predicate.loop_unroll_after_grad 2.06% : 0.000018s : 129: predicate.loop_unroll_before_grad 1.19% : 0.000011s : 67: predicate.make_slice_get_slice_eliminator 0.47% : 0.000004s : 29: predicate.merge_addn 1.78% : 0.000016s : 126: predicate.micro_step_allgather_replace 1.83% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.87% : 0.000008s : 55: predicate.minmaximum_grad 0.37% : 0.000003s : 6: predicate.mutable_eliminate 0.15% : 0.000001s : 6: predicate.opt_reshape 0.18% : 0.000002s : 6: predicate.parallel_virtual_node 1.99% : 0.000018s : 76: predicate.partial_defer_inline 2.34% : 0.000021s : 68: predicate.partial_eliminate 0.92% : 0.000008s : 55: predicate.print_const_string_wrapper 0.50% : 0.000005s : 29: predicate.reduce_all_const_elim 1.22% : 0.000011s : 55: predicate.reduce_eliminate 2.10% : 0.000019s : 129: predicate.redundant_stop_gradient_eliminater 0.36% : 0.000003s : 29: predicate.remove_not_recompute_node 1.99% : 0.000018s : 190: predicate.replace_applicator 0.90% : 0.000008s : 97: predicate.replace_old_param 0.08% : 0.000001s : 6: predicate.reset_defer_inline 0.96% : 0.000009s : 55: predicate.reshape_eliminate 1.90% : 0.000017s : 126: predicate.row_tensor_add_zeros_like 0.13% : 0.000001s : 6: predicate.row_tensor_eliminate 2.08% : 0.000019s : 130: predicate.same_eliminate 0.32% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.60% : 0.000005s : 29: predicate.shard_identity_eliminate 0.25% : 0.000002s : 12: predicate.special_op_eliminate 0.56% : 0.000005s : 29: predicate.specialize_transform 2.19% : 0.000020s : 126: predicate.split_environ_get_set_with_tuple_value 1.83% : 0.000016s : 97: predicate.stack_unstack_eliminate 0.13% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.27% : 0.000011s : 76: predicate.switch_defer_inline 3.14% : 0.000028s : 206: predicate.switch_layer_defer_inline 3.95% : 0.000035s : 240: predicate.switch_simplify 0.89% : 0.000008s : 55: predicate.tile_eliminate 0.95% : 0.000009s : 55: predicate.transpose_eliminate 1.30% : 0.000012s : 67: predicate.tuple_list_convert_item_index_to_positive 1.26% : 0.000011s : 67: predicate.tuple_list_get_item_const_eliminator 1.09% : 0.000010s : 67: predicate.tuple_list_get_item_depend_reorder 2.66% : 0.000024s : 103: predicate.tuple_list_get_item_eliminator 1.17% : 0.000010s : 67: predicate.tuple_list_get_set_item_eliminator 1.80% : 0.000016s : 96: predicate.tuple_list_set_item_eliminator 1.29% : 0.000012s : 74: predicate.tuple_to_list_eliminator_ 2.04% : 0.000018s : 129: predicate.updatestate_pure_node_eliminater 2.73% : 0.000024s : 159: predicate.updatestate_useless_node_eliminater 0.13% : 0.000001s : 6: predicate.value_based_eliminate 0.50% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.54% : 0.000005s : 29: predicate.virtual_output_eliminate 0.12% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.14% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003210 48 61.45% : 0.001973s : 24: func_graph_cloner_run.FuncGraphClonerGraph 38.55% : 0.001237s : 24: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.882153 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.47% : 0.004178s : 1: add_attr 0.47% : 0.004164s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000104s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.03% : 0.000309s : 1: auto_monad 0.01% : 0.000090s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.07% : 0.000576s : 1: bootstrap 0.00% : 0.000037s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000025s : 1: control_data_broadcast_order 0.00% : 0.000013s : 1: convert_after_rewriter 0.00% : 0.000040s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.00% : 0.000011s : 1: environ_conv 0.01% : 0.000086s : 1: event_method 0.00% : 0.000016s : 1: execute 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% : 0.000007s : 1: insert-virtual-dataset 0.00% : 0.000004s : 1: interleave_parallel_branches 0.00% : 0.000005s : 1: interleave_split_concat_branches 0.00% : 0.000006s : 1: label_fine_grained_interleaved_index 0.00% : 0.000008s : 1: label_micro_interleaved_index 0.06% : 0.000523s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.09% : 0.000800s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000024s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000026s : 1: opt.transform.mutable_eliminate 23.90% : 0.210842s : 125: opt.transform.opt_a 0.01% : 0.000061s : 1: opt.transform.opt_after_cconv 0.00% : 0.000039s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000184s : 28: opt.transform.opt_b 0.11% : 0.000930s : 2: opt.transform.opt_resolve 0.01% : 0.000068s : 2: opt.transform.opt_trans_graph 0.01% : 0.000062s : 4: opt.transform.symbol_engine_opt 26.04% : 0.229677s : 1: opt_a 0.02% : 0.000168s : 1: opt_after_cconv 0.07% : 0.000596s : 1: opt_after_jit_grad 0.04% : 0.000325s : 1: opt_b 26.37% : 0.232603s : 1: optimize 0.00% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000011s : 1: order_py_execute_after_rewriter 0.00% : 0.000033s : 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% : 0.000005s : 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% : 0.000079s : 1: pre_auto_parallel 0.01% : 0.000051s : 1: py_interpret_to_execute 0.00% : 0.000024s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000022s : 1: remove_dup_value 0.55% : 0.004849s : 2: renormalize.infer 0.34% : 0.003040s : 2: 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% : 0.000074s : 1: rewriter_after_opt_a 0.02% : 0.000168s : 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% : 0.000013s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000116s : 1: symbol_engine_optimizer 8.14% : 0.071824s : 1: task_emit 0.01% : 0.000099s : 1: tuple_transform 13.01% : 0.114802s : 1: type_inference 0.01% : 0.000090s : 1: validate [WARNING] ME(169920:281473205423920,MainProcess):2026-01-29-17:42:55.995.477 [mindspore/common/tensor.py:142] For 'Tensor', if 'dtype' is not None, 'input_data', 'shape' or 'init' must not be None. [WARNING] ME(169920:281473205423920,MainProcess):2026-01-29-17:42:56.184.854 [mindspore/common/tensor.py:142] For 'Tensor', if 'dtype' is not None, 'input_data', 'shape' or 'init' must not be None. Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. *****************End Generalize test PYNATIVE_MODE Deterministic case***************** *****************Start Generalize test GRAPH_MODE_O0 DiscontiguousInput case***************** [WARNING]: TEST_OP_GENERALIZE GRAPH_MODE_O0 'DiscontiguousInput' is skipped temporarily with 'GRAPH_MODE_O0' and 'GRAPH_MODE_GE', it's will cause accuracy issue after 'Tensor storage refactor'. *****************End Generalize test GRAPH_MODE_O0 DiscontiguousInput case***************** *****************Start Generalize test GRAPH_MODE_O0 ViewTensor case***************** [INFO]: TEST_OP_GENERALIZE generate view input with shape (2, 3), origin input with shape (2, 3) [INFO]: TEST_OP_GENERALIZE Start to test no view input [INFO]: TEST_OP_GENERALIZE Start to test view input [INFO]: TEST_OP_GENERALIZE Start to compare view out with no view out Compare ['Success'] for 0'th output. *****************End Generalize test GRAPH_MODE_O0 ViewTensor case***************** *****************Start Generalize test GRAPH_MODE_O0 EmptyTensor case***************** [WARNING]: TEST_OP_GENERALIZE GRAPH_MODE_O0 'EmptyTensor' in 'disable_case', EmptyTensor case is skipped. *****************End Generalize test GRAPH_MODE_O0 EmptyTensor case***************** *****************Start Generalize test GRAPH_MODE_O0 ScalarTensor case***************** [INFO]: TEST_OP_GENERALIZE Start to test scalar tensor [INFO]: TEST_OP_GENERALIZE Start to test 1d tensor [INFO]: TEST_OP_GENERALIZE Start to compare scalar tensor out with 1d tensor out Compare ['Success'] for 0'th output. *****************End Generalize test GRAPH_MODE_O0 ScalarTensor case***************** *****************Start Generalize test GRAPH_MODE_O0 GradByRequirement case***************** [INFO]: TEST_OP_GENERALIZE Start to test all grads [INFO]: TEST_OP_GENERALIZE Start to test grad by requirement on position 0 [INFO]: TEST_OP_GENERALIZE Start to compare grad by requirement with all grads Compare ['Success'] for 0'th output. *****************End Generalize test GRAPH_MODE_O0 GradByRequirement case***************** *****************Start Generalize test GRAPH_MODE_O0 Deterministic case***************** [INFO]: TEST_OP_GENERALIZE test deterministic case loop 0 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 1 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 2 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 3 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE test deterministic case loop 4 [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func forward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare forward second time out with first time out Compare ['Success'] for 0'th output. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic first time. [INFO]: TEST_OP_GENERALIZE Start to test func backward with deterministic second time. [INFO]: TEST_OP_GENERALIZE Start to compare backward second time out with first time out Compare ['Success'] for 0'th output. *****************End Generalize test GRAPH_MODE_O0 Deterministic case***************** [WARNING]: TEST_OP_GENERALIZE GRAPH_MODE_GE is skipped. ************************************************************************************************** **************************************TEST_OP_GENERALIZE End************************************** ************************************************************************************************** ************************************************************************************************* **************************************TEST_OP_DYNAMIC Start************************************** ************************************************************************************************* [WARNING]: TEST_DYNAMIC GRAPH_MODE_GE is skipped, Ge backend don't support dynamic situations any longer. *********************Begin DynamicTest for with PYNATIVE_MODE ********************* Start DynamicTest running on PYNATIVE_MODE with static shape with first inputs... End Start DynamicTest running on PYNATIVE_MODE with static shape with second inputs... End Start DynamicTest testing with [PYNATIVE_MODE] [Resize]... Compare ['Success'] for 0'th output of Resize. End Start DynamicTest testing with [PYNATIVE_MODE] [DYNAMIC_SHAPE] [None]... Compare ['Success'] for 0'th output of Forward. Compare ['Success'] for 0'th output of Grad. End Start DynamicTest testing with [PYNATIVE_MODE] [DYNAMIC_RANK] [None]... Compare ['Success'] for 0'th output of Forward. Compare ['Success'] for 0'th output of Grad. End *********************End DynamicTest for with PYNATIVE_MODE ********************* TotalTime = 0.10423, [24] [bootstrap]: 0.00043107 [type_inference]: 0.00764929 [event_method]: 2.043e-05 [auto_monad]: 0.00016716 [graph_reusing]: 6.34999e-06 [inline]: 2.98e-06 [add_attr]: 0.0036615, [1] [add_attr_with_inline]: 0.00365038, [1] [Cycle 1]: 6.672e-05, [2] [tag_attr]: 2.435e-05 [meta_addattr_fg_expand]: 6.30997e-06 [parallel-infer-symbol]: 4.02998e-06 [pre_auto_parallel]: 3.86e-05 [insert-virtual-dataset]: 2.41998e-06 [parallel-infer-symbol-second]: 7.50006e-07 [dataset_repeat_opt]: 2.07001e-06 [pipeline_split]: 2.04e-06 [optimize]: 0.0823036, [53] [py_interpret_to_execute]: 2.851e-05 [rewriter_before_opt_a]: 8.725e-05 [opt_a]: 0.00329019, [2] [Cycle 1]: 0.00244846, [45] [expand_dump_flag]: 3.04999e-06 [switch_simplify]: 4.844e-05 [loop_unroll]: 3.465e-05 [a_1]: 0.00063956 [with_stream_mark]: 1.744e-05 [recompute_prepare]: 1.164e-05 [updatestate_depend_eliminate]: 6.42001e-06 [updatestate_assign_eliminate]: 6.07001e-06 [updatestate_loads_eliminate]: 5.19e-06 [parameter_eliminate]: 1.79e-06 [a_2]: 0.00012582 [accelerated_algorithm]: 9.14e-06 [shard]: 1.86003e-06 [meta_shard_fg_expand]: 2.32999e-06 [shard_inline]: 9.77001e-06 [merge_send_recv]: 1.245e-05 [auto_parallel]: 8.38999e-06 [parallel]: 1.824e-05 [flash_sp]: 9.48002e-06 [merge_comm]: 5.79e-06 [allreduce_fusion]: 5.35001e-06 [matmul_add_comm_reduction]: 1.127e-05 [allreduce_slice_to_reducescatter]: 6.19999e-07 [virtual_shard_identity]: 1.06e-05 [virtual_dataset]: 8.77e-06 [get_grad_eliminate_]: 8.80999e-06 [virtual_output]: 9.17999e-06 [merge_forward]: 5.84999e-06 [cell_reuse_recompute_pass]: 1.07998e-06 [offload_activation]: 1.388e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.801e-05 [merge_recompute_call_nodes]: 1.37e-06 [before_grad]: 1.544e-05 [set_forward_comm_id_for_comm_node_pass]: 5.57001e-06 [meta_fg_expand]: 4.30999e-06 [flash_sp_send_recv_attached]: 2.36998e-06 [receive_attached]: 2.39001e-06 [after_resolve]: 1.286e-05 [a_after_grad]: 1.319e-05 [renormalize]: 0.00092123 [add_forward_monad_depend]: 5.32001e-06 [auto_monad_grad]: 2.84999e-06 [auto_monad_eliminator]: 2.369e-05 [cse]: 3.996e-05 [a_3]: 6.742e-05 [Cycle 2]: 0.00083115, [45] [expand_dump_flag]: 2.29999e-06 [switch_simplify]: 1.009e-05 [loop_unroll]: 8.11002e-06 [a_1]: 0.00019764 [with_stream_mark]: 1.481e-05 [recompute_prepare]: 8.74e-06 [updatestate_depend_eliminate]: 5.67999e-06 [updatestate_assign_eliminate]: 4.29002e-06 [updatestate_loads_eliminate]: 4.83001e-06 [parameter_eliminate]: 1.30999e-06 [a_2]: 0.00010901 [accelerated_algorithm]: 8.13999e-06 [shard]: 1.68002e-06 [meta_shard_fg_expand]: 1.84998e-06 [shard_inline]: 8.17e-06 [merge_send_recv]: 7.93001e-06 [auto_parallel]: 8.02e-06 [parallel]: 6.07999e-06 [flash_sp]: 4.16001e-06 [merge_comm]: 4.93001e-06 [allreduce_fusion]: 4.85999e-06 [matmul_add_comm_reduction]: 8.52998e-06 [allreduce_slice_to_reducescatter]: 7.49977e-07 [virtual_shard_identity]: 9.11002e-06 [virtual_dataset]: 7.92e-06 [get_grad_eliminate_]: 7.36001e-06 [virtual_output]: 8.02003e-06 [merge_forward]: 4.4e-06 [cell_reuse_recompute_pass]: 1.84e-06 [offload_activation]: 1.028e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.642e-05 [merge_recompute_call_nodes]: 8.50006e-07 [before_grad]: 1.408e-05 [set_forward_comm_id_for_comm_node_pass]: 5.65001e-06 [meta_fg_expand]: 3.21999e-06 [flash_sp_send_recv_attached]: 1.40001e-06 [receive_attached]: 1.54998e-06 [after_resolve]: 1.16e-05 [a_after_grad]: 1.215e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 1.30001e-06 [auto_monad_grad]: 1.24e-06 [auto_monad_eliminator]: 1.326e-05 [cse]: 2.2e-05 [a_3]: 4.967e-05 [py_interpret_to_execute_after_opt_a]: 1.196e-05 [slice_cell_reuse_recomputed_activation]: 2.29999e-06 [rewriter_after_opt_a]: 0.00018519 [convert_after_rewriter]: 1.012e-05 [order_py_execute_after_rewriter]: 6.69001e-06 [mutable_eliminate]: 0.00062382 [opt_b]: 0.00027947, [1] [Cycle 1]: 0.00027238, [7] [b_1]: 0.00018307 [b_2]: 1.057e-05 [updatestate_depend_eliminate]: 8.13001e-06 [updatestate_assign_eliminate]: 4.43001e-06 [updatestate_loads_eliminate]: 4.50999e-06 [renormalize]: 7.00005e-07 [cse]: 2.58e-05 [optimize_parallel_all_gather_comm]: 2.068e-05 [overlap_param_gather]: 2.02999e-06 [cconv]: 2.689e-05 [loop_unroll]: 0.00044334 [opt_after_cconv]: 0.0001398, [1] [Cycle 1]: 0.00013351, [7] [c_1]: 5.246e-05 [parameter_eliminate]: 3.09999e-06 [updatestate_depend_eliminate]: 8.92999e-06 [updatestate_assign_eliminate]: 5.18002e-06 [updatestate_loads_eliminate]: 4.60999e-06 [cse]: 2.386e-05 [renormalize]: 5.39992e-07 [remove_dup_value]: 1.478e-05 [tuple_transform]: 9.061e-05, [1] [Cycle 1]: 8.578e-05, [4] [d_1]: 5.591e-05 [none_parameter_eliminate]: 1.67999e-06 [renormalize]: 2.59985e-07 [switch_simplify]: 9.42001e-06 [partial_unused_args_eliminate]: 1.80001e-06 [add_recomputation]: 6.949e-05 [cse_after_recomputation]: 2.806e-05, [1] [Cycle 1]: 2.314e-05, [1] [cse]: 1.772e-05 [environ_conv]: 7.26999e-06 [swap_dp_allreduce_reducescatter]: 7.83999e-06 [bias_add_comm_swap]: 3.4e-06 [label_micro_interleaved_index]: 5.27999e-06 [label_fine_grained_interleaved_index]: 2.86999e-06 [merge_cast_opt]: 1.32999e-06 [slice_recompute_activation]: 2.06e-06 [micro_interleaved_order_control]: 2.31998e-06 [assign_add_opt]: 1.54e-06 [ForceFp32Comm]: 7.90023e-07 [remove_cast_before_assign_add]: 7.23e-06 [full_micro_interleaved_order_control]: 1.088e-05 [reorder_send_recv_between_fp_bp]: 3.11001e-06 [comm_op_add_attrs]: 1.14e-06 [add_comm_op_reuse_tag]: 1.00999e-06 [interleave_split_concat_branches]: 1.44e-06 [interleave_parallel_branches]: 1.05001e-06 [overlap_opt_shard_in_pipeline]: 1.62001e-06 [overlap_opt_shard_grad_in_pipeline]: 2.02001e-06 [control_data_broadcast_order]: 6.356e-05 [grouped_pairwise_exchange_alltoall]: 2.07999e-06 [offloading_packed_experts]: 8.42e-06 [overlap_recompute_and_grad_model_parallel]: 6.86999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.26002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.38002e-06 [overlap_recompute_comm]: 2.17999e-06 [overlap_grad_ring_attention]: 6.01e-06 [overlap_grad_flash_sp]: 3.828e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.17001e-06 [split_layernorm_comm]: 1.83002e-06 [handle_group_info]: 1.10001e-06 [symbol_engine_optimizer]: 0.0001621, [1] [Cycle 1]: 0.00015121, [6] [build]: 1.227e-05 [elim_shapecalc]: 3.554e-05 [elim_not_effective]: 2.94e-05 [opt_reshape]: 1.136e-05 [fold_const_symbol]: 1.527e-05 [renormalize]: 4.80009e-07 [detach_backward]: 2.29999e-06 [pipeline_parallel_scheduler]: 1.60999e-06 [auto_monad_reorder]: 3.741e-05 [get_jit_bprop_graph]: 2.11e-06 [rewriter_after_jit_bprop_graph]: 0.0002028 [opt_after_jit_grad]: 0.00083898 [validate]: 5.897e-05 [backend_pass]: 8.79983e-07 [task_emit]: 0.00853668 [execute]: 1.068e-05 Sums bootstrap : 0.000431s : 1.86% type_inference : 0.007649s : 33.06% event_method : 0.000020s : 0.09% auto_monad : 0.000167s : 0.72% graph_reusing : 0.000006s : 0.03% inline : 0.000003s : 0.01% add_attr.add_attr_with_inline.tag_attr : 0.000024s : 0.11% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000006s : 0.03% parallel-infer-symbol : 0.000004s : 0.02% pre_auto_parallel : 0.000039s : 0.17% insert-virtual-dataset : 0.000002s : 0.01% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.01% pipeline_split : 0.000002s : 0.01% optimize.py_interpret_to_execute : 0.000029s : 0.12% optimize.rewriter_before_opt_a : 0.000087s : 0.38% optimize.opt_a.expand_dump_flag : 0.000005s : 0.02% optimize.opt_a.switch_simplify : 0.000059s : 0.25% optimize.opt_a.loop_unroll : 0.000043s : 0.18% optimize.opt_a.a_1 : 0.000837s : 3.62% optimize.opt_a.with_stream_mark : 0.000032s : 0.14% optimize.opt_a.recompute_prepare : 0.000020s : 0.09% optimize.opt_a.updatestate_depend_eliminate : 0.000012s : 0.05% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.04% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.04% optimize.opt_a.parameter_eliminate : 0.000003s : 0.01% optimize.opt_a.a_2 : 0.000235s : 1.02% optimize.opt_a.accelerated_algorithm : 0.000017s : 0.07% optimize.opt_a.shard : 0.000004s : 0.02% optimize.opt_a.meta_shard_fg_expand : 0.000004s : 0.02% optimize.opt_a.shard_inline : 0.000018s : 0.08% optimize.opt_a.merge_send_recv : 0.000020s : 0.09% optimize.opt_a.auto_parallel : 0.000016s : 0.07% optimize.opt_a.parallel : 0.000024s : 0.11% optimize.opt_a.flash_sp : 0.000014s : 0.06% optimize.opt_a.merge_comm : 0.000011s : 0.05% optimize.opt_a.allreduce_fusion : 0.000010s : 0.04% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.09% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.01% optimize.opt_a.virtual_shard_identity : 0.000020s : 0.09% optimize.opt_a.virtual_dataset : 0.000017s : 0.07% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.07% optimize.opt_a.virtual_output : 0.000017s : 0.07% optimize.opt_a.merge_forward : 0.000010s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.01% optimize.opt_a.offload_activation : 0.000024s : 0.10% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000034s : 0.15% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.01% optimize.opt_a.before_grad : 0.000030s : 0.13% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.05% optimize.opt_a.meta_fg_expand : 0.000008s : 0.03% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.02% optimize.opt_a.receive_attached : 0.000004s : 0.02% optimize.opt_a.after_resolve : 0.000024s : 0.11% optimize.opt_a.a_after_grad : 0.000025s : 0.11% optimize.opt_a.renormalize : 0.000921s : 3.98% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.03% optimize.opt_a.auto_monad_grad : 0.000004s : 0.02% optimize.opt_a.auto_monad_eliminator : 0.000037s : 0.16% optimize.opt_a.cse : 0.000062s : 0.27% optimize.opt_a.a_3 : 0.000117s : 0.51% optimize.py_interpret_to_execute_after_opt_a : 0.000012s : 0.05% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.01% optimize.rewriter_after_opt_a : 0.000185s : 0.80% optimize.convert_after_rewriter : 0.000010s : 0.04% optimize.order_py_execute_after_rewriter : 0.000007s : 0.03% optimize.mutable_eliminate : 0.000624s : 2.70% optimize.opt_b.b_1 : 0.000183s : 0.79% optimize.opt_b.b_2 : 0.000011s : 0.05% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.04% optimize.opt_b.updatestate_assign_eliminate : 0.000004s : 0.02% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.02% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000026s : 0.11% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.09% optimize.overlap_param_gather : 0.000002s : 0.01% optimize.cconv : 0.000027s : 0.12% optimize.loop_unroll : 0.000443s : 1.92% optimize.opt_after_cconv.c_1 : 0.000052s : 0.23% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.01% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 0.04% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.02% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000005s : 0.02% optimize.opt_after_cconv.cse : 0.000024s : 0.10% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000015s : 0.06% optimize.tuple_transform.d_1 : 0.000056s : 0.24% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000009s : 0.04% optimize.partial_unused_args_eliminate : 0.000002s : 0.01% optimize.add_recomputation : 0.000069s : 0.30% optimize.cse_after_recomputation.cse : 0.000018s : 0.08% optimize.environ_conv : 0.000007s : 0.03% optimize.swap_dp_allreduce_reducescatter : 0.000008s : 0.03% optimize.bias_add_comm_swap : 0.000003s : 0.01% optimize.label_micro_interleaved_index : 0.000005s : 0.02% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.01% optimize.merge_cast_opt : 0.000001s : 0.01% optimize.slice_recompute_activation : 0.000002s : 0.01% optimize.micro_interleaved_order_control : 0.000002s : 0.01% optimize.assign_add_opt : 0.000002s : 0.01% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000007s : 0.03% optimize.full_micro_interleaved_order_control : 0.000011s : 0.05% 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 : 0.01% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000002s : 0.01% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.01% optimize.control_data_broadcast_order : 0.000064s : 0.27% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.01% optimize.offloading_packed_experts : 0.000008s : 0.04% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.03% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.01% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.01% optimize.overlap_recompute_comm : 0.000002s : 0.01% optimize.overlap_grad_ring_attention : 0.000006s : 0.03% optimize.overlap_grad_flash_sp : 0.000038s : 0.17% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000002s : 0.01% optimize.split_layernorm_comm : 0.000002s : 0.01% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000012s : 0.05% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000036s : 0.15% optimize.symbol_engine_optimizer.elim_not_effective : 0.000029s : 0.13% optimize.symbol_engine_optimizer.opt_reshape : 0.000011s : 0.05% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 0.07% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.01% pipeline_parallel_scheduler : 0.000002s : 0.01% auto_monad_reorder : 0.000037s : 0.16% get_jit_bprop_graph : 0.000002s : 0.01% rewriter_after_jit_bprop_graph : 0.000203s : 0.88% opt_after_jit_grad : 0.000839s : 3.63% validate : 0.000059s : 0.25% backend_pass : 0.000001s : 0.00% task_emit : 0.008537s : 36.90% execute : 0.000011s : 0.05% Time group info: ------[substitution.] 0.000237 60 2.96% : 0.000007s : 2: substitution.depend_value_elim 1.32% : 0.000003s : 5: substitution.elim_not_effective 1.05% : 0.000002s : 5: substitution.fold_const_symbol 3.10% : 0.000007s : 6: substitution.graph_param_transform 67.66% : 0.000161s : 4: substitution.inline 2.42% : 0.000006s : 10: substitution.j_node_and_user_rematch 3.35% : 0.000008s : 10: substitution.remove_not_recompute_node 1.71% : 0.000004s : 2: substitution.replace_old_param 4.90% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator 6.32% : 0.000015s : 6: substitution.updatestate_pure_node_eliminater 5.21% : 0.000012s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.007586 2 89.35% : 0.006778s : 1: type_inference.infer 10.65% : 0.000808s : 1: type_inference.specialize ------[replace.] 0.000060 6 69.35% : 0.000042s : 4: replace.inline 30.65% : 0.000018s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000169 6 93.83% : 0.000158s : 4: match.inline 6.17% : 0.000010s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000265 1652 0.85% : 0.000002s : 16: predicate.accumulaten_eliminater 1.34% : 0.000004s : 6: predicate.ad_related_special_op_eliminate 0.63% : 0.000002s : 12: predicate.addn_check_dump 0.88% : 0.000002s : 16: predicate.addn_zero_filter 0.78% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.15% : 0.000006s : 28: predicate.arithmetic_simplify 0.99% : 0.000003s : 16: predicate.cast_eliminate 0.73% : 0.000002s : 12: predicate.check_bprop_eliminate 0.62% : 0.000002s : 12: predicate.compare_switch_simplify 0.20% : 0.000001s : 6: predicate.const_output_eliminate 0.70% : 0.000002s : 12: predicate.depend_value_elim 0.91% : 0.000002s : 16: predicate.dict_get_item_const_eliminator 0.91% : 0.000002s : 16: predicate.dict_get_item_eliminator 0.87% : 0.000002s : 16: predicate.dict_set_item_eliminator 1.28% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.45% : 0.000001s : 6: predicate.elim_not_effective 0.79% : 0.000002s : 6: predicate.elim_shapecalc_of_broadcastargs 1.14% : 0.000003s : 22: predicate.environ_add_const_eliminate 1.10% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.09% : 0.000003s : 22: predicate.environ_get_depend_swap 1.84% : 0.000005s : 34: predicate.environ_get_eliminate 1.08% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.18% : 0.000003s : 22: predicate.exchange_switch_depend_value 1.78% : 0.000005s : 22: predicate.float_depend_g_call 0.64% : 0.000002s : 12: predicate.float_environ_get_switch 0.93% : 0.000002s : 18: predicate.float_tuple_getitem_switch 0.23% : 0.000001s : 6: predicate.fold_const_symbol 0.68% : 0.000002s : 12: predicate.get_grad_eliminate 0.28% : 0.000001s : 6: predicate.graph_param_transform 0.66% : 0.000002s : 12: predicate.incorporate_call 0.58% : 0.000002s : 12: predicate.incorporate_call_switch 5.88% : 0.000016s : 74: predicate.inline 0.88% : 0.000002s : 12: predicate.inline_without_move 0.33% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.84% : 0.000002s : 12: predicate.less_batch_normalization 1.62% : 0.000004s : 30: predicate.list_to_tuple_eliminator_ 2.40% : 0.000006s : 46: predicate.load_eliminater 0.94% : 0.000003s : 6: predicate.loop_unroll_after_grad 2.13% : 0.000006s : 38: predicate.loop_unroll_before_grad 1.73% : 0.000005s : 28: predicate.make_slice_get_slice_eliminator 0.65% : 0.000002s : 12: predicate.merge_addn 0.73% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.62% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.79% : 0.000002s : 16: predicate.minmaximum_grad 1.02% : 0.000003s : 6: predicate.mutable_eliminate 0.53% : 0.000001s : 6: predicate.opt_reshape 0.42% : 0.000001s : 6: predicate.parallel_virtual_node 1.48% : 0.000004s : 22: predicate.partial_defer_inline 1.39% : 0.000004s : 24: predicate.partial_eliminate 0.87% : 0.000002s : 16: predicate.print_const_string_wrapper 0.63% : 0.000002s : 12: predicate.reduce_all_const_elim 1.07% : 0.000003s : 16: predicate.reduce_eliminate 2.41% : 0.000006s : 46: predicate.redundant_stop_gradient_eliminater 0.39% : 0.000001s : 12: predicate.remove_not_recompute_node 1.24% : 0.000003s : 30: predicate.replace_applicator 0.58% : 0.000002s : 12: predicate.replace_old_param 0.28% : 0.000001s : 6: predicate.reset_defer_inline 0.88% : 0.000002s : 16: predicate.reshape_eliminate 0.74% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.56% : 0.000001s : 6: predicate.row_tensor_eliminate 1.01% : 0.000003s : 12: predicate.same_eliminate 0.43% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.85% : 0.000002s : 12: predicate.shard_identity_eliminate 0.83% : 0.000002s : 12: predicate.special_op_eliminate 0.83% : 0.000002s : 12: predicate.specialize_transform 1.06% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 0.86% : 0.000002s : 12: predicate.stack_unstack_eliminate 0.42% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.25% : 0.000003s : 22: predicate.switch_defer_inline 1.90% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.96% : 0.000013s : 78: predicate.switch_simplify 0.85% : 0.000002s : 16: predicate.tile_eliminate 0.91% : 0.000002s : 16: predicate.transpose_eliminate 1.59% : 0.000004s : 28: predicate.tuple_list_convert_item_index_to_positive 1.56% : 0.000004s : 28: predicate.tuple_list_get_item_const_eliminator 1.45% : 0.000004s : 28: predicate.tuple_list_get_item_depend_reorder 2.96% : 0.000008s : 42: predicate.tuple_list_get_item_eliminator 1.52% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.35% : 0.000006s : 40: predicate.tuple_list_set_item_eliminator 1.58% : 0.000004s : 30: predicate.tuple_to_list_eliminator_ 2.39% : 0.000006s : 46: predicate.updatestate_pure_node_eliminater 3.40% : 0.000009s : 58: predicate.updatestate_useless_node_eliminater 0.44% : 0.000001s : 6: predicate.value_based_eliminate 0.68% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.75% : 0.000002s : 12: predicate.virtual_output_eliminate 0.39% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.44% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000569 11 49.15% : 0.000280s : 5: func_graph_cloner_run.FuncGraphClonerGraph 50.85% : 0.000290s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.192874 196 39.55% : 0.076281s : 1: ForceFp32Comm 1.90% : 0.003668s : 1: add_attr 1.89% : 0.003655s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.04% : 0.000074s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.09% : 0.000175s : 1: auto_monad 0.02% : 0.000041s : 1: auto_monad_reorder 0.00% : 0.000007s : 1: backend_pass 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.24% : 0.000458s : 1: bootstrap 0.02% : 0.000030s : 1: cconv 0.00% : 0.000006s : 1: comm_op_add_attrs 0.04% : 0.000068s : 1: control_data_broadcast_order 0.01% : 0.000014s : 1: convert_after_rewriter 0.02% : 0.000031s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000006s : 1: detach_backward 0.01% : 0.000011s : 1: environ_conv 0.01% : 0.000027s : 1: event_method 0.01% : 0.000017s : 1: execute 0.01% : 0.000017s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.01% : 0.000011s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000007s : 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 0.23% : 0.000451s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.33% : 0.000632s : 1: mutable_eliminate 0.01% : 0.000021s : 1: offloading_packed_experts 0.01% : 0.000017s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000019s : 1: opt.transform.mutable_eliminate 0.75% : 0.001445s : 78: opt.transform.opt_a 0.03% : 0.000051s : 1: opt.transform.opt_after_cconv 0.02% : 0.000041s : 1: opt.transform.opt_after_jit_grad 0.09% : 0.000166s : 28: opt.transform.opt_b 0.03% : 0.000063s : 2: opt.transform.opt_trans_graph 0.04% : 0.000083s : 4: opt.transform.symbol_engine_opt 1.71% : 0.003294s : 1: opt_a 0.07% : 0.000144s : 1: opt_after_cconv 0.44% : 0.000851s : 1: opt_after_jit_grad 0.15% : 0.000283s : 1: opt_b 42.68% : 0.082311s : 1: optimize 0.01% : 0.000024s : 1: optimize_parallel_all_gather_comm 0.01% : 0.000010s : 1: order_py_execute_after_rewriter 0.02% : 0.000042s : 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% : 0.000005s : 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% : 0.000010s : 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 0.02% : 0.000043s : 1: pre_auto_parallel 0.02% : 0.000032s : 1: py_interpret_to_execute 0.01% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.02% : 0.000041s : 1: remove_cast_before_assign_add 0.01% : 0.000018s : 1: remove_dup_value 0.28% : 0.000536s : 1: renormalize.infer 0.20% : 0.000376s : 1: renormalize.specialize 0.00% : 0.000007s : 1: reorder_send_recv_between_fp_bp 0.11% : 0.000210s : 1: rewriter_after_jit_bprop_graph 0.10% : 0.000191s : 1: rewriter_after_opt_a 0.05% : 0.000092s : 1: rewriter_before_opt_a 0.00% : 0.000005s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.01% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.09% : 0.000165s : 1: symbol_engine_optimizer 4.44% : 0.008557s : 1: task_emit 0.05% : 0.000094s : 1: tuple_transform 3.98% : 0.007668s : 1: type_inference 0.05% : 0.000094s : 1: validate TotalTime = 0.438713, [24] [bootstrap]: 0.00044212 [type_inference]: 0.141761 [event_method]: 0.00011218 [auto_monad]: 0.00031042 [graph_reusing]: 1.017e-05 [inline]: 3.14001e-06 [add_attr]: 0.00367318, [1] [add_attr_with_inline]: 0.00366233, [1] [Cycle 1]: 0.00010969, [2] [tag_attr]: 4.286e-05 [meta_addattr_fg_expand]: 1.114e-05 [parallel-infer-symbol]: 3.98999e-06 [pre_auto_parallel]: 6.177e-05 [insert-virtual-dataset]: 2.73998e-06 [parallel-infer-symbol-second]: 6.69999e-07 [dataset_repeat_opt]: 2.39999e-06 [pipeline_split]: 1.71e-06 [optimize]: 0.158938, [53] [py_interpret_to_execute]: 4.55e-05 [rewriter_before_opt_a]: 0.00016326 [opt_a]: 0.156303, [3] [Cycle 1]: 0.152141, [45] [expand_dump_flag]: 4.08001e-06 [switch_simplify]: 8.187e-05 [loop_unroll]: 6.603e-05 [a_1]: 0.00150084 [with_stream_mark]: 3.147e-05 [recompute_prepare]: 2.482e-05 [updatestate_depend_eliminate]: 1.302e-05 [updatestate_assign_eliminate]: 1.044e-05 [updatestate_loads_eliminate]: 9.94999e-06 [parameter_eliminate]: 2.86e-06 [a_2]: 0.00028715 [accelerated_algorithm]: 1.874e-05 [shard]: 2.11e-06 [meta_shard_fg_expand]: 4.87e-06 [shard_inline]: 1.789e-05 [merge_send_recv]: 2.061e-05 [auto_parallel]: 1.442e-05 [parallel]: 2.017e-05 [flash_sp]: 1.241e-05 [merge_comm]: 1.16e-05 [allreduce_fusion]: 1.066e-05 [matmul_add_comm_reduction]: 3.437e-05 [allreduce_slice_to_reducescatter]: 8.00006e-07 [virtual_shard_identity]: 2.081e-05 [virtual_dataset]: 1.829e-05 [get_grad_eliminate_]: 1.759e-05 [virtual_output]: 1.773e-05 [merge_forward]: 1.117e-05 [cell_reuse_recompute_pass]: 1.29e-06 [offload_activation]: 2.18e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.455e-05 [merge_recompute_call_nodes]: 1.79e-06 [before_grad]: 3.379e-05 [set_forward_comm_id_for_comm_node_pass]: 1.239e-05 [meta_fg_expand]: 0.0066147 [flash_sp_send_recv_attached]: 5.16998e-06 [receive_attached]: 2.49001e-06 [after_resolve]: 0.00015866 [a_after_grad]: 0.00018402 [renormalize]: 0.141247 [add_forward_monad_depend]: 1.519e-05 [auto_monad_grad]: 8.18999e-06 [auto_monad_eliminator]: 7.446e-05 [cse]: 0.0002292 [a_3]: 0.00085862 [Cycle 2]: 0.00321487, [45] [expand_dump_flag]: 3.35e-06 [switch_simplify]: 5.648e-05 [loop_unroll]: 5.117e-05 [a_1]: 0.00124378 [with_stream_mark]: 2.411e-05 [recompute_prepare]: 1.316e-05 [updatestate_depend_eliminate]: 3.332e-05 [updatestate_assign_eliminate]: 6.16e-06 [updatestate_loads_eliminate]: 5.00999e-06 [parameter_eliminate]: 2.01e-06 [a_2]: 0.00011937 [accelerated_algorithm]: 1.009e-05 [shard]: 2.81999e-06 [meta_shard_fg_expand]: 3.05998e-06 [shard_inline]: 9.21002e-06 [merge_send_recv]: 1.128e-05 [auto_parallel]: 1.155e-05 [parallel]: 8.89e-06 [flash_sp]: 3.93001e-06 [merge_comm]: 5.42001e-06 [allreduce_fusion]: 5.40999e-06 [matmul_add_comm_reduction]: 1.267e-05 [allreduce_slice_to_reducescatter]: 1.02998e-06 [virtual_shard_identity]: 1.055e-05 [virtual_dataset]: 8.65999e-06 [get_grad_eliminate_]: 7.9e-06 [virtual_output]: 8.15999e-06 [merge_forward]: 5.47001e-06 [cell_reuse_recompute_pass]: 1.47999e-06 [offload_activation]: 1.352e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.669e-05 [merge_recompute_call_nodes]: 1.49e-06 [before_grad]: 1.424e-05 [set_forward_comm_id_for_comm_node_pass]: 5.35999e-06 [meta_fg_expand]: 0.00011687 [flash_sp_send_recv_attached]: 2.16e-06 [receive_attached]: 2.38998e-06 [after_resolve]: 1.519e-05 [a_after_grad]: 1.468e-05 [renormalize]: 0.00095041 [add_forward_monad_depend]: 4.85001e-06 [auto_monad_grad]: 1.85001e-06 [auto_monad_eliminator]: 2.099e-05 [cse]: 3.301e-05 [a_3]: 6.361e-05 [Cycle 3]: 0.00092942, [45] [expand_dump_flag]: 1.64998e-06 [switch_simplify]: 1.059e-05 [loop_unroll]: 9.35001e-06 [a_1]: 0.00029082 [with_stream_mark]: 1.379e-05 [recompute_prepare]: 9.12999e-06 [updatestate_depend_eliminate]: 5.42999e-06 [updatestate_assign_eliminate]: 4.27998e-06 [updatestate_loads_eliminate]: 4.32e-06 [parameter_eliminate]: 9.89996e-07 [a_2]: 0.00011479 [accelerated_algorithm]: 8.67e-06 [shard]: 1.64998e-06 [meta_shard_fg_expand]: 2.11e-06 [shard_inline]: 8.42e-06 [merge_send_recv]: 7.21999e-06 [auto_parallel]: 7.75e-06 [parallel]: 5.47001e-06 [flash_sp]: 1.06002e-06 [merge_comm]: 4.97e-06 [allreduce_fusion]: 5.02999e-06 [matmul_add_comm_reduction]: 8.08999e-06 [allreduce_slice_to_reducescatter]: 7.80012e-07 [virtual_shard_identity]: 9.26002e-06 [virtual_dataset]: 8.12e-06 [get_grad_eliminate_]: 7.48e-06 [virtual_output]: 8.17e-06 [merge_forward]: 4.57e-06 [cell_reuse_recompute_pass]: 1.49998e-06 [offload_activation]: 9.84999e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.637e-05 [merge_recompute_call_nodes]: 9.39996e-07 [before_grad]: 1.436e-05 [set_forward_comm_id_for_comm_node_pass]: 5.29e-06 [meta_fg_expand]: 3.11001e-06 [flash_sp_send_recv_attached]: 9.20001e-07 [receive_attached]: 1.37e-06 [after_resolve]: 1.049e-05 [a_after_grad]: 1.284e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 1.20001e-06 [auto_monad_grad]: 1.15001e-06 [auto_monad_eliminator]: 1.284e-05 [cse]: 2.253e-05 [a_3]: 5.279e-05 [py_interpret_to_execute_after_opt_a]: 1.57e-05 [slice_cell_reuse_recomputed_activation]: 2.12001e-06 [rewriter_after_opt_a]: 5.919e-05 [convert_after_rewriter]: 9.86e-06 [order_py_execute_after_rewriter]: 6.74001e-06 [mutable_eliminate]: 0.00071614 [opt_b]: 0.00029013, [1] [Cycle 1]: 0.00028253, [7] [b_1]: 0.00018868 [b_2]: 1.158e-05 [updatestate_depend_eliminate]: 8.81997e-06 [updatestate_assign_eliminate]: 4.81002e-06 [updatestate_loads_eliminate]: 4.26001e-06 [renormalize]: 4.00003e-07 [cse]: 2.761e-05 [optimize_parallel_all_gather_comm]: 2.163e-05 [overlap_param_gather]: 2.23002e-06 [cconv]: 2.701e-05 [loop_unroll]: 0.00045784 [opt_after_cconv]: 0.00013994, [1] [Cycle 1]: 0.00013401, [7] [c_1]: 5.293e-05 [parameter_eliminate]: 2.68e-06 [updatestate_depend_eliminate]: 8.20999e-06 [updatestate_assign_eliminate]: 4.70001e-06 [updatestate_loads_eliminate]: 4.43999e-06 [cse]: 2.648e-05 [renormalize]: 2.00002e-07 [remove_dup_value]: 1.784e-05 [tuple_transform]: 9.118e-05, [1] [Cycle 1]: 8.636e-05, [4] [d_1]: 5.792e-05 [none_parameter_eliminate]: 1.77999e-06 [renormalize]: 2.50002e-07 [switch_simplify]: 8.95999e-06 [partial_unused_args_eliminate]: 1.81e-06 [add_recomputation]: 8.376e-05 [cse_after_recomputation]: 3.195e-05, [1] [Cycle 1]: 2.707e-05, [1] [cse]: 2.119e-05 [environ_conv]: 6.84999e-06 [swap_dp_allreduce_reducescatter]: 8.76997e-06 [bias_add_comm_swap]: 2.80002e-06 [label_micro_interleaved_index]: 4.72998e-06 [label_fine_grained_interleaved_index]: 2.75997e-06 [merge_cast_opt]: 1.35999e-06 [slice_recompute_activation]: 2.81e-06 [micro_interleaved_order_control]: 2.21998e-06 [assign_add_opt]: 1.31998e-06 [ForceFp32Comm]: 8.50006e-07 [remove_cast_before_assign_add]: 1.02e-06 [full_micro_interleaved_order_control]: 2.32001e-06 [reorder_send_recv_between_fp_bp]: 2.69999e-06 [comm_op_add_attrs]: 1.14998e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.54e-06 [interleave_parallel_branches]: 1.15999e-06 [overlap_opt_shard_in_pipeline]: 1.33002e-06 [overlap_opt_shard_grad_in_pipeline]: 1.81998e-06 [control_data_broadcast_order]: 1.874e-05 [grouped_pairwise_exchange_alltoall]: 1.90001e-06 [offloading_packed_experts]: 5.47999e-06 [overlap_recompute_and_grad_model_parallel]: 5.69999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.51998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.39e-06 [overlap_recompute_comm]: 2.61e-06 [overlap_grad_ring_attention]: 5.78002e-06 [overlap_grad_flash_sp]: 2.901e-05 [begin_end_overlap_inline]: 5.39992e-07 [split_matmul_comm_elemetwise]: 2.16e-06 [split_layernorm_comm]: 2.19001e-06 [handle_group_info]: 9.80013e-07 [symbol_engine_optimizer]: 9.438e-05, [1] [Cycle 1]: 8.991e-05, [6] [build]: 4.85999e-06 [elim_shapecalc]: 1.334e-05 [elim_not_effective]: 1.776e-05 [opt_reshape]: 9.41e-06 [fold_const_symbol]: 1.429e-05 [renormalize]: 2.3999e-07 [detach_backward]: 2.43998e-06 [pipeline_parallel_scheduler]: 1.85001e-06 [auto_monad_reorder]: 6.915e-05 [get_jit_bprop_graph]: 2.15002e-06 [rewriter_after_jit_bprop_graph]: 5.29e-06 [opt_after_jit_grad]: 0.00049631 [validate]: 5.13e-05 [backend_pass]: 8.80013e-07 [task_emit]: 0.132491 [execute]: 8.98002e-06 Sums bootstrap : 0.000442s : 0.10% type_inference : 0.141761s : 32.69% event_method : 0.000112s : 0.03% auto_monad : 0.000310s : 0.07% graph_reusing : 0.000010s : 0.00% inline : 0.000003s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000043s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000011s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000062s : 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 : 0.000045s : 0.01% optimize.rewriter_before_opt_a : 0.000163s : 0.04% optimize.opt_a.expand_dump_flag : 0.000009s : 0.00% optimize.opt_a.switch_simplify : 0.000149s : 0.03% optimize.opt_a.loop_unroll : 0.000127s : 0.03% optimize.opt_a.a_1 : 0.003035s : 0.70% optimize.opt_a.with_stream_mark : 0.000069s : 0.02% optimize.opt_a.recompute_prepare : 0.000047s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000052s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000021s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000019s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000521s : 0.12% optimize.opt_a.accelerated_algorithm : 0.000038s : 0.01% optimize.opt_a.shard : 0.000007s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000010s : 0.00% optimize.opt_a.shard_inline : 0.000036s : 0.01% optimize.opt_a.merge_send_recv : 0.000039s : 0.01% optimize.opt_a.auto_parallel : 0.000034s : 0.01% optimize.opt_a.parallel : 0.000035s : 0.01% optimize.opt_a.flash_sp : 0.000017s : 0.00% optimize.opt_a.merge_comm : 0.000022s : 0.01% optimize.opt_a.allreduce_fusion : 0.000021s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000055s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000003s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000041s : 0.01% optimize.opt_a.virtual_dataset : 0.000035s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000033s : 0.01% optimize.opt_a.virtual_output : 0.000034s : 0.01% optimize.opt_a.merge_forward : 0.000021s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000045s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000068s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000062s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000023s : 0.01% optimize.opt_a.meta_fg_expand : 0.006735s : 1.55% optimize.opt_a.flash_sp_send_recv_attached : 0.000008s : 0.00% optimize.opt_a.receive_attached : 0.000006s : 0.00% optimize.opt_a.after_resolve : 0.000184s : 0.04% optimize.opt_a.a_after_grad : 0.000212s : 0.05% optimize.opt_a.renormalize : 0.142198s : 32.79% optimize.opt_a.add_forward_monad_depend : 0.000021s : 0.00% optimize.opt_a.auto_monad_grad : 0.000011s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000108s : 0.02% optimize.opt_a.cse : 0.000285s : 0.07% optimize.opt_a.a_3 : 0.000975s : 0.22% optimize.py_interpret_to_execute_after_opt_a : 0.000016s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000059s : 0.01% optimize.convert_after_rewriter : 0.000010s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000716s : 0.17% optimize.opt_b.b_1 : 0.000189s : 0.04% optimize.opt_b.b_2 : 0.000012s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000028s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000027s : 0.01% optimize.loop_unroll : 0.000458s : 0.11% optimize.opt_after_cconv.c_1 : 0.000053s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 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 : 0.000000s : 0.00% optimize.remove_dup_value : 0.000018s : 0.00% optimize.tuple_transform.d_1 : 0.000058s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000084s : 0.02% optimize.cse_after_recomputation.cse : 0.000021s : 0.00% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 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 : 0.000003s : 0.00% optimize.micro_interleaved_order_control : 0.000002s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000002s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000019s : 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 : 0.000002s : 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 : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000029s : 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 : 0.000005s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000013s : 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 : 0.000014s : 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 : 0.000069s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000496s : 0.11% validate : 0.000051s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.132491s : 30.55% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.001622 227 0.79% : 0.000013s : 4: substitution.depend_value_elim 0.17% : 0.000003s : 5: substitution.elim_not_effective 0.18% : 0.000003s : 3: substitution.float_depend_g_call 0.28% : 0.000005s : 2: substitution.float_tuple_getitem_switch 0.14% : 0.000002s : 5: substitution.fold_const_symbol 46.88% : 0.000760s : 4: substitution.getattr_setattr_resolve 0.45% : 0.000007s : 6: substitution.graph_param_transform 0.20% : 0.000003s : 2: substitution.incorporate_call 0.14% : 0.000002s : 2: substitution.incorporate_call_switch 30.64% : 0.000497s : 15: substitution.inline 2.34% : 0.000038s : 5: substitution.inline_without_move 0.78% : 0.000013s : 23: substitution.j_node_and_user_rematch 0.67% : 0.000011s : 6: substitution.minmaximum_grad 0.24% : 0.000004s : 3: substitution.partial_eliminate 0.97% : 0.000016s : 23: substitution.remove_not_recompute_node 3.41% : 0.000055s : 21: substitution.replace_applicator 0.90% : 0.000015s : 16: substitution.replace_old_param 0.18% : 0.000003s : 1: substitution.set_cell_output_no_recompute 1.33% : 0.000022s : 6: substitution.tuple_list_convert_item_index_to_positive 0.58% : 0.000009s : 6: substitution.tuple_list_get_item_const_eliminator 0.86% : 0.000014s : 6: substitution.tuple_list_get_item_depend_reorder 3.21% : 0.000052s : 16: substitution.tuple_list_get_item_eliminator 0.83% : 0.000013s : 6: substitution.tuple_list_get_set_item_eliminator 1.60% : 0.000026s : 18: substitution.updatestate_pure_node_eliminater 2.26% : 0.000037s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.141593 2 10.10% : 0.014296s : 1: type_inference.infer 89.90% : 0.127296s : 1: type_inference.specialize ------[replace.] 0.000840 28 66.92% : 0.000562s : 3: replace.getattr_setattr_resolve 14.67% : 0.000123s : 15: replace.inline 5.65% : 0.000047s : 1: replace.replace_applicator 10.05% : 0.000084s : 8: replace.tuple_list_get_item_eliminator 2.71% : 0.000023s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001247 28 57.36% : 0.000715s : 3: match.getattr_setattr_resolve 39.14% : 0.000488s : 15: match.inline 0.78% : 0.000010s : 1: match.replace_applicator 1.95% : 0.000024s : 8: match.tuple_list_get_item_eliminator 0.77% : 0.000010s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000811 5572 1.02% : 0.000008s : 57: predicate.accumulaten_eliminater 0.29% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.48% : 0.000004s : 29: predicate.addn_check_dump 0.96% : 0.000008s : 57: predicate.addn_zero_filter 0.90% : 0.000007s : 57: predicate.adjust_all_reduce_mul_add 1.97% : 0.000016s : 86: predicate.arithmetic_simplify 1.02% : 0.000008s : 57: predicate.cast_eliminate 2.06% : 0.000017s : 130: predicate.check_bprop_eliminate 0.51% : 0.000004s : 29: predicate.compare_switch_simplify 0.07% : 0.000001s : 6: predicate.const_output_eliminate 0.51% : 0.000004s : 29: predicate.depend_value_elim 1.01% : 0.000008s : 57: predicate.dict_get_item_const_eliminator 1.05% : 0.000009s : 57: predicate.dict_get_item_eliminator 0.96% : 0.000008s : 57: predicate.dict_set_item_eliminator 0.35% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.10% : 0.000001s : 6: predicate.elim_not_effective 0.16% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.09% : 0.000009s : 63: predicate.environ_add_const_eliminate 1.04% : 0.000008s : 63: predicate.environ_get_add_eliminate 1.08% : 0.000009s : 63: predicate.environ_get_depend_swap 1.55% : 0.000013s : 92: predicate.environ_get_eliminate 1.01% : 0.000008s : 63: predicate.environ_get_set_eliminate 1.36% : 0.000011s : 81: predicate.exchange_switch_depend_value 1.96% : 0.000016s : 81: predicate.float_depend_g_call 0.54% : 0.000004s : 29: predicate.float_environ_get_switch 0.59% : 0.000005s : 35: predicate.float_tuple_getitem_switch 0.06% : 0.000000s : 6: predicate.fold_const_symbol 0.53% : 0.000004s : 29: predicate.get_grad_eliminate 0.77% : 0.000006s : 20: predicate.getattr_setattr_resolve 0.10% : 0.000001s : 6: predicate.graph_param_transform 0.52% : 0.000004s : 29: predicate.incorporate_call 0.46% : 0.000004s : 29: predicate.incorporate_call_switch 4.89% : 0.000040s : 208: predicate.inline 2.38% : 0.000019s : 97: predicate.inline_without_move 0.25% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.64% : 0.000005s : 29: predicate.less_batch_normalization 1.29% : 0.000010s : 77: predicate.list_to_tuple_eliminator_ 2.19% : 0.000018s : 134: predicate.load_eliminater 0.36% : 0.000003s : 6: predicate.loop_unroll_after_grad 2.34% : 0.000019s : 142: predicate.loop_unroll_before_grad 1.21% : 0.000010s : 69: predicate.make_slice_get_slice_eliminator 0.52% : 0.000004s : 29: predicate.merge_addn 1.99% : 0.000016s : 126: predicate.micro_step_allgather_replace 1.96% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.94% : 0.000008s : 57: predicate.minmaximum_grad 0.38% : 0.000003s : 6: predicate.mutable_eliminate 0.13% : 0.000001s : 6: predicate.opt_reshape 0.15% : 0.000001s : 6: predicate.parallel_virtual_node 1.69% : 0.000014s : 81: predicate.partial_defer_inline 1.31% : 0.000011s : 71: predicate.partial_eliminate 0.92% : 0.000007s : 57: predicate.print_const_string_wrapper 0.49% : 0.000004s : 29: predicate.reduce_all_const_elim 1.28% : 0.000010s : 57: predicate.reduce_eliminate 2.21% : 0.000018s : 134: predicate.redundant_stop_gradient_eliminater 0.33% : 0.000003s : 29: predicate.remove_not_recompute_node 2.18% : 0.000018s : 193: predicate.replace_applicator 0.98% : 0.000008s : 97: predicate.replace_old_param 0.08% : 0.000001s : 6: predicate.reset_defer_inline 1.02% : 0.000008s : 57: predicate.reshape_eliminate 2.10% : 0.000017s : 126: predicate.row_tensor_add_zeros_like 0.18% : 0.000001s : 6: predicate.row_tensor_eliminate 2.25% : 0.000018s : 130: predicate.same_eliminate 0.33% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.59% : 0.000005s : 29: predicate.shard_identity_eliminate 0.26% : 0.000002s : 12: predicate.special_op_eliminate 0.61% : 0.000005s : 29: predicate.specialize_transform 2.24% : 0.000018s : 126: predicate.split_environ_get_set_with_tuple_value 2.07% : 0.000017s : 97: predicate.stack_unstack_eliminate 0.15% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.48% : 0.000012s : 81: predicate.switch_defer_inline 3.46% : 0.000028s : 211: predicate.switch_layer_defer_inline 4.54% : 0.000037s : 258: predicate.switch_simplify 1.00% : 0.000008s : 57: predicate.tile_eliminate 1.00% : 0.000008s : 57: predicate.transpose_eliminate 1.34% : 0.000011s : 69: predicate.tuple_list_convert_item_index_to_positive 1.42% : 0.000012s : 69: predicate.tuple_list_get_item_const_eliminator 1.24% : 0.000010s : 69: predicate.tuple_list_get_item_depend_reorder 2.43% : 0.000020s : 106: predicate.tuple_list_get_item_eliminator 1.24% : 0.000010s : 69: predicate.tuple_list_get_set_item_eliminator 1.82% : 0.000015s : 98: predicate.tuple_list_set_item_eliminator 1.37% : 0.000011s : 77: predicate.tuple_to_list_eliminator_ 2.28% : 0.000019s : 134: predicate.updatestate_pure_node_eliminater 2.91% : 0.000024s : 164: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 6: predicate.value_based_eliminate 0.54% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.52% : 0.000004s : 29: predicate.virtual_output_eliminate 0.12% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.15% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002925 52 66.64% : 0.001949s : 26: func_graph_cloner_run.FuncGraphClonerGraph 33.36% : 0.000976s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.750627 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.49% : 0.003679s : 1: add_attr 0.49% : 0.003666s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000088s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000320s : 1: auto_monad 0.01% : 0.000074s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.06% : 0.000476s : 1: bootstrap 0.00% : 0.000030s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000022s : 1: control_data_broadcast_order 0.00% : 0.000013s : 1: convert_after_rewriter 0.00% : 0.000035s : 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% : 0.000122s : 1: event_method 0.00% : 0.000016s : 1: execute 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 0.06% : 0.000466s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.10% : 0.000725s : 1: mutable_eliminate 0.00% : 0.000008s : 1: offloading_packed_experts 0.00% : 0.000018s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000021s : 1: opt.transform.mutable_eliminate 0.72% : 0.005441s : 125: opt.transform.opt_a 0.01% : 0.000051s : 1: opt.transform.opt_after_cconv 0.00% : 0.000034s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000171s : 28: opt.transform.opt_b 0.18% : 0.001378s : 2: opt.transform.opt_resolve 0.01% : 0.000064s : 2: opt.transform.opt_trans_graph 0.01% : 0.000051s : 4: opt.transform.symbol_engine_opt 20.82% : 0.156307s : 1: opt_a 0.02% : 0.000144s : 1: opt_after_cconv 0.07% : 0.000506s : 1: opt_after_jit_grad 0.04% : 0.000294s : 1: opt_b 21.17% : 0.158943s : 1: optimize 0.00% : 0.000025s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 1: order_py_execute_after_rewriter 0.00% : 0.000032s : 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% : 0.000006s : 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% : 0.000067s : 1: pre_auto_parallel 0.01% : 0.000050s : 1: py_interpret_to_execute 0.00% : 0.000019s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000021s : 1: remove_dup_value 18.62% : 0.139786s : 2: renormalize.infer 0.32% : 0.002392s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000063s : 1: rewriter_after_opt_a 0.02% : 0.000168s : 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% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000097s : 1: symbol_engine_optimizer 17.65% : 0.132512s : 1: task_emit 0.01% : 0.000095s : 1: tuple_transform 18.89% : 0.141792s : 1: type_inference 0.01% : 0.000079s : 1: validate TotalTime = 0.204608, [24] [bootstrap]: 0.00039986 [type_inference]: 0.0841445 [event_method]: 2.091e-05 [auto_monad]: 0.00016186 [graph_reusing]: 6.58e-06 [inline]: 2.19999e-06 [add_attr]: 0.00335877, [1] [add_attr_with_inline]: 0.00334819, [1] [Cycle 1]: 6.355e-05, [2] [tag_attr]: 2.294e-05 [meta_addattr_fg_expand]: 6.26998e-06 [parallel-infer-symbol]: 3.5e-06 [pre_auto_parallel]: 3.696e-05 [insert-virtual-dataset]: 2.66e-06 [parallel-infer-symbol-second]: 9.30013e-07 [dataset_repeat_opt]: 2.22001e-06 [pipeline_split]: 2.11e-06 [optimize]: 0.00663097, [53] [py_interpret_to_execute]: 2.604e-05 [rewriter_before_opt_a]: 8.023e-05 [opt_a]: 0.00377998, [2] [Cycle 1]: 0.0028598, [45] [expand_dump_flag]: 2.96999e-06 [switch_simplify]: 4.501e-05 [loop_unroll]: 2.953e-05 [a_1]: 0.00062443 [with_stream_mark]: 1.853e-05 [recompute_prepare]: 1.188e-05 [updatestate_depend_eliminate]: 7.15e-06 [updatestate_assign_eliminate]: 5.97999e-06 [updatestate_loads_eliminate]: 5.52001e-06 [parameter_eliminate]: 2.11e-06 [a_2]: 0.00012435 [accelerated_algorithm]: 1.029e-05 [shard]: 2.88e-06 [meta_shard_fg_expand]: 2.68e-06 [shard_inline]: 8.76002e-06 [merge_send_recv]: 1.395e-05 [auto_parallel]: 9.49999e-06 [parallel]: 1.979e-05 [flash_sp]: 9.76e-06 [merge_comm]: 5.30001e-06 [allreduce_fusion]: 5.71e-06 [matmul_add_comm_reduction]: 1.303e-05 [allreduce_slice_to_reducescatter]: 9.80013e-07 [virtual_shard_identity]: 1.17e-05 [virtual_dataset]: 8.99998e-06 [get_grad_eliminate_]: 9.25999e-06 [virtual_output]: 8.55999e-06 [merge_forward]: 6.40002e-06 [cell_reuse_recompute_pass]: 1.67001e-06 [offload_activation]: 1.432e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.936e-05 [merge_recompute_call_nodes]: 1.68002e-06 [before_grad]: 1.641e-05 [set_forward_comm_id_for_comm_node_pass]: 6.17001e-06 [meta_fg_expand]: 3.86001e-06 [flash_sp_send_recv_attached]: 2.97002e-06 [receive_attached]: 2.57001e-06 [after_resolve]: 1.301e-05 [a_after_grad]: 1.367e-05 [renormalize]: 0.00132762 [add_forward_monad_depend]: 7.12002e-06 [auto_monad_grad]: 2.66e-06 [auto_monad_eliminator]: 2.435e-05 [cse]: 4.098e-05 [a_3]: 6.713e-05 [Cycle 2]: 0.0009093, [45] [expand_dump_flag]: 2.02001e-06 [switch_simplify]: 9.89999e-06 [loop_unroll]: 8.47e-06 [a_1]: 0.0002452 [with_stream_mark]: 1.666e-05 [recompute_prepare]: 9.84001e-06 [updatestate_depend_eliminate]: 5.66998e-06 [updatestate_assign_eliminate]: 4.77998e-06 [updatestate_loads_eliminate]: 5.59e-06 [parameter_eliminate]: 1.16002e-06 [a_2]: 0.00011161 [accelerated_algorithm]: 8.95999e-06 [shard]: 1.44998e-06 [meta_shard_fg_expand]: 2.33002e-06 [shard_inline]: 8.69e-06 [merge_send_recv]: 8.35001e-06 [auto_parallel]: 7.97e-06 [parallel]: 6.19999e-06 [flash_sp]: 4.14002e-06 [merge_comm]: 5.09998e-06 [allreduce_fusion]: 4.99e-06 [matmul_add_comm_reduction]: 9.52001e-06 [allreduce_slice_to_reducescatter]: 7.09988e-07 [virtual_shard_identity]: 9.67001e-06 [virtual_dataset]: 7.98001e-06 [get_grad_eliminate_]: 8.23001e-06 [virtual_output]: 7.76001e-06 [merge_forward]: 5.43997e-06 [cell_reuse_recompute_pass]: 2.14e-06 [offload_activation]: 9.92001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.763e-05 [merge_recompute_call_nodes]: 1.10001e-06 [before_grad]: 1.335e-05 [set_forward_comm_id_for_comm_node_pass]: 5.36002e-06 [meta_fg_expand]: 3.56999e-06 [flash_sp_send_recv_attached]: 8.70001e-07 [receive_attached]: 1.97001e-06 [after_resolve]: 1.159e-05 [a_after_grad]: 1.248e-05 [renormalize]: 9.00181e-08 [add_forward_monad_depend]: 1.94e-06 [auto_monad_grad]: 1.62001e-06 [auto_monad_eliminator]: 1.379e-05 [cse]: 2.421e-05 [a_3]: 5.073e-05 [py_interpret_to_execute_after_opt_a]: 1.52e-05 [slice_cell_reuse_recomputed_activation]: 2.04e-06 [rewriter_after_opt_a]: 0.00023995 [convert_after_rewriter]: 1.263e-05 [order_py_execute_after_rewriter]: 7.23e-06 [mutable_eliminate]: 0.00063628 [opt_b]: 0.00033315, [1] [Cycle 1]: 0.00032573, [7] [b_1]: 0.00021346 [b_2]: 1.119e-05 [updatestate_depend_eliminate]: 2.188e-05 [updatestate_assign_eliminate]: 4.67e-06 [updatestate_loads_eliminate]: 4.42998e-06 [renormalize]: 6.09987e-07 [cse]: 3.025e-05 [optimize_parallel_all_gather_comm]: 5.553e-05 [overlap_param_gather]: 2.88e-06 [cconv]: 2.827e-05 [loop_unroll]: 0.00052676 [opt_after_cconv]: 0.00015201, [1] [Cycle 1]: 0.00014381, [7] [c_1]: 5.59e-05 [parameter_eliminate]: 3.9e-06 [updatestate_depend_eliminate]: 8.82e-06 [updatestate_assign_eliminate]: 4.62e-06 [updatestate_loads_eliminate]: 5.57999e-06 [cse]: 2.905e-05 [renormalize]: 6.49976e-07 [remove_dup_value]: 1.741e-05 [tuple_transform]: 9.26e-05, [1] [Cycle 1]: 8.746e-05, [4] [d_1]: 5.711e-05 [none_parameter_eliminate]: 1.89e-06 [renormalize]: 1.8999e-07 [switch_simplify]: 9.05001e-06 [partial_unused_args_eliminate]: 2.57001e-06 [add_recomputation]: 7.515e-05 [cse_after_recomputation]: 2.87e-05, [1] [Cycle 1]: 2.389e-05, [1] [cse]: 1.84e-05 [environ_conv]: 7.6e-06 [swap_dp_allreduce_reducescatter]: 7.81001e-06 [bias_add_comm_swap]: 2.79999e-06 [label_micro_interleaved_index]: 6.27001e-06 [label_fine_grained_interleaved_index]: 3.13e-06 [merge_cast_opt]: 1.49998e-06 [slice_recompute_activation]: 2.82002e-06 [micro_interleaved_order_control]: 2.33002e-06 [assign_add_opt]: 1.79998e-06 [ForceFp32Comm]: 1.19e-06 [remove_cast_before_assign_add]: 1.04998e-06 [full_micro_interleaved_order_control]: 2.27999e-06 [reorder_send_recv_between_fp_bp]: 2.66e-06 [comm_op_add_attrs]: 1.35999e-06 [add_comm_op_reuse_tag]: 1.08001e-06 [interleave_split_concat_branches]: 1.44e-06 [interleave_parallel_branches]: 1.22e-06 [overlap_opt_shard_in_pipeline]: 1.39e-06 [overlap_opt_shard_grad_in_pipeline]: 1.79e-06 [control_data_broadcast_order]: 1.989e-05 [grouped_pairwise_exchange_alltoall]: 1.74e-06 [offloading_packed_experts]: 5.84e-06 [overlap_recompute_and_grad_model_parallel]: 5.96e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.35001e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40999e-06 [overlap_recompute_comm]: 2.18002e-06 [overlap_grad_ring_attention]: 5.36998e-06 [overlap_grad_flash_sp]: 2.84e-05 [begin_end_overlap_inline]: 5.50004e-07 [split_matmul_comm_elemetwise]: 2.63e-06 [split_layernorm_comm]: 1.82001e-06 [handle_group_info]: 1.01002e-06 [symbol_engine_optimizer]: 0.00010111, [1] [Cycle 1]: 9.596e-05, [6] [build]: 5.84999e-06 [elim_shapecalc]: 1.518e-05 [elim_not_effective]: 1.829e-05 [opt_reshape]: 9.70002e-06 [fold_const_symbol]: 1.51e-05 [renormalize]: 2.09984e-07 [detach_backward]: 2.26e-06 [pipeline_parallel_scheduler]: 1.66e-06 [auto_monad_reorder]: 3.111e-05 [get_jit_bprop_graph]: 1.34e-06 [rewriter_after_jit_bprop_graph]: 5.07999e-06 [opt_after_jit_grad]: 0.00058144 [validate]: 9.071e-05 [backend_pass]: 1.25001e-06 [task_emit]: 0.108878 [execute]: 9.46e-06 Sums bootstrap : 0.000400s : 0.20% type_inference : 0.084145s : 42.05% event_method : 0.000021s : 0.01% auto_monad : 0.000162s : 0.08% graph_reusing : 0.000007s : 0.00% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000023s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000006s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000037s : 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 : 0.000026s : 0.01% optimize.rewriter_before_opt_a : 0.000080s : 0.04% optimize.opt_a.expand_dump_flag : 0.000005s : 0.00% optimize.opt_a.switch_simplify : 0.000055s : 0.03% optimize.opt_a.loop_unroll : 0.000038s : 0.02% optimize.opt_a.a_1 : 0.000870s : 0.43% optimize.opt_a.with_stream_mark : 0.000035s : 0.02% optimize.opt_a.recompute_prepare : 0.000022s : 0.01% 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 : 0.000011s : 0.01% optimize.opt_a.parameter_eliminate : 0.000003s : 0.00% optimize.opt_a.a_2 : 0.000236s : 0.12% optimize.opt_a.accelerated_algorithm : 0.000019s : 0.01% optimize.opt_a.shard : 0.000004s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.00% optimize.opt_a.shard_inline : 0.000017s : 0.01% optimize.opt_a.merge_send_recv : 0.000022s : 0.01% optimize.opt_a.auto_parallel : 0.000017s : 0.01% optimize.opt_a.parallel : 0.000026s : 0.01% optimize.opt_a.flash_sp : 0.000014s : 0.01% optimize.opt_a.merge_comm : 0.000010s : 0.01% optimize.opt_a.allreduce_fusion : 0.000011s : 0.01% 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 : 0.000021s : 0.01% optimize.opt_a.virtual_dataset : 0.000017s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000017s : 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 : 0.000024s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000037s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000030s : 0.01% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000012s : 0.01% 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 : 0.000025s : 0.01% optimize.opt_a.a_after_grad : 0.000026s : 0.01% optimize.opt_a.renormalize : 0.001328s : 0.66% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.00% optimize.opt_a.auto_monad_grad : 0.000004s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000038s : 0.02% optimize.opt_a.cse : 0.000065s : 0.03% optimize.opt_a.a_3 : 0.000118s : 0.06% optimize.py_interpret_to_execute_after_opt_a : 0.000015s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000240s : 0.12% optimize.convert_after_rewriter : 0.000013s : 0.01% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000636s : 0.32% optimize.opt_b.b_1 : 0.000213s : 0.11% optimize.opt_b.b_2 : 0.000011s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000022s : 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 : 0.000030s : 0.02% optimize.optimize_parallel_all_gather_comm : 0.000056s : 0.03% optimize.overlap_param_gather : 0.000003s : 0.00% optimize.cconv : 0.000028s : 0.01% optimize.loop_unroll : 0.000527s : 0.26% optimize.opt_after_cconv.c_1 : 0.000056s : 0.03% optimize.opt_after_cconv.parameter_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.cse : 0.000029s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000057s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000003s : 0.00% optimize.add_recomputation : 0.000075s : 0.04% optimize.cse_after_recomputation.cse : 0.000018s : 0.01% 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 : 0.000006s : 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 : 0.000002s : 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 : 0.000020s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 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 : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000006s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000015s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000015s : 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 : 0.000031s : 0.02% get_jit_bprop_graph : 0.000001s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000581s : 0.29% validate : 0.000091s : 0.05% backend_pass : 0.000001s : 0.00% task_emit : 0.108878s : 54.41% execute : 0.000009s : 0.00% Time group info: ------[substitution.] 0.000225 60 2.92% : 0.000007s : 2: substitution.depend_value_elim 1.25% : 0.000003s : 5: substitution.elim_not_effective 1.11% : 0.000002s : 5: substitution.fold_const_symbol 3.29% : 0.000007s : 6: substitution.graph_param_transform 66.30% : 0.000149s : 4: substitution.inline 2.26% : 0.000005s : 10: substitution.j_node_and_user_rematch 3.84% : 0.000009s : 10: substitution.remove_not_recompute_node 1.42% : 0.000003s : 2: substitution.replace_old_param 5.22% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator 6.32% : 0.000014s : 6: substitution.updatestate_pure_node_eliminater 6.08% : 0.000014s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.084081 2 99.13% : 0.083350s : 1: type_inference.infer 0.87% : 0.000731s : 1: type_inference.specialize ------[replace.] 0.000058 6 70.36% : 0.000041s : 4: replace.inline 29.64% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000157 6 93.31% : 0.000147s : 4: match.inline 6.69% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000267 1652 0.80% : 0.000002s : 16: predicate.accumulaten_eliminater 1.11% : 0.000003s : 6: predicate.ad_related_special_op_eliminate 0.60% : 0.000002s : 12: predicate.addn_check_dump 0.85% : 0.000002s : 16: predicate.addn_zero_filter 0.82% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.21% : 0.000006s : 28: predicate.arithmetic_simplify 0.98% : 0.000003s : 16: predicate.cast_eliminate 0.70% : 0.000002s : 12: predicate.check_bprop_eliminate 0.60% : 0.000002s : 12: predicate.compare_switch_simplify 0.20% : 0.000001s : 6: predicate.const_output_eliminate 0.65% : 0.000002s : 12: predicate.depend_value_elim 0.91% : 0.000002s : 16: predicate.dict_get_item_const_eliminator 1.01% : 0.000003s : 16: predicate.dict_get_item_eliminator 0.84% : 0.000002s : 16: predicate.dict_set_item_eliminator 1.27% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.23% : 0.000001s : 6: predicate.elim_not_effective 0.53% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.12% : 0.000003s : 22: predicate.environ_add_const_eliminate 1.06% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.12% : 0.000003s : 22: predicate.environ_get_depend_swap 1.86% : 0.000005s : 34: predicate.environ_get_eliminate 1.14% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.14% : 0.000003s : 22: predicate.exchange_switch_depend_value 2.01% : 0.000005s : 22: predicate.float_depend_g_call 0.60% : 0.000002s : 12: predicate.float_environ_get_switch 0.88% : 0.000002s : 18: predicate.float_tuple_getitem_switch 0.20% : 0.000001s : 6: predicate.fold_const_symbol 0.70% : 0.000002s : 12: predicate.get_grad_eliminate 0.22% : 0.000001s : 6: predicate.graph_param_transform 0.78% : 0.000002s : 12: predicate.incorporate_call 0.59% : 0.000002s : 12: predicate.incorporate_call_switch 5.90% : 0.000016s : 74: predicate.inline 0.81% : 0.000002s : 12: predicate.inline_without_move 0.33% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.88% : 0.000002s : 12: predicate.less_batch_normalization 1.73% : 0.000005s : 30: predicate.list_to_tuple_eliminator_ 2.66% : 0.000007s : 46: predicate.load_eliminater 1.13% : 0.000003s : 6: predicate.loop_unroll_after_grad 1.90% : 0.000005s : 38: predicate.loop_unroll_before_grad 1.72% : 0.000005s : 28: predicate.make_slice_get_slice_eliminator 0.63% : 0.000002s : 12: predicate.merge_addn 0.71% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.64% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.80% : 0.000002s : 16: predicate.minmaximum_grad 1.07% : 0.000003s : 6: predicate.mutable_eliminate 0.47% : 0.000001s : 6: predicate.opt_reshape 0.40% : 0.000001s : 6: predicate.parallel_virtual_node 1.41% : 0.000004s : 22: predicate.partial_defer_inline 1.42% : 0.000004s : 24: predicate.partial_eliminate 0.95% : 0.000003s : 16: predicate.print_const_string_wrapper 0.66% : 0.000002s : 12: predicate.reduce_all_const_elim 1.24% : 0.000003s : 16: predicate.reduce_eliminate 2.38% : 0.000006s : 46: predicate.redundant_stop_gradient_eliminater 0.54% : 0.000001s : 12: predicate.remove_not_recompute_node 1.26% : 0.000003s : 30: predicate.replace_applicator 0.52% : 0.000001s : 12: predicate.replace_old_param 0.31% : 0.000001s : 6: predicate.reset_defer_inline 0.88% : 0.000002s : 16: predicate.reshape_eliminate 0.66% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.40% : 0.000001s : 6: predicate.row_tensor_eliminate 0.84% : 0.000002s : 12: predicate.same_eliminate 0.44% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.84% : 0.000002s : 12: predicate.shard_identity_eliminate 0.94% : 0.000003s : 12: predicate.special_op_eliminate 0.82% : 0.000002s : 12: predicate.specialize_transform 1.00% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 1.01% : 0.000003s : 12: predicate.stack_unstack_eliminate 0.44% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.22% : 0.000003s : 22: predicate.switch_defer_inline 1.81% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.68% : 0.000012s : 78: predicate.switch_simplify 0.81% : 0.000002s : 16: predicate.tile_eliminate 0.87% : 0.000002s : 16: predicate.transpose_eliminate 1.59% : 0.000004s : 28: predicate.tuple_list_convert_item_index_to_positive 1.67% : 0.000004s : 28: predicate.tuple_list_get_item_const_eliminator 1.55% : 0.000004s : 28: predicate.tuple_list_get_item_depend_reorder 3.37% : 0.000009s : 42: predicate.tuple_list_get_item_eliminator 1.52% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.44% : 0.000007s : 40: predicate.tuple_list_set_item_eliminator 1.63% : 0.000004s : 30: predicate.tuple_to_list_eliminator_ 2.47% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.24% : 0.000009s : 58: predicate.updatestate_useless_node_eliminater 0.39% : 0.000001s : 6: predicate.value_based_eliminate 0.72% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.72% : 0.000002s : 12: predicate.virtual_output_eliminate 0.34% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.48% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000762 11 51.87% : 0.000395s : 5: func_graph_cloner_run.FuncGraphClonerGraph 48.13% : 0.000367s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.217735 196 0.00% : 0.000004s : 1: ForceFp32Comm 1.55% : 0.003364s : 1: add_attr 1.54% : 0.003352s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.04% : 0.000079s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.08% : 0.000169s : 1: auto_monad 0.02% : 0.000035s : 1: auto_monad_reorder 0.00% : 0.000007s : 1: backend_pass 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.02% : 0.000047s : 1: bias_add_comm_swap 0.20% : 0.000428s : 1: bootstrap 0.01% : 0.000032s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.01% : 0.000023s : 1: control_data_broadcast_order 0.01% : 0.000018s : 1: convert_after_rewriter 0.01% : 0.000032s : 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 0.01% : 0.000028s : 1: event_method 0.01% : 0.000016s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000005s : 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% : 0.000007s : 1: label_fine_grained_interleaved_index 0.01% : 0.000012s : 1: label_micro_interleaved_index 0.25% : 0.000536s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.30% : 0.000645s : 1: mutable_eliminate 0.00% : 0.000010s : 1: offloading_packed_experts 0.01% : 0.000020s : 1: opt.transform.loop_unroll_optimizer 0.01% : 0.000019s : 1: opt.transform.mutable_eliminate 0.68% : 0.001475s : 78: opt.transform.opt_a 0.02% : 0.000054s : 1: opt.transform.opt_after_cconv 0.02% : 0.000036s : 1: opt.transform.opt_after_jit_grad 0.09% : 0.000193s : 28: opt.transform.opt_b 0.03% : 0.000064s : 2: opt.transform.opt_trans_graph 0.03% : 0.000055s : 4: opt.transform.symbol_engine_opt 1.74% : 0.003783s : 1: opt_a 0.07% : 0.000156s : 1: opt_after_cconv 0.27% : 0.000592s : 1: opt_after_jit_grad 0.15% : 0.000337s : 1: opt_b 3.05% : 0.006636s : 1: optimize 0.03% : 0.000062s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000011s : 1: order_py_execute_after_rewriter 0.01% : 0.000032s : 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% : 0.000006s : 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% : 0.000007s : 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% : 0.000041s : 1: pre_auto_parallel 0.01% : 0.000030s : 1: py_interpret_to_execute 0.01% : 0.000019s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.01% : 0.000021s : 1: remove_dup_value 0.21% : 0.000452s : 1: renormalize.infer 0.40% : 0.000866s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.11% : 0.000248s : 1: rewriter_after_opt_a 0.04% : 0.000085s : 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% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.05% : 0.000104s : 1: symbol_engine_optimizer 50.01% : 0.108897s : 1: task_emit 0.04% : 0.000096s : 1: tuple_transform 38.65% : 0.084163s : 1: type_inference 0.06% : 0.000127s : 1: validate TotalTime = 0.455551, [24] [bootstrap]: 0.00058285 [type_inference]: 0.147572 [event_method]: 5.844e-05 [auto_monad]: 0.00027736 [graph_reusing]: 9.74e-06 [inline]: 2.46998e-06 [add_attr]: 0.0032818, [1] [add_attr_with_inline]: 0.00327296, [1] [Cycle 1]: 7.749e-05, [2] [tag_attr]: 3.678e-05 [meta_addattr_fg_expand]: 1.101e-05 [parallel-infer-symbol]: 3.46999e-06 [pre_auto_parallel]: 5.357e-05 [insert-virtual-dataset]: 2.37001e-06 [parallel-infer-symbol-second]: 7.09988e-07 [dataset_repeat_opt]: 2.22999e-06 [pipeline_split]: 2.01e-06 [optimize]: 0.147051, [53] [py_interpret_to_execute]: 3.991e-05 [rewriter_before_opt_a]: 0.00015358 [opt_a]: 0.144457, [3] [Cycle 1]: 0.140144, [45] [expand_dump_flag]: 3.73001e-06 [switch_simplify]: 7.842e-05 [loop_unroll]: 6.569e-05 [a_1]: 0.00164662 [with_stream_mark]: 3.004e-05 [recompute_prepare]: 2.777e-05 [updatestate_depend_eliminate]: 1.175e-05 [updatestate_assign_eliminate]: 1.101e-05 [updatestate_loads_eliminate]: 9.61e-06 [parameter_eliminate]: 3.30998e-06 [a_2]: 0.00028432 [accelerated_algorithm]: 1.894e-05 [shard]: 1.89e-06 [meta_shard_fg_expand]: 4.47e-06 [shard_inline]: 1.834e-05 [merge_send_recv]: 2.001e-05 [auto_parallel]: 1.328e-05 [parallel]: 1.909e-05 [flash_sp]: 1.034e-05 [merge_comm]: 1.149e-05 [allreduce_fusion]: 1.061e-05 [matmul_add_comm_reduction]: 3.119e-05 [allreduce_slice_to_reducescatter]: 7.7e-07 [virtual_shard_identity]: 2.008e-05 [virtual_dataset]: 1.806e-05 [get_grad_eliminate_]: 1.754e-05 [virtual_output]: 1.754e-05 [merge_forward]: 1.21e-05 [cell_reuse_recompute_pass]: 1.24e-06 [offload_activation]: 2.214e-05 [cell_reuse_handle_not_recompute_node_pass]: 0.124652 [merge_recompute_call_nodes]: 4.26001e-06 [before_grad]: 5.914e-05 [set_forward_comm_id_for_comm_node_pass]: 3.729e-05 [meta_fg_expand]: 0.00606559 [flash_sp_send_recv_attached]: 5.81e-06 [receive_attached]: 2.59001e-06 [after_resolve]: 0.00011313 [a_after_grad]: 0.00017644 [renormalize]: 0.00505498 [add_forward_monad_depend]: 1.37e-05 [auto_monad_grad]: 8.08999e-06 [auto_monad_eliminator]: 6.971e-05 [cse]: 0.00024931 [a_3]: 0.00082093 [Cycle 2]: 0.00347134, [45] [expand_dump_flag]: 3.18998e-06 [switch_simplify]: 5.447e-05 [loop_unroll]: 5.324e-05 [a_1]: 0.00155643 [with_stream_mark]: 2.205e-05 [recompute_prepare]: 1.354e-05 [updatestate_depend_eliminate]: 3.301e-05 [updatestate_assign_eliminate]: 6.04001e-06 [updatestate_loads_eliminate]: 5.29e-06 [parameter_eliminate]: 1.97001e-06 [a_2]: 0.00011775 [accelerated_algorithm]: 9.87001e-06 [shard]: 1.99e-06 [meta_shard_fg_expand]: 2.61e-06 [shard_inline]: 8.37e-06 [merge_send_recv]: 9.56e-06 [auto_parallel]: 1.199e-05 [parallel]: 1.086e-05 [flash_sp]: 4.47e-06 [merge_comm]: 5.07e-06 [allreduce_fusion]: 5.37001e-06 [matmul_add_comm_reduction]: 1.235e-05 [allreduce_slice_to_reducescatter]: 1.05999e-06 [virtual_shard_identity]: 1.145e-05 [virtual_dataset]: 9.49e-06 [get_grad_eliminate_]: 8.55001e-06 [virtual_output]: 7.94997e-06 [merge_forward]: 5.35999e-06 [cell_reuse_recompute_pass]: 1.34e-06 [offload_activation]: 1.277e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.818e-05 [merge_recompute_call_nodes]: 1.17999e-06 [before_grad]: 1.554e-05 [set_forward_comm_id_for_comm_node_pass]: 6.17999e-06 [meta_fg_expand]: 0.00010339 [flash_sp_send_recv_attached]: 1.52001e-06 [receive_attached]: 1.70001e-06 [after_resolve]: 1.503e-05 [a_after_grad]: 1.338e-05 [renormalize]: 0.00087636 [add_forward_monad_depend]: 5.49e-06 [auto_monad_grad]: 1.96998e-06 [auto_monad_eliminator]: 1.882e-05 [cse]: 2.834e-05 [a_3]: 6.165e-05 [Cycle 3]: 0.00082298, [45] [expand_dump_flag]: 1.27e-06 [switch_simplify]: 9.77001e-06 [loop_unroll]: 8.35999e-06 [a_1]: 0.00020431 [with_stream_mark]: 1.132e-05 [recompute_prepare]: 8.52998e-06 [updatestate_depend_eliminate]: 5.46e-06 [updatestate_assign_eliminate]: 4.51002e-06 [updatestate_loads_eliminate]: 4.1e-06 [parameter_eliminate]: 1.02e-06 [a_2]: 0.00011233 [accelerated_algorithm]: 8.53001e-06 [shard]: 1.28002e-06 [meta_shard_fg_expand]: 2.11e-06 [shard_inline]: 8.42e-06 [merge_send_recv]: 6.58e-06 [auto_parallel]: 7.51001e-06 [parallel]: 5.02e-06 [flash_sp]: 1.02998e-06 [merge_comm]: 5.00999e-06 [allreduce_fusion]: 4.72e-06 [matmul_add_comm_reduction]: 8.3e-06 [allreduce_slice_to_reducescatter]: 4.90021e-07 [virtual_shard_identity]: 8.88002e-06 [virtual_dataset]: 7.87e-06 [get_grad_eliminate_]: 7.48e-06 [virtual_output]: 7.9e-06 [merge_forward]: 4.31002e-06 [cell_reuse_recompute_pass]: 1.24e-06 [offload_activation]: 9.35001e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.528e-05 [merge_recompute_call_nodes]: 1.14998e-06 [before_grad]: 1.362e-05 [set_forward_comm_id_for_comm_node_pass]: 5.46e-06 [meta_fg_expand]: 2.94999e-06 [flash_sp_send_recv_attached]: 8.39995e-07 [receive_attached]: 9.29984e-07 [after_resolve]: 1.079e-05 [a_after_grad]: 1.232e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 1.07e-06 [auto_monad_grad]: 9.50007e-07 [auto_monad_eliminator]: 1.154e-05 [cse]: 2.136e-05 [a_3]: 5.218e-05 [py_interpret_to_execute_after_opt_a]: 3.218e-05 [slice_cell_reuse_recomputed_activation]: 2.71999e-06 [rewriter_after_opt_a]: 6.284e-05 [convert_after_rewriter]: 9.78002e-06 [order_py_execute_after_rewriter]: 6.36998e-06 [mutable_eliminate]: 0.00065471 [opt_b]: 0.00030776, [1] [Cycle 1]: 0.00030035, [7] [b_1]: 0.00018603 [b_2]: 1.091e-05 [updatestate_depend_eliminate]: 7.80998e-06 [updatestate_assign_eliminate]: 4.40999e-06 [updatestate_loads_eliminate]: 4e-06 [renormalize]: 5.3001e-07 [cse]: 4.837e-05 [optimize_parallel_all_gather_comm]: 2.137e-05 [overlap_param_gather]: 1.97999e-06 [cconv]: 2.408e-05 [loop_unroll]: 0.00044693 [opt_after_cconv]: 0.00013771, [1] [Cycle 1]: 0.00013153, [7] [c_1]: 5.284e-05 [parameter_eliminate]: 2.32001e-06 [updatestate_depend_eliminate]: 7.55e-06 [updatestate_assign_eliminate]: 4.62e-06 [updatestate_loads_eliminate]: 4.2e-06 [cse]: 2.523e-05 [renormalize]: 4.19997e-07 [remove_dup_value]: 1.726e-05 [tuple_transform]: 8.91e-05, [1] [Cycle 1]: 8.441e-05, [4] [d_1]: 5.557e-05 [none_parameter_eliminate]: 1.72001e-06 [renormalize]: 2.50002e-07 [switch_simplify]: 8.92e-06 [partial_unused_args_eliminate]: 2.01003e-06 [add_recomputation]: 6.522e-05 [cse_after_recomputation]: 4.354e-05, [1] [Cycle 1]: 3.882e-05, [1] [cse]: 3.298e-05 [environ_conv]: 6.76e-06 [swap_dp_allreduce_reducescatter]: 8.62e-06 [bias_add_comm_swap]: 2.57001e-06 [label_micro_interleaved_index]: 4.38001e-06 [label_fine_grained_interleaved_index]: 2.81e-06 [merge_cast_opt]: 1.49998e-06 [slice_recompute_activation]: 2.36e-06 [micro_interleaved_order_control]: 2.44001e-06 [assign_add_opt]: 1.30999e-06 [ForceFp32Comm]: 9.00007e-07 [remove_cast_before_assign_add]: 1.47999e-06 [full_micro_interleaved_order_control]: 2.44001e-06 [reorder_send_recv_between_fp_bp]: 2.78e-06 [comm_op_add_attrs]: 1.04998e-06 [add_comm_op_reuse_tag]: 1.00001e-06 [interleave_split_concat_branches]: 1.35999e-06 [interleave_parallel_branches]: 1.05999e-06 [overlap_opt_shard_in_pipeline]: 1.24e-06 [overlap_opt_shard_grad_in_pipeline]: 1.87999e-06 [control_data_broadcast_order]: 4.036e-05 [grouped_pairwise_exchange_alltoall]: 1.77999e-06 [offloading_packed_experts]: 6.11e-06 [overlap_recompute_and_grad_model_parallel]: 6.51999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.14998e-06 [overlap_recompute_allgather_and_fa_grad]: 1.35001e-06 [overlap_recompute_comm]: 2.22999e-06 [overlap_grad_ring_attention]: 5.78002e-06 [overlap_grad_flash_sp]: 2.868e-05 [begin_end_overlap_inline]: 5.3001e-07 [split_matmul_comm_elemetwise]: 2.44999e-06 [split_layernorm_comm]: 1.70001e-06 [handle_group_info]: 9.60019e-07 [symbol_engine_optimizer]: 9.688e-05, [1] [Cycle 1]: 9.164e-05, [6] [build]: 4.38001e-06 [elim_shapecalc]: 1.399e-05 [elim_not_effective]: 1.778e-05 [opt_reshape]: 1.014e-05 [fold_const_symbol]: 1.543e-05 [renormalize]: 1.8999e-07 [detach_backward]: 2.69999e-06 [pipeline_parallel_scheduler]: 1.54e-06 [auto_monad_reorder]: 5.93e-05 [get_jit_bprop_graph]: 1.65001e-06 [rewriter_after_jit_bprop_graph]: 4e-06 [opt_after_jit_grad]: 0.00048095 [validate]: 4.662e-05 [backend_pass]: 8.70001e-07 [task_emit]: 0.1558 [execute]: 9.67001e-06 Sums bootstrap : 0.000583s : 0.13% type_inference : 0.147572s : 32.73% event_method : 0.000058s : 0.01% auto_monad : 0.000277s : 0.06% graph_reusing : 0.000010s : 0.00% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000037s : 0.01% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000011s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000054s : 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 : 0.000040s : 0.01% optimize.rewriter_before_opt_a : 0.000154s : 0.03% optimize.opt_a.expand_dump_flag : 0.000008s : 0.00% optimize.opt_a.switch_simplify : 0.000143s : 0.03% optimize.opt_a.loop_unroll : 0.000127s : 0.03% optimize.opt_a.a_1 : 0.003407s : 0.76% optimize.opt_a.with_stream_mark : 0.000063s : 0.01% optimize.opt_a.recompute_prepare : 0.000050s : 0.01% optimize.opt_a.updatestate_depend_eliminate : 0.000050s : 0.01% optimize.opt_a.updatestate_assign_eliminate : 0.000022s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000019s : 0.00% optimize.opt_a.parameter_eliminate : 0.000006s : 0.00% optimize.opt_a.a_2 : 0.000514s : 0.11% optimize.opt_a.accelerated_algorithm : 0.000037s : 0.01% optimize.opt_a.shard : 0.000005s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000009s : 0.00% optimize.opt_a.shard_inline : 0.000035s : 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 : 0.000035s : 0.01% optimize.opt_a.flash_sp : 0.000016s : 0.00% optimize.opt_a.merge_comm : 0.000022s : 0.00% optimize.opt_a.allreduce_fusion : 0.000021s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000052s : 0.01% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000040s : 0.01% optimize.opt_a.virtual_dataset : 0.000035s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000034s : 0.01% optimize.opt_a.virtual_output : 0.000033s : 0.01% optimize.opt_a.merge_forward : 0.000022s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000004s : 0.00% optimize.opt_a.offload_activation : 0.000044s : 0.01% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.124706s : 27.66% optimize.opt_a.merge_recompute_call_nodes : 0.000007s : 0.00% optimize.opt_a.before_grad : 0.000088s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000049s : 0.01% optimize.opt_a.meta_fg_expand : 0.006172s : 1.37% 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 : 0.000139s : 0.03% optimize.opt_a.a_after_grad : 0.000202s : 0.04% optimize.opt_a.renormalize : 0.005931s : 1.32% optimize.opt_a.add_forward_monad_depend : 0.000020s : 0.00% optimize.opt_a.auto_monad_grad : 0.000011s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000100s : 0.02% optimize.opt_a.cse : 0.000299s : 0.07% optimize.opt_a.a_3 : 0.000935s : 0.21% optimize.py_interpret_to_execute_after_opt_a : 0.000032s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000003s : 0.00% optimize.rewriter_after_opt_a : 0.000063s : 0.01% optimize.convert_after_rewriter : 0.000010s : 0.00% optimize.order_py_execute_after_rewriter : 0.000006s : 0.00% optimize.mutable_eliminate : 0.000655s : 0.15% optimize.opt_b.b_1 : 0.000186s : 0.04% optimize.opt_b.b_2 : 0.000011s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 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 : 0.000048s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000024s : 0.01% optimize.loop_unroll : 0.000447s : 0.10% optimize.opt_after_cconv.c_1 : 0.000053s : 0.01% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.00% optimize.opt_after_cconv.cse : 0.000025s : 0.01% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.00% optimize.tuple_transform.d_1 : 0.000056s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000065s : 0.01% optimize.cse_after_recomputation.cse : 0.000033s : 0.01% optimize.environ_conv : 0.000007s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 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 : 0.000001s : 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 : 0.000040s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000002s : 0.00% optimize.overlap_grad_ring_attention : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000029s : 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 : 0.000014s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000018s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 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 : 0.000059s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000004s : 0.00% opt_after_jit_grad : 0.000481s : 0.11% validate : 0.000047s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.155800s : 34.56% execute : 0.000010s : 0.00% Time group info: ------[substitution.] 0.001584 227 0.72% : 0.000011s : 4: substitution.depend_value_elim 0.18% : 0.000003s : 5: substitution.elim_not_effective 0.17% : 0.000003s : 3: substitution.float_depend_g_call 0.26% : 0.000004s : 2: substitution.float_tuple_getitem_switch 0.16% : 0.000002s : 5: substitution.fold_const_symbol 45.60% : 0.000722s : 4: substitution.getattr_setattr_resolve 0.45% : 0.000007s : 6: substitution.graph_param_transform 0.24% : 0.000004s : 2: substitution.incorporate_call 0.16% : 0.000002s : 2: substitution.incorporate_call_switch 30.59% : 0.000485s : 15: substitution.inline 2.26% : 0.000036s : 5: substitution.inline_without_move 1.34% : 0.000021s : 23: substitution.j_node_and_user_rematch 0.68% : 0.000011s : 6: substitution.minmaximum_grad 0.22% : 0.000004s : 3: substitution.partial_eliminate 1.74% : 0.000028s : 23: substitution.remove_not_recompute_node 3.27% : 0.000052s : 21: substitution.replace_applicator 0.87% : 0.000014s : 16: substitution.replace_old_param 0.20% : 0.000003s : 1: substitution.set_cell_output_no_recompute 1.32% : 0.000021s : 6: substitution.tuple_list_convert_item_index_to_positive 0.63% : 0.000010s : 6: substitution.tuple_list_get_item_const_eliminator 0.89% : 0.000014s : 6: substitution.tuple_list_get_item_depend_reorder 3.36% : 0.000053s : 16: substitution.tuple_list_get_item_eliminator 0.86% : 0.000014s : 6: substitution.tuple_list_get_set_item_eliminator 1.49% : 0.000024s : 18: substitution.updatestate_pure_node_eliminater 2.37% : 0.000037s : 23: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.147473 2 98.80% : 0.145704s : 1: type_inference.infer 1.20% : 0.001769s : 1: type_inference.specialize ------[replace.] 0.000321 28 16.81% : 0.000054s : 3: replace.getattr_setattr_resolve 38.31% : 0.000123s : 15: replace.inline 10.44% : 0.000033s : 1: replace.replace_applicator 27.34% : 0.000088s : 8: replace.tuple_list_get_item_eliminator 7.11% : 0.000023s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001200 28 56.67% : 0.000680s : 3: match.getattr_setattr_resolve 39.74% : 0.000477s : 15: match.inline 0.76% : 0.000009s : 1: match.replace_applicator 2.04% : 0.000025s : 8: match.tuple_list_get_item_eliminator 0.79% : 0.000009s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.000965 5572 0.80% : 0.000008s : 57: predicate.accumulaten_eliminater 0.25% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.40% : 0.000004s : 29: predicate.addn_check_dump 0.81% : 0.000008s : 57: predicate.addn_zero_filter 0.75% : 0.000007s : 57: predicate.adjust_all_reduce_mul_add 1.67% : 0.000016s : 86: predicate.arithmetic_simplify 0.88% : 0.000008s : 57: predicate.cast_eliminate 1.70% : 0.000016s : 130: predicate.check_bprop_eliminate 0.43% : 0.000004s : 29: predicate.compare_switch_simplify 0.05% : 0.000001s : 6: predicate.const_output_eliminate 0.44% : 0.000004s : 29: predicate.depend_value_elim 0.84% : 0.000008s : 57: predicate.dict_get_item_const_eliminator 0.97% : 0.000009s : 57: predicate.dict_get_item_eliminator 0.86% : 0.000008s : 57: predicate.dict_set_item_eliminator 0.26% : 0.000002s : 12: predicate.dumpgradient_eliminate 0.06% : 0.000001s : 6: predicate.elim_not_effective 0.14% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 0.87% : 0.000008s : 63: predicate.environ_add_const_eliminate 0.86% : 0.000008s : 63: predicate.environ_get_add_eliminate 0.85% : 0.000008s : 63: predicate.environ_get_depend_swap 1.35% : 0.000013s : 92: predicate.environ_get_eliminate 0.90% : 0.000009s : 63: predicate.environ_get_set_eliminate 1.18% : 0.000011s : 81: predicate.exchange_switch_depend_value 1.74% : 0.000017s : 81: predicate.float_depend_g_call 0.43% : 0.000004s : 29: predicate.float_environ_get_switch 0.58% : 0.000006s : 35: predicate.float_tuple_getitem_switch 0.05% : 0.000001s : 6: predicate.fold_const_symbol 0.46% : 0.000004s : 29: predicate.get_grad_eliminate 0.59% : 0.000006s : 20: predicate.getattr_setattr_resolve 0.06% : 0.000001s : 6: predicate.graph_param_transform 0.43% : 0.000004s : 29: predicate.incorporate_call 0.38% : 0.000004s : 29: predicate.incorporate_call_switch 4.25% : 0.000041s : 208: predicate.inline 1.81% : 0.000017s : 97: predicate.inline_without_move 0.23% : 0.000002s : 29: predicate.j_node_and_user_rematch 0.49% : 0.000005s : 29: predicate.less_batch_normalization 1.13% : 0.000011s : 77: predicate.list_to_tuple_eliminator_ 1.87% : 0.000018s : 134: predicate.load_eliminater 0.28% : 0.000003s : 6: predicate.loop_unroll_after_grad 1.95% : 0.000019s : 142: predicate.loop_unroll_before_grad 1.01% : 0.000010s : 69: predicate.make_slice_get_slice_eliminator 0.41% : 0.000004s : 29: predicate.merge_addn 1.63% : 0.000016s : 126: predicate.micro_step_allgather_replace 1.70% : 0.000016s : 126: predicate.mini_step_allgather_replace 0.80% : 0.000008s : 57: predicate.minmaximum_grad 0.30% : 0.000003s : 6: predicate.mutable_eliminate 0.11% : 0.000001s : 6: predicate.opt_reshape 0.12% : 0.000001s : 6: predicate.parallel_virtual_node 1.44% : 0.000014s : 81: predicate.partial_defer_inline 1.21% : 0.000012s : 71: predicate.partial_eliminate 0.79% : 0.000008s : 57: predicate.print_const_string_wrapper 0.49% : 0.000005s : 29: predicate.reduce_all_const_elim 1.03% : 0.000010s : 57: predicate.reduce_eliminate 1.93% : 0.000019s : 134: predicate.redundant_stop_gradient_eliminater 0.51% : 0.000005s : 29: predicate.remove_not_recompute_node 1.86% : 0.000018s : 193: predicate.replace_applicator 0.81% : 0.000008s : 97: predicate.replace_old_param 0.06% : 0.000001s : 6: predicate.reset_defer_inline 0.84% : 0.000008s : 57: predicate.reshape_eliminate 1.71% : 0.000017s : 126: predicate.row_tensor_add_zeros_like 0.11% : 0.000001s : 6: predicate.row_tensor_eliminate 1.83% : 0.000018s : 130: predicate.same_eliminate 0.28% : 0.000003s : 31: predicate.set_cell_output_no_recompute 0.48% : 0.000005s : 29: predicate.shard_identity_eliminate 0.22% : 0.000002s : 12: predicate.special_op_eliminate 0.50% : 0.000005s : 29: predicate.specialize_transform 1.81% : 0.000017s : 126: predicate.split_environ_get_set_with_tuple_value 1.71% : 0.000016s : 97: predicate.stack_unstack_eliminate 0.11% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.31% : 0.000013s : 81: predicate.switch_defer_inline 2.93% : 0.000028s : 211: predicate.switch_layer_defer_inline 3.93% : 0.000038s : 258: predicate.switch_simplify 0.80% : 0.000008s : 57: predicate.tile_eliminate 0.80% : 0.000008s : 57: predicate.transpose_eliminate 1.10% : 0.000011s : 69: predicate.tuple_list_convert_item_index_to_positive 1.14% : 0.000011s : 69: predicate.tuple_list_get_item_const_eliminator 1.02% : 0.000010s : 69: predicate.tuple_list_get_item_depend_reorder 2.01% : 0.000019s : 106: predicate.tuple_list_get_item_eliminator 1.06% : 0.000010s : 69: predicate.tuple_list_get_set_item_eliminator 1.57% : 0.000015s : 98: predicate.tuple_list_set_item_eliminator 1.08% : 0.000010s : 77: predicate.tuple_to_list_eliminator_ 2.03% : 0.000020s : 134: predicate.updatestate_pure_node_eliminater 18.14% : 0.000175s : 164: predicate.updatestate_useless_node_eliminater 0.15% : 0.000001s : 6: predicate.value_based_eliminate 0.45% : 0.000004s : 29: predicate.virtual_dataset_eliminate 0.43% : 0.000004s : 29: predicate.virtual_output_eliminate 0.09% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.13% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.002824 52 60.59% : 0.001711s : 26: func_graph_cloner_run.FuncGraphClonerGraph 39.41% : 0.001113s : 26: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.743302 247 0.00% : 0.000004s : 1: ForceFp32Comm 0.44% : 0.003287s : 1: add_attr 0.44% : 0.003277s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000069s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000288s : 1: auto_monad 0.01% : 0.000064s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.08% : 0.000616s : 1: bootstrap 0.00% : 0.000027s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.01% : 0.000046s : 1: control_data_broadcast_order 0.00% : 0.000013s : 1: convert_after_rewriter 0.01% : 0.000047s : 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% : 0.000067s : 1: event_method 0.00% : 0.000017s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000005s : 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% : 0.000007s : 1: label_micro_interleaved_index 0.06% : 0.000455s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.09% : 0.000664s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000018s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 17.54% : 0.130370s : 125: opt.transform.opt_a 0.01% : 0.000051s : 1: opt.transform.opt_after_cconv 0.00% : 0.000031s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000169s : 28: opt.transform.opt_b 0.11% : 0.000827s : 2: opt.transform.opt_resolve 0.01% : 0.000062s : 2: opt.transform.opt_trans_graph 0.01% : 0.000054s : 4: opt.transform.symbol_engine_opt 19.43% : 0.144460s : 1: opt_a 0.02% : 0.000141s : 1: opt_after_cconv 0.07% : 0.000491s : 1: opt_after_jit_grad 0.04% : 0.000311s : 1: opt_b 19.78% : 0.147056s : 1: optimize 0.00% : 0.000025s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000009s : 1: order_py_execute_after_rewriter 0.00% : 0.000032s : 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% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.01% : 0.000058s : 1: pre_auto_parallel 0.01% : 0.000044s : 1: py_interpret_to_execute 0.01% : 0.000038s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000021s : 1: remove_dup_value 0.49% : 0.003641s : 2: renormalize.infer 0.31% : 0.002273s : 2: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000067s : 1: rewriter_after_opt_a 0.02% : 0.000159s : 1: rewriter_before_opt_a 0.00% : 0.000006s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000005s : 1: slice_recompute_activation 0.00% : 0.000004s : 1: split_layernorm_comm 0.00% : 0.000005s : 1: split_matmul_comm_elemetwise 0.00% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.01% : 0.000100s : 1: symbol_engine_optimizer 20.96% : 0.155821s : 1: task_emit 0.01% : 0.000092s : 1: tuple_transform 19.86% : 0.147592s : 1: type_inference 0.01% : 0.000074s : 1: validate [WARNING] ME(169920:281473205423920,MainProcess):2026-01-29-17:43:04.763.570 [mindspore/common/tensor.py:142] For 'Tensor', if 'dtype' is not None, 'input_data', 'shape' or 'init' must not be None. TotalTime = 15.3628, [24] [bootstrap]: 0.00047766 [type_inference]: 13.3801 [event_method]: 8.223e-05 [auto_monad]: 0.00033841 [graph_reusing]: 1.776e-05 [inline]: 5.94999e-06 [add_attr]: 0.0038908, [1] [add_attr_with_inline]: 0.00388032, [1] [Cycle 1]: 0.00013328, [2] [tag_attr]: 7.27e-05 [meta_addattr_fg_expand]: 1.861e-05 [parallel-infer-symbol]: 3.84002e-06 [pre_auto_parallel]: 9.56e-05 [insert-virtual-dataset]: 2.73e-06 [parallel-infer-symbol-second]: 1.34998e-06 [dataset_repeat_opt]: 2.11e-06 [pipeline_split]: 1.92999e-06 [optimize]: 0.805148, [53] [py_interpret_to_execute]: 8.086e-05 [rewriter_before_opt_a]: 0.00033722 [opt_a]: 0.801072, [2] [Cycle 1]: 0.798536, [45] [expand_dump_flag]: 8.08999e-06 [switch_simplify]: 0.00028525 [loop_unroll]: 0.00011487 [a_1]: 0.791749 [with_stream_mark]: 3.962e-05 [recompute_prepare]: 2.027e-05 [updatestate_depend_eliminate]: 8.99e-06 [updatestate_assign_eliminate]: 6.31e-06 [updatestate_loads_eliminate]: 5.85002e-06 [parameter_eliminate]: 2.69001e-06 [a_2]: 0.00026919 [accelerated_algorithm]: 9.99001e-06 [shard]: 2.74999e-06 [meta_shard_fg_expand]: 8.54e-06 [shard_inline]: 8.28001e-06 [merge_send_recv]: 1.2e-05 [auto_parallel]: 1.279e-05 [parallel]: 3.022e-05 [flash_sp]: 1.277e-05 [merge_comm]: 5.61003e-06 [allreduce_fusion]: 5.20999e-06 [matmul_add_comm_reduction]: 1.273e-05 [allreduce_slice_to_reducescatter]: 9.49978e-07 [virtual_shard_identity]: 1.072e-05 [virtual_dataset]: 9.56e-06 [get_grad_eliminate_]: 8.28999e-06 [virtual_output]: 8.84e-06 [merge_forward]: 5.77999e-06 [cell_reuse_recompute_pass]: 1.44e-06 [offload_activation]: 1.415e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.045e-05 [merge_recompute_call_nodes]: 1.74e-06 [before_grad]: 1.478e-05 [set_forward_comm_id_for_comm_node_pass]: 5.40999e-06 [meta_fg_expand]: 6.33002e-06 [flash_sp_send_recv_attached]: 2.53e-06 [receive_attached]: 2.41e-06 [after_resolve]: 1.446e-05 [a_after_grad]: 1.48e-05 [renormalize]: 0.00527741 [add_forward_monad_depend]: 8.77999e-06 [auto_monad_grad]: 3.25e-06 [auto_monad_eliminator]: 3.081e-05 [cse]: 4.351e-05 [a_3]: 7.741e-05 [Cycle 2]: 0.00251866, [45] [expand_dump_flag]: 2.34001e-06 [switch_simplify]: 1.111e-05 [loop_unroll]: 9.14998e-06 [a_1]: 0.00021485 [with_stream_mark]: 1.806e-05 [recompute_prepare]: 8.42e-06 [updatestate_depend_eliminate]: 5.61e-06 [updatestate_assign_eliminate]: 5.53997e-06 [updatestate_loads_eliminate]: 5.24998e-06 [parameter_eliminate]: 1.34e-06 [a_2]: 0.00011894 [accelerated_algorithm]: 8.74e-06 [shard]: 1.87001e-06 [meta_shard_fg_expand]: 2.51998e-06 [shard_inline]: 8.35001e-06 [merge_send_recv]: 8.33999e-06 [auto_parallel]: 8.89998e-06 [parallel]: 6.26998e-06 [flash_sp]: 3.79002e-06 [merge_comm]: 5.20001e-06 [allreduce_fusion]: 4.94e-06 [matmul_add_comm_reduction]: 9.31002e-06 [allreduce_slice_to_reducescatter]: 4.30009e-07 [virtual_shard_identity]: 9.15999e-06 [virtual_dataset]: 8.28999e-06 [get_grad_eliminate_]: 8.10999e-06 [virtual_output]: 8.54e-06 [merge_forward]: 5.42999e-06 [cell_reuse_recompute_pass]: 6.06998e-06 [offload_activation]: 2.425e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.183e-05 [merge_recompute_call_nodes]: 1.84998e-06 [before_grad]: 1.68e-05 [set_forward_comm_id_for_comm_node_pass]: 8.87999e-06 [meta_fg_expand]: 4.33999e-06 [flash_sp_send_recv_attached]: 1.30999e-06 [receive_attached]: 2.26998e-06 [after_resolve]: 1.412e-05 [a_after_grad]: 1.353e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 3.76001e-06 [auto_monad_grad]: 4.656e-05 [auto_monad_eliminator]: 2.235e-05 [cse]: 3.149e-05 [a_3]: 5.798e-05 [py_interpret_to_execute_after_opt_a]: 1.935e-05 [slice_cell_reuse_recomputed_activation]: 2.06e-06 [rewriter_after_opt_a]: 0.00056809 [convert_after_rewriter]: 1.746e-05 [order_py_execute_after_rewriter]: 8.3e-06 [mutable_eliminate]: 0.00071873 [opt_b]: 0.00032114, [1] [Cycle 1]: 0.00031291, [7] [b_1]: 0.00019549 [b_2]: 1.019e-05 [updatestate_depend_eliminate]: 7.98999e-06 [updatestate_assign_eliminate]: 4.85001e-06 [updatestate_loads_eliminate]: 4.56002e-06 [renormalize]: 7.30011e-07 [cse]: 5.008e-05 [optimize_parallel_all_gather_comm]: 2.346e-05 [overlap_param_gather]: 4.80001e-06 [cconv]: 3.149e-05 [loop_unroll]: 0.00046923 [opt_after_cconv]: 0.00013748, [1] [Cycle 1]: 0.00013105, [7] [c_1]: 5.535e-05 [parameter_eliminate]: 2.46998e-06 [updatestate_depend_eliminate]: 7.28e-06 [updatestate_assign_eliminate]: 4.28001e-06 [updatestate_loads_eliminate]: 4.07e-06 [cse]: 2.29e-05 [renormalize]: 4.59986e-07 [remove_dup_value]: 1.51e-05 [tuple_transform]: 9.097e-05, [1] [Cycle 1]: 8.642e-05, [4] [d_1]: 5.78e-05 [none_parameter_eliminate]: 1.62001e-06 [renormalize]: 2.09984e-07 [switch_simplify]: 8.65999e-06 [partial_unused_args_eliminate]: 2.43e-06 [add_recomputation]: 6.952e-05 [cse_after_recomputation]: 2.68e-05, [1] [Cycle 1]: 2.234e-05, [1] [cse]: 1.713e-05 [environ_conv]: 7.16001e-06 [swap_dp_allreduce_reducescatter]: 7.6e-06 [bias_add_comm_swap]: 2.50002e-06 [label_micro_interleaved_index]: 4.60999e-06 [label_fine_grained_interleaved_index]: 2.75002e-06 [merge_cast_opt]: 1.46002e-06 [slice_recompute_activation]: 2.52001e-06 [micro_interleaved_order_control]: 2.64001e-06 [assign_add_opt]: 1.54e-06 [ForceFp32Comm]: 9.49978e-07 [remove_cast_before_assign_add]: 1.18001e-06 [full_micro_interleaved_order_control]: 2.42001e-06 [reorder_send_recv_between_fp_bp]: 3.14999e-06 [comm_op_add_attrs]: 1.39998e-06 [add_comm_op_reuse_tag]: 1.34e-06 [interleave_split_concat_branches]: 1.65001e-06 [interleave_parallel_branches]: 1.28002e-06 [overlap_opt_shard_in_pipeline]: 3.95e-06 [overlap_opt_shard_grad_in_pipeline]: 1.89999e-06 [control_data_broadcast_order]: 1.881e-05 [grouped_pairwise_exchange_alltoall]: 1.81998e-06 [offloading_packed_experts]: 5.12999e-06 [overlap_recompute_and_grad_model_parallel]: 6.14001e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.27e-06 [overlap_recompute_allgather_and_fa_grad]: 1.45001e-06 [overlap_recompute_comm]: 2.72001e-06 [overlap_grad_ring_attention]: 0.00023784 [overlap_grad_flash_sp]: 3.124e-05 [begin_end_overlap_inline]: 6.39993e-07 [split_matmul_comm_elemetwise]: 2.89999e-06 [split_layernorm_comm]: 2.01998e-06 [handle_group_info]: 1.19e-06 [symbol_engine_optimizer]: 0.00044967, [1] [Cycle 1]: 0.0004441, [6] [build]: 0.0003039 [elim_shapecalc]: 1.851e-05 [elim_not_effective]: 4.822e-05 [opt_reshape]: 1.016e-05 [fold_const_symbol]: 2.435e-05 [renormalize]: 3.00002e-07 [detach_backward]: 2.65002e-06 [pipeline_parallel_scheduler]: 1.73002e-06 [auto_monad_reorder]: 3.072e-05 [get_jit_bprop_graph]: 1.47001e-06 [rewriter_after_jit_bprop_graph]: 4.57998e-06 [opt_after_jit_grad]: 0.0005642 [validate]: 5.154e-05 [backend_pass]: 1.33002e-06 [task_emit]: 1.17158 [execute]: 1.169e-05 Sums bootstrap : 0.000478s : 0.00% type_inference : 13.380123s : 87.13% event_method : 0.000082s : 0.00% auto_monad : 0.000338s : 0.00% graph_reusing : 0.000018s : 0.00% inline : 0.000006s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000073s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000019s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000096s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000081s : 0.00% optimize.rewriter_before_opt_a : 0.000337s : 0.00% optimize.opt_a.expand_dump_flag : 0.000010s : 0.00% optimize.opt_a.switch_simplify : 0.000296s : 0.00% optimize.opt_a.loop_unroll : 0.000124s : 0.00% optimize.opt_a.a_1 : 0.791964s : 5.16% optimize.opt_a.with_stream_mark : 0.000058s : 0.00% optimize.opt_a.recompute_prepare : 0.000029s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000015s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000012s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000011s : 0.00% optimize.opt_a.parameter_eliminate : 0.000004s : 0.00% optimize.opt_a.a_2 : 0.000388s : 0.00% 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 : 0.000020s : 0.00% optimize.opt_a.auto_parallel : 0.000022s : 0.00% optimize.opt_a.parallel : 0.000036s : 0.00% optimize.opt_a.flash_sp : 0.000017s : 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 : 0.000022s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000020s : 0.00% optimize.opt_a.virtual_dataset : 0.000018s : 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 : 0.000011s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000008s : 0.00% optimize.opt_a.offload_activation : 0.000038s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000052s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000004s : 0.00% optimize.opt_a.before_grad : 0.000032s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000014s : 0.00% optimize.opt_a.meta_fg_expand : 0.000011s : 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 : 0.000029s : 0.00% optimize.opt_a.a_after_grad : 0.000028s : 0.00% optimize.opt_a.renormalize : 0.005277s : 0.03% optimize.opt_a.add_forward_monad_depend : 0.000013s : 0.00% optimize.opt_a.auto_monad_grad : 0.000050s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000053s : 0.00% optimize.opt_a.cse : 0.000075s : 0.00% optimize.opt_a.a_3 : 0.000135s : 0.00% optimize.py_interpret_to_execute_after_opt_a : 0.000019s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000568s : 0.00% optimize.convert_after_rewriter : 0.000017s : 0.00% optimize.order_py_execute_after_rewriter : 0.000008s : 0.00% optimize.mutable_eliminate : 0.000719s : 0.00% optimize.opt_b.b_1 : 0.000195s : 0.00% optimize.opt_b.b_2 : 0.000010s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000008s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.000050s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000023s : 0.00% optimize.overlap_param_gather : 0.000005s : 0.00% optimize.cconv : 0.000031s : 0.00% optimize.loop_unroll : 0.000469s : 0.00% optimize.opt_after_cconv.c_1 : 0.000055s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000002s : 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 : 0.000023s : 0.00% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000015s : 0.00% optimize.tuple_transform.d_1 : 0.000058s : 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 : 0.000009s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000070s : 0.00% optimize.cse_after_recomputation.cse : 0.000017s : 0.00% 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 : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000001s : 0.00% optimize.slice_recompute_activation : 0.000003s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000002s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000004s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000019s : 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 : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000238s : 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 : 0.000003s : 0.00% optimize.split_layernorm_comm : 0.000002s : 0.00% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000304s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000019s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000048s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000024s : 0.00% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000003s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000031s : 0.00% get_jit_bprop_graph : 0.000001s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000564s : 0.00% validate : 0.000052s : 0.00% backend_pass : 0.000001s : 0.00% task_emit : 1.171584s : 7.63% execute : 0.000012s : 0.00% Time group info: ------[substitution.] 0.001130 198 1.13% : 0.000013s : 5: substitution.depend_value_elim 2.69% : 0.000030s : 5: substitution.elim_not_effective 0.88% : 0.000010s : 5: substitution.fold_const_symbol 0.66% : 0.000007s : 6: substitution.graph_param_transform 66.66% : 0.000753s : 28: substitution.inline 0.53% : 0.000006s : 10: substitution.j_node_and_user_rematch 0.67% : 0.000008s : 5: substitution.minmaximum_grad 0.93% : 0.000011s : 10: substitution.remove_not_recompute_node 0.52% : 0.000006s : 2: substitution.replace_old_param 1.84% : 0.000021s : 8: substitution.switch_simplify 4.34% : 0.000049s : 18: substitution.tuple_list_convert_item_index_to_positive 2.17% : 0.000025s : 19: substitution.tuple_list_get_item_const_eliminator 4.53% : 0.000051s : 19: substitution.tuple_list_get_item_depend_reorder 6.27% : 0.000071s : 24: substitution.tuple_list_get_item_eliminator 2.99% : 0.000034s : 19: substitution.tuple_list_get_set_item_eliminator 1.62% : 0.000018s : 6: substitution.updatestate_pure_node_eliminater 1.57% : 0.000018s : 9: substitution.updatestate_useless_node_eliminater ------[type_inference.] 13.378962 2 99.94% : 13.371120s : 1: type_inference.infer 0.06% : 0.007842s : 1: type_inference.specialize ------[replace.] 0.000377 43 2.79% : 0.000011s : 1: replace.depend_value_elim 55.63% : 0.000210s : 28: replace.inline 24.34% : 0.000092s : 8: replace.switch_simplify 4.37% : 0.000017s : 1: replace.tuple_list_get_item_depend_reorder 12.88% : 0.000049s : 5: replace.tuple_list_get_item_eliminator ------[match.] 0.000785 43 0.06% : 0.000000s : 1: match.depend_value_elim 93.91% : 0.000737s : 28: match.inline 2.03% : 0.000016s : 8: match.switch_simplify 2.32% : 0.000018s : 1: match.tuple_list_get_item_depend_reorder 1.67% : 0.000013s : 5: match.tuple_list_get_item_eliminator ------[predicate.] 0.000573 3425 1.08% : 0.000006s : 42: predicate.accumulaten_eliminater 0.45% : 0.000003s : 6: predicate.ad_related_special_op_eliminate 0.43% : 0.000002s : 19: predicate.addn_check_dump 1.09% : 0.000006s : 42: predicate.addn_zero_filter 1.01% : 0.000006s : 42: predicate.adjust_all_reduce_mul_add 2.28% : 0.000013s : 61: predicate.arithmetic_simplify 1.30% : 0.000007s : 42: predicate.cast_eliminate 0.35% : 0.000002s : 12: predicate.check_bprop_eliminate 0.44% : 0.000003s : 19: predicate.compare_switch_simplify 0.10% : 0.000001s : 6: predicate.const_output_eliminate 0.52% : 0.000003s : 20: predicate.depend_value_elim 1.15% : 0.000007s : 42: predicate.dict_get_item_const_eliminator 1.18% : 0.000007s : 42: predicate.dict_get_item_eliminator 1.11% : 0.000006s : 42: predicate.dict_set_item_eliminator 0.50% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.13% : 0.000001s : 6: predicate.elim_not_effective 0.26% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.31% : 0.000007s : 48: predicate.environ_add_const_eliminate 1.19% : 0.000007s : 48: predicate.environ_get_add_eliminate 1.19% : 0.000007s : 48: predicate.environ_get_depend_swap 1.69% : 0.000010s : 67: predicate.environ_get_eliminate 2.41% : 0.000014s : 48: predicate.environ_get_set_eliminate 2.10% : 0.000012s : 76: predicate.exchange_switch_depend_value 2.66% : 0.000015s : 76: predicate.float_depend_g_call 0.45% : 0.000003s : 19: predicate.float_environ_get_switch 0.64% : 0.000004s : 25: predicate.float_tuple_getitem_switch 0.09% : 0.000001s : 6: predicate.fold_const_symbol 0.35% : 0.000002s : 12: predicate.get_grad_eliminate 0.13% : 0.000001s : 6: predicate.graph_param_transform 0.48% : 0.000003s : 19: predicate.incorporate_call 0.41% : 0.000002s : 19: predicate.incorporate_call_switch 6.24% : 0.000036s : 161: predicate.inline 0.44% : 0.000003s : 12: predicate.inline_without_move 0.15% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.41% : 0.000002s : 12: predicate.less_batch_normalization 1.62% : 0.000009s : 60: predicate.list_to_tuple_eliminator_ 2.51% : 0.000014s : 102: predicate.load_eliminater 0.53% : 0.000003s : 6: predicate.loop_unroll_after_grad 3.57% : 0.000020s : 140: predicate.loop_unroll_before_grad 1.57% : 0.000009s : 55: predicate.make_slice_get_slice_eliminator 0.50% : 0.000003s : 19: predicate.merge_addn 0.32% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.42% : 0.000002s : 12: predicate.mini_step_allgather_replace 1.15% : 0.000007s : 42: predicate.minmaximum_grad 0.46% : 0.000003s : 6: predicate.mutable_eliminate 0.17% : 0.000001s : 6: predicate.opt_reshape 0.21% : 0.000001s : 6: predicate.parallel_virtual_node 2.71% : 0.000016s : 76: predicate.partial_defer_inline 1.51% : 0.000009s : 54: predicate.partial_eliminate 1.10% : 0.000006s : 42: predicate.print_const_string_wrapper 0.47% : 0.000003s : 18: predicate.reduce_all_const_elim 1.75% : 0.000010s : 42: predicate.reduce_eliminate 2.53% : 0.000015s : 102: predicate.redundant_stop_gradient_eliminater 0.26% : 0.000002s : 12: predicate.remove_not_recompute_node 1.21% : 0.000007s : 60: predicate.replace_applicator 0.20% : 0.000001s : 12: predicate.replace_old_param 0.12% : 0.000001s : 6: predicate.reset_defer_inline 1.20% : 0.000007s : 42: predicate.reshape_eliminate 0.39% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.18% : 0.000001s : 6: predicate.row_tensor_eliminate 0.45% : 0.000003s : 12: predicate.same_eliminate 0.22% : 0.000001s : 13: predicate.set_cell_output_no_recompute 0.36% : 0.000002s : 12: predicate.shard_identity_eliminate 0.44% : 0.000003s : 12: predicate.special_op_eliminate 0.58% : 0.000003s : 19: predicate.specialize_transform 0.56% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 0.50% : 0.000003s : 12: predicate.stack_unstack_eliminate 0.20% : 0.000001s : 6: predicate.switch_call_monad_eliminater 2.13% : 0.000012s : 76: predicate.switch_defer_inline 2.32% : 0.000013s : 88: predicate.switch_layer_defer_inline 7.40% : 0.000042s : 257: predicate.switch_simplify 1.16% : 0.000007s : 42: predicate.tile_eliminate 1.08% : 0.000006s : 42: predicate.transpose_eliminate 1.65% : 0.000009s : 54: predicate.tuple_list_convert_item_index_to_positive 1.65% : 0.000009s : 55: predicate.tuple_list_get_item_const_eliminator 1.56% : 0.000009s : 55: predicate.tuple_list_get_item_depend_reorder 2.94% : 0.000017s : 79: predicate.tuple_list_get_item_eliminator 1.59% : 0.000009s : 55: predicate.tuple_list_get_set_item_eliminator 2.38% : 0.000014s : 74: predicate.tuple_list_set_item_eliminator 1.55% : 0.000009s : 60: predicate.tuple_to_list_eliminator_ 2.46% : 0.000014s : 102: predicate.updatestate_pure_node_eliminater 3.12% : 0.000018s : 121: predicate.updatestate_useless_node_eliminater 0.21% : 0.000001s : 6: predicate.value_based_eliminate 0.32% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.33% : 0.000002s : 12: predicate.virtual_output_eliminate 0.18% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.25% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.008776 71 76.34% : 0.006699s : 39: func_graph_cloner_run.FuncGraphClonerGraph 23.66% : 0.002077s : 32: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 16.970479 213 0.00% : 0.000004s : 1: ForceFp32Comm 0.02% : 0.003897s : 1: add_attr 0.02% : 0.003885s : 1: add_attr_with_inline 0.00% : 0.000006s : 1: add_comm_op_reuse_tag 0.00% : 0.000074s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.00% : 0.000347s : 1: auto_monad 0.00% : 0.000035s : 1: auto_monad_reorder 0.00% : 0.000008s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000005s : 1: bias_add_comm_swap 0.00% : 0.000510s : 1: bootstrap 0.00% : 0.000035s : 1: cconv 0.00% : 0.000029s : 1: comm_op_add_attrs 0.00% : 0.000022s : 1: control_data_broadcast_order 0.00% : 0.000024s : 1: convert_after_rewriter 0.00% : 0.000030s : 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 0.00% : 0.000091s : 1: event_method 0.00% : 0.000020s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000004s : 1: get_jit_bprop_graph 0.00% : 0.000023s : 1: graph_reusing 0.00% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000010s : 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 0.00% : 0.000478s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000005s : 1: micro_interleaved_order_control 0.00% : 0.000728s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000019s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000019s : 1: opt.transform.mutable_eliminate 4.67% : 0.793062s : 95: opt.transform.opt_a 0.00% : 0.000054s : 1: opt.transform.opt_after_cconv 0.00% : 0.000037s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000176s : 28: opt.transform.opt_b 0.00% : 0.000064s : 2: opt.transform.opt_trans_graph 0.00% : 0.000097s : 4: opt.transform.symbol_engine_opt 4.72% : 0.801076s : 1: opt_a 0.00% : 0.000141s : 1: opt_after_cconv 0.00% : 0.000576s : 1: opt_after_jit_grad 0.00% : 0.000325s : 1: opt_b 4.74% : 0.805154s : 1: optimize 0.00% : 0.000027s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000012s : 1: order_py_execute_after_rewriter 0.00% : 0.000037s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000247s : 1: overlap_grad_ring_attention 0.00% : 0.000004s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000007s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000008s : 1: overlap_param_gather 0.00% : 0.000004s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000009s : 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 0.00% : 0.000101s : 1: pre_auto_parallel 0.00% : 0.000085s : 1: py_interpret_to_execute 0.00% : 0.000023s : 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 0.02% : 0.003149s : 1: renormalize.infer 0.01% : 0.002114s : 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% : 0.000606s : 1: rewriter_after_opt_a 0.00% : 0.000347s : 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% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000007s : 1: split_matmul_comm_elemetwise 0.00% : 0.000010s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000453s : 1: symbol_engine_optimizer 6.90% : 1.171607s : 1: task_emit 0.00% : 0.000094s : 1: tuple_transform 78.84% : 13.380153s : 1: type_inference 0.00% : 0.000133s : 1: validate TotalTime = 0.307716, [24] [bootstrap]: 0.00050616 [type_inference]: 0.00587636 [event_method]: 2.039e-05 [auto_monad]: 0.00017151 [graph_reusing]: 6.93e-06 [inline]: 2.29999e-06 [add_attr]: 0.13813, [1] [add_attr_with_inline]: 0.138117, [1] [Cycle 1]: 7.559e-05, [2] [tag_attr]: 2.6e-05 [meta_addattr_fg_expand]: 6.22001e-06 [parallel-infer-symbol]: 4.2e-06 [pre_auto_parallel]: 4.576e-05 [insert-virtual-dataset]: 2.59999e-06 [parallel-infer-symbol-second]: 7.80012e-07 [dataset_repeat_opt]: 2.05002e-06 [pipeline_split]: 1.64e-06 [optimize]: 0.154105, [53] [py_interpret_to_execute]: 3.474e-05 [rewriter_before_opt_a]: 9.326e-05 [opt_a]: 0.128065, [2] [Cycle 1]: 0.00256773, [45] [expand_dump_flag]: 3.13e-06 [switch_simplify]: 4.734e-05 [loop_unroll]: 3.16e-05 [a_1]: 0.00068023 [with_stream_mark]: 1.966e-05 [recompute_prepare]: 1.091e-05 [updatestate_depend_eliminate]: 6.04001e-06 [updatestate_assign_eliminate]: 5.62001e-06 [updatestate_loads_eliminate]: 5.12e-06 [parameter_eliminate]: 1.89e-06 [a_2]: 0.00013043 [accelerated_algorithm]: 9.83002e-06 [shard]: 1.79e-06 [meta_shard_fg_expand]: 2.60997e-06 [shard_inline]: 9.18002e-06 [merge_send_recv]: 1.107e-05 [auto_parallel]: 9.05999e-06 [parallel]: 2.041e-05 [flash_sp]: 9.49e-06 [merge_comm]: 5.52001e-06 [allreduce_fusion]: 5.36002e-06 [matmul_add_comm_reduction]: 1.178e-05 [allreduce_slice_to_reducescatter]: 8.99978e-07 [virtual_shard_identity]: 1.068e-05 [virtual_dataset]: 9.10001e-06 [get_grad_eliminate_]: 8.54e-06 [virtual_output]: 8.59e-06 [merge_forward]: 5.91003e-06 [cell_reuse_recompute_pass]: 1.48002e-06 [offload_activation]: 1.406e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.872e-05 [merge_recompute_call_nodes]: 1.76e-06 [before_grad]: 1.588e-05 [set_forward_comm_id_for_comm_node_pass]: 5.87001e-06 [meta_fg_expand]: 3.75998e-06 [flash_sp_send_recv_attached]: 2.98e-06 [receive_attached]: 2.32001e-06 [after_resolve]: 1.227e-05 [a_after_grad]: 1.356e-05 [renormalize]: 0.00101059 [add_forward_monad_depend]: 5.61998e-06 [auto_monad_grad]: 2.77002e-06 [auto_monad_eliminator]: 2.414e-05 [cse]: 3.999e-05 [a_3]: 6.623e-05 [Cycle 2]: 0.125485, [45] [expand_dump_flag]: 1.15999e-06 [switch_simplify]: 1.019e-05 [loop_unroll]: 8.52998e-06 [a_1]: 0.124607 [with_stream_mark]: 4.43e-05 [recompute_prepare]: 1.913e-05 [updatestate_depend_eliminate]: 7.53999e-06 [updatestate_assign_eliminate]: 5.40001e-06 [updatestate_loads_eliminate]: 6.63e-06 [parameter_eliminate]: 2.88e-06 [a_2]: 0.00012482 [accelerated_algorithm]: 8.95999e-06 [shard]: 3.8e-06 [meta_shard_fg_expand]: 4.54002e-06 [shard_inline]: 8.87e-06 [merge_send_recv]: 1.311e-05 [auto_parallel]: 1.231e-05 [parallel]: 1.207e-05 [flash_sp]: 4.86002e-06 [merge_comm]: 5.54e-06 [allreduce_fusion]: 5.44e-06 [matmul_add_comm_reduction]: 1.635e-05 [allreduce_slice_to_reducescatter]: 1.09003e-06 [virtual_shard_identity]: 1.075e-05 [virtual_dataset]: 8.65999e-06 [get_grad_eliminate_]: 8.35001e-06 [virtual_output]: 8.2e-06 [merge_forward]: 6.62002e-06 [cell_reuse_recompute_pass]: 3.32997e-06 [offload_activation]: 1.423e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.744e-05 [merge_recompute_call_nodes]: 1.50001e-06 [before_grad]: 1.468e-05 [set_forward_comm_id_for_comm_node_pass]: 7.21999e-06 [meta_fg_expand]: 4.06001e-06 [flash_sp_send_recv_attached]: 2.49001e-06 [receive_attached]: 2.32001e-06 [after_resolve]: 1.653e-05 [a_after_grad]: 1.246e-05 [renormalize]: 9.00181e-08 [add_forward_monad_depend]: 3.11999e-06 [auto_monad_grad]: 3.65998e-06 [auto_monad_eliminator]: 2.664e-05 [cse]: 4.938e-05 [a_3]: 5.501e-05 [py_interpret_to_execute_after_opt_a]: 2.367e-05 [slice_cell_reuse_recomputed_activation]: 2.21998e-06 [rewriter_after_opt_a]: 0.00025652 [convert_after_rewriter]: 1.184e-05 [order_py_execute_after_rewriter]: 7.13e-06 [mutable_eliminate]: 0.00082449 [opt_b]: 0.0228832, [1] [Cycle 1]: 0.0228693, [7] [b_1]: 0.00019765 [b_2]: 1.021e-05 [updatestate_depend_eliminate]: 1.011e-05 [updatestate_assign_eliminate]: 4.92999e-06 [updatestate_loads_eliminate]: 5.12e-06 [renormalize]: 9.09989e-07 [cse]: 0.0225566 [optimize_parallel_all_gather_comm]: 4.56e-05 [overlap_param_gather]: 2.41998e-06 [cconv]: 4.608e-05 [loop_unroll]: 0.00080024 [opt_after_cconv]: 0.00017092, [1] [Cycle 1]: 0.00016213, [7] [c_1]: 6.721e-05 [parameter_eliminate]: 7.26999e-06 [updatestate_depend_eliminate]: 1.07e-05 [updatestate_assign_eliminate]: 5.61e-06 [updatestate_loads_eliminate]: 5.11002e-06 [cse]: 2.923e-05 [renormalize]: 8.70001e-07 [remove_dup_value]: 1.665e-05 [tuple_transform]: 9.496e-05, [1] [Cycle 1]: 8.979e-05, [4] [d_1]: 6.091e-05 [none_parameter_eliminate]: 1.67999e-06 [renormalize]: 2.3999e-07 [switch_simplify]: 9.51e-06 [partial_unused_args_eliminate]: 1.96e-06 [add_recomputation]: 8.024e-05 [cse_after_recomputation]: 2.828e-05, [1] [Cycle 1]: 2.25e-05, [1] [cse]: 1.719e-05 [environ_conv]: 7.95e-06 [swap_dp_allreduce_reducescatter]: 8.26002e-06 [bias_add_comm_swap]: 3.49001e-06 [label_micro_interleaved_index]: 5.37999e-06 [label_fine_grained_interleaved_index]: 2.71999e-06 [merge_cast_opt]: 1.50001e-06 [slice_recompute_activation]: 2.81e-06 [micro_interleaved_order_control]: 2.67001e-06 [assign_add_opt]: 1.40999e-06 [ForceFp32Comm]: 9.70002e-07 [remove_cast_before_assign_add]: 1.06002e-06 [full_micro_interleaved_order_control]: 2.34001e-06 [reorder_send_recv_between_fp_bp]: 3.06001e-06 [comm_op_add_attrs]: 1.12999e-06 [add_comm_op_reuse_tag]: 1.04e-06 [interleave_split_concat_branches]: 1.20001e-06 [interleave_parallel_branches]: 1.20999e-06 [overlap_opt_shard_in_pipeline]: 1.19998e-06 [overlap_opt_shard_grad_in_pipeline]: 2.07001e-06 [control_data_broadcast_order]: 2.027e-05 [grouped_pairwise_exchange_alltoall]: 1.74e-06 [offloading_packed_experts]: 5.66e-06 [overlap_recompute_and_grad_model_parallel]: 6.58e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.22999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.41002e-06 [overlap_recompute_comm]: 2.98e-06 [overlap_grad_ring_attention]: 5.91e-06 [overlap_grad_flash_sp]: 3.045e-05 [begin_end_overlap_inline]: 5.19998e-07 [split_matmul_comm_elemetwise]: 2.32001e-06 [split_layernorm_comm]: 1.86e-06 [handle_group_info]: 1.14998e-06 [symbol_engine_optimizer]: 0.0002162, [1] [Cycle 1]: 0.00021134, [6] [build]: 0.00010737 [elim_shapecalc]: 1.369e-05 [elim_not_effective]: 2.107e-05 [opt_reshape]: 1.036e-05 [fold_const_symbol]: 2.559e-05 [renormalize]: 2.30008e-07 [detach_backward]: 2.48e-06 [pipeline_parallel_scheduler]: 1.94999e-06 [auto_monad_reorder]: 3.116e-05 [get_jit_bprop_graph]: 2.41e-06 [rewriter_after_jit_bprop_graph]: 6.91001e-06 [opt_after_jit_grad]: 0.00062209 [validate]: 0.00011199 [backend_pass]: 9.30013e-07 [task_emit]: 0.00781063 [execute]: 8.95999e-06 Sums bootstrap : 0.000506s : 0.30% type_inference : 0.005876s : 3.49% event_method : 0.000020s : 0.01% auto_monad : 0.000172s : 0.10% graph_reusing : 0.000007s : 0.00% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000026s : 0.02% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000006s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000046s : 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 : 0.000035s : 0.02% optimize.rewriter_before_opt_a : 0.000093s : 0.06% optimize.opt_a.expand_dump_flag : 0.000004s : 0.00% optimize.opt_a.switch_simplify : 0.000058s : 0.03% optimize.opt_a.loop_unroll : 0.000040s : 0.02% optimize.opt_a.a_1 : 0.125287s : 74.41% optimize.opt_a.with_stream_mark : 0.000064s : 0.04% optimize.opt_a.recompute_prepare : 0.000030s : 0.02% 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 : 0.000012s : 0.01% optimize.opt_a.parameter_eliminate : 0.000005s : 0.00% optimize.opt_a.a_2 : 0.000255s : 0.15% optimize.opt_a.accelerated_algorithm : 0.000019s : 0.01% optimize.opt_a.shard : 0.000006s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000007s : 0.00% optimize.opt_a.shard_inline : 0.000018s : 0.01% optimize.opt_a.merge_send_recv : 0.000024s : 0.01% optimize.opt_a.auto_parallel : 0.000021s : 0.01% optimize.opt_a.parallel : 0.000032s : 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 : 0.000011s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000028s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000021s : 0.01% optimize.opt_a.virtual_dataset : 0.000018s : 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 : 0.000005s : 0.00% optimize.opt_a.offload_activation : 0.000028s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000036s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000003s : 0.00% optimize.opt_a.before_grad : 0.000031s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000013s : 0.01% optimize.opt_a.meta_fg_expand : 0.000008s : 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 : 0.02% optimize.opt_a.a_after_grad : 0.000026s : 0.02% optimize.opt_a.renormalize : 0.001011s : 0.60% optimize.opt_a.add_forward_monad_depend : 0.000009s : 0.01% optimize.opt_a.auto_monad_grad : 0.000006s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000051s : 0.03% optimize.opt_a.cse : 0.000089s : 0.05% optimize.opt_a.a_3 : 0.000121s : 0.07% optimize.py_interpret_to_execute_after_opt_a : 0.000024s : 0.01% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000257s : 0.15% optimize.convert_after_rewriter : 0.000012s : 0.01% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000824s : 0.49% optimize.opt_b.b_1 : 0.000198s : 0.12% optimize.opt_b.b_2 : 0.000010s : 0.01% optimize.opt_b.updatestate_depend_eliminate : 0.000010s : 0.01% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000001s : 0.00% optimize.opt_b.cse : 0.022557s : 13.40% optimize.optimize_parallel_all_gather_comm : 0.000046s : 0.03% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000046s : 0.03% optimize.loop_unroll : 0.000800s : 0.48% optimize.opt_after_cconv.c_1 : 0.000067s : 0.04% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000011s : 0.01% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.cse : 0.000029s : 0.02% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000017s : 0.01% optimize.tuple_transform.d_1 : 0.000061s : 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 : 0.000010s : 0.01% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000080s : 0.05% optimize.cse_after_recomputation.cse : 0.000017s : 0.01% 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 : 0.000005s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000003s : 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 : 0.000002s : 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 : 0.000020s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000030s : 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 : 0.000107s : 0.06% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.01% optimize.symbol_engine_optimizer.elim_not_effective : 0.000021s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.01% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000026s : 0.02% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.00% pipeline_parallel_scheduler : 0.000002s : 0.00% auto_monad_reorder : 0.000031s : 0.02% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000007s : 0.00% opt_after_jit_grad : 0.000622s : 0.37% validate : 0.000112s : 0.07% backend_pass : 0.000001s : 0.00% task_emit : 0.007811s : 4.64% execute : 0.000009s : 0.01% Time group info: ------[substitution.] 0.000296 60 3.20% : 0.000009s : 2: substitution.depend_value_elim 1.82% : 0.000005s : 5: substitution.elim_not_effective 3.90% : 0.000012s : 5: substitution.fold_const_symbol 2.89% : 0.000009s : 6: substitution.graph_param_transform 60.01% : 0.000177s : 4: substitution.inline 1.97% : 0.000006s : 10: substitution.j_node_and_user_rematch 3.00% : 0.000009s : 10: substitution.remove_not_recompute_node 2.54% : 0.000008s : 2: substitution.replace_old_param 4.16% : 0.000012s : 2: substitution.tuple_list_get_item_eliminator 10.21% : 0.000030s : 6: substitution.updatestate_pure_node_eliminater 6.29% : 0.000019s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.005815 2 87.39% : 0.005082s : 1: type_inference.infer 12.61% : 0.000734s : 1: type_inference.specialize ------[replace.] 0.000061 6 71.83% : 0.000044s : 4: replace.inline 28.17% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000186 6 94.02% : 0.000175s : 4: match.inline 5.98% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000306 1652 0.87% : 0.000003s : 16: predicate.accumulaten_eliminater 0.93% : 0.000003s : 6: predicate.ad_related_special_op_eliminate 0.64% : 0.000002s : 12: predicate.addn_check_dump 1.05% : 0.000003s : 16: predicate.addn_zero_filter 0.85% : 0.000003s : 16: predicate.adjust_all_reduce_mul_add 2.19% : 0.000007s : 28: predicate.arithmetic_simplify 0.88% : 0.000003s : 16: predicate.cast_eliminate 0.66% : 0.000002s : 12: predicate.check_bprop_eliminate 0.53% : 0.000002s : 12: predicate.compare_switch_simplify 0.15% : 0.000000s : 6: predicate.const_output_eliminate 0.71% : 0.000002s : 12: predicate.depend_value_elim 0.84% : 0.000003s : 16: predicate.dict_get_item_const_eliminator 1.00% : 0.000003s : 16: predicate.dict_get_item_eliminator 0.90% : 0.000003s : 16: predicate.dict_set_item_eliminator 1.07% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.27% : 0.000001s : 6: predicate.elim_not_effective 0.38% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.29% : 0.000004s : 22: predicate.environ_add_const_eliminate 1.03% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.12% : 0.000003s : 22: predicate.environ_get_depend_swap 1.81% : 0.000006s : 34: predicate.environ_get_eliminate 1.10% : 0.000003s : 22: predicate.environ_get_set_eliminate 1.10% : 0.000003s : 22: predicate.exchange_switch_depend_value 2.23% : 0.000007s : 22: predicate.float_depend_g_call 0.62% : 0.000002s : 12: predicate.float_environ_get_switch 0.86% : 0.000003s : 18: predicate.float_tuple_getitem_switch 0.16% : 0.000000s : 6: predicate.fold_const_symbol 0.62% : 0.000002s : 12: predicate.get_grad_eliminate 0.25% : 0.000001s : 6: predicate.graph_param_transform 0.62% : 0.000002s : 12: predicate.incorporate_call 0.53% : 0.000002s : 12: predicate.incorporate_call_switch 6.38% : 0.000020s : 74: predicate.inline 0.74% : 0.000002s : 12: predicate.inline_without_move 0.29% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.85% : 0.000003s : 12: predicate.less_batch_normalization 2.73% : 0.000008s : 30: predicate.list_to_tuple_eliminator_ 2.19% : 0.000007s : 46: predicate.load_eliminater 1.12% : 0.000003s : 6: predicate.loop_unroll_after_grad 1.80% : 0.000006s : 38: predicate.loop_unroll_before_grad 1.86% : 0.000006s : 28: predicate.make_slice_get_slice_eliminator 0.76% : 0.000002s : 12: predicate.merge_addn 0.61% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.72% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.76% : 0.000002s : 16: predicate.minmaximum_grad 1.16% : 0.000004s : 6: predicate.mutable_eliminate 0.35% : 0.000001s : 6: predicate.opt_reshape 0.37% : 0.000001s : 6: predicate.parallel_virtual_node 2.87% : 0.000009s : 22: predicate.partial_defer_inline 1.28% : 0.000004s : 24: predicate.partial_eliminate 0.82% : 0.000003s : 16: predicate.print_const_string_wrapper 0.66% : 0.000002s : 12: predicate.reduce_all_const_elim 1.06% : 0.000003s : 16: predicate.reduce_eliminate 2.43% : 0.000007s : 46: predicate.redundant_stop_gradient_eliminater 0.42% : 0.000001s : 12: predicate.remove_not_recompute_node 1.09% : 0.000003s : 30: predicate.replace_applicator 0.51% : 0.000002s : 12: predicate.replace_old_param 0.23% : 0.000001s : 6: predicate.reset_defer_inline 0.89% : 0.000003s : 16: predicate.reshape_eliminate 0.72% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.41% : 0.000001s : 6: predicate.row_tensor_eliminate 0.89% : 0.000003s : 12: predicate.same_eliminate 0.45% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.81% : 0.000002s : 12: predicate.shard_identity_eliminate 0.79% : 0.000002s : 12: predicate.special_op_eliminate 0.74% : 0.000002s : 12: predicate.specialize_transform 0.83% : 0.000003s : 12: predicate.split_environ_get_set_with_tuple_value 0.87% : 0.000003s : 12: predicate.stack_unstack_eliminate 0.38% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.18% : 0.000004s : 22: predicate.switch_defer_inline 1.80% : 0.000006s : 34: predicate.switch_layer_defer_inline 4.26% : 0.000013s : 78: predicate.switch_simplify 1.11% : 0.000003s : 16: predicate.tile_eliminate 0.94% : 0.000003s : 16: predicate.transpose_eliminate 1.49% : 0.000005s : 28: predicate.tuple_list_convert_item_index_to_positive 1.60% : 0.000005s : 28: predicate.tuple_list_get_item_const_eliminator 1.59% : 0.000005s : 28: predicate.tuple_list_get_item_depend_reorder 3.21% : 0.000010s : 42: predicate.tuple_list_get_item_eliminator 1.33% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.31% : 0.000007s : 40: predicate.tuple_list_set_item_eliminator 1.66% : 0.000005s : 30: predicate.tuple_to_list_eliminator_ 2.21% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.54% : 0.000011s : 58: predicate.updatestate_useless_node_eliminater 0.41% : 0.000001s : 6: predicate.value_based_eliminate 0.70% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.73% : 0.000002s : 12: predicate.virtual_output_eliminate 0.32% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.50% : 0.000002s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000579 11 49.86% : 0.000289s : 5: func_graph_cloner_run.FuncGraphClonerGraph 50.14% : 0.000290s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.727229 196 0.00% : 0.000004s : 1: ForceFp32Comm 19.00% : 0.138138s : 1: add_attr 18.99% : 0.138122s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000084s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.02% : 0.000179s : 1: auto_monad 0.00% : 0.000035s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.07% : 0.000539s : 1: bootstrap 0.01% : 0.000050s : 1: cconv 0.00% : 0.000005s : 1: comm_op_add_attrs 0.00% : 0.000023s : 1: control_data_broadcast_order 0.00% : 0.000016s : 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% : 0.000011s : 1: environ_conv 0.00% : 0.000026s : 1: event_method 0.00% : 0.000016s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000011s : 1: graph_reusing 0.00% : 0.000004s : 1: grouped_pairwise_exchange_alltoall 0.00% : 0.000004s : 1: handle_group_info 0.00% : 0.000006s : 1: inline 0.00% : 0.000006s : 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 0.11% : 0.000809s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.12% : 0.000836s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000028s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000025s : 1: opt.transform.mutable_eliminate 17.32% : 0.125924s : 78: opt.transform.opt_a 0.01% : 0.000066s : 1: opt.transform.opt_after_cconv 0.00% : 0.000036s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000176s : 28: opt.transform.opt_b 0.01% : 0.000068s : 2: opt.transform.opt_trans_graph 0.01% : 0.000067s : 4: opt.transform.symbol_engine_opt 17.61% : 0.128070s : 1: opt_a 0.02% : 0.000174s : 1: opt_after_cconv 0.09% : 0.000632s : 1: opt_after_jit_grad 3.15% : 0.022892s : 1: opt_b 21.19% : 0.154111s : 1: optimize 0.01% : 0.000053s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 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% : 0.000009s : 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% : 0.000006s : 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 0.01% : 0.000050s : 1: pre_auto_parallel 0.01% : 0.000039s : 1: py_interpret_to_execute 0.00% : 0.000028s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000021s : 1: remove_dup_value 0.08% : 0.000584s : 1: renormalize.infer 0.06% : 0.000417s : 1: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000011s : 1: rewriter_after_jit_bprop_graph 0.04% : 0.000263s : 1: rewriter_after_opt_a 0.01% : 0.000099s : 1: rewriter_before_opt_a 0.00% : 0.000007s : 1: slice_cell_reuse_recomputed_activation 0.00% : 0.000006s : 1: slice_recompute_activation 0.00% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.03% : 0.000219s : 1: symbol_engine_optimizer 1.08% : 0.007830s : 1: task_emit 0.01% : 0.000099s : 1: tuple_transform 0.81% : 0.005895s : 1: type_inference 0.02% : 0.000145s : 1: validate *********************Begin DynamicTest for with GRAPH_MODE_O0 ********************* Start DynamicTest running on GRAPH_MODE_O0 with static shape with first inputs... End Start DynamicTest running on GRAPH_MODE_O0 with static shape with second inputs... End Start DynamicTest testing with [GRAPH_MODE_O0] [Resize]... Compare ['Success'] for 0'th output of Resize. End Start DynamicTest testing with [GRAPH_MODE_O0] [DYNAMIC_SHAPE] [None]... TotalTime = 51.2682, [24] [bootstrap]: 0.00084111 [type_inference]: 0.740815 [event_method]: 6.701e-05 [auto_monad]: 0.00027092 [graph_reusing]: 9.97999e-06 [inline]: 2.36e-06 [add_attr]: 0.00416097, [1] [add_attr_with_inline]: 0.00414984, [1] [Cycle 1]: 9.622e-05, [2] [tag_attr]: 5.076e-05 [meta_addattr_fg_expand]: 1.166e-05 [parallel-infer-symbol]: 4.40999e-06 [pre_auto_parallel]: 6.413e-05 [insert-virtual-dataset]: 2.75002e-06 [parallel-infer-symbol-second]: 7.09988e-07 [dataset_repeat_opt]: 2.02999e-06 [pipeline_split]: 1.69e-06 [optimize]: 1.00637, [53] [py_interpret_to_execute]: 4.714e-05 [rewriter_before_opt_a]: 0.00019904 [opt_a]: 1.00325, [4] [Cycle 1]: 0.537278, [45] [expand_dump_flag]: 4.27e-06 [switch_simplify]: 0.00010154 [loop_unroll]: 7.315e-05 [a_1]: 0.0019418 [with_stream_mark]: 3.149e-05 [recompute_prepare]: 2.661e-05 [updatestate_depend_eliminate]: 1.302e-05 [updatestate_assign_eliminate]: 1.451e-05 [updatestate_loads_eliminate]: 1.074e-05 [parameter_eliminate]: 3.24001e-06 [a_2]: 0.00031057 [accelerated_algorithm]: 2.108e-05 [shard]: 2.26998e-06 [meta_shard_fg_expand]: 5.25001e-06 [shard_inline]: 1.956e-05 [merge_send_recv]: 2.177e-05 [auto_parallel]: 1.55e-05 [parallel]: 2.042e-05 [flash_sp]: 1.205e-05 [merge_comm]: 1.296e-05 [allreduce_fusion]: 1.241e-05 [matmul_add_comm_reduction]: 3.36e-05 [allreduce_slice_to_reducescatter]: 8.00006e-07 [virtual_shard_identity]: 2.236e-05 [virtual_dataset]: 1.961e-05 [get_grad_eliminate_]: 1.899e-05 [virtual_output]: 1.885e-05 [merge_forward]: 1.259e-05 [cell_reuse_recompute_pass]: 1.29e-06 [offload_activation]: 2.315e-05 [cell_reuse_handle_not_recompute_node_pass]: 3.801e-05 [merge_recompute_call_nodes]: 1.76e-06 [before_grad]: 5.531e-05 [set_forward_comm_id_for_comm_node_pass]: 1.557e-05 [meta_fg_expand]: 0.524861 [flash_sp_send_recv_attached]: 5.97001e-06 [receive_attached]: 3.08e-06 [after_resolve]: 0.0001258 [a_after_grad]: 0.00018874 [renormalize]: 0.00670594 [add_forward_monad_depend]: 1.696e-05 [auto_monad_grad]: 8.59e-06 [auto_monad_eliminator]: 9.2e-05 [cse]: 0.00030838 [a_3]: 0.00163753 [Cycle 2]: 0.463405, [45] [expand_dump_flag]: 3.93999e-06 [switch_simplify]: 6.628e-05 [loop_unroll]: 6.06e-05 [a_1]: 0.00294425 [with_stream_mark]: 3.765e-05 [recompute_prepare]: 2.463e-05 [updatestate_depend_eliminate]: 1.266e-05 [updatestate_assign_eliminate]: 1.217e-05 [updatestate_loads_eliminate]: 1.112e-05 [parameter_eliminate]: 4.51002e-06 [a_2]: 0.457792 [accelerated_algorithm]: 2.639e-05 [shard]: 6.07001e-06 [meta_shard_fg_expand]: 7.36999e-06 [shard_inline]: 1.175e-05 [merge_send_recv]: 2.481e-05 [auto_parallel]: 1.962e-05 [parallel]: 1.197e-05 [flash_sp]: 7.62998e-06 [merge_comm]: 7.28999e-06 [allreduce_fusion]: 6.63998e-06 [matmul_add_comm_reduction]: 3.37e-05 [allreduce_slice_to_reducescatter]: 9.60019e-07 [virtual_shard_identity]: 1.342e-05 [virtual_dataset]: 1.721e-05 [get_grad_eliminate_]: 1.087e-05 [virtual_output]: 1.101e-05 [merge_forward]: 7.26001e-06 [cell_reuse_recompute_pass]: 3.04001e-06 [offload_activation]: 1.646e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.81e-05 [merge_recompute_call_nodes]: 1.66e-06 [before_grad]: 2.157e-05 [set_forward_comm_id_for_comm_node_pass]: 8.12e-06 [meta_fg_expand]: 0.00016048 [flash_sp_send_recv_attached]: 1.99e-06 [receive_attached]: 2.74001e-06 [after_resolve]: 1.979e-05 [a_after_grad]: 1.82e-05 [renormalize]: 0.00131025 [add_forward_monad_depend]: 8.13001e-06 [auto_monad_grad]: 2.561e-05 [auto_monad_eliminator]: 3.968e-05 [cse]: 7.841e-05 [a_3]: 8.717e-05 [Cycle 3]: 0.00164485, [45] [expand_dump_flag]: 2.71999e-06 [switch_simplify]: 1.258e-05 [loop_unroll]: 1.141e-05 [a_1]: 0.0003256 [with_stream_mark]: 1.811e-05 [recompute_prepare]: 1.126e-05 [updatestate_depend_eliminate]: 5.649e-05 [updatestate_assign_eliminate]: 6.45997e-06 [updatestate_loads_eliminate]: 5.30999e-06 [parameter_eliminate]: 2.17999e-06 [a_2]: 0.00012502 [accelerated_algorithm]: 9.99001e-06 [shard]: 1.57999e-06 [meta_shard_fg_expand]: 2.20002e-06 [shard_inline]: 8.82e-06 [merge_send_recv]: 9.33002e-06 [auto_parallel]: 8.84e-06 [parallel]: 6.94001e-06 [flash_sp]: 1.42e-06 [merge_comm]: 5.24e-06 [allreduce_fusion]: 5.34e-06 [matmul_add_comm_reduction]: 9.14998e-06 [allreduce_slice_to_reducescatter]: 4.00003e-07 [virtual_shard_identity]: 1.017e-05 [virtual_dataset]: 8.97999e-06 [get_grad_eliminate_]: 9.51003e-06 [virtual_output]: 8.56997e-06 [merge_forward]: 5.14e-06 [cell_reuse_recompute_pass]: 1.55999e-06 [offload_activation]: 1.173e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.794e-05 [merge_recompute_call_nodes]: 8.80013e-07 [before_grad]: 1.589e-05 [set_forward_comm_id_for_comm_node_pass]: 6.02999e-06 [meta_fg_expand]: 3.40998e-06 [flash_sp_send_recv_attached]: 1.04998e-06 [receive_attached]: 1.08001e-06 [after_resolve]: 1.267e-05 [a_after_grad]: 1.381e-05 [renormalize]: 0.00052665 [add_forward_monad_depend]: 4.2e-06 [auto_monad_grad]: 1.91e-06 [auto_monad_eliminator]: 1.705e-05 [cse]: 3.263e-05 [a_3]: 6.451e-05 [Cycle 4]: 0.00089598, [45] [expand_dump_flag]: 1.00999e-06 [switch_simplify]: 1.177e-05 [loop_unroll]: 9.39e-06 [a_1]: 0.00023153 [with_stream_mark]: 1.134e-05 [recompute_prepare]: 9.72999e-06 [updatestate_depend_eliminate]: 5.57999e-06 [updatestate_assign_eliminate]: 4.75999e-06 [updatestate_loads_eliminate]: 4.45999e-06 [parameter_eliminate]: 9.60019e-07 [a_2]: 0.00012012 [accelerated_algorithm]: 9.49e-06 [shard]: 1.15999e-06 [meta_shard_fg_expand]: 2.08998e-06 [shard_inline]: 9.20999e-06 [merge_send_recv]: 7.43999e-06 [auto_parallel]: 7.67998e-06 [parallel]: 4.32e-06 [flash_sp]: 1.09e-06 [merge_comm]: 5.72999e-06 [allreduce_fusion]: 5.22e-06 [matmul_add_comm_reduction]: 8.69e-06 [allreduce_slice_to_reducescatter]: 3.19997e-07 [virtual_shard_identity]: 1.027e-05 [virtual_dataset]: 8.78001e-06 [get_grad_eliminate_]: 9.04998e-06 [virtual_output]: 8.55001e-06 [merge_forward]: 4.49002e-06 [cell_reuse_recompute_pass]: 1.27e-06 [offload_activation]: 9.70002e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.817e-05 [merge_recompute_call_nodes]: 7.09988e-07 [before_grad]: 1.524e-05 [set_forward_comm_id_for_comm_node_pass]: 5.82001e-06 [meta_fg_expand]: 3.30998e-06 [flash_sp_send_recv_attached]: 8.80013e-07 [receive_attached]: 1.07e-06 [after_resolve]: 1.189e-05 [a_after_grad]: 1.372e-05 [renormalize]: 1.00001e-07 [add_forward_monad_depend]: 1.50999e-06 [auto_monad_grad]: 9.39996e-07 [auto_monad_eliminator]: 1.29e-05 [cse]: 2.571e-05 [a_3]: 5.871e-05 [py_interpret_to_execute_after_opt_a]: 1.524e-05 [slice_cell_reuse_recomputed_activation]: 1.92001e-06 [rewriter_after_opt_a]: 6.987e-05 [convert_after_rewriter]: 1.018e-05 [order_py_execute_after_rewriter]: 6.86001e-06 [mutable_eliminate]: 0.00073672 [opt_b]: 0.00032362, [1] [Cycle 1]: 0.00031534, [7] [b_1]: 0.00021094 [b_2]: 1.2e-05 [updatestate_depend_eliminate]: 8.88002e-06 [updatestate_assign_eliminate]: 4.72e-06 [updatestate_loads_eliminate]: 5.07999e-06 [renormalize]: 2.3999e-07 [cse]: 3.43e-05 [optimize_parallel_all_gather_comm]: 2.809e-05 [overlap_param_gather]: 2.24001e-06 [cconv]: 3.007e-05 [loop_unroll]: 0.00055569 [opt_after_cconv]: 0.0001636, [1] [Cycle 1]: 0.00015683, [7] [c_1]: 6.458e-05 [parameter_eliminate]: 3.4e-06 [updatestate_depend_eliminate]: 9.40001e-06 [updatestate_assign_eliminate]: 4.94e-06 [updatestate_loads_eliminate]: 5.22999e-06 [cse]: 3.168e-05 [renormalize]: 5.19998e-07 [remove_dup_value]: 1.844e-05 [tuple_transform]: 0.0001093, [1] [Cycle 1]: 0.00010423, [4] [d_1]: 7.097e-05 [none_parameter_eliminate]: 2.56998e-06 [renormalize]: 1.40019e-07 [switch_simplify]: 1.097e-05 [partial_unused_args_eliminate]: 1.98997e-06 [add_recomputation]: 8.549e-05 [cse_after_recomputation]: 3.474e-05, [1] [Cycle 1]: 2.973e-05, [1] [cse]: 2.378e-05 [environ_conv]: 7.78001e-06 [swap_dp_allreduce_reducescatter]: 9.32001e-06 [bias_add_comm_swap]: 2.97002e-06 [label_micro_interleaved_index]: 5.24998e-06 [label_fine_grained_interleaved_index]: 2.96999e-06 [merge_cast_opt]: 1.64e-06 [slice_recompute_activation]: 2.79001e-06 [micro_interleaved_order_control]: 2.81999e-06 [assign_add_opt]: 1.69e-06 [ForceFp32Comm]: 1.03001e-06 [remove_cast_before_assign_add]: 1.16997e-06 [full_micro_interleaved_order_control]: 2.42001e-06 [reorder_send_recv_between_fp_bp]: 2.81e-06 [comm_op_add_attrs]: 1.14998e-06 [add_comm_op_reuse_tag]: 1.15999e-06 [interleave_split_concat_branches]: 1.24e-06 [interleave_parallel_branches]: 1.42e-06 [overlap_opt_shard_in_pipeline]: 1.19998e-06 [overlap_opt_shard_grad_in_pipeline]: 2.02999e-06 [control_data_broadcast_order]: 2.055e-05 [grouped_pairwise_exchange_alltoall]: 1.69998e-06 [offloading_packed_experts]: 5.91e-06 [overlap_recompute_and_grad_model_parallel]: 6.84999e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.23002e-06 [overlap_recompute_allgather_and_fa_grad]: 1.50999e-06 [overlap_recompute_comm]: 2.79001e-06 [overlap_grad_ring_attention]: 6.46999e-06 [overlap_grad_flash_sp]: 3.512e-05 [begin_end_overlap_inline]: 7.2e-07 [split_matmul_comm_elemetwise]: 2.63e-06 [split_layernorm_comm]: 1.82999e-06 [handle_group_info]: 1.07e-06 [symbol_engine_optimizer]: 0.00028889, [1] [Cycle 1]: 0.000284, [6] [build]: 0.00017074 [elim_shapecalc]: 1.595e-05 [elim_not_effective]: 2.411e-05 [opt_reshape]: 1.123e-05 [fold_const_symbol]: 2.786e-05 [renormalize]: 4.69998e-07 [detach_backward]: 2.68998e-06 [pipeline_parallel_scheduler]: 1.94e-06 [auto_monad_reorder]: 3.393e-05 [get_jit_bprop_graph]: 2.26e-06 [rewriter_after_jit_bprop_graph]: 4.85001e-06 [opt_after_jit_grad]: 0.00057887 [validate]: 5.629e-05 [backend_pass]: 1.13001e-06 [task_emit]: 49.5146 [execute]: 1.083e-05 Sums bootstrap : 0.000841s : 0.00% type_inference : 0.740815s : 1.45% event_method : 0.000067s : 0.00% auto_monad : 0.000271s : 0.00% graph_reusing : 0.000010s : 0.00% inline : 0.000002s : 0.00% add_attr.add_attr_with_inline.tag_attr : 0.000051s : 0.00% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000012s : 0.00% parallel-infer-symbol : 0.000004s : 0.00% pre_auto_parallel : 0.000064s : 0.00% insert-virtual-dataset : 0.000003s : 0.00% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.00% pipeline_split : 0.000002s : 0.00% optimize.py_interpret_to_execute : 0.000047s : 0.00% optimize.rewriter_before_opt_a : 0.000199s : 0.00% optimize.opt_a.expand_dump_flag : 0.000012s : 0.00% optimize.opt_a.switch_simplify : 0.000192s : 0.00% optimize.opt_a.loop_unroll : 0.000155s : 0.00% optimize.opt_a.a_1 : 0.005443s : 0.01% optimize.opt_a.with_stream_mark : 0.000099s : 0.00% optimize.opt_a.recompute_prepare : 0.000072s : 0.00% optimize.opt_a.updatestate_depend_eliminate : 0.000088s : 0.00% optimize.opt_a.updatestate_assign_eliminate : 0.000038s : 0.00% optimize.opt_a.updatestate_loads_eliminate : 0.000032s : 0.00% optimize.opt_a.parameter_eliminate : 0.000011s : 0.00% optimize.opt_a.a_2 : 0.458348s : 0.89% optimize.opt_a.accelerated_algorithm : 0.000067s : 0.00% optimize.opt_a.shard : 0.000011s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000017s : 0.00% optimize.opt_a.shard_inline : 0.000049s : 0.00% optimize.opt_a.merge_send_recv : 0.000063s : 0.00% optimize.opt_a.auto_parallel : 0.000052s : 0.00% optimize.opt_a.parallel : 0.000044s : 0.00% optimize.opt_a.flash_sp : 0.000022s : 0.00% optimize.opt_a.merge_comm : 0.000031s : 0.00% optimize.opt_a.allreduce_fusion : 0.000030s : 0.00% optimize.opt_a.matmul_add_comm_reduction : 0.000085s : 0.00% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000056s : 0.00% optimize.opt_a.virtual_dataset : 0.000055s : 0.00% optimize.opt_a.get_grad_eliminate_ : 0.000048s : 0.00% optimize.opt_a.virtual_output : 0.000047s : 0.00% optimize.opt_a.merge_forward : 0.000029s : 0.00% optimize.opt_a.cell_reuse_recompute_pass : 0.000007s : 0.00% optimize.opt_a.offload_activation : 0.000061s : 0.00% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000102s : 0.00% optimize.opt_a.merge_recompute_call_nodes : 0.000005s : 0.00% optimize.opt_a.before_grad : 0.000108s : 0.00% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000036s : 0.00% optimize.opt_a.meta_fg_expand : 0.525028s : 1.02% optimize.opt_a.flash_sp_send_recv_attached : 0.000010s : 0.00% optimize.opt_a.receive_attached : 0.000008s : 0.00% optimize.opt_a.after_resolve : 0.000170s : 0.00% optimize.opt_a.a_after_grad : 0.000234s : 0.00% optimize.opt_a.renormalize : 0.008543s : 0.02% optimize.opt_a.add_forward_monad_depend : 0.000031s : 0.00% optimize.opt_a.auto_monad_grad : 0.000037s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000162s : 0.00% optimize.opt_a.cse : 0.000445s : 0.00% optimize.opt_a.a_3 : 0.001848s : 0.00% optimize.py_interpret_to_execute_after_opt_a : 0.000015s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000070s : 0.00% optimize.convert_after_rewriter : 0.000010s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000737s : 0.00% optimize.opt_b.b_1 : 0.000211s : 0.00% optimize.opt_b.b_2 : 0.000012s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000034s : 0.00% optimize.optimize_parallel_all_gather_comm : 0.000028s : 0.00% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000030s : 0.00% optimize.loop_unroll : 0.000556s : 0.00% optimize.opt_after_cconv.c_1 : 0.000065s : 0.00% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000009s : 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 : 0.000032s : 0.00% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000018s : 0.00% optimize.tuple_transform.d_1 : 0.000071s : 0.00% optimize.tuple_transform.none_parameter_eliminate : 0.000003s : 0.00% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000011s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000085s : 0.00% optimize.cse_after_recomputation.cse : 0.000024s : 0.00% optimize.environ_conv : 0.000008s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 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 : 0.000003s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000002s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 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 : 0.000021s : 0.00% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000006s : 0.00% optimize.overlap_grad_flash_sp : 0.000035s : 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 : 0.000171s : 0.00% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000016s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000024s : 0.00% optimize.symbol_engine_optimizer.opt_reshape : 0.000011s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000028s : 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 : 0.000034s : 0.00% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000005s : 0.00% opt_after_jit_grad : 0.000579s : 0.00% validate : 0.000056s : 0.00% backend_pass : 0.000001s : 0.00% task_emit : 49.514621s : 96.59% execute : 0.000011s : 0.00% Time group info: ------[substitution.] 0.002513 320 0.96% : 0.000024s : 8: substitution.depend_value_elim 0.23% : 0.000006s : 6: substitution.elim_not_effective 0.20% : 0.000005s : 5: substitution.float_depend_g_call 0.19% : 0.000005s : 3: substitution.float_tuple_getitem_switch 0.50% : 0.000013s : 6: substitution.fold_const_symbol 31.39% : 0.000789s : 4: substitution.getattr_setattr_resolve 0.38% : 0.000010s : 7: substitution.graph_param_transform 0.13% : 0.000003s : 2: substitution.incorporate_call 0.08% : 0.000002s : 2: substitution.incorporate_call_switch 45.20% : 0.001136s : 22: substitution.inline 1.54% : 0.000039s : 5: substitution.inline_without_move 0.75% : 0.000019s : 35: substitution.j_node_and_user_rematch 0.74% : 0.000018s : 11: substitution.minmaximum_grad 0.28% : 0.000007s : 5: substitution.partial_eliminate 1.04% : 0.000026s : 35: substitution.remove_not_recompute_node 2.68% : 0.000067s : 22: substitution.replace_applicator 0.75% : 0.000019s : 17: substitution.replace_old_param 0.24% : 0.000006s : 2: substitution.set_cell_output_no_recompute 1.53% : 0.000038s : 11: substitution.tuple_list_convert_item_index_to_positive 0.68% : 0.000017s : 11: substitution.tuple_list_get_item_const_eliminator 0.93% : 0.000023s : 11: substitution.tuple_list_get_item_depend_reorder 3.55% : 0.000089s : 30: substitution.tuple_list_get_item_eliminator 2.08% : 0.000052s : 11: substitution.tuple_list_get_set_item_eliminator 1.32% : 0.000033s : 20: substitution.updatestate_pure_node_eliminater 2.66% : 0.000067s : 29: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.740710 2 99.76% : 0.738939s : 1: type_inference.infer 0.24% : 0.001772s : 1: type_inference.specialize ------[replace.] 0.001244 43 5.33% : 0.000066s : 3: replace.getattr_setattr_resolve 25.21% : 0.000314s : 22: replace.inline 54.77% : 0.000681s : 1: replace.replace_applicator 11.79% : 0.000147s : 16: replace.tuple_list_get_item_eliminator 2.90% : 0.000036s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001932 43 38.52% : 0.000744s : 3: match.getattr_setattr_resolve 58.13% : 0.001123s : 22: match.inline 0.55% : 0.000011s : 1: match.replace_applicator 2.24% : 0.000043s : 16: match.tuple_list_get_item_eliminator 0.56% : 0.000011s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.001260 8144 1.00% : 0.000013s : 88: predicate.accumulaten_eliminater 0.20% : 0.000002s : 7: predicate.ad_related_special_op_eliminate 0.62% : 0.000008s : 56: predicate.addn_check_dump 0.96% : 0.000012s : 88: predicate.addn_zero_filter 0.92% : 0.000012s : 88: predicate.adjust_all_reduce_mul_add 2.28% : 0.000029s : 139: predicate.arithmetic_simplify 0.96% : 0.000012s : 88: predicate.cast_eliminate 1.69% : 0.000021s : 165: predicate.check_bprop_eliminate 0.65% : 0.000008s : 56: predicate.compare_switch_simplify 0.05% : 0.000001s : 7: predicate.const_output_eliminate 0.86% : 0.000011s : 51: predicate.depend_value_elim 1.04% : 0.000013s : 88: predicate.dict_get_item_const_eliminator 1.03% : 0.000013s : 88: predicate.dict_get_item_eliminator 0.98% : 0.000012s : 88: predicate.dict_set_item_eliminator 0.23% : 0.000003s : 14: predicate.dumpgradient_eliminate 0.08% : 0.000001s : 7: predicate.elim_not_effective 0.11% : 0.000001s : 7: predicate.elim_shapecalc_of_broadcastargs 1.08% : 0.000014s : 95: predicate.environ_add_const_eliminate 1.05% : 0.000013s : 95: predicate.environ_get_add_eliminate 1.02% : 0.000013s : 95: predicate.environ_get_depend_swap 2.22% : 0.000028s : 146: predicate.environ_get_eliminate 1.04% : 0.000013s : 95: predicate.environ_get_set_eliminate 1.44% : 0.000018s : 126: predicate.exchange_switch_depend_value 1.96% : 0.000025s : 126: predicate.float_depend_g_call 0.66% : 0.000008s : 56: predicate.float_environ_get_switch 0.69% : 0.000009s : 63: predicate.float_tuple_getitem_switch 0.05% : 0.000001s : 7: predicate.fold_const_symbol 0.50% : 0.000006s : 42: predicate.get_grad_eliminate 0.51% : 0.000006s : 20: predicate.getattr_setattr_resolve 0.07% : 0.000001s : 7: predicate.graph_param_transform 0.58% : 0.000007s : 51: predicate.incorporate_call 0.52% : 0.000007s : 51: predicate.incorporate_call_switch 6.32% : 0.000080s : 326: predicate.inline 1.63% : 0.000021s : 110: predicate.inline_without_move 0.25% : 0.000003s : 42: predicate.j_node_and_user_rematch 0.66% : 0.000008s : 42: predicate.less_batch_normalization 1.35% : 0.000017s : 118: predicate.list_to_tuple_eliminator_ 2.25% : 0.000028s : 206: predicate.load_eliminater 0.27% : 0.000003s : 7: predicate.loop_unroll_after_grad 1.87% : 0.000024s : 174: predicate.loop_unroll_before_grad 1.19% : 0.000015s : 102: predicate.make_slice_get_slice_eliminator 0.67% : 0.000008s : 56: predicate.merge_addn 1.66% : 0.000021s : 161: predicate.micro_step_allgather_replace 1.69% : 0.000021s : 161: predicate.mini_step_allgather_replace 0.98% : 0.000012s : 88: predicate.minmaximum_grad 0.24% : 0.000003s : 7: predicate.mutable_eliminate 0.11% : 0.000001s : 7: predicate.opt_reshape 0.09% : 0.000001s : 7: predicate.parallel_virtual_node 1.89% : 0.000024s : 126: predicate.partial_defer_inline 1.40% : 0.000018s : 111: predicate.partial_eliminate 0.97% : 0.000012s : 88: predicate.print_const_string_wrapper 0.74% : 0.000009s : 51: predicate.reduce_all_const_elim 1.31% : 0.000016s : 88: predicate.reduce_eliminate 2.19% : 0.000028s : 206: predicate.redundant_stop_gradient_eliminater 0.31% : 0.000004s : 42: predicate.remove_not_recompute_node 2.03% : 0.000026s : 267: predicate.replace_applicator 0.76% : 0.000010s : 110: predicate.replace_old_param 0.07% : 0.000001s : 7: predicate.reset_defer_inline 0.98% : 0.000012s : 88: predicate.reshape_eliminate 1.96% : 0.000025s : 161: predicate.row_tensor_add_zeros_like 0.12% : 0.000001s : 7: predicate.row_tensor_eliminate 1.87% : 0.000024s : 165: predicate.same_eliminate 0.33% : 0.000004s : 49: predicate.set_cell_output_no_recompute 0.55% : 0.000007s : 42: predicate.shard_identity_eliminate 0.20% : 0.000002s : 14: predicate.special_op_eliminate 0.85% : 0.000011s : 56: predicate.specialize_transform 1.81% : 0.000023s : 161: predicate.split_environ_get_set_with_tuple_value 1.56% : 0.000020s : 110: predicate.stack_unstack_eliminate 0.10% : 0.000001s : 7: predicate.switch_call_monad_eliminater 1.51% : 0.000019s : 126: predicate.switch_defer_inline 3.15% : 0.000040s : 291: predicate.switch_layer_defer_inline 4.17% : 0.000053s : 363: predicate.switch_simplify 0.93% : 0.000012s : 88: predicate.tile_eliminate 1.01% : 0.000013s : 88: predicate.transpose_eliminate 1.26% : 0.000016s : 102: predicate.tuple_list_convert_item_index_to_positive 1.35% : 0.000017s : 102: predicate.tuple_list_get_item_const_eliminator 1.24% : 0.000016s : 102: predicate.tuple_list_get_item_depend_reorder 2.64% : 0.000033s : 169: predicate.tuple_list_get_item_eliminator 2.28% : 0.000029s : 102: predicate.tuple_list_get_set_item_eliminator 1.90% : 0.000024s : 153: predicate.tuple_list_set_item_eliminator 1.30% : 0.000016s : 118: predicate.tuple_to_list_eliminator_ 2.23% : 0.000028s : 206: predicate.updatestate_pure_node_eliminater 3.02% : 0.000038s : 259: predicate.updatestate_useless_node_eliminater 0.12% : 0.000001s : 7: predicate.value_based_eliminate 0.94% : 0.000012s : 42: predicate.virtual_dataset_eliminate 0.50% : 0.000006s : 42: predicate.virtual_output_eliminate 0.08% : 0.000001s : 7: predicate.virtual_view_grad_eliminate 0.10% : 0.000001s : 7: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003562 58 62.90% : 0.002240s : 26: func_graph_cloner_run.FuncGraphClonerGraph 4.95% : 0.000176s : 3: func_graph_cloner_run.FuncGraphClonerNode 32.15% : 0.001145s : 29: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 52.755264 305 0.00% : 0.000005s : 1: ForceFp32Comm 0.01% : 0.004166s : 1: add_attr 0.01% : 0.004154s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.00% : 0.000090s : 1: add_recomputation 0.00% : 0.000005s : 1: assign_add_opt 0.00% : 0.000280s : 1: auto_monad 0.00% : 0.000038s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000004s : 1: begin_end_overlap_inline 0.00% : 0.000006s : 1: bias_add_comm_swap 0.00% : 0.000873s : 1: bootstrap 0.00% : 0.000034s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000024s : 1: control_data_broadcast_order 0.00% : 0.000013s : 1: convert_after_rewriter 0.00% : 0.000038s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.00% : 0.000011s : 1: environ_conv 0.00% : 0.000075s : 1: event_method 0.00% : 0.000023s : 1: execute 0.00% : 0.000006s : 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% : 0.000007s : 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 0.00% : 0.000565s : 1: loop_unroll 0.00% : 0.000005s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.00% : 0.000746s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.00% : 0.000020s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000020s : 1: opt.transform.mutable_eliminate 0.88% : 0.466726s : 181: opt.transform.opt_a 0.00% : 0.000063s : 1: opt.transform.opt_after_cconv 0.00% : 0.000037s : 1: opt.transform.opt_after_jit_grad 0.00% : 0.000192s : 28: opt.transform.opt_b 0.00% : 0.000915s : 2: opt.transform.opt_resolve 0.00% : 0.000080s : 2: opt.transform.opt_trans_graph 0.00% : 0.000075s : 4: opt.transform.symbol_engine_opt 1.90% : 1.003254s : 1: opt_a 0.00% : 0.000167s : 1: opt_after_cconv 0.00% : 0.000589s : 1: opt_after_jit_grad 0.00% : 0.000327s : 1: opt_b 1.91% : 1.006374s : 1: optimize 0.00% : 0.000032s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 1: order_py_execute_after_rewriter 0.00% : 0.000039s : 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% : 0.000005s : 1: overlap_opt_shard_in_pipeline 0.00% : 0.000006s : 1: overlap_param_gather 0.00% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.00% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000007s : 1: overlap_recompute_comm 0.00% : 0.000008s : 1: parallel-infer-symbol 0.00% : 0.000004s : 1: parallel-infer-symbol-second 0.00% : 0.000005s : 1: partial_unused_args_eliminate 0.00% : 0.000005s : 1: pipeline_parallel_scheduler 0.00% : 0.000005s : 1: pipeline_split 0.00% : 0.000069s : 1: pre_auto_parallel 0.00% : 0.000051s : 1: py_interpret_to_execute 0.00% : 0.000019s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000022s : 1: remove_dup_value 0.01% : 0.005646s : 3: renormalize.infer 0.01% : 0.002869s : 3: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000008s : 1: rewriter_after_jit_bprop_graph 0.00% : 0.000074s : 1: rewriter_after_opt_a 0.00% : 0.000206s : 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% : 0.000005s : 1: split_layernorm_comm 0.00% : 0.000006s : 1: split_matmul_comm_elemetwise 0.00% : 0.000013s : 1: swap_dp_allreduce_reducescatter 0.00% : 0.000292s : 1: symbol_engine_optimizer 93.86% : 49.514660s : 1: task_emit 0.00% : 0.000113s : 1: tuple_transform 1.40% : 0.740841s : 1: type_inference 0.00% : 0.000090s : 1: validate [WARNING] ME(169920:281473205423920,MainProcess):2026-01-29-17:45:16.136.781 [mindspore/common/tensor.py:142] For 'Tensor', if 'dtype' is not None, 'input_data', 'shape' or 'init' must not be None. TotalTime = 0.0282311, [24] [bootstrap]: 0.00064961 [type_inference]: 0.00728015 [event_method]: 2.154e-05 [auto_monad]: 0.00018794 [graph_reusing]: 5.97001e-06 [inline]: 2.07001e-06 [add_attr]: 0.00457164, [1] [add_attr_with_inline]: 0.00455971, [1] [Cycle 1]: 6.377e-05, [2] [tag_attr]: 2.389e-05 [meta_addattr_fg_expand]: 5.74e-06 [parallel-infer-symbol]: 4.07003e-06 [pre_auto_parallel]: 3.839e-05 [insert-virtual-dataset]: 2.57001e-06 [parallel-infer-symbol-second]: 7.99977e-07 [dataset_repeat_opt]: 2.22999e-06 [pipeline_split]: 1.66e-06 [optimize]: 0.00619855, [53] [py_interpret_to_execute]: 2.836e-05 [rewriter_before_opt_a]: 8.335e-05 [opt_a]: 0.00328433, [2] [Cycle 1]: 0.00245672, [45] [expand_dump_flag]: 2.94999e-06 [switch_simplify]: 4.308e-05 [loop_unroll]: 2.94e-05 [a_1]: 0.00074012 [with_stream_mark]: 1.75e-05 [recompute_prepare]: 1.074e-05 [updatestate_depend_eliminate]: 6.47001e-06 [updatestate_assign_eliminate]: 6.06998e-06 [updatestate_loads_eliminate]: 5.29e-06 [parameter_eliminate]: 1.91998e-06 [a_2]: 0.00012113 [accelerated_algorithm]: 9.52001e-06 [shard]: 2.39001e-06 [meta_shard_fg_expand]: 2.71999e-06 [shard_inline]: 8.42e-06 [merge_send_recv]: 1.136e-05 [auto_parallel]: 8.21002e-06 [parallel]: 4.917e-05 [flash_sp]: 8.62998e-06 [merge_comm]: 6.02001e-06 [allreduce_fusion]: 5.18002e-06 [matmul_add_comm_reduction]: 1.197e-05 [allreduce_slice_to_reducescatter]: 7.50006e-07 [virtual_shard_identity]: 1.073e-05 [virtual_dataset]: 8.72998e-06 [get_grad_eliminate_]: 8.35999e-06 [virtual_output]: 8.52e-06 [merge_forward]: 5.62999e-06 [cell_reuse_recompute_pass]: 1.34e-06 [offload_activation]: 1.379e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.792e-05 [merge_recompute_call_nodes]: 1.46002e-06 [before_grad]: 1.492e-05 [set_forward_comm_id_for_comm_node_pass]: 5.49e-06 [meta_fg_expand]: 4.03999e-06 [flash_sp_send_recv_attached]: 2.87002e-06 [receive_attached]: 2.39001e-06 [after_resolve]: 1.224e-05 [a_after_grad]: 1.316e-05 [renormalize]: 0.00083832 [add_forward_monad_depend]: 5.64998e-06 [auto_monad_grad]: 2.37001e-06 [auto_monad_eliminator]: 2.161e-05 [cse]: 4.023e-05 [a_3]: 6.518e-05 [Cycle 2]: 0.00081711, [45] [expand_dump_flag]: 1.50999e-06 [switch_simplify]: 9.84001e-06 [loop_unroll]: 8.12e-06 [a_1]: 0.00019601 [with_stream_mark]: 1.418e-05 [recompute_prepare]: 8.38999e-06 [updatestate_depend_eliminate]: 4.80999e-06 [updatestate_assign_eliminate]: 4.17003e-06 [updatestate_loads_eliminate]: 4.57e-06 [parameter_eliminate]: 9.00007e-07 [a_2]: 0.00010873 [accelerated_algorithm]: 8.53001e-06 [shard]: 1.15001e-06 [meta_shard_fg_expand]: 2.16998e-06 [shard_inline]: 8.52998e-06 [merge_send_recv]: 6.52001e-06 [auto_parallel]: 7.35e-06 [parallel]: 4.58999e-06 [flash_sp]: 3.75e-06 [merge_comm]: 4.79002e-06 [allreduce_fusion]: 4.49002e-06 [matmul_add_comm_reduction]: 7.81001e-06 [allreduce_slice_to_reducescatter]: 3.60014e-07 [virtual_shard_identity]: 8.96002e-06 [virtual_dataset]: 7.71001e-06 [get_grad_eliminate_]: 7.61999e-06 [virtual_output]: 7.39002e-06 [merge_forward]: 4.30999e-06 [cell_reuse_recompute_pass]: 1.25001e-06 [offload_activation]: 8.69003e-06 [cell_reuse_handle_not_recompute_node_pass]: 1.508e-05 [merge_recompute_call_nodes]: 7.2e-07 [before_grad]: 2.343e-05 [set_forward_comm_id_for_comm_node_pass]: 5.54e-06 [meta_fg_expand]: 3.33e-06 [flash_sp_send_recv_attached]: 9.39996e-07 [receive_attached]: 1.15001e-06 [after_resolve]: 1.082e-05 [a_after_grad]: 1.262e-05 [renormalize]: 2.30008e-07 [add_forward_monad_depend]: 1.29998e-06 [auto_monad_grad]: 9.20001e-07 [auto_monad_eliminator]: 1.222e-05 [cse]: 1.971e-05 [a_3]: 4.928e-05 [py_interpret_to_execute_after_opt_a]: 1.182e-05 [slice_cell_reuse_recomputed_activation]: 2.41e-06 [rewriter_after_opt_a]: 0.00020535 [convert_after_rewriter]: 1.024e-05 [order_py_execute_after_rewriter]: 6.77002e-06 [mutable_eliminate]: 0.00059742 [opt_b]: 0.00030352, [1] [Cycle 1]: 0.00029628, [7] [b_1]: 0.00020009 [b_2]: 1.139e-05 [updatestate_depend_eliminate]: 8.68001e-06 [updatestate_assign_eliminate]: 4.79e-06 [updatestate_loads_eliminate]: 4.35e-06 [renormalize]: 4.80009e-07 [cse]: 2.57e-05 [optimize_parallel_all_gather_comm]: 2.18e-05 [overlap_param_gather]: 2.69001e-06 [cconv]: 2.795e-05 [loop_unroll]: 0.00063551 [opt_after_cconv]: 0.00015309, [1] [Cycle 1]: 0.00014601, [7] [c_1]: 6.049e-05 [parameter_eliminate]: 3.46001e-06 [updatestate_depend_eliminate]: 9.57001e-06 [updatestate_assign_eliminate]: 4.53001e-06 [updatestate_loads_eliminate]: 4.16001e-06 [cse]: 2.681e-05 [renormalize]: 3.70026e-07 [remove_dup_value]: 1.745e-05 [tuple_transform]: 0.00010164, [1] [Cycle 1]: 9.61e-05, [4] [d_1]: 6.438e-05 [none_parameter_eliminate]: 1.67999e-06 [renormalize]: 2.3999e-07 [switch_simplify]: 1.077e-05 [partial_unused_args_eliminate]: 2.61999e-06 [add_recomputation]: 8.043e-05 [cse_after_recomputation]: 3.051e-05, [1] [Cycle 1]: 2.54e-05, [1] [cse]: 1.958e-05 [environ_conv]: 7.63001e-06 [swap_dp_allreduce_reducescatter]: 8.22e-06 [bias_add_comm_swap]: 2.87002e-06 [label_micro_interleaved_index]: 5.32001e-06 [label_fine_grained_interleaved_index]: 3.28e-06 [merge_cast_opt]: 1.83002e-06 [slice_recompute_activation]: 2.44999e-06 [micro_interleaved_order_control]: 2.27001e-06 [assign_add_opt]: 1.64e-06 [ForceFp32Comm]: 1.02998e-06 [remove_cast_before_assign_add]: 1.20001e-06 [full_micro_interleaved_order_control]: 2.31998e-06 [reorder_send_recv_between_fp_bp]: 2.74001e-06 [comm_op_add_attrs]: 1.24e-06 [add_comm_op_reuse_tag]: 1.00999e-06 [interleave_split_concat_branches]: 1.37e-06 [interleave_parallel_branches]: 1.14e-06 [overlap_opt_shard_in_pipeline]: 1.27e-06 [overlap_opt_shard_grad_in_pipeline]: 2.02999e-06 [control_data_broadcast_order]: 2.368e-05 [grouped_pairwise_exchange_alltoall]: 1.92001e-06 [offloading_packed_experts]: 6.33e-06 [overlap_recompute_and_grad_model_parallel]: 6.68e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.20999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.72001e-06 [overlap_recompute_comm]: 2.49999e-06 [overlap_grad_ring_attention]: 6.10002e-06 [overlap_grad_flash_sp]: 2.741e-05 [begin_end_overlap_inline]: 5.79981e-07 [split_matmul_comm_elemetwise]: 2.54001e-06 [split_layernorm_comm]: 2.06003e-06 [handle_group_info]: 1.11002e-06 [symbol_engine_optimizer]: 0.00021255, [1] [Cycle 1]: 0.00020757, [6] [build]: 0.00010164 [elim_shapecalc]: 1.43e-05 [elim_not_effective]: 2.127e-05 [opt_reshape]: 1.022e-05 [fold_const_symbol]: 2.429e-05 [renormalize]: 2.70025e-07 [detach_backward]: 2.14e-06 [pipeline_parallel_scheduler]: 1.70001e-06 [auto_monad_reorder]: 2.953e-05 [get_jit_bprop_graph]: 1.76e-06 [rewriter_after_jit_bprop_graph]: 5.06002e-06 [opt_after_jit_grad]: 0.00055018 [validate]: 0.00012017 [backend_pass]: 1.17e-06 [task_emit]: 0.00831417 [execute]: 8.77e-06 Sums bootstrap : 0.000650s : 2.87% type_inference : 0.007280s : 32.21% event_method : 0.000022s : 0.10% auto_monad : 0.000188s : 0.83% graph_reusing : 0.000006s : 0.03% inline : 0.000002s : 0.01% add_attr.add_attr_with_inline.tag_attr : 0.000024s : 0.11% add_attr.add_attr_with_inline.meta_addattr_fg_expand : 0.000006s : 0.03% parallel-infer-symbol : 0.000004s : 0.02% pre_auto_parallel : 0.000038s : 0.17% insert-virtual-dataset : 0.000003s : 0.01% parallel-infer-symbol-second : 0.000001s : 0.00% dataset_repeat_opt : 0.000002s : 0.01% pipeline_split : 0.000002s : 0.01% optimize.py_interpret_to_execute : 0.000028s : 0.13% optimize.rewriter_before_opt_a : 0.000083s : 0.37% optimize.opt_a.expand_dump_flag : 0.000004s : 0.02% optimize.opt_a.switch_simplify : 0.000053s : 0.23% optimize.opt_a.loop_unroll : 0.000038s : 0.17% optimize.opt_a.a_1 : 0.000936s : 4.14% optimize.opt_a.with_stream_mark : 0.000032s : 0.14% optimize.opt_a.recompute_prepare : 0.000019s : 0.08% optimize.opt_a.updatestate_depend_eliminate : 0.000011s : 0.05% optimize.opt_a.updatestate_assign_eliminate : 0.000010s : 0.05% optimize.opt_a.updatestate_loads_eliminate : 0.000010s : 0.04% optimize.opt_a.parameter_eliminate : 0.000003s : 0.01% optimize.opt_a.a_2 : 0.000230s : 1.02% optimize.opt_a.accelerated_algorithm : 0.000018s : 0.08% optimize.opt_a.shard : 0.000004s : 0.02% optimize.opt_a.meta_shard_fg_expand : 0.000005s : 0.02% optimize.opt_a.shard_inline : 0.000017s : 0.07% optimize.opt_a.merge_send_recv : 0.000018s : 0.08% optimize.opt_a.auto_parallel : 0.000016s : 0.07% optimize.opt_a.parallel : 0.000054s : 0.24% optimize.opt_a.flash_sp : 0.000012s : 0.05% optimize.opt_a.merge_comm : 0.000011s : 0.05% optimize.opt_a.allreduce_fusion : 0.000010s : 0.04% optimize.opt_a.matmul_add_comm_reduction : 0.000020s : 0.09% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000001s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000020s : 0.09% optimize.opt_a.virtual_dataset : 0.000016s : 0.07% optimize.opt_a.get_grad_eliminate_ : 0.000016s : 0.07% optimize.opt_a.virtual_output : 0.000016s : 0.07% optimize.opt_a.merge_forward : 0.000010s : 0.04% optimize.opt_a.cell_reuse_recompute_pass : 0.000003s : 0.01% optimize.opt_a.offload_activation : 0.000022s : 0.10% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000033s : 0.15% optimize.opt_a.merge_recompute_call_nodes : 0.000002s : 0.01% optimize.opt_a.before_grad : 0.000038s : 0.17% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000011s : 0.05% optimize.opt_a.meta_fg_expand : 0.000007s : 0.03% optimize.opt_a.flash_sp_send_recv_attached : 0.000004s : 0.02% optimize.opt_a.receive_attached : 0.000004s : 0.02% optimize.opt_a.after_resolve : 0.000023s : 0.10% optimize.opt_a.a_after_grad : 0.000026s : 0.11% optimize.opt_a.renormalize : 0.000839s : 3.71% optimize.opt_a.add_forward_monad_depend : 0.000007s : 0.03% optimize.opt_a.auto_monad_grad : 0.000003s : 0.01% optimize.opt_a.auto_monad_eliminator : 0.000034s : 0.15% optimize.opt_a.cse : 0.000060s : 0.27% optimize.opt_a.a_3 : 0.000114s : 0.51% optimize.py_interpret_to_execute_after_opt_a : 0.000012s : 0.05% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.01% optimize.rewriter_after_opt_a : 0.000205s : 0.91% optimize.convert_after_rewriter : 0.000010s : 0.05% optimize.order_py_execute_after_rewriter : 0.000007s : 0.03% optimize.mutable_eliminate : 0.000597s : 2.64% optimize.opt_b.b_1 : 0.000200s : 0.89% optimize.opt_b.b_2 : 0.000011s : 0.05% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.04% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.02% optimize.opt_b.updatestate_loads_eliminate : 0.000004s : 0.02% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000026s : 0.11% optimize.optimize_parallel_all_gather_comm : 0.000022s : 0.10% optimize.overlap_param_gather : 0.000003s : 0.01% optimize.cconv : 0.000028s : 0.12% optimize.loop_unroll : 0.000636s : 2.81% optimize.opt_after_cconv.c_1 : 0.000060s : 0.27% optimize.opt_after_cconv.parameter_eliminate : 0.000003s : 0.02% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000010s : 0.04% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000005s : 0.02% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000004s : 0.02% optimize.opt_after_cconv.cse : 0.000027s : 0.12% optimize.opt_after_cconv.renormalize : 0.000000s : 0.00% optimize.remove_dup_value : 0.000017s : 0.08% optimize.tuple_transform.d_1 : 0.000064s : 0.28% optimize.tuple_transform.none_parameter_eliminate : 0.000002s : 0.01% optimize.tuple_transform.renormalize : 0.000000s : 0.00% optimize.tuple_transform.switch_simplify : 0.000011s : 0.05% optimize.partial_unused_args_eliminate : 0.000003s : 0.01% optimize.add_recomputation : 0.000080s : 0.36% optimize.cse_after_recomputation.cse : 0.000020s : 0.09% optimize.environ_conv : 0.000008s : 0.03% optimize.swap_dp_allreduce_reducescatter : 0.000008s : 0.04% optimize.bias_add_comm_swap : 0.000003s : 0.01% optimize.label_micro_interleaved_index : 0.000005s : 0.02% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.01% optimize.merge_cast_opt : 0.000002s : 0.01% optimize.slice_recompute_activation : 0.000002s : 0.01% optimize.micro_interleaved_order_control : 0.000002s : 0.01% optimize.assign_add_opt : 0.000002s : 0.01% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.01% optimize.full_micro_interleaved_order_control : 0.000002s : 0.01% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.01% optimize.comm_op_add_attrs : 0.000001s : 0.01% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.01% optimize.interleave_parallel_branches : 0.000001s : 0.01% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.01% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.01% optimize.control_data_broadcast_order : 0.000024s : 0.10% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.01% optimize.offloading_packed_experts : 0.000006s : 0.03% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.03% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.01% optimize.overlap_recompute_allgather_and_fa_grad : 0.000002s : 0.01% optimize.overlap_recompute_comm : 0.000002s : 0.01% optimize.overlap_grad_ring_attention : 0.000006s : 0.03% optimize.overlap_grad_flash_sp : 0.000027s : 0.12% optimize.begin_end_overlap_inline : 0.000001s : 0.00% optimize.split_matmul_comm_elemetwise : 0.000003s : 0.01% optimize.split_layernorm_comm : 0.000002s : 0.01% optimize.handle_group_info : 0.000001s : 0.00% optimize.symbol_engine_optimizer.build : 0.000102s : 0.45% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000014s : 0.06% optimize.symbol_engine_optimizer.elim_not_effective : 0.000021s : 0.09% optimize.symbol_engine_optimizer.opt_reshape : 0.000010s : 0.05% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000024s : 0.11% optimize.symbol_engine_optimizer.renormalize : 0.000000s : 0.00% detach_backward : 0.000002s : 0.01% pipeline_parallel_scheduler : 0.000002s : 0.01% auto_monad_reorder : 0.000030s : 0.13% get_jit_bprop_graph : 0.000002s : 0.01% rewriter_after_jit_bprop_graph : 0.000005s : 0.02% opt_after_jit_grad : 0.000550s : 2.43% validate : 0.000120s : 0.53% backend_pass : 0.000001s : 0.01% task_emit : 0.008314s : 36.78% execute : 0.000009s : 0.04% Time group info: ------[substitution.] 0.000247 60 2.66% : 0.000007s : 2: substitution.depend_value_elim 1.96% : 0.000005s : 5: substitution.elim_not_effective 4.18% : 0.000010s : 5: substitution.fold_const_symbol 3.12% : 0.000008s : 6: substitution.graph_param_transform 65.80% : 0.000162s : 4: substitution.inline 2.20% : 0.000005s : 10: substitution.j_node_and_user_rematch 3.30% : 0.000008s : 10: substitution.remove_not_recompute_node 1.29% : 0.000003s : 2: substitution.replace_old_param 5.09% : 0.000013s : 2: substitution.tuple_list_get_item_eliminator 5.66% : 0.000014s : 6: substitution.updatestate_pure_node_eliminater 4.73% : 0.000012s : 8: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.007221 2 89.33% : 0.006451s : 1: type_inference.infer 10.67% : 0.000770s : 1: type_inference.specialize ------[replace.] 0.000059 6 71.20% : 0.000042s : 4: replace.inline 28.80% : 0.000017s : 2: replace.tuple_list_get_item_eliminator ------[match.] 0.000171 6 93.34% : 0.000160s : 4: match.inline 6.66% : 0.000011s : 2: match.tuple_list_get_item_eliminator ------[predicate.] 0.000263 1652 0.82% : 0.000002s : 16: predicate.accumulaten_eliminater 0.91% : 0.000002s : 6: predicate.ad_related_special_op_eliminate 0.60% : 0.000002s : 12: predicate.addn_check_dump 0.80% : 0.000002s : 16: predicate.addn_zero_filter 0.85% : 0.000002s : 16: predicate.adjust_all_reduce_mul_add 2.07% : 0.000005s : 28: predicate.arithmetic_simplify 0.98% : 0.000003s : 16: predicate.cast_eliminate 0.72% : 0.000002s : 12: predicate.check_bprop_eliminate 0.60% : 0.000002s : 12: predicate.compare_switch_simplify 0.20% : 0.000001s : 6: predicate.const_output_eliminate 0.69% : 0.000002s : 12: predicate.depend_value_elim 0.89% : 0.000002s : 16: predicate.dict_get_item_const_eliminator 0.94% : 0.000002s : 16: predicate.dict_get_item_eliminator 0.84% : 0.000002s : 16: predicate.dict_set_item_eliminator 1.08% : 0.000003s : 12: predicate.dumpgradient_eliminate 0.24% : 0.000001s : 6: predicate.elim_not_effective 0.43% : 0.000001s : 6: predicate.elim_shapecalc_of_broadcastargs 1.25% : 0.000003s : 22: predicate.environ_add_const_eliminate 1.12% : 0.000003s : 22: predicate.environ_get_add_eliminate 1.31% : 0.000003s : 22: predicate.environ_get_depend_swap 1.87% : 0.000005s : 34: predicate.environ_get_eliminate 1.39% : 0.000004s : 22: predicate.environ_get_set_eliminate 1.13% : 0.000003s : 22: predicate.exchange_switch_depend_value 1.86% : 0.000005s : 22: predicate.float_depend_g_call 0.60% : 0.000002s : 12: predicate.float_environ_get_switch 0.95% : 0.000002s : 18: predicate.float_tuple_getitem_switch 0.17% : 0.000000s : 6: predicate.fold_const_symbol 0.72% : 0.000002s : 12: predicate.get_grad_eliminate 0.22% : 0.000001s : 6: predicate.graph_param_transform 0.68% : 0.000002s : 12: predicate.incorporate_call 0.59% : 0.000002s : 12: predicate.incorporate_call_switch 6.10% : 0.000016s : 74: predicate.inline 0.86% : 0.000002s : 12: predicate.inline_without_move 0.34% : 0.000001s : 12: predicate.j_node_and_user_rematch 0.83% : 0.000002s : 12: predicate.less_batch_normalization 1.81% : 0.000005s : 30: predicate.list_to_tuple_eliminator_ 2.49% : 0.000007s : 46: predicate.load_eliminater 1.19% : 0.000003s : 6: predicate.loop_unroll_after_grad 1.94% : 0.000005s : 38: predicate.loop_unroll_before_grad 1.86% : 0.000005s : 28: predicate.make_slice_get_slice_eliminator 0.65% : 0.000002s : 12: predicate.merge_addn 0.76% : 0.000002s : 12: predicate.micro_step_allgather_replace 0.67% : 0.000002s : 12: predicate.mini_step_allgather_replace 0.88% : 0.000002s : 16: predicate.minmaximum_grad 0.94% : 0.000002s : 6: predicate.mutable_eliminate 0.48% : 0.000001s : 6: predicate.opt_reshape 0.46% : 0.000001s : 6: predicate.parallel_virtual_node 1.41% : 0.000004s : 22: predicate.partial_defer_inline 1.43% : 0.000004s : 24: predicate.partial_eliminate 0.83% : 0.000002s : 16: predicate.print_const_string_wrapper 0.67% : 0.000002s : 12: predicate.reduce_all_const_elim 1.14% : 0.000003s : 16: predicate.reduce_eliminate 2.46% : 0.000006s : 46: predicate.redundant_stop_gradient_eliminater 0.43% : 0.000001s : 12: predicate.remove_not_recompute_node 1.18% : 0.000003s : 30: predicate.replace_applicator 0.52% : 0.000001s : 12: predicate.replace_old_param 0.30% : 0.000001s : 6: predicate.reset_defer_inline 0.89% : 0.000002s : 16: predicate.reshape_eliminate 0.67% : 0.000002s : 12: predicate.row_tensor_add_zeros_like 0.47% : 0.000001s : 6: predicate.row_tensor_eliminate 0.89% : 0.000002s : 12: predicate.same_eliminate 0.42% : 0.000001s : 12: predicate.set_cell_output_no_recompute 0.87% : 0.000002s : 12: predicate.shard_identity_eliminate 0.92% : 0.000002s : 12: predicate.special_op_eliminate 0.83% : 0.000002s : 12: predicate.specialize_transform 0.87% : 0.000002s : 12: predicate.split_environ_get_set_with_tuple_value 0.96% : 0.000003s : 12: predicate.stack_unstack_eliminate 0.41% : 0.000001s : 6: predicate.switch_call_monad_eliminater 1.21% : 0.000003s : 22: predicate.switch_defer_inline 1.80% : 0.000005s : 34: predicate.switch_layer_defer_inline 4.59% : 0.000012s : 78: predicate.switch_simplify 0.81% : 0.000002s : 16: predicate.tile_eliminate 0.84% : 0.000002s : 16: predicate.transpose_eliminate 1.75% : 0.000005s : 28: predicate.tuple_list_convert_item_index_to_positive 1.64% : 0.000004s : 28: predicate.tuple_list_get_item_const_eliminator 1.48% : 0.000004s : 28: predicate.tuple_list_get_item_depend_reorder 3.03% : 0.000008s : 42: predicate.tuple_list_get_item_eliminator 1.48% : 0.000004s : 28: predicate.tuple_list_get_set_item_eliminator 2.41% : 0.000006s : 40: predicate.tuple_list_set_item_eliminator 1.62% : 0.000004s : 30: predicate.tuple_to_list_eliminator_ 2.70% : 0.000007s : 46: predicate.updatestate_pure_node_eliminater 3.46% : 0.000009s : 58: predicate.updatestate_useless_node_eliminater 0.44% : 0.000001s : 6: predicate.value_based_eliminate 0.69% : 0.000002s : 12: predicate.virtual_dataset_eliminate 0.69% : 0.000002s : 12: predicate.virtual_output_eliminate 0.43% : 0.000001s : 6: predicate.virtual_view_grad_eliminate 0.54% : 0.000001s : 6: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.000623 11 54.26% : 0.000338s : 5: func_graph_cloner_run.FuncGraphClonerGraph 45.74% : 0.000285s : 6: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.041719 196 0.01% : 0.000004s : 1: ForceFp32Comm 10.97% : 0.004577s : 1: add_attr 10.94% : 0.004564s : 1: add_attr_with_inline 0.01% : 0.000004s : 1: add_comm_op_reuse_tag 0.20% : 0.000085s : 1: add_recomputation 0.01% : 0.000004s : 1: assign_add_opt 0.47% : 0.000196s : 1: auto_monad 0.08% : 0.000034s : 1: auto_monad_reorder 0.02% : 0.000007s : 1: backend_pass 0.01% : 0.000004s : 1: begin_end_overlap_inline 0.01% : 0.000006s : 1: bias_add_comm_swap 1.63% : 0.000681s : 1: bootstrap 0.08% : 0.000031s : 1: cconv 0.01% : 0.000005s : 1: comm_op_add_attrs 0.07% : 0.000027s : 1: control_data_broadcast_order 0.03% : 0.000014s : 1: convert_after_rewriter 0.08% : 0.000034s : 1: cse_after_recomputation 0.01% : 0.000006s : 1: dataset_repeat_opt 0.01% : 0.000006s : 1: detach_backward 0.03% : 0.000012s : 1: environ_conv 0.07% : 0.000028s : 1: event_method 0.03% : 0.000014s : 1: execute 0.01% : 0.000005s : 1: full_micro_interleaved_order_control 0.01% : 0.000005s : 1: get_jit_bprop_graph 0.03% : 0.000010s : 1: graph_reusing 0.01% : 0.000005s : 1: grouped_pairwise_exchange_alltoall 0.01% : 0.000004s : 1: handle_group_info 0.01% : 0.000005s : 1: inline 0.01% : 0.000006s : 1: insert-virtual-dataset 0.01% : 0.000004s : 1: interleave_parallel_branches 0.01% : 0.000005s : 1: interleave_split_concat_branches 0.02% : 0.000007s : 1: label_fine_grained_interleaved_index 0.02% : 0.000008s : 1: label_micro_interleaved_index 1.55% : 0.000646s : 1: loop_unroll 0.01% : 0.000005s : 1: merge_cast_opt 0.01% : 0.000005s : 1: micro_interleaved_order_control 1.46% : 0.000608s : 1: mutable_eliminate 0.03% : 0.000010s : 1: offloading_packed_experts 0.05% : 0.000021s : 1: opt.transform.loop_unroll_optimizer 0.04% : 0.000018s : 1: opt.transform.mutable_eliminate 3.67% : 0.001532s : 78: opt.transform.opt_a 0.14% : 0.000059s : 1: opt.transform.opt_after_cconv 0.09% : 0.000036s : 1: opt.transform.opt_after_jit_grad 0.43% : 0.000180s : 28: opt.transform.opt_b 0.17% : 0.000073s : 2: opt.transform.opt_trans_graph 0.16% : 0.000066s : 4: opt.transform.symbol_engine_opt 7.88% : 0.003287s : 1: opt_a 0.38% : 0.000157s : 1: opt_after_cconv 1.34% : 0.000561s : 1: opt_after_jit_grad 0.74% : 0.000307s : 1: opt_b 14.87% : 0.006203s : 1: optimize 0.06% : 0.000026s : 1: optimize_parallel_all_gather_comm 0.02% : 0.000010s : 1: order_py_execute_after_rewriter 0.07% : 0.000031s : 1: overlap_grad_flash_sp 0.01% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.02% : 0.000009s : 1: overlap_grad_ring_attention 0.01% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.01% : 0.000004s : 1: overlap_opt_shard_in_pipeline 0.02% : 0.000006s : 1: overlap_param_gather 0.01% : 0.000005s : 1: overlap_recompute_allgather_and_fa_grad 0.02% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.01% : 0.000005s : 1: overlap_recompute_comm 0.02% : 0.000008s : 1: parallel-infer-symbol 0.01% : 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 0.10% : 0.000043s : 1: pre_auto_parallel 0.08% : 0.000033s : 1: py_interpret_to_execute 0.04% : 0.000015s : 1: py_interpret_to_execute_after_opt_a 0.01% : 0.000004s : 1: remove_cast_before_assign_add 0.05% : 0.000021s : 1: remove_dup_value 1.11% : 0.000465s : 1: renormalize.infer 0.88% : 0.000366s : 1: renormalize.specialize 0.01% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.02% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.51% : 0.000212s : 1: rewriter_after_opt_a 0.21% : 0.000088s : 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 0.02% : 0.000006s : 1: split_matmul_comm_elemetwise 0.03% : 0.000011s : 1: swap_dp_allreduce_reducescatter 0.52% : 0.000216s : 1: symbol_engine_optimizer 19.97% : 0.008330s : 1: task_emit 0.25% : 0.000105s : 1: tuple_transform 17.49% : 0.007297s : 1: type_inference 0.37% : 0.000154s : 1: validate TotalTime = 0.420244, [24] [bootstrap]: 0.0004295 [type_inference]: 0.164159 [event_method]: 6.796e-05 [auto_monad]: 0.00029177 [graph_reusing]: 1.077e-05 [inline]: 2.65002e-06 [add_attr]: 0.0041029, [1] [add_attr_with_inline]: 0.00409256, [1] [Cycle 1]: 9.849e-05, [2] [tag_attr]: 4.955e-05 [meta_addattr_fg_expand]: 1.179e-05 [parallel-infer-symbol]: 3.31001e-06 [pre_auto_parallel]: 6.419e-05 [insert-virtual-dataset]: 2.84999e-06 [parallel-infer-symbol-second]: 7.39994e-07 [dataset_repeat_opt]: 2.14e-06 [pipeline_split]: 1.62999e-06 [optimize]: 0.240824, [53] [py_interpret_to_execute]: 4.688e-05 [rewriter_before_opt_a]: 0.00017244 [opt_a]: 0.141061, [4] [Cycle 1]: 0.133162, [45] [expand_dump_flag]: 4.13001e-06 [switch_simplify]: 9.514e-05 [loop_unroll]: 8.168e-05 [a_1]: 0.00169831 [with_stream_mark]: 2.938e-05 [recompute_prepare]: 2.832e-05 [updatestate_depend_eliminate]: 1.332e-05 [updatestate_assign_eliminate]: 1.151e-05 [updatestate_loads_eliminate]: 1.053e-05 [parameter_eliminate]: 2.82002e-06 [a_2]: 0.00032252 [accelerated_algorithm]: 2.273e-05 [shard]: 2.14e-06 [meta_shard_fg_expand]: 5.46002e-06 [shard_inline]: 2.365e-05 [merge_send_recv]: 2.13e-05 [auto_parallel]: 1.597e-05 [parallel]: 1.93e-05 [flash_sp]: 1.216e-05 [merge_comm]: 1.261e-05 [allreduce_fusion]: 1.226e-05 [matmul_add_comm_reduction]: 3.42e-05 [allreduce_slice_to_reducescatter]: 7.09988e-07 [virtual_shard_identity]: 2.198e-05 [virtual_dataset]: 2.236e-05 [get_grad_eliminate_]: 2.087e-05 [virtual_output]: 2.166e-05 [merge_forward]: 1.383e-05 [cell_reuse_recompute_pass]: 1.44e-06 [offload_activation]: 2.41e-05 [cell_reuse_handle_not_recompute_node_pass]: 4.228e-05 [merge_recompute_call_nodes]: 1.57999e-06 [before_grad]: 3.972e-05 [set_forward_comm_id_for_comm_node_pass]: 1.603e-05 [meta_fg_expand]: 0.00628486 [flash_sp_send_recv_attached]: 4.75001e-06 [receive_attached]: 2.43e-06 [after_resolve]: 0.00011528 [a_after_grad]: 0.00017662 [renormalize]: 0.122128 [add_forward_monad_depend]: 1.606e-05 [auto_monad_grad]: 8.48001e-06 [auto_monad_eliminator]: 9.45e-05 [cse]: 0.00028539 [a_3]: 0.00100011 [Cycle 2]: 0.00522414, [45] [expand_dump_flag]: 3.43999e-06 [switch_simplify]: 6.611e-05 [loop_unroll]: 6.116e-05 [a_1]: 0.00240202 [with_stream_mark]: 3.241e-05 [recompute_prepare]: 2.304e-05 [updatestate_depend_eliminate]: 1.164e-05 [updatestate_assign_eliminate]: 1.159e-05 [updatestate_loads_eliminate]: 1.167e-05 [parameter_eliminate]: 3.86999e-06 [a_2]: 0.00038104 [accelerated_algorithm]: 1.34e-05 [shard]: 1.99e-06 [meta_shard_fg_expand]: 3.76999e-06 [shard_inline]: 1.103e-05 [merge_send_recv]: 1.312e-05 [auto_parallel]: 1.41e-05 [parallel]: 9.49e-06 [flash_sp]: 5.42999e-06 [merge_comm]: 6.93e-06 [allreduce_fusion]: 7.05e-06 [matmul_add_comm_reduction]: 1.357e-05 [allreduce_slice_to_reducescatter]: 8.2e-07 [virtual_shard_identity]: 1.242e-05 [virtual_dataset]: 1.123e-05 [get_grad_eliminate_]: 1.06e-05 [virtual_output]: 1.05e-05 [merge_forward]: 6.88e-06 [cell_reuse_recompute_pass]: 1.32999e-06 [offload_activation]: 1.626e-05 [cell_reuse_handle_not_recompute_node_pass]: 2.348e-05 [merge_recompute_call_nodes]: 1.45999e-06 [before_grad]: 1.92e-05 [set_forward_comm_id_for_comm_node_pass]: 7.78001e-06 [meta_fg_expand]: 0.00013653 [flash_sp_send_recv_attached]: 2.03002e-06 [receive_attached]: 2.43002e-06 [after_resolve]: 1.706e-05 [a_after_grad]: 1.836e-05 [renormalize]: 0.00137494 [add_forward_monad_depend]: 6.17999e-06 [auto_monad_grad]: 2.56998e-06 [auto_monad_eliminator]: 2.371e-05 [cse]: 6.104e-05 [a_3]: 8.252e-05 [Cycle 3]: 0.0016909, [45] [expand_dump_flag]: 1.86e-06 [switch_simplify]: 1.229e-05 [loop_unroll]: 1.139e-05 [a_1]: 0.00029073 [with_stream_mark]: 1.285e-05 [recompute_prepare]: 1.102e-05 [updatestate_depend_eliminate]: 4.455e-05 [updatestate_assign_eliminate]: 5.68002e-06 [updatestate_loads_eliminate]: 4.47e-06 [parameter_eliminate]: 1.10999e-06 [a_2]: 0.00012657 [accelerated_algorithm]: 9.61e-06 [shard]: 1.29003e-06 [meta_shard_fg_expand]: 2.09999e-06 [shard_inline]: 9.15001e-06 [merge_send_recv]: 8.80001e-06 [auto_parallel]: 8.29002e-06 [parallel]: 5.31002e-06 [flash_sp]: 1.23002e-06 [merge_comm]: 5.65001e-06 [allreduce_fusion]: 5.12e-06 [matmul_add_comm_reduction]: 9.31002e-06 [allreduce_slice_to_reducescatter]: 4.10015e-07 [virtual_shard_identity]: 1.031e-05 [virtual_dataset]: 9.54e-06 [get_grad_eliminate_]: 9.77001e-06 [virtual_output]: 8.58001e-06 [merge_forward]: 5.20001e-06 [cell_reuse_recompute_pass]: 1.34998e-06 [offload_activation]: 1.256e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.773e-05 [merge_recompute_call_nodes]: 8.59989e-07 [before_grad]: 1.492e-05 [set_forward_comm_id_for_comm_node_pass]: 5.64e-06 [meta_fg_expand]: 3.68e-06 [flash_sp_send_recv_attached]: 1.08001e-06 [receive_attached]: 1.22e-06 [after_resolve]: 1.182e-05 [a_after_grad]: 3.214e-05 [renormalize]: 0.00060492 [add_forward_monad_depend]: 4.37e-06 [auto_monad_grad]: 1.39e-06 [auto_monad_eliminator]: 1.833e-05 [cse]: 3.263e-05 [a_3]: 6.661e-05 [Cycle 4]: 0.00096158, [45] [expand_dump_flag]: 1.50999e-06 [switch_simplify]: 1.067e-05 [loop_unroll]: 9.24e-06 [a_1]: 0.00022238 [with_stream_mark]: 1.049e-05 [recompute_prepare]: 9.19e-06 [updatestate_depend_eliminate]: 5.52999e-06 [updatestate_assign_eliminate]: 4.60001e-06 [updatestate_loads_eliminate]: 4.33001e-06 [parameter_eliminate]: 9.39996e-07 [a_2]: 0.00012276 [accelerated_algorithm]: 9.44e-06 [shard]: 1.22999e-06 [meta_shard_fg_expand]: 2.32001e-06 [shard_inline]: 8.96998e-06 [merge_send_recv]: 7.38e-06 [auto_parallel]: 7.77998e-06 [parallel]: 4.18999e-06 [flash_sp]: 1.12e-06 [merge_comm]: 5.42999e-06 [allreduce_fusion]: 4.94998e-06 [matmul_add_comm_reduction]: 8.07e-06 [allreduce_slice_to_reducescatter]: 4.19997e-07 [virtual_shard_identity]: 1.027e-05 [virtual_dataset]: 9.12001e-06 [get_grad_eliminate_]: 1.004e-05 [virtual_output]: 8.82e-06 [merge_forward]: 5.97999e-06 [cell_reuse_recompute_pass]: 1.55001e-06 [offload_activation]: 1.085e-05 [cell_reuse_handle_not_recompute_node_pass]: 1.765e-05 [merge_recompute_call_nodes]: 6.59988e-07 [before_grad]: 1.55e-05 [set_forward_comm_id_for_comm_node_pass]: 6.05002e-06 [meta_fg_expand]: 3.73999e-06 [flash_sp_send_recv_attached]: 1.07e-06 [receive_attached]: 9.20001e-07 [after_resolve]: 1.192e-05 [a_after_grad]: 1.383e-05 [renormalize]: 8.9989e-08 [add_forward_monad_depend]: 1.25999e-06 [auto_monad_grad]: 9.09989e-07 [auto_monad_eliminator]: 1.275e-05 [cse]: 8.183e-05 [a_3]: 6.388e-05 [py_interpret_to_execute_after_opt_a]: 1.501e-05 [slice_cell_reuse_recomputed_activation]: 2.27001e-06 [rewriter_after_opt_a]: 6.317e-05 [convert_after_rewriter]: 9.61e-06 [order_py_execute_after_rewriter]: 6.99001e-06 [mutable_eliminate]: 0.0007368 [opt_b]: 0.00031588, [1] [Cycle 1]: 0.00030845, [7] [b_1]: 0.00020632 [b_2]: 1.167e-05 [updatestate_depend_eliminate]: 8.76997e-06 [updatestate_assign_eliminate]: 4.83001e-06 [updatestate_loads_eliminate]: 4.95999e-06 [renormalize]: 3.30008e-07 [cse]: 3.302e-05 [optimize_parallel_all_gather_comm]: 2.134e-05 [overlap_param_gather]: 2.41e-06 [cconv]: 2.713e-05 [loop_unroll]: 0.0972659 [opt_after_cconv]: 0.00021799, [1] [Cycle 1]: 0.00020712, [7] [c_1]: 7.747e-05 [parameter_eliminate]: 7.08e-06 [updatestate_depend_eliminate]: 1.724e-05 [updatestate_assign_eliminate]: 6.04001e-06 [updatestate_loads_eliminate]: 5.45001e-06 [cse]: 5.533e-05 [renormalize]: 8.40024e-07 [remove_dup_value]: 1.863e-05 [tuple_transform]: 0.00010585, [1] [Cycle 1]: 0.00010057, [4] [d_1]: 6.969e-05 [none_parameter_eliminate]: 1.39e-06 [renormalize]: 2.00002e-07 [switch_simplify]: 1.059e-05 [partial_unused_args_eliminate]: 2.01e-06 [add_recomputation]: 8.874e-05 [cse_after_recomputation]: 3.475e-05, [1] [Cycle 1]: 2.881e-05, [1] [cse]: 2.243e-05 [environ_conv]: 1.027e-05 [swap_dp_allreduce_reducescatter]: 9.16002e-06 [bias_add_comm_swap]: 3.98999e-06 [label_micro_interleaved_index]: 7.43e-06 [label_fine_grained_interleaved_index]: 3.01001e-06 [merge_cast_opt]: 1.50001e-06 [slice_recompute_activation]: 2.41e-06 [micro_interleaved_order_control]: 2.85998e-06 [assign_add_opt]: 1.37e-06 [ForceFp32Comm]: 9.09989e-07 [remove_cast_before_assign_add]: 1.37999e-06 [full_micro_interleaved_order_control]: 2.17999e-06 [reorder_send_recv_between_fp_bp]: 2.76999e-06 [comm_op_add_attrs]: 1.09998e-06 [add_comm_op_reuse_tag]: 1.15999e-06 [interleave_split_concat_branches]: 1.17e-06 [interleave_parallel_branches]: 1.10999e-06 [overlap_opt_shard_in_pipeline]: 1.38002e-06 [overlap_opt_shard_grad_in_pipeline]: 1.94e-06 [control_data_broadcast_order]: 2.299e-05 [grouped_pairwise_exchange_alltoall]: 1.99e-06 [offloading_packed_experts]: 6.43998e-06 [overlap_recompute_and_grad_model_parallel]: 6.70998e-06 [overlap_grad_matmul_and_grad_allreduce]: 1.10999e-06 [overlap_recompute_allgather_and_fa_grad]: 1.40001e-06 [overlap_recompute_comm]: 2.96001e-06 [overlap_grad_ring_attention]: 6.64001e-06 [overlap_grad_flash_sp]: 2.896e-05 [begin_end_overlap_inline]: 5.50004e-07 [split_matmul_comm_elemetwise]: 2.19001e-06 [split_layernorm_comm]: 1.70001e-06 [handle_group_info]: 1.05999e-06 [symbol_engine_optimizer]: 0.0002191, [1] [Cycle 1]: 0.00021441, [6] [build]: 0.00010676 [elim_shapecalc]: 1.512e-05 [elim_not_effective]: 2.271e-05 [opt_reshape]: 1.068e-05 [fold_const_symbol]: 2.684e-05 [renormalize]: 3.39991e-07 [detach_backward]: 2.79999e-06 [pipeline_parallel_scheduler]: 1.59e-06 [auto_monad_reorder]: 3.121e-05 [get_jit_bprop_graph]: 2.28002e-06 [rewriter_after_jit_bprop_graph]: 5.72999e-06 [opt_after_jit_grad]: 0.00052098 [validate]: 5.558e-05 [backend_pass]: 1.02998e-06 [task_emit]: 0.00928835 [execute]: 8.41002e-06 Sums bootstrap : 0.000429s : 0.10% type_inference : 0.164159s : 39.62% event_method : 0.000068s : 0.02% auto_monad : 0.000292s : 0.07% 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 : 0.000012s : 0.00% parallel-infer-symbol : 0.000003s : 0.00% pre_auto_parallel : 0.000064s : 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 : 0.000047s : 0.01% optimize.rewriter_before_opt_a : 0.000172s : 0.04% optimize.opt_a.expand_dump_flag : 0.000011s : 0.00% optimize.opt_a.switch_simplify : 0.000184s : 0.04% optimize.opt_a.loop_unroll : 0.000163s : 0.04% optimize.opt_a.a_1 : 0.004613s : 1.11% optimize.opt_a.with_stream_mark : 0.000085s : 0.02% optimize.opt_a.recompute_prepare : 0.000072s : 0.02% optimize.opt_a.updatestate_depend_eliminate : 0.000075s : 0.02% optimize.opt_a.updatestate_assign_eliminate : 0.000033s : 0.01% optimize.opt_a.updatestate_loads_eliminate : 0.000031s : 0.01% optimize.opt_a.parameter_eliminate : 0.000009s : 0.00% optimize.opt_a.a_2 : 0.000953s : 0.23% optimize.opt_a.accelerated_algorithm : 0.000055s : 0.01% optimize.opt_a.shard : 0.000007s : 0.00% optimize.opt_a.meta_shard_fg_expand : 0.000014s : 0.00% optimize.opt_a.shard_inline : 0.000053s : 0.01% optimize.opt_a.merge_send_recv : 0.000051s : 0.01% optimize.opt_a.auto_parallel : 0.000046s : 0.01% optimize.opt_a.parallel : 0.000038s : 0.01% optimize.opt_a.flash_sp : 0.000020s : 0.00% optimize.opt_a.merge_comm : 0.000031s : 0.01% optimize.opt_a.allreduce_fusion : 0.000029s : 0.01% optimize.opt_a.matmul_add_comm_reduction : 0.000065s : 0.02% optimize.opt_a.allreduce_slice_to_reducescatter : 0.000002s : 0.00% optimize.opt_a.virtual_shard_identity : 0.000055s : 0.01% optimize.opt_a.virtual_dataset : 0.000052s : 0.01% optimize.opt_a.get_grad_eliminate_ : 0.000051s : 0.01% optimize.opt_a.virtual_output : 0.000050s : 0.01% optimize.opt_a.merge_forward : 0.000032s : 0.01% optimize.opt_a.cell_reuse_recompute_pass : 0.000006s : 0.00% optimize.opt_a.offload_activation : 0.000064s : 0.02% optimize.opt_a.cell_reuse_handle_not_recompute_node_pass : 0.000101s : 0.02% optimize.opt_a.merge_recompute_call_nodes : 0.000005s : 0.00% optimize.opt_a.before_grad : 0.000089s : 0.02% optimize.opt_a.set_forward_comm_id_for_comm_node_pass : 0.000036s : 0.01% optimize.opt_a.meta_fg_expand : 0.006429s : 1.55% optimize.opt_a.flash_sp_send_recv_attached : 0.000009s : 0.00% optimize.opt_a.receive_attached : 0.000007s : 0.00% optimize.opt_a.after_resolve : 0.000156s : 0.04% optimize.opt_a.a_after_grad : 0.000241s : 0.06% optimize.opt_a.renormalize : 0.124108s : 29.95% optimize.opt_a.add_forward_monad_depend : 0.000028s : 0.01% optimize.opt_a.auto_monad_grad : 0.000013s : 0.00% optimize.opt_a.auto_monad_eliminator : 0.000149s : 0.04% optimize.opt_a.cse : 0.000461s : 0.11% optimize.opt_a.a_3 : 0.001213s : 0.29% optimize.py_interpret_to_execute_after_opt_a : 0.000015s : 0.00% optimize.slice_cell_reuse_recomputed_activation : 0.000002s : 0.00% optimize.rewriter_after_opt_a : 0.000063s : 0.02% optimize.convert_after_rewriter : 0.000010s : 0.00% optimize.order_py_execute_after_rewriter : 0.000007s : 0.00% optimize.mutable_eliminate : 0.000737s : 0.18% optimize.opt_b.b_1 : 0.000206s : 0.05% optimize.opt_b.b_2 : 0.000012s : 0.00% optimize.opt_b.updatestate_depend_eliminate : 0.000009s : 0.00% optimize.opt_b.updatestate_assign_eliminate : 0.000005s : 0.00% optimize.opt_b.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_b.renormalize : 0.000000s : 0.00% optimize.opt_b.cse : 0.000033s : 0.01% optimize.optimize_parallel_all_gather_comm : 0.000021s : 0.01% optimize.overlap_param_gather : 0.000002s : 0.00% optimize.cconv : 0.000027s : 0.01% optimize.loop_unroll : 0.097266s : 23.47% optimize.opt_after_cconv.c_1 : 0.000077s : 0.02% optimize.opt_after_cconv.parameter_eliminate : 0.000007s : 0.00% optimize.opt_after_cconv.updatestate_depend_eliminate : 0.000017s : 0.00% optimize.opt_after_cconv.updatestate_assign_eliminate : 0.000006s : 0.00% optimize.opt_after_cconv.updatestate_loads_eliminate : 0.000005s : 0.00% optimize.opt_after_cconv.cse : 0.000055s : 0.01% optimize.opt_after_cconv.renormalize : 0.000001s : 0.00% optimize.remove_dup_value : 0.000019s : 0.00% optimize.tuple_transform.d_1 : 0.000070s : 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 : 0.000011s : 0.00% optimize.partial_unused_args_eliminate : 0.000002s : 0.00% optimize.add_recomputation : 0.000089s : 0.02% optimize.cse_after_recomputation.cse : 0.000022s : 0.01% optimize.environ_conv : 0.000010s : 0.00% optimize.swap_dp_allreduce_reducescatter : 0.000009s : 0.00% optimize.bias_add_comm_swap : 0.000004s : 0.00% optimize.label_micro_interleaved_index : 0.000007s : 0.00% optimize.label_fine_grained_interleaved_index : 0.000003s : 0.00% optimize.merge_cast_opt : 0.000002s : 0.00% optimize.slice_recompute_activation : 0.000002s : 0.00% optimize.micro_interleaved_order_control : 0.000003s : 0.00% optimize.assign_add_opt : 0.000001s : 0.00% optimize.ForceFp32Comm : 0.000001s : 0.00% optimize.remove_cast_before_assign_add : 0.000001s : 0.00% optimize.full_micro_interleaved_order_control : 0.000002s : 0.00% optimize.reorder_send_recv_between_fp_bp : 0.000003s : 0.00% optimize.comm_op_add_attrs : 0.000001s : 0.00% optimize.add_comm_op_reuse_tag : 0.000001s : 0.00% optimize.interleave_split_concat_branches : 0.000001s : 0.00% optimize.interleave_parallel_branches : 0.000001s : 0.00% optimize.overlap_opt_shard_in_pipeline : 0.000001s : 0.00% optimize.overlap_opt_shard_grad_in_pipeline : 0.000002s : 0.00% optimize.control_data_broadcast_order : 0.000023s : 0.01% optimize.grouped_pairwise_exchange_alltoall : 0.000002s : 0.00% optimize.offloading_packed_experts : 0.000006s : 0.00% optimize.overlap_recompute_and_grad_model_parallel : 0.000007s : 0.00% optimize.overlap_grad_matmul_and_grad_allreduce : 0.000001s : 0.00% optimize.overlap_recompute_allgather_and_fa_grad : 0.000001s : 0.00% optimize.overlap_recompute_comm : 0.000003s : 0.00% optimize.overlap_grad_ring_attention : 0.000007s : 0.00% optimize.overlap_grad_flash_sp : 0.000029s : 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 : 0.000107s : 0.03% optimize.symbol_engine_optimizer.elim_shapecalc : 0.000015s : 0.00% optimize.symbol_engine_optimizer.elim_not_effective : 0.000023s : 0.01% optimize.symbol_engine_optimizer.opt_reshape : 0.000011s : 0.00% optimize.symbol_engine_optimizer.fold_const_symbol : 0.000027s : 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 : 0.000031s : 0.01% get_jit_bprop_graph : 0.000002s : 0.00% rewriter_after_jit_bprop_graph : 0.000006s : 0.00% opt_after_jit_grad : 0.000521s : 0.13% validate : 0.000056s : 0.01% backend_pass : 0.000001s : 0.00% task_emit : 0.009288s : 2.24% execute : 0.000008s : 0.00% Time group info: ------[substitution.] 0.002068 320 0.97% : 0.000020s : 8: substitution.depend_value_elim 0.28% : 0.000006s : 6: substitution.elim_not_effective 0.22% : 0.000005s : 5: substitution.float_depend_g_call 0.26% : 0.000005s : 3: substitution.float_tuple_getitem_switch 0.60% : 0.000012s : 6: substitution.fold_const_symbol 36.89% : 0.000763s : 4: substitution.getattr_setattr_resolve 0.41% : 0.000008s : 7: substitution.graph_param_transform 0.15% : 0.000003s : 2: substitution.incorporate_call 0.10% : 0.000002s : 2: substitution.incorporate_call_switch 37.67% : 0.000779s : 22: substitution.inline 1.64% : 0.000034s : 5: substitution.inline_without_move 0.83% : 0.000017s : 35: substitution.j_node_and_user_rematch 0.82% : 0.000017s : 11: substitution.minmaximum_grad 0.35% : 0.000007s : 5: substitution.partial_eliminate 1.21% : 0.000025s : 35: substitution.remove_not_recompute_node 2.88% : 0.000059s : 22: substitution.replace_applicator 0.70% : 0.000015s : 17: substitution.replace_old_param 0.27% : 0.000006s : 2: substitution.set_cell_output_no_recompute 1.65% : 0.000034s : 11: substitution.tuple_list_convert_item_index_to_positive 0.78% : 0.000016s : 11: substitution.tuple_list_get_item_const_eliminator 2.28% : 0.000047s : 11: substitution.tuple_list_get_item_depend_reorder 3.82% : 0.000079s : 30: substitution.tuple_list_get_item_eliminator 1.04% : 0.000021s : 11: substitution.tuple_list_get_set_item_eliminator 1.57% : 0.000032s : 20: substitution.updatestate_pure_node_eliminater 2.60% : 0.000054s : 29: substitution.updatestate_useless_node_eliminater ------[type_inference.] 0.164061 2 98.67% : 0.161886s : 1: type_inference.infer 1.33% : 0.002174s : 1: type_inference.specialize ------[replace.] 0.000484 43 10.35% : 0.000050s : 3: replace.getattr_setattr_resolve 45.20% : 0.000219s : 22: replace.inline 10.68% : 0.000052s : 1: replace.replace_applicator 28.38% : 0.000137s : 16: replace.tuple_list_get_item_eliminator 5.39% : 0.000026s : 1: replace.updatestate_useless_node_eliminater ------[match.] 0.001541 43 46.50% : 0.000717s : 3: match.getattr_setattr_resolve 49.72% : 0.000766s : 22: match.inline 0.76% : 0.000012s : 1: match.replace_applicator 2.46% : 0.000038s : 16: match.tuple_list_get_item_eliminator 0.55% : 0.000008s : 1: match.updatestate_useless_node_eliminater ------[predicate.] 0.001216 8144 1.06% : 0.000013s : 88: predicate.accumulaten_eliminater 0.23% : 0.000003s : 7: predicate.ad_related_special_op_eliminate 0.60% : 0.000007s : 56: predicate.addn_check_dump 1.00% : 0.000012s : 88: predicate.addn_zero_filter 0.95% : 0.000012s : 88: predicate.adjust_all_reduce_mul_add 2.01% : 0.000024s : 139: predicate.arithmetic_simplify 1.02% : 0.000012s : 88: predicate.cast_eliminate 1.78% : 0.000022s : 165: predicate.check_bprop_eliminate 0.66% : 0.000008s : 56: predicate.compare_switch_simplify 0.05% : 0.000001s : 7: predicate.const_output_eliminate 0.63% : 0.000008s : 51: predicate.depend_value_elim 1.10% : 0.000013s : 88: predicate.dict_get_item_const_eliminator 1.10% : 0.000013s : 88: predicate.dict_get_item_eliminator 1.02% : 0.000012s : 88: predicate.dict_set_item_eliminator 0.26% : 0.000003s : 14: predicate.dumpgradient_eliminate 0.05% : 0.000001s : 7: predicate.elim_not_effective 0.10% : 0.000001s : 7: predicate.elim_shapecalc_of_broadcastargs 1.05% : 0.000013s : 95: predicate.environ_add_const_eliminate 1.05% : 0.000013s : 95: predicate.environ_get_add_eliminate 1.05% : 0.000013s : 95: predicate.environ_get_depend_swap 1.70% : 0.000021s : 146: predicate.environ_get_eliminate 1.04% : 0.000013s : 95: predicate.environ_get_set_eliminate 1.49% : 0.000018s : 126: predicate.exchange_switch_depend_value 1.98% : 0.000024s : 126: predicate.float_depend_g_call 0.64% : 0.000008s : 56: predicate.float_environ_get_switch 0.75% : 0.000009s : 63: predicate.float_tuple_getitem_switch 0.05% : 0.000001s : 7: predicate.fold_const_symbol 0.55% : 0.000007s : 42: predicate.get_grad_eliminate 0.50% : 0.000006s : 20: predicate.getattr_setattr_resolve 0.07% : 0.000001s : 7: predicate.graph_param_transform 0.60% : 0.000007s : 51: predicate.incorporate_call 0.53% : 0.000006s : 51: predicate.incorporate_call_switch 5.08% : 0.000062s : 326: predicate.inline 3.00% : 0.000036s : 110: predicate.inline_without_move 0.25% : 0.000003s : 42: predicate.j_node_and_user_rematch 0.68% : 0.000008s : 42: predicate.less_batch_normalization 1.40% : 0.000017s : 118: predicate.list_to_tuple_eliminator_ 2.39% : 0.000029s : 206: predicate.load_eliminater 0.67% : 0.000008s : 7: predicate.loop_unroll_after_grad 2.05% : 0.000025s : 174: predicate.loop_unroll_before_grad 1.25% : 0.000015s : 102: predicate.make_slice_get_slice_eliminator 0.71% : 0.000009s : 56: predicate.merge_addn 1.71% : 0.000021s : 161: predicate.micro_step_allgather_replace 1.73% : 0.000021s : 161: predicate.mini_step_allgather_replace 0.98% : 0.000012s : 88: predicate.minmaximum_grad 0.22% : 0.000003s : 7: predicate.mutable_eliminate 0.10% : 0.000001s : 7: predicate.opt_reshape 0.12% : 0.000001s : 7: predicate.parallel_virtual_node 1.81% : 0.000022s : 126: predicate.partial_defer_inline 1.45% : 0.000018s : 111: predicate.partial_eliminate 1.00% : 0.000012s : 88: predicate.print_const_string_wrapper 0.62% : 0.000008s : 51: predicate.reduce_all_const_elim 1.28% : 0.000016s : 88: predicate.reduce_eliminate 2.29% : 0.000028s : 206: predicate.redundant_stop_gradient_eliminater 0.29% : 0.000004s : 42: predicate.remove_not_recompute_node 2.01% : 0.000024s : 267: predicate.replace_applicator 0.75% : 0.000009s : 110: predicate.replace_old_param 0.06% : 0.000001s : 7: predicate.reset_defer_inline 1.01% : 0.000012s : 88: predicate.reshape_eliminate 1.77% : 0.000021s : 161: predicate.row_tensor_add_zeros_like 0.12% : 0.000001s : 7: predicate.row_tensor_eliminate 2.06% : 0.000025s : 165: predicate.same_eliminate 0.36% : 0.000004s : 49: predicate.set_cell_output_no_recompute 0.56% : 0.000007s : 42: predicate.shard_identity_eliminate 0.22% : 0.000003s : 14: predicate.special_op_eliminate 0.80% : 0.000010s : 56: predicate.specialize_transform 1.94% : 0.000024s : 161: predicate.split_environ_get_set_with_tuple_value 1.48% : 0.000018s : 110: predicate.stack_unstack_eliminate 0.12% : 0.000001s : 7: predicate.switch_call_monad_eliminater 1.58% : 0.000019s : 126: predicate.switch_defer_inline 3.32% : 0.000040s : 291: predicate.switch_layer_defer_inline 4.40% : 0.000053s : 363: predicate.switch_simplify 1.01% : 0.000012s : 88: predicate.tile_eliminate 0.97% : 0.000012s : 88: predicate.transpose_eliminate 1.36% : 0.000016s : 102: predicate.tuple_list_convert_item_index_to_positive 1.36% : 0.000017s : 102: predicate.tuple_list_get_item_const_eliminator 1.22% : 0.000015s : 102: predicate.tuple_list_get_item_depend_reorder 2.49% : 0.000030s : 169: predicate.tuple_list_get_item_eliminator 1.27% : 0.000015s : 102: predicate.tuple_list_get_set_item_eliminator 1.94% : 0.000024s : 153: predicate.tuple_list_set_item_eliminator 1.30% : 0.000016s : 118: predicate.tuple_to_list_eliminator_ 2.31% : 0.000028s : 206: predicate.updatestate_pure_node_eliminater 3.09% : 0.000038s : 259: predicate.updatestate_useless_node_eliminater 0.11% : 0.000001s : 7: predicate.value_based_eliminate 0.58% : 0.000007s : 42: predicate.virtual_dataset_eliminate 0.52% : 0.000006s : 42: predicate.virtual_output_eliminate 0.09% : 0.000001s : 7: predicate.virtual_view_grad_eliminate 0.11% : 0.000001s : 7: predicate.zero_like_fill_zero ------[func_graph_cloner_run.] 0.003492 58 55.73% : 0.001946s : 26: func_graph_cloner_run.FuncGraphClonerGraph 3.62% : 0.000126s : 3: func_graph_cloner_run.FuncGraphClonerNode 40.65% : 0.001419s : 29: func_graph_cloner_run.FuncGraphSpecializer ------[meta_graph.] 0.000000 0 ------[manager.] 0.000000 0 ------[pynative] 0.000000 0 ------[others.] 0.798409 305 0.00% : 0.000005s : 1: ForceFp32Comm 0.51% : 0.004108s : 1: add_attr 0.51% : 0.004097s : 1: add_attr_with_inline 0.00% : 0.000004s : 1: add_comm_op_reuse_tag 0.01% : 0.000093s : 1: add_recomputation 0.00% : 0.000004s : 1: assign_add_opt 0.04% : 0.000302s : 1: auto_monad 0.00% : 0.000035s : 1: auto_monad_reorder 0.00% : 0.000006s : 1: backend_pass 0.00% : 0.000003s : 1: begin_end_overlap_inline 0.00% : 0.000007s : 1: bias_add_comm_swap 0.06% : 0.000467s : 1: bootstrap 0.00% : 0.000031s : 1: cconv 0.00% : 0.000004s : 1: comm_op_add_attrs 0.00% : 0.000026s : 1: control_data_broadcast_order 0.00% : 0.000013s : 1: convert_after_rewriter 0.00% : 0.000038s : 1: cse_after_recomputation 0.00% : 0.000005s : 1: dataset_repeat_opt 0.00% : 0.000007s : 1: detach_backward 0.00% : 0.000014s : 1: environ_conv 0.01% : 0.000077s : 1: event_method 0.02% : 0.000130s : 1: execute 0.00% : 0.000005s : 1: full_micro_interleaved_order_control 0.00% : 0.000006s : 1: get_jit_bprop_graph 0.00% : 0.000016s : 1: graph_reusing 0.00% : 0.000005s : 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% : 0.000010s : 1: label_micro_interleaved_index 12.18% : 0.097284s : 1: loop_unroll 0.00% : 0.000004s : 1: merge_cast_opt 0.00% : 0.000006s : 1: micro_interleaved_order_control 0.09% : 0.000746s : 1: mutable_eliminate 0.00% : 0.000009s : 1: offloading_packed_experts 0.01% : 0.000042s : 1: opt.transform.loop_unroll_optimizer 0.00% : 0.000018s : 1: opt.transform.mutable_eliminate 0.99% : 0.007886s : 181: opt.transform.opt_a 0.01% : 0.000076s : 1: opt.transform.opt_after_cconv 0.00% : 0.000036s : 1: opt.transform.opt_after_jit_grad 0.02% : 0.000188s : 28: opt.transform.opt_b 0.11% : 0.000863s : 2: opt.transform.opt_resolve 0.01% : 0.000078s : 2: opt.transform.opt_trans_graph 0.01% : 0.000072s : 4: opt.transform.symbol_engine_opt 17.67% : 0.141065s : 1: opt_a 0.03% : 0.000223s : 1: opt_after_cconv 0.07% : 0.000530s : 1: opt_after_jit_grad 0.04% : 0.000320s : 1: opt_b 30.16% : 0.240830s : 1: optimize 0.00% : 0.000025s : 1: optimize_parallel_all_gather_comm 0.00% : 0.000010s : 1: order_py_execute_after_rewriter 0.00% : 0.000032s : 1: overlap_grad_flash_sp 0.00% : 0.000004s : 1: overlap_grad_matmul_and_grad_allreduce 0.00% : 0.000010s : 1: overlap_grad_ring_attention 0.00% : 0.000005s : 1: overlap_opt_shard_grad_in_pipeline 0.00% : 0.000006s : 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% : 0.000010s : 1: overlap_recompute_and_grad_model_parallel 0.00% : 0.000007s : 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% : 0.000069s : 1: pre_auto_parallel 0.01% : 0.000051s : 1: py_interpret_to_execute 0.00% : 0.000018s : 1: py_interpret_to_execute_after_opt_a 0.00% : 0.000004s : 1: remove_cast_before_assign_add 0.00% : 0.000022s : 1: remove_dup_value 15.11% : 0.120679s : 3: renormalize.infer 0.43% : 0.003402s : 3: renormalize.specialize 0.00% : 0.000006s : 1: reorder_send_recv_between_fp_bp 0.00% : 0.000009s : 1: rewriter_after_jit_bprop_graph 0.01% : 0.000067s : 1: rewriter_after_opt_a 0.02% : 0.000177s : 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% : 0.000012s : 1: swap_dp_allreduce_reducescatter 0.03% : 0.000222s : 1: symbol_engine_optimizer 1.17% : 0.009304s : 1: task_emit 0.01% : 0.000109s : 1: tuple_transform 20.56% : 0.164186s : 1: type_inference 0.01% : 0.000086s : 1: validate Compare ['Success'] for 0'th output of Forward. Compare ['Success'] for 0'th output of Grad. End Start DynamicTest testing with [GRAPH_MODE_O0] [DYNAMIC_RANK] [None]... Compare ['Success'] for 0'th output of Forward. Compare ['Success'] for 0'th output of Grad. End *********************End DynamicTest for with GRAPH_MODE_O0 ********************* ************************************************************************************************* ***************************************TEST_OP_DYNAMIC End*************************************** ************************************************************************************************* . [hook] pytest_runtest_teardown:test_sigmoid__dynamic_shape tests/st/tensor/test_sigmoid_.py::test_sigmoid__dynamic_shape,max_mem:4.0M =============================== 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_sigmoid_.py::test_sigmoid__dynamic_shape /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 ================== 1 passed, 26 warnings in 507.15s (0:08:27) ==================